Host-state biomarkers and clinical factors for stratification and optimization of cell therapy trials in chronic limb-threatening ischemia: A review

Yusuke Shimizu; Edward Hosea Ntege; Naoki Matsuura; Tatsuya Ishii; Hiroshi Sunami; Kojiro Furukawa; Moriyasu Nakaema; Akihiro Tokushige; Yoshikazu Inoue

doi:10.1016/j.reth.2026.101162

. 2026 Jul 29;33:101162. doi: 10.1016/j.reth.2026.101162

Host-state biomarkers and clinical factors for stratification and optimization of cell therapy trials in chronic limb-threatening ischemia: A review

Yusuke Shimizu ^a,^⁎, Edward Hosea Ntege ^a, Naoki Matsuura ^a, Tatsuya Ishii ^a, Hiroshi Sunami ^b, Kojiro Furukawa ^c, Moriyasu Nakaema ^c, Akihiro Tokushige ^d, Yoshikazu Inoue ^e

PMCID: PMC13451802 PMID: 42571258

Abstract

Background

Chronic limb-threatening ischemia (CLTI) is a severe manifestation of peripheral arterial disease associated with high risks of amputation and mortality. Trials of cell-based therapies in patients unsuitable for revascularization (“no-option” or “poor-option” CLTI) have produced inconsistent results, underscoring the need for improved prognostic enrichment, baseline risk balancing, and protocol-defined stratification.

Main body

This structured narrative review synthesizes evidence linking host-state markers—including laboratory biomarkers, clinical host-state factors, and physiologic/perfusion indices—to clinically meaningful CLTI outcomes. Inflammatory, nutritional, renal, metabolic, lipid, coagulation-related, infection, dialysis, frailty, smoking, etiology, and perfusion-related variables may shape the regenerative microenvironment and contribute to heterogeneity in cell therapy trial outcomes. However, current evidence primarily supports prognostic risk stratification rather than validated prediction of differential treatment response. We therefore interpret treated-cohort associations and post hoc subgroup observations as hypothesis-generating candidate variables for prospective host-state × treatment interaction testing, not as validated predictive biomarkers. We propose a conceptual trial-design framework: first, confirm no-option or poor-option CLTI through multidisciplinary review, guideline-aligned assessment, objective hemodynamic testing, vascular imaging, and multidomain assessment; second, provisionally categorize patients into favorable, intermediate, and unfavorable host-state profiles using approximate prognostic ranges for prospective validation rather than clinical eligibility cutoffs, treatment-selection rules, or gatekeeping thresholds; third, ensure clinically necessary stabilization before eligibility confirmation and treat any run-in or post-randomization optimization strategy only as an optional, secondary, protocol-justified design component; and fourth, apply stratified randomization, covariate adjustment, blinded outcome adjudication, and prespecified host-state × treatment interaction analyses.

Conclusions

Integrating host-state biomarkers and clinical factors into CLTI cell therapy trial design may support prognostic enrichment, stratified enrollment, baseline risk adjustment, and hypothesis-driven evaluation of treatment-effect heterogeneity. The proposed profiles are hypothesis-generating trial-design strata, not validated clinical eligibility criteria, treatment-selection rules, or biomarker-based gatekeeping thresholds. Prospective validation will require standardized biomarker assays, harmonized endpoints, transparent reporting of cell products and concomitant care, and randomized testing of host-state × treatment interactions to advance precision regenerative medicine in this setting.

Keywords: Chronic limb-threatening ischemia, Cell therapy, Host-state markers, Prognostic enrichment, Stratified randomization, No-option disease

Highlights

•
Host-state biology may contribute to CLTI cell therapy trial heterogeneity.
•
Laboratory, clinical, and perfusion markers support prognostic stratification.
•
Host-state profiles are trial-design strata, not clinical gatekeeping cutoffs.
•
Optional optimization/run-in strategies remain secondary and bias-sensitive.
•
Standardized assays, endpoints, product reporting, and interaction tests are needed.

Abbreviations

ABI: ankle-brachial index
ABRFC: anatomy-biology-risk-function-context
ACC/AHA: American College of Cardiology/American Heart Association
ACSL4: acyl-CoA synthetase long-chain family member 4
AFS: amputation-free survival
ASO: arteriosclerosis obliterans
AUC: area under the receiver operating characteristic curve
BALI: Bone Marrow Autograft in Limb Ischemia
BMMNC: bone marrow mononuclear cell
BSC: best supportive care
CD34+: cluster of differentiation 34-positive
CFS: Clinical Frailty Scale
CI: confidence interval
CKD: chronic kidney disease
CKD-EPI: Chronic Kidney Disease Epidemiology Collaboration
CLI: critical limb ischemia
CLTI: chronic limb-threatening ischemia
CRP: C-reactive protein
CTD: connective tissue disease
CXCR3: C-X-C motif chemokine receptor 3
DSMB: Data and Safety Monitoring Board
ECFC: endothelial colony-forming cell
eGFR: estimated glomerular filtration rate
EPO: erythropoietin
ESRD: end-stage renal disease
ESC: European Society of Cardiology
ESVS: European Society for Vascular Surgery
EV: extracellular vesicle
EVT: endovascular therapy
FDA: US Food and Drug Administration
FDA-NIH BEST: Biomarkers, EndpointS, and other Tools Resource
G-CSF: granulocyte colony-stimulating factor
GLASS: Global Limb Anatomic Staging System
GPX4: glutathione peroxidase 4
HbA1c: glycated hemoglobin
HR: hazard ratio
IA: intra-arterial
IL-6: interleukin-6
IM: intramuscular
IP-10: interferon-γ-inducible protein-10 (CXCL10)
ITT: intention-to-treat
IWGDF: International Working Group on the Diabetic Foot
LLOQ: lower limit of quantification
Lp(a): lipoprotein(a)
MACE: major adverse cardiovascular events
MALE: major adverse limb events
MDT: multidisciplinary team
mITT: modified intention-to-treat
MSC: mesenchymal stromal cell
NLR: neutrophil-to-lymphocyte ratio
NRT: nicotine replacement therapy
OR: odds ratio
PAD: peripheral arterial disease
PBMNC: peripheral blood mononuclear cell
QoL: quality of life
RCT: randomized controlled trial
SLE: systemic lupus erythematosus
SVS: Society for Vascular Surgery
TAO: thromboangiitis obliterans
TBI: toe-brachial index
TcPO₂: transcutaneous oxygen pressure
TFRC: transferrin receptor
TNF-α: tumor necrosis factor-α
TTF: time-to-treatment failure
WIfI: Wound-Ischemia-foot Infection classification.

1. Background

Chronic limb-threatening ischemia (CLTI) is the most advanced stage of peripheral arterial disease (PAD) and is characterized by ischemic rest pain, nonhealing ulcers, or gangrene due to chronic arterial insufficiency [1]. PAD affects more than 200 million people worldwide, and patients with CLTI are at the highest risk of adverse outcomes [2]. Despite advances in revascularization and secondary prevention, recent series report 1-year major-amputation rates of 10%–40% and mortality rates of 20%–25%, with 5-year mortality approaching 50%–60% [1,3,4].

Throughout this review, we use CLTI as the preferred contemporary term for the clinical syndrome of PAD associated with ischemic rest pain, gangrene, or lower-extremity ulceration of at least 2 weeks’ duration. This terminology follows current vascular guidelines and is preferred over the older term critical limb ischemia (CLI), which implies fixed hemodynamic thresholds rather than a continuum of limb threat [1,[5], [6], [7]]. The term CLI is retained only when referring to historical studies, trial titles, or source-specific eligibility criteria.

Approximately 20%–30% of patients are classified as “no-option”—those unsuitable for, or who have failed, further surgical or endovascular revascularization. Even with optimal wound care and medical therapy, only about 60% achieve 12-month amputation-free survival (AFS) [1,3,4]. For the purposes of this review, no-option or poor-option CLTI refers to limb-threatening ischemia confirmed by a qualified multidisciplinary vascular team using guideline-aligned, consensus-informed assessment after objective hemodynamic testing, vascular imaging, and multidomain assessment of anatomy, biology, procedural risk, limb function, and clinical context, and for which conventional endovascular or surgical revascularization is not technically feasible, has failed, or offers an unacceptable risk-benefit profile [1,[5], [6], [7], [8]]. No-option or poor-option status should be documented explicitly and may reflect anatomic factors, such as absence of a suitable distal target, poor pedal runoff, or “desert foot”; technical or procedural factors, such as repeated failed revascularization or lack of useable bypass conduit; or patient-level factors, such as prohibitive procedural risk, nonfunctional limb, limited life expectancy, or other contextual considerations. No-option status should not be inferred simply because revascularization was not performed.

Because definitions of CLTI, CLI, and no-option disease vary across older trials and contemporary studies, this review preserves original study terminology where necessary but interprets the evidence within a guideline-aligned framework. Host-state assessment and stratification are proposed as trial-design adjuncts to guideline-directed assessment—including WIfI limb-threat staging, GLASS anatomic assessment, objective perfusion testing, infection control, wound care, and multidisciplinary review—and not as substitutes for determining revascularization feasibility.

Real-world analyses highlight a trial-to-practice gap: trial-eligible patients represent a minority of those encountered in routine CLTI care. In an analysis aligned with BEST-CLI criteria, only 11% of consecutive CLTI patients met eligibility criteria, whereas excluded patients—commonly due to adherence concerns, high surgical risk, or limited life expectancy—experienced substantially higher amputation rates (33% vs 0%, P = 0.02) [9]. This excluded group overlaps with the population most likely to be considered for regenerative approaches.

Multisocietal guidelines and comprehensive reviews emphasize multidisciplinary, pathway-based CLTI care and stress the need to consider patient factors, limb threat, and anatomy together when selecting treatments [1,10]. Nevertheless, outcomes remain heterogeneous across real-world cohorts and trials, and anatomy alone does not fully explain prognosis or treatment responsiveness [[11], [12], [13]].

The pathobiology of CLTI involves not only large-vessel occlusion but also microvascular rarefaction, immunothrombosis, and impaired wound healing—processes that are often exacerbated by neuropathy and infection. Systemic factors such as chronic inflammation, malnutrition, renal dysfunction, frailty, and active infection increase risks of amputation and mortality even after technically successful revascularization [3,4,11,12]. Importantly, these host-state features may also affect the effectiveness of cell-based therapies by modulating the local tissue environment.

A systematic review of 22 studies reported that inflammatory biomarkers—including C-reactive protein (CRP), fibrinogen, neutrophil-to-lymphocyte ratio (NLR), interleukin-6 (IL-6), and tumor necrosis factor-α—are consistently elevated in diabetic CLTI and are prognostically associated with adverse limb outcomes [14]. Another synthesis reported that markers of inflammation, thrombosis, and endothelial activation, including CRP, IL-6, NLR, and adhesion molecules, increase with disease severity and are associated with worse outcomes [15]. Mechanistic studies further support a causal link: peripheral blood-derived endothelial colony-forming cells (ECFCs) from diabetic patients show TNF-α pathway enrichment and altered microRNA profiles, and TNF-α exposure inhibits ECFC proliferation and angiogenic function in vitro [16]. Existing reviews generally lack validated operational ranges and practical host-state stratification frameworks for clinical trials, highlighting a key knowledge gap. Most biomarker data in CLTI are prognostic, and evidence that specific biomarkers predict differential benefit from cell therapy remains limited and requires prospective testing.

These observations have introduced the concept of a “hostile” tissue milieu. Severe systemic inflammation, ongoing infection, and impaired nutritional status may blunt the regenerative potential of cell-based therapies [11,12,17]. Reviews and risk-model syntheses call for risk-stratification models that incorporate host-state factors such as systemic inflammation, renal function, and frailty [18,19]. Validation studies in “no-option” cohorts confirm the prognostic value of staging systems developed for broader CLTI populations [20]. In practice, clinicians continue to use WIfI and GLASS to describe limb threat and anatomy. However, these classifications do not directly quantify systemic biological capacity for therapeutic angiogenesis and wound repair.

Cell-based therapies, such as mesenchymal stromal cells (MSCs), bone marrow mononuclear cells (BMMNCs), peripheral blood mononuclear cells (PBMNCs), and selected CD34⁺ cells, have been studied for “no-option” CLTI. These therapies are thought to act mainly via paracrine mechanisms—secreting pro-angiogenic mediators and extracellular vesicles and modulating immune responses—rather than via long-term engraftment [11,21]. Preclinical studies show that MSCs and ECFCs can each promote neovascularization through distinct pathways, underscoring that therapeutic efficacy depends on both cell-product characteristics and the host microenvironment [22]. Accordingly, different cell products should not be treated as biologically interchangeable; product source, manufacturing method, dose, delivery route, release criteria, and potency assays must be reported and interpreted separately.

The randomized trial landscape for cell therapy in CLTI is heterogeneous. The early TACT study provided proof-of-concept evidence that intramuscular autologous BMMNC implantation could improve physiologic measures such as ABI and TcPO₂ and reduce rest pain, but the study relied heavily on surrogate endpoints and included a favorable etiology mix, including patients with thromboangiitis obliterans [23]. Subsequent placebo-controlled or randomized studies tested diverse products and delivery strategies. PROVASA evaluated intra-arterial autologous BMMNCs in a randomized-start, placebo-controlled design and reported ulcer-healing and perfusion signals, but it was small and not powered to demonstrate a definitive AFS benefit [24]. RESTORE-CLI evaluated ixmyelocel-T, a culture-expanded autologous multicellular product, and showed favorable trends in time-to-treatment failure without definitive statistical evidence for improved AFS [25]. ACT34-CLI tested intramuscular autologous CD34⁺ cells in a randomized, double-blind, placebo-controlled pilot study and demonstrated safety, with exploratory efficacy signals [26]. JUVENTAS, the largest blinded autologous BMMNC trial at the time, found no significant reduction in major amputation or improvement in AFS [27]. The BALI trial, although randomized and double-blind, was small and exploratory [28]. IMPACT evaluated G–CSF–mobilized PBMNCs in a broader PAD population that included, but was not limited to, CLTI, limiting direct extrapolation to no-option CLTI [29]. Most recently, the phase III PACE trial of allogeneic placental-derived stromal cells failed to show an overall AFS benefit in Rutherford 5 no-option CLTI; although a post hoc subgroup with HbA1c <6.5% showed improved 12-month AFS (HR 0.46, 95% CI 0.21–0.99), this observation remains hypothesis-generating and does not establish HbA1c as a validated predictive biomarker of differential cell-therapy benefit [30]. Collectively, these studies demonstrate feasibility and biological plausibility but also underscore the absence of consistent benefit on hard clinical endpoints in unselected populations.

A recent multicenter, open-label, phase II dose-ranging proof-of-concept report evaluated intra-arterial autologous BMMNC therapy in diabetic CLTI patients unsuitable for surgical revascularization and reported a wound-healing signal at 12 months [31]. However, this evidence should be interpreted with caution because the report is available only as a supplement abstract, the design was open-label, the control group did not undergo a sham or placebo procedure, and no difference in amputation rates was observed across arms. We therefore treat this study as preliminary contextual evidence supporting feasibility and hypothesis generation, rather than as definitive evidence of clinical efficacy.

Across these trials, interpretation is limited by differences in cell source, manufacturing, dose, delivery route, comparator, blinding, endpoint definition, follow-up duration, and baseline patient risk. No completed major randomized trial incorporated prospective host-state profiling for enrichment or stratified randomization. Thus, the mixed trial results should not be interpreted simply as evidence for or against “cell therapy” as a uniform intervention; rather, they highlight the need for product-specific evaluation, contemporary control benchmarks, standardized endpoints, and prespecified host-state stratification. Supplementary Table S1 summarizes key randomized and controlled studies alongside contemporary conservative-care benchmarks, and recent trial landscape reviews emphasize population and endpoint heterogeneity, strengthening the need for risk- and biology-aware trial design and reporting [32,33].

These mixed results have prompted renewed focus on prognostic enrichment and host-state-based stratification in trial design. Observational cohorts and biomarker studies consistently link host-state factors with outcomes under standard care and within cell therapy cohorts, suggesting that baseline biology may contribute to outcome heterogeneity. For example, the Pan et al. [34] retrospective treated-cohort nomogram study in patients receiving mononuclear-cell therapy identified age, baseline TcPO₂, fibrinogen, arterial occlusion pattern, and log-transformed CD34⁺ cell count as variables associated with 6-month responder/remission status. In that treated cohort, fibrinogen >4 g/L was associated with lower odds of responder/remission status (OR 0.176, P = 0.003), each 1-mmHg increase in baseline TcPO₂ was associated with higher odds of responder/remission status (OR 1.062, P = 0.006), and the multivariable nomogram achieved an AUC of 0.851 [34]. Because all patients received cell therapy and no randomized comparator or treatment-by-marker interaction was tested, these findings should be interpreted as treated-cohort associations and candidate stratification variables, not as validated predictive biomarkers of differential cell-therapy benefit.

Accordingly, the present review integrates evidence on host-state biomarkers and clinical factors to inform a hypothesis-generating conceptual framework for host-state stratification of no-option or poor-option CLTI. It also discusses consequences for the design, reporting, and interpretation of cell therapy trials, with a focus on prognostic enrichment, stratified randomization, covariate adjustment, and prespecified host-state × treatment interaction testing, rather than clinical gatekeeping or biomarker-based treatment selection before prospective validation.

1.1. Literature search, study selection, and evidence handling

This review was designed as a structured narrative review and hypothesis-generating synthesis rather than a formal systematic review or meta-analysis. MEDLINE/PubMed, Embase, and Google Scholar were searched for English-language human studies published between January 2000 and October 2025; key articles published thereafter were added when directly relevant to the evolving trial landscape, guideline context, or host-state marker framework. A final targeted update was performed on May 15, 2026, to capture major guideline, consensus, and trial-design publications published after October 2025. Reference lists of selected key guidelines, randomized trials, systematic reviews, and recent narrative reviews were screened to identify additional relevant studies and contextual sources. Search terms combined disease descriptors (“chronic limb-threatening ischemia,” “critical limb ischemia,” “peripheral arterial disease,” “no-option limb ischemia”) with cell-therapy terms (“cell therapy,” “stem cell,” “mesenchymal stromal cell,” “bone marrow mononuclear cell,” “CD34-positive cell,” “peripheral blood mononuclear cell”), study-design terms (“randomized,” “placebo,” “sham,” “trial,” “cohort,” “registry”), outcome terms (“amputation-free survival,” “limb salvage,” “major amputation,” “wound healing,” “mortality”), and candidate host-state terms (“CRP,” “IL-6,” “NLR,” “albumin,” “anemia,” “eGFR,” “dialysis,” “fibrinogen,” “Lp(a),” “TcPO₂,” “WIfI,” “GLASS,” “infection,” “frailty,” and “etiology”).

Study selection prioritized multisociety or national guidelines, randomized, placebo- or sham-controlled trials, multicenter cohorts, registry studies, systematic reviews, and biomarker studies that reported clinically meaningful CLTI outcomes. Different study types were used for different inferential purposes. Randomized trials were used to summarize the cell-therapy trial landscape and treatment-effect evidence; conservative-care and natural-history cohorts were used to contextualize baseline prognosis and event-rate assumptions; observational biomarker cohorts were used to identify prognostic associations; treated cell-therapy cohorts and single-arm/proof-of-concept studies were used only for contextual or hypothesis-generating signals; and mechanistic or preclinical studies were used to support biological plausibility rather than clinical efficacy claims.

