The study, led by researchers at Hubei Cancer Hospital in Wuhan and published in Frontiers in Oncology, focused on a particularly vulnerable group: lung cancer patients who had previously received immune checkpoint inhibitors (ICIs) and were scheduled for thoracic radiotherapy at doses of at least 50 Gy. This sequential treatment paradigm, in which immunotherapy is followed by radiation, has improved survival outcomes in recent years, but it comes with a hidden cost. Prior ICI therapy can prime the pulmonary immune microenvironment, expanding effector T cells and elevating baseline levels of proinflammatory mediators. When radiation is then delivered to the chest, the primed lung responds with an exaggerated inflammatory and fibrotic reaction. Recent clinical data indicate that symptomatic RILI affects more than 45 percent of ICI-pretreated lung cancer patients undergoing thoracic radiotherapy, yet no approved preventive drug exists, leaving clinicians to rely on reactive high-dose corticosteroids once symptoms appear.

The researchers reasoned that if the biological cascade driving RILI could be intercepted early, the injury might be blunted before irreversible damage occurs. Their target was the transforming growth factor-beta (TGF-β) signaling pathway, the central molecular circuit controlling both the inflammatory and fibrotic phases of radiation injury. Radiation triggers excessive production of reactive oxygen species and activates TGF-β, which gradually drives fibroblast activation and deposition of collagen, culminating in fibrotic damage that is notoriously difficult to reverse. Pirfenidone, an orally administered antifibrotic agent with high bioavailability that is officially approved for idiopathic pulmonary fibrosis, suppresses TGF-β production and had already shown promise in preclinical models of radiation-induced pulmonary fibrosis, as well as in earlier trials treating established grade 2 to 3 RILI. The question was whether starting the drug before radiation could prevent the cascade altogether.

To answer it, the team designed a single-center, single-arm, open-label phase II trial using Simon’s optimal two-stage statistical design, a framework that allows early termination for futility while minimizing the number of patients exposed to an ineffective treatment. Between June 2022 and October 2023, 52 patients were screened and 39 were enrolled. All had histologically confirmed lung cancer, had received at least two cycles of an anti-PD-1 or anti-PD-L1 antibody within six months of planned radiotherapy, and maintained adequate baseline pulmonary function, with requirements including an FEV1 above 1.0 liter and a diffusing capacity of the lungs for carbon monoxide (DLCO) above 40 percent of predicted. Patients with a history of grade 2 or higher checkpoint inhibitor pneumonitis, severe interstitial lung disease, or significant hepatic or renal impairment were excluded.

The prophylactic regimen was carefully structured around tolerability. Oral pirfenidone began one week before radiotherapy at 600 mg per day, was up-titrated through 1200 mg per day in the second week, and then escalated to the target dose of 1800 mg per day, taken as 600 mg three times daily with meals. Treatment continued throughout the radiotherapy course and for four weeks afterward, a planned duration of roughly eleven weeks. Thoracic irradiation was delivered with modern conformal techniques, either intensity-modulated radiotherapy or volumetric modulated arc therapy, under strict lung dose-volume constraints including a V20 of at most 30 percent and a mean lung dose of no more than 15 Gy. To counter the assessment bias inherent in an open-label design, all RILI diagnoses and grades were adjudicated by an independent panel of at least two radiation oncologists and one pulmonologist, who reviewed symptoms and high-resolution CT imaging while actively excluding alternative causes such as infection, tumor progression, and cardiogenic pulmonary edema.

The primary endpoint was the cumulative incidence of grade 2 or higher RILI within six months of radiotherapy, benchmarked against a historical control rate of 40 percent derived from two contemporary studies of ICI-era lung cancer populations. The result crossed the trial’s predefined success threshold. Nine of 39 patients, or 23.1 percent (95 percent confidence interval 11.1 to 39.3 percent), developed grade 2 or higher RILI, a rate statistically significantly lower than the 40 percent control (one-sided exact binomial p = 0.0205). Severity was also constrained: only three patients (7.7 percent) experienced grade 3 events, and no grade 4 or 5 events occurred at all. Among those who did develop significant injury, the median time to onset was 9.1 weeks after radiotherapy, with events appearing between roughly five and seventeen weeks, consistent with the expected temporal window for radiation pneumonitis.

