Durability of complete response after blinatumomab therapy for relapsed/refractory diffuse large B-cell lymphoma

Despite advances in standards of care, the prognosis of relapsed/refractory (R/R) diffuse large B-cell lymphoma (DLBCL) remains poor. In these patients, 50–74% fail to respond to next line therapy, and median overall survival (OS) is 6–10 months [1,2]. Roughly, 30–50% of patients with R/R DLBCL who achieve disease remission with salvage therapies relapse within 1 year of treatment [3,4]. Thus, salvage therapies providing durable responses are needed for patients with R/R DLBCL.

Blinatumomab, CD19/CD3 BiTE^® (bispecific T-cell engager) immuno-oncotherapy, has demonstrated activity in patients with R/R DLBCL in 3 clinical studies, with complete response (CR) rates of 16%, 26%, and 29% [5–7]. In this pooled analysis, we assessed the duration of CR (DOCR), OS, and safety profile of patients with R/R DLBCL treated with blinatumomab.

Descriptions of patients and designs of 3 open-label, multicenter, single-arm trials were previously reported: phase 1 R/R B-cell non-Hodgkin’s lymphoma (B-NHL) study (NCT00274742; n = 14) [6], phase 2 R/R DLBCL study (NCT01741792; n = 25) [5], and phase 2 R/R B-NHL study (NCT02910063; n = 34) [7] are described in Supplemental Materials Table S1. In phase 1 R/R B-NHL study, 4 (29%) of 14 patients with R/R DLBCL achieved CR (defined as disappearance of all disease symptoms for at least 4 weeks); 4 (16%) of 25 patients in phase 2 R/R DLBCL study achieved CR; and 9 (26%) of 34 patients with R/R DLBCL in phase 2 R/R B-NHL study achieved complete metabolic response (CMR) per Lugano classification (defined as no new lesions, score 1–3 of staging primary lymph nodes, and no evidence of disease in bone marrow by positron emission tomography [PET]/computed tomography [CT]). All patients provided written, informed consent before enrollment. Institutional review board approval was obtained for each study.

Table 1. Patient baseline demographics and disease characteristics.

CR					Non-CR
Characteristic	Phase 2 B-NHL study (n = 9)	Phase 2 DLBCL study (n = 4)	Phase 1 study (n = 4)^a	Total (n = 17)	Total (n = 63)
Sex, n (%)
Male	5 (55.6)	2 (50.0)	1 (25.0)	8 (47.1)	41 (65.1)
Female	4 (44.4)	2 (50.0)	3 (75.0)	9 (52.9)	22 (34.9)
Age, median (range), years	57 (51–68)	70 (62–75)	56 (35–65)	59 (51–68)	56 (47–67)
Primary disease status, n (%)
Refractory	4 (44.4)	2 (50.0)	NA	6 (35.3)	38 (60.3)
Relapsed	5 (55.6)	2 (50.0)	4 (100.0)	11 (64.7)	25 (39.7)
Prior HSCT, n (%)
Yes	0 (0)	2 (50.0)	2 (50.0)	4 (23.5)	11 (17.5)
No	9 (100.0)	2 (50.0)	2 (50.0)	13 (76.5)	52 (82.5)
Double hit, n (%)
Yes	1 (11.1)	1 (25.0)	NA	2 (11.8)	6 (9.5)
No	8 (88.9)	3 (75.0)	NA	11 (64.7)	47 (74.6)
Prior anticancer regimens, n (%)
1	0 (0)	1 (25.0)	0 (0)	1 (5.9)	1 (1.6)
2	9 (100.0)	1 (25.0)	2 (50.0)	12 (70.6)	34 (54.0)
≥3	0 (0)	2 (50.0)	2 (50.0)	4 (23.5)	28 (44.4)
Tumor stage, n (%)
II	2 (22.2)	0 (0)	2 (50.0)	4 (23.5)	19 (30.2)
III	4 (44.4)	1 (25.0)	1 (25.0)	6 (35.3)	18 (28.6)
IIIB	0 (0)	1 (25.0)	0 (0)	1 (5.9)	3 (4.8)
IV	3 (33.3)	2 (50.0)	1 (25.0)	6 (35.3)	33 (52.4)

B-NHL: B-cell non-Hodgkin’s lymphoma; CR: complete response; DLBCL: diffuse large B-cell lymphoma; HSCT: hematopoietic stem cell transplant; NA: not applicable.

^aIn phase 1 study, unconfirmed complete response was used as a response, and the patient was included as a complete responder (n = 1).

Objective response rate (ORR), CR plus partial response (PR), was a study endpoint for phase 1 R/R B-NHL [6] and phase 2 R/R DLBCL studies [5]. CMR was the primary endpoint for phase 2 R/R B-NHL study [7]. ORR was assessed by locally reviewed PET/CT, whereas CMR was assessed by centrally reviewed PET. Response was defined according to International Working Group criteria for phase 1 R/R B-NHL study (CR or unconfirmed CR); by Cheson revised criteria for phase 2 R/R DLBCL study (CR); and by Lugano classification (CMR by PET) for phase 2 R/R B-NHL study. DOCR was defined as time from first assessment of CR until progression of disease, initiation of new antitumor therapy, or death. All adverse events (AEs) were recorded and graded per National Cancer Institute Common Terminology Criteria for Adverse Events [8,9].

