ODAC Votes STARGLO Trial Not Applicable to US Patients With DLBCL

Lymphoma: © David A Litman - www.stock.adobe.com

The FDA’s Oncologic Drug Advisory Committee (ODAC) has voted 8 to 1 that findings from the phase 3, confirmatory STARGLO trial (NCT04408638) evaluating glofitamab-gxbm (Columvi) in addition to gemcitabine and oxaliplatin (GemOx) for the treatment of relapsed/refractory diffuse large B-cell lymphoma (DLBCL) are not applicable to the US patient population.

STARGLO was a global trial, with 9% of patients enrolled in the US and 48% in Asian regions, leading the FDA to question whether the results were sufficiently applicable to a US patient population.

“We’re sitting here at the US FDA, and 25 patients were enrolled from the US,” said Daniel Spratt, MD, ODAC member and chairman and professor of radiation oncology at University Hospital Seidman Cancer Center, when explaining the reason why he voted no.

“Another word I kept coming back to was ‘sufficient,’ that sponsors should plan to enroll a sufficient number of US participants,” Ravi A. Madan, MD, ODAC member and senior clinician at the National Cancer Institute, when commenting on his no vote. Based on the US enrollment, Madan concluded that “we can’t know if this therapy may be effective.”

About the STARGLO Trial

Glofitamab, the bispecific CD20-directed CD3 T-cell engager, was granted accelerated approval by the FDA on June 15, 2023, for the treatment of adult patients with relapsed/refractory DLBCL not otherwise specified or arising from follicular lymphoma, after 2 or more prior lines of therapy. This approval was based on findings from the NP30179 study (NCT03075696).

STARGLO was designed as a confirmatory trial.² Patients were eligible if they had received at least 1 prior line of therapy and were not eligible for autologous stem cell transplant. A total of 274 patients were enrolled and randomized 2:1 to receive 12 cycles of glofitamab plus 8 cycles of GemOx (n = 183) or 8 cycles of rituximab (Rituxan) and 8 cyles of GemOx (n = 91).

The primary end point was overall survival (OS), and key secondary end points included progression-free survival (PFS), best complete response (CR), and duration of CR.

STARGLO met the end points of OS, PFS, and best CR. Glofitamab plus GemOx led to a median OS of 25.5 months (95% CI, 18.3-not estimable) compared with 12.9 plus with rituximab plus GemOx (range,7.9-18.5; HR, 0.62; 95% CI, 0.43-0.88). The median PFS by independent review committee (IRC) was 13.8 months (95% CI, 8.7-20.5) in the glofitamab arm vs 3.6 months (range, 2.5-7.1; HR, 0.40; 95% CI, 0.28-0.57), and the IRC-assessed CR rate was 58.5% (95% CI, 51.0%-65.7%) with glofitamab vs 25.3% (range, 168%-35.5%).

Notably, however, efficacy differed between Asian and non-Asian regions. In the Asia-Pacific region, the HR for OS was 0.41 (95% CI, 0.27-0.64) compared with 1.09 (range, 0.54-2.18) in Europe and 2.62 (range, 0.56-12.34) in North America. Further, HR for PFS was 0.27 (95% CI, 0.17-0.42) in the Asia-Pacific region vs 0.84 (range, 0.44-2.59) in Europe and 2.25 (range, 0.48-10.54) in North America.¹

Significantly, patients in the Asia region comprised 48% of the overall population (n = 161), leading the FDA to question whether the results are robust and can be applied to the US patient population. Nicole Gormley, MD, director, Division of Hematologic Malignancies II, Officer of Oncologic Diseases, Office of New Drugs at the FDA, who delivered the opening remarks for the FDA, highlighted that in global drug development, patients from the US typically comprise 20% of the overall patient population, compared with only 9% in STARGLO.

Applicant’s Stance

Jeremy Abramson, MD, professor of medicine at Harvard Medical School and director of the Jon and Jo Ann Hagler Center for Lymphoma at Massachusetts General Hospital, provided background on DLBCL and the unmet needs on behalf of the applicant, Genentech. He explained that about 40% of patients with DLBCL have disease progression following first-line therapy, and cellular therapy like chimeric antigen receptor (CAR) T-cell therapy is not a treatment option for about 75% of patients in the US. Using claims records from private and Medicare insurance from 2024, Abramson noted that 7% of patients received CAR T-cell therapy and 5% received transplant, leaving 88% of patients with DLBCL with a noncurative treatment option.

Michelle Boyer, PhD, global head of lymphoma and chronic lymphocytic leukemia development at Genentech, discussed the efficacy and safety findings from STARGLO. She explained that the observed regional differences were unexpected, and there was no apparent pharmacokinetic reason why this would be the case. Boyer presented analyses to identify why glofitamab performed differently across regions.

