番茄社区

PHOENIX -- Use of systemic corticosteroids did not increase the risk of nonmelanoma skin cancer in a prospective study, contradicting previous evidence from several retrospective studies, investigators reported here.

Treatment with oral prednisone increased the hazard for squamous-cell carcinoma by 3% to 5% and the hazard for basal cell carcinoma by 7% to 26%, depending on the definition of corticosteroid use, according to a presentation at the Society for Investigative Dermatology meeting.

All of the values were associated with overlapping confidence intervals that did not exclude the possibility of a chance finding.

"There is no consistent evidence that use of corticosteroids increases the risk of dermal malignancies," Ake T. Baibergenova, MD, of the University of Toronto, said in conclusion.

"In fact, corticosteroids are often used to manage the side effects of chemotherapy in cancer patients. There have also been some reports of direct antitumor activity of corticosteroids."

Several retrospective cohort studies have implicated corticosteroids in an increased risk of skin cancer, both squamous-cell and basal-cell carcinoma. For example, an analysis of a large pharmacy claims database in Denmark revealed hazard ratios of 1.52 for squamous-cell carcinoma and 2.45 for basal-cell carcinoma among patients who filled at least 15 prescriptions for corticosteroids versus patients who had no steroid prescriptions (J Natl Cancer Inst 2004; 96: 709-711).

Other studies had yielded hazard ratios ranging from 1.1 to 2.5, some of which had overlapping confidence intervals.

All of the earlier studies used retrospective data, said Baibergenova. To address that limiting factor, he and his colleagues analyzed data from the prospective Veterans Affairs Topical Tretinoin Chemoprevention trial.

The study involved 1,051 patients who had a recent history of at least two keratinocyte carcinomas. They were randomized to 0.1% topical tretinoin or placebo and followed for as long as six years.

The primary outcomes of the trial were time to first diagnosis of a squamous-cell carcinoma lesion and time to first diagnosis of a first basal-cell carcinoma lesion.

On the basis of pharmacy records, the investigators determined that 148 of the study population filled at least one prescription for oral prednisone during the trial. The records also showed that 63 of those 148 used the drug for 30 days or longer.

Prednisone was the only corticosteroid used by the study participants.

Baibergenova reported that 44% of study participants developed at least one basal-cell carcinoma lesion during the study, and 29% developed at least one squamous-cell lesion.

In a multivariable analysis, any use of oral prednisone was associated with hazard ratios of 1.07 for basal-cell carcinoma and 1.05 for squamous-cell carcinoma, neither of which achieved statistical significance.

The hazards changed slightly but remained nonsignificant in an analysis of patients who used oral prednisone for at least 30 days. The hazard ratios in that subgroup were 1.26 for basal-cell carcinoma and 1.03 for squamous-cell carcinoma.

Acknowledging limitations of the study, Baibergenova pointed out that investigators could not separate the effect of oral prednisone from the effect of the condition for which the prednisone was prescribed.

Additionally, the study involved only patients who had an increased risk of skin cancer, in the form of keratinocyte carcinoma.

"We found no evidence to support the existence of an association between use of oral prednisone and risk of basal cell carcinoma or squamous-cell carcinoma."

Comment