Because the included studies were heterogeneous in population definition, no-option criteria, cell product, delivery route, comparator, biomarker assay, timing of measurement, endpoint definition, and follow-up duration, formal PRISMA methods, quantitative pooling, structured risk-of-bias grading, and GRADE or certainty-of-evidence assessment were not performed. Where effect estimates were reported, they were interpreted within the original study context rather than pooled across studies. Biomarker units were standardized where possible, and proposed threshold ranges were treated as approximate prognostic ranges for protocol development and prospective validation, not as statistically optimized predictive cut points, validated clinical eligibility criteria, treatment-selection thresholds, or gatekeeping rules. Treated-cohort associations and post hoc subgroup observations were interpreted as hypothesis-generating candidate variables for prospective host-state × treatment interaction testing rather than as evidence of biomarker-defined treatment response, unless supported by prespecified randomized treatment-by-marker interaction analyses. Descriptive source-context terms used in tables and supplementary materials are not formal evidence-certainty ratings.

In this review, we use “host-state markers” as an umbrella term for variables that characterize the patient's biological, clinical, and physiologic condition at the time of trial screening or enrollment. Within this framework, “laboratory biomarkers” refer to measurable analytes or laboratory-derived indices, including CRP, NLR, albumin, hemoglobin, creatinine/eGFR, IL-6, fibrinogen, HbA1c, Lp(a), and CXCL10/IP-10. This usage is aligned with biomarker terminology in the FDA-NIH BEST Resource [35]. By contrast, WIfI infection grade, dialysis dependence, disease etiology, smoking status, frailty, and perfusion measures such as TcPO₂, ABI, and TBI are treated as clinical or physiologic host-state factors rather than biomarkers in the strict laboratory sense. The proposed favorable, intermediate, and unfavorable profiles therefore represent composite host-state strata for protocol development and prospective validation, not biomarker-only categories, validated clinical cutoffs, or automatic eligibility/gatekeeping rules.

2. Evidence for prognostic host-state markers in CLTI

Several observational studies have identified associations between readily available host-state markers and outcomes in CLTI under contemporary standard care. These markers span domains of inflammation, nutritional status, renal function, infection, perfusion, metabolism, lipid risk, functional reserve, behavioral factors, and disease etiology. While most are prognostic under standard care, their consistent association with limb loss and death provides a biologically plausible rationale for prognostic enrichment, baseline risk stratification, stratified randomization, and covariate adjustment in cell therapy trials.

Throughout this review, we distinguish prognostic markers from predictive biomarkers. Prognostic markers are associated with clinical outcomes, such as amputation, mortality, wound healing, or AFS, irrespective of treatment assignment. Predictive biomarkers, in contrast, identify differential treatment benefit and require evidence of a treatment-by-marker interaction, ideally from randomized trials [35,36]. Because most available CLTI biomarker studies are observational, single-arm, or treated-cohort analyses, the host-state markers discussed here should be interpreted primarily as prognostic enrichment and stratification variables. Their predictive value for cell-therapy benefit remains unproven and should be tested prospectively through prespecified analyses of host-state × treatment interactions. Accordingly, treated-cohort associations and post hoc subgroup observations are described in this review as hypothesis-generating candidate variables for prospective host-state × treatment interaction testing, not as validated predictive biomarkers of differential cell-therapy benefit.

Table 1 summarizes reported host-state marker ranges and prognostic associations in CLTI, and Supplementary Fig. S1 presents a qualitative forest-plot overview of representative associations with adverse limb outcomes. Supplementary Table S2 maps each plotted estimate in Supplementary Fig. S1 to its source reference, study context, population, outcome, and interpretive caveat to make clear that no formal meta-analysis or quantitative pooling was performed. The numerical ranges summarized in Table 1 and the estimates displayed in Supplementary Fig. S1 should be interpreted as approximate, context-specific prognostic information for protocol development and prospective validation, not as statistically optimized predictive cut points, clinical eligibility criteria, treatment-selection thresholds, or gatekeeping rules. Table 1 classifies each numerical boundary as study-derived, standard/convention-derived, or an illustrative pragmatic value; pragmatic boundaries are explicitly identified as not marker-specific evidence-based thresholds. The “Boundary provenance, source context, and interpretive caveat” column in Table 1 is descriptive and should not be interpreted as a formal evidence-grading, risk-of-bias, or certainty-of-evidence assessment.

Table 1.

Summary of host-state markers and clinical factors, approximate prognostic ranges for prospective validation, boundary provenance, and reported associations in CLTI studies.

Domain/trial role	Host-state marker or factor	Unit/scale	Approximate favorable prognostic range/profile	Approximate intermediate prognostic range/profile	Approximate unfavorable prognostic range/profile	Reported association and interpretation caveat	Modifiability/trial handling	Boundary provenance, source context, and interpretive caveat	References
Inflammatory; core pragmatic	CRP; CAR	CRP: mg/L; CAR: ratio	<30 mg/L (CRP)	30-100 mg/L	>100 mg/L	Elevated CRP is associated with adverse CLTI/PAD outcomes in biomarker syntheses. CAR was associated with long-term all-cause mortality after below-the-knee EVT; current cited evidence does not support CAR as an independent predictor of AFS.	Potentially modifiable through infection control, wound care, smoking cessation, and cardiometabolic therapy; repeat after acute infection/debridement.	Illustrative pragmatic: CRP 30 and 100 mg/L are inflammatory-burden bins; cited CLTI/PAD biomarker syntheses support prognostic direction but do not validate these exact cutoffs. Multiple observational cohorts and biomarker syntheses; no cell-therapy-specific predictive validation.	[14,15,37]
Inflammatory; core pragmatic	NLR	Unitless	<5	5-10	>10	High NLR is associated with all-cause mortality, MALE, and MACE in PAD meta-analysis; individual CLI/CLTI cohorts report risk around NLR >4.6-8. Thresholds are not meta-analytically validated CLTI cut points.	Partially modifiable; interpret in context of infection, steroids, perioperative stress, and timing of sampling.	Illustrative pragmatic: <5 is a rounded approximation informed by reported PAD/CLTI thresholds around 4.6–8; >10 is a very-high-inflammatory-burden tier, not a validated CLTI cutoff. PAD meta-analysis plus CLI/CLTI observational cohorts; thresholds vary and require prospective validation.	[14,15,17,[38], [39], [40]]
Inflammatory; exploratory/mechanistic	IL-6	pg/mL; assay-dependent	No validated CLTI threshold	No validated CLTI threshold	No validated CLTI threshold	IL-6 signaling has genetic and mechanistic links to PAD/atherothrombosis, but current evidence does not validate IL-6 thresholds for CLTI limb outcomes or cell-therapy response.	Use as mechanistic substudy marker if assay platform and LLOQ are prespecified and constant.	Inflammatory biomarker syntheses plus mechanistic/genetic PAD evidence; exploratory for CLTI trial stratification.	[14,15,41]
Inflammatory/coagulation; pragmatic adjunct	Fibrinogen	g/L	<4 g/L	≥4 g/L	No validated upper tier; interpret continuously and with overall inflammatory profile	In the Pan et al. treated-cohort nomogram, fibrinogen >4 g/L was associated with lower odds of 6-month responder/remission status (OR 0.176, P = 0.003), together with age, baseline TcPO₂, arterial occlusion pattern, and log CD34⁺ cell count; AUC was 0.851. This is a treated-cohort association, not a validated predictive biomarker. No marker-specific source supports an additional >6 g/L tier.	Partially modifiable through treatment of inflammation/infection; prespecify as a stratification or interaction-test variable rather than an eligibility rule.	Study-derived: the ≥4 g/L boundary comes from a single retrospective treated mononuclear-cell cohort (n = 103); no randomized comparator or treatment-by-marker interaction was tested. No validated upper fibrinogen tier has been established.	[34]
Nutritional/inflammatory; core pragmatic	Albumin	g/dL	≥3.5	3.0-3.5	<3.0	Albumin reflects nutrition, inflammation, and frailty. Current cited evidence supports mortality, frailty, functional vulnerability, and CAR-related mortality associations; albumin alone should not be interpreted as a validated independent cutoff for delayed wound healing or amputation in the cited sources.	Potentially modifiable with nutrition, infection control, and anemia management; interpret with inflammatory status.	Clinical/pragmatic convention: albumin 3.5 and 3.0 g/dL are nutritional-risk conventions supported indirectly by nutritional, inflammatory, frailty, and wound-care/procedural observational evidence; they are not CLTI- or cell-therapy-specific validated thresholds.	[37,[42], [43], [44]]
Nutritional/hematologic; pragmatic adjunct	Hemoglobin/anemia	g/dL; sex-specific	≥12 female/≥13 male	10-12 female/10-13 male	<10	Baseline anemia was identified as an adverse prognostic factor for amputation and mortality in the non-revascularizable CLTI/JUVENTAS-derived cohort. Sex-specific no-anemia thresholds follow WHO 2024 hemoglobin cutoffs for adult nonpregnant women (≥12 g/dL) and men (≥13 g/dL). The <10 g/dL tier is an illustrative pragmatic high-risk anemia tier, not a WHO-derived CLTI threshold.	Modifiable depending on cause; investigate iron deficiency, CKD anemia, bleeding, inflammation; prespecify transfusion/anemia strategy.	Standard/convention-derived: sex-specific no-anemia thresholds follow WHO hemoglobin cutoffs; the <10 g/dL tier is an illustrative pragmatic trial-risk/safety tier. Nonrevascularizable CLTI/JUVENTAS-derived evidence supports anemia as an adverse prognostic factor but does not validate a cell-therapy-specific predictive threshold.	[4,45]
Renal; core pragmatic clinical/laboratory factor	eGFR and dialysis status	mL/min/1.73 m²; dialysis yes/no	≥45 and not dialysis-dependent	30-44	<30 or dialysis-dependent	Lower eGFR and dialysis dependence are adverse prognostic host-state factors; uremia, vascular calcification, and chronic inflammation provide biologic plausibility for poor regenerative milieu. The eGFR categories are adapted from KDIGO GFR-category conventions and grouped pragmatically for trial-design purposes.	Limited reversibility; optimize volume status, avoid nephrotoxins, document CKD-EPI or other equation; stratify or set safety rules.	Standard/convention-derived: eGFR strata are adapted from KDIGO GFR categories (≥45 mL/min/1.73 m² broadly G1–G3a; 30–44 G3b; <30 G4–G5). Dialysis dependence is a clinical safety and competing-risk category. CLTI renal-risk cohorts support adverse prognosis but not a validated cell-therapy interaction threshold.	[4,9,[46], [47], [48], [49]]
Infection; core clinical host-state factor	WIfI infection grade	0-3	0-1	2	3	Higher infection grade, especially grade 3 with tissue loss, predicts major amputation/poor limb outcomes in WIfI validation and no-option CLTI cohorts. Treat as clinical factor, not laboratory biomarker.	Modifiable with debridement, source control, culture-directed antibiotics, offloading, and osteomyelitis evaluation.	Standard/convention-derived: WIfI infection-grade categories 0–1, 2, and 3 follow WIfI classification/guideline conventions and are supported by WIfI validation and no-option cohort context; this is a clinical host-state factor, not a laboratory biomarker.	[1,20,[50], [51], [52]]
Metabolic; pragmatic adjunct	HbA1c	%	<6.5%	≥6.5%	No validated HbA1c-defined unfavorable tier	PACE was neutral overall; a post hoc subgroup with HbA1c <6.5% showed improved 12-month AFS (HR 0.46, 95% CI 0.21–0.99). This finding is exploratory, hypothesis-generating, and requires prospective validation; HbA1c should not be used as a treatment-selection rule. No validated HbA1c-defined unfavorable tier has been established for CLTI cell-therapy trial selection.	Modifiable over weeks to months; prespecify glycemic-management approach and avoid acute trial delays when limb status is unstable.	Study-derived/no adverse tier: the <6.5% boundary comes from the PACE post hoc subgroup context. The ≥6.5% category indicates absence of the favorable post hoc subgroup signal, not an adverse-response threshold. The previous >8.0% tier was removed because it was not CLTI/cell-therapy-derived.	[30]
Lipid/thrombo-inflammatory; pragmatic adjunct	Lipoprotein(a)	mg/dL or nmol/L (assay-reported; fixed conversion not recommended)	<30 mg/dL (<75 nmol/L)	30–50 mg/dL (75–125 nmol/L)	≥50 mg/dL (≥125 nmol/L)	High Lp(a) (>30 mg/dL) was associated with delayed wound healing after EVT in CLTI; it is not a cell-therapy response marker. The <30, 30–50, and ≥50 mg/dL risk zones follow broader European Atherosclerosis Society consensus conventions; fixed conversion between mg/dL and nmol/L is not recommended.	Largely nonmodifiable with standard care; report units/assay; use for risk stratification and counseling.	Study-derived plus consensus convention: the CLTI EVT study supports risk beginning around >30 mg/dL; broader EAS consensus provides pragmatic <30, 30–50, and ≥50 mg/dL risk zones. No validated CLTI cell-therapy interaction evidence exists.	[[53], [54], [55]]
Physiologic/perfusion; pragmatic adjunct	TcPO₂	mmHg	≥30	20-30	<20	In Pan et al., each 1-mmHg increase in baseline TcPO₂ was associated with higher odds of 6-month responder/remission status (OR 1.062, P = 0.006) within a treated cohort. This may reflect limb viability/prognosis rather than treatment-effect modification.	Partially modifiable via wound care, edema reduction, oxygenation, or revascularization where feasible; standardize measurement conditions and use for stratification/interaction testing rather than eligibility gating.	Guideline/pragmatic context: TcPO₂ is supported by guideline perfusion context and a single retrospective treated-cohort association. The ≥30, 20–30, and <20 mmHg categories are illustrative pragmatic perfusion bands for trial design; they may reflect limb viability/prognosis rather than treatment-effect modification and are not validated cell-therapy thresholds.	[1,34]
Inflammatory/anti-angiogenic; exploratory	CXCL10/IP-10	pg/mL; platform-dependent	No validated CLTI threshold	No validated CLTI threshold	No validated CLTI threshold	In vitro, IP-10 can inhibit VEGF-induced endothelial cell motility and tube formation. No validated CLTI prognostic thresholds or trial-stratification cutoffs exist.	Exploratory mechanistic marker only; centralize assay if included.	Mechanistic/in vitro anti-angiogenic evidence only; no validated clinical CLTI threshold.	[56]
Etiology; clinical host-state factor	Disease etiology	Clinical category	TAO with verified smoking cessation and controlled infection	Typical atherosclerotic CLTI or mixed etiology	Active inflammatory/CTD-associated disease, severe tissue loss, or high competing-risk phenotype	TAO cohorts often show more favorable limb outcomes than ASO; CTD/inflammatory nonatherosclerotic CLTI may have higher event burden. These are prognostic patterns, not validated treatment-etiology interactions.	Not directly modifiable except disease control and tobacco cessation; stratify by etiology.	Observational etiology-specific cohorts and subgroup reports; no randomized treatment-by-etiology interaction evidence.	[[57], [58], [59], [60], [61], [62]]
Behavioral; clinical host-state factor	Smoking status	Current/former/quit verified	Nonsmoker or sustained verified quitter	Current smoker engaged in cessation	Active smoker unwilling/unable to quit	Continued smoking is a dominant adverse factor in TAO/Buerger disease and undermines interpretation of cell-therapy response.	Modifiable; document cessation support and verification when relevant.	Guideline and TAO observational context; smoking cessation should be documented but should not be interpreted as a validated cell-therapy response predictor.	[1,57]
Functional reserve; clinical host-state factor	Frailty/ambulatory status	CFS; ambulatory status	CFS 1-4 and ambulatory	CFS 5-6 or limited ambulation	CFS ≥7, nonambulatory, or bedbound	Frailty/functional vulnerability affects discharge and competing-risk context; non-home discharge after EVT reflects vulnerability and resource burden. Use as covariate/stratifier rather than laboratory biomarker.	Partially modifiable through rehabilitation, pain control, nutrition, and care planning.	Standard/convention-derived: CFS 1–4, 5–6, and ≥7 follow Clinical Frailty Scale conventions, with functional-status and frailty observational evidence supporting prognostic relevance; no direct cell-therapy interaction evidence exists.	[42,43]

Open in a new tab

Abbreviations: AFS, amputation-free survival; AUC, area under the receiver operating characteristic curve; CAR, C-reactive protein-to-albumin ratio; CFS, Clinical Frailty Scale; CKD-EPI, Chronic Kidney Disease Epidemiology Collaboration; CLTI, chronic limb-threatening ischemia; CRP, C-reactive protein; CTD, connective tissue disease; eGFR, estimated glomerular filtration rate; EVT, endovascular therapy; HbA1c, glycated hemoglobin; IL-6, interleukin-6; IP-10, interferon-γ-inducible protein-10; LLOQ, lower limit of quantification; Lp(a), lipoprotein(a); MACE, major adverse cardiovascular events; MALE, major adverse limb events; NLR, neutrophil-to-lymphocyte ratio; OR, odds ratio; TAO, thromboangiitis obliterans; TcPO₂, transcutaneous oxygen pressure; WIfI, Wound-Ischemia-foot Infection classification.

Notes.

Numerical ranges represent approximate prognostic ranges for protocol development and prospective validation. They are not statistically optimized predictive cut points, validated clinical eligibility criteria, treatment-selection thresholds, or gatekeeping rules, and should not be used to automatically include or exclude individual patients. Continuous marker values should be retained for analysis whenever possible, with prespecified categories used only for stratified randomization, covariate adjustment, sensitivity analyses, and prospective validation.

Effect estimates are reported only where directly supported by the cited source and should be interpreted within the source-study context. HRs, ORs, treated-cohort associations, and post hoc subgroup observations were not pooled.

Treated-cohort associations and post hoc subgroup findings are hypothesis-generating and do not establish validated predictive biomarkers of differential cell-therapy benefit.

Boundary provenance: Numerical boundaries in this table were classified as study-derived, standard/convention-derived, or illustrative pragmatic values. Study-derived boundaries were taken from the cited CLTI/PAD outcome study, treated-cohort analysis, or post hoc trial subgroup. Standard/convention-derived boundaries follow established clinical classifications or guideline conventions, including WHO hemoglobin cutoffs [45], KDIGO eGFR categories [47], SVS/WIfI infection-grade conventions [50], European Atherosclerosis Society Lp(a) risk-zone conventions [55], and the Clinical Frailty Scale [42]. Illustrative pragmatic boundaries are explicitly identified as such and should not be cited as evidence-based marker-specific thresholds. The numerical ranges cannot be interpreted independently of the cited source context, original references, and interpretive caveats. Any reproduction of this table should include the boundary-provenance/source-context column and the accompanying notes to avoid misinterpretation as validated clinical cutoffs, treatment-selection thresholds, or eligibility/gatekeeping rules.

Modifiability indicates whether the marker or factor could plausibly change with clinical stabilization or supportive intervention; it does not imply that altering the marker improves cell-therapy efficacy or clinical outcomes.

Report CRP in mg/L; state the eGFR equation; specify assay platforms and LLOQ for IL-6/CXCL10; report Lp(a) in assay-reported units without fixed mg/dL-to-nmol/L conversion; standardize TcPO₂ measurement conditions.