Secondary endpoints revealed a functional signature of protection. Mean DLCO, a measure of how efficiently oxygen diffuses from the lungs into the blood, declined by an average of 7.6 percent at three months and 6.5 percent at six months, figures the authors note are substantially lower than the 10 to 22 percent declines typically observed in patients with symptomatic pneumonitis. Forced vital capacity, a marker of lung volume, fell only 2.9 percent at three months and 1.7 percent at six months. The stratification was telling: patients who developed grade 2 or higher RILI lost an average of 14.6 percent of DLCO at three months, compared with just 5.4 percent among those who avoided significant injury, and declines exceeding ten percentage points on individual waterfall plots concentrated almost entirely in the patients who developed RILI. Exploratory subgroup analyses hinted that lung V20 of 20 percent or higher, concurrent chemotherapy, and male sex were associated with numerically higher injury rates, though none reached statistical significance with so few events.

One finding carries particular clinical weight: the history of prior checkpoint inhibitor pneumonitis. Among the five patients who had previously experienced even a grade 1 episode of immune-related pneumonitis, three, a striking 60 percent, went on to develop grade 2 or higher RILI despite prophylaxis, compared with 17.6 percent of patients without such a history (Fisher’s exact p = 0.070). Although limited by the small sample, this signal suggests that a prior CIP history marks a state of immune priming that may respond poorly to standard preventive antifibrotic therapy, and that these patients warrant closer surveillance and individualized protective strategies when thoracic radiotherapy is planned.

Tolerability proved workable in an oncology population already contending with the burdens of cancer treatment. Fatigue affected 43.6 percent of patients, nausea 41.0 percent, and appetite loss 33.3 percent, with the overwhelming majority of events graded as mild to moderate. Photosensitivity and rash, hallmark side effects of pirfenidone, were managed with sun-avoidance counseling and dermatological support. Overall, 84.6 percent of patients completed the full pirfenidone course, with a median relative dose intensity of 89.3 percent, and the radiotherapy completion rate reached 89.7 percent. Critically, no patient required interruption of radiotherapy or modification of concurrent chemotherapy because of the study drug, and no patient received ICI during the treatment window. The daily 1800 mg dose therefore proved feasible to integrate into routine outpatient care without compromising standard anti-tumor treatment.

The authors are careful to frame these results as hypothesis-generating rather than definitive. The single-arm, open-label design, while standard for phase II oncology, invites selection and evaluation bias when compared against historical cohorts, and unmeasured confounders, including the continuous refinement of high-precision conformal radiotherapy, cannot be excluded. The permitted interval of up to six months between the last ICI dose and radiotherapy introduced heterogeneity in immune-priming status, and the exploratory comparison of intervals of 60 days or fewer versus longer, at 25.0 versus 22.2 percent RILI incidence, rested on too few events to be conclusive. Enrolled patients also had relatively preserved baseline lung function, which may limit extrapolation to those with severe respiratory dysfunction, and the absence of longitudinal biomarker data such as circulating TGF-β and interleukin-6 levels leaves the mechanistic relationship among pirfenidone, radiation, and the immune-activated tumor microenvironment incompletely explained. What the trial does deliver is proof of concept: a biologically grounded, statistically significant reduction in a feared complication, achieved with a well-characterized drug given before harm occurs rather than after. Confirmation in a randomized, multicenter phase III trial, ideally with prespecified stratification by ICI-to-radiotherapy interval and prior pneumonitis history, is the necessary next step toward making prophylactic antifibrotic therapy a routine part of thoracic oncology.

Subject of Research: Prophylactic pirfenidone for prevention of radiation-induced lung injury in ICI-pretreated lung cancer patients receiving thoracic radiotherapy

Article Title: Prophylactic pirfenidone to prevent radiation- induced lung injury in ICI-pretreated lung cancer patients: a phase II study

Article References: Huang, Y., Ma, J., Han, G., Bi, J., Pi, G., Li, Y., Peng, Y., Xiao, C., & Tuo, X. (2026). Prophylactic pirfenidone to prevent radiation- induced lung injury in ICI-pretreated lung cancer patients: a phase II study. Frontiers in Oncology, 16, Article 1876642. https://doi.org/10.3389/fonc.2026.1876642

Image Credits: AI Generated

DOI: 10.3389/fonc.2026.1876642

Keywords: pirfenidone, radiation-induced lung injury, lung cancer, immune checkpoint inhibitors, thoracic radiotherapy, pneumonitis, TGF-beta, antifibrotic, phase II trial, pulmonary function, DLCO, radiotherapy