Aromatase inhibitor musculoskeletal toxicity (AIMT) in patients (pts) with early breast cancer (EBC): Prevalence, management, and association with non-adherence to AI.

12122 Background: AIMT may lead to treatment discontinuation and detriment on clinical outcomes in pts with EBC. Recommended supportive care (SC) management include adequate physical activity (PA) levels, acupuncture, and physical therapy. Reasonable pharmacologic approaches include duloxetine use and switching AI. We assessed prevalence of AIMT, SC use, and non-adherence to AI. Methods: Postmenopausal pts with EBC treated with adjuvant AI were included from the longitudinal CANTO cohort (NCT01993498). AIMT was defined as any grade (G) articular or muscular pain (CTCAE v4.0) after 3-6 months (Y0), 1 (Y1), 3 (Y3) and 5 years (Y5) of AI. Non-pharmacologic SC included adherence to PA recommendations (≥10 MET-h/week), consulting with an acupuncturist, physical therapist, or osteopath. Pharmacological SC included use of duloxetine, oral complementary-alternative medicine (OCAM), and switching to a different AI. Non-adherence was defined as any interruption and/or permanent discontinuation of AI. Multivariable logistic regressions tested associations of reporting AIMT with subsequent SC use and with non-adherence to AI. Results: Among 4854 pts, 85.9% reported AIMT overall (G3 17.5%): 61.0% at Y0 (G3 9.9%), 68.5% at Y1 (G3 8.7%), 65.4% at Y3 (G3 9.0%), and 57.0% at Y5 (G3 8.8%). Pts with AIMT were younger, with higher rate of previous musculoskeletal problems, and more frequently received chemotherapy. First and second prescribed AIs were mostly letrozole (54%) and exemestane (56%), respectively. Pts reporting AIMT at Y0 were less likely, by Y1, to adhere to PA recommendations (59.3% vs 62.6%), and slightly more likely to consult with an acupuncturist (9.0% vs 7.6%), physical therapist (46.0% vs 35.0%), or osteopath (22.4 vs 16.3%) compared to pts without AIMT. Results were consistent at remaining time-points. In multivariable models, reporting AIMT was consistently associated over time only with subsequent physical therapy consultations (adjusted aOR 95%CI: Y1, 1.48 1.24-1.78; Y3, 1.40 1.13-1.74; Y5, 1.45 1.10-1.91), but not with other non-pharmacological SC. Only 1.4% of pts with AIMT reported duloxetine use, while 25.3% OCAM use. Switching AI was reported by 19.7% of pts with AIMT (aOR of AI switch vs no AIMT 2.47 1.59-3.83). 18.4% of pts reporting AIMT were non-adherent to AI: 6.0% had interruptions and 17.8% permanent discontinuations. Reporting overall AIMT was associated with non-adherence to AI (aOR vs no AIMT 2.38 1.49-3.79). Conclusions: Despite AIMT was highly prevalent and associated with non-adherence to AI in this prospective cohort, uptake of recommended SC strategies to manage AIMT seemed suboptimal and inconsistent over time. This large study highlights gaps in the implementation of guideline-concordant SC and warrants efforts to maximize treatment retention among pts with EBC treated with adjuvant AI.