Study design and objectives
PROTECT trial is a phase II “proof of concept”, multicentre, randomized 1:1, open label, parallel-groups study, designed to evaluate if dapagliflozin reduces chemotherapy-induced cardiotoxicity (CTRCD) in participants with breast cancer treated with (neo-) adjuvant Anthracycline-based chemotherapy +/- trastuzumab. The study aims to describe an efficacy for dapagliflozin compared to standard of care.
The study design is shown in Fig. 2.
The study involves two cancer centers in Italy (Table 1) and it started on November 2023.
This study was approved by the ethic board of the Italian Medicine Agency (AIFA), and by that of each recruiting hospitals, such as EudraCT Number 2022-003377-28. The study has been registered with ClinicalTrials.gov (registration date and number: 2024-03-19 and NCT06341842).
Table 1
List of participating Centers
Center | Role | PI |
Fondazione IRCCS Policlinico San Matteo | Coordinator | Laura Scelsi |
Istituto Nazionale Tumori - IRCCS - Fondazione "G.Pascale" | Participant | Nicola Maurea |
The new 2022 ESC Cardio-Oncology Guidelines (10) classified Cancer therapy-related cardiac dysfunction (CTRCD) in:
Asymptomatic CTRCD
-
Severe: New LVEF reduction to < 40%
-
Moderate : New LVEF reduction by ≥ 10 percentage points to an LVEF of 40–49% OR New LVEF reduction by < 10 percentage points to an LVEF of 40– 49% AND either new relative decline in GLS by > 15% from baseline OR new rise in cardiac biomarkers (cTnI/cTnT > 99th percentile, BNP ≥ 35 pg/mL, NT-proBNP ≥ 125 pg/mL or new significant rise from baseline beyond the biological and analytical variation of the assay used).
-
Mild: LVEF ≥ 50% AND new relative decline in GLS by > 15% from baseline AND/OR new rise in cardiac biomarkers (cTnI/cTnT > 99th percentile, BNP ≥ 35 pg/mL, NT-proBNP ≥ 125 pg/mL or new significant rise from baseline beyond the biological and analytical variation of the assay used).
Symptomatic CTRCD: specifically HF, which is a clinical syndrome consisting of cardinal symptoms (e.g. breathlessness, ankle swelling, and fatigue) that may be accompanied by signs (e.g. elevated jugular venous pressure, pulmonary crackles, and peripheral oedema) and has traditionally been divided into distinct phenotypes based on the measurement of LVEF: ≤40%= HFrEF; 41–49%=HFmrEF; ≥50%=HFpEF.
The purpose of this study is to evaluate whether dapagliflozin reduces chemotherapy induced cardiotoxicity in participants with breast cancer treated with (neo-)adjuvant anthracycline-based chemotherapy +/- trastuzumab.
The study seeks primarily to assess whether the administration of dapagliflozin is associated with a lower rate of asymptomatic and symptomatic CTRCD during 18 months. The key secondary objective is to assess whether the administration of dapagliflozin is associated with a lower rate of asymptomatic CTRCD during 18 months.
The primary and key secondary outcome will be measured using transthoracic echocardiography and cardiac biomarkers. For this purpose we calculate during transthoracic echocardiography LVEF (%) and GLS (%), according to definition of CRTCD in ESC Cardio-Oncology Guidelines (10).
The study secondary objectives will be: 1) Difference in severe, moderate and mild asymptomatic CTRCD between the two groups during 18 months according to the background therapy with AC with or without TZ and with or without the use of any of angiotensin-converting enzyme inhibitors, angiotensin receptor blockers, or beta-blockers for their potential cardioprotective role on cardiotoxicity development (subgroup analysis) (11–15); 2) Difference in symptomatic CTRCD between the two groups during 18 months according to the background therapy with AC with or without TZ (subgroup analysis) 3) Change from baseline of at least of one grade of diastolic disfunction (according to ESC guidelines (27), appendix 1) during 18 months. 4) Change from baseline in end diastolic and systolic left ventricular volumes and in left atrial volume during 18 months. 5) Change in plasma levels of the bio-humoral markers between baseline and follow-up at three, six, twelve and eighteen months: • NT-pro-BNP • hsTNI • CKD-EPI eGFR • hsCRP
The safety outcome include: decline in eGFR under 25 ml/min/1.7 mq; documented clinical hypoglycaemic events; premature interruption of trial drug due to recurrent genito-urinary tract infections; premature interruption of trial drug due to symptomatic hypotension; cancer relapse; new cancer events.
Assessment measures
Evaluations will take place at baseline (T1), after chemotherapy and at 3-6-12-18 months.
The timing and frequency of the study visits, including assessment time windows, are presented in the table of study procedures and assessments (Table 2).
