Prospective clinical and biomarker validation of the ASTCT consensus definition for transplant-associated thrombotic microangiopathy (TA-TMA)

Abstract

Introduction:

Transplantation-associated thrombotic microangiography (TA-TMA) is a disorder that causes severe complications post-allogeneic hematopoietic cell transplantation (allo-HCT). Diagnosing TA-TMA is challenging due to non-standardized criteria. In this study, we aimed to evaluate the new TA-TMA consensus definition from the American Society for Transplantation and Cellular Therapy (ASTCT) panel as part of an ongoing prospective pediatric cohort study and to compare the impact and outcomes of employing the current clinical TMA definition (cTMA) versus the new consensus definition.

Methods:

We included patients aged 0 to 18 years who underwent their first allo-HCT from May 2021 to January 2023 at Texas Children’s Hospital. We compared the incidence, biomarkers, and outcomes of TA-TMA applying the previous and recently proposed screening algorithms and definitions.

Results:

While the classic microangiopathic hemolytic anemia (MAHA)-based cTMA definition led to an incidence of 12.7%, the ASTCT-HR definition doubled the incidence to 28.5% by day 100. In contrast to patients with concordant diagnosis (+/+) who had significantly worse post-transplant survival, those reclassified as TA-TMA by the new definition only (−/+) had significantly different prognosis (100% survival at day 100) despite the lack of TMA-directed therapy. Furthermore, biomarkers of the terminal and alternative complement pathways (sC5b9 and Ba, respectively) were significantly elevated around day 15 in the concordant group (+/+) but not in the discordant group (−/+) when compared to the non-TMA patients.

Conclusions:

The recently proposed ASTCT consensus TA-TMA diagnosis is more sensitive and allows earlier recognition of manifestation that requires closer clinical monitoring but risks over-diagnosis and over-treatment. We recommend additional prospective validation.

Introduction:

Transplantation-associated thrombotic microangiography (TA-TMA) is a disorder characterized by microvascular endothelial dysfunction leading to microangiopathic hemolytic anemia and severe end-organ damage after allogeneic hematopoietic cell transplantation (allo-HCT)¹. This condition has been associated with various risk factors, including potential genetic susceptibility, conditioning regimen, and post-transplant complications like graft-versus-host disease (GVHD) and infections.^2–4

TA-TMA is currently a challenging clinical diagnosis made by experienced HCT providers. Due to the lack of harmonization for diagnostic criteria, the actual incidence of TA-TMA in research studies remains ambiguous. The reported incidence varies considerably, with estimates ranging from 0.8% to 39%^2,3,5–8 and a pooled incidence of 12% based on a recent systematic review and meta-analysis³. The introduction of a new TA-TMA consensus definition by the American Society for Transplantation and Cellular Therapy (ASTCT) holds the potential to considerably influence clinical practice and data interpretation in TA-TMA-related research⁹. To date, no published studies have prospectively employed this recently proposed criteria. Furthermore, diagnostic complement biomarker correlation used in previous TA-TMA studies, such as levels of the terminal complement activation product sC5b-9^10,11 and the alternative complement pathway activation product Factor Ba^12–14, are lacking with the new consensus definition. The ease of understanding and implementing the definition, along with the ambiguity regarding whether the high-risk TA-TMA definition alone is intended to guide treatment, remains to be clarified.

In this study, we aimed to evaluate the new TA-TMA consensus definition from the ASTCT panel as part of an ongoing prospective pediatric cohort study and to compare the impact and outcomes of employing the current clinical definition versus the new TA-TMA consensus definition. Our objectives were to better understand the implementations of the novel consensus criteria and examine the impact of prognostic biomarkers longitudinally.

Methods

Study design and population

In this ongoing prospective cohort study, we studied pediatric and adolescent patients aged 0 to 18 years who underwent their first allo-HCT from May 2021 to January 2023 at Texas Children’s Hospital (TCH). We excluded patients undergoing autologous HCT to maintain the uniformity of data collection. We excluded patients who did not have at least 100 days post-HCT follow-up unless TA-TMA, death, or subsequent HCT had occurred. This study was approved by the Institutional Review Board at Baylor College of Medicine.

