Abstract
Abstract 4587
Chronic Lymphocytic Leukemia (CLL) is the most common lymphoprolipherative disorder of the elderly population in Western countries. It shows a highly variable clinical course. In fact, some patients may die within few months from the diagnosis because of CLL itself or disease-related complications. Other patients do not require any treatment for many years and have a long-standing disease. Many of them could die because of disease different from CLL. The identification of subgroups of patients with peculiar features predictive of the clinical behaviour of the disease is important. This retrospective analysis has the purpose to study patients affected by CLL, diagnosed and followed at our single centre of Haematology, focusing our attention on their causes of death.
We selected 340 patients affected by CLL from our data-base, diagnosed from January 1999 to December 2010 and followed until March 2012. We distinguished the causes of death in two groups: one related to CLL (as progression of the disease, evolution to Richter's Syndrome, infections due to chemotherapy) and the other not related to CLL (i.e. cardiovascular diseases, solid tumours, old age). Statistical analysis, conducted using SPSS version 16.0 for Windows and “GraphPad Prism” GraphPad Software Inc., compared these two groups and tried to select other subgroups.
We recorded 69 deaths: 47 related to CLL (68.1%) and 22 unrelated to CLL (31.9%). The median age of death of our cohort of patients was 76 years (range 40–92); those patients with a CLL-related death (related pts) died at a median age of 76 years (range 40–89) and the patients with a CLL-unrelated death (unrelated pts) died at a median age of 76 years (range 61–92). No differences in terms of median age of death were found analysing the data by gender.
Also, considering the overall survival time from diagnosis to death, it was 58 months (range 9–155) in the related group and 43 months (range 14–121) in the unrelated group (p=0.185).
When divided our population by the disease behaviour, we obtained 3 subgroups of patients: patients who progressed and died for CLL related causes (group A), patients who progressed and died for CLL unrelated causes (group B) and patients who did not progress and died for CLL unrelated causes (group C) (Table 1). The only statistical significant difference was found among the median overall survival times in un-progressive patients who died for non CLL related causes (p=0.043).
The median age of death was not affected by the cause of death in our CLL population. Moreover, patients with un-progressive CLL showed an overall survival time shorter than the progressive CLL subgroups. Patients with un-progressive CLL probably had a shorter survival because unrelated CLL diseases could are more difficult aggressive than CLL itself.
Table 1. group A group B group C p N° patients 47 13 9 Median age at diagnosis (years) (range) 70 (39–85) 71 (56–83) 77 (67–86) n.s. Median age at death (years) (range) 76 (40–89) 75 (61–84) 79 (68–92) n.s. Overall survival time (months) (range) 58 (9–155) 48 (18–121) 22 (14–78) 0.043
Disclosures:
No relevant conflicts of interest to declare.
Machine Learning and Radiomic Features to Predict Overall Survival Time for Glioblastoma Patients