2.1. Nutritional and renal markers

Serum albumin is a negative acute-phase reactant that reflects the interplay among nutrition, inflammation, and frailty. In CLTI patients undergoing below-the-knee endovascular therapy, a higher C-reactive protein-to-albumin ratio has been independently associated with increased long-term all-cause mortality [37]. Albumin overlaps with frailty biology [42], and adverse functional status and periprocedural complications predict non-home discharge after endovascular therapy in contemporary cohorts [43]. The JUVENTAS trial identified baseline anemia as an independent risk factor for amputation and mortality [4,27], consistent with malnutrition-inflammation-anemia syndromes observed in chronic vascular disease. For tabulation, sex-specific no-anemia thresholds were aligned with WHO 2024 hemoglobin cutoffs for adult nonpregnant women (≥12 g/dL) and men (≥13 g/dL); the <10 g/dL category is retained only as an illustrative pragmatic high-risk anemia tier for trial-design discussion [45]. Nutritional status is potentially modifiable and should be assessed and optimized within guideline-directed multidisciplinary CLTI care, including dietary counseling and correction of contributory factors such as anemia and infection [1,4]. For trial design, albumin, CAR, and anemia may help identify high-risk patients for stratified randomization, covariate adjustment, and sensitivity analyses. Any nutritional or anemia-focused optimization strategy should remain an optional, secondary, protocol-justified component and should not be interpreted as evidence that improving laboratory values alone normalizes baseline risk or enhances cell-therapy responsiveness.

Renal dysfunction is an adverse prognostic factor in CLTI. CKD—particularly ESRD requiring dialysis—worsens limb salvage and survival and increases periprocedural risk [46]. For tabulation, eGFR categories were adapted from KDIGO GFR-category conventions: ≥45 mL/min/1.73 m² broadly corresponds to G1–G3a, 30–44 mL/min/1.73 m² to G3b, and <30 mL/min/1.73 m² to G4–G5 or dialysis-risk strata [47]. Uremia promotes vascular calcification, chronic inflammation, and endothelial dysfunction [48,49] and may contribute to an impaired regenerative milieu. In JUVENTAS, lower baseline eGFR was independently associated with higher mortality [4,27], and conditional survival analyses show persistently high death rates among dialysis-dependent CLTI patients [9]. Advanced renal dysfunction is therefore a host-state factor that is often difficult to modify within trial timeframes and may define a population at high baseline risk of poor outcomes after cell therapy. Accordingly, kidney function should be prespecified as a stratification variable and, where necessary, as a protocol-defined eligibility or safety consideration rather than treated as a post hoc confounder. Dialysis dependence should be handled as a competing-risk and safety variable, not as a validated marker of differential cell-therapy benefit.

2.2. Inflammatory and coagulation markers

CRP is a classic acute-phase reactant reflecting IL-6-driven systemic inflammation. Elevated CRP is consistently associated with poorer cardiovascular and limb outcomes in CLTI cohorts [14,15]. Composite indices that incorporate CRP, such as the CAR, have been associated with increased long-term mortality after below-the-knee endovascular therapy [37]. Mechanistically, inflammatory cytokines promote endothelial dysfunction, thrombosis, and impaired angiogenesis, plausibly blunting regenerative responses [11,12,17]. Importantly, CRP is potentially modifiable through infection control and cardiovascular risk optimization, including intensified lipid-lowering therapy and smoking cessation; early MSC trials have reported reductions in inflammatory cytokines, including IL-1β and IFN-γ, alongside clinical signals such as improvements in rest pain and ulcer healing [63]. These biomarker shifts are supportive but should be interpreted with caution, given the small sample sizes and heterogeneity across products and endpoints. CRP and CAR should therefore be treated as prognostic and context-sensitive inflammatory indices, not as validated treatment-response biomarkers or cell-therapy eligibility cutoffs. In addition to reflecting risk, a high-CRP phenotype plausibly denotes an IL-6/TNF-α-driven environment that suppresses endothelial repair and promotes immunothrombosis, thereby reducing responsiveness to paracrine pro-angiogenic signaling [11,12,17]. This mechanistic rationale supports prospective stratification and interaction testing, but it does not by itself establish predictive utility.

IL-6 is an upstream inflammatory mediator implicated causally in atherothrombotic risk. Human genetic evidence implicates IL-6 receptor signaling in susceptibility to peripheral artery disease [41]. Assay performance varies across platforms, so reporting should specify the methods and regional reference ranges used. At present, IL-6 has greater mechanistic than routine clinical utility but may serve as an exploratory inflammatory marker when validated assays are available [11,12]. Where feasible, centralized testing and prespecified timing relative to infection control/debridement can reduce interpretability bias. Because no validated CLTI staging threshold or cell-therapy interaction threshold exists for IL-6, IL-6 should generally be reserved for mechanistic substudies unless prospectively validated for broader trial use.

The NLR is an inexpensive index of systemic inflammation and immune dysregulation. A systematic review and meta-analysis in peripheral artery disease cohorts found that high NLR was associated with at least a twofold increased risk of all-cause mortality, major adverse limb events, and major adverse cardiovascular events [38]. Values around 4.6–5 have been identified as marking high-risk states in CLTI cohorts [14], with very high values, such as >10, associated with markedly increased risk of major adverse events in individual studies [39]. In a retrospective CLI cohort, ROC-derived thresholds, including NLR ≥8 for 1-year mortality and ≥6 for MALE, identified patients at substantially higher risk of adverse outcomes [40]. Raised NLR reflects neutrophil-predominant inflammation—with release of reactive oxygen species, proteases, and neutrophil extracellular traps—alongside relative lymphopenia, which together promote microthrombosis and impaired angiogenesis [17]. Mechanistically, this neutrophil-dominant inflammation may increase microvascular obstruction and protease burden in chronic wounds, thereby limiting the effectiveness of secreted growth factors and extracellular vesicles, which are central to MSC-based therapies [17]. Because NLR is sensitive to acute infection, steroid exposure, and perioperative stress, trials should prespecify the timing of measurements and interpret extreme values in the context of infection status. NLR thresholds should be regarded as approximate, cohort-dependent prognostic ranges for prospective validation, not meta-analytically validated CLTI cutoffs or treatment-selection rules.

Fibrinogen is an acute-phase reactant reflecting systemic inflammation and coagulation activation. In a retrospective treated-cohort analysis of 103 patients with “no-option” CLI receiving autologous mononuclear cell transplantation, fibrinogen >4 g/L was associated with lower odds of 6-month responder/remission status (OR 0.176, P = 0.003), together with age, TcPO₂, arterial occlusion pattern, and log CD34⁺ cell count; the multivariable nomogram achieved an AUC of 0.851 [34]. Because all patients received cell therapy and no randomized comparator or treatment-by-marker interaction was tested, this finding should be interpreted as a treated-cohort association and candidate stratification variable, not as a validated predictive biomarker of differential cell-therapy benefit. Elevated fibrinogen may reflect a prothrombotic inflammatory phenotype that could limit microcirculatory perfusion and reduce the physiologic substrate for therapeutic angiogenesis. This hypothesis supports prespecified fibrinogen stratification and prospective randomized host-state × treatment interaction testing, but does not justify using fibrinogen as a treatment-selection cutoff or eligibility gate.

CXCL10 (IP-10) is an interferon-γ-inducible chemokine with anti-angiogenic activity. In vitro, IP-10 can inhibit VEGF-induced endothelial cell motility and capillary-like tube formation, in part by inhibiting calpain [56]. Acting via CXCR3, CXCL10 may therefore antagonize angiogenic signaling, providing a mechanistic rationale to explore it as a negative marker of angiogenic responsiveness. However, IP-10 assays are largely research-grade with substantial inter-platform variability, so at present it remains an exploratory mechanistic marker rather than a clinically validated prognostic tool. Current evidence for CXCL10/IP-10 is mechanistic and in vitro; no validated CLTI prognostic threshold, trial-stratification cutoff, or predictive role as a cell-therapy biomarker has been established.

More broadly, because biomarker-outcome associations are influenced by care processes, trial procedures should standardize wound care pathways, including debridement, offloading, and infection surveillance, and incorporate nutritional assessment into baseline care. Outcomes in CLTI also reflect care delivery; streamlined multidisciplinary pathways can reduce major amputation rates in contemporary cohorts [64]. Nutritional optimization is recognized as a component of wound care and may support healing in patients at risk of malnutrition [44]. These care-delivery measures should be considered part of guideline-directed CLTI management and trial standardization, not evidence that biomarker optimization itself improves cell-therapy efficacy.

2.3. Infection and limb-threat staging

Active infection amplifies systemic inflammation, impairs wound healing, and increases the risk of amputation. The WIfI classification integrates wound extent, ischemia, and foot infection; higher infection grades are associated with increased risk of major amputation, particularly when combined with extensive tissue loss and severe ischemia [1,50]. In “no-option” cohorts, WIfI stages 3–4—especially with infection grade 3—are associated with poor limb salvage and survival despite intensive wound care [20]. Adherence to international guideline-based infection management, including debridement, culture-directed antibiotics, and offloading, is therefore a prerequisite before considering advanced therapies [7,51]. In the context of cell therapy trials, active infection is both a safety concern and a major confounder of inflammatory biomarkers; therefore, protocols should prespecify minimum infection-control criteria and standardized wound-care requirements at enrollment. WIfI infection grade should be treated as a clinical host-state factor, not as a laboratory biomarker, and should support protocol-defined safety assessment, stratification, and standardized care rather than automatic biomarker-based gatekeeping.

A JUVENTAS subanalysis validated WIfI in non-revascularized CLTI patients [20], and BEST-CLI confirmed that advanced WIfI stage independently predicts poor outcomes in contemporary revascularized cohorts [52]. An observed anomaly was a higher 1-year amputation rate in WIfI stage 2 than in stage 3, partly driven by W0–I3-fI0, representing severe ischemia with rest pain but no wound or infection, which was assigned to stage 2; reclassification of this phenotype has been suggested. This finding is relevant to “no-option” CLTI, in which rest-pain phenotypes may carry substantial risk despite limited tissue loss. Accordingly, limb-threat staging should be interpreted alongside objective perfusion testing, imaging, infection control status, and MDT judgment, and should not be replaced by host-state marker profiles.

2.4. Lipoprotein(a) and disease etiology

Lipoprotein(a) is a prothrombotic, pro-inflammatory lipoprotein associated with atherosclerotic risk [53]. In CLTI patients undergoing endovascular therapy, elevated Lp(a) has been linked to delayed wound healing and adverse limb outcomes [54]. Concentrations above approximately 30 mg/dL have been used to identify higher-risk patients [54]. For tabulation, Lp(a) categories were aligned with European Atherosclerosis Society consensus risk zones—<30 mg/dL (<75 nmol/L), 30–50 mg/dL (75–125 nmol/L), and ≥50 mg/dL (≥125 nmol/L)—while recognizing that fixed conversion between mg/dL and nmol/L is not recommended [55]. Assays report Lp(a) in mg/dL or nmol/L, which are not directly interchangeable; careful reporting of units is essential [53,55]. Because available therapies produce only modest reductions in Lp(a) [53], Lp(a) may function primarily as a relatively fixed risk marker for counseling, baseline risk stratification, and trial enrichment in this setting. In trials, prespecifying units and the analytic approach (continuous versus threshold-based) is particularly important for interpretability. Lp(a) is not currently a validated CLTI cell-therapy response marker, and any threshold-based use should be prespecified and prospectively validated.

Underlying etiology strongly modifies prognosis and apparent outcomes after therapy. TAO is associated with a more favorable course, with 5-year limb-salvage rates of approximately 90%–95% and a lower risk of major amputation after autologous cell therapy compared with atherosclerotic CLTI [57]. These observations should be interpreted as etiology-associated prognostic patterns and treated-cohort signals, not as proof of a randomized treatment-by-etiology interaction. In contrast, inflammatory nonatherosclerotic CLTI, including CTD-associated vasculitis, often demonstrates poorer wound healing and higher rates of reintervention and amputation than typical atherosclerotic CLTI, particularly when tissue loss is present [58]. Data on CTD-specific EVT are mixed, with acceptable limb salvage in a small case series [59] but poorer AFS and frequent repeat revascularization in a multicenter registry [60]. Nevertheless, these patients may represent biologically plausible candidates for therapeutic angiogenesis. Clinical data from autologous BMMNC implantation in collagen diseases, including systemic sclerosis, suggest feasibility and potentially favorable limb salvage, warranting prospective evaluation [61,62]. Nomogram analyses also highlight the interplay among macrovascular anatomy, microcirculatory reserve, systemic inflammatory state, and age in determining outcome probability within treated cohorts [34]. Etiology-specific patterns are summarized in Supplementary Table S3. All etiology-specific associations should be interpreted as prognostic unless prospective randomized treatment-by-etiology interaction analyses demonstrate differential cell-therapy benefit.

Smoking status is a particularly important behavioral host-state factor, especially in TAO/Buerger disease. Continued smoking is strongly associated with worse limb prognosis and undermines interpretation of apparent cell-therapy benefit in TAO cohorts [1,57]. Smoking cessation should be documented and supported as part of standard care and trial conduct. However, smoking status should not be interpreted as a validated predictor of differential cell-therapy response; rather, it should be treated as a protocol-defined clinical factor, an adherence variable, and an etiology-relevant stratification or sensitivity-analysis variable.

2.5. Perfusion, metabolic, functional, and other adjunct markers

TcPO₂ reflects microcirculatory perfusion and tissue oxygenation. In the same retrospective treated-cohort nomogram study, each 1-mmHg increase in baseline TcPO₂ was associated with higher odds of 6-month responder/remission status (OR 1.062, P = 0.006), and values ≥ 30 mmHg were associated with higher responder/remission probability [34]. However, because TcPO₂ also reflects baseline limb viability, wound severity, and perfusion reserve, this association may represent prognosis rather than treatment-effect modification. TcPO₂ should therefore be treated as a physiologic prognostic index and candidate variable for prospective stratification and host-state × treatment interaction testing, not as a validated predictive biomarker of cell-therapy benefit. TcPO₂ availability varies across centers. Where included in protocols, measurement conditions, including site selection, temperature, infection status, oxygen supplementation, edema, and timing relative to debridement/offloading, should be standardized to support reproducible stratification and interaction testing.

HbA1c reflects chronic glycemic exposure and may capture aspects of metabolic host state relevant to wound healing, infection risk, and microvascular dysfunction. In PACE, the overall phase III trial was neutral for AFS benefit in Rutherford 5 no-option CLTI; although a post hoc subgroup with HbA1c <6.5% showed improved 12-month AFS (HR 0.46, 95% CI 0.21–0.99), this observation remains exploratory and hypothesis-generating [30]. HbA1c should therefore not be interpreted as a validated predictive biomarker, treatment-selection cutoff, or clinical eligibility rule. For future trials, glycemic management should be standardized and reported, but acute trial delays for glycemic optimization should be avoided when limb status is unstable. HbA1c-defined hypotheses should be prospectively tested through prespecified stratification and treatment-interaction analyses.

Frailty and ambulatory status are clinical host-state factors that reflect physiologic reserve, rehabilitation potential, competing mortality risk, discharge disposition, and capacity to benefit from limb-salvage interventions. Frailty overlaps with malnutrition, inflammation, renal dysfunction, pain, and functional decline [42,43]. Frailty and ambulatory status are not laboratory biomarkers and should not be used as biomarker-based exclusion rules. Instead, they should be prespecified as clinical covariates, stratification variables, or sensitivity-analysis factors, particularly in trials where AFS, wound healing, quality of life, or functional limb salvage are key endpoints.

Taken together, the host-state markers discussed above provide a biologically and clinically plausible basis for improving the design and interpretation of no-option or poor-option CLTI cell therapy trials. Their current evidentiary role is primarily prognostic: they may support baseline risk stratification, prognostic enrichment, covariate adjustment, stratified randomization, standardized care pathways, and prespecified host-state × treatment interaction testing. They should not be used as validated predictive biomarkers, treatment-selection rules, or gatekeeping thresholds unless and until prospective randomized evidence demonstrates differential cell-therapy benefit across marker-defined strata.

3. Existing risk stratification approaches and their limitations

Several risk-stratification tools have been developed for CLTI, including WIfI and GLASS, which have improved standardization of limb-threat and anatomic assessment and help predict outcomes [1,50]. However, systematic reviews highlight constraints in capturing host-state biology and call for models that include systemic inflammation, renal function, and frailty [18,19]. WIfI and GLASS were designed to stage limb threat and anatomy; they do not directly quantify systemic physiologic reserve, including inflammation, nutritional status, renal dysfunction, frailty, dialysis dependence, or behavioral and etiologic factors that may influence wound healing, competing mortality risk, and regenerative capacity. Accordingly, host-state assessment should be viewed as complementary to, not a replacement for, WIfI/GLASS staging, objective perfusion testing, vascular imaging, infection control, wound care, and MDT determination of revascularization feasibility.

In “no-option” cohorts, WIfI retains important prognostic value, and advanced WIfI stage, severe infection, extensive tissue loss, and severe ischemia identify patients at high risk of limb loss and death [20]. For cell therapy trials, these features should be handled as protocol-defined clinical host-state and safety factors for baseline stratification, covariate adjustment, standardized care, and sensitivity analyses. They should not be interpreted as validated predictors of differential cell-therapy benefit, nor should they be used as automatic biomarker-based gatekeeping criteria. In some trials, uncontrolled sepsis, rapidly progressive tissue loss, planned imminent major amputation, or urgent revascularization reassessment may appropriately defer eligibility confirmation; however, such decisions should reflect clinical safety and MDT judgment rather than isolated host-state marker values. When patients are clinically stable, continuous host-state variables should be retained for analysis wherever possible, with any prespecified categories used to balance baseline risk and support prospective validation rather than to mandate inclusion or exclusion.

Streamlined multidisciplinary care pathways can reduce major amputation rates, suggesting that optimization of care delivery—including standardized wound care, infection control, debridement, offloading, nutrition assessment, cardiovascular risk management, and reassessment of revascularization feasibility—may be as important as novel therapies [64]. These care-delivery measures should be distinguished from a formal trial run-in or optimization intervention. Urgent clinical stabilization before eligibility confirmation represents guideline-directed care, whereas any pre-randomization run-in or post-randomization optimization strategy should remain an optional, secondary, protocol-justified design component for selected clinically stable patients. Such strategies should be accompanied by transparent reporting of screened patients, pre-randomization exclusions, run-in failures, clinical deterioration, and events occurring before randomization to avoid selection, attrition, survivorship, regression-to-the-mean, co-intervention, and immortal-time biases.

Conditional survival analyses show that mortality risk remains highest in the early years after CLTI diagnosis, especially in patients with renal dysfunction, heart failure, or severe tissue loss [9]. Contemporary conservative-care outcomes of approximately 60% 12-month AFS provide a realistic benchmark for trial design [3,4]. These data are useful for event-rate assumptions, safety monitoring, competing-risk planning, and stratified randomization, but they do not establish host-state profiles as validated treatment-selection tools. Because early mortality is a major competing risk, enrolling very high-risk host-state phenotypes without prespecified stratification can dilute limb endpoints and complicate interpretation of treatment effects. Conversely, excluding such patients without transparent justification may reduce generalizability and obscure the population most likely to be considered for regenerative approaches in practice.

Overall, integration of inflammatory, nutritional, renal, infection, perfusion, functional, behavioral, and etiology-related host-state markers with WIfI/GLASS assessment represents a logical extension of current CLTI trial design [13,18]. The purpose of this integration is prognostic enrichment, baseline risk balancing, covariate adjustment, blinded outcome interpretation, and prespecified host-state × treatment interaction testing—not clinical gatekeeping or unvalidated prediction of cell-therapy response. Prospective validation will require standardized assays, harmonized endpoints, consistent background care, transparent reporting of cell products and concomitant therapies, and trial designs that support stratified randomization and cautious interpretation of host-state × treatment interaction analyses.

