Dexamethasone and Diphenhydramine Withdrawal and Quality of Life Assessment in Breast Cancer Patients Receiving Weekly Paclitaxel

The current was a single-center, prospective, phase-2 study conducted at the Oncology Center at Hospital São Lucas, located in the city of Porto Alegre, Brazil. The eligibility criteria for enrollment were female breast cancer patients who would receive a standard weekly dose of paclitaxel (80mg/m²) as part of the neoadjuvant, adjuvant or palliative treatments, aged between 18 and 75 years, with performance status scores from 0 to 2, and ability to provide informed consent prior to enrollment. The exclusion criteria were chronic use of corticosteroids, prior use of taxane, combination of paclitaxel with another chemotherapy agent simultaneously, a taxane regimen other than on a weekly schedule, and other neoplasms. The combination of paclitaxel and trastuzumab was allowed.

The clinical trial was approved by the institutional Ethics Committee according to local regulation, and all patients provided written informed consent before enrollment. The trial was registered at the publicly-available Brazilian study registry Plataforma Brasil (under registration number 56605322.7.0000.5336).

Patients were included in the study from April 2022 to April 2024. All patients received intravenous dexamethasone 10 mg and diphenhydramine 50 mg 30 minutes prior to paclitaxel infusion as prophylaxis to HSR in the first 3 of the 12 planned weeks of treatment. If no HSR occurred after the third week of paclitaxel, premedication was allowed to be discontinued for the remaining 9 weeks of treatment, as long as agreed upon by the attending physician. Therefore, the patients were divided split into two groups for subsequent comparisons using the per protocol principle: one group that discontinued premedication and one group that did not, regardless of the reason for premedication continuation.

We interviewed patients on weeks 1, 2, 3, 4, 5, 9, and 12 of treatment in face-to-face appointments to assess QoL using self-reported descriptive and evaluative measurements. We used three different questionnaires at every time point, as described below.

We selected the validated Portuguese version of the QLQ-C30 to address health-related QoL in cancer patients in general. This questionnaire is composed of a global health/QoL (GH/QoL) scale, as well as functioning and symptoms scales. We used the combination of items 29 and 30 as the main QLQ-30 measurement of QoL, according to EORTC guidance. The higher the score, the higher the QoL.9

The validated Portuguese version of the QLQ-BR23 was selected to evaluate specific QoL factors in breast cancer patients. The multi-item scales are divided into 2 groups: functional (body image, sexual functioning, sexual enjoyment, and future perspective) and symptom-focused (systemic therapy side effects, breast symptoms, arm symptoms, and upset by hair loss). We used the combination of items 1, 2, 3, 4, 6, 7, and 8 (that is, the systemic therapy side effects [BRST] scale) as the main QLQ-BR23 measurement of QoL. Higher scores represent greater symptom burden.9 10

Items from the QLQ-C30 and QLQ-BR23 were scaled and scored according to the scoring manual methodology: patients provided their answers to both questionnaires on a 4-point scale (from 1 [“not at all”] to 4 [“very much”]), except for the GH/QoL, which is a 7-point scale (from 1 [“very poor”] to 7 [“excellent”]). Linear transformation was used to standardize the raw score, so that the overall scores ranged from 0 to 100.9 10

The 14-item modified Dexamethasone Symptom Questionnaire (DSQ) was developed by the University of Toronto to evaluate the frequency of adverse events related to the use of dexamethasone5 (Supplementary Material Figure S3). We evaluated the frequency of highest-grade adverse events between the 2 groups through the comparison of related symptoms reported as 3 (“moderately”) or 4 (“very much”).

The patients were physically evaluated and treated for HSRs by trained oncology fellows on call at the Oncology Center. The severity of the HSRs was graded according to the National Cancer Institute (NCI) Common Terminology Criteria for Adverse Events (CTCAE), version 5.0.

The primary endpoint of the current study was the comparison of QoL between the group that kept and the group that omitted premedication using scores for the main QLQ-C30 and QLQ-BR23 measurements, as aforementioned. We analyzed questionnaires from interviews conducted in weeks 3, 4, 5, 9, and 12, when the patients had already been split into the 2 groups.

The study secondary endpoints were the evaluation of the frequency of severe HSRs according to the CTCAE in patients who withdrew premedication, the frequency of HSRs of any grade during the 12 weeks of treatment in both groups, and the frequency of highest-grade adverse events in both groups according to the DSQ.

We conducted an exploratory analysis comparing GH/QoL scores in weeks 1 and 12 in each patient to analyze if there was a detriment in QoL over time. We also analyzed weight gain during the 12 weeks of treatment.

For the GH/QoL scale of the QLQ-C30 and the BRST of the QLQ-B23, the statistical analyses were based on a linear mixed effects model comparing the 2 groups of patients from weeks 3 to 12 of treatment. The models were developed with treatment, a time effect, and treatment-time interactions as fixed effects, as well as a patient-specific random effect. Parameter estimates were based on a restricted maximum likelihood method.