Table 2
study procedures and assessments
Cardiovascular parameters | Screening | Basal - Randomization (prior CT) | | Treatment period (during and post-CT) | | End of trial |
Visit | V1 | V2 | V3 | V4 | V5 | V6 |
Days | -30 to -1 | -5 to -1 | 90 +/- 6 | 180 +/-6 | 360 +/- 6 | 540 +/- 6 |
Informed Consent | X | | | | | |
Examination of inclusion and exclusion criteria | X | | | | | |
Randomization | | X | | | | |
Medical History Collection Of The Following Data: • Demography • Age • Sex • Height • Weight • Evaluation Of Any Previous Reported Allergy • Cardiovascular Risk Factors* • Oncological History, Any Other Chemo-Radiotherapy Administered In The Past Medical History • Other Comorbidities A warning collection of the whole patient’s past medical history | X | | | | | |
Clinical Procedures • Physical Examination • Height • Weight • Pressure Arteriosue • Heart Rate Twelve-Lead Electrocardiogram | X | | X | X | X | X |
Blood sample for laboratory tests examined in the local laboratory **: | X | | X | X | X | X |
Transthoracic ecocardiography | X | | X | X | X | X |
Concomitant therapy and intervention | X | | X | X | X | X |
SAFETY EVALUATION Adverse events related to the study medication | X | X | X | X | X | X |
Investigational Product | | X | X | X | X | |
Outcomes • Primary outcome • Secondary outcome | | | X | X | X | X |
Outcome of interest | | | | | Xa | |
*Arterial Hypertension, Smoking Habits, Family History For Cardiovascular Disease, Dyslipidemia And Diabetes
** Full blood count; Creatinine; Urea; CKD-EPI; eGFR; Sodium;Potassium; GOT; GPT; Total bilirubin; Direct and indirect bilirubin; Alkaline phosphatase; Creatine Kinase (CK); NT-pro-BNP; TNI-hs I;CRP-hs;Cytokines
Xa: post breast surgery
Participant timeline
Recruitment started in October 2023 at Fondazione IRCCS Policlinico San Matteo, Pavia. Enrollment is about 12 months with follow-up period of 18 months. Total duration from first in last out is 30 months.
The data collected from the participants and the follow-up timeline are presented in Table 2.
Sample size
The sample size is computed based on the primary endpoint and makes use of the data reported in the literature [10]. At 18 months, we expect a cumulative incidence of CTRCD of 35% in the control arm, corresponding to an event-free survival rate of 65%. With 316 patients (158 per arm) we will be able to elicit an increase in event-free survival in the dapaglifozin arm up to 80% (hazard ratio, HR 0.52, 78 events) with a power of 80 and type I error (2-tailed) of 5%. This sample size accounts for a 10% dropout rate and the log rank test to compare event-free survivals. For calculation we used the Stata command: power log rank .65 .80, wdprob(.1) power(.80) alpha(0.05)
Statistical analyses
All analyses will be performed using the Stata software (release 18, StataCorp, College Station, TX, USA); they will be detailed in the statistical analysis plan. Continuous data will be described with the mean and standard deviation or the median and quartiles, if skewed. Categorical data will be described as counts and percent for each categories. Descriptions will be performed separately for each treatment group. A 2-sided p-value < 0.05 will be considered statistically significant. Variables may be log-transformed for the purpose of the analysis. Briefly, the logrank test will be used to compare cardiac toxicity free survival. Cox regression will be used to compute HR and 95%CI. Regression models for repeated measures will be used to compare changes in biomarkers over time; details are provided in table X.
Table 3
Primary Endpoint | Analysis |
Cumulative incidence of CTRCD | The cumulative incidence of CTRCD per 100 person year will be reported by arm. Cumulative event-free survival will be computed and plotted using the Kaplan Meier method and compared with the logrank test. Cox regression models will be fitted to derive the HR and its 95% confidence interval. Huber-White robust standard errors will be computed to account for intra-centre lack of independence. In a sensitivity analysis of the primary endpoint, we will evaluate the modifying effect of trastuzumab, by including an interaction term of trastuzumab and treatment in the Cox model |
Key secondary endpoint | |
Cumulative incidence of asymptomatic CTRCD | This endpoint will be analyzed as described for the primary endpoint |
Secondary endpoints | |
1 Difference in severe, moderate and mild asymptomatic CTRCD between the two groups during 18 months according to the background therapy with AC with or without TZ (subgroup analysis) and according to the background therapy with cardioprotective drugs | We will evaluate the modifying effect of background therapy, by including an interaction term of background therapy and treatment in the Cox model described in the sensitivity analysis of the primary endpoint. |
2 Difference in symptomatic CTRCD between the two groups during 18 months according to the background therapy with AC with or without TZ (subgroup analysis) | We will evaluate the modifying effect of background therapy, by including an interaction term of background therapy and treatment in the Cox model described in the sensitivity analysis of the primary endpoint. |
3 Change from baseline of at least of one grade of diastolic dysfunction (according to ESC guidelines (41, appendix 2) during 18 months. | The proportion of patients with a change at least one grade will be compared using a random effect regression model for longitudinal binomial data |
4 Change from baseline in end diastolic and systolic left ventricular volumes and in left atrial volume during 18 months. | The two group will be compared using a random effect regression model for longitudinal data. Log-transformation may be applied if needed |
5 Change in plasma levels of the bio-humoral markers between baseline and follow-up at three, six, twelve and eighteen months: NT-pro-BNP, hsTNI, CKD-EPI eGFR, hsCRP | The two group will be compared using a random effect regression model for longitudinal data. Log-transformation may be applied if needed |