This study indicated that co-administration of pertuzumab increased the incidence and severity of infusion reactions caused by the administration of trastuzumab. In addition, this study showed that high lymphocyte counts just before therapy were a risk factor for infusion reactions in patients receiving Tra+Per combination therapy. These findings provide valuable information for the prevention of infusion reactions during Tra+Per combination therapy.
The incidence of infusion reactions in patients receiving Tra+Per was 47.7%, which was significantly higher than that in patients receiving Tra alone (18.9%), with a risk ratio of 2.53 (Fig. 1). The incidence of infusion reactions for trastuzumab monotherapy (18.9%) was similar to the results in a previous report (16.2%) [6], indicating that the incidence of infusion reactions during trastuzumab monotherapy was reasonable in this study. In patients receiving Tra+Per, infusion reactions were not observed during pertuzumab treatment, but they appeared during or after trastuzumab administration (Table 2). The incidence (13.0%) of infusion reactions for Tra+Per combination therapy in the CLEOPATRA trial for metastatic recurrent breast cancer was lower than that in the neoadjuvant and adjuvant trials (18.6%-25.0%) for early-stage breast cancer [9] and in our study (47.7%). One reason for the low incidence in the CLEOPATRA trial is the long interval between pertuzumab (day 1) and Tra+DTX (day 2). Thus, pertuzumab did not increase the risk of infusion reactions associated with trastuzumab in the CLEOPATRA trial, since it was not administered on the same day. Therefore, our results suggest that co-administration of pertuzumab increases the incidence and severity of infusion reactions caused by trastuzumab.
Our results indicate that the pretreatment lymphocyte counts in patients who showed infusion reactions during the Tra+Per combination therapy were significantly higher than those in patients who did not show infusion reactions (Table 4). Multivariate analysis identified lymphocyte counts just before treatment as a risk factor for the development of infusion reactions in Tra+Per combination therapy (Table 5). Patients taking rituximab, another biologic drug used for chronic lymphocytic leukemia, and showing lymphocyte counts greater than 5.0 × 104/µL have been reported to show a significantly higher rate of severe symptoms than patients taking rituximab with low lymphocyte counts [10]. Thus, the infusion reactions caused by rituximab are more likely to become severe in patients with high lymphocyte counts. This is thought to be because the number of lymphocytes increases as the disease progresses, and the incidence and severity of infusion reactions increase accordingly because CD20-positive B cells can release cytokines during death. In contrast, pertuzumab and trastuzumab activate antibody-dependent cytotoxicity (ADCC) through effector cells such as macrophages and natural killer cells (NK cells) [11]. Activated NK cells, one of the lymphocytes, can produce cytokines such as tumor necrosis factor alpha, so higher numbers of NK cells release more cytokines, which may increase the incidence and severity of infusion reactions. Thus, a novel finding of the present study is that the number of circulating lymphocytes is important for the incidence of infusion reactions not only in hematopoietic tumors but also in solid tumors.
All patients in the Tra+Per combination therapy group who showed infusion reactions during the second and subsequent doses also showed infusion reactions after the first dose. In the Tra alone group, 93.8% of patients showing infusion reactions have been reported to develop infusion reactions at the first dose [6]. In addition, high-dose patients (maintenance therapy at 4 mg/kg after the initial dose of 8 mg/kg) had higher rates of fever (45% vs. 36%) and chills (40% vs. 22%) than low-dose patients (maintenance therapy at 2 mg/kg after an initial dose of 4 mg/kg) [12]. On the basis of these results, we speculate that the incidence of infusion reactions is higher after the first dose because of the use of loading doses of both trastuzumab and pertuzumab. Therefore, special attention should be paid to infusion reactions at the first dose in the Tra+Per combination therapy.
Our study did not extract BMI as a risk factor of infusion reactions (Table 4), but high BMI has been reported to increase the risk of infusion reactions with trastuzumab monotherapy [6]. Only 23.1% of the patients in our study had a high BMI (25 or more), which was markedly less than the proportion in a previous study (69.5%), and one reason why the influence of BMI was not observed may be the difference in patient background. In addition, this study did not extract stage classification as a risk factor for infusion reaction (Table 4), unlike previous reports on trastuzumab monotherapy [6]. The incidence of infusion reactions in patients with stage IV disease was 50% (7/14). The incidence of infusion reactions in stage IV patients pretreated with trastuzumab (1/6, 16.8%) tended to be lower than that in patients not pretreated with trastuzumab (6/8, 75%; p = 0.103) in this study. Therefore, one reason why the influence of stage classification is different could be pretreatment with trastuzumab.
Prediction of infusion reactions on the basis of the lymphocyte count, which was identified as a risk factor for infusion reactions in this study, will enable consideration of prophylactic antipyretic analgesic medication and enhanced care tailored to individual patient risk. In the future, we hope to conduct a study with a larger patient population and investigate the benefit of premedication in Tra+Per combination therapy and the relationship between infusion reaction occurrence and therapeutic effect so that chemotherapy for breast cancer patients can be provided more safely and effectively.
This study had several limitations. This was a retrospective study based on medical records, and because of the limited number of patients, the background characteristics of patients in the IR and non-IR groups were not standardized. Future studies should aim to evaluate more cases and evaluate patients at multiple institutions. Another limitation of our study is that since most of the patients were outpatients undergoing chemotherapy, we may not have accurately assessed the infusion reactions that occurred after the patients returned home.