Neoadjuvant trastuzumab deruxtecan (T-DXd) with response-directed definitive therapy in early stage HER2-positive breast cancer: a phase II study protocol (SHAMROCK study)

The SHAMROCK study is a phase 2 open label, single-arm, adaptive multicentre trial. Patients with early stage (stage 2–3) HER2-positive breast cancer are treated with neoadjuvant T-DXd every 3 weeks for up to 6 cycles. A repeat biopsy after 2 cycles (at cycle 2 day 14) of T-DXd is performed for the RDI score assessment. As a safety measure patients will undergo clinical examination before each cycle of T-DXd to enable early identification of on-treatment locoregional progression. If progression is seen, then T-DXd will be stopped, and the patient taken off study treatment. In the absence of early progression, those patients with a high chance of pCR based on the RDI score undergo repeat breast imaging after four cycles of T-DXd. Patients who have a high chance of pCR based on the RDI score proceed to surgery after four cycles of T-DXd if they also have imaging complete response (iCR) at that point. Other patients who have a high chance of pCR based on the RDI score but iCR is not attained after four cycles or who have a low/intermediate chance of pCR based on the RDI score undergo repeat breast imaging after six cycles of T-DXd. Patients with iCR after six cycles of T-DXd, regardless of RDI score, proceed to surgery. Patients who have a low/intermediate chance of pCR based on the RDI score and residual disease on imaging after six cycles of T-DXd will undergo either further systemic therapy or proceed to surgery (at the discretion of their treating physician). Regardless of whether the patient will proceed to surgery or undergo further systemic therapy the maximum number of cycles of T-DXd is six cycles. Based on the surgical specimen, patients who achieve a pCR will undergo further trastuzumab post-operatively as per standard of care to complete a total of 52 weeks of systemic therapy from the first cycle of T-DXd. Patients with residual disease at surgery will receive adjuvant chemotherapy as decided by the treating physician. The study design is illustrated in Fig. 1.

The main inclusion criteria for this study are as follows: newly diagnosed histologically confirmed HER2-positive breast cancer, age ≥ 18 years, stage 2–3 breast cancer, no prior therapy received for breast cancer, ECOG performance status 0–1 and adequate organ function. Patients with lung-specific intercurrent clinically significant illnesses, any autoimmune or inflammatory disorders with pulmonary involvement or prior pneumonectomy are excluded. Key inclusion and exclusion criteria are detailed in Table 1.

Trastuzumab deruxtecan: 5.4 mg/kg administered as an intravenous infusion on day 1, every 3 weeks.

Eligible patients will receive a total of either 4 or 6 cycles of T-DXd. Dose modifications and treatment interruptions are allowed and must be done according to the guidelines in the summary of product characteristics [15]. If treatment has been delayed for more than 4 weeks from the planned date of administration, the patient is withdrawn from the study treatment.

Study data will be handled confidentially. The study data will be collected using electronic Case Report Forms (eCRFs). Clinical trial data will be entered by authorised site personnel into the electronic data capture system. The investigator must maintain accurate documentation (source data) that supports the information entered in the eCRF. The investigator is responsible for verifying that data entries are accurate and correct by signing the eCRF. All data will be reviewed for completeness and logical consistency queries will be generated via the electronic data capture system to correct or clarify data or request missing information. The designated site staff will be required to respond to these queries in accordance with data entry and data query timelines for the study. Ongoing source data verification will be performed to confirm that data entered into the eCRF by authorised site personnel are accurate, complete, and verifiable from source documents; that the safety and rights of participants are being protected; and that the study is being conducted in accordance with the currently approved protocol and any other study agreements, ICH GCP, and all applicable regulatory requirements. A Project Management File (PMF) will be maintained for the duration of the study.

All adverse events will be reported routinely on the eCRF. Intensity for each adverse event will be determined by using the current version 5.0 of the National Cancer Institute (NCI) Common Terminology Criteria for Adverse Events (CTCAE) [16], wherever possible. In cases where these criteria do not apply, intensity should be defined according to the general clinical description of grades as per CTCAE. Adverse events of special interest (AESIs) and serious adverse events (SAEs) will be reported to the sponsor within 24 h of the study site becoming aware of the event. Suspected Unexpected Serious Adverse Reactions (SUSARs) will be reported to the Ethics Committee, EudraVigilance and local regulatory authorities where required. The study will be reviewed on a regular basis by the Safety Monitoring Committee. The Safety Monitoring Committee will report its assessment of the continuing benefit-risk of the study and any recommended actions to the Chief Investigator and the Clinical Lead for consideration and appropriate action.

