Effect of Total Psoas Muscle Area and Serum Albumin on Outcomes After Lung Transplantation

ABSTRACT

Background

Sarcopenia of the psoas muscle and hypoalbuminemia indicate poor nutritional status, inflammation, and frailty in lung transplant (LT) candidates, correlating with worse post‐transplant outcomes.

Methods

Retrospective study of LT recipients (2015–2023) examining the association of total psoas muscle area (TPA) and serum albumin with hospital stay, survival, and pulmonary function.

Results

One hundred thirty‐two LT recipients (mean age 59.56 ± 10.65 years, BMI 26.73 ± 5.55 kg/m², 65% males), 95% underwent bilateral LT. Higher TPA was associated with shorter hospital and ICU stays (p = 0.001). Similarly, higher albumin levels were associated with reduced hospital and ICU stays (p < 0.001). Hospital survivors had higher TPA (17.5 ± 6.1 vs. 14.6 ± 5.2 cm², p = 0.02) and higher albumin levels (3.25 ± 0.73 vs. 2.75 ± 0.85 mg/dL, p = 0.01). Long‐term survivors had higher TPA (17.8 ± 6.35 vs. 15.9 ± 5.51 cm², p = 0.07) and higher albumin levels (3.29 ± 0.75 vs. 2.97 ± 0.78 mg/dL, p = 0.01). On multivariate analysis, albumin and male gender remained independent predictors of hospital and long‐term survival. TPA was positively associated with post‐transplant pulmonary function based on FVC and FEV1 (p < 0.001), while albumin levels showed no association.

Conclusion

In the present study of LT recipients, higher TPA and albumin levels were linked to shorter hospitalization, and albumin independently predicted survival. TPA, but not albumin, was associated with pulmonary function post‐transplant.

Abbreviations

AM‐PAC 6cBM

activity measure for post‐acute care "6 clicks” basic mobility

BMI

body mass index

CSA

cross‐sectional area

CT

computed tomography

FEV1

forced expiratory volume in one second

FVC

forced vital capacity

ICU

intensive care unit

IPR

inpatient pulmonary rehabilitation

SPM

sarcopenia of the psoas muscle

TPA

total psoas muscle area

1. Introduction

Determining LT candidacy involves evaluating predicted risks, benefits, and the limited availability of donor organs. The lung allocation score and the composite allocation score use multiple variables, including body mass index (BMI), to predict transplant urgency and post‐LT survival. However, BMI does not consistently correlate with body composition. Frailty, often linked to chronic illness and aging, is associated with poor clinical outcomes in chronic lung disease patients and can lead to turn down, delisting, and increased mortality in LT candidates [1, 2, 3, 4, 5].

Traditional frailty assessment tools, designed for community‐dwelling elderly individuals, may not be applicable to those with advanced lung disease, especially critically ill patients [6]. Guidelines recommend evaluating muscle quantity, quality, strength, and physical performance for sarcopenia diagnosis [7, 8, 9]. The International Society for Heart and Lung Transplantation consensus document on the selection of LT candidates recommends that programs select candidates based on risk of death without LT and a high predicted likelihood of survival after transplant. To that end, evaluation of frailty has recently been suggested [9]. Sarcopenia, a progressive skeletal disorder associated with frailty, is prevalent in 30%–40% of LT candidates, potentially reaching 72% depending on assessment methods [9, 10, 11, 12, 13, 14, 15, 16, 17].

Computed tomography (CT) scans are used to measure skeletal muscle cross‐sectional area (CSA) in pre‐ and post‐LT populations [10, 11, 12, 13, 14, 15, 16, 17]. The data on sarcopenia's impact on LT outcomes are mixed. Furthermore, assessing muscle function in critically ill patients is challenging due to medications and hemodynamic instability. In the most vulnerable, sicker patients admitted to the intensive care unit (ICU), the prevalence of sarcopenia is likely higher and associated with increased mortality [12, 14]. The effect of acute, or acute‐on‐chronic, respiratory failure requiring ICU hospitalization for life‐sustaining therapies in the development of sarcopenia is not well studied, with conflicting findings [12, 15]. Additionally, patients with end‐stage lung disease on mechanical ventilation or extracorporeal membrane oxygenation (ECMO) often have unstable hemodynamics, which can hinder a complete evaluation of their physical function or limited physical therapy while they await assessment or transplantation. The evaluation and participation in physical therapy of the sickest patients awaiting LT requires a dedicated multidisciplinary team including intensivists, perfusionists, nurses, respiratory therapists, and highly trained physical therapists [18, 19].

As low psoas muscle mass has been the one muscle group associated with clinical outcomes in previous studies with good inter‐rater reliability [12, 13], we chose to study the effect of the psoas muscle mass on clinical outcomes post‐LT.

Serum albumin level, affected by malnutrition and inflammation, is a marker of frailty and predicts 1‐year post‐LT mortality [17, 20, 21].

This study investigates the impact of total psoas muscle area (TPA) and serum albumin levels on survival, length of stay, and pulmonary function tests in LT recipients.

2. Materials and Methods

2.1. Study Design

Single‐center retrospective study of patients who underwent LT between January 24, 2015, and October 16, 2023.

The institutional review board approved this study as minimal‐risk research using data collected for routine clinical practice (#16796). The study was performed in accordance with the ethical standards of the 1964 Declaration of Helsinki and its later amendments or comparable ethical standards.

2.2. Inclusion Criteria

Patients older than 18 years who underwent a LT during the study period and who had a CT scan within 6 months peri‐transplant. The initial diagnosis point was defined as the time of CT scan evaluation. All LT candidates included had albumin levels measured in serum, an assessment of functionality by physical therapists, and had a CT scan of the abdomen and pelvis within 6 months of transplant or decision not to transplant.

The survival time was defined as the duration in years from the date of LT to the last documented follow‐up visit on surviving patients or the date of death. After the last follow‐up visit, the data were censored.

2.3. Exclusion Criteria

Patients not transplanted, those without available CT scans, and individuals younger than 18 years were excluded from the study.

2.4. Variables Measured

Baseline demographics abstracted from the medical record included age, sex, BMI, and indication for LT. Ambulatory patients were defined as those who were not hospitalized prior to admission for elective LT surgery.

Psoas muscle evaluation: CT scans were routinely obtained 6 months before or after transplant. CT scans, both with and without intravenous contrast‐enhancement, that were closest to the date of LT were considered. As the psoas muscle surface does not distinguish between relative proportions of fat and lean muscle mass, the psoas muscle quality was assessed by measuring its density in Hounsfield units. The value of radiographic sarcopenia was compared to previously reported cutoffs of sarcopenia in thoracic surgery, including LT patients. In case multiple data points of the same variable were available, the data point closest to the time of the diagnostic CT was chosen. CSA measurements of the psoas muscle at the L3 (lumbar spine) level were used for sarcopenia evaluation using the total psoas area (TPA), the psoas muscle index (PsI), and the density of the psoas muscle (PsD). The following parameters were employed: 120 kV, 40 mAs, 32 × 0.6 mm detector collimation, pitch 1. Images were reconstructed with slice thickness between 1.25 and 5.0 mm using standard and high‐resolution reconstruction algorithms. CSAs were obtained utilizing Vitrea 7.14.2 (Canon Medical Informatics, Minnetonka, MN), a software used for radiographic image analysis.

Two independent radiologists, blinded to patients’ clinical data and to the outcomes of interest, measured the right and left psoas muscles at the level of the third lumbar vertebra on axial CT images. The chosen single slice was located halfway between the superior and inferior borders of the L3 vertebral body. When assessing muscle mass, the TPA measures the surface areas of the right and left psoas muscles in square centimeters (cm²) added together (TPA = right psoas area + left psoas area). TPA in cm² corrected by height in m² describes the PsI (PsI = TPA/height). In addition, the values obtained were used to measure psoas density (PsD) by measuring the mean attenuation values in Hounsfield units and its standard deviations (SD). PsD was estimated adding the product of the right mean psoas Hounsfield unit density and RPA with that of the product of the left mean psoas Hounsfield unit density and LPA, then dividing by the TPA: PsD = ([right psoas attenuation*right PsA] + [left psoas attenuation*left PsA])/TPA. We did not conduct a reliability analysis as an excellent inter‐rater agreement has been previously reported [22].

Sarcopenia of the psoas muscle (SPM): TPA measurements were used to define sarcopenia of the psoas muscle (SPM), with a TPA below the first quartile indicating significant muscle mass wasting.

Physical function: Baseline assessment was during the first days after LT and before hospital discharge. Physical function is described by the Activity Measure for Post‐Acute Care “6 clicks” Basic Mobility (AM‐PAC 6cBM), the Johns Hopkins Highest Level of Mobility (JH‐HLM) and the sit‐to‐stand test. Physical function assessment was utilized to recommend discharge disposition after LT (home vs. inpatient pulmonary rehabilitation [IPR]). Based on this scale, a JH‐HLM < 2 is seen in fully dependent patients, JH‐HLM 3–6 and AM‐PAC 6cBM < 17 in partially dependent patients, and JH‐HLM > 8, AM‐PAC 6cBM > 22 in independent patients without significant physical limitation.

Hypoalbuminemia: Defined by a serum albumin level <3.5 mg/dL. The value used for analysis was that obtained closest to the CT scan used for evaluation of psoas CSA.

Pulmonary function assessment: Spirometric values were obtained post‐LT: best forced vital capacity or FVC in mililiters (mL) and best forced expiratory volume in 1 second (FEV1, in mL), and last FVC and FEV1 obtained during the last clinic visit.

2.5. Outcomes

To report on the clinical outcomes of interest, we collected data on ICU and hospital lengths of stay (LOS), hospital survival, long‐term survival, and pulmonary function tests post‐transplantation as indicators of long‐term allograft function.

2.6. Management of LT

LT recipients received induction immunosuppression with basiliximab, mycophenolate, and methylprednisolone, unless contraindicated. Maintenance immunosuppression started between Days 2 and 4 after LT, with a calcineurin inhibitor (tacrolimus preferred), with a target trough level goal of 10–12 ng/mL for tacrolimus or 250–350 ng/mL for cyclosporin A, during the first year. The level of maintenance immunosuppression is based on post‐LT timing, history of rejection and infections, as well as tolerance. Most patients receive a combination of three drugs including a calcineurin inhibitor (cyclosporin A or tacrolimus), systemic corticosteroids, and a cell‐cycle inhibitor (azathioprine or mycophenolate). Prophylaxis included azithromycin for bronchiolitis obliterans, trimethoprim‐sulfamethoxazole for Pneumocystis jirovecii, valganciclovir for cytomegalovirus, and inhaled amphotericin for fungal infections.

Nutrition was initiated soon after transplant, either enterally or orally post‐swallow assessment. A transplant nutritionist tailored the nutritional plan. Physical rehabilitation began within 24 h postoperatively for stable patients, including reassessment of supplemental oxygen, activity balancing, and early in‐hospital rehabilitation. Daily assessments of lower extremity strength and gait speed determined discharge readiness.

Allograft function was regularly evaluated during clinic visits using spirometry and chest X‐rays, with surveillance bronchoscopies at specified intervals up to 24 months post‐LT.

Recipients attended pulmonary rehabilitation, either outpatient or inpatient, with at least three sessions per week, typically up to 36 sessions, or more if needed to achieve goals.

2.7. Statistical Analyses

Patients’ demographic, clinical, radiological, and laboratory data were collected for statistical analysis. Data are presented as mean and SD if normally distributed and median (25% and 75% percentiles) or interquartile range (IQR) if not. For parametric data, differences in the mean were compared using Student's t‐test or analysis of variance. The Wilcoxon‐Mann‐Whitney/Kruskal‐Wallis test was used for highly skewed data. The normality of the data was assessed using a frequency distribution histogram. Differences in proportions were assessed by the chi‐square or Fisher's exact test. A regression model was performed to study the association between length of stay and pulmonary function tests. Cox proportional hazards models were used to study the associations and adjust for confounding factors. Variables with p ≤ 0.05 in univariate analyses were included. Statistical significance was set at p < 0.05. All analyses were performed using the JMP statistical software version 16 (SAS Institute, Cary, NC).

3. Results

Baseline demographic and clinical characteristics of transplanted patients are shown in Table 1. One hundred and forty‐seven patients underwent pre‐transplant evaluation; of those, 132 patients underwent LT during the study period (123 bilateral LT, 6 single LT, 3 patients with liver‐bilateral LT) (Figure 1).

TABLE 1.

Clinical characteristics and outcomes of transplanted patients.

Variable	Response	All (N = 132)	ICU prior to LT (N = 28)	Ambulatory prior to LT (N = 104)	p value
Age (mean ± SD), years		59.5 ± 10.3	52.8 ± 12.6	61.3 ± 8.78	<0.001
Sex	Female	45 (34%)	11 (39%)	34 (33%)	0.51
	Male	87 (66%)	17 (61%)	70 (67%)
Race	White	109 (83%)	23 (82%)	86 (83%)	0.14
	Black	17 (13%)	2 (7%)	15 (14%)
	Other	6 (5%)	3 (11%)	3 (3%)
BMI (mean ± SD), kg/m2		26.73 ± 5.55	27.56 ± 6.26	26.51 ± 5.35	0.42
LT indication	Pulmonary fibrosis	71 (53.7%)	15 (53.5%)	56 (53.8%)	0.33
	Emphysema	15 (11.3%)	0 (0%)	15 (14.4%)
	CPFE	11 (8.3%)	0 (0%)	11 (8.3%)
	ARDS	8 (6%)	8 (28.6%)	0 (0%)
	Sarcoidosis	7 (5.3%)	1 (3.5%)	6 (5.7%)
	CTD‐ILD	6 (4.5%)	2 (7.14%)	4 (3.85%)
	PAH	3 (2.2%)	0 (0%)	3 (2.2%)
	Bronchiectasis	5 (3.7%)	0 (0%)	5 (4.8%)
	Others	6 (4.5%)	2 (7.1%)	4 (3.8%)
LAS (median, IQR [25%–75%])		65 [41.9–87.3]	87.7 [85.3–89.7]	57.1 [41.4–85]	<0.01
Charlson Comorbidity index (mean ± SD)		1.72 ± 1.63	1.86 ± 1.78	1.68 ± 1.59	0.62
Serum albumin, mg/dL (mean ± SD)		3.17 ± 0.78	3.10 ± 0.77	3.18 ± 0.78	0.63
RPA, cm2 (mean ± SD)		8.32 ± 3.01	8.08 ± 2.72	8.38 ± 3.10	0.64
LPA, cm2 (mean ± SD)		8.66 ± 3.15	8.57 ± 3.04	8.68 ± 3.19	0.87
TPA, cm2 (mean ± SD)		16.97 ± 6.04	16.65 ± 5.65	17.06 ± 6.16	0.28
PsI, cm2/m2 (mean ± SD)		5.57 ± 1.73	5.46 ± 1.49	5.60 ± 1.80	0.25
PsD (mean ± SD)		36.64 ± 13.82	37.02 ± 13.19	36.54 ± 14.05	0.87
LOS, days (mean ± SD)		68.07 ± 62.42	94.46 ± 54.07	60.96 ± 62.84	0.01
ICU, days (mean ± SD)		52.15 ± 65.06	74.68 ± 42.38	46.03 ± 68.87	<0.01
Disposition	Home	70 (53%)	11 (44%)	59 (57.8%)	<0.01
	IPR	43 (32.5%)	11 (44%)	32 (31%)
	Other	14 (10.6%)	3 (12%)	11 (10.8%)
	Dead	4 (3%)	4 (14%)	0 (0%)
Physical function at hospital discharge	JH<2	10 (8%)	3 (11%)	7 (7%)	0.02
AM‐PAC 6cBM and JH‐HLM a	6c<17, JH3‐6	27 (20%)	11 (39%)	16 (15%)
	6c18‐21, JH = 7	53 (40%)	10 (36%)	43 (41%)
	6c>22, JH>8	34 (26%)	2 (7%)	32 (31%)
	UNK	8 (6%)	2 (7%)	6 (6%)
Survival years (mean ± SD)		2.80 ± 2.18	2.21 ± 1.93	2.96 ± 2.23	0.11
Alive at last follow‐up		75 (57.7%)	17 (62.9%)	58 (56.3%)	0.66

Abbreviations: AM‐PAC 6cBM (6c), activity measure for post‐acute care “6 clicks” basic mobility; ARDS, acute respiratory distress syndrome; CPFE, combined pulmonary fibrosis and emphysema; CTD‐ILD, connective tissue disease‐interstitial lung disease; ICU, intensive care unit; IPR, inpatient pulmonary rehabilitation; JH‐HLM (JH), Johns Hopkins Highest Level of Mobility; LAS, lung allocation score; LOS, length of stay; LPA, left psoas area; LT, lung transplant; PAH, pulmonary arterial hypertension; PsI, psoas muscle index (cm²/m²); PsD, psoas density; RPA, right psoas area; SD, standard deviation; TPA, total psoas area (cm²); UNK, unknown; 25%–75% IQR (interquartile range).

^a JH<2: fully dependent; 6c<17, JH3‐6: partially dependent; 6c18‐21, JH = 7: partially independent; 6c>22, JH>8: independent.

FIGURE 1.

Patients evaluated for sarcopenia of the psoas muscle with CT scan. ICU, intensive care unit; SPM, sarcopenia of the psoas muscle; TPA, total psoas area (cm²).

LT was performed in 28 patients admitted to the ICU and 104 patients who waited for LT at home. The most common reason for LT was fibrotic interstitial lung disease (54%). Besides higher lung allocation scores among ICU patients and older age among ambulatory patients, other baseline characteristics were similar between the groups.

For ambulatory patients prior to LT, the median walk distance during a 6‐min walk test was 248 meters (IQR, 300–190 m). The walking distance recorded for patients in the ICU at the time of listing was 0.

The median perioperative time from CT scan to LT was 36.5 days (IQR, 13–80.75 days). Forty percent of CT scans were performed before LT or the decision not to transplant, and the rest were completed after LT.

There was an excellent correlation between TPA with RPA (0.97), LPA (0.98), and PsI (0.96), p < 0.001, so TPA was used in statistical models.

Among the 28 lung transplant ICU patients, bridging strategies included noninvasive ventilation in four, invasive mechanical ventilation in four, and VV‐ECMO in 20 patients.

Seventy‐one percent of the patients were able to complete a sit‐to‐stand test prior to LT. Except for eight patients with acute respiratory distress syndrome or ARDS supported on mechanical ventilation or ECMO, most patients were able to receive prehabilitation prior to LT. There was a significant difference in the degree of physical disability post‐LT between those transplanted in the ICU (AM‐PAC 6cMB < 21 and JH‐HLM < 7, respectively) versus those who were waiting at home prior to LT. This level of disability was associated with a higher rate of discharge to IPR (42%) in patients transplanted out of the ICU. Sixty eight percent of ambulatory patients before LT had a functional level after LT that allowed them to be partially or fully independent upon home discharge (AM‐PAC 6cBM > 18 and JH‐HLM > 7).

The mean ± SD ICU stay for the 28 ICU transplanted patients was 74.68 ± 42.38 days, and their hospital LOS was 94.46 ± 54.07 days. These durations were significantly longer compared to ambulatory patients (p 0.01).

The mean ± SD TPA in the entire cohort was 16.9 ± 6.04 cm². The value of TPA was further classified into quartiles (<11.98 cm², 11.98–16.10 cm², 16.10–21.22 cm² and >21.22 cm²) for the entire cohort. Based on our definition of SPM, with TPA values below the first quartile, we found SPM in 32 patients for a prevalence of sarcopenia of 24% in the entire cohort of transplanted patients. There was no difference in TPA values when comparing underlying LT indication (Table S1).

The TPA mean ± SD for females was 11.7 ± 2.82 versus 19.7 ± 5.42 cm² for males; the PsI was 4.39 ± 1.06 for females versus 6.18 ± 1.70 cm²/m² for males (p < 0.001). Given the notable difference in muscle mass between sexes and based on TPA/PsI values falling below the sex‐specific first quartile, we identified SPM in 21/87 (24.1%) males versus 11/45 (24.4%) females, p = 0.099 (Table S2).

TPA > 21.22 cm² (fourth quartile) was more commonly present in males than females (38% vs. 0%, p < 0.001). There was no difference in BMI between males, females, or in patients with SPM.

SPM based on TPA less than the first quartile in the sarcopenic range (<11.98 cm²) was associated with increased ICU and hospital LOS when compared to TPA > 21.22 cm² (fourth quartile). In addition, TPA > 21.22 cm² was associated with shorter LOS compared to lower TPAs in the first to third quartiles (LS‐mean 32.10, p < 0.01) and ICU stay (LS‐mean 22.83, p < 0.01) (Table 2).

TABLE 2.

Association of albumin and TPA with length of stay.

	Estimate	Standard error	t ratio	p value
Hospital length of stay
TPA	−2.93	0.86	−3.38	0.001
Albumin	−33.2	6.41	−5.19	<.0001
ICU length of stay
TPA	−2.96	0.91	−3.26	0.001
Albumin	−32	6.81	−4.69	<.0001

Abbreviation: TPA, total psoas area (cm²).

Tracheostomy was performed in 56% of patients with SPM versus 36% of non‐sarcopenic patients (p = 0.043).

Sarcopenia was not associated with physical function prior to hospital discharge (independent 14% vs. 26%, and fully dependent 14% vs. 5%, p = 0.087) or increase in IPR discharges (33% vs. 40%, p = 0.292) when compared to patients with TPA in higher quartiles.

The PsD, a measurement of muscle quality (mean ± SD 36.64 ± 13.82 Hounsefield units), was no different among patients in the ICU and those ambulatory prior to LT. As the quality of the psoas muscles, described as PsD, was not different in our patients, we looked at a subgroup of patients with BMI > 30 kg/m² and TPA less than the first quartile (n = 6). In this subgroup analysis, we observed worse overall survival in obese sarcopenic patients, an association that was statistically significant (hazard ratio, 4.5, 95% CI 1.69–12.07).

Hypoalbuminemia was observed in 81 (61%) of the patients. The presence of hypoalbuminemia was not different among males and females, underlying diagnosis (Table S3), ICU or ambulatory patients. Preoperative serum albumin >3.5 g/dL had significantly shorter LOS (least square‐mean 36.07 vs. 51.51 days, p < 0.05) or ICU stay (25.86 vs. 41.67, p < 0.05). Serum albumin and TPA were weakly correlated (0.25, p = 0.001). On univariate analysis, patients discharged alive after LT had higher TPA (17.5 ± 6.1 vs. 14.6 ± 5.2 cm², p = 0.02). On multivariate analyses, only albumin and male sex remained significantly associated with hospital and long‐term survival (Table 3).

TABLE 3.

Cox proportional hazards model. Survival analyses.

	HR	95% CI	p value
Hospital survival
Male sex	0.24	0.07–0.8	0.02
Albumin	0.52	0.27–0.97	0.03
TPA	1.03	0.91–1.15	0.57
Long‐term survival
Male sex	0.24	0.07–0.8	0.02
Albumin	0.53	0.28–0.96	0.04
TPA	1.03	0.91–1.15	0.57

Abbreviations: HR, hazard ratio; TPA, total psoas area (cm²).

The median follow‐up of patients that survived hospital discharge post‐LT was 3.3 ± 2.2 years. For 108 patients discharged alive post‐LT and with complete follow‐up data, the 1‐year and 3‐years survival post‐LT were 88% and 72%, respectively. Fifty three patients died after LT (53/132, 40%) by the time of data censoring. The overall mortality at last follow‐up was lower among ambulatory patients that waited at home prior to LT when compared to patients transplanted out of the ICU (hazard ratio = 0.41, 95% CI 0.19‐0.86). Out of 106 patients with complete lung function data beyond 1‐year post‐LT, 2 patients with bronchostenosis were not included in the analysis of CLAD. Thirty‐eight patients (34%) developed any grade of CLAD, which was diagnosed at 2.89 ± 1.85 years post‐LT. There was no association between gender, sarcopenia or albumin level with CLAD diagnosis.

In this cohort, the mean ± SD for the number of TBBx post‐LT was 2.43 ± 2.47. Notably, patients with SPM underwent less TBBx compared to non‐sarcopenic patients (2.68 ± 2.51 vs. 1.75 ± 2.38, p = 0.37), but this difference was not statistically significant.

Pulmonary function tests as measured by FVC and FEV1 were better in patients with higher TPAs (Table 4). Patients with SPM had a lower lung function as determined by best post‐LT FEV1 (2.06 ± 0.70 vs 2.70 ± 0.87, p = 0.001) and lower FEV1 at last clinic follow‐up (1.58 ± 0.79 vs 2.16 ± 0.94, p = 0.007) when compared to patients with higher TPA above the first quartile. On the other hand, there was no association between albumin levels and post‐transplant pulmonary function tests.

TABLE 4.

Association of albumin and TPA with post‐transplant FEV1 and FVC.

	Estimate	Std. error	t ratio	p value
FEV1
TPA	0.05	0.01	4.36	<0.001
Albumin	0.02	0.11	0.22	0.82
FVC
TPA	0.07	0.01	5.08	<0.001
Albumin	0.13	0.13	1.02	0.31

Abbreviations: FEV1, forced expiratory volume in 1 s; FVC, forced vital capacity; TPA, total psoas area (cm²).

Out of 15 ICU patients who did not undergo LT, 12 died after being delisted, all of whom were on ECMO. These patients had lower mean ± SD values for TPA and PsI (14.53 ± 5.23 cm² and 5.11 ± 1.71 cm²/m², respectively), but the differences were not statistically significant (p = 0.13 for TPA and p = 0.33 for PsI) compared to those who underwent LT or survived without LT. Three patients with COVID‐ARDS on venovenous ECMO recovered and were removed from the LT list, surviving to hospital discharge without LT.

4. Discussion

The psoas muscles are the main flexors of the hips and provide postural support of the lumbar spine and sacroiliac and hip joints. Weakness in these muscles can impair balance and mobility. Few studies have detailed SPM using CT scan measurements of the CSA of the psoas muscles and its association with outcomes following LT [10, 11, 12, 13, 14, 15, 16]. Various authors have proposed different cutoffs for TPA and PsI. In patients that underwent LT, Hsu et al. [12] described PsI mean values for males 7.8 and 6.4 cm²/m² for females, while Weig et al. found the mean lean psoas area to be 22.3 ± 8.3 cm² [13]. Bahat et al. [19] examined liver transplant patients and found that PsI was 5.4 ± 4.64 in males and 3.56 ± 2.66 cm²/m² in females.

We observed TPA and PsI values that were lower than the ones described by other authors, likely due to differences in race and age. Of note, our cohort was older than the patients in the studies by Bahat et al. (32.5 ± 9 years old), Weig et al. (49 ± 13 years old), and Hsu et al. (50.1 ± 15.2 years old).

Since there is no established definition for SPM and previous studies have used varying criteria for sarcopenia, we examined the effects of defining SPM as TPA below the first quartile—an approach previously used by others—to better represent extremely low muscle mass. Using this definition, we observed a sarcopenia prevalence of 24%, which is lower than previously reported estimates of 30%–72% [3, 8, 9].

Prior studies [11, 12] showed that SPM was associated with lower survival post‐LT, lower pulmonary function, and graft failure. In those studies, SPM is defined as any PsI value under the mean or under the first quartile. In our study, we did not find an association between SPM, CLAD, and mortality after adjusting for multiple variables.

In our study, we did not see a difference in survival or development of CLAD between SPM and non‐sarcopenic patients over a follow‐up period of 3.3 ± 2.2 years. Despite the difference in demographics between patients included in the different studies that could have negatively affected our cohort of older patients, the presence of SPM did not directly affect our patients’ survival or the development of CLAD.

The differences in the demographics, definition of SPM, and type of transplant in the different single‐center studies are worth mentioning, as they could explain in part the difference in reported outcomes. For instance, 95% of our patients underwent bilateral LT compared to 73.7% in Hsu et al.’s study [12]. Females represented 34% of our cohort versus 49.5% in their study.

Higher morbidity based on LOS and tracheostomy rate was observed in patients with SPM, similar to previous reports [11, 12, 13, 14] but no association with mortality was seen [12, 14].

The lack of association of SPM with post‐LT survival is consistent with the findings of Rozenberg et al. [7], who reported in their systematic review that most included studies did not show a significant link between post‐LT survival and the variables examined.

Although pre‐LT frailty is a strong risk factor for poor outcomes [4], sarcopenia is just one factor contributing to frailty [4, 5, 10]. Swaminathan et al. found poor correlation among frailty measurements, including thoracic sarcopenia (paraspinal muscles, including the trapezius at the level of the inferior end plate of the sixth thoracic vertebrae), and no significant association with clinical outcomes [10].

Frailty is better evaluated by utilizing multiple frailty constructs, as different tests predict different outcomes, as described by the Lung Transplant Frailty Scale developed by Singer et al. [4], and supported by the findings in the study by Swaminathan et al. [10].

Besides the differences in baseline characteristics of the patients studied by other centers, the variable effect of sarcopenia on outcomes post‐LT is likely due to the different muscle groups used to define sarcopenia by CT scan and the different cutoff values used, as well as transplant center practices. Most of our patients underwent bilateral LT. While single LT is thought to result in lower morbidity, a recent study using data from the Scientific Registry of Transplant Recipients (involving around 10 000 LT recipients aged 65 years and older) found no significant difference in short‐term survival rates between single and bilateral LT groups. However, bilateral LT recipients had lower graft rejection rates and better 5‐year survival outcomes [23].

We observed lower lung function in patients with lower TPAs, including those with SPM, as measured by various lung function tests post‐LT. Patients with SPM had less objective assessment of allograft function due to factors like tracheostomy status, muscle weakness, and fatigue. Bronchoscopy was often precluded by logistical challenges related to anticoagulation and concerns about procedural complications.

Hypoalbuminemia prior to LT (<3.5 g/dL) has been associated with increased morbidity after LT, which we also observed in our study. This effect, compared to sarcopenia, seems to be better established [20, 21, 24]. A prior study by Halpern et al. [17] did not find any association between sarcopenia and hypoalbuminemia, and notably, in our study, we only found a weak correlation between the two.

In our cohort, ICU hospitalization prior to LT was associated with longer ICU and total hospital LOS post‐LT, worse physical function, greater tracheostomy rates, and a higher number of patients discharged to IPR (42% of patients). These findings are concordant with a United Network for Organ Sharing (UNOS) database analysis of more than 7000 LT recipients showing that preoperative ICU status was the strongest predictor of poor postoperative function and death at 1 year [25]. In addition, Zhang et al. [14] performed a systematic review and meta‐analysis of ICU patients with sarcopenia, using CT scans of different muscle groups, finding an increased risk of mortality in critically ill patients.

Studies by Hsu et al. (n = 44) and Weig et al. (n = 18) showed varying prevalence of sarcopenia in ICU patients [12, 13, 15]. In our study, the prevalence of SPM was not different between ICU patients and ambulatory patients prior to LT. Those who did not undergo LT had lower TPA and PsI values, but these differences were not statistically significant.

Low BMI can over‐ or under‐estimate muscle mass, as it does not measure body composition [26, 27, 28]. Sarcopenia can be present in normal‐weight and overweight patients, making low BMI an inappropriate parameter to describe sarcopenia and frailty [15, 24, 27], although sarcopenia was more prevalent in lower BMI patients in several studies [15, 29].