4. Toward host-state-informed stratification: A conceptual framework

The evidence synthesized in this review supports the biological plausibility of using host-state markers to inform prognostic enrichment and stratified enrollment for trial-design purposes, rather than to mandate individual eligibility decisions or clinical treatment selection, in cell therapy trials for MDT-confirmed no-option or poor-option CLTI. Most candidate markers remain prognostic rather than validated predictors of differential treatment benefit. Treated-cohort associations and post hoc subgroup observations should therefore be interpreted as hypothesis-generating candidate variables for prospective host-state × treatment interaction testing, not as validated predictive biomarkers of cell-therapy response.

Accordingly, we propose that host-state markers be used to characterize baseline risk, prespecify stratified randomization and covariate adjustment, and enable hypothesis-driven host-state × treatment interaction analyses in prospectively designed trials. Fig. 1 illustrates a conceptual host-state assessment pathway for trial-design stratification after MDT confirmation of no-option or poor-option CLTI. The pathway is intended to support protocol development, baseline risk balancing, and prospective validation; it should not be interpreted as a prescriptive clinical algorithm. Action labels in the pathway indicate protocol-dependent trial-design options rather than automatic clinical instructions.

The numerical ranges displayed in Table 1, Fig. 1, and Supplementary Table S4 should be interpreted as approximate prognostic ranges for protocol development, prospective trial stratification, and future validation, rather than as validated clinical cutoffs, treatment-selection rules, or eligibility/gatekeeping criteria. These ranges were synthesized from heterogeneous observational cohorts, trial subgroups, treated-cohort analyses, standard clinical classification systems, and guideline-informed risk conventions. They are assay-dependent, context-specific, and require local calibration and prospective validation. The provenance of each numerical boundary in Table 1 and Supplementary Table S4 is classified as study-derived, standard/convention-derived, or illustrative pragmatic; boundaries without marker-specific outcome sources are labeled as pragmatic trial-design examples, not evidence-based thresholds. In future studies, host-state variables should be captured continuously whenever possible, with any prespecified categories used for stratified randomization, covariate adjustment, sensitivity analyses, and exploratory host-state × treatment interaction testing rather than for automatic patient exclusion or clinical decision-making. The descriptive boundary-provenance/source-context information provided in Table 1 and Supplementary Table S4 should not be interpreted as a formal evidence-grading, risk-of-bias, or certainty-of-evidence assessment.

4.1. Potential core host-state marker panel

A minimum core panel could include CRP, NLR, albumin, eGFR/dialysis status, and WIfI infection grade. These measures are widely available, inexpensive, and feasible in routine vascular practice. Reporting should standardize units, including CRP in mg/L, specify the eGFR equation used (e.g., CKD-EPI), and prespecify sampling settings and timing (e.g., before or after debridement and during or after infection treatment), because acute infection and periprocedural stress can transiently shift inflammatory indices. Interpretation should be anchored to contemporaneous clinical assessment and should avoid rigid “gatekeeping” cutoffs; where feasible, markers should be captured as continuous variables, with prespecified threshold categories used for stratification, covariate adjustment, and sensitivity analyses.

For trial implementation, host-state markers can be grouped by intended use. Core pragmatic markers include routinely available measures suitable for multicenter stratification, such as CRP, NLR, albumin, eGFR/dialysis status, and WIfI infection grade, supplemented by clinical host-state factors such as etiology and infection status. Pragmatic adjunct markers, including hemoglobin/anemia, HbA1c, fibrinogen, Lp(a), frailty score, and TcPO₂, may be incorporated when clinically relevant and standardized. Exploratory mechanistic markers, such as IL-6, CXCL10/IP-10, omics-derived signatures, or cell-product potency correlates, should be reserved for mechanistic substudies unless prospectively validated for broader trial use.

Adjuncts may include IL-6, where validated assays exist; fibrinogen, given its association with outcomes within a treated cohort, interpreted as a hypothesis-generating treated-cohort association rather than as a validated predictive biomarker of differential treatment benefit [34]; and HbA1c and Lp(a) for risk characterization, trial enrichment, and prespecified subgroup hypotheses. HbA1c should not be treated as a treatment-selection biomarker on the basis of post hoc subgroup findings alone. CXCL10/IP-10 is best reserved for mechanistic studies, and TcPO₂ is valuable where available as a physiologic/perfusion index. TcPO₂ may reflect baseline limb viability, wound severity, and perfusion reserve; therefore, it should be treated as a physiologic prognostic index and candidate stratification variable, not as a validated predictive biomarker of cell-therapy benefit.

When markers are incorporated into trial protocols, investigators should define minimum safety and stabilization requirements at enrollment, such as the absence of uncontrolled systemic infection, while avoiding automatic exclusion based on isolated marker values. Protocols should also prespecify whether each marker is used as a continuous covariate, a stratification variable, a sensitivity-analysis variable, or an exploratory interaction-testing variable. Supplementary Table S4 presents a hypothetical implementation framework intended as a starting point for protocol development and prospective validation. It should not be interpreted as a validated clinical decision tool.

4.2. Conceptual host-state profiling for trial-design stratification

On the basis of current evidence, patients with MDT-confirmed no-option or poor-option CLTI might be provisionally categorized into three host-state profiles—favorable, intermediate, and unfavorable—to support trial enrichment, stratified randomization, covariate adjustment, and prospective validation (Table 1; Supplementary Table S4). These profiles are pattern-based trial-design strata. The proposed numerical values are approximate prognostic examples, not definitive clinical cutoffs, treatment-selection thresholds, or eligibility/gatekeeping rules. They require assay calibration, standardized measurement timing, local validation, and prospective testing before any clinical implementation.

For patients with discordant host-state markers, profile assignment should be based on prespecified safety-critical overrides and the overall host-state pattern rather than a simple count of abnormal variables. Uncontrolled sepsis, rapidly progressive tissue loss, planned imminent major amputation, or urgent revascularization reassessment should override marker-based profiling and defer eligibility confirmation. Such deferral reflects clinical safety and MDT judgment, not automatic biomarker-based exclusion. In otherwise clinically stable patients, continuous marker values should be retained for analysis, while categorical profiles should be used for stratification, covariate adjustment, and sensitivity analyses rather than rigid exclusion.

Favorable prognostic profiles may include low inflammatory markers, such as CRP <30 mg/L and NLR <5; adequate nutrition, such as albumin ≥3.5 g/dL; preserved renal function, such as eGFR ≥45 mL/min/1.73 m²; no active or only mild infection, such as WIfI infection grade 0–1; and a favorable etiology. These values are approximate examples for prospective validation. Patients with a favorable profile may be considered for protocol-defined trial enrollment after MDT confirmation and application of trial-specific safety criteria, particularly in early-phase signal-detection studies, while preserving appropriate controls, stratification, and blinded outcome assessment to enable unbiased estimation of efficacy.

Intermediate prognostic profiles may include moderately abnormal but potentially modifiable markers, such as CRP 30–100 mg/L, NLR 5–10, albumin 3.0–3.5 g/dL, WIfI infection grade 2, and eGFR 30–44 mL/min/1.73 m². For this group, clinical stabilization, reassessment, or an optional protocol-defined optimization strategy may be considered only when clinically appropriate and when urgent infection control, wound management, or revascularization reassessment is not delayed. A time-limited run-in or post-randomization optimization strategy may reduce transient inflammatory misclassification and host-state heterogeneity at the time of randomization or dosing, but whether such strategies improve cell-therapy efficacy remains unproven. The conservative-care benchmark of approximately 60% 12-month AFS provides context for the ethical feasibility of brief stabilization windows in clinically stable patients [3,4], but this should not be used to justify delaying urgent limb-management decisions.

Unfavorable prognostic profiles may include persistently hostile biology, such as CRP >100 mg/L, NLR >10, albumin <3.0 g/dL, WIfI infection grade 3, eGFR <30 mL/min/1.73 m², or dialysis dependence, often with extensive tissue loss or aggressive etiology. These features identify high baseline risk and competing-risk burden; they do not establish a lack of cell-therapy benefit or validated treatment resistance. In such patients, immediate priorities typically include infection control, symptom management, reassessment of revascularization feasibility, wound-care optimization, goals-of-care discussion, and MDT decision-making regarding limb salvage versus amputation. Cell-based interventions in this setting should be evaluated cautiously in carefully monitored trials, with explicit stopping rules, realistic counseling regarding competing risks and expected effect sizes, and consideration of separate safety and feasibility cohorts, if appropriate, rather than automatic exclusion or automatic assignment to a primary efficacy cohort.

4.3. Optional optimization and run-in strategies as secondary trial-design considerations

Optimization or run-in strategies should be viewed as optional, secondary protocol-level considerations after MDT confirmation, baseline host-state assessment, prognostic enrichment, and stratification procedures have been defined. They are not required components of the proposed framework and should not supersede guideline-directed CLTI care, timely infection control, revascularization reassessment, or unbiased estimation of treatment effect. The central purpose of the framework remains prognostic enrichment, stratified randomization, covariate adjustment, blinded outcome adjudication, and prespecified host-state × treatment interaction testing.

The concept of a pretreatment optimization phase for selected clinically stable intermediate-risk patients is supported by several considerations: partial modifiability of CRP, NLR, albumin, and infection grade through infection control, lipid-lowering therapy, smoking cessation, glycemic optimization, and nutritional support [11,12,37]; observed reductions in inflammatory markers in early MSC trials that coincided with clinical improvement [63]; mechanistic plausibility given the predominantly paracrine mode of action of many cell therapies [11,12,21]; and ethical feasibility of a short stabilization window in clinically stable patients, supported by contemporary conservative-care outcomes [3,4]. From a trial-design perspective, a protocolized optimization phase may also reduce biomarker misclassification due to transient inflammation and improve the interpretability of host-state-stratified analyses. However, these considerations support only hypothesis generation and protocol planning; they do not demonstrate that improving host-state marker profiles enhances cell-therapy efficacy.

An optimization bundle may target infection control, including debridement, culture-directed antibiotics, and offloading following international guidelines [7,51]; intensified cardiometabolic care; nutrition, including protein intake of approximately 1.2–1.5 g/kg/day, dietitian involvement, and anemia treatment [4,44]; and pain management and rehabilitation. Frailty assessment, such as the Clinical Frailty Scale, can support realistic goal-setting and contextualize biomarker abnormalities [42]. Host-state markers may be reassessed after stabilization or optimization to determine whether the host profile has changed. Reassessment should be domain-specific: inflammatory and infection markers may change over days to weeks, whereas nutritional, glycemic, renal, and functional measures may require longer follow-up. Protocols should prespecify objective criteria for eligibility confirmation, randomization, dosing, or reassessment and should report adherence to the optimization bundle and its components as potential determinants of outcomes.

Although a time-limited optimization run-in may improve clinical stability and reduce biomarker misclassification, it may also introduce bias if implemented before randomization. Patients who deteriorate, undergo major amputation, die, withdraw, or fail to achieve predefined stability criteria during the run-in may be excluded from randomization, producing selection, attrition, and survivorship bias [65]. Similarly, if outcome follow-up begins only after the run-in period is complete, the design may introduce immortal-time bias by excluding events that occur during the optimization period [66]. Biomarker improvement during the run-in may also reflect regression-to-the-mean, resolution of transient infection, or intensified background care rather than a durable change in regenerative potential or treatment responsiveness. Therefore, any pre-randomization run-in should be limited to clinically stable patients, should not delay urgent infection control or limb-management decisions, and should be accompanied by prospective reporting of all screened patients, run-in failures, reasons for non-randomization, amputations, deaths, withdrawals, and clinical events occurring during the run-in.

We distinguish three related but methodologically different concepts. First, clinical stabilization before eligibility confirmation, such as urgent infection control, debridement, analgesia, or reassessment of revascularization feasibility, represents ethical standard care and should not be framed as a trial intervention. Second, a pre-randomization run-in may reduce transient host-state heterogeneity but can enrich for patients who survive or stabilize before randomization; such designs must capture all screened patients and all events occurring before randomization. Third, a post-randomization optimization strategy can evaluate the optimization concept more directly when all randomized participants are included in the intention-to-treat analysis. Accordingly, run-in or optimization strategies should remain ancillary to the main trial-design objectives and should be implemented only with prespecified safeguards against selection, attrition, co-intervention, regression-to-the-mean, and immortal-time biases.

5. Consequences for trial design

5.1. Enrichment and stratification strategies

The host-state evidence reviewed here suggests several consequences for cell therapy trials in MDT-confirmed no-option or poor-option CLTI. Early-phase studies may consider protocol-defined enrichment for more favorable prognostic host-state profiles to maximize safety and improve early signal detection [21]. Definitive randomized trials could incorporate provisional host-state profile–based stratified randomization, with prespecified host-state × treatment interaction analyses to test for differential treatment effects by risk category [3,4,[23], [24], [25], [26], [27], [28], [29], [30], [31],36]. For definitive efficacy trials, placebo- or sham-controlled designs with blinded outcome adjudication should be preferred where ethically and operationally feasible, particularly when endpoints include wound healing, pain, functional status, or quality-of-life measures.

Because most candidate markers in CLTI are prognostic rather than validated predictors of treatment benefit, these approaches should be framed primarily as prognostic enrichment, baseline risk balancing, covariate adjustment, and stratified enrollment. Interaction testing should be treated as hypothesis-driven unless the trial is specifically powered for host-state × treatment interaction analyses. Treated-cohort associations and post hoc subgroup observations should not be used to define validated treatment-selection rules. When intermediate profiles are included, protocols should clearly distinguish urgent clinical stabilization before eligibility confirmation, optional pre-randomization run-in designs, and post-randomization optimization strategies. Any optimization or run-in component should remain ancillary to the primary trial-design objectives of unbiased treatment-effect estimation, prognostic enrichment, stratified randomization, covariate adjustment, blinded outcome adjudication, and prespecified host-state × treatment interaction testing.

Supplementary Fig. S2 illustrates one possible host-state-stratified design in which host-state strata are distinguished from randomized treatment arms. This figure should be interpreted as an optional, hypothesis-generating trial-design example, not as a required structure for CLTI cell-therapy trials or a clinical algorithm. Collectively, these approaches move beyond treating no-option or poor-option CLTI as biologically homogeneous and align with precision-medicine paradigms that complement anatomy-based staging [13].

Stratification could be implemented using a prespecified panel (Table 1; Supplementary Table S4) with centralized laboratory testing where feasible, or with harmonized local assays, including calibration, quality-control procedures, and predefined rules for repeat testing after infection treatment. Numerical host-state ranges should be retained as approximate prognostic ranges for protocol development and prospective validation, not as validated eligibility cutoffs, treatment-selection thresholds, or gatekeeping rules. Operationally, trials could choose among several strategies: enrichment for more favorable prognostic profiles in first-in-human or early signal-detection studies; stratified randomization across provisional host-state profiles to balance baseline risk; and prognostic enrichment that defers eligibility confirmation only for clearly unsafe or clinically unstable conditions, such as uncontrolled sepsis, while preserving real-world heterogeneity when clinically appropriate.

Protocols should predefine marker timing, including baseline and post-stabilization or post-optimization sampling; the use of central versus local laboratories; and the handling of transient inflammatory surges, such as acute infection, recent debridement, or perioperative stress. Where possible, marker sampling should align with clinically meaningful stability time points, such as after initial debridement and initiation of antimicrobial therapy, to improve interpretability. Host-state variables should be analyzed continuously whenever possible, with categorical profiles used for stratified randomization, covariate adjustment, sensitivity analyses, and prospective validation rather than for automatic inclusion or exclusion.

Interaction analyses should be prespecified and interpreted cautiously, recognizing that most trials are powered for overall treatment effects rather than subgroup effects. To reduce false-positive subgroup inferences, trials should prespecify the statistical estimand, define the primary analysis population, estimate the overall treatment effect before interaction analyses, use hierarchical testing where appropriate, and limit the number of interaction tests or adjust for multiplicity when multiple markers or profiles are examined. Analysis plans should include intention-to-treat analyses, with prespecified per-protocol sensitivity analyses that account for marker shifts between baseline, stabilization, dosing, and follow-up. Given uncertainty in event rates, blinded sample-size re-estimation may be considered while preserving allocation concealment.

Stratification can be layered with limb-based variables, such as WIfI stage, and key comorbidities to minimize imbalance, and analysis should adjust for prespecified prognostic factors. Host-state panels should be analyzed both continuously and as categorical profiles to reduce information loss. Finally, because cell therapy products are biologically heterogeneous, trials should avoid pooling fundamentally different products, such as BMMNCs versus MSCs, within a single primary efficacy analysis unless explicitly modeled. Product type, dose, delivery route, potency characteristics, and delivery schedule should be prespecified and reported as potential effect modifiers.

Alternative trial designs may be preferable depending on the study objective. For definitive efficacy trials, stratified randomization without a pre-randomization run-in may provide the least biased estimate of treatment effect while balancing favorable, intermediate, and unfavorable host-state profiles across study arms. Where ethically and operationally feasible, placebo or sham procedures and blinded endpoint adjudication should be incorporated to reduce expectation, performance, and assessor bias. Prognostic enrichment without biomarker conversion may be appropriate when the goal is to defer or exclude only patients with clearly unsafe or unstable biology, such as uncontrolled sepsis, while preserving real-world heterogeneity.

A uniform stabilization period before randomization may improve safety and standardize background care, but it requires transparent reporting of pre-randomization attrition, deaths, amputations, clinical deterioration, withdrawals, and reasons for non-randomization. If the optimization strategy itself is a scientific question, post-randomization or factorial designs can test optimization versus usual care more directly, provided all randomized participants are followed according to intention-to-treat principles. Adaptive enrichment or platform designs may be useful in later-phase development, provided that interim decision rules, multiplicity control, sample-size re-estimation procedures, and analyses of host-state × treatment interactions are prespecified. Patients with persistently unfavorable prognostic profiles may be most appropriately studied in separate safety and feasibility cohorts with explicit stopping rules, realistic counseling regarding competing risks, and conservative effect-size assumptions, rather than being automatically pooled into primary efficacy cohorts or excluded from all research.

5.2. Endpoint considerations

A systematic review of 49 CLTI trials involving 11,667 patients found substantial heterogeneity in outcome definitions: only 20.4% used quantitative wound assessment, MALE definitions varied considerably, and no validated CLTI-specific quality-of-life instrument exists [67]. This evidence supports adopting 12-month AFS with blinded adjudication as a primary endpoint in definitive trials, with standardized secondary outcomes including wound healing, limb events, pain, quality of life, functional status, and perfusion indices [67]. Event-rate assumptions should be based on contemporary conservative-care, placebo, or best-supportive-care risks rather than older historical assumptions [3,30]. Because mortality is a major competing risk in CLTI, protocols should prespecify how death is handled in limb-endpoint analyses, for example as part of AFS and/or through competing-risk methods for limb-specific outcomes.

Beyond AFS, trials should standardize definitions of major amputation, reintervention, wound healing, and time-to-closure, and should specify the approach to blinded wound assessment, such as serial photography and planimetry, where feasible. Patient-centered outcomes, including pain, functional status, ambulation, quality of life, and health-economic endpoints, are important given the morbidity and resource intensity of CLTI care. These outcomes should be interpreted within the context of baseline host-state profile, background care, and competing mortality risk.