This analysis assessed DOCR of patients with R/R DLBCL who achieved CR or CMR within cycle 1 (12 weeks) of blinatumomab treatment in the 3 studies. DOCR was estimated as a time-to-event variable using the Kaplan–Meier method and 2-sided 95% confidence interval (CI).

Seventeen of 73 (23%) patients with R/R DLBCL achieved CMR/CR within 12 weeks of blinatumomab and included in the assessment of DOCR. Generally, demographics and baseline disease characteristics were consistent across the 3 studies (Table 1). Most patients in the phase 2 studies had DLBCL (83% and 100%), whereas indolent lymphoma (68%) was the most common subtype in the phase 1 R/R B-NHL study. Across the 3 studies, 11 (65%) patients had relapsed disease; 6 (35%) were primary refractory; 16 (94%) had ≥2 lines of prior treatment; and 4 (24%) received previous hematopoietic stem cell transplant (HSCT).

Among 17 patients with R/R DLBCL who achieved CR/CMR, 16 (94%) completed cycle 1 of blinatumomab and 11 (65%) received ≥80% of their total intended cycle 1 dose (Table S2). In total, 16 (94%) patients had treatment interruptions, including 5 (29%) for AEs and 9 (53%) for other reasons (e.g. imaging scans, pump error). Of 6 patients who started cycle 2, three completed the cycle and two received ≥80% of the intended dose. Treatment interruptions during cycle 2 occurred in 2 (12%) patients due to AE (n = 1) and technical reason (n = 1).

In this pooled analysis, the median DOCR (months) was not reached (95% CI, 6.0–not estimable [NE]), with a median follow-up of 15.6 months. At 6, 12, 18, and 21 months, 79.0%, 62.2%, 62.2%, and 62.2% of patients were estimated to remain in response, respectively (Figure 1(a)). Of 17 patients with R/R DLBCL who achieved CR/CMR with blinatumomab, 4 (23.5%) patients relapsed and 1 (5.9%) died during follow-up. Of 7 patients who were still in CR >12 months, 3 received autoHSCT and 4 did not receive HSCT.

Figure 1. (a) Kaplan–Meier analysis of DOCR for patients who achieved CMR/CR within 12 weeks. DOCR is calculated from the date of response of achieving CMR/CR until earliest date of relapse or death, whichever occurs first. Patients who have not relapsed and are alive are censored on their last disease assessment date. (b) Kaplan–Meier analysis of overall survival for patients who achieved CMR/CR within 12 weeks. Overall survival is calculated from the time of first dose of blinatumomab until death due to any cause. Patients alive are censored at the date last known to be alive. In the phase 1 study, a patient with unconfirmed CR was included as a complete responder (n = 1). CI: confidence interval; CMR: complete metabolic response; CR: complete response; DOCR: duration of CR; NE: not estimable.

In addition, 13 patients achieved PMR/PR; median duration of PR (months) was 2.1 (95% CI, 0.9–NE) with a median follow-up of 7.6 months. At 6, 12, 18, and 24 months, 32.7%, 21.8%, 21.8%, and NE were estimated to remain in response, respectively. Moreover, of 13 patients who partially responded to blinatumomab, 6 (46.2%) relapsed, 2 (15.4%) died, and 4 (30.8%) patients received HSCT (alloHSCT, n = 1; autoHSCT, n = 3).

Median OS (months) was not reached (95% CI, 13.1–NE), with median follow-up of 16.4 months. The OS probability at 6, 12, 18, and 24 months was 100%, 86.5%, 77.9%, and 77.9%, respectively (Figure 1(b)). Furthermore, of 17 patients who responded to blinatumomab after treatment, 3 (17.6%) died during follow-up and 9 (52.9%) proceeded to HSCT (alloHSCT, n = 1; autoHSCT, n = 8).

All 17 (100%) patients had any-grade treatment-emergent AEs (TEAEs), of whom 14 (82.4%) had grade ≥3 TEAEs and 8 (47.1%) had serious AEs (Table S3). Overall, 3 (17.6%) patients discontinued treatment due to TEAEs. The most common grade 3 TEAEs were neurologic events (5.3%), infections (35.3%), cytopenias (29.4%), and elevated liver enzyme (17.6%). Overall, 16 (94.1%) patients had treatment-related AEs (TRAEs); 11 (64.7%) patients had grade ≥3 and 1 (5.9%) had grade ≥4 TRAEs; 4 (23.5%) TRAEs were deemed serious; 11 (64.7%) patients developed severe TRAEs and 2 (11.8%) discontinued treatment. Altogether, no life-threatening or fatal AEs were reported.