Outcomes definitions and capture

For patients who consented and enrolled in the prospective cohort study, we applied two different outcome screening algorithms and definitions for TA-TMA. First, we developed a prospective weekly screening trigger embedded in our institutional electronic health record system to screen for potential microangiopathic hemolytic anemia (MAHA).^5,15,16 Patients with 4/4 of the following criteria: schistocytes on automatic or manual differential review (1+ or higher), severe thrombocytopenia (platelet <50 K/uL or platelet <100 K/uL with preceding transfusion), severe anemia (Hb <9 g/dL) within 24 hours, along with elevated lactate dehydrogenase (LDH > age-adjusted upper limit of normal [ULN]) within 7 days of meeting the other 3/4 criteria were flagged. Those meeting this screening algorithm at least twice in the preceding 14 days were triaged to undergo a two-hematologist medical record adjudication to determine clinical cTMA diagnosis. Specifically, the clinical reviewers were asked to make a clinical diagnosis of TA-TMA while excluding known mimics or alternative causes of TMA that would merit differential intervention; this list included autoimmune hemolytic anemia, severe coagulopathy, disseminated intravascular coagulation (DIC), overt pre-transplant disease relapse (e.g. leukemic blast crisis), transplant graft failure, or transient abnormalities immediately post-conditioning (Table 2)¹⁷. Notably, both the prospective screening and diagnostic criteria for MAHA-based cTMA required the presence of schistocytes.

Table 2.

Details of the Screening and Diagnostic Criteria for TA-TMA

Parameter	Screening
	Current MAHA screening	ASTCT standard-risk
Criteria	1. Meet all criteria (1) to (4) concurrently within 24 hours (an elevated LDH within 7 days is allowed) 2. Above occur ≥2 times in 14 days	1. Meet ≥4 of the following 7 criteria (does not need to be concurrent) 2. Above occur ≥2 times in 14 days
(1) Schistocytes	Present as 1+ or higher	Present
(2) Thrombocytopenia	Defined as one of the following: 1. Plt <50 K/uL 2. Plt <100 K/uL with preceding Plt transfusion	Defined as one of the following: 1. ≥50% Plt reduction from baseline after full platelet engraftment 2. Failure to achieve Plt engraftment 3. Higher than expected Plt transfusion needs or refractoriness
(3) Anemia	Hb <9 g/dL	Defined as one of the following: 1. ≥1 g/dL Hb reduction from baseline after engraftment 2. Failure to achieve RBC transfusion independence 3. New onset of RBC transfusion dependence
(4) Elevated LDH	LDH >ULN	LDH >ULN
(5) Hypertension		1. >99th percentile for age (<18 yr) 2. SBP ≥140 or DBP ≥90 (≥18 yr)
(6) Proteinuria		Spot rUPCR ≥1 mg/mg [Modification: urine protein 2+ or higher on dipstick if rUPCR not available]
(7) Elevated sC5b-9		sC5b9 >ULN
	Diagnosis
	Current cTA-TMA diagnosis	ASTCT high-risk
Criteria	Defined as all of the following: 1. Meeting MAHA screening criteria above 2. No mimics or alternative causes of TA-TMA: - not autoimmune hemolysis - not severe coagulopathy or DIC - not overt pre-transplant disease relapse - not graft failure - not related to initial conditioning unless delayed engraftment 3. Agreed upon by 2 independent reviewers; adjudicated by a 3rd reviewer if disagreement	Defined as all of the following: 1. Meeting ASTCT standard-risk criteria above 2. ANY of the following parameters
(1) Elevated LDH		Peak LDH >2x ULN
(2) Proteinuria		Spot rUPCR ≥1 mg/mg [Modification: urine protein 2+ or higher on dipstick if rUPCR not available]
(3) Elevated sC5b-9		sCb59 >ULN
(4) Renal dysfunction		Serum creatinine ≥2x baseline
(5) aGVHD		aGVHD grade II-IV
(6) Infection		Systemic bacterial or viral infection
(7) Organ dysfunction		Any organ dysfunction developing in the setting of TA-TMA except stage I AKI