Perfusion measures, including toe pressure, TcPO₂, and imaging-derived microcirculatory indices, can support mechanistic interpretation but should not substitute for hard clinical endpoints. Quantitative wound metrics, including wound area, wound volume, time-to-closure, and recurrence, should be prespecified, because modest perfusion changes without durable healing may not translate into patient-relevant benefit. Where mechanistic endpoints are included, such as TcPO₂, they should be measured at standardized time points and analyzed as supportive secondary outcomes rather than surrogates for AFS. TcPO₂ and other perfusion indices should therefore be treated as physiologic prognostic or mechanistic measures, not as validated predictive biomarkers of differential cell-therapy benefit unless prospectively tested in randomized interaction analyses.

To improve reproducibility, trials should also prespecify wound-care co-interventions, including debridement frequency, offloading strategy, infection surveillance, antibiotic protocols, vascular reassessment triggers, and nutritional or rehabilitation support, and should capture adherence to these measures. Variation in background care can materially influence wound endpoints and may interact with host-state biology. If a stabilization, run-in, or optimization strategy is used, the timing of endpoint follow-up should be defined carefully so that events occurring before randomization or dosing are not obscured, excluded, or misclassified. Pre-randomization deaths, amputations, withdrawals, clinical deterioration, and screen failures should be reported transparently.

For host-state-stratified trials, endpoint analyses should distinguish three related questions: the overall treatment effect in the randomized population; the prognostic association between baseline host-state profile and outcome; and the host-state × treatment interaction, if prespecified. A favorable or unfavorable profile should not be interpreted as proof of treatment sensitivity or resistance unless a randomized treatment-by-profile interaction is demonstrated. Similarly, post hoc subgroup findings, including marker-defined signals from neutral trials, should be treated as hypothesis-generating and used to design prospective validation rather than to define clinical treatment-selection thresholds.

5.3. Product and safety reporting

Standardized reporting of cell-product characteristics is essential for comparison across studies and to separate host-level from product-level drivers of heterogeneity. This applies to both cellular products and emerging cell-free approaches. MSC-derived extracellular vesicles show pro-angiogenic and wound-healing signals in preclinical diabetic wound models and meta-analyses [68,69], and a GMP-manufactured UC-MSC product demonstrated pro-angiogenic activity in vitro [70]. These findings support biological plausibility and translational development but should not be interpreted as evidence of clinical efficacy in no-option or poor-option CLTI without prospective trial validation.

For MSC-based products, adherence to the International Society for Cellular Therapy minimal criteria [71] should be documented, along with transparent reporting of the cell source, expansion conditions, release criteria, dose, schedule, and delivery route. For allogeneic products, donor screening, donor eligibility criteria, banking conditions, lot release procedures, immunologic considerations, and storage and transport conditions should be described. At minimum, release criteria should include sterility and endotoxin testing, viability, identity, and a reproducible potency readout aligned with the proposed mechanism, reported alongside delivered dose and injection mapping. Additional reporting that improves interpretability includes cryopreservation and thaw conditions, passage number where applicable, lot-to-lot variability, and a clearly defined dose metric, including total dose and, when relevant, dose per kilogram.

Open reporting of co-interventions, including wound-care pathways, antibiotic and offloading protocols, analgesia, vascular reassessment, nutritional support, and rehabilitation, is also essential, as variation in background care may obscure treatment effects and may interact with host-state biology. For extracellular vesicle or exosome products, standardized characterization, dose metrics, stability, storage conditions, and, where carriers are used, release kinetics should be reported to enable comparison and reproducibility [72]. Because host-state markers may shift with infection control, stabilization, and optimization bundles, trials should report background-care adherence and timing relative to marker sampling, randomization, dosing, and outcome assessment to contextualize biomarker trajectories.

Safety monitoring should capture systemic events, including major adverse cardiovascular events, infections, thrombotic events, immune-mediated reactions, and death, as well as local limb events, including compartment syndrome, acute ischemia, worsening infection, and progression to major amputation. Independent oversight should be used for first-in-class products, higher-risk host-state strata, allogeneic or engineered products, and studies involving biomaterials or combination approaches. For studies combining revascularization or surgical procedures with cell therapy, procedural timing, perioperative infection control, wound debridement, and interactions with antithrombotic regimens should be clearly described. Where biomaterial carriers are used to improve local retention, their composition, degradation profile, release kinetics, and safety monitoring plan should be specified [72].

Scalable GMP manufacturing platforms are feasible, including industrial-scale expansion workflows for CD34⁺ cell products, supporting the wider translational pathway for standardized regenerative therapies [73]. However, manufacturing feasibility should be distinguished from clinical efficacy, and product-level standardization should not obscure the need for host-state-aware design and reporting. Because no-option or poor-option CLTI carries high competing risks, protocols should prespecify treatment-hold or stopping criteria for sepsis progression, rapidly worsening tissue loss, planned or urgent major amputation, unexpected cardiovascular event imbalance, immune-mediated toxicity, or other safety signals. For trials enrolling higher-risk host-state profiles, predefined safety stopping boundaries and an independent data and safety monitoring board are particularly important to protect participants and to ensure credible interpretation of early signals.

Finally, host-state and product-level factors should be reported together rather than interpreted in isolation. A trial that fails to show benefit may reflect product characteristics, dose, delivery route, host-state distribution, background care, endpoint selection, or competing-risk burden. Conversely, an apparent subgroup signal should not be interpreted as a validated predictive biomarker or product-response signature unless supported by a prespecified randomized treatment-by-host-state or product-by-host-state interaction analysis. This integrated reporting approach may improve reproducibility, clarify sources of heterogeneity, and guide prospective validation of precision regenerative-therapy strategies in no-option or poor-option CLTI.

6. Limitations and validation needs

Several limitations should be acknowledged. This review was structured as a narrative, hypothesis-generating synthesis rather than a systematic review with formal PRISMA methods, a formal risk-of-bias assessment, a GRADE or certainty-of-evidence assessment, or a meta-analysis, and may therefore be subject to incomplete study capture and selection bias. Descriptive source-context terms used in Table 1 and Supplementary Table S4 are intended only to summarize study type, population, and inferential limitations; they should not be interpreted as evidence-certainty ratings. The host-state ranges discussed are derived from heterogeneous observational studies, trial subgroups, treated-cohort analyses, mechanistic studies, and guideline-informed risk conventions. They are approximate prognostic ranges for protocol development and prospective validation, not statistically optimized predictive cut points, validated clinical eligibility criteria, treatment-selection thresholds, or gatekeeping rules. Thresholds are assay-dependent, context-specific, and influenced by timing relative to infection control, debridement, wound-care interventions, renal status, nutritional status, and acute clinical deterioration. Accordingly, future studies should capture host-state variables continuously whenever possible, use prespecified categories only for stratified randomization, covariate adjustment, sensitivity analyses, and prospective validation, and avoid automatic patient exclusion based on isolated marker values.

Prospective validation should model host-state markers both continuously and categorically, report marker distributions by treatment arm, and test whether host-state profiles improve discrimination, calibration, and clinical interpretability beyond WIfI/GLASS staging, objective perfusion testing, vascular imaging, established clinical risk models, and care-pathway effects. Validation studies should prespecify assay platforms, units, lower limits of quantification, sampling windows, repeat-testing rules, and handling of acute infection, recent debridement, steroid exposure, dialysis timing, transfusion, nutritional intervention, and perioperative stress. Because the proposed favorable, intermediate, and unfavorable profiles are provisional and pattern-based, external validation should assess both individual markers and composite profiles across centers, care pathways, cell products, and disease etiologies. Calibration should be repeated in independent datasets and in populations underrepresented in prior trials, including dialysis-dependent patients, frail patients, patients with active or recently controlled infection, patients with CTD-associated or inflammatory nonatherosclerotic CLTI, and patients excluded from conventional revascularization trials.

Evidence linking host-state markers to differential cell-therapy benefit remains limited. The Pan et al. [34] nomogram provides the most directly relevant treated-cohort signal, including associations of fibrinogen and TcPO₂ with 6-month responder/remission status, but it was retrospective, all patients received cell therapy, and no randomized comparator or treatment-by-marker interaction was tested. These findings should therefore be interpreted as treated-cohort associations and candidate stratification variables, not as validated predictive biomarkers. Similarly, post hoc subgroup observations, including the HbA1c <6.5% signal reported in the neutral phase III PACE trial, remain exploratory and require prospective randomized validation before being interpreted as treatment-selection biomarkers [30]. At present, host-state-informed stratification should be viewed as prognostic enrichment, baseline risk balancing, covariate adjustment, and stratified trial design rather than as a proven predictive selection tool. Demonstrating predictive utility will require trials designed to estimate host-state × treatment interactions, with prespecified hypotheses, adequate power where feasible, appropriate control of multiplicity, and transparent reporting of both overall treatment effects and interaction estimates.

Heterogeneity in cell products, manufacturing processes, release criteria, potency assays, dosing, delivery routes, delivery schedules, and concomitant care will continue to influence outcomes independently of host biology and may interact with host-state markers in ways that are not yet defined. Therefore, validation should not indiscriminately pool biologically distinct products. Product source, manufacturing method, potency characteristics, dose, route, injection mapping, and background-care adherence should be reported alongside host-state profiles. Future studies should distinguish host-state × treatment interactions from product-by-host-state interactions, because an apparent marker-defined signal may reflect the biology of a specific product or delivery strategy rather than a generalizable host-state predictor of cell-therapy benefit.

This review focused on pragmatic laboratory, clinical, and physiologic markers that could plausibly be incorporated into multicenter CLTI trial protocols. It did not comprehensively evaluate advanced imaging, omics-based biomarkers, transcriptomic or proteomic signatures, cell-product potency correlates, microbiome data, imaging-derived microcirculatory measures, or in-depth formal frailty instruments, all of which may provide incremental prognostic or mechanistic information. Evidence for some domains, such as Lp(a), CXCL10/IP-10, disease etiology, smoking status, frailty, and CTD-associated or inflammatory nonatherosclerotic CLTI, remains limited and often derives from small cohorts, subgroup analyses, surrogate endpoints, or mechanistic studies. These domains should therefore be treated as exploratory or context-specific unless prospectively validated in appropriately designed clinical cohorts or randomized trials.

No published data demonstrate that improving host-state marker profiles during an optimization phase improves cell-therapy efficacy. Any optimization or run-in strategy should therefore remain an optional, secondary, protocol-level design consideration, not a required component of the framework and not a basis for assuming improved treatment responsiveness. Apparent benefits of optimization may reflect improved infection control, wound care, nutrition, cardiometabolic management, pain control, rehabilitation, or regression-to-the-mean rather than a true increase in regenerative responsiveness. A pre-randomization run-in may also introduce selection, attrition, survivorship, co-intervention, regression-to-the-mean, and immortal-time biases unless all screened patients, run-in failures, deteriorations, amputations, deaths, withdrawals, and reasons for non-randomization are prospectively captured and reported [65,66]. Where the optimization strategy itself is a scientific question, post-randomization or factorial designs may provide a less biased approach, provided that all randomized participants are followed according to intention-to-treat principles.

Finally, “no-option” and “poor-option” CLTI remain variably defined across historical trials, contemporary cohorts, and real-world practice. Patients excluded from large revascularization trials may differ systematically from those included, limiting generalizability and complicating event-rate assumptions [9]. Future studies should use MDT-confirmed, guideline-aligned, consensus-informed definitions of no-option or poor-option CLTI, document the anatomic, biologic, procedural-risk, functional, and contextual reasons for this designation, and avoid inferring no-option status simply because revascularization was not performed. Prospective validation should standardize background care, prespecify the timing of host-state marker assessment, track pre-randomization exclusions and clinical deterioration during any run-in period, report events occurring before randomization or dosing, and externally calibrate approximate prognostic ranges to minimize confounding, selection bias, and immortal-time bias. Until such validation is available, the proposed framework should be interpreted as a hypothesis-generating trial-design resource for prognostic enrichment, stratified randomization, covariate adjustment, and prespecified host-state × treatment interaction testing, not as a validated clinical decision tool.

7. Future directions

Prospective validation will require host-state profile–aware observational cohorts and randomized trials with prespecified analyses of baseline prognosis and treatment-effect heterogeneity by host-state profile, applying standardized outcome measures and blinded adjudication [67]. Stratified randomization, adaptive enrichment designs, or platform trials may be particularly efficient approaches for evaluating host-state × treatment and product-by-host-state interactions while preserving an unbiased estimate of the overall treatment effect. Because current host-state profiles are provisional and hypothesis-generating, future trials should retain continuous marker values whenever possible and use categorical profiles primarily for risk balancing, covariate adjustment, sensitivity analyses, and prospective validation rather than for automatic patient exclusion or clinical treatment selection. Harmonized reporting across diverse health systems, including low- and middle-income countries with high CLTI burden, will be critical for equity and transferability [10]. Only after prospective validation in appropriately designed cohorts and randomized trials could host-state markers be considered as adjuncts to trial counseling, risk communication, and clinical decision-making; even then, they should remain complementary to MDT judgment, WIfI/GLASS assessment, objective perfusion testing, vascular imaging, infection control, wound care, and assessment of revascularization feasibility.

Mechanistic studies may refine biomarker discovery. In a murine CLTI model, proteomic analyses identified 4019 proteins and indicated that transplanted human cells were undetectable by day 21, yet perfusion benefits persisted; cell therapy was associated with downregulation of inflammatory infiltrates and modulation of ferroptosis-related pathways, including GPX4, ACSL4, and TFRC [22]. Translation of such signatures into clinically measurable panels may support mechanistically informed stratification, including the development of parsimonious markers that are feasible for multicenter trials. However, mechanistic signatures should be treated as exploratory until linked to clinically meaningful outcomes and prospectively tested in designs capable of distinguishing prognosis from differential treatment benefit.

Emerging analytical tools may accelerate this translation. Multiomics and multidimensional phenotyping, including proteomics and transcriptomics, may complement pragmatic laboratory panels by identifying latent endotypes and candidate interaction hypotheses. Machine learning could help integrate biomarker, clinical, perfusion, anatomic, product-level, and care-pathway data into calibrated risk models, but it will require transparency, external validation, bias assessment, calibration monitoring, and transportability testing, particularly given the limited representation of trial-ineligible populations. Where applied, models should be evaluated across host-state profiles and across centers with different care pathways. Algorithmic outputs should not be used as clinical gatekeeping rules unless their safety, calibration, fairness, and incremental value beyond established CLTI assessment have been prospectively demonstrated.

Beyond patient stratification, product-level strategies may help overcome hostile ischemic microenvironments. Engineered or gene-modified MSC approaches are designed to enhance angiogenesis and persistence [74]. In parallel, MSC-derived extracellular vesicles improve diabetic wound healing and angiogenesis in preclinical models [68], with meta-analytic signals for augmented wound closure [69], and biomaterial carriers are being developed to improve local delivery and retention [72]. These next-generation strategies remain translational and require product-specific clinical validation. Host-state profiles may provide a rational framework for prospectively testing whether specific products, delivery systems, or potency characteristics interact with high-inflammatory, poor-healing, renal-risk, or infection-prone phenotypes. Such analyses should be prespecified as product-by-host-state interaction tests and interpreted cautiously unless adequately powered and replicated.

Future studies should also clarify how stabilization and optimization strategies are best incorporated into trial design. Clinical stabilization before eligibility confirmation, such as infection control, debridement, pain management, wound care, and reassessment of revascularization feasibility, should remain guideline-directed care rather than a trial intervention. Any pre-randomization run-in or post-randomization optimization strategy should be considered only as an optional, secondary, protocol-level design component for selected clinically stable patients. If a run-in is used, investigators should prospectively report all screened patients, run-in failures, deteriorations, amputations, deaths, withdrawals, and reasons for non-randomization to minimize selection, attrition, survivorship, regression-to-the-mean, co-intervention, and immortal-time biases [65,66]. Where the optimization concept itself is being tested, post-randomization or factorial designs may provide a less biased approach, provided all randomized participants are followed according to intention-to-treat principles.

Finally, future research should prioritize reproducibility and clinical interpretability. Trials should standardize host-state marker assays, define sampling windows relative to infection control and debridement, harmonize endpoints, use contemporary conservative-care or placebo benchmarks for sample-size assumptions, and transparently report cell-product characteristics, delivery procedures, concomitant wound care, and safety events. The key validation question is not whether host-state markers can divide patients into visually intuitive categories, but whether adding these markers improves trial interpretation, risk balancing, endpoint adjudication, and prespecified testing of host-state × treatment interactions beyond existing MDT-based CLTI assessment.

8. Conclusion

This structured narrative review integrates evidence that host-state factors—including inflammation, nutrition, renal function, infection, perfusion status, disease etiology, frailty, smoking status, and selected metabolic, lipid, and coagulation-related markers—are associated with outcomes in CLTI. The biological plausibility of these associations is supported by the predominantly paracrine mechanism of many cell-based therapies and the key role of the regenerative microenvironment. However, most candidate markers are currently supported primarily as prognostic markers. Their value for predicting differential benefit from cell therapy remains unproven and requires prospective randomized host-state × treatment interaction testing. Treated-cohort associations and post hoc subgroup observations should therefore be interpreted as hypothesis-generating candidate variables, not as validated predictive biomarkers or treatment-selection rules.

A conceptual framework emerges that distinguishes provisional favorable, intermediate, and unfavorable prognostic host-state profiles based on readily available laboratory biomarkers, clinical host-state factors, and physiologic/perfusion indices. These profiles are intended as hypothesis-generating trial-design strata for protocol development, prognostic enrichment, stratified randomization, covariate adjustment, sensitivity analyses, and prospective validation. They are not validated clinical eligibility criteria, treatment-selection cutoffs, or biomarker-based gatekeeping thresholds. Continuous host-state variables should be retained for analysis whenever possible, with categorical profiles used to inform trial design rather than to automatically include or exclude patients.

The central purpose of the proposed framework is to support unbiased and interpretable regenerative-therapy trials in MDT-confirmed no-option or poor-option CLTI. For selected clinically stable intermediate-risk patients, optional stabilization, run-in, or post-randomization optimization strategies may be considered to standardize background care and reduce transient host-state misclassification. These strategies remain secondary design features, not required components of the framework and not evidence that improving host-state marker profiles enhances cell-therapy efficacy. Any run-in or optimization approach should include safeguards against selection, attrition, survivorship, co-intervention, regression-to-the-mean, and immortal-time biases, and should not delay urgent infection control, wound management, revascularization reassessment, or other guideline-directed CLTI care.

Patients with persistently unfavorable prognostic profiles may carry high competing risks of amputation, death, infection progression, and poor wound healing. These profiles should not be interpreted as proof of treatment resistance or lack of potential benefit. Rather, trials enrolling such patients may require separate safety and feasibility cohorts, explicit stopping rules, conservative effect-size assumptions, transparent counseling regarding competing risks, and careful reporting of all screen failures, deteriorations, amputations, deaths, and reasons for non-randomization.

Prospective host-state profile–aware trials, standardized assays, harmonized endpoints, blinded outcome adjudication, transparent cell-product and concomitant-care reporting, and prespecified host-state × treatment and product-by-host-state interaction analyses will be essential to determine whether host-state-informed enrichment and stratification improve the interpretability, impact, and transferability of regenerative treatments in no-option or poor-option CLTI. Until such validation is available, the framework should be used as a cautious, hypothesis-generating trial-design resource rather than as a clinical decision algorithm.

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Availability of data and materials

Not applicable.

