Niraparib/Dostarlimab Plus CRT Displays Efficacy but Has Toxicity Concerns in Locally Advanced HNSCC

The addition of niraparib (Zejula) and dostarlimab-gxly (Jemperli) to definitive chemoradiotherapy (CRT) improved 1-year progression-free survival (PFS) vs historical control in cisplatin-fit patients with locally advanced head and neck squamous cell carcinoma (HNSCC), but not in cisplatin-unfit patients who received the combination with radiotherapy alone, according to findings from the phase 1b/2 RADIAN trial (NCT05784012) presented at the 2026 ESMO Congress.¹ Additionally, treatment-related toxicity led to frequent discontinuation in both cohorts.

Findings showed that at a median follow-up of 17 months (95% CI, 16-25) in cohort A (cisplatin fit; n = 17) and 13.6 months (95% CI, 7.8-16.4) in cohort B (cisplatin unfit; n = 17), the 1-year PFS rates were 88.2% (95% CI, 74.2%-100.0%) for cohort A and 60.6% (95% CI, 40.2%-91.4%) in cohort B. Compared with historical controls, the HRs were 0.19 (P = .01) in cohort A and 0.72 (P = .19) in cohort B. The prespecified efficacy criterion, defined as a 1-year PFS rate of 75.9% vs a historical 65% in cohort A and 68.2% vs 55% in cohort B, was met only in cohort A.

Regarding safety, adverse effects (AEs) led to niraparib dose reductions and interruptions in 65% and 71% of patients, respectively, in cohort A. These respective rates were 67% and 88% in cohort B. In cohort A, treatment-related AEs (TRAEs) led to discontinuation of maintenance dostarlimab and/or niraparib in 41% of patients, with the most common being upper respiratory tract edema (12% and grade 2 pneumonitis (18%). In cohort B, TRAEs led to permanent discontinuation of dostarlimab or niraparib in 59% of patients, including 24% during the concurrent phase. The most common TRAEs leading to discontinuation in this group were mucositis (12% and odynodysphagia (6%).

"Dostarlimab plus niraparib showed promising efficacy in patients [who were] fit and unfit for cisplatin with high disease burden, but with substantial treatment-related toxicity, especially in the maintenance phase, resulting in treatment discontinuation," said lead study author Marc Oliva, MD, PhD, of the Institut Català d'Oncologia in Barcelona, Spain.

How was the RADIAN trial designed?

The nonrandomized, noncomparative study, sponsored by the Spanish Group for the Treatment of Head and Neck Cancer (TTCC), enrolled patients at least 18 years of age with stage III to IVB oropharyngeal, hypopharyngeal, or laryngeal squamous cell carcinoma, a centrally confirmed PD-L1 combined positive score of at least 1, and an ECOG performance status of 0 or 1.^1,2

All patients received 1 dose of dostarlimab at 500 mg followed by 14 days of niraparib at 200 mg to 300 mg daily before radiotherapy (70 Gy in 35 fractions), given with cisplatin at 100 mg/m² every 3 weeks in cohort A or with daily niraparib in cohort B. Maintenance niraparib plus dostarlimab at 500 mg every 21 days followed for up to 14 cycles.

The primary end point was 1-year PFS rate. Safety and tolerability was a secondary end point, along with circulating tumor DNA (ctDNA) dynamics and other efficacy outcomes such as overall response rate (ORR), disease control rate, and others.

Compared with cohort A, patients in cohort B were older (median age, 73 vs 65 years) and more often had stage IVB disease (47% vs 12%).

What other efficacy data were reported?

The ORR 12 weeks after radiotherapy was 88.2% in both cohorts, with complete response rates of 65% in cohort A and 59% in cohort B. Median PFS was not reached in cohort A and was 15.33 months (95% CI, 8-not reached) in cohort B; median overall survival was not reached in either cohort. The 1-year locoregional failure rate was 0% in cohort A and 7.6% (95% CI, 0.4%-30%) in cohort B.

All patients completed neoadjuvant therapy, and 88% and 100% of patients in cohort A and cohort B, respectively, received the full radiotherapy dose. Maintenance was initiated by 82% and 76% of patients, respectively, but was completed at respective rates of 57% and 8%, with a median of 13 (range, 3-14) vs 3 (range, 1-5) dostarlimab cycles. In cohort A, 2 patients died due to AEs; in cohort B, 5 patients died (2 due to disease, 3 due to AEs).

What other data did the safety analysis show?

The most common grade 3 or higher TRAEs in cohort A comprised neutropenia (71%), dysphagia (29%), lymphopenia (29%), anemia (24%), mucositis (24%), dysgeusia (18%), asthenia (12%), radiodermatitis (12%), and thrombocytopenia (12%). One grade 5 TRAE (febrile neutropenia related to cisplatin and niraparib) occurred.

In cohort B, the most common grade 3 or higher TRAEs included mucositis (59%), dysphagia (35%), anemia (18%), asthenia (18%), odynophagia (18%), anorexia (12%), pharyngeal necrosis (12%), and thrombocytopenia (12%). Two grade 5 TRAEs of aspiration pneumonia and radionecrosis were attributed to radiotherapy and possibly to niraparib

What are the next steps for research following RADIAN?

The investigators cautioned that the findings should be interpreted carefully given the single-arm design and small sample size.

In cisplatin-unfit patients, increased and prolonged radiation-mediated toxicity drove high discontinuation rates without a PFS gain. Further investigation of niraparib at reduced doses during the neoadjuvant and concurrent phases may be warranted, particularly in this population.

Tumor next-generation sequencing and ctDNA analyses for biomarker evaluation are ongoing.

References

1. Oliva M, Vidales Z, Cirauqui B, et al. TTCC-2022-01 RADIAN: niraparib and dostarlimab plus (chemo)radiotherapy in patients with locally advanced head and neck squamous cell carcinoma. Presented at: 2026 ESMO Congress; October 23-27, 2026; Madrid, Spain. Abstract LBA46.
2. Two-cohort study of niraparib and dostarlimab plus (chemo)radIotherapy in locally-advanced head and neck squamous cell carcinoma (RADIAN). ClinicalTrials.gov. Updated June 4, 2026. Accessed October 26, 2026. https://clinicaltrials.gov/study/NCT05784012