Injecting nivolumab (Opdivo) directly into precancerous oral lesions led to a reduction in lesion size and allowed some patients to avoid surgery, according to research from a phase I clinical trial presented at the American Association for Cancer Research (AACR) 2026 annual meeting, held April 17-22.
About 5 percent of the general population has precancerous lesions in their mouths that can carry a 1 percent to 36 percent risk of developing oral cancer, depending on the extent of the dysplasia (that is, how abnormal the cells in the lesion are) and other factors, explained presenter Moran Amit, MD, PhD, a surgeon and assistant professor of cancer at the University of Texas. Since there are no reliable biomarkers to predict the risk of progression, many patients usually undergo surgical excision of their lesions, but this procedure is associated with high morbidity, he said.
“The mouth is the main conduit for so many different functions, such as speaking, eating, drinking and breathing. Think about the last time you had a sore spot in your mouth, how debilitating that was. Now imagine a patient who has to undergo surgery in multiple places in their mouth, over and over again as lesions recur,” said Amit, noting that 60% of patients relapse. after surgery it can reach 40%.
Every time a patient has to undergo surgery, they lose volume in their oral cavity, most often in the tongue. Once you lose a certain amount of your tongue, you can no longer articulate, you can’t swallow effectively. Due to precancerous lesions, patients may lose their ability to speak and eat. The goal of our study was to find a way to relieve patients of this often debilitating surgery.”
Moran Amit, MD, PhD, surgeon and assistant professor, University of Texas MD Anderson Cancer Center
Previous research has shown that treatment with the immune checkpoint inhibitor nivolumab could reduce the size and risk of progression of precancerous oral lesions, but this treatment, which is given systemically via intravenous infusion, comes with serious toxicities, Amit said.
“Although effective, systemic nivolumab can lead to toxicities that would not be acceptable for patients who do not yet have cancer,” he noted.
Amit and colleagues suggested that injecting a lower dose (2% to 4% of the intravenous dose) of nivolumab directly into the oral lesion could effectively treat the lesion without causing the systemic toxicities associated with intravenous nivolumab administration.
To test this hypothesis, they conducted a phase I clinical trial to assess the safety and efficacy of intratracheal nivolumab.
The trial enrolled 29 patients with at least one histologically confirmed, untreated premalignant oral lesion that was at high risk of progression to oral cancer due to the size, location, morphology, or extent of the malformation, or the patient’s age or medical history. More than half of the lesions were located on the tongue. Fifteen patients had lesions with high-grade (moderate or severe) dysplasia. The remaining 14 patients had low-grade (mild) dysplasia.
Patients received either 10 mg or 20 mg of nivolumab injected directly into one of their oral lesions every three weeks for a total of four cycles. The trial protocol allowed treatment of only one lesion per patient in order to assess whether the effects of intralesional nivolumab would be systemic or confined to the injection site. In cases where patients had more than one lesion, the largest lesion was selected for treatment.
At a median follow-up of 14.5 months after the first injection, 25 of 29 patients (85%) had experienced a clinical response, defined as a reduction in lesion size. Lesion area was reduced by an average of 60%, and 19 patients had reductions greater than 50%. Clinical responses were observed for lesions that were high-grade and those that were low-grade at baseline. Twelve patients (41%) experienced histological deterioration of the treated lesion, and six patients had a complete pathological response, meaning that their treated lesion had no signs of dysplasia at the time of follow-up. Of the six patients with complete pathologic response, four had moderate dysplasia before treatment and two had mild dysplasia.
Twelve months after treatment, 82.13% of treated lesions continued to be cancer-free. For the six patients whose treated lesions developed into cancer, the progression was detected early and the lesions were surgically removed. None of the patients whose lesions did not progress required or chose to have surgical resection of the treated lesions during the follow-up time.
Nivolumab serum levels were consistently 10-fold lower than typically observed with systemic administration. No dose-limiting toxicities occurred with intralesional therapy. The most common side effects were fatigue, diarrhea and rash. Mild injection site reactions occurred in 40% of injections but resolved within 48 hours without intervention. Most adverse events were grade 1 or 2, but there was one case each of grade 3 diarrhea, grade 3 hyperglycemia, and grade 4 acidosis.
All but four enrolled patients completed all treatment cycles and follow-up. Patient symptoms, such as those related to swallowing, mouth and throat pain, voice, communication, taste, and eating, either improved or remained stable during treatment and follow-up, according to patient-reported outcomes. Patients reported greater enjoyment of life and increased physical activity after treatment compared to baseline.
The researchers also examined tissue samples from 23 patients to determine how intralesional nivolumab affected the immune microenvironment of treated and untreated lesions. They observed immune activation exclusively in the treated lesions, as indicated by greater infiltration of CD4+ T cells, CD8+ T cells, and activated dendritic cells, as well as immune cell interactions indicative of adaptive immune responses. Untreated lesions from the same patients showed no immunological changes, which, according to Amit, suggests that the intralesional administration effectively limited the action of nivolumab at the target sites.
“Our findings demonstrate that intralesional administration of nivolumab is safe, well-tolerated and results in efficacy rates unmatched by other non-surgical methods, which have allowed us to limit surgery for the majority of patients – removing pieces of their mouth, whether it’s their tongue, cheek, floor of mouth or palate. “Even if a patient ends up having surgery later, the average 60% reduction in lesion size from intralesional nivolumab means we can substantially minimize the amount of surgery they’ll need down the road, which hopefully translates to a much lower impact on their quality of life.”
Amit noted that the findings may have implications beyond oral lesions. “Many types of cancer are preceded by precursor lesions, such as those occurring in the skin, cervix or colon. Our results raise the possibility that local immunotherapy could be an effective strategy to intercept these precancerous lesions as well.”
Limitations of the study include the single-arm design, the short follow-up time, and the fact that the study was not statistically powered to assess efficacy.
The study was supported by the Cancer Prevention and Research Institute of Texas.