Abbreviations: TA-TMA, transplant-associated thrombotic microangiopathy; DIC, disseminated intravascular coagulation; ISTH, International Society on Thrombosis and Haemostasis; ASTCT, American Society for Transplantation and Cellular Therapy; LDH, lactate dehydrogenase; ULN, upper limit of normal; Plt, platelets; Hb, hemoglobin; aGVHD, acute graft-versus-host disease; AKI, acute kidney injury; MAHA, microangiopathic hemolytic anemia; cTA-TMA, clinical transplant-associated TMA; rUPCR, random urine protein-to-creatinine ratio

Since the publication of the recent TMA Harmonization Panel Consensus,⁹ we also reapplied the ASTCT standard-risk (SR) and high-risk (HR) criteria to our consented patients. Specifically, patients were determined to have SR (i.e. positive screening) if they met 4 or more of the 7 criteria (schistocytosis, thrombocytopenia, anemia, elevated LDH, elevated blood pressure, elevated urinary protein excretion, and elevated sC5b9) at least twice in 14 days (Table 2). Notably, these criteria did not need to be concurrent and did not mandate the presence of schistocytosis, as long as 4 or more of them occurred in a sliding 14-day window. All of the above criteria were routinely measured by the treating clinicians for patients in the prospective cohort study except for spot urine protein to creatine ratio where we used dipstick urinalysis 2+ protein as a substitute. Furthermore, patients were classified as meeting the ASTCT-HR definition (i.e. positive diagnosis) if they met the SR and any of the end-organ damage features listed in Table 2. To differentiate the two study definitions, we categorized those who were MAHA/ASTCT-SR doubly negative as −/− screening, MAHA negative/ASTCT-SR positive as −/+ screening, and MAHA/ASTCT-SR doubly positive as +/+ screening. Similarly, we classified cTMA/ASTCT-HR doubly negative as −/− diagnosis, cTMA negative/ASTCT-HR positive as −/+ diagnosis, and cTMA/ASTCT-HR doubly positive as +/+ diagnosis.

Sampling and data collection

For a subset of patients enrolled in the prospective cohort study who also consented to the biospecimen study, we collected serial sodium citrate plasma samples at six pre-determined time points: day −30 to 0 (pre-transplant), 15, 30, 45, 60, and 90. Samples were collected from existing indwelling central venous catheters and centrifuged at 2500 rpm at 4 °C for 15 minutes before aliquoting and storage at −80 °C. Samples were maintained on ice and processed within 2 hours of collection to avoid in-vitro complement activation. Only samples that had undergone a maximum of 1 freeze-thaw cycle were used for analysis. The concentrations of sC5b-9 and Ba were measured using MicroVue sC5b-9 Plus EIA and MicroVue Ba EIA (Quidel, San Diego, CA) kits, respectively.

Statistical methods

The incidence of TA-TMA during the first year after allo-HCT was determined by the cumulative incidence function while treating death and conditioning for subsequent transplant as competing risks. Overall survival (OS) was plotted using the Kaplan-Meier method and compared by the log-rank test stratified by the presence of TA-TMA outcome or the use of eculizumab. Locally weighted scatterplot smoothing regression (LOWESS) analysis was plotted to fit a smooth curve and identify potential trends of biomarkers in TA-TMA and non-TMA groups. For samples collected around day 15 (+/− 14 days), two-sample independent t-tests were used to compare biomarker levels between different groups. Analyses were performed using R 4.2.2 (R Foundation for Statistical Computing, Vienna, Austria).

