The data were extracted from the medical hospital records of patients with evidence of MBC during the period of nine years (January 2011 to December 2019) treated at Medical Oncology Department, South Egypt Cancer Institute, Assiut University. The inclusion criteria included the evidence of MBC, received PLD, age of 18 years old or more, no evidence of heart failure before treatment, frequent assessments by echocardiography were present. These data included the age of the patient, Eastern Cooperative Oncology Group (ECOG) performance status, height, weight, medical comorbidities (i.e. diabetes mellitus and hypertension), type of previously received conventional anthracycline and its cumulative dose, prior exposure to trastuzumab, prior exposure to locoregional radiotherapy to left breast cancer, the cumulative dose and number of cycles of PLD, left ventricular ejection fraction (LVEF) as measured by echocardiography (baseline and follow up measurements), and manifestations of congestive heart failure (CHF). Body mass index (BMI) was calculated from patient’s height and weight to outline whether the patient was obese (BMI ≥ 30 kg/m²) or not.

Patients were grouped according to age group (less than 65 years and 65 years or more according to definition of World Health Organization of the elderly), ECOG performance status (grade 0/I and grade II), obesity (obese or not), medical comorbidities (No, diabetes mellitus, hypertension or both diabetes mellitus and hypertension),prior conventional anthracyclines (doxorubicin or epirubicin), cumulative dose of prior conventional anthracyclines in mg/m², prior exposure to trastuzumab (yes or no), prior exposure to locoregional radiotherapy to left breast cancer (yes or no), the cumulative dose of PLD in mg/m², degree of the cumulative dose (high ≥ 400 mg/m² versus not high, based on definition reported by Kesterson et al. [12]), grades of the decline in LVEF according to the Common Terminology Criteria for Adverse Events (CTCAE) version 4.03 [13], grades of congestive heart failure according to the CTCAE 4.03 [13].

Grade 1: not applicable;

Grade 2: Resting ejection fraction (EF) 40% - 50%; or 10% - 19% drop from baseline;

Grade 3: Resting ejection fraction (EF) 20% - 39%; >20% drop from baseline;

Grade 4: Resting ejection fraction (EF) < 20%;

Grade 5: not applicable.

Grade 1: Asymptomatic with cardiac imaging abnormalities;

Grade 2: Symptoms with mild to moderate activity or exertion;

Grade 3: Severe with symptoms at rest or with minimal activity or exertion; intervention indicated;

Grade 4: Life-threatening consequences; urgent intervention indicated (e.g., continuous IV therapy or mechanical hemodynamic support);

Grade 5: Death.

PLD was given as 50 mg/m² as a 30- or 60-minute intravenous infusion on day 1, repeated every 28 days. The medical history and clinical examination were taken every visit specially cardiac symptoms and signs suggesting cardiac failure and other toxicity. Baseline echocardiography was done, then every 3 - 4 courses of PLD. Treatment was continued until maximum response, unacceptable toxicity, or patient refusal to continue.

The primary endpoint was PLD-related cardiotoxicity though outlining the grades of the decline of LVEF and grades of heart failure. This assessment is based on the Common Terminology Criteria for Adverse Events (CTCAE 4.03) [13]. The secondary endpoint were PLD-related other toxicities. Shapiro-Wilk test was used to define the normality of distribution of our sample. Univariate analysis was used through the presentation of continuous variables as mean ± standard deviation. Categorical variables (nominal and ordinal) are presented as frequency and percentage. Bivariate analysis was done to compare categorical variables using Chi-Square test or Fisher Exact test when appropriate. A p-value less than 0.05 is considered as a cut off of significance. SPSS version 21.0 (SPSS Inc. Chicago, IL, USA) was used in the storage and analysis of data [14].

CONSORT diagram (Figure 1) shows flow of the examined and the analyzed cases through this retrospective trial. During a period of nine years (January 2011 to December 2019), eighty-one patients met our eligibility criteria.

On initial statistical assessment, we found that our sample was normally distributed (Shapiro-Wilk test is not significant, p = 0.076). Table 1 illustrates the patients’ demographics and disease characteristics. The mean age at the time PLD initiation was 43.9 years with the standard deviation of 13.2 years. About one-tenth (9.9%) of our sample were elderly (65 or more years old). Twenty-one patients were hypertensive, while seven patients were diabetic. Only three cases had both hypertension and diabetes mellitus (DM). The majority (93.8%) had ECOG performance status grade II. About one-fifth (21%) of the subjects had a history of prior exposure to trastuzumab. Regarding the prior anthracyclines, less than half (39.5%) of the cases received doxorubicin (mean cumulative dose ± SD: 253.1 ± 25.2 mg/m²), whereas the remaining cases (60.5%) received epirubicin (mean cumulative dose ± SD: 588.0 ± 97.3 mg/m²). The mean number of PLD cycles was 7.57 (range 2 - 24). The mean cumulative dose of PLD was 378.4 mg/m² (±SD of 250.2) and a range of 100 - 1200 mg/m². With taking of 400 mg/m² as the cutoff of the HCD of PLD, thirty-one (38.3%) patients received HCD of PLD (400 mg/m² or more), while the remaining 50 patients did not.

Abbreviations: DM, diabetes mellitus; ECOG, Eastern Cooperative Oncology Group; SD, Standard deviation.

Table 2 shows that the decline in LVEF was relatively rare; and of low grade. Grade 2 decline in LVEF occurred in two patients who received HCD of PLD, and only one patient who did not reach HCD (p = 0.555). Grade 3 or 4 decline in LVEF did not occur in patients either with or without HCD. Also, there were no reported grade 3/4 congestive heart failure in both groups.

Regarding other non-hematological toxicities, there was a significant increase in incidence of all grades palmar plantar erythrodysesthesia (PPE) in patients received HCD of PLD when compared to those did not reach the HCD (38.7% versus 16% respectively; p = 0.021). Also, the grades 3/4 PPE significantly increased with HCD (16.1% for patients with HCD versus 2.0% for those without HCD; P = 0.028) (Table 2). In addition to previous findings, patients who received HCD of PLD experienced higher incidences of grades 3/4 stomatitis (p = 0.039), all grades constipation (p = 0.049), all grades anorexia (p = 0.004), grades 3/4 anorexia (0.028), and all grades fatigue (p = 0.007) when compared to those

Abbreviation: PPE, palmar plantar erythrodysesthesia; LVEF, left ventricular ejection fraction.

who not reaching the HCD. From data about hematological toxicities, HCD was associated with a significant increase in the incidence of allgrades leukopenia (p = 0.026) and grades 3/4 neutropenia (p = 0.019) (Table 2). Regarding the other toxicities, there was no statistical difference between patients who received HCD of PLD and those who did not (Table 2). There was no treatment-related death.