ABSTRACT

Aim

This study aimed to evaluate the impact of sarcopenia and myosteatosis on progression-free survival (PFS) in patients with aggressive non-Hodgkin lymphoma (NHL).

Methods

A retrospective analysis was conducted on 33 high-grade NHL patients diagnosed between 01.01.2018 and 31.12.2019, who had pre- and post-treatment ¹⁸F-fluorodeoxyglucose positron emission tomography/computed tomography (CT) imaging. Body composition was semiautomatically measured using axial CT slices at the lumbar L3 vertebral level, focusing on the psoas muscle (PM). The right and left average hounsfield unit (HU) values [right psoas HU corrected (RPHUc) and left psoas HU corrected (LPHUc)] were recorded as myosteatosis parameters. The cross-sectional area of the PM was adjusted for body size by dividing it by the square of the patient’s height in meters, resulting in the PM index, expressed in cm²/m².

Results

The average age of the patients was 57±16.2 years, and sarcopenia was identified in 39.4% of the cohort. During follow-up, relapse occurred in 9 patients (27.3%). A significant association was found between relapse and age (p=0.002), Eastern Cooperative Oncology Group (ECOG) performance status (p=0.003), and RPHUc (p=0.015). Receiver operating characteristic analysis for RPHUc (cut-off >33.550) in predicting PFS showed an area under the curve of 0.778 (p=0.015), with 66.7% sensitivity and 33% specificity. Univariate analysis identified age (p=0.001), ECOG score (p=0.000), and RPHUc (p=0.017) as significant prognostic factors for PFS. In multivariate analysis, only age remained an independent prognostic factor (p=0.04).

Conclusion

Our study demonstrated that age and RPHUc values have prognostic significance for PFS in aggressive lymphoma patients. These parameters, easily obtainable from routine imaging, may aid in guiding clinical management strategies.

Introduction

Aggressive lymphomas are fast-growing subtypes of non-Hodgkin lymphoma (NHL) characterized by high proliferation rates. The most common type is diffuse large B-cell lymphoma (DLBCL). Others include mantle cell lymphoma (MCL), Burkitt lymphoma, high-grade B-cell lymphomas, primary mediastinal large B-cell lymphoma, and peripheral T-cell lymphomas [1]. Treatment typically involves intensive chemotherapy, targeted therapies, immunotherapy, and, in some cases, stem cell transplantation [2]. The heterogeneity in DLBCL’s immunophenotype, genetic profile, and histology influences treatment response and long-term prognosis [3, 4]. Although the MCL International Prognostic Index (IPI) has been introduced as a tool for risk stratification, its prognostic utility continues to be a subject of discussion [5, 6]. Identifying new prognostic factors is crucial for disease management and survival. Established indices like the IPI, revised IPI, and National Comprehensive Cancer Network (NCCN)-IPI underscore the need for novel biomarkers to predict aggressive disease courses.

Sarcopenia, defined as reduced skeletal muscle mass, quality, and function, is often age-related but may also result from an underlying disease or treatment side effects [7]. Studies in various malignancies suggest sarcopenia adversely impacts survival [8]. Current evidence supports low skeletal muscle mass as a prognostic biomarker in cancer patients [9], including hematologic malignancies [10]. Sarcopenia has been associated with an increased risk of adverse outcomes, including falls, bone fractures, functional impairment, and overall mortality [7, 11]. Cancer patients may lose 15-50% of skeletal muscle mass, reducing chemotherapy tolerance and quality of life [12-14]. In recent years, it has also garnered attention as a potential prognostic marker in NHL [11, 15].

This study aimed to expand the current literature by evaluating the prognostic significance of both sarcopenia and myosteatosis in patients with high-grade lymphoma.

