Abstract: PB1708
Type: Publication Only
Background
Current prognostic scores are not sufficient to define high risk patients with diffuse large B cell lymphoma (DLBCL). Besides parameters included in the International Prognostic Index (IPI), other clinical and laboratory parameters have been investigated as potential prognostic markers. However, contradictory data have been reported.
Aims
The aim of this study was to evaluate prognostic significance of clinical and laboratory parameters on the overall survival (OS) of patients with DLBCL.
Methods
A total of 393 patients (188 females/205 males) with the median age of 60 years (range 18-84) were analyzed. All patients were initially treated with rituximab plus CHOP (Cyclophosphamide, Doxorubicine, Vincristine, Prednisone) or CHOP like protocols.
Results
Ann Arbor stage I, II, III and IV had 56 patients (14.2%), 142 (36.1%), 71 (18.1%) and 124 (31.6%), respectively. Bulky disease had 99 patients (25.2%), B symptoms 263 patients (66.9%), and poor performance status according to the European Cooperative Oncology Group (ECOG) ≥2 had 82 (20.9%). Bone marrow involvement was present in 68 patients (17.3%). Low IPI risk was present in 194 patients (49.4%), low intermediate in 86 (21.9%), high intermediate in 77 (19.6%), and high in 36 (9.2%). Median absolute lymphocyte count (ALC) at diagnosis was 1.35x109/L (range 0.07-60.76x109/l), absolute monocyte count (AMC) was 0.64x109/l (range 0.06-8.58x109/l), ALC/AMC was 2.3 (range 0.07-37.0x109/l), hemoglobin level was 125g/l (range 57-421g/l), platelet level was 274x109/l (range 50-584x109/l), C-reactive protein was 10.2 mg/l (range 0.10-438mg/l), erythrocyte sedimentation rate (ESR) was 30mm/h (range 2-636mm/h), and albumin level was 38g/l (range 20-51g/l).
Complete remission (CR) was achieved in 288 patients (73.3%), partial remission (PR) in 58 (14.8%), stable disease (SD) in 5 (1.3%) and progressive disease in 42 (10.7%). Disease relapse was confirmed in 59/346 patients (17.0%). OS was influenced by the presence of B symptoms (p=0.004, 95% CI 1.263-3.549), ECOG≥2 (p<0.0001, 95% CI, 1.827-4.290), Ann Arbor clinical stage (p<0.0001, 95% CI 1.601-3.883), and albumin level (p<0.0001, 95% CI 0.905-0.953). Optimal cut off point for albumin level was 34g/l, and was determined by Receiver operating characteristic (ROC) curve (AUC 0.699, 95% CI, 0.629-0.770, p<0.0001). The prognostic value of IPI was highly statistically significant for OS (p<0.0001, 95% CI, 1.545-2.236). However, other analyzed parameters did not influence OS. Multivariate analysis among significant parameters (presence of B symptoms, IPI, and albumin), has pointed to IPI (HR 1.81, p<0.0001, 95% CI, 1.489-2.222), and albumin level (HR 1.77, 95% CI, 1.164-2.69, p=0.008) as the most important parameters that influenced survival.
Conclusion
Although IPI is widely used as a prognostic index in DLBCL, it cannot fully recognize high-risk patients. Pretreatment albumin level may represent a useful tool in order to discriminate high-risk patients and is likely to add significant information to the IPI.
Session topic: 20. Aggressive Non-Hodgkin lymphoma - Clinical
