Predictive factors for early and late local toxicities in anal cancer treated by radiotherapy in combination with or without chemotherapy.

BACKGROUND

The treatment of anal cancer is based on concomitant radiotherapy and chemotherapy and is associated with a nonnegligible rate of local severe toxicities that can strongly impair the quality of life.

OBJECTIVE

A retrospective analysis was performed to screen the following factors as potential predictive factors for local skin and digestive toxicities, and as potential prognostic factors for cumulative colostomy incidence: sex, age, tumor size, clinical T and N stage, circumferential extension, invasion of anal margin, HIV status, type of chemotherapy, and type of radiotherapy and dose delivered.

METHODS

One hundred five patients in our database treated between January 2000 and February 2010 met the eligibility criteria.

RESULTS

Median follow-up was 54.1 months (range, 1-133). Early and late severe local toxicities occurred in 33 patients (31.4%) and 18 patients (17.1%). The 5-year cumulative rate of colostomy was 26.6%. Predictive factors for local severe early toxicities were as follows: clinical stage III/IV (p = 0.01), no brachytherapy boost (p = 0.003), and use of chemotherapy (p = 0.01). Only brachytherapy retained its independence in multivariate analysis (OR = 4.8 (1.4-16.3), p = 0.01). Human immunodeficiency virus positivity (p = 0.04) was the only predictive factor for late toxicities in univariate analysis; it was linked independently to the occurrence of ulcer (OR = 0.1 (0.01-0.66), p = 0.01). Tumor size ≥4 cm (p < 0.001) and occurrence of grade 2 to 3 ulcers (p < 0.001) were correlated with greater cumulative colostomy incidence.

CONCLUSIONS

In this cohort, nonuse of brachytherapy was an independent predictive factor for local acute toxicity. Human immunodeficiency virus positivity was the only predictive factor for local late toxicities and strongly influenced the onset of ulcer.