Results

Baseline characteristics

A total of 38 patients participated in the prospective cohort study over 18 months. After limiting to first allo-HCT and at least 100 days of follow-up, the final analytic cohort consisted of 32 patients (Figure S1). Baseline patient-, disease- and transplant-related characteristics are shown in Table 1. The median age at the time of HCT was 5.7 years (range 1.1–18.8 years), the majority of patients were White (68.8%), Hispanic or Latino (56.2%) with a male predominance (65.6%) and high-performance status at HCT admission (78.1% with Lansky Scale ≥ 90%). Half of the patients had malignant conditions, and the rest had non-malignant disorders. The most common graft source was bone marrow (68.8%). Most conditioning regimens were myeloablative (93.7%). Median follow-up was 221 days after transplantation (range: 62–626 days).

Table 1.

Patient Characteristics at Baseline

Characteristic	Values
Age, yr, median (range)	5.7 (1.1–18.8)
Sex
Male	21.0 (65.6%)
Female	11.0 (34.4%)
Race
White	22.0 (68.8%)
Black or African American	5.0 (15.6%)
Not reported	5.0 (15.6%)
Ethnicity
Hispanic or Latino	18.0 (56.2%)
Not Hispanic or Latino	14.0 (43.8%)
BMI, kg/m 2 , median (range)	17.9 (11.2–37.5)
Primary disease
Malignant*	16.0 (50.0%)
Non-malignant†	16.0 (50.0%)
Stem cell source
Bone marrow	22.0 (68.8%)
Peripheral blood	3.0 (9.4%)
Umbilical cord	7.0 (21.9%)
HLA match status
Matched	16.0 (50.0%)
Haploidentical	7.0 (21.9%)
Mismatched	9.0 (28.1%)
Donor source
Related	13.0 (40.6%)
Unrelated	19.0 (59.4%)
Conditioning group
Myeloablative with high-dose TBI (≥1200 cGy)	9.0 (28.1%)
Myeloablative without high-dose TBI	21.0 (65.6%)
Nonmyeloablative (reduced intensity)	2.0 (6.2%)
Karnofsky Scale (16+) or Lansky Scale (<16)
100	17.0 (53.1%)
90	8.0 (25.0%)
80	5.0 (15.6%)
70 or below	2.0 (6.3%)
GVHD prophylaxis
Cyclosporine-based	14.0 (43.8%)
Others	3.0 (9.4%)
Tacrolimus-based	15.0 (46.9%)
VOD/SOS prophylaxis
Ursodiol	32.0 (100.0%)

Abbreviations: BMI, body mass index; HLA, human leukocyte antigens; GVHD, graft-versus-host disease; TBI, total body irradiation; VOD, hepatic veno-occlusive disease; SOS, Sinusoidal Obstruction Syndrome

^*Malignant disease includes: Acute lymphatic leukemia, Acute myeloid leukemia, Myelodysplastic syndromes;

^†Non-malignant disease includes: Aplastic anemia, Autoimmune diseases, Disorders of the immune system, Hemoglobinopathies (including sickle cell), Histiocytic disorders, Inherited abnormalities of platelets (including Glanzmann), Inherited bone marrow failure syndromes, Inherited disorders of metabolism

Cumulative incidence based on different TA-TMA definitions

Detailed information on individual time to TA-TMA screening and diagnosis based on the two different diagnostic criteria is summarized in Table S1. Figure 1 depicts the cumulative incidence of TA-TMA stratified by each definition. Based on the existing criteria, 22.3% of patients (n=9) had a positive consecutive MAHA screening and 12.7% (n=5) had a cTMA diagnosis by the classical TMA definition by 100 days. In contrast, 34.4% of patients (n=12) had an ASTCT-SR diagnosis and 28.5% (n=10) had an ASTCT-HR diagnosis by 100 days. The median time to cTMA was 77 days while the median time to ASTCT-HR was 61 days. The most common reason for the 12 patients meeting the ASTCT-SR diagnosis were schistocytes (n=11), hypertension (n=10), elevated LDH (n=9), thrombocytopenia (n=8), anemia (n=7), and proteinuria (n=5). The most common reason for the 10 patients meeting ASTCT-HR diagnosis was “peak LDH >2x ULN” (n=4). As shown in Figure 2, there were 3 distinct groups of non-overlapping patients by 100 days, cTMA/ASTCT-HR doubly negative (−/−, n=22), cTMA negative/ASTCT-HR positive (−/+, n=5), and cTMA/ASTCT-HR doubly positive (+/+, n=5).

