Immune checkpoint inhibitor therapies (ICIs) pembrolizumab and nivolumab show high overall response rates (ORR) in the treatment of relapsed or refractory (R/R) primary mediastinal B-cell lymphoma (PMBCL), regardless of whether consolidation therapy was used, new data from a real-world setting showed. 

This finding “indicates that ICIs may be used with curative intent in a significant subset of patients,” said the authors in their study published in the American Journal of Hematology.

PMBCL, a rare subtype of BCL predominantly affecting young women, has high cure rates with chemoimmunotherapy; however, patients with R/R disease had worse outcomes until only recently. The advent of immunotherapy changed that, with ICI drugs pembrolizumab and nivolumab combined with brentuximab vedotin (nivo-BV) having shown high efficacy and favorable safety in phase 1 and 2 clinical trials involving patients with R/R PMBCL. Key trials included the KEYNOTE-170 trial, which showed an ORR of 41.5% with pembrolizumab, and the CheckMate 436 trial, which showed an ORR of 73.3% with nivo-BV.

The authors of the new research conducted the multicenter, retrospective, PRIMICI trial because of the lack of real-world studies of ICIs, including data on outcomes with consolidation strategies. Seventy-four patients with R/R PMBCL enrolled in the study, including 42 treated with pembrolizumab and 32 with nivo-BV.

Methods and Results

Treatment consisted either of 200 mg pembrolizumab given intravenously once every 3 weeks up to 35 cycles until disease progression or unacceptable toxicity or 240 mg nivolumab and 1.8 mg/kg BV given intravenously once every 3 weeks until disease progression or unacceptable toxicity.

The patients had a median age of 31 years (range, 18-66) at the time of ICI treatment, and their median number of prior therapies was two, including 10 patients having prior autologous stem-cell therapy (SCT), one having allogeneic SCT, and three having prior CAR T-cell therapy. 

Over a median follow-up of 34 months, high ORRs were observed, with an average ORR of 64% in both groups combined and with a trend toward a higher ORR with nivo-BV vs pembrolizumab (75% vs 55%; P = .09).

The median time to best response was just 2.8 months with nivo-BV vs 5.5 months with pembrolizumab (P = .003).

The average complete remission (CR) rate overall was 50%, with 56% for nivo-BV and 45% for pembrolizumab.

Among 37 patients achieving CR, eight of 18 patients treated with nivo-BV and three of 19 patients in the pembrolizumab group received consolidation; however, the difference was not significant (P = .079).

Importantly, the stability of CR rates persisted independent of the use of consolidation, with 4-year rates of disease-free survival of 100% among all study participants.

The estimated 5-year progression-free survival (PFS) was 60.4%, while overall survival rate was 77.5%.

The results were similar regardless of the number of prior lines of therapy, “suggesting that ICIs have an efficacy that is in part independent of the number of prior treatments, probably because of the completely different mechanism of action compared to standard chemotherapy,” the authors noted.

Regarding safety, the rates of serious or grade 2 or higher adverse events (AEs) were comparable to those reported in other studies, with only 9% of patients having had to permanently discontinue treatment because of an AE. 

Grade 3 or higher BV-­related peripheral neuropathy was observed in 9% of patients, which was similar to the rate reported in the CheckMate 436 study.

Faster Response With Nivo-BV

The authors underscore that the differences observed between nivo-BV and pembrolizumab in terms of a faster response were important, and “remarkable.” 

“In an aggressive disease like PMBCL, promptly identifying patients who are not responding to treatment is crucial for timely referral to alternative therapies like CAR T,” they explain. 

While the reasons for a faster response with nivo-­BV are unclear, one theory is that “despite the very limited activity of BV as single agent, the combination with nivolumab may allow a better tumor control in the early phases while the anti-­lymphoma immune activity is still mounting,” the authors said.

“In addition, a chemo-­sensitizing effect of ICIs, as reported in Hodgkin lymphoma, may be supposed,” they add.

The reasons for the higher ORR observed with pembrolizumab in the study compared with the rate reported in the KEYNOTE-170 (55% vs 41.5%, respectively) are also not clear, but the difference may reflect increasing experience with the drug in real-world settings, the authors speculated.

In addition, the current study had more favorable patient selection, with a lower median number of prior treatments and fewer patients having had a previous SCT (14% vs 26%) as well as a lower median number of prior treatments (two vs three) than the KEYNOTE-­170 trial.

Overall, the study “provides the largest published dataset of patients obtaining a CR after ICIs in R/R PMBCL, [with] 37 patients compared with the cumulative 23 patients reported in the CheckMate 436 and KEYNOTE-­170 studies, representing an interesting point of reference,” the authors wrote.

“Taken together, our data reinforce the concept that patients that reach a CR after ICIs may be probably cured even without a consolidation treatment.”

“For these patients, the ideal length of treatment has not been determined yet,” they continued.

The study “adds to the current literature by confirming findings that have been suggested by clinical trials, [showing] that these findings are also clear in the noninvestigational ‘real-world’ population with both manageable toxicity and comparable efficacy,” said Catherine Diefenbach, MD, director of the Clinical Lymphoma Program at NYU Langone’s Perlmutter Cancer Center, New York City.

She noted that previous studies have shown nivo-BV to be superior to both BV and nivo alone in relapsed Hodgkin lymphoma and in PMBCL expressing CD30. Therefore, “combining two therapies that target complimentary aspects of PMBCL biology seems like a rational strategy and likely superior to single-agent therapy,” she said.

Ultimately, “both regimens, nivo-BV and pembrolizumab, should be considered for patients with relapsed PMBCL, and both regimens offer CR patients the possibility for long-term disease control,” she said.

The authors had no disclosures to report. Diefenbach reported consulting and serving on advisory boards for Bristol Myers Squibb and Pfizer.