Association between cycline antibiotic and development of pseudotumor cerebri syndrome

Background: Cycline antibiotics (CAs) are commonly used to treat acne, blepharitis, and dry eye syndrome. Prescribers or patients may hesitate to use Cas because they may increase the risk of pseudotumor cerebri syndrome (PTCS). Objective: We sought to assess whether CA use is associated with an increased risk of PTCS or papilledema and whether the risk depends upon dosage or duration of CA intake. Methods: We studied patients 12 to 65 years of age who were diagnosed with acne, blepharitis, or dry eye syndrome, who were enrolled in a nationwide managed care network between January 1, 2001 and December 31, 2015, and who had no preexisting diagnosis of papilledema or PTCS. Multivariable Cox regression modeling was used to assess the risk of developing papilledema or PTCS from exposure to CAs. Results: Among the 728,811 eligible enrollees (mean age, 34.7 years; 72% female), 42.0% filled ≥1 CA prescription. Of the 305,823 CA users, 170 (0.06%) were diagnosed with papilledema or PTCS. By comparison, of the 57.0% with no record of CA use, 121 (0.03%) were diagnosed with papilledema or PTCS (P < .0001). In the unadjusted model, every additional year of CA use was associated with a 70% (doxycycline: hazard ratio, 1.70 [95% confidence interval 0.98-2.97]; P = .06) or 91% (minocycline: hazard ratio, 1.91 [95% confidence interval 1.11-3.29]; P = .02) increased hazard of papilledema/PTCS relative to nonusers of CAs. After adjustment for confounders, the increased hazard of PTCS/papilledema with CA use was no longer statistically significant (P = .06, doxycycline; P = .08, minocycline). Limitations: This study relies on claims data, which lack clinical data. Conclusion: This study offers some evidence that CAs may increase the risk of PTCS/papilledema. However, after accounting for confounding factors in our multivariable models, we found no statistically significant association between CA use and the development of PTCS. Moreover, there was no dose-response effect whereby greater CA use was associated with a higher PTCS risk.