Figure 1. Cumulative incidence of TA-TMA.

Abbreviations: MAHA, microangiopathic hemolytic anemia; ASTCT, American Society for Transplantation and Cellular Therapy; cTA-TMA, clinical TA-TMA; allo-HCT, allogeneic hematopoietic cell transplantation.

Figure 2. A diagram demonstrating the distribution of each definition of TA-TMA.

Abbreviations: ASTCT, American Society for Transplantation and Cellular Therapy, cTA-TMA, clinical TA-TMA; allo-HCT, allogeneic hematopoietic cell transplantation.

Prognosis based on different TA-TMA definitions

The survival prognosis was different based on different diagnostic criteria. Patients meeting both MAHA and ASTCT-SR screening criteria (+/+) demonstrated a non-significant trend toward worse survival while those meeting only SR criteria (−/+) showed no differences compared to the non-TMA group (−/−) (Figure 3A). This survival distinction was clearer when comparing diagnostic criteria. Patients diagnosed with TA-TMA by both current cTMA and ASTCT-HR criteria (+/+) exhibited significantly worse survival while those meeting only HR criteria (−/+) showed no differences compared to the non-TMA group (−/−) (Figure 3B).

Figure 3. Overall survival stratified by:

(A) TA-TMA screening. +/− group (purple line): screen positive by current MAHA screening criteria but negative by ASTCT standard-risk criteria; −/− group (red line): screen negative by both current MAHA screening and ASTCT standard-risk criteria. −/+ group (green line): screen negative by current MAHA screening criteria but positive by ASTCT standard-risk criteria; +/+ group (yellow line): screen positive by both current MAHA screening and ASTCT standard-risk criteria. (B) TA-TMA diagnosis. −/− group (red line): patients identified as negative by both current cTA-TMA diagnosis and ASTCT high-risk criteria. −/+ group (green line): patients identified as negative by current cTA-TMA diagnosis criteria but positive by ASTCT high-risk criteria; +/+ group (yellow line): patients identified as positive by both current cTA-TMA diagnosis and ASTCT high-risk criteria. (C) eculizumab administration after meeting high-risk TA-TMA criteria. Abbreviation: allo-HCT, allogeneic hematopoietic cell transplantation

TMA-directed intervention and survival at last follow-up are detailed in Table S1. Among 10 patients meeting ASTCT-HR, 4 received eculizumab from 2, 25, 64, and 156 days after meeting HR while 6 did not (1 shortly died after TMA presentation; 5 did not meet cTMA diagnosis). Among 4 patients who were treated with eculizumab, 2 also received therapeutic plasma exchange (TPE). In the unadjusted log-rank test, there was a non-significant trend towards worse survival after ASTCT-HR diagnosis in the eculizumab group (P=0.36, Figure 3C).

Biomarker analysis based on different TA-TMA definitions

Longitudinal biomarker analysis was conducted on 29 patients who consented to the biospecimen protocol. The highest average plasma levels of sC5b-9 and Ba during the first 100 days were observed in patients meeting both cTMA and ASTCT-HR criteria (+/+). In contrast, patients who did not meet either of these criteria (−/−) displayed the lowest levels of these tests (Figure 4). Given the noticeable difference among the groups around day 15, we conducted additional analyses focusing on samples collected within 30 days post-transplant. Figure 5 depicts significantly elevated levels of sC5b-9 (P=0.041) and Ba plasma (P=0.004) in the doubly positive (+/+) as compared to the non-TMA (−/−) group. We did not observe significantly different biomarker distributions for ASTCT-HR singly positive (−/+) versus the non-TMA (−/−) group.

