A CS including sarcopenia and adipopenia, assessed by a single CT scan slice, predicts outcome independent of BMI and the IPI.

Not available.

No abstract

No abstract

Background Cancer Cachexia is a metabolic syndrome that can be present even in absence of weight loss and associated with significantly impaired survival. Muscle wasting represents a key-symptom of this syndrome and we recently demonstrated the strong prognosis impact of sarcopenia assessed by computed tomography (CT)-scan in diffuse large B-cell lymphoma (DLBCL) (Lanic et al. Leukemia & Lymphoma 2013). Conversely, the clinical relevance of loss of fat mass (adipopenia) remains unclear. The aim of this study was (i) to investigate the prognostic impact of a multidimensional tool combining a nutritional parameter (albuminemia) and body composition measurements (skeletal muscle and body fat composition) in elderly patients with DLBCL treated by chemotherapy and rituximab (R) (ii) to document the evolution of sarcopenia after immunochemotherapy. Methods This retrospective analysis included 80 DLBCL patients older than 70 years (y) and treated by R-CHOP or R-miniCHOP. Skeletal muscle (SM), visceral (V) and subcutaneal (S) adipose (A) tissues were measured by analysis of stored CT images at the Lumbar vertebrae 3 (L3) level. The surface of the muscular and adipose tissues was selected according to CT Hounsfield unit. Values were normalized for stature to calculate the L3 SM index (LSMI, in cm2/m2), the LVAI and the LSAI and used to define sarcopenia and visceral/subcutaneal adipopenia. Results The characteristics of the patients were as follows: median age = 78 y [70-95]; 36 males; IPI 0-2 = 22, 3-5 = 58; treatment by R-CHOP (n = 45) or R-miniCHOP (n = 35); median body mass index (BMI; in kg/m2) = 23.9. According to the sex-specific defined cut-offs for LSMI (< 55.8 cm²/m² for men and 38.9 cm²/m² for women), 44 DLBCL patients (55 %, 23 males) were considered as sarcopenic. With a median follow-up of 39 months, the 2y overall survival (OS) in the sarcopenic population was 46% as compared to 84% in the non-sarcopenic group (HR = 3.12; CI95%, 1.66-5.88; p=0.0004). The median LSAI was 76.3 cm2/m2 [10-167] in females and 47.4 cm2/m2[22-100] in males. The median LVSAI was 43.5 cm2/m2[3-141] in females and 50.4 cm2/m2[14-159] in males. Adipopenia, defined by a low LVAI and/or a low LSAI was also highly predictive of the outcome. The 2y OS of the low LVAI population was 48% as compared to 82% for the non-adipopenic group (HR = 2.20; CI95%, 1.19-4.05; p=0.01). The 2y OS in the low LSAI population was 48% as compared to 80% in the non-adipopenic group (HR = 2.28; CI95%, 1.23-4.21; p=0.008). A Three-point cachexia score (CS) including adipopenia, sarcopenia and hypo-albuminemia (defined by an albuminemia < 35 g/L) was build and delineates three distinct risk-groups (Figure 1). More importantly the CS remains predictive of the prognosis in a multivariate analysis including BMI (< or >= 25 kg/m2), age (< or >= 80y), IPI and gender (HR=2.5; CI95%= 1.14-5.39; p =0.02). LMSI was subsequently reassessed in thirty seven patients during the routine CT scan follow-up [mean = 10 months after pre-treatment CT scan (range 2.8-19.2)]. 15 (40%) patients displayed a 5% decrease of their LSMI, whereas 13 (35%) and 9 (25%) displayed no significant change or increase (>5%) of the LMSI respectively. Conclusion Our study demonstrates that sarcopenia and adipopenia estimated by CT-scan define cachexia more accurately than BMI or weight loss in elderly DLBCL patients. These factors can be integrated in a cachexia scoring tool which predicts the outcome independently of the BMI and of the IPI. CT scan follow-up indicates that cachexia is a reversible process that should be integrated as part of the therapeutic target in combination with lymphoma treatment. A prospective multicentric trial (registered as NCT01715961/Clinical.gov) is ongoing to validate these anthropometric and nutritional parameters and compare to geriatric assessment scales. Disclosures: No relevant conflicts of interest to declare.