ODAC Votes Against Risk:Benefit Profile of Frontline Anti–PD-1 Therapy in Metastatic/Unresectable ESCC With a PD-L1 Expression Under 1

US FDA

The FDA’s Oncologic Drugs Advisory Committee (ODAC) voted 11 to 1 with 1 abstention against the risk:benefit profile of anti–PD-1 antibodies in the first-line treatment of patients with metastatic or unresectable esophageal squamous cell carcinoma (ESCC) with a PD-L1 expression of less than 1.¹

The vote was preceded by a discussion of the following question:

DISCUSSION: FDA would like the committee to discuss the risk and benefits of the treatment with anti PD-1 antibodies for the first line treatment of patients with metastatic or unresectable esophageal squamous cell carcinoma with PD-L1 expression < 1.²

“I voted no,” Lori Dodd, PhD, of the National Institute of Health in Bethesda, Maryland, said following the vote. “There was no evidence of benefit [with PD-L1 inhibitors] in [patients with PD-L1 expression lower than 1], and in particular, there was evidence of a clear treatment by PD-L1 status interaction…There needs to be more done. I also tried to approach this [decision] as: If we were presented with these data de novo today, would we approve this in this group? There’s just simply not evidence to suggest that it would provide a benefit in this subgroup.”

During the meeting, representatives from Merck Sharp and Dohme, Bristol-Myers Squibb, and BeiGene, presented data from the phase 3 KEYNOTE-590 (NCT03189719), CheckMate 648 (NCT03143153), and RATIONALE-306 (NCT03783442) trials, respectively.¹

“Although the OS results were statistically significant for the anti–PD-1–containing arm in all 3 trials…in subgroup analyses, the estimates for the treatment effect appear marginally or not favorable in patients with PD-L1 less than 1 tumors and intermediate in patients with PD-L1 less than 10 tumors,” Sandra Casak, MD, acting clinical team leader of Gastrointestinal Malignancies at the FDA, stated during the meeting’s FDA Introductory Remarks. “Although these results are exploratory, and uncertainty exists for each trial, the data do not appear to support the use of anti–PD-1 drugs in patients with PD-L1 less than 1 tumors, and benefit appears to be of higher magnitude in patients with PD-L1 10 or higher–expressing tumors…PD-1 monoclonal antibody treatments have toxicity, and benefit to patients relies on efficacy weighing against that risk. In PD-L1–low populations, efficacy has come into question, and with it, whether favorable risk:benefit remains for those patients with tumors with PD-L1 less than 1.”

KEYNOTE-590

In March 2021, the FDA approved pembrolizumab (Keytruda) in combination with platinum- and fluoropyrimidine-based chemotherapy for the treatment of patients with metastatic or locally advanced esophageal or gastroesophageal carcinoma who are not eligible for surgical resection or definitive chemoradiation, based on findings from KEYNOTE-590.³

The trial enrolled patients with treatment-naive, locally advanced, unresectable or metastatic esophageal adenocarcinoma, ESCC, or esophagogastric junction Siewert type 1 adenocarcinoma with measurable disease and an ECOG performance status (PS) of 0 or 1, regardless of PD-L1 status.⁴ Patients were randomly assigned 1:1 to receive either pembrolizumab plus chemotherapy or placebo plus chemotherapy. Among patients in the pembrolizumab arm, 86% had a PD-L1 combined positive score (CPS) of at least 1, and 50% had a PD-L1 CPS of at least 10. These respective rates were 87% and 52% in the placebo arm.

Overall survival (OS) and progression-free survival (PFS) served as the trial’s dual primary end points. OS was assessed in patients with ESCC with a PD-L1 CPS of at least 10, all patients with ESCC, all patients with a PD-L1 CPS of at least 10, and all patients. PFS was assessed in all patients with ESCC, all patients with a PD-L1 CPS of at least 10, and all patients.

In the intention-to-treat (ITT) population, the median OS favored the pembrolizumab arm (n = 373) at 12.4 months (95% CI, 10.5-14.0) vs 9.8 months (95% CI, 8.8-10.8) with placebo plus chemotherapy (n = 376; HR, 0.73; 95% CI, 0.62-0.86; P < .0001). Median PFS also favored the pembrolizumab arm at 6.3 months (95% CI, 6.2-6.9) vs 5.8 months (95% CI, 5.0-6.0) for the placebo arm (HR, 0.65; 95% CI, 0.55-0.76; P < .0001).

Among patients in the ESCC cohort, the median OS was 12.6 months (95% CI, 10.2-14.3) with pembrolizumab plus chemotherapy (n = 274) vs 9.8 months (95% CI, 8.6-11.1) with placebo plus chemotherapy (n = 274; HR, 0.72; 95% CI, 0.60-0.88; P = .0006). In these arms, the median PFS was 6.3 months (95% CI, 6.2-6.9) and 5.8 months (95% CI, 5.0-6.1), respectively (HR, 0.65; 95% CI, 0.54-0.78; P < .0001).

