Abstract
Complete responses (CR) of up to 95% have been reported with both pentostatin and cladribine in patients with hairy cell leukemia (HCL). Although responses are durable, relapses occur even 10 years after treatment. No trial has compared pentostatin to cladribine for HCL. Furthermore, few series achieve sufficient maturity to answer the question: can these agents induce a cure in HCL? We reviewed retrospectively 219 patients with HCL (median follow up (FU) from diagnosis: 12.6 years) to compare pentostatin with cladribine in the treatment of HCL, and to assess the potential for cure in this disease. Diagnosis of HCL and exclusion of HCL-variant were confirmed by central review. Overall response to 1st line pentostatin (n=185) was 96% with CR of 81% and median FU of 10.8 (range 0.3–17.9) years. Response to cladribine was 100% with CR of 82%. No significant difference in response was seen with cladribine (n=34) at median FU of 7.2 (range 0.5–11.5) years. Median disease free survival (DFS) for both agents was 10 years. 38% of patients relapsed 4.9 (range 1–16) years and 4.4 (range 1–10) years after treatment with pentostatin and cladribine, respectively. Although responses were maintained for pentostatin and cladribine when used at 2nd (94% and 100% respectively) and 3rd line (100% and 100% respectively) treatment, CR decreased significantly with each sequential relapse through 70% to 45% (p≤0.01). Attainment of CR at 1st line treatment was significantly associated with increased DFS (p=0.000) as compared to those achieving only partial response (PR).
Figure
A similar result was seen at 2nd line therapy (p=0.000). DFS also showed a significant decline with sequential treatment (p=0.001) mirroring the reduced likelihood of achieving CR with increasing number of courses. There was some crossover of patients. At 1st relapse, 20 patients received pentostatin of whom 17 had previously received this agent, and 53 patients received cladribine of whom 44 were previously treated with pentostatin. Patients relapsing after an initial CR showed no significant difference in ability to re-attain CR as opposed to PR or no response, whether retreated with the same agent, or switched to the other. However, for patients who initially failed to achieve CR, switching to the alternative agent was associated with an increased rate of CR although this difference did not achieve statistical significance. Known 2nd malignancies (excluding basal cell carcinoma) occurred in 20 patients (9 %). In a separate group of 7 patients who were never treated with either agent, 2 patients developed 2nd malignancies. We demonstrated equivalent efficacies of pentostatin and cladribine in the treatment of HCL. Median survival has not been reached for either agent. At 10 years FU, the survival is 83% (pentostatin) and 90% (cladribine). DFS curves show no plateau for either agent with relapses occurring up to 16 years after pentostatin treatment. True cure in this disease thus remains elusive, however, our results suggest that first line CR with purine analogue therapy should be the prime aim of HCL treatment.
