To our knowledge, this is the first study to establish single-agent activity of temsirolimus in patients with B-cell lymphomas other than MCL, and validates mTOR inhibition as a rational target across lymphomas. Patients were stratified by histology (DLBCL, FL, other indolent lymphomas), with clinical activity observed in aggressive and indolent subtypes. Responses occurred in heavily pretreated patients, including patients relapsing after prior autologous stem-cell transplantation. Greater than 80% of patients had relapsed after prior rituximab-containing therapy. Patients with FL had the best outcome, with an overall response rate of 54% and a CR rate of 26%; responses were durable, with a median duration of response approximately 13 months. Patients with DLBCL had a 28% response rate but the duration of response was only 2.4 months. There was minimal single-agent activity in chronic lymphocytic leukemia/small lymphocytic lymphoma. Temsirolimus was generally well-tolerated with weekly dosing, with the expected metabolic changes and mild stomatitis being the most common nonhematologic toxicities. Possible drug-related pneumonitis occurred in four patients. Although it is difficult to determine predisposing factors in such a small group of patients, we note that two patients with possible pneumonitis experienced treatment failure a prior autologous transplant, and that radiation and/or radioimmunotherapy had been used in three of the four patients. Cytopenias were common, but reversible in all cases. The high incidence of observed grade 3 and 4 hematopoietic toxicities partially reflects the eligibility criteria, which allowed patients with baseline platelets as low as 50,000/μL (20,000/μL if there was marrow involvement) and absolute neutrophil count of 1,000/μL to enroll.
It is of interest that four of the nine responders in group A (DLBCL, transformed FL) had transformed from a prior FL. It is possible that temsirolimus is more active in follicle center-derived lymphomas, and the germinal center (GC) phenotype of DLBCL may predict for a higher response rate compared to the non-GC phenotype. Indeed, six of the seven responders who had immunohistochemistry performed (as per Hans et al22
) were CD10 positive compared with only six of 23 group A patients who did not respond to single-agent temsirolimus (data not shown). This is an intriguing hypothesis that will need prospective validation. In this study, a significant limitation is the paucity of primary tumor tissue prospectively collected.
It is unclear if the efficacy observed with temsirolimus in non-Hodgkin's lymphoma is a class effect and whether it can be replicated by other mTOR inhibitors. Everolimus, an oral prodrug of rapamycin, was tested in a phase I trial in 27 patients with advanced hematologic malignancies by the M. D. Anderson Cancer Center.23
Although there were no objective responses in patients with lymphoid malignancies, four of six patients with chronic lymphocytic leukemia had either a decrease in lymphocytosis or lymphadenopathy. A second brief report of single-agent everolimus in seven patients with heavily pretreated chronic lymphocytic leukemia showed significant infectious complications without objective responses, although all patients were heavily pretreated and had received significant prior immunosuppressive therapy.24
The largest study of everolimus in lymphomas was a phase II study conducted by the Mayo Clinic group, and recently reported in abstract format.25
One hundred forty-five patients with aggressive and indolent lymphomas were accrued, and the study included an exploratory arm with T-cell lymphomas and Hodgkin's lymphomas. This was an elderly population with a median age of 67 years (although the exploratory group was younger) and a median of four prior treatment regimens. The overall response rate was 33%, and similar to our findings, was heavily influenced by histology. The median PFS was 4.3 months, and median duration of response was 6.8 months. These positive results have prompted an ongoing randomized phase III trial of adjuvant everolimus in patients with DLBCL. Finally, Ghobrial et al26
recently reported a 42% PR rate in patients with Waldenstrom's macroglobulinemia treated with everolimus. Although it remains to be seen if a specific mTOR inhibitor is superior to another, these promising data suggest that a variety of mTOR inhibitors are active in lymphomas.
A major unresolved question is the optimal dose of temsirolimus in lymphomas. Dosing in this trial derives from solid tumor literature that 25 mg weekly carries similar efficacy and diminished toxicity compared to 250 mg weekly, which was close to the maximum-tolerated dose identified in advanced solid tumors.27,28
In support of the lower dose are sequential publications by the Mayo Clinic phase II consortium and the North Central Cancer Treatment Group (NCCTG) in patients with MCL, both of which were published after this study was underway.29,30
This group initially investigated 250 mg weekly in relapsed MCL, but found that dose reductions, primarily due to hematologic toxicity, were required in 30 of 34 patients. The NCCTG subsequently showed that 25-mg weekly preserved the response rate and PFS and improved tolerability in a very similar population of heavily pretreated patients with MCL. However, both were phase II trials, and a true phase I trial has not been performed in lymphoma. Of note, a recently published international phase III trial readdresses the issue of dose in patients with relapsed MCL.31
There were three treatment arms: temsirolimus 175 mg weekly for 3 weeks followed by 75 mg weekly (175/75 mg), temsirolimus 175 mg weekly for 3 weeks followed by 25 mg (175/25 mg), or investigator's choice of single-agent chemotherapy. In contrast to the phase II data, improved response rates and PFS were seen for the high-dose but not the low-dose temsirolimus arm when compared to investigator's choice therapy. Specifically, the 175/75-mg group showed a superior overall response rate (22% v
= .0019) and PFS (median 4.8 v
1.9 months; P
= .0009) compared to investigator's choice; in contrast, the 175/25-mg group only trended toward improved response rate (6% v
2%) and PFS (median 3.4 v
1.9 months). Neither temsirolimus arm improved OS, although survival was not the primary end point. The suggestion that higher temsirolimus dosing may be more effective contrasts with the two phase II studies, and it remains unclear if these findings are relevant in other lymphoma histologies.
In summary, to our knowledge, this is the first report of substantial activity of temsirolimus in lymphomas other than MCL. Durable responses were observed in heavily pretreated patients, including patients who experienced treatment failure with a prior autologous stem-cell transplant. In addition, many patients were successfully bridged to transplant. The significant activity and general tolerability of single-agent temsirolimus in patients with FL and DLBCL is encouraging, and warrants investigation in combination with other agents. Critical to further development of these agents, especially when placed in combination settings, is a better understanding of mechanisms of resistance. Several groups have hypothesized that combining mTOR inhibitors with other PAM pathway inhibitors (ie, PI3K inhibitors, Akt inhibitors) or with agents that block parallel pathways will be important. In addition, the heterogeneity of response implies that underlying biologic features may predict outcome and this study should prompt further investigations into defining these features.