High Expression Of Cereblon (CRBN) Is Associated With Achievement Of Complete Response To Thalidomide Plus Fludarabine Regimen In Chronic Lymphocytic Leukemia

Cereblon (CRBN) has recently been identified as a target for immunomodulatory drugs (IMiDs), thalidomide, lenalidomide and pomalidomide, and its downregulation has been linked to resistance to IMiDs in multiple myeloma (MM). However, little is known on the role of CRBN as a resistance/response biomarker in other diseases in which treatment with IMiDs is investigated, such as chronic lymphocytic leukemia (CLL). Although lenalidomide has recently proven to be more promising drug in CLL, we had previously performed a clinical trial that assessed efficacy of thalidomide (100mg p.o on days 1-28) combined with fludarabine (25mg/m2 i.v. on days 7–11) given for six 4-week cycles in patients with CLL. The objective of the present study was to correlate retrospectively pre-treatment CRBN expression in CLL cells with response to thalidomide/fludarabine regimen as well as to cellular and molecular changes that were monitored in vivo after initial 7-day thalidomide monotherapy. CRBN and interferon regulatory factor 4 (IRF4) expressions were studied by quantitative RT-PCR using mRNA extracted from frozen pre-treatment CLL cells. Tumor necrosis factor (TNF) and TNF receptor expression were analyzed by ELISA. T-cell populations were estimated using flow cytometry during the clinical part of the study. Frozen biological pre-treatment samples were available for 27 out of 40 patients included to the thalidomide/fludarabine clinical trial. First, we verified that CRBN expression was independent from all established prognostic parameters in CLL including IGVH mutational status, high-risk FISH, ZAP-70 and CD38 expression and beta-2 microglobulin level (p>0.05 for all parameters). Subsequently, we analyzed the CRBN expression in regard to response and CLL prognosis. Interestingly, we found that patients with CRBN expression in the highest quartile tended to have significantly higher probability of complete response (CR) (50% vs. 10%, Fisher exact test p=0.56). However, although median CRBN expression was highest in patients with CR (median=0.80), it was lower in patients achieving partial remission (PR) (0.24) than in patients who did not responded (0.50), and the response categories did not correlate directly with CRBN expression in CLL cells, (r=0.07, p=0.71). Furthermore, no significant association was found between CRBN expression above and below median and duration of response after thalidomide/fludarabine treatment (28 vs. 29 month, p=0.88). Interestingly, in contrast to previous data in MM, we found that CRBN expression did not correlate with IRF4 expression (r=0.25, p=0.22). Furthermore, analysis of molecular and cellular parameters assessed after initial 7-day thalidomide monotherapy including the changes in TNF levels and regulatory T-cell populations did not reveal any significant relation to basal CRBN levels. In conclusion, we found that high pre-treatment CRBN expression in CLL cells is associated with CR achievement on thalidomide-containing chemotherapy. However, in contrast to published data on MM, CRBN expression in CLL cells does not correlate directly with CLL prognosis or molecular and cytological effects caused by thalidomide. Our results may indicate complex mechanisms of IMiDs activity in CLL and that basal CRBN levels in CLL microenvironment and immune effector T and NK cells may be more important for CLL outcome.