Figure 4. LOWESS (Locally Weighted Scatterplot Smoothing) regression analysis for biomarker analysis.

−/− group (orange): patients identified as negative by both current cTA-TMA diagnosis and ASTCT high-risk criteria. −/+ group (green): patients identified as negative by current cTA-TMA diagnosis criteria but positive by ASTCT high-risk criteria; +/+ group (blue): patients identified as positive by both current cTA-TMA diagnosis and ASTCT high-risk criteria. (B) Ba. Abbreviation: allo-HCT, allogeneic hematopoietic cell transplantation.

Figure 5. Beeswarm plots for biomarker analysis at d15 after allo-HCT in each group.

−/− group (orange): patients identified as negative by both current cTA-TMA diagnosis and ASTCT high-risk criteria. −/+ group (green): patients identified as negative by current cTA-TMA diagnosis criteria but positive by ASTCT high-risk criteria; +/+ group (blue): patients identified as positive by both current cTA-TMA diagnosis and ASTCT high-risk criteria. Each dot represents a sample. Diamonds and solid lines indicate the mean and median values for each group, respectively. (A) sC5b9; (B) Ba. * <0.05; ** < 0.01.

Discussion

To our knowledge, this is the first prospective cohort study to compare and validate the recently proposed ASTCT consensus criteria for TA-TMA using a combination of clinical and biomarker endpoints. While the classic MAHA-based cTMA definition reported an incidence of 12.7%, the newly proposed ASTCT-HR definition doubled the incidence to 28.5% by day 100. In contrast to patients with concordant diagnosis (+/+) who had significantly worse post-transplant survival, those reclassified as TA-TMA by the new definition only (−/+) had significantly different prognosis (100% survival at day 100) despite the lack of TMA-directed therapy. Furthermore, biomarkers of the terminal and alternative complement pathways (sC5b9 and Ba, respectively) were significantly elevated around day 15 in the concordant group (+/+) but not in the discordant group (−/+) when compared to the non-TMA patients. Taken together, the recently proposed ASTCT consensus TA-TMA diagnosis (both standard and high-risk) is more sensitive and can lead to earlier recognition of manifestation that requires closer clinical monitoring; however, it could also lead to potential over-diagnosis and over-treatment as many patients met the 4/7 laboratory criteria plus a singular high risk-feature without overt clinical manifestation of TA-TMA.

In the process of implementing the new ASTCT consensus criteria, we observed benefits and potential deficits associated with the new definition. First, the new diagnostic criteria offered significant clarity on defining anemia and thrombocytopenia, which must occur after neutrophil engraftment and after excluding alternative causes. This clarification obviated the need to rely on singular laboratory testing such as direct antiglobulin test, haptoglobin, prothrombin time, or international normalized ratio for TA-TMA diagnostic criteria and instead advocated for the reliance on clinical diagnosis. Second, the two consecutive time points ensured the filtering of isolated non-clinically relevant events, though it remains unclear what “consecutive” means as the text describes it as any two time points in a 14-day window. Third, the addition of age-adjusted hypertension and proteinuria helped identify certain patients at risk for TA-TMA a few days to weeks before they had full onset of hematologic criteria; though it would be helpful to specify anti-hypertensive agents as an additional requirement for hypertension to preclude two isolated blood pressure measurement in a 14-day span.

We would like to highlight a few challenges encountered. First, it is important to recognize that the consensus panel recommended closer monitoring rather than TMA-directed therapy for ASTCT-SR (meeting 4/7 criteria in 2 separate times within 14 days); therefore, we interpreted it as a positive screening rather than disease diagnosis per se, similar to our existing MAHA screening (meeting 4/4 criteria in 2 separate times within 14 days). Second, in contrast to previous definitions, the 4/7 criteria in the new criteria did not have to occur concurrently; therefore, it was unclear if the same 4 positive criteria needed to repeat at least twice in the 14-day time frame, or any combination of 4/7 would suffice as long as two of such events occurred in the 14-day window. For the current study, we used the former interpretation. Third, consistent with our previous retrospective studies¹⁶ and systematic review³, we observed that schistocytes were present in all patients meeting ASTCT-SR criteria. Schistocytes have been previously reported to be a critical early marker for TA-TMA before the full hematologic criteria became apparent.¹⁸ While the accurate measurement of schistocyte count remains a topic of debate^19,20, five^{15,16,21–23} out of the six⁵ recognized criteria (Table 1 of the ASTCT manuscript) necessitated the presence of schistocytes for TA-TMA diagnosis, advocating for the pivotal role of schistocyte count in the diagnostic process. Finally, in our study, the ASTCT-HR criteria were met in nearly all patients who also met the SR criteria (10 out of 12). This was often attributed to isolated elevation of LDH >2 times ULN or having concurrent bacterial/viral infection or GVHD. Since all patients meeting ASTCT-HR without cTMA had excellent survival outcomes even without TMA-directed therapy, a more specific definition for HR disease would help reduce potential over-treatment.

Our biomarker analysis was preliminary and exploratory; however, it supported our clinical observation as both sC5b9 and Ba exhibited the highest levels in the concordant “+/+” group and the lowest levels in the concordant “−/−” group, especially early post-transplant around day 15. Several studies have demonstrated that early elevation of sC5b9 or Ba is prognostic for the development of TA-TMA.^5,11,13 Our findings in this prospective study confirmed those reports but only for the patients with cTMA diagnosis and not those with isolated ASTCT-HR, a pattern strongly correlated with the stratified survival trends. This is of clinical relevance as sC5b9 was one of the 7 diagnostic criteria and a reason for the initiation of TMA-directed therapy in the new consensus definition.

Our current study has several strengths and limitations. The prospective nature of our study permitted a more controlled investigation, reducing potential biases and allowing for a more rigorous evaluation of the defined hypotheses. The continuous assessment of biomarkers at each time point provided a robust and detailed dataset for analysis. Limitations include the relatively small sample size, the lack of enrollment of autologous HCT patients, and the primary focus on pediatric patients that might limit broader generalization of findings. The follow-up time was relatively short, which could have led to an underestimation of late-onset disease. Lastly, very few patients had a biopsy or autopsy for histologic confirmation of either study definition, which led to the reliance on clinician adjudication of outcomes.

In conclusion, we found that the recently proposed ASTCT consensus definition for TA-TMA may lead to earlier diagnosis but also over-diagnosis in a significant proportion of patients. Patients who solely met the ASTCT-HR diagnostic criteria but not the cTMA criteria had benign biomarker profiles and excellent survival prognosis without TMA-directed therapy. While clinicians need to be vigilant for infrequent yet potentially severe complications like TA-TMA, it is crucial that diagnostic criteria accurately identify affected patients given the expense and potential complications of treatment. We recommend ongoing validations of the new consensus diagnostic criteria from other prospective cohort studies with larger sample sizes and extended follow-up.

Highlights.

• The ASTCT consensus increased TA-TMA diagnosis from 12.7% to 28.5% on day 100.

• Patients with discordant TA-TMA diagnosis by ASTCT did not have adverse prognoses.

• Complement biomarkers were significantly elevated in concordant TA-TMA diagnosis.

• Complement biomarkers were similar in discordant TA-TMA diagnosis.

• The ASTCT consensus definition may lead to over-diagnosis of TA-TMA.

Data sharing statement:

The patient-level data are protected under study agreement with TCH and could not be shared publicly. For detailed questions regarding the modeling specifications, please contact Dr. Ang Li at ang.li2@bcm.edu.