Toward optimizing pomalidomide therapy in MM patients

In this issue of Blood, Sehgal et al report on the clinical and pharmacodynamic analysis of pomalidomide dosing strategies in multiple myeloma (MM) and their impact on immune activation and cereblon targets. The particular novelty of this study lies in the direct correlation of immune effects triggered by pomalidomide with clinical responses in MM patients. Results of this study will stimulate many additional studies.¹ 

Because of remarkable responses, the second-generation immunomodulatory drug (IMiD) pomalidomide (Pomalyst; Celgene) has been granted accelerated US Food and Drug Administration and European Medicines Agency approval in February 2013 and in August 2013, respectively, for the treatment of adult patients with relapsed/refractory MM who have received ≥2 prior treatment regimens, including lenalidomide and bortezomib, and have demonstrated disease progression on the last therapy. Functionally, preclinical data have shown that pomalidomide mediates direct anti-MM cell activity as well as immunomodulatory and antistroma support activities.²  Data on pomalidomide-mediated immune activation in vivo are limited. The optimal dosing schedules of both pomalidomide and concurrent steroids need to be clarified.

Sehgal et al randomized 39 patients with relapsed MM refractory to lenalidomide and bortezomib into 2 treatment cohorts using either continuous (2 mg for 28 of 28 days) or intermittent (4 mg for 21 of 28 days) pomalidomide dosing schedules. Using extensive analyses, the authors demonstrate that pomalidomide induces rapid changes in immune profile and function. Specifically, they show pomalidomide-induced increases of both T and natural killer (NK) cells and their activation in the peripheral blood as well as in the bone marrow. Objective treatment responses correlated with pomalidomide-induced increases in interferon γ- and tumor necrosis factor α-producing CD8⁺ and multifunctional T cells but not NK cells. Moreover, gene expression profiling analysis showed that genes altered by pomalidomide treatment in CD138-negative cells predominantly belong to immune pathways, ie, in lymphocytes. These data highlight the eminent role of immunomodulation as part of the pomalidomide-induced in vivo anti-MM activity. They also indicate that the biology of innate lymphocytes in the context of pomalidomide treatment requires further investigation. Comparable responses were observed under both dosing schedules, with significantly greater incidence of grade 3/4 side effects in the intermittent dosing schedule. These results suggest that the lower dosing schedule of pomalidomide may be sufficient in the clinical setting. However, future studies need to explore whether treatment holidays as part of intermittent dosing schedules restore sensitivity to pomalidomide. Importantly, baseline levels of CD4/CD8⁺ T cells did not predict response rates to pomalidomide treatment. Therefore, a high need remains for those biomarkers that determine which patients are likely to benefit from pomalidomide-containing immunotherapies. Moreover, previous data demonstrate that dexamethasone synergizes with pomalidomide in terms of tumor regression.²  Based on the present data, dexamethasone also blunts pomalidomide-mediated T-cell activation. Optimization of pomalidomide/dexamethasone combination therapy using intermittent or modified low-dose dexamethasone is therefore required.

Another important finding of this study is that pomalidomide treatment increases the expression of immune checkpoint modulators B- and T-lymphocyte attenuator (BTLA) on T cells and T-cell immunoglobulin and mucin-domain containing-3 (Tim-3) on NK cells in MM. These data indicate a potential therapeutic role for combining pomalidomide with next-generation immune checkpoint inhibitors directed against Tim-3 and BTLA in MM. Importantly, pomalidomide induced changes of Tim-3 and BTLA expression, but not PD-1, currently the best investigated immune checkpoint protein in T cells. In contrast to normal plasma cells, PD-1 ligand is expressed on primary MM cells and induced by cytokines and bone marrow stromal cells, highlighting the intimate functional interrelation of the nonimmune and immune compartments within the MM bone marrow microenvironment.^3,4  The efficacy of inhibiting the PD-1/PD-L1 pathway has been demonstrated in several preclinical studies of MM.^4,5  A clinical study that evaluates the activity of pomalidomide in combination with the anti–PD-1 inhibitor pembrolizumab (formerly lambrolizumab and MK-3475) in relapsed/refractory MM has recently been initiated (https://clinicaltrials.gov, #NCT02289222). Moreover, preclinical coblockade of PD-L1 and TIM-3 together with low-dose whole body irradiation has led to synergistic increases in survival.⁶  Several companies actively develop agents directed against these proteins.⁷ 

A critical role for the mechanism of action of IMiDs has been attributed to the ubiquitously expressed E3 ligase protein cereblon (CRBN). Binding of IMiDs to CRBN and the recombinant CRBN-DNA damage-binding protein-1 (DDB1) complex potentiates the ubiquitination and proteolysis of 2 transcription factors, Ikaros zinc finger proteins IKZF1/Ikaros and IKZF3/Aiolos, followed by downregulation of IRF4 and Myc.^8,9  In the final part of the present paper, the authors investigate whether the expression of CRBN, IKZF1, and IKZF3 in primary cells correlates with the clinical response to pomalidomide. A modest, but statistically significant correlation between CRBN RNA/protein levels (but not IKZF1 and IKZF3 RNA/protein levels) and the percentage of reduction in measurable disease was observed. Based on data presented in this paper, IKZF1 and IKZF3 are unreliable biomarkers for the selection of MM patients for pomalidomide therapy. However, additional studies are needed to confirm these results in a larger patient cohort. Sehgal et al also address the potential role of CRBN as a biomarker of response or resistance to pomalidomide. Similar to previous studies, the authors emphasize that measuring CRBN levels rather than CRBN mutations requires the use of standardized reagents and a thorough understanding of gene complexity. In addition, future investigations should also evaluate the existence of CRBN-independent mechanisms of intrinsic resistance to IMiDs.^9,10 

The addition of pomalidomide to the armamentarium of anti-MM agents has significantly improved the outcome in MM patients, even in those resistant against bortezomib and/or lenalidomide. The present study considerably extends our understanding of in vivo anti-MM effects of pomalidomide. It demonstrates that pomalidomide triggers strong and rapid immunomodulatory effects involving both innate and adaptive immunity, even in heavily pretreated MM, and that these effects correlate with clinical responses. Most importantly, Sehgal et al provide the basis for new avenues of future investigation. Specifically, joint efforts should aim to further optimize pomalidomide/dexamethasone dosing schedules, to identify reliable biomarkers for IMiD/pomalidomide responders, and to rationally enhance immune effects of pomalidomide by combination therapies with immune checkpoint inhibitors in particular, to further improve MM patient outcome.