'No' to Anti-VEGF-Chemo Combo in NSCLC Sans Driver Alterations

Retrospective studies have touted a benefit for bevacizumab-pemetrexed as maintenance therapy in patients with metastatic non-squamous non-small cell lung cancer (NSCLC). But the pemetrexed plus anti-VEGF therapy approach was upended by the 2019 ECOG-ACRIN 5508 study, which showed that pemetrexed alone, bevacizumab alone, or the combination, did not offer a survival benefit in these patients -- not to mention hit those patients with higher toxicities.

Given those results, "anti-VEGF therapy should not be added to pemetrexed and platinum therapy or given as maintenance," surmised Dwight H. Owen, MD, MS, of the Ohio State University in Columbus, and colleagues, in their Journal of Clinical Oncology (JCO) update to the ASCO Living Guideline for advanced NSCLC without driver alterations.

What exactly did ECOG-ACRIN 5508 reveal? Suresh Ramalingam, MD, of Emory University in Atlanta, and colleagues tested the regimen in previously untreated patients with advanced nonsquamous NSCLC. They received carboplatin, paclitaxel, and bevacizumab for up to four cycles. Patients without progression after four cycles were randomly assigned to maintenance therapy with bevacizumab (controls), pemetrexed, or a combination. The trial's primary endpoint was overall survival (OS).

Ramalingam's group reported that at a median follow-up of 50.6 months, median OS with pemetrexed was 15.9 months versus 14.4 months with bevacizumab, and 16.4 months with the combination.

In terms of adverse events (AEs), more than half of those in the combination arm had grade 3-4 AEs compared with 37% with pemetrexed alone and 29% with bevacizumab alone. These AEs included anemia, fever with neutropenia, neutropenia, thrombocytopenia, fatigue, and hypertension.

Ramalingam and colleagues concluded that "single-agent bevacizumab or pemetrexed is efficacious as maintenance therapy for advanced nonsquamous NSCLC. Because of a lack of survival benefit and higher toxicity, the combination of bevacizumab and pemetrexed cannot be recommended."

But not everyone was in agreement with the ECOG-ACRIN 5508 investigators. In two letters in JCO, other experts shared their concerns with the trial; and Ramalingam and co-investigator Suzanne Dahlberg, PhD, of Boston Children's Hospital, responded to each.

How should the ECOG-ACRIN 5508 findings be considered in terms of other trials?

Nasser Hanna, MD and Shadia Jalal, MD, both of Indiana University School of Medicine: The benefit of maintenance bevacizumab monotherapy in NSCLC remains elusive, as no randomized trial has addressed this issue directly. The trial by Ramalingam et al. included a switch maintenance strategy on two of the three arms, with different chemotherapy backbones used in induction and in the maintenance setting ... the bevacizumab arm also was numerically inferior with respect to survival compared with pemetrexed or the combination of pemetrexed plus bevacizumab ... although this difference was not statistically significant.

These results should be placed in context with other trial results addressing a similar question [POINTBREAK, AVAPERL, COMPASS] ... Taken collectively, these three trials report consistent results; namely, maintenance pemetrexed plus bevacizumab results in superior PFS compared with bevacizumab alone, and the OS of those reaching the maintenance phase numerically favors the combination.

Ramalingam and Dahlberg: The results of ECOG-ACRIN 5508 are consistent with observations from other trials. The hazard ratio for [OS] was 0.9 in this trial. The AVAPERL and COMPASS studies ... reported hazard ratios for overall survival of 0.87 in each. Our article about ECOG-ACRIN 5508 had been accepted for publication ... before the presentation of the results of the [2019] COMPASS study; hence, we did not have the opportunity to cite those results. The fact that all of the trials demonstrate a modest numerical improvement in survival for the combination regimen without a statistically significant [OS] benefit cannot be accepted as sufficient grounds to recommend this strategy.

The adoption of bevacizumab and pemetrexed as combination maintenance therapy was based on a phase II study [by Jyoti Patel, MD, and colleagues] that demonstrated favorable [OS], which was published in 2009. That it has taken until 2019 to definitely conclude that the combination approach lacks merit should serve as an important lesson against adopting promising approaches without documented proof of benefit.

Was ACRIN-ECOG 5508 underpowered?

Biyun Qian, MD, PhD, and Li Xie, MD, both of Shanghai Jiao Tong University School of Medicine; and Xingwen Fan, MD, Fudan University Shanghai Cancer Center:

[ECOG-ACRIN 5508] might be underpowered as the result of inaccurate estimation of the OS improvement for both treatment and control arms. The study was designed to detect a median OS improvement from 12 months (bevacizumab arm) to 16 months (pemetrexed arm or combination arm), corresponding to a hazard ratio of 0.75. However, a previous phase II study [by Patel et al.] reported a median OS for the combination of pemetrexed and bevacizumab maintenance therapy in nonsquamous NSCLC of 14.1 months.

We do not agree with the author that post-study treatment does not affect the ability to assess the impact of maintenance therapy on OS in the first-line maintenance therapy setting ... When OS is used as the primary endpoint in first-line maintenance therapy, the effects of post-progression treatment should be considered. We believe the information of treatment choice after disease progression during maintenance therapy is essential to help us understand the whole study.

[The ECOG-ACRIN 5508 investigators] might need a data analysis check for median follow-up time in this study ... a median follow-up of 50.6 months -- with a median OS of 14.4 to 16.4 months and a survival rate lower than 20% at 48 months -- is impossible.

Ramalingam and Dahlberg: Xie et al. raise questions about the statistical assumptions used for the ECOG-ACRIN 5508 study; in doing so, they cite a phase II study [by Patel et al.] that was not designed to evaluate maintenance therapy. Our statistical assumptions were based on the data from the ECOG 4599 trial that demonstrated improvement in [OS] with the addition of bevacizumab to chemotherapy.

ECOG-ACRIN 5508 did not specify post-progression therapies, and those therapy decisions were up to the treating physicians. Although imbalances in post-study therapy between one treatment arm and the other could influence the survival outcomes, no specific attribute of this study would have favored such imbalances. The fact that the results of the ECOG-ACRIN 5508 study are consistent with prior reports ... argues against a random bias introduced by post-study therapies.

The point ... regarding the median follow-up reported in the ECOG-ACRIN 5508 trial is inaccurate. We used the reverse Kaplan-Meier method to estimate a median follow-up time of 50.6 months ... the censoring distribution and length of follow-up displayed in the Kaplan-Meier estimate of [OS] support this estimate, particularly when one considers the 8-year total study duration and the protocol-specified 5 years of survival follow-up for all patients.

Read the Guideline update here.