Authors’ contributions

YS, EHN, NM, KF, and AT conceived and designed the narrative review, including development of the conceptual framework, and led the literature search and evidence synthesis. YS, EHN, NM, TI, HS, KF, MN, AT, and YI contributed to the review design, interpretation of the literature, manuscript drafting, and critical revision. YS, HS, KF, MN, AT, and YI provided specialized input on prognostic marker interpretation, CLTI management, regenerative-therapy trial methodology, and refinement of the conceptual framework. All authors approved the final manuscript and agreed to be accountable for all aspects of the work.

Funding

None.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

None.

Footnotes

Peer review under responsibility of the Japanese Society for Regenerative Medicine.

^{Appendix A}

Supplementary data to this article can be found online at https://doi.org/10.1016/j.reth.2026.101162.

Appendix A. Supplementary data

The following are the Supplementary data to this article.

Multimedia component 1

mmc1.docx^{(33.6KB, docx)}

Multimedia component 2

mmc2.docx^{(32.7KB, docx)}

Multimedia component 3

mmc3.docx^{(30.4KB, docx)}

Multimedia component 4

mmc4.docx^{(37.4KB, docx)}

Multimedia component 5

Supplementary Figure S2. Optional host-state-stratified trial-design concept for investigational cell therapy in MDT-confirmed no-option or poor-option chronic limb-threatening ischemia

This figure depicts one optional, hypothesis-generating trial-design concept for evaluating an investigational cell therapy in MDT-confirmed no-option or poor-option CLTI. It should not be interpreted as a required structure for CLTI cell-therapy trials, a current or approved trial, or a clinical decision algorithm. The design incorporates provisional host-state profiles to support prognostic enrichment, baseline risk balancing, stratified randomization, covariate adjustment, and prespecified host-state × treatment interaction testing. Host-state profiles are distinguished from randomized treatment arms. After screening and any clinical stabilization required before eligibility confirmation, participants undergo baseline host-state assessment, including CRP, NLR, albumin, eGFR/dialysis status, WIfI infection grade, and etiology, with optional adjuncts where available. Stabilization before eligibility confirmation represents standard ethical care rather than a trial intervention. Participants may then be provisionally categorized into favorable, intermediate, or unfavorable host-state strata for protocol-defined trial purposes. Within the favorable and unfavorable strata, randomization may compare cell therapy plus BSC versus placebo/sham plus BSC, with stratum-specific efficacy or safety and feasibility objectives. Within the intermediate stratum, a post-randomization optimization strategy may be tested by comparing optimization plus cell therapy plus BSC versus optimization plus placebo/sham plus BSC. In the illustrative design shown, the favorable stratum uses 12-month AFS as the primary outcome; the intermediate stratum has a protocol-defined endpoint hierarchy in which 12-month AFS is shown as the primary clinical endpoint and host-state profile improvement is treated as an exploratory or secondary endpoint; and the unfavorable stratum uses a 6-month safety and feasibility endpoint as the primary outcome. This endpoint hierarchy is illustrative and should be prespecified in the protocol; host-state profile improvement should not be interpreted as a validated surrogate for cell-therapy efficacy unless prospectively validated. This intermediate-stratum comparison illustrates a post-randomization optimization strategy rather than an unreported pre-randomization survivor-enrichment run-in. Optimization or run-in strategies remain secondary design features and should not supersede unbiased treatment-effect estimation, guideline-directed CLTI care, or transparent reporting of all screened patients and pre-randomization events. Adaptive elements may include prespecified interim analysis, early stopping for harm or futility, sample-size re-estimation, prespecified subgroup analyses, and a final prespecified host-state × treatment interaction assessment. Sample sizes shown are illustrative and require formal power calculations. Host-state profiles are intended for trial design and prospective validation and should not be interpreted as validated clinical eligibility criteria, treatment-selection cutoffs, or gatekeeping thresholds.

Abbreviations: AFS, amputation-free survival; BSC, best supportive care; CLTI, chronic limb-threatening ischemia; CRP, C-reactive protein; eGFR, estimated glomerular filtration rate; MDT, multidisciplinary team; NLR, neutrophil-to-lymphocyte ratio; WIfI, Wound-Ischemia-foot Infection classification.

mmc5.zip^{(3.3MB, zip)}

Fig. S1 — Supplementary Figure S1. Qualitative forest-plot overview of host-state marker associations with adverse limb outcomes in chronic limb-threatening ischemia

This figure provides a qualitative forest-plot overview of representative, non-pooled associations between host-state markers and adverse limb outcomes in CLTI, including major amputation, mortality, and wound-healing failure, based on observational studies, trial cohorts, and treated-cohort analyses reviewed in this article. Markers are grouped by domain. Point estimates, including HRs or ORs with 95% confidence intervals, are plotted on a logarithmic scale; the vertical line indicates the null. Estimates to the right of the null indicate increased risk of adverse outcomes, whereas estimates to the left indicate reduced risk. This display is intended as a qualitative visual summary of prognostic gradients, not as a meta-analysis or quantitative evidence synthesis. Formal pooling was not performed because of heterogeneity in study design, populations, assay methods, outcome definitions, effect measures, and follow-up. HRs and ORs should therefore be interpreted within their source-study contexts rather than compared quantitatively across rows. The lower panel highlights associations within treated cohorts and is hypothesis-generating; these associations do not establish validated predictive biomarkers of treatment effect or differential cell-therapy benefit. Lower-panel ORs are plotted in the original treated-cohort outcome direction: fibrinogen OR <1 denotes lower responder/remission odds, whereas TcPO₂ OR >1 denotes higher responder/remission odds. The Pan et al. [34] fibrinogen/TcPO₂ findings should be interpreted as hypothesis-generating candidate variables for prospective host-state × treatment interaction testing, not as validated treatment-selection biomarkers. Supplementary Table S2 provides source-level and estimate-basis mapping for each plotted estimate.

Abbreviations: AFS, amputation-free survival; CI, confidence interval; CLTI, chronic limb-threatening ischemia; CRP, C-reactive protein; eGFR, estimated glomerular filtration rate; HR, hazard ratio; IL-6, interleukin-6; NLR, neutrophil-to-lymphocyte ratio; OR, odds ratio; TAO, thromboangiitis obliterans; TcPO₂transcutaneous oxygen pressure; WIfI, Wound-Ischemia-foot Infection classification.

References

1.Conte M.S., Bradbury A.W., Kolh P., et al. Global vascular guidelines on the management of chronic limb-threatening ischemia. Eur J Vasc Endovasc Surg. 2019 Jul;58(1S):S1–S109.e33. doi: 10.1016/j.ejvs.2019.05.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
2.Song P., Rudan D., Zhu Y., et al. Global, regional, and national prevalence and risk factors for peripheral artery disease in 2015: an updated systematic review and analysis. Lancet Glob Health. 2019 Aug;7(8):e1020–e1030. doi: 10.1016/S2214-109X(19)30255-4. [DOI] [PubMed] [Google Scholar]
3.van Reijen N.S., Hensing T., Santema T.K.B., et al. Outcomes of conservative treatment in patients with chronic limb threatening ischemia: a systematic review and meta-analysis. Eur J Vasc Endovasc Surg. 2021 Aug;62(2):214–224. doi: 10.1016/j.ejvs.2021.01.005. [DOI] [PubMed] [Google Scholar]
4.Verwer M.C., Wijnand J.G.J., Teraa M., et al. Long-term survival and limb salvage in patients with non-revascularisable chronic limb threatening ischemia. Eur J Vasc Endovasc Surg. 2021 Aug;62(2):225–232. doi: 10.1016/j.ejvs.2021.04.003. [DOI] [PubMed] [Google Scholar]
5.Gornik H.L., Aronow H.D., Goodney P.P., et al. 2024 ACC/AHA/AACVPR/APMA/ABC/SCAI/SVM/SVN/SVS/SIR/VESS guideline for the management of lower extremity peripheral artery disease: a report of the American college of Cardiology/American Heart Association Joint Committee on clinical practice guidelines. Circulation. 2024;149(24):e1313–e1410. doi: 10.1161/CIR.0000000000001251. [DOI] [PMC free article] [PubMed] [Google Scholar]
6.Mazzolai L., Teixido-Tura G., Lanzi S., et al. ESC Scientific Document Group 2024 ESC guidelines for the management of peripheral arterial and aortic diseases. Eur Heart J. 2024 Sep 29;45(36):3538–3700. doi: 10.1093/eurheartj/ehae179. [DOI] [PubMed] [Google Scholar]
7.Fitridge R., Chuter V.H., Mills J.L., et al. The intersocietal IWGDF, ESVS, SVS guidelines on peripheral artery disease in people with diabetes mellitus and a foot ulcer. J Vasc Surg. 2023 Nov;78(5):1101–1131. doi: 10.1016/j.jvs.2023.07.020. [DOI] [PubMed] [Google Scholar]
8.Fabiani M.A., van den Berg J.C., de la Torre O.A., Verastegui A., et al. Development of a standard definition of 'no-option' and 'poor-option' for revascularization in chronic limb-threatening ischemia. Br J Surg. 2026;113 doi: 10.1093/bjs/znag040. znag040. [DOI] [PubMed] [Google Scholar]
9.Schwarz S.M., Warren A.S., Hemingway J.F., et al. Patients excluded from BEST-CLI were more likely to undergo a major amputation at a single urban safety net hospital. J Vasc Surg. 2025 Oct;82(4):1359–1365. doi: 10.1016/j.jvs.2025.04.019. [DOI] [PubMed] [Google Scholar]
10.Daher G., Upadhyay S., Li J. Chronic limb-threatening ischemia: a comprehensive review paper. Interv Cardiol Clin. 2025 Apr;14(2):257–272. doi: 10.1016/j.iccl.2024.11.011. [DOI] [PubMed] [Google Scholar]
11.Athavale A., Fukaya E., Leeper N.J. Peripheral artery disease: molecular mechanisms and novel therapies. Arterioscler Thromb Vasc Biol. 2024 Jun;44(6):1165–1170. doi: 10.1161/ATVBAHA.124.320195. [DOI] [PMC free article] [PubMed] [Google Scholar]
12.Huerta C.T., Voza F.A., Ortiz Y.Y., et al. Mesenchymal stem cell-based therapy for non-healing wounds due to chronic limb-threatening ischemia: a review of preclinical and clinical studies. Front Cardiovasc Med. 2023 Feb 1;10 doi: 10.3389/fcvm.2023.1113982. [DOI] [PMC free article] [PubMed] [Google Scholar]
13.McGinigle K.L., Freeman N.L.B., Marston W.A., et al. Precision medicine enables more TNM-like staging in patients with chronic limb-threatening ischemia. Front Cardiovasc Med. 2021 Jul 16;8 doi: 10.3389/fcvm.2021.709904. [DOI] [PMC free article] [PubMed] [Google Scholar]
14.Qin Q., Haba D., Nakagami G. Which biomarkers predict hard-to-heal diabetic foot ulcers? A scoping review. Drug Discov Ther. 2024 Jan 12;17(6):368–377. doi: 10.5582/ddt.2023.01086. [DOI] [PubMed] [Google Scholar]
15.Popescu A.I., Rata A.L., Barac S., et al. Narrative review of biological markers in chronic limb-threatening ischemia. Biomedicines. 2024 Apr 3;12(4):798. doi: 10.3390/biomedicines12040798. [DOI] [PMC free article] [PubMed] [Google Scholar]
16.Liu Y., Lyons C.J., Keane A., et al. TNF-α pathway activation and altered microRNA expression in peripheral blood-derived endothelial colony-forming cells: insights into diabetes-induced dysfunction. Stem Cell Rev Rep. 2025 Oct;21(7):2272–2286. doi: 10.1007/s12015-025-10908-6. [DOI] [PubMed] [Google Scholar]
17.Buso G., Faggin E., Rosenblatt-Velin N., et al. The role of neutrophils in lower limb peripheral artery disease: state of the art and future perspectives. Int J Mol Sci. 2023 Jan 6;24(2):1169. doi: 10.3390/ijms24021169. [DOI] [PMC free article] [PubMed] [Google Scholar]
18.Chung J., Modrall J.G., Valentine R.J. The need for improved risk stratification in chronic critical limb ischemia. J Vasc Surg. 2014 Dec;60(6):1677–1685. doi: 10.1016/j.jvs.2014.07.104. [DOI] [PubMed] [Google Scholar]
19.Song P., Xiong H., Quan X., et al. Predictive factors and models for major adverse cardiovascular and limb events in patients with peripheral arterial disease. Vasa. 2025 Sep 2 doi: 10.1024/0301-1526/a001228. [DOI] [PubMed] [Google Scholar]
20.van Haelst S.T.W., Teraa M., Moll F.L., et al. Prognostic value of the society for vascular surgery wound, ischemia, and foot infection (WIfI) classification in patients with no-option chronic limb-threatening ischemia. J Vasc Surg. 2018 Oct;68(4):1104–1113.e1. doi: 10.1016/j.jvs.2018.02.028. [DOI] [PubMed] [Google Scholar]
21.Shimizu Y., Ntege E.H., Inoue Y., et al. Optimizing mesenchymal stem cell extracellular vesicles for chronic wound healing: bioengineering, standardization, and safety. Regen Ther. 2024 Jun 15;26:260–274. doi: 10.1016/j.reth.2024.06.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
22.Rojas-Torres M., Beltran-Camacho L., Martinez-Val A., et al. Unraveling the differential mechanisms of revascularization promoted by MSCs & ECFCs from adipose tissue or umbilical cord in a murine model of critical limb-threatening ischemia. J Biomed Sci. 2024 Jul 15;31(1):71. doi: 10.1186/s12929-024-01059-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
23.Tateishi-Yuyama E., Matsubara H., Murohara T., et al. Therapeutic Angiogenesis using cell transplantation (TACT) study investigators. Therapeutic angiogenesis for patients with limb ischemia by autologous transplantation of bone-marrow cells: a pilot study and a randomized controlled trial. Lancet. 2002 Aug 10;360(9331):427–435. doi: 10.1016/S0140-6736(02)09670-8. [DOI] [PubMed] [Google Scholar]
24.Walter D.H., Krankenberg H., Balzer J.O., et al. Provasa Investigators. Intra-arterial administration of bone marrow mononuclear cells in patients with critical limb ischemia: a randomized-start, placebo-controlled pilot trial (PROVASA) Circ Cardiovasc Interv. 2011 Feb 1;4(1):26–37. doi: 10.1161/CIRCINTERVENTIONS.110.958348. [DOI] [PubMed] [Google Scholar]
25.Powell R.J., Marston W.A., Berceli S.A., et al. Cellular therapy with Ixmyelocel-T to treat critical limb ischemia: the randomized, double-blind, placebo-controlled RESTORE-CLI trial. Mol Ther. 2012 Jun;20(6):1280–1286. doi: 10.1038/mt.2012.52. [DOI] [PMC free article] [PubMed] [Google Scholar]
26.Losordo D.W., Kibbe M.R., Mendelsohn F., Marston W., Driver V., Sharafuddin M., et al. A randomized, controlled pilot study of autologous CD34+ cell therapy for critical limb ischemia. Circ Cardiovasc Interv. 2012;5(6):821–830. doi: 10.1161/CIRCINTERVENTIONS.112.968321. [DOI] [PMC free article] [PubMed] [Google Scholar]
27.Teraa M., Sprengers R.W., Schutgens R.E., et al. Effect of repetitive intra-arterial infusion of bone marrow mononuclear cells in patients with no-option limb ischemia: the randomized, double-blind, placebo-controlled rejuvenating endothelial progenitor cells via transcutaneous intra-arterial supplementation (JUVENTAS) trial. Circulation. 2015 Mar 10;131(10):851–860. doi: 10.1161/CIRCULATIONAHA.114.012913. [DOI] [PubMed] [Google Scholar]
28.Pignon B., Sevestre M.A., Kanagaratnam L., Pernod G., Stephan D., Emmerich J., et al. Autologous bone marrow mononuclear cell implantation and its impact on the outcome of patients with critical limb ischemia: results of a randomized, double-blind, placebo-controlled trial. Circ J. 2017;81(11):1713–1720. doi: 10.1253/circj.CJ-17-0045. [DOI] [PubMed] [Google Scholar]
29.Horie T., Yamazaki S., Hanada S., et al. Outcome from a randomized controlled clinical trial: improvement of peripheral arterial disease by Granulocyte colony-stimulating factor-mobilized autologous peripheral-blood-mononuclear cell transplantation (IMPACT) Circ J. 2018;82(8):2165–2174. doi: 10.1253/circj.CJ-17-1220. [DOI] [PubMed] [Google Scholar]
30.Norgren L., Weiss N., Nikol S., et al. PACE: randomized, controlled, multicentre, multinational, phase III study of PLX-PAD for critical limb ischemia in patients unsuitable for revascularization: randomized clinical trial. Br J Surg. 2024 Jan 31;111(2) doi: 10.1093/bjs/znad437. znad437. [DOI] [PMC free article] [PubMed] [Google Scholar]
31.Chacón Quevedo A., García-Revillo García J., Sagrario Lombardo Galera M., et al. Autologous bone marrow mononuclear cells in the treatment of chronic limb-threatening ischemia: a proof-of-concept trial. Cardiovasc Res. 2022;118(Suppl 1):cvac066–cvac236. doi: 10.1093/cvr/cvac066.236. [DOI] [Google Scholar]
32.Kwiatkowski T., Zbierska-Rubinkiewicz K., Krzywon J.W., et al. Cellular therapies in no-option critical limb ischemia: present status and future directions. Postepy Kardiol Interwencyjnej. 2022 Dec;18(4):340–349. doi: 10.5114/aic.2022.120962. [DOI] [PMC free article] [PubMed] [Google Scholar]
33.Liang G., Zhang Y., Chen C., et al. Clinical trial landscape of stem cell therapy for peripheral arterial disease. Stem Cell Rev Rep. 2026;22:1042–1044. doi: 10.1007/s12015-025-11026-z. [DOI] [PubMed] [Google Scholar]
34.Pan T., Liu H., Fang Y., et al. Predictors of responders to mononuclear stem cell-based therapeutic angiogenesis for no-option critical limb ischemia. Stem Cell Res Ther. 2019 Jan 11;10(1):15. doi: 10.1186/s13287-018-1117-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
35.FDA-NIH Biomarker Working Group . Silver Spring (MD): Food and Drug Administration (US) National Institutes of Health (US); Bethesda (MD): 2016. BEST (Biomarkers, EndpointS, and other tools) resource.https://www.ncbi.nlm.nih.gov/books/NBK326791/ Available from: [PubMed] [Google Scholar]
36.US Food and Drug Administration Enrichment strategies for clinical trials to support determination of effectiveness of human drugs and biological products: guidance for industry. Silver Spring (MD): US Food Drug Administration. 2019 [Google Scholar]
37.Panc C., Guler A., Gurbak I., et al. Association between CRP/Albumin ratio and long-term mortality in patients with chronic limb-threatening ischemia undergoing endovascular therapy below the knee: the ACHILES-BTK registry. Ann Vasc Surg. 2022 May;82:172–180. doi: 10.1016/j.avsg.2021.11.008. [DOI] [PubMed] [Google Scholar]
38.Kurniawan R.B., Siahaan P.P., Saputra P.B., et al. Neutrophil-to-lymphocyte ratio as a prognostic biomarker in patients with peripheral artery disease: a systematic review and meta-analysis. Vasc Med. 2024 Dec;29(6):687–699. doi: 10.1177/1358863X241281699. [DOI] [PMC free article] [PubMed] [Google Scholar]
39.Gary T., Pichler M., Belaj K., et al. Neutrophil-to-lymphocyte ratio and its association with critical limb ischemia in patients with PAOD. PLoS One. 2013;8(2) doi: 10.1371/journal.pone.0056745. [DOI] [PMC free article] [PubMed] [Google Scholar]
40.Su M.I., Liu C.W. Neutrophil-to-lymphocyte ratio associated with an increased risk of mortality in patients with critical limb ischemia. PLoS One. 2021 May 27;16(5) doi: 10.1371/journal.pone.0252030. [DOI] [PMC free article] [PubMed] [Google Scholar]
41.Levin M.G., Klarin D., Georgakis M.K., et al. A missense variant in the IL-6 receptor and protection from peripheral artery disease. Circ Res. 2021 Oct 29;129(10):968–970. doi: 10.1161/CIRCRESAHA.121.319589. [DOI] [PMC free article] [PubMed] [Google Scholar]
42.Rockwood K., Song X., MacKnight C., et al. A global clinical measure of fitness and frailty in elderly people. CMAJ (Can Med Assoc J) 2005 Aug 30;173(5):489–495. doi: 10.1503/cmaj.050051. [DOI] [PMC free article] [PubMed] [Google Scholar]
43.Hata Y., Iida O., Takahara M., et al. Non-home discharge after endovascular therapy for chronic limb-threatening ischemia - insights from a Japanese nationwide registry. Circ J. 2026;90(5):559–561. doi: 10.1253/circj.CJ-25-0698. [DOI] [PubMed] [Google Scholar]
44.Ju M., Kim Y., Seo K.W. Role of nutrition in wound healing and nutritional recommendations for promotion of wound healing: a narrative review. Ann Clin Nutr Metab. 2023 Dec 1;15(3):67–71. doi: 10.15747/ACNM.2023.15.3.67. [DOI] [Google Scholar]
45.World Health Organization . World Health Organization; Geneva: 2024. Guideline on haemoglobin cutoffs to define anaemia in individuals and populations.https://www.who.int/publications/i/item/9789240088542 Available from: [PubMed] [Google Scholar]
46.O'Hare A.M., Feinglass J., Sidawy A.N., et al. Impact of renal insufficiency on short-term morbidity and mortality after lower extremity revascularization: data from the department of veterans affairs' National Surgical Quality Improvement Program. J Am Soc Nephrol. 2003 May;14(5):1287–1295. doi: 10.1097/01.asn.0000061776.60146.02. [DOI] [PubMed] [Google Scholar]
47.Kidney Disease: Improving Global Outcomes (KDIGO) CKD Work Group KDIGO 2024 Clinical practice guideline for the evaluation and management of chronic kidney disease. Kidney Int. 2024 Apr;105(4S):S117–S314. doi: 10.1016/j.kint.2023.10.018. [DOI] [PubMed] [Google Scholar]
48.Lanzer P., Boehm M., Sorribas V., Thiriet M., Janzen J., Zeller T., St Hilaire C., Shanahan C. Medial vascular calcification revisited: review and perspectives. Eur Heart J. 2014 Jun 14;35(23):1515–1525. doi: 10.1093/eurheartj/ehu163. [DOI] [PMC free article] [PubMed] [Google Scholar]
49.Stenvinkel P. Inflammation in end-stage renal disease: the hidden enemy. Nephrology. 2006 Feb;11(1):36–41. doi: 10.1111/j.1440-1797.2006.00541.x. [DOI] [PubMed] [Google Scholar]
50.Mills JL Sr, Conte M.S., Armstrong D.G., et al. Society for Vascular Surgery Lower Extremity Guidelines Committee The society for vascular surgery lower extremity threatened limb classification system: risk stratification based on wound, ischemia, and foot infection (WIfI) J Vasc Surg. 2014 Jan;59(1):220–234. doi: 10.1016/j.jvs.2013.08.003. e1-2. [DOI] [PubMed] [Google Scholar]
51.Game F.L., Hinchliffe R.J., Apelqvist J., et al. A systematic review of interventions to enhance the healing of chronic ulcers of the foot in diabetes. Diabetes Metab Res Rev. 2012 Feb;28(Suppl 1):119–141. doi: 10.1002/dmrr.2246. [DOI] [PubMed] [Google Scholar]
52.Siracuse J.J., Farber A., Menard M.T., et al. Advanced wound, ischemia, and foot infection stage is associated with poor outcomes in the BEST-CLI trial. J Vasc Surg. 2025 Mar;81(3):720–729.e1. doi: 10.1016/j.jvs.2024.11.027. [DOI] [PubMed] [Google Scholar]
53.Tsimikas S. A test in context: Lipoprotein(a): diagnosis, prognosis, controversies, and emerging therapies. J Am Coll Cardiol. 2017 Feb 14;69(6):692–711. doi: 10.1016/j.jacc.2016.11.042. [DOI] [PubMed] [Google Scholar]
54.Tomoi Y., Takahara M., Soga Y., et al. Impact of high Lipoprotein(a) levels on delayed wound healing in patients with chronic limb-threatening ischemia after peripheral endovascular therapy. Circ J. 2025 Apr 25;89(5):550–556. doi: 10.1253/circj.CJ-24-0383. [DOI] [PubMed] [Google Scholar]
55.Kronenberg F., Mora S., Stroes E.S.G., et al. Lipoprotein(a) in atherosclerotic cardiovascular disease and aortic stenosis: a European Atherosclerosis Society consensus statement. Eur Heart J. 2022 Oct 14;43(39):3925–3946. doi: 10.1093/eurheartj/ehac361. [DOI] [PMC free article] [PubMed] [Google Scholar]
56.Bodnar R.J., Yates C.C., Wells A. IP-10 blocks vascular endothelial growth factor-induced endothelial cell motility and tube formation by inhibiting calpain. Circ Res. 2006 Mar 17;98(5):617–625. doi: 10.1161/01.RES.0000209968.66606.10. [DOI] [PMC free article] [PubMed] [Google Scholar]
57.Idei N., Soga J., Hata T., et al. Autologous bone-marrow mononuclear cell implantation reduces long-term major amputation risk in patients with critical limb ischemia: a comparison of atherosclerotic peripheral arterial disease and Buerger disease. Circ Cardiovasc Interv. 2011 Feb 1;4(1):15–25. doi: 10.1161/CIRCINTERVENTIONS.110.955724. [DOI] [PubMed] [Google Scholar]
58.Iida O., Takahara M., Kodama A., et al. Clinical outcomes of chronic limb-threatening ischemia due to inflammatory nonatherosclerotic versus atherosclerotic etiologies undergoing revascularization. Vasc Med. 2023 Aug;28(4):299–307. doi: 10.1177/1358863X231164659. [DOI] [PubMed] [Google Scholar]
59.Obara H., Matsubara K., Fujimura N., et al. Preliminary report of endovascular treatment for critical limb ischemia patients with connective tissue disease: case series and review of the literature. Int J Angiol. 2015 Jun;24(2):137–142. doi: 10.1055/s-0035-1547516. [DOI] [PMC free article] [PubMed] [Google Scholar]
60.Ouchi S., Takehisa K., Murata N., et al. Long-term outcomes of endovascular therapy for chronic limb-threatening ischemia patients with connective tissue disease. Cardiovasc Interv Ther. 2025 Oct;40(4):860–867. doi: 10.1007/s12928-025-01157-y. [DOI] [PubMed] [Google Scholar]
61.Kondo K., Yanishi K., Hayashida R., et al. Long-term clinical outcomes survey of bone marrow-derived cell therapy in critical limb ischemia in Japan. Circ J. 2018 Mar 23;82(4):1168–1178. doi: 10.1253/circj.CJ-17-0510. [DOI] [PubMed] [Google Scholar]
62.Shoji K., Yanishi K., Yoshimi R., et al. Impact of therapeutic angiogenesis using autologous bone marrow-derived mononuclear cells implantation in critical limb ischemia with scleroderma - subanalysis of the long-term clinical outcomes survey. Circ J. 2019 Feb 25;83(3):662–671. doi: 10.1253/circj.CJ-18-1044. [DOI] [PubMed] [Google Scholar]
63.Shirbaghaee Z., Heidari Keshel S., et al. Report of a phase 1 clinical trial for safety assessment of human placental mesenchymal stem cells therapy in patients with critical limb ischemia (CLI) Stem Cell Res Ther. 2023 Jul 5;14(1):174. doi: 10.1186/s13287-023-03390-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
64.Houghton J.S.M., Meffen A., Gray L.J., et al. Leg ischemia management Collaboration. Streamlined clinical management pathways may reduce major amputations in patients with chronic limb threatening ischemia: a prospective cohort study with historical controls. Eur J Vasc Endovasc Surg. 2025 Mar;69(3):465–473. doi: 10.1016/j.ejvs.2024.09.005. [DOI] [PubMed] [Google Scholar]
65.Laursen D.R.T., Paludan-Muller A.S., Hrobjartsson A. Randomized clinical trials with run-in periods: frequency, characteristics and reporting. Clin Epidemiol. 2019;11:169–184. doi: 10.2147/CLEP.S188752. [DOI] [PMC free article] [PubMed] [Google Scholar]
66.Suissa S. Immortal time bias in pharmacoepidemiology. Am J Epidemiol. 2008 Feb 15;167(4):492–499. doi: 10.1093/aje/kwm324. [DOI] [PubMed] [Google Scholar]
67.Lakhter V., Weinberg M.D., Galmer A., et al. Objective outcome measures for trials in patients with chronic limb-threatening ischemia across 2 decades: analysis and recommendations. JACC Cardiovasc Interv. 2021 Dec 13;14(23):2584–2597. doi: 10.1016/j.jcin.2021.08.079. [DOI] [PubMed] [Google Scholar]
68.Wei Q., Wang Y., Ma K., et al. Extracellular vesicles from human umbilical cord Mesenchymal stem cells facilitate diabetic wound healing through MiR-17-5p-mediated enhancement of angiogenesis. Stem Cell Rev Rep. 2022 Mar;18(3):1025–1040. doi: 10.1007/s12015-021-10176-0. [DOI] [PubMed] [Google Scholar]
69.Bailey A.J.M., Li H., Kirkham A.M., et al. MSC-derived extracellular vesicles to heal diabetic wounds: a systematic review and meta-analysis of preclinical animal studies. Stem Cell Rev Rep. 2022 Mar;18(3):968–979. doi: 10.1007/s12015-021-10164-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
70.Carcopino C., Rossi E., Mebarki M., et al. Understanding the angiogenic characteristics of clinical-grade mesenchymal stromal cells isolated from human umbilical cord. Stem Cell Rev Rep. 2024 Jul;20(5):1353–1356. doi: 10.1007/s12015-024-10712-8. [DOI] [PubMed] [Google Scholar]
71.Dominici M., Le Blanc K., Mueller I., et al. Minimal criteria for defining multipotent mesenchymal stromal cells. The international society for cellular therapy position statement. Cytotherapy. 2006;8(4):315–317. doi: 10.1080/14653240600855905. [DOI] [PubMed] [Google Scholar]
72.Shimizu Y., Inoue Y., Matsuura N., et al. Mesenchymal stromal cell-derived extracellular vesicles in regenerative medicine: standardization, bioengineering, and clinical translation. Regen Ther. 2026;31 doi: 10.1016/j.reth.2025.101058. [DOI] [PMC free article] [PubMed] [Google Scholar]
73.Henon P., Kowalczyk M., Aries A., et al. Industrialized GMP production of CD34+ cells (ProtheraCytes) at clinical scale for treatment of ischemic cardiac diseases is feasible and safe. Stem Cell Rev Rep. 2022 Jun;18(5):1614–1626. doi: 10.1007/s12015-022-10373-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
74.Hegde M., Singh A.K., Kannan S., et al. Therapeutic applications of engineered mesenchymal stromal cells for enhanced angiogenesis in cardiac and cerebral ischemia. Stem Cell Rev Rep. 2024 Nov;20(8):2138–2154. doi: 10.1007/s12015-024-10787-3. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Multimedia component 1

