Patients with chronic rhinosinusitis with nasal polyps who received higher annual oral corticosteroid exposure had greater odds of systemic adverse events in a nationwide nested case-control study.
Researchers analyzed data from the Korean Health Insurance Review and Assessment Service database from 2010 to 2023. The study included 165,361 adult patients with incident chronic rhinosinusitis with nasal polyps diagnosed from 2010 to 2016 who received at least 1 oral corticosteroid prescription. Patients with prior immune-mediated inflammatory diseases or prior diagnoses of the study outcomes were excluded.
Among the cohort, 55,468 patients developed at least 1 adverse event during follow-up. After matching, the analysis included 51,647 case patients and 472,369 controls, with a mean midpoint-approximated age of 47 years and mean follow-up of roughly 5.5 to 5.7 years. The composite outcome included multiple predefined events, including cardiometabolic diagnoses such as hypertension and dyslipidemia, which should be considered when interpreting the overall adverse-event rate.
Oral corticosteroid exposure was assessed according to annual treatment duration, annual cumulative prednisolone-equivalent dose, and prescription frequency. To reduce confounding from steroid use for other indications, researchers counted oral corticosteroid prescriptions issued during encounters coded for chronic rhinosinusitis with nasal polyps. A daily dose of at least 5 mg prednisolone-equivalents was considered clinically relevant systemic exposure.
For the composite outcome of any systemic adverse event, cumulative annual dose was the exposure metric that met the researchers’ threshold for clinical importance. Patients receiving at least 1.0 g of prednisolone-equivalent oral corticosteroids per year had 23% higher adjusted odds of any systemic adverse event compared with those receiving less than 0.5 g per year. The researchers prespecified a relative risk increase of at least 10% as the minimum clinically meaningful effect; this high-dose composite finding met that threshold, whereas the clinical importance of the corresponding composite association for the highest annual-duration group could not be determined because the lower confidence bound fell below the prespecified threshold.
However, the highest-exposure groups were small. Among patients in the matched analysis, 441 case patients and 3,279 controls were in the at-least-1.0-g/year cumulative-dose group. For annual treatment duration, 256 case patients and 1,969 controls received oral corticosteroids for more than 90 days per year.
In adverse-event–specific analyses, patients receiving at least 1.0 g/year had 2.6 times the odds of avascular bone necrosis and 1.5 times the odds of pneumonia compared with those receiving less than 0.5 g/year. The avascular bone necrosis estimates should be interpreted cautiously because they were based on small exposed case counts, including 20 case patients in the highest cumulative-dose group.
For annual treatment duration, the adverse-event–specific findings differed from those for cumulative dose. Patients receiving oral corticosteroids for more than 90 days per year had higher adjusted odds of avascular bone necrosis, osteoporosis, and pneumonia compared with those receiving 30 days or less per year. However, the investigators noted that the clinical importance of the overall composite association for the more-than-90-day group could not be determined because the confidence interval extended below the 10% relative-risk threshold they had set for a minimally meaningful clinical effect.
High cumulative dose also was associated with dyslipidemia, heart failure, hypertension, and type 2 diabetes, although the researchers noted that the clinical importance of these cardiometabolic associations was uncertain. Osteoporosis was not significantly associated with the highest cumulative-dose category but was associated with oral corticosteroid exposure lasting more than 90 days per year.
Prescription frequency was not associated with overall adverse-event risk. Frequent use was associated with avascular bone necrosis in the adverse-event–specific analysis, but the researchers cautioned that the clinical importance of this finding remained uncertain.
The researchers used conditional logistic regression and adjusted for Charlson Comorbidity Index score, which remained imbalanced after matching. Because controls were selected through risk-set sampling, the investigators noted that the odds ratios approximated hazard ratios from the full cohort.
The study was limited by its observational design and reliance on administrative claims data from a single national health care system in South Korea, which may limit generalizability to other populations and prescribing settings. Researchers could not assess chronic rhinosinusitis with nasal polyps severity, asthma severity, the exact clinical indication for each prescription, medication adherence, over-the-counter steroid use, or exposure to inhaled, topical, or intravenous corticosteroids. Residual confounding remained possible despite matching and statistical adjustment.
The authors described the exposure cutoffs as pragmatic, prescription-pattern–based categories rather than empirically optimized safety thresholds. They reported that risk escalation was observed beyond more than 90 days of annual exposure or at least 1.0 g prednisolone-equivalents per year and concluded that annual exposure should be limited to less than 30 days or below 1.0 g whenever possible.
“These findings support limiting annual OCS exposure to less than 30 days or below 1.0 g prednisolone-equivalents whenever possible and reinforce the need for evidence-based, steroid-sparing treatment approaches in the long-term management of CRSwNP,” the study authors wrote.
Disclosures: The researchers reported no conflicts of interest.