AUTHORS
Songphol Tungjitviboonkun, MD, MS, Matee Rehman, MD, MS2, Lea Daran, MS3, Elly Gardev, MS4
AFFILIATIONS
1 UCSF Helen Diller Family Comprehensive Cancer Center, San Francisco, California, United States, University of California, San Francisco, San Francisco, California, United States
2 University of California, San Francisco, San Francisco, California, United States
3 University of California, San Francisco, San Francisco, California, United States
4 University of California, San Francisco, San Francisco, California, United States
ABSTRACT
Background
Adrenal insufficiency (AI) is an underrecognized complication of repeated high-dose corticosteroid exposure during lymphoma treatment. However, the incidence, timing, corticosteroid exposure thresholds, and prognostic variables associated with AI across lymphoma subtypes remain poorly defined. We evaluated the incidence, clinical characteristics, corticosteroid exposure, and survival outcomes of AI in a large real-world cohort of patients with lymphoma.
Methods
We conducted a retrospective cohort study using the University of California, San Francisco electronic health record database (2010–2025). Adult patients with lymphoma receiving at least one chemotherapy were included. AI was defined by the ICD-10 diagnosis code for adrenal insufficiency and a confirmed morning serum cortisol level <5 µg/dL. Cumulative corticosteroid exposure was standardized as prednisone-equivalent dose and calculated within 3, 6, and 12 months after lymphoma diagnosis. Baseline characteristics were compared using Fisher's exact test or the Mann-Whitney U test. Overall survival was estimated using Kaplan-Meier analysis. To eliminate immortal time bias, a time-dependent Cox proportional hazards regression was used to evaluate the association between AI and mortality after adjustment for baseline demographics (age, sex) and baseline comorbidities (hypertension, diabetes mellitus, HIV, and chronic obstructive pulmonary disease), with stratification by lymphoma subtype.
Results
Among 3,342 eligible patients with lymphoma, 206 (6.2%) developed AI. AI prevalence varied by race, with the highest rates observed among Black/African American (9.3%) and Asian (8.6%) patients, compared with White patients (5.2%). Across lymphoma subtypes, AI was most frequent in mature T/NK-cell lymphomas (9.7%) and diffuse large B-cell lymphoma (6.3%), whereas lower rates were observed in Hodgkin and indolent lymphomas (3.7%).
Patients without AI received significantly lower cumulative glucocorticoid exposure than those with AI, likely reflecting greater use of steroid-sparing chemotherapy regimens, whereas patients who developed AI experienced higher median prednisone-equivalent doses within 3 months (469.0 vs. 298.0 mg, P=0.007), 6 months (1,012.0 vs. 597.5 mg, P<0.001), and 12 months (1,397.2 vs. 813.5 mg, P<0.001). Stem cell transplantation was also more common among patients who developed AI (29.1% vs. 21.6%, P=0.01).
The median time from lymphoma diagnosis to AI was 7.8 months (IQR, 2.1–23.2) and did not differ significantly by age (P=0.424), sex (P=0.595), or lymphoma subtype (P=0.332).
Patients with AI had significantly inferior overall survival compared with those without AI (8-year survival, 49% vs. 79%; log-rank P<0.001). In time-dependent stratified Cox regression, AI remained independently associated with increased mortality (hazard ratio [HR], 2.68; 95% CI, 2.10–3.42; P<0.001). Additional independent predictors of mortality included older age (HR, 1.02 per year; 95% CI, 1.01–1.03), male sex (HR, 1.33; 95% CI, 1.13–1.56), diabetes mellitus (HR, 1.35; 95% CI, 1.07–1.69), and hypertension (HR, 1.25; 95% CI, 1.05–1.50).
Conclusions
Secondary AI is a common (6.2%) yet delayed complication of corticosteroid-containing lymphoma therapy, developing a median of 7.8 months post-diagnosis. AI is strongly associated with high cumulative steroid exposure and a >2.5-fold increase in mortality risk. These findings suggest that steroid-sparing chemotherapy and/or dose reduction strategies may reduce cumulative corticosteroid exposure and mitigate the risk of AI.
DISCLOSURES
Songphol Tungjitviboonkun, MD, MS: No relevant conflicts of interest to declare.