Background. Chronic Lymphocytic Leukemia is the most frequent leukemia in the Western countries (Rozman, 1995) but its incidence is not well established. Epidemiological data are based on reports from tumour registries, with an estimated incidence lower than 3-cases/100000 pts-year. Tumour registries can however underestimate the incidence of CLL as highlighted by the 37% higher incidence reported in a selected population of Veterans (Zent, 2000). We aimed to assessing the incidence of CLL in a population-based cohort from an epidemiological survey on familial thrombosis, (Tosetto, 2003) which was followed for at least 8 years.

Material and Methods. 15109 healthy people from 18 to 65 years of age in the Vicenza township were enrolled in the VITA Project (8017 females and 7092 males, with a median age of 43 years). A clinical history was collected and blood and DNA samples were withdrawn and stored at the time of enrolment. The study begun in 1993 and ended in 1996. Patients with a diagnosis of CLL (ICD-9 code 204) were traced using data from the Veneto Region Health Service database, which records all hospital admission for residents in the enrolment area. Moreover, the database encompasses data of all the subjects visited in the Vicenza Hematology Department, the unique hematological facility of the area, even if not subsequently admitted to the hospital. Diagnosis of CLL was validated if based on morphological and immunophenotypic data according to currently accepted criteria (Cheson, 1988; IWCC, 1989; Rozman, 1995).

Results. Six out 15109 people were diagnosed with a CLL (4 males). The median follow-up for the entire cohort was 8.2 years and the total time of exposure was 121000 pts-year. Incidence of CLL is estimated to be 4.9 cases/100000 pts-year (CI: 1.8–10.7). Diagnosis of CLL was done from 36 to 97 months after the enrolment. Mean age at diagnosis was 66 years (61 – 70). At the time of enrolment in the study, all of the patients had a normal WBC. Three persons presented an immunoglobulin VH gene rearrangement, 32, 54 and 57 months before diagnosis. All of the patients were asymptomatic at diagnosis. The median follow-up of all the subjects was 28 months (7 – 57 months). Only one of them became symptomatic and needed a treatment, 40 months after the diagnosis.

Conclusions. According to our data, the incidence of CLL appears to be higher than assessed so far. In our opinion it’s noteworthy that the estimation is based on data of subjects evaluated in a clinical setting and not from tumour or disease registries. In fact it could be still underestimated comparing the young median age of our cohort with that usually reported at diagnosis (Rozman, 1995). Moreover, the comparison of our data with those available could suggest that only a quote of the cases might be usually diagnosed. Finally, the presence of the Ig VH gene rearrangement at the time of the enrolment suggests that the disease may be present a very long time before any clinical evidence.

Author notes

Corresponding author