PHOENIX – Definitive radiation therapy with or without chemotherapy was less tolerated and less effective in HIV-positive patients with head and neck squamous cell carcinoma than in HIV-negative patients in a single-institution retrospective study of 71 HIV-positive patients.
"Head and neck squamous cell carcinoma with coexisting HIV remains a challenging clinical problem. ... Due to the advances in highly active antiretroviral therapy (HAART) – which prolongs HIV patients’ survival – the likelihood to develop HIV-related malignancy increases. It is of paramount importance to establish better-tolerated treatment strategies and regimens to improve tolerance, toxicity, and outcomes in this growing patient population," said Dr. Waleed Mourad, a radiation oncologist at Beth Israel Medical Center, New York.
The 71 HIV-positive patients with HNSCC were treated January 1997 through 2010. They had a median age of 34 years at the time of HIV diagnosis (range 25-50 years) with a median 11 years’ duration of seropositivity (6-20 years). Their median age at the time of radiation therapy was 51 years (32-72 years). All but one patient had squamous cell carcinoma, with the other having submandibular salivary duct carcinoma. That patient was treated with definitive surgery and received adjuvant radiation therapy without chemotherapy, Dr. Mourad noted.
Approximately one-third of the patients had cancer of the oropharynx (32%) and larynx (35%), and another 13% had cancer of the oral cavity. Other cancers among the patients included those located in the hypopharynx, nasopharynx, occult primary, and nasal cavity. American Joint Committee on Cancer 7th edition stages I-II, III and IVa/b were 22%, 27%, 51% respectively.
All patients were treated definitively with radiation therapy, with or without chemotherapy (cisplatin, carboplatin, or cetuximab). A total of 50 patients (70%) were on HAART during treatment, with a median CD4 count of 290 (range, 203-1,142). A median dose of 70 Gy (66-70) was delivered to the gross disease; high-risk neck 60-63 Gy; low-risk neck; and lateral retropharyngeal nodes 54 Gy. All fractions were given at the rate of 1.8-2 Gy/fraction. The median duration of treatment was 52 (49-64) days. A total of 12 patients (17%) underwent planned neck dissection for N3 disease.
In all, local control was achieved in 69% (49) and locoregional failure occurred in 31% (22). By site, local control rates were 70% (16) for the oropharynx, 76% (19) for the larynx, 78% (7) for the oral cavity, 0% (of 5) for hypopharynx, 67% (2) for the nasopharynx, 100% (3) for occult primary, 50% (1) for the nasal cavity, and 100% (1) for the submandibular duct.
Thus, locoregional failure rates ranged from 100% (all five patients with hypopharyngeal cancer) to 0 (for the three patients with occult primary cancer and the one with submandibular duct cancer). Seven patients developed second primary cancers, Dr. Mourad reported.
After a median follow up of 47 months (7-140), there were no fatalities related to radiation or chemotherapy. Treatment breaks in excess of 10, 7, and 5 days occurred in 6%, 13%, and 14% of patients, respectively. Acute dysphagia and odynophagia grades 1, 2, and 3 occurred in 31%, 52%, and 17%, respectively. All of the patients experienced dysgeusia, dysphagia, and xerostomia of grades 1-3. Acute desquamation of the skin of grades 1, 2, and 3 occurred in 66%, 20%, and 14%.
One patient was hospitalized for grade 4 mucositis, dysphagia, and fever. Another developed osteoradionecrosis during concurrent chemoradiotherapy. Late dysphagia of grades 1-4 occurred in 46%, 28%, 15%, and 11%. Xerostomia of grades 1-3 occurred in 45%, 32%, and 23%, respectively, he said.
At 4 years, locoregional control was 69% and overall survival 55%. In an interview, Dr. Mourad noted that at his institution, the comparable rates for non-HIV patients were 85%-90% and 75%-85%, respectively.
The median CD4 count prior to chemotherapy was 370, and viral load was undetectable. The CD4 count dropped as chemoradiotherapy progressed to a nadir at 11 weeks, at which point it began to rise again and viral load decreased.
There was a statistically significant relationship between locoregional control and the duration of radiation therapy (P less than .001), and positive trends with weight loss of greater than 10% and absence of second malignancy. However, multivariate analysis did not show any statistically significant relationship because of the relatively small sample size with diverse subsites, Dr. Mourad noted.
"We believe that our data provide important information regarding treatment outcomes for newly diagnosed head and neck cancer with HIV. This is particularly important given the growing population of patients who live with HIV long term because of effective antiretroviral maintenance treatments," Dr. Mourad said in an interview.