Photosensitizing Antihypertensive Drug Use and Risk of Cutaneous Squamous Cell Carcinoma

K.A. Su; L.A. Habel; N.S. Achacoso; G.D. Friedman; M.M. Asgari

Disclosures

The British Journal of Dermatology. 2018;179(5):1088-1094. 

In This Article

Results

We identified a cohort of 28 357 NHW members of the RPGEH with hypertension who were followed for a mean of 5·09 years after cohort entry. Table 1 shows the baseline demographic characteristics of the cohort members. The mean age at cohort entry was 69·1 years (SD 10·6 years) and was similar across the treatment groups. Women constituted 56·3% of the cohort and 59·5% of users of photosensitizing ADs. After baseline, 17 946 patients used photosensitizing ADs; 19 092 patients used nonphotosensitizing ADs; 17 345 patients used ADs with unknown photosensitizing potential; and 1530 patients did not use any ADs. There was considerable overlap between the treatment groups; many patients received prescriptions for multiple different classes of ADs after cohort entry. CCI score was higher among users of unknown photosensitizing ADs after cohort entry and lower among nonusers of ADs after cohort entry. Prior cSCC and AK were more common among nonusers of ADs after cohort entry (prior cSCC 14·2%; prior AK 40·1%) compared with the overall cohort (prior cSCC 7·9%; prior AK 35·2%). Prior photosensitizing AD use was less common among nonusers of ADs after cohort entry (51·7%) compared with the overall cohort (74·0%).

We identified 3010 cSCCs over 144 215 person–years of follow–up. Table 2 summarizes the HRs for cSCC among those exposed to ADs compared with those with no AD use. Compared with nonuse of ADs, the adjusted hazard ratio (aHR) for ever use of photosensitizing ADs after baseline was 1·17 (95% CI 1·07–1·28). The increase in cSCC risk associated with ever use of photosensitizing ADs was driven in large part by combination thiazide diuretics (aHR 1·32, 95% CI 1·19–1·46). Alpha–2 receptor agonists (aHR 1·30, 95% CI 0·62–2·74); potassium–sparing diuretics (aHR 1·24, 95% CI 0·96–1·61); loop diuretics (aHR 1·10, 95% CI 0·98–1·24); and thiazide diuretics (aHR 1·09, 95% CI 0·99–1·19) all had HRs above 1·0. The aHR for ever use of nonphotosensitizing ADs was 0·99 (95% CI 0·91–1·07). Alpha blockers (aHR 0·93, 95% CI 0·83–1·03), beta blockers (aHR 1·04, 95% CI 0·96–1·12), central agonists (aHR 1·07, 95% CI 0·85–1·34) and ARBs (aHR 0·97, 95% CI 0·87–1·08) had HRs very close to 1·0.

The aHR for ever use of ADs with unknown photosensitizing potential was 1·11 (95% CI 1·02–1·20). The increase in cSCC risk associated with ever use of ADs with unknown photosensitizing potential was driven by ACE inhibitors (aHR 1·10, 95% CI 1·01–1·19). CCBs had an HR very close to 1·0 (aHR 1·06, 95% CI 0·98–1·15). Vasodilators (aHR 0·96, 95% CI 0·76–1·21) and other combinations (aHR 0·81, 95% CI 0·64–1·03) were infrequently prescribed and not associated with increased cSCC risk.

Table 3 summarizes dose–response analyses based on the number of prescriptions for ADs. There was a modest increase in the risk of cSCC with an increased number of prescriptions for photosensitizing ADs (one to seven fills aHR 1·12, 95% CI 1·02–1·24; eight to 15 fills aHR 1·19, 95% CI 1·06–1·34; ≥ 16 fills aHR 1·41, 95% CI 1·20–1·67). In contrast, no dose–response relationship was observed with use of nonphotosensitizing ADs (one to seven fills: aHR 0·99, 95% CI 0·89–1·09; eight to 15 fills aHR 1·02, 95% CI 0·92–1·14; ≥ 16 fills aHR 0·95, 95% CI 0·83–1·07) or ADs of unknown photosensitizing potential (one to seven fills aHR 1·12, 95% CI 1·02–1·23; eight to 15 fills aHR 1·12, 95% CI 1·01–1·25; ≥ 16 fills aHR 1·06 95% CI 0·93–1·21).

Consistent with the main results, when outcomes were restricted to invasive disease, users of photosensitizing ADs had a very modest increased risk of invasive cSCC compared with nonusers (aHR 1·14, 95% CI 1·02–1·27); users of ADs in the 'unknown' photosensitizing category also had a very modest increased risk of invasive cSCC (aHR 1·08, 95% CI 0·99–1·19). In contrast, users of nonphotosensitizing ADs had a very modest decreased risk of invasive cSCC (aHR 0·92, 95% CI 0·84–1·02).

The association of photosensitizing ADs with risk of cSCC did not vary by history of actinic lesions (aHR 1·16, 95% CI 1·05–1·28 among those with history; aHR 1·17, 95% CI 0·97–1·41 among those without history). There was an increased risk of cSCC among users of ADs in the 'unknown' photosensitizing category with a past history of actinic lesions (aHR 1·15, 95% CI 1·05–1·26), but not among those without a past history of actinic lesions (aHR 0·99, 95% CI 0·84–1·16).

Table 4 summarizes the results of additional subgroup analyses of new users of ADs after cohort entry and nonusers of photosensitizing ADs before cohort entry. Analyses of new users of ADs after cohort entry were inconclusive because of the small number of new users (n = 405). Compared with nonuse of ADs, new users of photosensitizing ADs had an elevated cSCC risk (aHR 1·46, 95% CI 0·70–3·02), new users of nonphotosensitizing ADs had a reduced cSCC risk (aHR 0·49, 95% CI 0·19–1·30) and risk for new users of ADs with unknown photosensitizing potential fell between photosensitizing and nonphotosensitizing new users (aHR 0·92, 95% CI 0·42–2·00). Owing to the small number of new users, we were unable to demonstrate statistically significant differences in cSCC risk for the different exposure categories.

Analyses restricted to individuals without use of photosensitizing ADs prior to baseline (n = 7363) showed that users of photosensitizing ADs and ADs with unknown photosensitizing potential during follow–up had an increased cSCC risk compared with users of nonphotosensitizing ADs during follow–up, although the results were not statistically significant (use of photosensitizing ADs during follow–up: aHR 1·20, 95% CI 0·97–1·48; use of ADs with unknown photosensitizing potential during follow–up: aHR 1·16, 95% CI 0·97–1·39).

Comments

3090D553-9492-4563-8681-AD288FA52ACE

processing....