Background: The I-SPY2 TRIAL has shown that the addition of pembrolizumab to standard neoadjuvant chemotherapy results in improved rates of estimated pathologic complete response (pCR) for hormone-receptor positive (HR+) as well as triple negative breast cancer (TNBC). I-SPY2 has tested a number of novel immunotherapy combinations including the combination of pembrolizumab and SD-101, a synthetic oligonucleotide with cytidine-phospho-guanosine (CpG) motifs that stimulates plasmacytoid dendritic cells (pDC) through engagement of toll-like receptor 9 (TLR9). Biomarkers of response to immunotherapy may be different than chemotherapy, and pathologic response may not fully reflect the benefit from checkpoint inhibitors. Early markers of response to immunotherapy are needed to minimize the risk of immune-related toxicity in patients who are unlikely to benefit. The objective of this study is to assess the change in regional lymph nodes (LNs) by breast MRI in I-SPY2 patients receiving neoadjuvant chemotherapy +/- immunotherapy, and to explore if changes in regional LNs correlate with pCR and residual cancer burden (RCB).
Methods: This is a retrospective study of the prospective multicenter I-SPY2 adaptive neoadjuvant trial investigating all patients enrolled at UCSF in 3 study arms between Dec 2015 and April 2021: 1) Control (weekly paclitaxel 80 mg/m2 x 12 weeks followed by AC x 4); 2) weekly paclitaxel + pembrolizumab 200 mg IV every 3 weeks, followed by AC x 4; 3) weekly paclitaxel + every 3 week pembrolizumab + intra-tumoral SD-101 weekly x 4 then every 3 weeks x 2 followed by AC x 4. Serial pre-operative breast MRIs were performed in all patients at baseline, 3 weeks, 12 weeks, and 20 weeks. A single breast radiologist blinded to treatment arm reviewed all MRIs for all patients included in this study and assessed longest diameter and cortical thickness of the largest ipsilateral axillary LN, number and location of abnormal LNs, and development of new abnormal LNs over time. Rates of pCR and RCB were evaluated in patients who showed evidence of LN growth and/or development of new lymphadenopathy (LAD) compared to those who did not. A two-sample test for equality of proportions to measure the statistical significance at alpha=.5 between control and immunotherapy arms and LN change was used.
Results: A total of 43 patients were included, of whom 16 were in the control group, 11 received pembrolizumab, and 16 received pembrolizumab + SD-101. Median age was 45 years, mean tumor size was 4.8 cm, and 60% of tumors were HR+HER2-negative with the remaining 40% being TNBC. Baseline patient and tumor characteristics were similar between the 3 study arms. LN enlargement and/or development of new LAD over time was significantly more common in patients receiving immunotherapy than patients in the control arm (48.1% versus 6.3%, p=0.006). This was seen in both HR+ and TN subtypes. While there was a numerically higher rate of pCR/RCB-1 pathology in patients with increased and/or new LAD compared to those without (64% vs 51%), this difference was not statistically significant. The association of LN change and immune-related toxicities will be reported.
Conclusion: The addition of immunotherapy to standard neoadjuvant chemotherapy in the I-SPY2 trial was associated with an increase in size of ipsilateral axillary LN and/or development of new LAD on serial breast MRI imaging during the course of neoadjuvant treatment. These changes were not associated with worse pathologic response at surgery and should not be assumed to be due to disease progression. Whether these changes could reflect immunotherapy benefit needs to be investigated in a larger trial with longer follow-up.