Increased Incidence of Cancer Observed in HIV/hepatitis C Virus-Coinfected Patients Versus HIV-Monoinfected

Héctor Meijide; Sonia Pértega; Iria Rodríguez-Osorio; Ángeles Castro-Iglesias; Josefa Baliñas; Guillermo Rodríguez-Martínez; Álvaro Mena; Eva Poveda


AIDS. 2017;31(8):1099-1107. 

In This Article


Data Collection

All PLWH older than 18 years and above at Complexo Hospitalario Universitario de A Coruña (CHUAC) between 1993 and 2014 were included in the cohort. Chronic HCV coinfection was defined by the presence of a measurable viral load by PCR, without acute hepatitis. Epidemiological, demographic, clinical, immunovirological data were recorded. Chronic hepatitis B virus (HBV) infection was defined by the presence of positive hepatitis B surface antigen for more than 6 months. Patients in care were defined as at least two visits to the outpatient clinic and a measure of CD4+ cell count. Loss of care was defined by the nonappearance for reasons other than death.

All cancers in PLWH were obtained through the hospital-coding department. The method of selection was based on the review of the medical records of patients recognized by the encoding unit with a previous diagnosis of HIV infection and cancer. Malignancies registered during the follow-up in the clinical record were also included. Cancers were classified into two groups: ADC and NADC. Second malignancy was defined as the appearance, simultaneously or not, of a different tumor with malignant histology and the possibility of metastasis due to the first cancer was excluded. The end of the observation period was determined by date of cancer diagnosis, death or last follow-up visit for patients lost to follow-up, whichever occurred first. Patients with cancer were followed until their last regular clinical visit, death or lost to follow-up, to analyze the development of second malignancies.

Patients who are diagnosed of HCV infection during the follow-up contributed in person-years to the HIV-monoinfected group from HIV diagnosis to HCV diagnosis and to the HIV/HCV-coinfected group after. In coinfected patients who received anti-HCV treatment, the follow-up was censored when achieved sustained viral response (if applicable).

Ethical Considerations

The research protocol was reviewed and approved by the regional ethics committee (register code 2015/164). All clinical data were anonymized and de-identified prior to analysis and the identification numbers of the patients were blinded.

Statistical Analyses

A descriptive analysis was performed for all the variables recorded. Quantitative variables are reported as mean ± SD or median (interquartile range). Qualitative variables are expressed as frequencies and percentages.

A comparison of HIV-monoinfected and HIV/HCV-coinfected patients was performed and the percentages were compared using the χ2 test. Quantitative parameters were compared by means of Student's t test.

Cancer incidence rates were estimated for both ADC and NADC. Crude incidence rates were expressed as the number of cases per 100 000 person-years of follow-up, and the follow-up was determined from the date of HIV diagnosis to the end of the observation period. Cancer incidence rates in this cohort were then compared with that observed in the Spanish general population, computing the standardized incidence ratios (SIRs) for coinfection groups and their 95% confidence intervals, with the Byar's approximation of Poisson model. Age-SIRs were also determined for each group using the age strata of the reference population. For this purpose, cancer incidence data published for Spain in the GLOBOCAN 2012 statistics were used.[25] A comparison of mortality outcomes between the coinfection groups was evaluated using the χ2 test. Logistic regression analyses were also performed to predict the death rate after 1 year.

A competing risk approach was used to estimate the probability of cancer at different time points in the follow-up after HIV diagnosis. Death before cancer was considered a competing risk event. The death-adjusted cumulative incidence for the marginal probability of cancer was obtained and the cumulative incidences in the competing risk data were compared using the modified log-rank test.[26]

To compare the subdistribution hazard ratios (SHRs) for cancer occurrence between HIV-monoinfected and HIV/HCV-coinfected patients, multivariable analyses were conducted using modified Cox regression hazard models.[27] This analysis has been carried out for both ADC and NADC (all cancers and excluding HCC).

Data management and analyses were performed using SPSS, version 19.0 for Windows (IBM Corp., Armonk, New York, USA). The cumulative incidence in competing risk analyses was calculated using the cmprsk package of R (Gray B. cmprsk: subdistribution analysis of competing risks. Available from Accessed 2 February 2016). P < 0.05 (two-sided) was considered statistically significant.