mmc1.docx^{(33.6KB, docx)}

Multimedia component 2

mmc2.docx^{(32.7KB, docx)}

Multimedia component 3

mmc3.docx^{(30.4KB, docx)}

Multimedia component 4

mmc4.docx^{(37.4KB, docx)}

Multimedia component 5

Supplementary Figure S2. Optional host-state-stratified trial-design concept for investigational cell therapy in MDT-confirmed no-option or poor-option chronic limb-threatening ischemia

mmc5.zip^{(3.3MB, zip)}

Data Availability Statement

Not applicable.

[bib1] 1.Conte M.S., Bradbury A.W., Kolh P., et al. Global vascular guidelines on the management of chronic limb-threatening ischemia. Eur J Vasc Endovasc Surg. 2019 Jul;58(1S):S1–S109.e33. doi: 10.1016/j.ejvs.2019.05.006. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib2] 2.Song P., Rudan D., Zhu Y., et al. Global, regional, and national prevalence and risk factors for peripheral artery disease in 2015: an updated systematic review and analysis. Lancet Glob Health. 2019 Aug;7(8):e1020–e1030. doi: 10.1016/S2214-109X(19)30255-4. [DOI] [PubMed] [Google Scholar]

[bib3] 3.van Reijen N.S., Hensing T., Santema T.K.B., et al. Outcomes of conservative treatment in patients with chronic limb threatening ischemia: a systematic review and meta-analysis. Eur J Vasc Endovasc Surg. 2021 Aug;62(2):214–224. doi: 10.1016/j.ejvs.2021.01.005. [DOI] [PubMed] [Google Scholar]

[bib4] 4.Verwer M.C., Wijnand J.G.J., Teraa M., et al. Long-term survival and limb salvage in patients with non-revascularisable chronic limb threatening ischemia. Eur J Vasc Endovasc Surg. 2021 Aug;62(2):225–232. doi: 10.1016/j.ejvs.2021.04.003. [DOI] [PubMed] [Google Scholar]

[bib5] 5.Gornik H.L., Aronow H.D., Goodney P.P., et al. 2024 ACC/AHA/AACVPR/APMA/ABC/SCAI/SVM/SVN/SVS/SIR/VESS guideline for the management of lower extremity peripheral artery disease: a report of the American college of Cardiology/American Heart Association Joint Committee on clinical practice guidelines. Circulation. 2024;149(24):e1313–e1410. doi: 10.1161/CIR.0000000000001251. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib6] 6.Mazzolai L., Teixido-Tura G., Lanzi S., et al. ESC Scientific Document Group 2024 ESC guidelines for the management of peripheral arterial and aortic diseases. Eur Heart J. 2024 Sep 29;45(36):3538–3700. doi: 10.1093/eurheartj/ehae179. [DOI] [PubMed] [Google Scholar]

[bib7] 7.Fitridge R., Chuter V.H., Mills J.L., et al. The intersocietal IWGDF, ESVS, SVS guidelines on peripheral artery disease in people with diabetes mellitus and a foot ulcer. J Vasc Surg. 2023 Nov;78(5):1101–1131. doi: 10.1016/j.jvs.2023.07.020. [DOI] [PubMed] [Google Scholar]

[bib8] 8.Fabiani M.A., van den Berg J.C., de la Torre O.A., Verastegui A., et al. Development of a standard definition of 'no-option' and 'poor-option' for revascularization in chronic limb-threatening ischemia. Br J Surg. 2026;113 doi: 10.1093/bjs/znag040. znag040. [DOI] [PubMed] [Google Scholar]

[bib9] 9.Schwarz S.M., Warren A.S., Hemingway J.F., et al. Patients excluded from BEST-CLI were more likely to undergo a major amputation at a single urban safety net hospital. J Vasc Surg. 2025 Oct;82(4):1359–1365. doi: 10.1016/j.jvs.2025.04.019. [DOI] [PubMed] [Google Scholar]

[bib10] 10.Daher G., Upadhyay S., Li J. Chronic limb-threatening ischemia: a comprehensive review paper. Interv Cardiol Clin. 2025 Apr;14(2):257–272. doi: 10.1016/j.iccl.2024.11.011. [DOI] [PubMed] [Google Scholar]

[bib11] 11.Athavale A., Fukaya E., Leeper N.J. Peripheral artery disease: molecular mechanisms and novel therapies. Arterioscler Thromb Vasc Biol. 2024 Jun;44(6):1165–1170. doi: 10.1161/ATVBAHA.124.320195. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib12] 12.Huerta C.T., Voza F.A., Ortiz Y.Y., et al. Mesenchymal stem cell-based therapy for non-healing wounds due to chronic limb-threatening ischemia: a review of preclinical and clinical studies. Front Cardiovasc Med. 2023 Feb 1;10 doi: 10.3389/fcvm.2023.1113982. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib13] 13.McGinigle K.L., Freeman N.L.B., Marston W.A., et al. Precision medicine enables more TNM-like staging in patients with chronic limb-threatening ischemia. Front Cardiovasc Med. 2021 Jul 16;8 doi: 10.3389/fcvm.2021.709904. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib14] 14.Qin Q., Haba D., Nakagami G. Which biomarkers predict hard-to-heal diabetic foot ulcers? A scoping review. Drug Discov Ther. 2024 Jan 12;17(6):368–377. doi: 10.5582/ddt.2023.01086. [DOI] [PubMed] [Google Scholar]

[bib15] 15.Popescu A.I., Rata A.L., Barac S., et al. Narrative review of biological markers in chronic limb-threatening ischemia. Biomedicines. 2024 Apr 3;12(4):798. doi: 10.3390/biomedicines12040798. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib16] 16.Liu Y., Lyons C.J., Keane A., et al. TNF-α pathway activation and altered microRNA expression in peripheral blood-derived endothelial colony-forming cells: insights into diabetes-induced dysfunction. Stem Cell Rev Rep. 2025 Oct;21(7):2272–2286. doi: 10.1007/s12015-025-10908-6. [DOI] [PubMed] [Google Scholar]

[bib17] 17.Buso G., Faggin E., Rosenblatt-Velin N., et al. The role of neutrophils in lower limb peripheral artery disease: state of the art and future perspectives. Int J Mol Sci. 2023 Jan 6;24(2):1169. doi: 10.3390/ijms24021169. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib18] 18.Chung J., Modrall J.G., Valentine R.J. The need for improved risk stratification in chronic critical limb ischemia. J Vasc Surg. 2014 Dec;60(6):1677–1685. doi: 10.1016/j.jvs.2014.07.104. [DOI] [PubMed] [Google Scholar]

[bib19] 19.Song P., Xiong H., Quan X., et al. Predictive factors and models for major adverse cardiovascular and limb events in patients with peripheral arterial disease. Vasa. 2025 Sep 2 doi: 10.1024/0301-1526/a001228. [DOI] [PubMed] [Google Scholar]

[bib20] 20.van Haelst S.T.W., Teraa M., Moll F.L., et al. Prognostic value of the society for vascular surgery wound, ischemia, and foot infection (WIfI) classification in patients with no-option chronic limb-threatening ischemia. J Vasc Surg. 2018 Oct;68(4):1104–1113.e1. doi: 10.1016/j.jvs.2018.02.028. [DOI] [PubMed] [Google Scholar]