A study analyzing the Scientific Registry of Transplant Recipients database found increased mortality risk at BMI ≤ 24 and ≥28 [28]. Another study found that cystic fibrosis patients who were underweight (BMI < 17 kg/m²) had post‐LT survival rates comparable to other groups [27]. Ezponda et al. demonstrated that lower psoas muscle density on CT scan (a marker of poor muscle quality) was independently associated with long‐term mortality, but in our patients, psoas muscle density did not impact clinical outcomes [22]. In our study, many patients had intravenous contrast‐enhanced CT scans, which are known to affect muscle density depending on phase, age, and sex of the patient.

We did an exploratory analysis in a subgroup of patients with BMI > 30 kg/m² and TPA less than the first quartile (n = 6). The analysis of the combined effect in this subgroup showed a 4.5 increase in overall mortality in recipients of LT when their BMI > 30 kg/m² and their TPA was less than the first quartile (sarcopenic obese) when compared to patients without this combined diagnosis. A similar finding by Suh et al. described the presence of sarcopenia in overweight patients (BMI > 23 kg/m² or “sarcopenic overweight patients”) and its association with lower overall survival based on a volumetric assessment of the psoas muscles using 3‐dimensional CT scans [15].

Outpatient rehabilitation programs, including supervised exercise training, improve limb muscle dysfunction, exercise capacity, and quality of life before and after LT [30, 31]. A study of sarcopenia post‐LT showed significant improvement in muscle mass area by 36 months post‐LT, with negligible differences in muscle strength and physical performance compared to non‐sarcopenic patients [31].

Considering the lower‐than‐recommended number of allograft function assessments based on our protocol, noninvasive or less invasive diagnostic tests could enhance the detection of acute rejection or CLAD. While their clinical utility is still under evaluation, these tests show promise in guiding management based on clinical findings and suspicion of rejection versus infection [32, 33, 34].

4.1. Study Limitations

This retrospective single‐center study had a lower prevalence of sarcopenia compared to other studies. Using a different definition of SPM could have increased the prevalence of SPM. Muscle CSA was assessed using CT studies within 6 months perioperative, with 40% performed before LT or decision not to transplant, and the rest after LT. In 60 patients, a contrast‐enhanced CT scan was evaluated as it was the nearest test to the time of LT. When evaluating psoas muscle density, it is important to mention the influence of age, sex, and the phase of the contrast on the Hounsfield unit, which we did not analyze.The use of contrast has little or no effect on the muscle mass as measured by the CSA [35, 36]. Regarding the method used for evaluation of SPM with CT scan, we did not conduct a reliability analysis, but others have reported excellent inter‐rater agreement [11, 13, 25].

Given the limited number of patients diagnosed with sarcopenia in our cohort, we did not conduct a sensitivity analysis. A subgroup analyses of the effect of TPA based on underlying lung disease was not performed given the small number of patients in some groups.

Additional multicenter studies with a larger number of patients that assess sarcopenia with standardization of measurement techniques and incorporation of clinical outcomes are needed. It is possible that a study of a higher number of sarcopenic patients will show the effect of SPM on mortality. We excluded from the mortality analysis 12 critically ill patients with lower TPA who were not transplanted and subsequently expired.

Despite these limitations, our study is one of the largest studies describing the impact of serum albumin and TPA in LT recipients and includes a higher number of ICU patients, with 28 patients, 20 of them on ECMO.

5. Conclusion

In the present study of LT recipients, TPA and albumin levels were associated with length of hospitalization. On multivariate analyses, albumin remained independently associated with survival. TPA, but not albumin, was associated with post‐transplant pulmonary function tests. Based on our findings, sarcopenia alone should not be considered a contraindication for LT, and additional clinical factors—such as hypoalbuminemia—should be included in the prediction of post‐transplant mortality.

Author Contributions

D.J.F.‐P., C.R.F.‐P., S.C., and T.S. conceived the study. D.J.F.‐P., H.B., R.L., A.M., S.C., and T.S. performed data extraction. D.J.F.‐P., C.R.F.‐P., L.S., S.C., K.O., J.S., and T.S. drafted the manuscript. C.R.F.‐P., M. L., and Y.W. performed statistical analysis. All authors revised the manuscript for intellectual content, discussed the results, and contributed substantially to the interpretation and analysis of data. All authors approved the final version of the manuscript.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting Table 1: TPA values by underlying diagnosis.

Supporting Table 2: Quartile values of TSA and PSI by gender.

Supporting Table 3: Albumin values by underlying diagnosis.

Franco‐Palacios D. J., Franco‐Palacios C. R., Crowley S., et al. “Effect of Total Psoas Muscle Area and Serum Albumin on Outcomes After Lung Transplantation.” Clinical Transplantation 39, no. 9 (2025): e70308. 10.1111/ctr.70308

Data Availability Statement

The datasets generated and/or analyzed during the current study are available from the corresponding author on reasonable request.