Role of social status and social environment on net survival in patients with chronic lymphocytic leukemia: A high-resolution population-based study

1Introduction

Chronic lymphocytic leukemia (CLL) is the most frequent hematological malignancy in France [1] . The median age at diagnosis was 71 years for men and 74 for women. The evolution is heterogeneous, with a chronic course; 70% of patients are asymptomatic at diagnosis with indolent lymphocytosis and do not need treatment. One-third of patients will never be treated. Therapeutic improvement in recent years has substantially changed CLL evolution, which is now a chronic disease with long-term treatment in half of patients and only close clinical monitoring and follow-up for the rest of the patients. The prognostic role of genetic factors 234 and environmental exposure is known [56] , but the impact of socioeconomic status on survival in CLL is unclear.

The pejorative role of social inequalities on cancer screening [7] , incidence [8] and survival has already been established in several studies in patients with solid cancer 91011 . The role of social inequalities in survival in hematological malignancies has been less studied but seems to be less important 10111213141516 .

Social inequalities in cancer survival have been investigated with several indicators at the individual level, including socioprofessional category (SCP), employment status, health care insurance level, educational level, income, access to water or electricity, etc. [1014] Social inequalities can also be investigated at a contextual level, for example, with aggregated ecological indicators such as the European deprivation index (EDI) [17] . The EDI is an ecological index measuring relative poverty in a small geographical area. Individual and environmental factors bring complementary but nonidentical information. Socioeconomic status is multidimensional, and individual factors bring great information, but the importance of the environment (such as residency and workplace) is lacking, which is known as ecological bias. Individual data are also harder to collect. However, ecological factors include information about environmental exposition and permit an evaluation in the absence of individual factors but remain incomplete. Ecologic factors alone cannot make a difference between the individual socioeconomic status and geographical socioeconomic status, even when the studied geographic zone is small.

Worldwide, the role of socioeconomic status on survival in CLL has been poorly studied. No effect of social status on CLL incidence was detected in the United States of America [18] or in Sweden [19] . In this last study, the criteria for social status was SPC. Inequality in access to health care services for CLL patients has already been suggested in France [20] . The impact of social inequalities on CLL prognosis is still unknown. A recent publication in France suggested a lower net survival in women with CLL living in the most deprived areas than in women living in the least deprived areas, but no effect was highlighted in men. However, this study was based only on nonparametric methods and did not account for clinical factors (stage, treatments, etc.) [21] .

In France, the cancer register network ( Réseau français des registres du cancer , France) permits the collection of high-quality real-life data about cancer in the whole population. The main strengths of the registries are the large population coverage, the data certification and the exhaustivity of diagnosis in a selected area. These tools allow a population-based analysis ideal for survival studies.

Survival studies are generally based on overall survival. The concept of net survival and excess mortality allows an estimation of survival related only to the studied disease, with the theoretical hypothesis that the patient dies only from the studied disease (independently from other possible causes of death). The cause of death is not needed. The excess mortality is estimated as the difference between “observed mortality” in registries and “expected mortality” from lifetables. Lifetables contain the mortality of the general population (here in France) by age, sex, department (administrative unit), and year [2223] . Net survival can be useful for diseases such as CLL, with an indolent course and an elderly population, where overall survival can be affected by comorbidities.

The objective of this study was to evaluate the influence of individual social status (with SPC) and social environment (with EDI) on net survival in patients with CLL, including a high-resolution population-based study in France.