Tumour tissue will be collected from the baseline biopsy and surgical pathological specimen. This tissue will be used for DNA/RNA sequencing and proteomic analysis. A fresh tumour tissue biopsy will also be collected after completing 2 cycles of therapy, which will be used to calculate the patient’s RDI score. Patient blood samples will be collected at screening and prior to each cycle of treatment. These samples will be used for circulating tumour DNA and RNA analysis and protein biomarker analysis. This translational analysis aims to identify clinically relevant biomarkers to predict benefit or resistance to treatment, and to facilitate future studies.

Sample size calculation was performed for the primary outcome. The primary outcome is the percentage of patients who are successfully treated (achieve pCR at surgery) with only T-DXd and trastuzumab. The preliminary data for the statistical design of our study came from our prior JNCI publication [13]. In this publication, the proportion of HER2 + tumours that were RDI score high with neoadjuvant TCH chemo (6 cycles total) was 41%. The pCR rate with a high RDI score was 83% and the pCR rate with a non-high RDI score was 10% [13]. The former percentage was rounded down a little (on the conservative side), and the latter was rounded up to 20% to allow for the effect of the receipt of optional additional preoperative chemotherapy for the purposes of the design of our study herein. There are no pCR data with T-DXd yet. The pCR rate with T-DM1 (four cycles) in the German ADAPT trial [17] was 41%. These are the data points that we used as a foundation for the statistical design of our trial herein.

Given expected pCR rates of 80% and 20% for the RDI high and non-RDI high groups respectively, the expected pCR rate after surgery for the study as a whole is 44%. Because reported pCR rates in unselected patients with standard HER2 directed neoadjuvant therapy are generally up to 50%, 60% is the minimum acceptable limit for likelihood of pCR in patients selected by high RDI score after 2 cycles of neoadjuvant T-DXd. Similarly, the minimum acceptable limit for likelihood of pCR in patients with non-high RDI score is 10%. Given minimum acceptable limits for expected pCR of 60% and 10% for the RDI high and non-RDI high groups respectively, the minimum acceptable limit for expected pCR rate after surgery for the study as a whole is 30%.

For an expected pCR rate after surgery 44%, a sample size of 74 evaluable patients would provide a 95% confidence interval (Wilson method) with a lower bound of 30% for the pCR rate (equivalent to providing statistical power of 80% to test an expected pCR rate of 44% versus an unacceptable pCR of 30%, using a one-sided significance level of 5%). Allowing for a non-evaluable rate of 5–10%, a total sample size of 80 patients should provide 74 evaluable patients, where an evaluable patient is defined as a patient who is evaluated for pCR after surgery. Patients should also have biopsy at Cycle 2 Day 14 ± 4 days for the RDI score assessment, receive a minimum of 4 T-DXd infusions and undergo surgery in order to be considered as evaluable for the primary endpoint.

The primary outcome will be presented along with the 95% exact binomial CI. If the lower bound of the 95% confidence interval exceeds 30%, we can conclude that an unacceptable pCR rate of 30% can be rejected.

EFS and OS will be estimated using the Kaplan–Meier method, with 95% CI for the set of patients treated with only T-DXd followed by trastuzumab. Comparisons between the two sets of patients in DFS and OS will be made using the Log-rank test, presenting the estimated hazard ratio together with 95% CI.

The SHAMROCK trial was approved by the National Office for Research Ethics Committee for Clinical Trials (NREC-CT), the NREC for Clinical Investigations of Medical Devices and Performance Studies of In Vitro Diagnostic Medical Devices (NREC-MD), the Health Products Regulatory Authority (HPRA) and the HPRA Medical Devices (HPRA-MD). The trial was registered in the European drug regulatory affairs Clinical Trials database (Eudra-CT 2022–002485-32, June 30, 2022) and ClinicalTrials.gov (ClinicalTrials.gov Identifier: NCT05710666).