If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
). These phase II studies aimed to determine whether oral nicotinamide, at different doses, reduced AKs in sun-damaged individuals.
Healthy, immune-competent volunteers with ≥4 palpable AKs (face, scalp and upper limbs) were recruited from Royal Prince Alfred Hospital Dermatology Clinics, Sydney, Australia. The study protocols (ACTRN12609000490279; ACTRN12610000689077; http://www.anzctr.org.au) adhered to Helsinki Guidelines and were approved by the Sydney South West Area Health Service and University of Sydney ethics committees. All volunteers provided written informed consent.
Participants were randomly assigned (1:1) to take nicotinamide 500mg (Nature's Own, Virginia, Queensland, Australia) or matched placebo (Australian Custom Pharmaceuticals, Sydney, New South Wales, Australia) twice daily (Study 1) or once daily (Study 2) for 4 months. The treatment allocation sequence was determined by a computer-generated randomization list prepared using a permuted blocks method (block size 6) by an investigator (DLD) not involved in AK assessment. Participants underwent complete skin examination before randomization, were encouraged to use daily sunscreen, and remained blinded throughout the study.
At baseline, 2 and 4 months, palpable AKs were identified visually and by touch by a blinded observer (DS), counted and documented on a body grid chart. At baseline and 2 months, full blood count, creatinine, and liver function were assessed.
A target of 36 patients was selected for Study 1 based on clinical judgement as this was our first pilot trial of oral nicotinamide. A conservative interpretation of the Study 1 results was used to inform the assumptions used in the sample size calculation for Study 2, where a sample size was selected to provide ≥80% power to detect a standardized effect size of 0.4 at the two-sided 5% level of significance, based on an analysis of covariance model given a correlation of 0.90 between baseline and follow-up assessments (
All randomized patients were eligible for inclusion in the efficacy analysis (intention to treat). The primary endpoint obtained for each patient was the AK count at 4 months. We also noted all histologically confirmed skin cancers during the study. The right-skewed distribution of the AK data was rectified by applying a loge transformation (with results back-transformed for reporting). The relative difference between groups on AK count was estimated using an analysis of covariance adjusting for baseline (
). Logistic and Poisson regression was used to compare treatment groups from both studies combined on skin cancer incidence with the number of previous cancers and study designation fitted as covariates.
A total of 35 patients were enrolled in Study 1 (June–October 2009) (Table 1). One withdrew (nicotinamide) at 2 months because of invasive squamous cell carcinoma (SCC), but returned for his 4-month AK count. A total of 41 patients were enrolled in Study 2 (August–November 2010); two withdrew from treatment (placebo) soon after their baseline counts because of nursing-home placement, but agreed to follow up AK counts. Two nicotinamide participants withdrew from follow-up for personal reasons soon after enrolment; their baseline AK counts were carried forward and included in the primary analysis (Figures 1 and 2).
Table 1Baseline characteristics and AKs during treatment with nicotinamide or placebo
Back-transformed estimates from an analysis of covariance performed on the loge(AK count) data at the time point indicated with treatment group included as a factor and the baseline loge(AK count) as a covariate. The absolute difference between the groups on loge(AK count) corresponds to the relative difference between the groups on AK count.
Back-transformed estimates from an analysis of covariance performed on the loge(AK count) data at the time point indicated with treatment group included as a factor and the baseline loge(AK count) as a covariate. The absolute difference between the groups on loge(AK count) corresponds to the relative difference between the groups on AK count.
Back-transformed estimates from an analysis of covariance performed on the loge(AK count) data at the time point indicated with treatment group included as a factor and the baseline loge(AK count) as a covariate. The absolute difference between the groups on loge(AK count) corresponds to the relative difference between the groups on AK count.
Back-transformed estimates from an analysis of covariance performed on the loge(AK count) data at the time point indicated with treatment group included as a factor and the baseline loge(AK count) as a covariate. The absolute difference between the groups on loge(AK count) corresponds to the relative difference between the groups on AK count.
Very similar P-values were obtained in a sensitivity analysis when the groups were compared on percentage change from baseline at month 4 in AK count (statistical problems associated with using percentage change as an endpoint and rationale for statistical adjustment of baseline scores via analysis of covariance as the optimal analysis approach reviewed in Bonate (2000)).
0.0006
0.005
Abbreviations: AKs, actinic keratoses; CI, confidence interval; LS, least squares.
Bold values indicate means and significant P-values.
1 Back-transformed estimates from an analysis of covariance performed on the loge(AK count) data at the time point indicated with treatment group included as a factor and the baseline loge(AK count) as a covariate. The absolute difference between the groups on loge(AK count) corresponds to the relative difference between the groups on AK count.
2 Very similar P-values were obtained in a sensitivity analysis when the groups were compared on percentage change from baseline at month 4 in AK count (statistical problems associated with using percentage change as an endpoint and rationale for statistical adjustment of baseline scores via analysis of covariance as the optimal analysis approach reviewed in
AK counts at baseline and follow-up are shown in the Table 1. A 35% relative reduction in AK count at 4 months (95% confidence interval (CI): 18–48%; P=0.0006) was estimated from Study 1 (with similar results at 2 months). A 29% relative reduction in AK count at 4 months (95% CI: 11–44%; P=0.005) was estimated from Study 2 (with smaller but significant differences observed at 2 months). There was no evidence that the relative effect of nicotinamide was modified by baseline AK count (treatment-by-baseline interaction P-value was nonsignificant).
For Studies 1 and 2 combined, 37 patients were randomized to placebo and 37 to nicotinamide. Eighty-one and 79% of placebo and nicotinamide patients, respectively, had previous, histologically confirmed skin cancers. During the 4-month trials, 11 placebo patients developed 20 new skin cancers (12 basal cell carcinoma (BCC) and 8 SCC) and 2 nicotinamide patients developed 4 cancers (2 BCC and 2 SCC). The odds of developing at least one skin cancer was significantly lower with nicotinamide (odds ratio=0.14; 95% CI: 0.03–0.73, P=0.019) as was the rate of new skin cancers (relative rate=0.24; 95% CI: 0.08–0.71, P=0.010) as estimated, respectively, by Logistic and Poisson regression models and adjusting for study and number of previous skin cancers. The treatment effect remained significant when analysis of skin cancer rates was repeated using a negative binomial model (P=0.038), although we note the unplanned nature of this combined analysis. Compliance, measured by counts of returned tablets, was 94–98%. One patient, who was also taking aspirin, described nausea while taking nicotinamide. No other potential side effects were reported and no clinically significant changes in blood profiles were observed.
The mechanisms by which nicotinamide might prevent skin cancer or reduce progression of subclinical lesions are unclear. Nicotinamide is a substrate and inhibitor of the nuclear enzyme poly-ADP-ribose polymerase, which is centrally involved in DNA repair (
), consistent with the 13–15% reduction from baseline observed in our placebo groups. Our randomized, double-blinded design enabled detection of AK reductions with nicotinamide relative to any background variations in AKs due to seasonal and behavioral fluctuations in UV doses.
Nicotinamide is well tolerated and costs $5–$10 per month at the doses used here. The results of these phase II studies suggest nicotinamide is effective in reducing AKs and shows promise for skin cancer chemoprevention. A longer phase III trial in a larger cohort, with new skin cancers as the primary endpoint, is now warranted.
Acknowledgments
We are most thankful to our study participants, and gratefully acknowledge the funding of the Cancer Council New South Wales, Epiderm, and the Dermatology Research Foundation.
35732-8&id=gr1.jpg){kind=link}
35732-8&id=gr2.jpg){kind=link}