Boyer explained that there was a clear imbalance in high-risk features between treatment arms, with a higher proportion of patients with primary refractory disease—a poor prognosticator with worse outcomes—in the North American glofitamab cohort. All deaths in the North American region occurred in patients with primary refractory disease.

Exploring the patient subgroups, Boyer showed that benefits with glofitamab were consistent regardless of age, refractory disease status, previous lines of therapy, and history of primary refractory disease.

Looking at safety, Boyer noted that the safety risks with the glofitamab regimen were “aligned with clinical expectations and supported by established management guidelines.” Specifically, cytokine release syndrome [CRS], which was reported in 44.2% of patients in the glofitamab arm and a known risk associated with glofitamab, is also associated with other lymphoma therapies. A higher incidence of infection (55.2% vs 29.5%) was also reported in the glofitamab arm. Boyer commented that this risk can be mitigated through a proactive approach, which is routinely applied by physicians managing similar therapies.

Boyer highlighted the significance of the STARGLO study.

“This was the first phase 3 study to demonstrate an overall survival benefit in patients with relapsed or refractory DLBCL who are ineligible for transplant, a major milestone for this high-risk population,” Boyer said.

FDA’s Stance

Lymphomas: © David A Litman - stock.adobe.com

The FDA highlighted discrepancies in regional demographics in the STARGLO trial. Patients in Asian regions were overall younger, with a median age of 62 vs 71 in non-Asian regions, and most patients (80%) in non-Asian regions were White, with only 2% Black patients and 13% unknown.

Additionally, when looking into why patients in Asian regions did not undergo transplant, it was identified that the primary reason was patient refusal, with 65% (n = 80) of patients in Asian regions refusing transplant compared with 7% (n = 10) in non-Asian regions. This served to confound the findings as it did not allow for a complete analysis of patient fitness.

The FDA also identified issues with the applicant’s discussion of new antilymphoma therapy, explaining that new therapy would typically be administered after PFS data is gathered, so it should not affect PFS and does not affect response rate results. The standards of care for new antilymphoma therapy also differ in the US from other regions, potentially confounding results.

Other factors that the FDA assessed that could throw the STARGLO findings’ robustness into question included cell of origin variations between regions, exposure differences due to body weight, quality verification of locally sourced gemcitabine and oxaliplatin, delayed regional enrollment, and safety profile differences.

ODAC’s Discussion and Decision

During the clarifying questions, Gormley cautioned on cross-trial comparisons to interpret these findings.

“We should be able to rely on the results from within the trial, because that’s controlled by randomization. If we have to rely on external data to interpret the trial, that really calls into question the robustness and interpretability of that [data],” Gormley said.

“…Cross-trial comparisons are fraught…but if we’re going to do it, we want to use end points that are incontrovertible, I would suggest that, as compared to assessed response, mortality is an indisputable end point,” said Charles Fuchs, MD, MPH, senior vice president and global head of oncology and hematology drug development at Genentech/Roche, in response to Gormley.

After the discussion portion, eight of 9 members of the ODAC voted that the STARGLO findings could not be translated reliably to a US population.

“When we think about enrolling to a US population, we're trying to account for any biological differences that we see,” said Christopher Lieu, MD, ODAC member and co-director of gastrointestinal medical oncology at the University of Colorado. “We have to be very careful…because we can’t make any firm conclusions, so then the question becomes, are the results generalizable?”

“It boils down to one thing: we don’t know,” Lieu added.

The ODAC also raised concern with the safety of glofitamab and whether it is truly as accessible as the applicant represented.

“We all focused on the 25 patients that were enrolled in North America, with 11 sites activated but 50-plus approached with this study. Several declined, and so that still bothers me about what were the reasons for decline. I don’t believe it’s all just COVID-related,” said Heidi McKean, MD, ODAC member and oncologist at Avera Health in Sioux Falls, South Dakota. “I come from a state of 400,000 people, and there’s 1 hub that does CAR T and transplant. Quite frankly, even if there’s a 40% CRS risk, they’d still have to come to our main hub to get treatment. We just simply don’t have enough oncologists and hematologists out there in the community…so we as community oncologists need to make sure this is safe…and 25 patients is tough to make that call.”

“This should be an eye-opener for all of us in the future to start with the representative patient population to be included in the clinical trials,” said Ajay Nooka, MD, MPH, FACP, ODAC member and hematologist at Winship Cancer Center of Emory University.

REFERENCES:

1. Meeting of the Oncologic Drug Advisory Committee. FDA. May 20, 2025. Accessed May 20, 2025. https://tinyurl.com/5n7mwvjt

2. Abramson JS, Ku M, Hertzberg M, et al. Glofitamab plus gemcitabine and oxaliplatin (GemOx) versus rituximab-GemOx for relapsed or refractory diffuse large B-cell lymphoma (STARGLO): a global phase 3, randomised, open-label trial. Lancet. 2024;404(10466):1940-1954. doi:10.1016/S0140-6736(24)01774-4