Among patients with ESCC with a PD-L1 CPS of less than 1, there was no significant difference in OS between the 2 arms (HR, 1.00; 95% CI, 0.54-1.85). However, among patients with ESCC with a PD-L1 CPS of at least 10, the OS data favored the pembrolizumab arm (HR, 0.57; 95% CI, 0.44-0.75).

Health-related quality-of-life outcomes remained stable during treatment and were generally similar between arms and across PD-L1 CPS subgroups. The safety profile of pembrolizumab plus chemotherapy was manageable and similar across PD-L1 CPS subgroups.

CheckMate 648

In May 2022, the FDA approved nivolumab (Opdivo) plus fluoropyrimidine- and platinum-containing chemotherapy and nivolumab plus ipilimumab (Yervoy) for the frontline treatment of adult patients with unresectable advanced or metastatic ESCC, regardless of PD-L1 status, based on findings from CheckMate 648.⁵

The study enrolled patients with treatment-naive unresectable, advanced, recurrent or metastatic ESCC who had measurable disease and an ECOG PS of 0 or 1.⁶ Patients were randomly assigned 1:1:1 to receive nivolumab at 240 mg plus fluorouracil and cisplatin every 4 weeks; nivolumab at 2 mg/kg every 2 weeks plus ipilimumab at 1 mg/kg every 6 weeks; or fluorouracil plus cisplatin alone every 4 weeks.

OS and PFS in the population of patients with a PD-L1 tumor proportion score (TPS) of at least 1% served as the trial’s coprimary end points. Secondary end points included OS in all patients, PFS in all patients, overall response rate (ORR) in the population of patients with a PD-L1 TPS of at least 1%, and ORR in all patients.

The addition of nivolumab to chemotherapy elicited a statistically significant and clinically meaningful OS benefit vs chemotherapy alone. The median OS was 13.2 months (95% CI, 11.1-15.7) in the nivolumab plus chemotherapy arm (n = 321) vs 10.7 months (95% CI, 9.4-11.9) in the chemotherapy alone arm (n = 324; HR, 0.74; 99.1% CI, 0.58-0.96; P = .0021). Among patients with a PD-L1 TPS of at least 1%, the median OS was 15.4 months (95% CI, 11.9-19.5) with nivolumab plus chemotherapy (n = 158) vs 9.1 months (95% CI, 7.7-10.0) with chemotherapy alone (n = 157; HR, 0.54; 95% CI, 0.37-0.80; P < .0001).

In an exploratory analysis of OS by PD-L1 CPS, outcomes favored nivolumab plus chemotherapy vs chemotherapy alone across CPS subgroups. However, the median OS with nivolumab plus chemotherapy was lowest in the subgroup of patients with a CPS of less than 1 at 9.86 months vs 12.09 months with chemotherapy alone (HR, 0.98; 95% CI, 0.50-1.95).

In the nivolumab plus ipilimumab arm (n = 325), the median OS was 12.75 months vs 10.71 months with chemotherapy alone (HR, 0.78; 95% CI, 0.65-0.95). In an exploratory analysis of OS by PD-L1 CPS, outcomes favored nivolumab plus ipilimumab over chemotherapy alone across all CPS subgroups except for patients with a CPS of less than 1, in whom the median OS was 11.47 months with the ipilimumab combination vs 12.09 months with chemotherapy alone (HR, 1.00; 95% CI, 0.65-0.95).

The safety profiles of nivolumab plus chemotherapy and nivolumab plus ipilimumab were consistent with the known safety profiles of the individual agents. The safety profiles were also consistent regardless of PD-L1 status.

RATIONALE-306

In September 2023, the FDA accepted for review a biologics license application (BLA) seeking the approval of tislelizumab-jsgr (Tevmbra) for the frontline treatment of patients with unresectable, recurrent, locally advanced, or metastatic ESCC, based on data from RATIONALE-306.⁷ The BLA has a target action date in the second half of 2024.

RATIONALE-306 enrolled patients with histologically confirmed ESCC that was stage IV unresectable at first diagnosis or unresectable, advanced recurrent, or metastatic.8 Patients were randomly assigned 1:1 to receive tislelizumab at 200 mg plus investigator’s choice of platinum-based chemotherapy (cisplatin or oxaliplatin) plus either fluoropyrimidine (5-fluorouracil or capecitabine) or paclitaxel once every 3 weeks or placebo in combination with the same chemotherapy regimens.

The primary end point was OS in the ITT population. Secondary end points included PFS, ORR, OS in the PD-L1 expression of at least 10% population, duration of response, and safety.

In the ITT population, the tislelizumab combination generated a clinically meaningful OS benefit vs placebo plus chemotherapy. The median OS was 17.2 months (95% CI, 15.8-20.1) in the tislelizumab arm (n = 326) vs 10.6 months (95% CI, 9.3-12.1) in the placebo arm (n = 323; HR, 0.66; 95% CI, 0.54-0.80; 1-sided P < .0001). Furthermore, an OS benefit was observed with tislelizumab plus chemotherapy vs chemotherapy alone across most subgroups of patients with a PD-L1 expression of at least 1%. However, OS outcomes did not favor the tislelizumab combination in the subgroup of patients with a PD-L1 expression of less than 5% (HR, 1.04; 95% CI, 0.74-1.46), although they did favor the combination in the subgroup of patients with a PD-L1 expression of at least 1% (HR, 0.66; 95% CI, 0.53-0.82).