The Chi-squared test of independence was used to evaluate the association between the two groups and the incidence of subjects who had at least 1 toxicity score of “3” or “4” in the DSQ, indicating high toxicity. The analysis was conducted based on data collected from weeks 4 to 12 of treatment.

All statistical tests were performed on an exploratory basis, and all p-values lower than 0.05 were deemed statistically significant. No correction for multiplicity was made. Data analysis was performed using the R (R Foundation for Statistical Computing) software, version 4.3.3.

We recruited 60 patients. Data were analyzed according to the per protocol principle: 32 patients were allocated to the group that omitted premedication after the first 3 weeks of treatment, and 28 patients were allocated to the group that kept premedication throughout the 12 weeks of treatment. The baseline characteristics of all patients (and according to group) are summarized in Table 1.

All patients completed weeks 1 to 12 of paclitaxel, with a total of 418 infusions. One patient needed a paclitaxel dose reduction to 60 mg/m² from weeks 9 to 12 due to grade-3 fatigue. All patients received dexamethasone and diphenhydramine as premedication in the first 3 weeks of treatment, and HSRs occurred in 10 of the 60 patients (16.7%) in this period, all grade 1. Therefore, premedication was kept in those patients. After the third week of paclitaxel, the remaining 50 patients met the criteria for premedication omission given the absence of HSR. However, premedication was kept in 18 of these patients (36%). The reason to keep premedication in patients who had not experienced HSRs were chemotherapy-related symptoms such as nausea (24%), rash (12%), and preference of the attending physician (20%). No additional episodes of HSRs occurred on weeks 3 to 12 of paclitaxel in either group.

Compliance to the QoL assessment was high. Missing data from 2 patients at weeks 10 and 12 were excluded from the analysis.

We first assessed global health using the combined score from items 29 and 30 of the QLQ-C30 (first primary endpoint). The scores at weeks 3, 4, 5, 9, and 12 for each group are presented in Fig. 1. Despite a minor separation of the curves, there were no statistically significant differences in the QoL scores during treatment between the two groups using the linear mixed effects model (p = 0.542). The mean scores for each group at weeks 1, 2, 3, 4, 5, 9, and 12 are presented in Supplementary Material Table S1.

Next, we evaluated systemic symptoms using the BRST scores of the QLQ-BR23 (second primary endpoint). There were no statistically significant differences regarding the average QoL scores in the comparison of the 2 groups during weeks 3, 4, 5, 9, and 12 (p = 0.278; linear mixed effects model) (Fig. 2).

As exploratory analyses, we evaluated global changes in QoL throughout the combined scores on items 29 and 30 of the QLQ-C30 at weeks 1 (baseline) and 12 (end of treatment) in the 2 groups (Table 2). Over time, we found a 1.3-point increase in the QoL score in the group without premedication, and a 3.1-point decrease in the score of the group with premedication. Neither of these exceed the 5-point change usually considered clinically meaningful.

We also analyzed, in an exploratory way, the different subdomains and symptoms from the QLQ-C30 and QLQ-BR23 (Supplementary Material Figures S2–S5). Some of the functioning and symptoms scales seemed to show clinically significant differences within each of the groups over time between weeks 3 and 12 of treatment. Specifically for the breast symptoms score (BRBS) of the QLQ-BR23, the change in average score in the group of subjects receiving premedication was of less than 0.001 (95%CI: -5.83–5.83; p = 1.000). Nevertheless, the group of subjects without premedication showed an average BRBS change of -8.33 (95%CI: -13.6–-3.05; p = 0.003), indicating that premedication omission resulted in clinically and statistically significant improvements in this QoL parameter over time.

Another difference was observed in the sexual enjoyment (BRSEF) score. For the group of patients receiving premedication, there was an 8.93-point decrease in the average score from weeks 3 to 12 (95%CI: -18.3–0.459; p 0.0614). On the other hand, the group of subjects not receiving premedication showed an increase in the BRSEF score of 5.21 points (95%CI: -2.34–12.8; p = 0.169), raising the hypothesis that premedication may contribute to worsening sexual health in breast cancer patients.

We also described QoL measures according to patient subgroups, which were devided according to age, clinical stage, and chemotherapy protocol received (Supplementary Material Figures S6–S8). As the sample size became too small in the subgroups, no statistical tests were performed.

The incidence of high toxicity scores according to the DSQ (secondary endpoint) is presented in Table 3. Although there were no statistically significant differences between the two groups, some symptoms were numerically more common in the group with premedication compared to the one with premediction, such as heartburn (53.6% versus 37.5% respectively) and acne (21.4% versus 9.4% respectively). In contrast, the incidence of nausea (25.0% versus 17.9%) was numerically higher in the group without premedication. The average weight gain over time was similar between the two groups throughout the treatment period (Supplementary Material Figure S1).