(047) Anaphylaxis Risk and Long-Term Central Venous Catheter Use: a Health Plan Claims Database Analysis Comparing Subcutaneous and Intravenous Pertuzumab / Trastuzumab Administration in the US

Background: Intravenous (IV) pertuzumab (P) and trastuzumab (H) are approved in the US to treat HER2-positive breast cancer. They are also available in a subcutaneous (SC) fixed-dose combination (PH FDC SC), potentially allowing treatment outside hospitals. However, anaphylactic reactions are a concern with both the SC and IV forms, especially when used with chemotherapy. The SC version may also reduce the need for long-term central venous catheters with port (CVCp) that may be associated with increased morbidity. This study examines the risk of anaphylaxis and the use of CVCp with PH in real-world settings.

Objectives: To evaluate per administration route the risk of anaphylaxis associated with the administration of P and to describe the concomitant use of CVCp.

Methods: Cohorts of incident users were selected using retrospective records from a US health plan claims database (IQVIA PharMetrics® Plus). Patients included were privately insured in the US, aged 18 to 65 years, and initiated P between 2012 and 2021. Possible and severe anaphylactic events (PAEs) were identified using the algorithms developed by the US Food and Drug Administration (FDA) Sentinel Initiative and updated with ICD10 codes (PMID: 34181291). PAEs were attributed to a drug if recorded on the day of administration or the day after. The prevalence proportion (PP) of PAE was used to approximate the risk of anaphylaxis at each cycle. A sensitivity analysis accounted for a potential depletion of susceptibles (P IV and H IV may be administered sequentially in any order, patients with a PAE from H would not receive scheduled P IV and thus be excluded from the P IV cohort). CVCp use was defined using procedure codes recorded within the year prior to the first administration of PH and up to 90 days after.

Results: A total of 12,264 patients and 184 initiated P IV and PH FDC SC, respectively. In the P IV cohort, seven PAEs were observed up to cycle 6 and none thereafter. The PP at cycle 1 of P IV was 1.63 per 10,000 patients (95% CI = [0.45; 5.94]) and 2.45/10,000 (95% CI = [0.83; 7.19]) in the primary and sensitivity analysis, respectively. No PAE was reported in patients treated with PH FDC SC. The use of CVCp was observed in 85.2% and 57.6% of patients with P IV and PH FDC SC, respectively.

Conclusions: The risk of PAE with P appeared to be very low but needs to be further contextualized. No PAE was observed with PH FDC. All PAE were observed when patients are expected to receive treatment at the hospital due to chemotherapy administration. PH FDC SC use was associated with a nearly 30% reduction in the use of CVCp. This reduction could be underestimated because of the look-back period and did not account for potential switch from P IV warranting further investigations. PH FDC SC, may enable less invasive treatment administration.