At 3 years of follow-up, OS outcomes favored the tislelizumab arm in the ITT population (HR, 0.70; 95% CI, 0.59-0.83) and across all subgroups of patients with a PD-L1 expression of at least 1%. In the subgroup of patients with a PD-L1 expression of at least 1%, the median OS was 17.2 months (95% CI, 15.8-20.1) in the tislelizumab arm vs 10.6 months (95% CI, 9.3-12.0) in the placebo arm.

Tislelizumab plus chemotherapy was associated with a manageable safety profile.

FDA Position

In a presentation on behalf of the FDA, Geetika Srivastava, MD, MSPH, a clinical reviewer, highlighted data on behalf of the regulatory agent from a patient-level analysis of an FDA ESCC-modified population of patients from each of the 3 trials. This population included 548, 629, and 648 patients from KEYNOTE-590, CheckMate 648, and RATIONALE-306, respectively, all of whom had ESCC disease histology.¹

Among all patients with ESCC enrolled in these 3 trials, the FDA analysis demonstrated an OS benefit with PD-L1 inhibition vs the comparator arms. In this analysis of KEYNOTE-590, the OS data favored pembrolizumab plus chemotherapy over placebo plus chemotherapy (HR, 0.72; 95% CI, 0.59-0.87). However, in the subgroup of patients with a PD-L1 score of less than 1, the OS outcomes were not significantly different between the 2 arms (HR, 1.00; 95% CI, 0.54-1.85). In this analysis of CheckMate 648, the OS data favored nivolumab plus chemotherapy over chemotherapy alone (HR, 0.73; 95% CI, 0.60-0.88). However, in the subgroup of patients with a PD-L1 score of less than 1, the OS outcomes were not significantly different between the 2 arms (HR, 0.93; 95% CI, 0.46-1.91). In this analysis of RATIONALE-306, the OS data favored tislelizumab plus chemotherapy over placebo plus chemotherapy (HR, 0.68; 95% CI, 0.56-0.82). However, in the subgroup of patients with a PD-L1 score of less than 1, the OS outcomes did not favor the tislelizumab arm (HR, 1.34; 95% CI, 0.73-2.46).

“In spite of the heterogeneity amongst different trials, use of different PD-L1 assays and testing algorithms, as well as different prespecified PD-L1 cutoffs for OS, in this FDA patient-level analysis, over 3 independently conducted trials, it appears that the overall efficacy of anti–PD-1 [therapy] in this setting is driven predominantly by PD-L1–high subgroups, as defined by each study and regardless of the method used to determine PD-L1 expression, there is replication of results over 3 trials and a consistent pattern suggesting a lack of benefit in PD-L1 less than 1 tumors,” Srivastava said.

References

1. September 26, 2024, Meeting of the Oncologic Drugs Advisory Committee (ODAC). FDA. https://www.youtube.com/live/ELA3JDqtcFw
2. Oncologic Drugs Advisory Committee (ODAC) Meeting. Final questions. Accessed September 26, 2024. https://www.fda.gov/media/182135/download
3. FDA approves pembrolizumab for esophageal or GEJ carcinoma. FDA. March 22, 2021. Accessed September 26, 2024. https://www.fda.gov/drugs/resources-information-approved-drugs/fda-approves-pembrolizumab-esophageal-or-gej-carcinoma
4. Oncologic Drugs Advisory Committee (ODAC) Meeting. PM Merck presentations. Accessed September 26, 2024. https://www.fda.gov/media/182215/download
5. US Food and Drug Administration approves two Opdivo (nivolumab)-based regimens as first-line treatments for unresectable advanced or metastatic esophageal squamous cell carcinoma. News release. Bristol Myers Squibb. May 27, 2022. Accessed September 26, 2024. https://news.bms.com/news/details/2022/U.S.-Food-and-Drug-Administration-Approves-Two-Opdivo-nivolumab-Based-Regimens-as-First-Line-Treatments-for-Unresectable-Advanced-or-Metastatic-Esophageal-Squamous-Cell-Carcinoma/default.aspx
6. Oncologic Drugs Advisory Committee (ODAC) Meeting. PM BMS presentations. Accessed September 26, 2024. https://www.fda.gov/media/182212/download
7. BeiGene announces positive regulatory updates in Europe and the US after recently regaining global rights for Tevimbra. News release. BeiGene, Ltd. September 19, 2023. Accessed September 26, 2024. https://ir.beigene.com/news/beigene-announces-positive-regulatory-updates-in-europe-and-the-u-s-after-recently-regaining-global-rights/5e4692d6-d726-4989-a659-49dc97d4d7ed/
8. Oncologic Drugs Advisory Committee (ODAC) Meeting. PM BeiGene presentations. Accessed September 26, 2024. https://www.fda.gov/media/182214/download