Of 17 patients, 15 (88.2%) had neurologic adverse events (NAEs) of any grade, including tremor (52.9%), headache (47.1%), and paresthesia (23.5%). Grade ≥3 NAEs were reported for 6 (35.3%) patients. NAEs led to blinatumomab interruption in 4 (23.5%) patients and discontinuation in 2 (11.8%). No patients had grade 4 or fatal NAEs. One patient reported grade 3 cytokine release syndrome.

The results of this pooled analysis suggest treatment with blinatumomab salvage therapy may lead to durable CR and survival benefit in patients with R/R DLBCL. Although median DOCR and median OS were not reached at the time of analysis, 62.2% of patients were still responding and 86.5% of patients were alive at 12 months. Moreover, 3 patients in phase 2 R/R DLBCL study achieved late CRs that were not captured in this pooled analysis and had longer DOCRs [5]. Furthermore, in a subset of patients with R/R B-NHL from phase 1 study (NCT00274742), remission is ongoing after more than 7 years [10]. There was no difference between CR and non-CR patients in rate of prior transplant or double hit disease. In non-CR patients, there was a trend for more refractory and late stage tumor baseline characteristics. However, about one-third of CR patients were refractory and had late stage. Considering the highly aggressive disease in this pooled population, these results are especially noteworthy.

The ZUMA-1 and JULIET studies reported high response rates and favorable OS with chimeric antigen receptor (CAR) T-cell therapies in patients with B-cell lymphomas. However, as these studies do not report on intent-to-treat populations, a relatively high number of patients excluded prior to treatment (drug production failures, tumor progression prior to treatment, deterioration of performance status) likely bias the results and preclude efficacy comparisons between CAR T-cell therapy and blinatumomab [11].

Although blinatumomab requires stepwise dose escalation over 3 weeks to mitigate neurotoxicity, it is an ‘off-the-shelf’ product which can be started immediately, allowing the treatment of more acutely ill patients in stark contrast to CAR T-cell therapy. The reported incidence of NAEs with blinatumomab is similar to that reported from ZUMA-1 and JULIET whereas incidence of cytokine release syndrome was lower with blinatumomab [12–14].

The retrospective pooled analysis was limited by small sample size and an intrinsic inability to compare patients who received or did not receive HSCT. Therefore, results should be interpreted with caution, and additional studies with larger sample size, longer follow-up, and investigation of how to identify subsets of responsive patients are needed. Despite these limitations, the DOCR and OS rates observed with blinatumomab are particularly encouraging given the aggressive nature of disease in this heavily pretreated patient population.

Considering how depth and duration of response to blinatumomab in the acute lymphoblastic leukemia population increased substantially when used earlier as consolidation therapy in patients with minimal residual disease [15] compared with use in the R/R setting [16], our results may warrant investigation of blinatumomab in earlier settings in patients with high-risk DLBCL and minimal residual disease.

In conclusion, salvage treatment with blinatumomab can lead to sustained responses with potential of long-term disease control and with a manageable safety profile in heavily pretreated patients with R/R DLBCL.

Acknowledgments

Beatrice Chiang (Amgen Inc.) provided medical writing assistance funded by Amgen Inc. and Ben Scott (Scott Medical Communications, LLC) provided editorial assistance funded by Amgen Inc. Bob Dawson (Cactus Communications, on behalf of Amgen Inc.) provided graphics support.

Disclosure statement

A.V. has received honoraria from and has served as a consultant/advisor for Roche, Bristol-Myers Squibb, Amgen Inc., Kite/Gilead, and Pfizer; and has received travel expenses from Roche, Celgene, Bristol-Myers Squibb, AbbVie, and Kite/Gilead. G.H. has served as a consultant/advisor for Roche, Janssen, Celgene, Gilead, Servier, and AbbVie; has received research funding from Roche, Servier, and Celgene; and has received travel expenses from Celgene, Roche, and Janssen. R.C.B. has received honoraria, has served as a consultant/advisor, and has received travel expenses from Amgen Inc., Cellex, GEMoaB, Genmab, Molecular Partners, and Novartis; and is a patent holder for blinatumomab and royalty payments. N.J.M. has received honoraria from Amgen Inc. and Janssen; has served as a consultant/advisor for Amgen Inc. and Roche; and has served on the speakers’ bureau for Amgen Inc. G.G. has served as a consultant/advisor for Autolus Ltd; and has received travel expenses from Roche, Takeda, and Gilead. M-E.G. has received honoraria from Novartis, Roche, Bristol-Myers Squibb, and Janssen-Cilag; and has served as a consultant/advisor for GEMoaB, Roche, and Novartis. K.I., A.Z., and J.F. are employees of and have stock ownership in Amgen Inc. D.C. was an employee of Amgen Inc. at the time of the study and has continued stock ownership in Amgen Inc. L.C. has received honoraria and travel expenses from Amgen Inc.

Data availability statement

Qualified researchers may request data from Amgen clinical trials. Complete details are available at http://www.amgen.com/datasharing.

Additional information

Funding

This study was funded by Amgen Inc.