Methods

This study was designed as a retrospective analysis of patients diagnosed with high-grade lymphoma between January 1, 2018, and December 31, 2022. Patients who had pre-treatment, post-treatment, and relapse, ¹⁸F-fluorodeoxyglucose (FDG) positron emission tomography/computed tomography (PET/CT) imaging available were included in the study. Patients with missing data, incomplete treatment records, or insufficient follow-up information were excluded from the analysis. Patient data, including age, sex, disease stage, ECOG performance score, lactate dehydrogenase levels, sedimentation rate, treatment regimens, relapse dates, last follow-up dates, and other relevant clinical findings, were recorded. Body composition parameters were obtained from the CT component of each patient’s ¹⁸F-FDG PET/CT scans. The study was conducted according to the principles of the Declaration of Helsinki, and approval was obtained from the University of Health Sciences Türkiye, Gazi Yaşargil Training and Research Hospital (decision no: 345, date: 07.02.2025).

Body Composition Assessment

All ¹⁸F-FDG PET/CT images were evaluated using the AW 4.7 workstation (Advantage Workstation software version 4.7; GE Healthcare, Milwaukee, WI, USA). The body composition of the psoas muscle (PM) was semiautomatically measured on axial CT slices at the level of the L3 lumbar vertebra from pre-treatment, post-treatment, and relapse ¹⁸F-FDG PET/CT scans. The hounsfield unit (HU) thresholds for PM were set between -29 and +150 HU [16]. Separate regions of interest were manually drawn around the right and left PM, avoiding bone and adipose tissue, and the average HU values [right psoas hounsfield unit corrected (RPHUc) and left psoas hounsfield unit corrected (LPHUc)] were recorded as myosteatosis parameters. To calculate the PM index (PMI), the PM area was normalized by dividing it by the square of the patient’s height in meters (m²) (Figure 1) [17, 18]. Sarcopenia was defined as PMI ≤5.1 cm²/m² in men and ≤43 cm²/m² in women [19].

Statistical Analysis

Statistical analysis was performed using Statistical Package for the Social Sciences 25.0 (IBM Corporation, Armonk, New York, United States). The normality of continuous variables was assessed using the Kolmogorov-Smirnov test. The Mann-Whitney U test was used to compare quantitative variables between two independent groups. The Kaplan-Meier (product-limit method) and log-rank (Mantel-Cox) tests were used to evaluate the impact of factors on progression-free survival (PFS). Cox regression analysis was applied to assess the prognostic effects of variables on PFS, with significant independent variables entered into the model both as single (individually) and multiple (collectively) variables. The relationship between predicted classification based on calculated cut-off values and actual classification was evaluated using sensitivity and specificity rates derived from receiver operating characteristic curve analysis. Variables were analyzed at a 95% confidence level, with a p<0.05 considered statistically significant.

Results

The study included 13 female and 20 male patients, with a mean age of 57±16.2 years (range: 20-85). Disease staging revealed one patient with stage 1, 5 with stage 2, 10 with stage 3, and 17 with stage 4 disease. ECOG performance status was 1 in 22 patients, 2 in 10 patients, and 3 in 1 patient. The median right psoas average HU (RPHUc) was 36.5±10.9. During follow-up, relapse occurred in 9 patients (27.3%), and one patient (3.03%) died. The median lactate dehydrogenase level prior to treatment initiation was 296.0 U/L, with values ranging from 139 to 4900 U/L. A statistically significant association was found between relapse and age (p=0.002), ECOG performance status (p=0.003), and RPHUc (p=0.015). Relapse was more frequently observed in patients of advanced age, with compromised ECOG performance status and reduced RPHUc levels. Among the 9 relapsed patients, 5 had sarcopenia in the PM, but no statistically significant relationship was found between relapse and sarcopenia. Other descriptive parameters are presented in Table 1. An RPHUc cut-off value of >33.55 predicted better PFS (area under the curve: 0.778, p=0.015) with 66.7% sensitivity, and 33.0% specificity (Figure 2). For RPHUc <33.550, the median PFS was 76.9 months at 1 year, while it was 27.4 months at 4 years. For RPHUc >33.550, the median 1-year and 4-year PFS were both 88 months. The median PFS was 10.03 months (range: 5-48) in relapsed patients and 22.72 months (range: 5-63) in non-relapsed patients (Figure 3). Univariate Cox regression analysis identified age (p=0.001), ECOG performance score (p=0.000), and RPHUc (p=0.017) as significant prognostic factors for PFS (Table 2). In multivariate Cox regression analysis, only age remained an independent prognostic factor for PFS (p=0.04).