[bib21] 21.Shimizu Y., Ntege E.H., Inoue Y., et al. Optimizing mesenchymal stem cell extracellular vesicles for chronic wound healing: bioengineering, standardization, and safety. Regen Ther. 2024 Jun 15;26:260–274. doi: 10.1016/j.reth.2024.06.001. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib22] 22.Rojas-Torres M., Beltran-Camacho L., Martinez-Val A., et al. Unraveling the differential mechanisms of revascularization promoted by MSCs & ECFCs from adipose tissue or umbilical cord in a murine model of critical limb-threatening ischemia. J Biomed Sci. 2024 Jul 15;31(1):71. doi: 10.1186/s12929-024-01059-w. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib23] 23.Tateishi-Yuyama E., Matsubara H., Murohara T., et al. Therapeutic Angiogenesis using cell transplantation (TACT) study investigators. Therapeutic angiogenesis for patients with limb ischemia by autologous transplantation of bone-marrow cells: a pilot study and a randomized controlled trial. Lancet. 2002 Aug 10;360(9331):427–435. doi: 10.1016/S0140-6736(02)09670-8. [DOI] [PubMed] [Google Scholar]

[bib24] 24.Walter D.H., Krankenberg H., Balzer J.O., et al. Provasa Investigators. Intra-arterial administration of bone marrow mononuclear cells in patients with critical limb ischemia: a randomized-start, placebo-controlled pilot trial (PROVASA) Circ Cardiovasc Interv. 2011 Feb 1;4(1):26–37. doi: 10.1161/CIRCINTERVENTIONS.110.958348. [DOI] [PubMed] [Google Scholar]

[bib25] 25.Powell R.J., Marston W.A., Berceli S.A., et al. Cellular therapy with Ixmyelocel-T to treat critical limb ischemia: the randomized, double-blind, placebo-controlled RESTORE-CLI trial. Mol Ther. 2012 Jun;20(6):1280–1286. doi: 10.1038/mt.2012.52. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib26] 26.Losordo D.W., Kibbe M.R., Mendelsohn F., Marston W., Driver V., Sharafuddin M., et al. A randomized, controlled pilot study of autologous CD34+ cell therapy for critical limb ischemia. Circ Cardiovasc Interv. 2012;5(6):821–830. doi: 10.1161/CIRCINTERVENTIONS.112.968321. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib27] 27.Teraa M., Sprengers R.W., Schutgens R.E., et al. Effect of repetitive intra-arterial infusion of bone marrow mononuclear cells in patients with no-option limb ischemia: the randomized, double-blind, placebo-controlled rejuvenating endothelial progenitor cells via transcutaneous intra-arterial supplementation (JUVENTAS) trial. Circulation. 2015 Mar 10;131(10):851–860. doi: 10.1161/CIRCULATIONAHA.114.012913. [DOI] [PubMed] [Google Scholar]

[bib28] 28.Pignon B., Sevestre M.A., Kanagaratnam L., Pernod G., Stephan D., Emmerich J., et al. Autologous bone marrow mononuclear cell implantation and its impact on the outcome of patients with critical limb ischemia: results of a randomized, double-blind, placebo-controlled trial. Circ J. 2017;81(11):1713–1720. doi: 10.1253/circj.CJ-17-0045. [DOI] [PubMed] [Google Scholar]

[bib29] 29.Horie T., Yamazaki S., Hanada S., et al. Outcome from a randomized controlled clinical trial: improvement of peripheral arterial disease by Granulocyte colony-stimulating factor-mobilized autologous peripheral-blood-mononuclear cell transplantation (IMPACT) Circ J. 2018;82(8):2165–2174. doi: 10.1253/circj.CJ-17-1220. [DOI] [PubMed] [Google Scholar]

[bib30] 30.Norgren L., Weiss N., Nikol S., et al. PACE: randomized, controlled, multicentre, multinational, phase III study of PLX-PAD for critical limb ischemia in patients unsuitable for revascularization: randomized clinical trial. Br J Surg. 2024 Jan 31;111(2) doi: 10.1093/bjs/znad437. znad437. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib31] 31.Chacón Quevedo A., García-Revillo García J., Sagrario Lombardo Galera M., et al. Autologous bone marrow mononuclear cells in the treatment of chronic limb-threatening ischemia: a proof-of-concept trial. Cardiovasc Res. 2022;118(Suppl 1):cvac066–cvac236. doi: 10.1093/cvr/cvac066.236. [DOI] [Google Scholar]

[bib32] 32.Kwiatkowski T., Zbierska-Rubinkiewicz K., Krzywon J.W., et al. Cellular therapies in no-option critical limb ischemia: present status and future directions. Postepy Kardiol Interwencyjnej. 2022 Dec;18(4):340–349. doi: 10.5114/aic.2022.120962. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib33] 33.Liang G., Zhang Y., Chen C., et al. Clinical trial landscape of stem cell therapy for peripheral arterial disease. Stem Cell Rev Rep. 2026;22:1042–1044. doi: 10.1007/s12015-025-11026-z. [DOI] [PubMed] [Google Scholar]

[bib34] 34.Pan T., Liu H., Fang Y., et al. Predictors of responders to mononuclear stem cell-based therapeutic angiogenesis for no-option critical limb ischemia. Stem Cell Res Ther. 2019 Jan 11;10(1):15. doi: 10.1186/s13287-018-1117-5. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib35] 35.FDA-NIH Biomarker Working Group . Silver Spring (MD): Food and Drug Administration (US) National Institutes of Health (US); Bethesda (MD): 2016. BEST (Biomarkers, EndpointS, and other tools) resource.https://www.ncbi.nlm.nih.gov/books/NBK326791/ Available from: [PubMed] [Google Scholar]

[bib36] 36.US Food and Drug Administration Enrichment strategies for clinical trials to support determination of effectiveness of human drugs and biological products: guidance for industry. Silver Spring (MD): US Food Drug Administration. 2019 [Google Scholar]

[bib37] 37.Panc C., Guler A., Gurbak I., et al. Association between CRP/Albumin ratio and long-term mortality in patients with chronic limb-threatening ischemia undergoing endovascular therapy below the knee: the ACHILES-BTK registry. Ann Vasc Surg. 2022 May;82:172–180. doi: 10.1016/j.avsg.2021.11.008. [DOI] [PubMed] [Google Scholar]

[bib47] 38.Kurniawan R.B., Siahaan P.P., Saputra P.B., et al. Neutrophil-to-lymphocyte ratio as a prognostic biomarker in patients with peripheral artery disease: a systematic review and meta-analysis. Vasc Med. 2024 Dec;29(6):687–699. doi: 10.1177/1358863X241281699. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib48] 39.Gary T., Pichler M., Belaj K., et al. Neutrophil-to-lymphocyte ratio and its association with critical limb ischemia in patients with PAOD. PLoS One. 2013;8(2) doi: 10.1371/journal.pone.0056745. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib49] 40.Su M.I., Liu C.W. Neutrophil-to-lymphocyte ratio associated with an increased risk of mortality in patients with critical limb ischemia. PLoS One. 2021 May 27;16(5) doi: 10.1371/journal.pone.0252030. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib46] 41.Levin M.G., Klarin D., Georgakis M.K., et al. A missense variant in the IL-6 receptor and protection from peripheral artery disease. Circ Res. 2021 Oct 29;129(10):968–970. doi: 10.1161/CIRCRESAHA.121.319589. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib38] 42.Rockwood K., Song X., MacKnight C., et al. A global clinical measure of fitness and frailty in elderly people. CMAJ (Can Med Assoc J) 2005 Aug 30;173(5):489–495. doi: 10.1503/cmaj.050051. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib39] 43.Hata Y., Iida O., Takahara M., et al. Non-home discharge after endovascular therapy for chronic limb-threatening ischemia - insights from a Japanese nationwide registry. Circ J. 2026;90(5):559–561. doi: 10.1253/circj.CJ-25-0698. [DOI] [PubMed] [Google Scholar]

[bib52] 44.Ju M., Kim Y., Seo K.W. Role of nutrition in wound healing and nutritional recommendations for promotion of wound healing: a narrative review. Ann Clin Nutr Metab. 2023 Dec 1;15(3):67–71. doi: 10.15747/ACNM.2023.15.3.67. [DOI] [Google Scholar]

[bib40] 45.World Health Organization . World Health Organization; Geneva: 2024. Guideline on haemoglobin cutoffs to define anaemia in individuals and populations.https://www.who.int/publications/i/item/9789240088542 Available from: [PubMed] [Google Scholar]

[bib41] 46.O'Hare A.M., Feinglass J., Sidawy A.N., et al. Impact of renal insufficiency on short-term morbidity and mortality after lower extremity revascularization: data from the department of veterans affairs' National Surgical Quality Improvement Program. J Am Soc Nephrol. 2003 May;14(5):1287–1295. doi: 10.1097/01.asn.0000061776.60146.02. [DOI] [PubMed] [Google Scholar]

[bib42] 47.Kidney Disease: Improving Global Outcomes (KDIGO) CKD Work Group KDIGO 2024 Clinical practice guideline for the evaluation and management of chronic kidney disease. Kidney Int. 2024 Apr;105(4S):S117–S314. doi: 10.1016/j.kint.2023.10.018. [DOI] [PubMed] [Google Scholar]

[bib43] 48.Lanzer P., Boehm M., Sorribas V., Thiriet M., Janzen J., Zeller T., St Hilaire C., Shanahan C. Medial vascular calcification revisited: review and perspectives. Eur Heart J. 2014 Jun 14;35(23):1515–1525. doi: 10.1093/eurheartj/ehu163. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib44] 49.Stenvinkel P. Inflammation in end-stage renal disease: the hidden enemy. Nephrology. 2006 Feb;11(1):36–41. doi: 10.1111/j.1440-1797.2006.00541.x. [DOI] [PubMed] [Google Scholar]

[bib53] 50.Mills JL Sr, Conte M.S., Armstrong D.G., et al. Society for Vascular Surgery Lower Extremity Guidelines Committee The society for vascular surgery lower extremity threatened limb classification system: risk stratification based on wound, ischemia, and foot infection (WIfI) J Vasc Surg. 2014 Jan;59(1):220–234. doi: 10.1016/j.jvs.2013.08.003. e1-2. [DOI] [PubMed] [Google Scholar]

[bib54] 51.Game F.L., Hinchliffe R.J., Apelqvist J., et al. A systematic review of interventions to enhance the healing of chronic ulcers of the foot in diabetes. Diabetes Metab Res Rev. 2012 Feb;28(Suppl 1):119–141. doi: 10.1002/dmrr.2246. [DOI] [PubMed] [Google Scholar]

[bib55] 52.Siracuse J.J., Farber A., Menard M.T., et al. Advanced wound, ischemia, and foot infection stage is associated with poor outcomes in the BEST-CLI trial. J Vasc Surg. 2025 Mar;81(3):720–729.e1. doi: 10.1016/j.jvs.2024.11.027. [DOI] [PubMed] [Google Scholar]

[bib56] 53.Tsimikas S. A test in context: Lipoprotein(a): diagnosis, prognosis, controversies, and emerging therapies. J Am Coll Cardiol. 2017 Feb 14;69(6):692–711. doi: 10.1016/j.jacc.2016.11.042. [DOI] [PubMed] [Google Scholar]

[bib57] 54.Tomoi Y., Takahara M., Soga Y., et al. Impact of high Lipoprotein(a) levels on delayed wound healing in patients with chronic limb-threatening ischemia after peripheral endovascular therapy. Circ J. 2025 Apr 25;89(5):550–556. doi: 10.1253/circj.CJ-24-0383. [DOI] [PubMed] [Google Scholar]

[bib58] 55.Kronenberg F., Mora S., Stroes E.S.G., et al. Lipoprotein(a) in atherosclerotic cardiovascular disease and aortic stenosis: a European Atherosclerosis Society consensus statement. Eur Heart J. 2022 Oct 14;43(39):3925–3946. doi: 10.1093/eurheartj/ehac361. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib50] 56.Bodnar R.J., Yates C.C., Wells A. IP-10 blocks vascular endothelial growth factor-induced endothelial cell motility and tube formation by inhibiting calpain. Circ Res. 2006 Mar 17;98(5):617–625. doi: 10.1161/01.RES.0000209968.66606.10. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib59] 57.Idei N., Soga J., Hata T., et al. Autologous bone-marrow mononuclear cell implantation reduces long-term major amputation risk in patients with critical limb ischemia: a comparison of atherosclerotic peripheral arterial disease and Buerger disease. Circ Cardiovasc Interv. 2011 Feb 1;4(1):15–25. doi: 10.1161/CIRCINTERVENTIONS.110.955724. [DOI] [PubMed] [Google Scholar]

[bib60] 58.Iida O., Takahara M., Kodama A., et al. Clinical outcomes of chronic limb-threatening ischemia due to inflammatory nonatherosclerotic versus atherosclerotic etiologies undergoing revascularization. Vasc Med. 2023 Aug;28(4):299–307. doi: 10.1177/1358863X231164659. [DOI] [PubMed] [Google Scholar]

[bib61] 59.Obara H., Matsubara K., Fujimura N., et al. Preliminary report of endovascular treatment for critical limb ischemia patients with connective tissue disease: case series and review of the literature. Int J Angiol. 2015 Jun;24(2):137–142. doi: 10.1055/s-0035-1547516. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib62] 60.Ouchi S., Takehisa K., Murata N., et al. Long-term outcomes of endovascular therapy for chronic limb-threatening ischemia patients with connective tissue disease. Cardiovasc Interv Ther. 2025 Oct;40(4):860–867. doi: 10.1007/s12928-025-01157-y. [DOI] [PubMed] [Google Scholar]

[bib63] 61.Kondo K., Yanishi K., Hayashida R., et al. Long-term clinical outcomes survey of bone marrow-derived cell therapy in critical limb ischemia in Japan. Circ J. 2018 Mar 23;82(4):1168–1178. doi: 10.1253/circj.CJ-17-0510. [DOI] [PubMed] [Google Scholar]

[bib64] 62.Shoji K., Yanishi K., Yoshimi R., et al. Impact of therapeutic angiogenesis using autologous bone marrow-derived mononuclear cells implantation in critical limb ischemia with scleroderma - subanalysis of the long-term clinical outcomes survey. Circ J. 2019 Feb 25;83(3):662–671. doi: 10.1253/circj.CJ-18-1044. [DOI] [PubMed] [Google Scholar]

[bib45] 63.Shirbaghaee Z., Heidari Keshel S., et al. Report of a phase 1 clinical trial for safety assessment of human placental mesenchymal stem cells therapy in patients with critical limb ischemia (CLI) Stem Cell Res Ther. 2023 Jul 5;14(1):174. doi: 10.1186/s13287-023-03390-9. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib51] 64.Houghton J.S.M., Meffen A., Gray L.J., et al. Leg ischemia management Collaboration. Streamlined clinical management pathways may reduce major amputations in patients with chronic limb threatening ischemia: a prospective cohort study with historical controls. Eur J Vasc Endovasc Surg. 2025 Mar;69(3):465–473. doi: 10.1016/j.ejvs.2024.09.005. [DOI] [PubMed] [Google Scholar]

[bib65] 65.Laursen D.R.T., Paludan-Muller A.S., Hrobjartsson A. Randomized clinical trials with run-in periods: frequency, characteristics and reporting. Clin Epidemiol. 2019;11:169–184. doi: 10.2147/CLEP.S188752. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib66] 66.Suissa S. Immortal time bias in pharmacoepidemiology. Am J Epidemiol. 2008 Feb 15;167(4):492–499. doi: 10.1093/aje/kwm324. [DOI] [PubMed] [Google Scholar]

[bib67] 67.Lakhter V., Weinberg M.D., Galmer A., et al. Objective outcome measures for trials in patients with chronic limb-threatening ischemia across 2 decades: analysis and recommendations. JACC Cardiovasc Interv. 2021 Dec 13;14(23):2584–2597. doi: 10.1016/j.jcin.2021.08.079. [DOI] [PubMed] [Google Scholar]

[bib68] 68.Wei Q., Wang Y., Ma K., et al. Extracellular vesicles from human umbilical cord Mesenchymal stem cells facilitate diabetic wound healing through MiR-17-5p-mediated enhancement of angiogenesis. Stem Cell Rev Rep. 2022 Mar;18(3):1025–1040. doi: 10.1007/s12015-021-10176-0. [DOI] [PubMed] [Google Scholar]

[bib69] 69.Bailey A.J.M., Li H., Kirkham A.M., et al. MSC-derived extracellular vesicles to heal diabetic wounds: a systematic review and meta-analysis of preclinical animal studies. Stem Cell Rev Rep. 2022 Mar;18(3):968–979. doi: 10.1007/s12015-021-10164-4. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib70] 70.Carcopino C., Rossi E., Mebarki M., et al. Understanding the angiogenic characteristics of clinical-grade mesenchymal stromal cells isolated from human umbilical cord. Stem Cell Rev Rep. 2024 Jul;20(5):1353–1356. doi: 10.1007/s12015-024-10712-8. [DOI] [PubMed] [Google Scholar]

[bib71] 71.Dominici M., Le Blanc K., Mueller I., et al. Minimal criteria for defining multipotent mesenchymal stromal cells. The international society for cellular therapy position statement. Cytotherapy. 2006;8(4):315–317. doi: 10.1080/14653240600855905. [DOI] [PubMed] [Google Scholar]

[bib72] 72.Shimizu Y., Inoue Y., Matsuura N., et al. Mesenchymal stromal cell-derived extracellular vesicles in regenerative medicine: standardization, bioengineering, and clinical translation. Regen Ther. 2026;31 doi: 10.1016/j.reth.2025.101058. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib73] 73.Henon P., Kowalczyk M., Aries A., et al. Industrialized GMP production of CD34+ cells (ProtheraCytes) at clinical scale for treatment of ischemic cardiac diseases is feasible and safe. Stem Cell Rev Rep. 2022 Jun;18(5):1614–1626. doi: 10.1007/s12015-022-10373-5. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib74] 74.Hegde M., Singh A.K., Kannan S., et al. Therapeutic applications of engineered mesenchymal stromal cells for enhanced angiogenesis in cardiac and cerebral ischemia. Stem Cell Rev Rep. 2024 Nov;20(8):2138–2154. doi: 10.1007/s12015-024-10787-3. [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

Host-state biomarkers and clinical factors for stratification and optimization of cell therapy trials in chronic limb-threatening ischemia: A review

Yusuke Shimizu

Edward Hosea Ntege

Naoki Matsuura

Tatsuya Ishii

Hiroshi Sunami

Kojiro Furukawa

Moriyasu Nakaema

Akihiro Tokushige

Yoshikazu Inoue

Abstract

Background

Main body

Conclusions

Highlights

Abbreviations

1. Background

1.1. Literature search, study selection, and evidence handling

2. Evidence for prognostic host-state markers in CLTI

Table 1.

2.1. Nutritional and renal markers

2.2. Inflammatory and coagulation markers

2.3. Infection and limb-threat staging

2.4. Lipoprotein(a) and disease etiology

2.5. Perfusion, metabolic, functional, and other adjunct markers

3. Existing risk stratification approaches and their limitations

4. Toward host-state-informed stratification: A conceptual framework

Fig. 1.

4.1. Potential core host-state marker panel

4.2. Conceptual host-state profiling for trial-design stratification

4.3. Optional optimization and run-in strategies as secondary trial-design considerations

5. Consequences for trial design

5.1. Enrichment and stratification strategies

5.2. Endpoint considerations

5.3. Product and safety reporting

6. Limitations and validation needs

7. Future directions

8. Conclusion

Ethics approval and consent to participate

Consent for publication

Availability of data and materials

Authors’ contributions

Funding

Declaration of competing interest

Acknowledgments

Footnotes

Appendix A. Supplementary data

Fig. S1.

References

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases