MM-398 Benefit Further Supported in Metastatic Pancreatic Cancer

Photo by © ASCO/Todd Buchanan 2015

Li-Tzong Chen, MD, PhD

An analysis of the per-protocol population in the phase III NAPOLI-1 trial supports the benefits of adding MM-398 to 5-fluorouracil plus leucovorin (5-FU/LV) for the treatment of patients with metastatic pancreatic cancer who were previously treated with gemcitabine.

In the per-protocol assessment, treatment with the nanoliposomal encapsulation of irinotecan MM-398 plus 5-FU/LV improved overall survival (OS) by 53% compared with 5-FU/LV alone. In the full intent-to-treat (ITT) population of NAPOLI-1, MM-398 plus 5-FU/LV improved OS by 43% as reported in 2014.

“In the per-protocol population, the MM-398 plus 5-FU/LV combination regimen achieved a median overall survival of 8.9 months,” said Li-Tzong Chen, MD, PhD, of the National Institute of Cancer Research in Tainan City, Taiwan, who reported the expanded pre-specified per-protocol and sensitivity analyses of the NAPOLI-1 data at the 2015 Gastrointestinal Cancers Symposium.

MM-398 contains about 80,000 molecules of irinotecan stably encapsulated in a 100-mm liposome. MM-398 at 120 mg/m²has extended circulation and achieves 9.6 ng/g of the active metabolite, SN-38, in the tumor, compared with 1.7 ng/mL in blood at 72 hours.

The open-label NAPOLI-1 study randomized 417 patients to receive MM-398 as a single agent, 5-FU/LV (control), or MM-398 prior to 5-FU and racemic LV. The primary endpoint was overall survival.

In the monotherapy arm, MM-398 was administered at 120 mg/m²every 3 weeks. In the control, 5-FU was administered at 2000 mg/m²over 24 hours with racemic LV at 200 mg/m²over 30 minutes for 4 weeks followed by 2 weeks off. In the combination arm, MM-398 was administered at 80 mg/m²every 3 weeks prior to 5-FU at 2400 mg/m²over 46 hours and racemic LV at 400 mg/m²over 30 minutes.

The ITT analysis explored all patients randomized to MM-398 plus 5-FU/LV (n - 117) and those in the control (n = 119). The per-protocol population included patients who received at least 80% of the target dose in the first 6 weeks and did not violate any inclusion or exclusion criteria. There were 66 patients in the per-protocol MM-398 arm and 71 in the control. The non-per-protocol population (n = 99) included those not qualifying for per-protocol analysis.

In the ITT population, the median PFS was 3.1 months for the combination compared with 1.5 months with the control (HR = 0.56; 95% CI, 0.41-0.75;P= .0001). The partial response rate was 16% versus 1% (P<.001). The disease control rate with MM-398 was 57%.

Table. Median OS in Per-Protocol and ITT Populations

MM-398 + 5-FU/LV

5-FU/LV

Stratified HR

Intent-to-Treat

6.1 months

4.2 months

0.57 (95% CI, 0.41-0.80)

P= .0009

Per-Protocol

8.9 months

5.1 months

0.47 (95% CI, 0.29-0.77)

P= .0018

Non-Per-Protocol

4.4 months

2.8 months

0.56 (95% CI, 0.33-0.97)

P= .0365

Analysis of the per-protocol population confirmed the favorable OS found with the combination in the ITT analysis. Additionally, the median OS appeared longer in the per-protocol arm than in the ITT population (Table). This was also reflected by the progression-free survival (PFS), objective response rate, and CA19-9 level for the combination over the control arm, Chen reported.&ldquo;Not surprisingly, the combination achieved a highly significant improvement in overall survival in the per-protocol analysis. But more surprisingly, it also significantly improved overall survival in the non-per-protocol population, versus the control arm,&rdquo; Chen said. &ldquo;Analysis of the subgroups consistently favored overall survival for the MM-398 plus 5-FU/LV arm over 5-FU/LV alone.&rdquo;

Analysis by pretreatment characteristics included age at diagnosis, time since initial histological diagnosis, prior lines of therapy, time since last prior therapy, and CA1909 levels. In the ITT arms, CA19-9 levels were decreased by &ge;50% in 36% versus 12% of patients in the combination and control arms, respectively.

No benefit was observed for MM-398 as a single agent, with median OS of 4.9 months, compared to 4.2 months with 5-FU/LV (P= .5545) in the ITT population.

Dose reductions were necessary for 33% of the MM-398 arm, versus 3% of the control arm, and dose delays occurred in 61% versus 21%, respectively. In spite of this, the average duration of exposure was 21 weeks for the combination, versus 13 weeks for 5-FU/LV alone, and the average relative dose intensity exceeded 85%, Chen indicated.

The safety profile of the combination regimen was considered manageable. In the per-protocol population, the most frequent grade &ge; 3 adverse events (versus the control arm) were neutropenia (15% vs 3%), fatigue (14% vs 6%), diarrhea (12% vs 7%), and vomiting (8% vs 3%).

Study discussant Laura Goff, MD, of Vanderbilt University Medical Center, Nashville, commented that the partial response rate of 16% in a second-line population was &ldquo;provocative,&rdquo; and the median survival times were similar to those previously reported for FOLFIRI (fluorouracil, leucovorin, irinotecan) in second-line pancreatic cancer.

&ldquo;MM-398 appears to have activity in combination with 5-FU/leucovorin in previously treated pancreas cancer, and it has received Fast Track status from the FDA,&rdquo; Goff indicated.

In November 2014, the FDA granted MM-398 plus 5-FU and leucovorin a Fast Track designation as a second-line treatment for patients with metastatic pancreatic cancer, based on data from the NAPOLI-1 study. As part of this review program, the company developing MM-298, Merrimack Pharmaceuticals, can submit information for a New Drug Application for MM-398 to the FDA in sections. The company plans to complete this submission process within the first few months of 2015.

Chen L-T, von Hoff DD, Li C-P, et al. Expanded analyses of napoli-1: Phase 3 study of MM-398 (nal-IRI), with or without 5-fluorouracil and leucovorin, versus 5-fluorouracil and leucovorin, in metastatic pancreatic cancer (mPAC) previously treated with gemcitabine-based therapy. 2015 Gastrointestinal Cancers Symposium; January 15-17, 2015; San Francisco, CA. Abstract 234.