Discussion

Studies have demonstrated that NHL patients may lose up to 31% of their total body weight [20]. In advanced-stage cancers, including both solid tumors and hematologic malignancies, cancer cachexia affects approximately 60-80% of patients [7, 10]. This condition can lead to adverse clinical outcomes such as reduced tolerance to chemotherapy and diminished quality of life [12, 13].

Aging represents one of the most significant risk factors for sarcopenia development, contributing to progressive muscle mass loss. Additionally, various pro-inflammatory cytokines released by tumors, including interleukin-1, interleukin-6, tumor necrosis factor, and interferon gamma, accelerate muscle tissue catabolism [21]. Tumor-induced abnormalities in protein and amino acid metabolism, combined with malnutrition and reduced physical activity during treatment, result in more pronounced muscle mass reduction [22]. Cancer treatments may also induce loss of fat and bone mass alongside muscle depletion [23]. The age-dependent increase in sarcopenia prevalence has been well-documented [24, 25]. In our study of aggressive lymphomas, univariate analysis identified age as a prognostic factor for PFS (p=0.001), with multivariate analysis confirming its independent prognostic value (p=0.04). This finding contrasts with Albano et al. [26] study of older MCL patients (mean age 72.7±5.6 years), which found no significant association between age and PFS. Our study’s inclusion of relatively younger patients provides valuable insights into the independent effect of age on PFS in sarcopenic patients.

While one MCL study in elderly patients reported significantly higher sarcopenia prevalence in women (93% vs. 47%, p=0.001) [25], Xu et al. [27] DLBCL study found no gender-specific association regarding gender differences between sarcopenia and prognosis. The lack of significant gender-PFS correlation in our study may reflect our inclusion of various high-grade lymphoma subtypes and relatively younger patients.

Saglam et al. [28] study of 112 patients identified ECOG performance status as a prognostic factor for PFS in both univariate and multivariate analyses. While our univariate Cox regression confirmed this association (p=0.000), multivariate analysis did not. This discrepancy may result from our limited sample size, cohort differences, or potential confounding effects of other variables in the model.

In our study, sarcopenia was identified in 36% of patients, aligning with the findings of Xiao et al. [29], who reported a prevalence exceeding 30% based on pretreatment CT evaluations in individuals with DLBCL. However, we found no significant association between PMI and PFS (p=0.128).

Myosteatosis, defined as muscle weakening due to fat infiltration and measured by HU [30, 31], has been associated with worse OS in various malignancies including lymphoma [32]. In our cohort, PFS differed significantly based on RPHUc cut-off values: median PFS was 10.03 months in relapsed patients, versus 22.72 months in non-relapsed patients. These findings suggest that muscle quality rather than quantity may influence PFS. Univariate analysis confirmed RPHUc as a prognostic factor (p=0.017), aligning with reports of significantly worse PFS in patients with low skeletal muscle density (hazard ratio: 2.28, p=0.002) [33].

Sarcopenia assessment may prove particularly valuable when deciding between standard R-CHOP chemoimmunotherapy and dose-reduced regimens for elderly patients or those with poor performance status, comorbidities.

Study Limitations

One of the key strengths of this study is the use of an objective radiological parameter (RPHUc) to assess muscle quality and the investigation of its prognostic significance in a relatively younger and clinically heterogeneous high-grade lymphoma cohort. The limited number of prior publications exploring RPHUc in this setting enhances the novelty and potential clinical relevance of our results. Study limitations include the small sample size, retrospective design, lack of gender-specific analysis, and absence of DLBCL subtype and treatment toxicity data. Nevertheless, the study by Lanic et al. [34] reported no significant association between sarcopenia and the GCB or non-GCB subtypes.

Conclusion

Our findings suggest that both age and RPHUc are valuable prognostic indicators in high-grade lymphoma. This simple, routinely accessible imaging parameter may enhance risk stratification and guide individualized treatment approaches.