Hairy Cell Leukemia: Evaluation of Long-Term Outcomes in 487 Patients

Hairy cell leukemia (HCL) is a rare hematologic disorder, and the place of emerging treatments, such as BRAF inhibitors, remains to be defined. We conducted a large, multi-center, retrospective survey in France to determine the frequency of malignancy in HCL patients (pts) and their families, and to analyze the long-term effects of the established purine nucleotide analogs (PNA) cladribine (C) and pentostatin (P).

Physician members of the Société Française d'Hématologie were surveyed on their management of HCL pts over the past 30 years. Clinicians completed a form that collected data concerning personal and familial medical history of pts, with particular focus on hematologic malignancies, solid tumors, clinical and biological presentation at HCL diagnosis, therapeutic options, response to treatment, time to relapse, secondary malignancies and cause of death.

The survey included 36 French clinical centers and 487 HCL cases (mean patient age 59y; range 29–90y). In the centers surveyed, HCL diagnosis was established after examination of peripheral blood and/or bone marrow and by immunophenotyping. Solid tumors and hematologic malignancies were present before HCL diagnosis in 45 cases (9.2%). A further 38 pts (7.8%) with pre-existing cancers were diagnosed with HCL after a median time of 89mo (range 1–529mo). At diagnosis, three HCL pts presented with follicular lymphoma, monoclonal gammopathy of undetermined significance, or large cell B cell lymphoma. Four pts developed HCL within a median time of 33mo (range 30–38mo) after diagnosis of B-cell lymphoproliferative disorders. In 93 pts (19.1%), at least one family member developed neoplasia: solid tumors=75 pts (15.4%); acute or chronic hematologic malignancies=18 pts (3.7%) with one familial HCL case. Twenty-three pts (4.7%) received no treatment for a median follow-up of 32mo (1–293). Three hundred and forty-five pts (70.8%) received just one first-line treatment (C=68.1%, P=23.8%, interferon [IFN]=3.8%, other=4.3%), and 119 pts (16.6%) received two-to-seven lines of therapy. After first-line treatment, complete responses were observed in 365 pts (74.9%; P=85.1%, C=85.0%, IFN=61.1%). Median duration of response was 41 months (1–270; C=40.5mo [1–188; 234 pts], P=43.0mo [1–209; 80 pts], IFN=74mo [24–270; 13 pts], other=21mo [2–186; 15 pts]). After second line treatment, the median duration of response decreased to 33mo (1–261; 80 pts) (C=27mo [1–159; 39 pts], P=38.5mo [0–206; 24 pts]) and also tended to differ according to first-line regimen (Table).

After third-line treatment, the median duration of response decreased again to 24mo (2–224; 24 pts) (C=46mo [2–92], P=20mo [2–166], other=22mo [2–224]). After 60 months follow-up (range 1–384mo), overall survival at 5 years was 95%. Twenty-nine pts (6%) had died (11 deaths were HCL-related), and a further 53 pts (10.9%) developed second malignancies (39 solid tumors and 14 hematologic malignancies). Second malignancies occurred 6 months after HCL diagnosis in 50 pts.

This study highlights the high frequency of cancers in HCL pts and their family members, suggesting a role for genetic factors in the development of the disease. The survey also confirms the high efficacy of PNA, and we therefore need to be cautious in adopting innovative new therapies treatment such as BRAF inhibitors or immunotoxins into first-line care.

No relevant conflicts of interest to declare.