Chemotherapy combined with anti-HER2 therapy remains the standard treatment for HER2 positive breast cancer. Due to the limited clinical data, the biomarkers for predicting response to neoadjuvant chemotherapy with dual anti-HER2 therapy remain unclear and require further investigation. In our present study, multivariate analyses revealed that PR expression, Ki67 index, and HER2 status were independent predictors of pCR. Based on these results, we constructed a nomogram to predict pCR. According to the nomogram, we found that patients with PR negative, higher Ki67 index and HER2 3 + are more likely to achieve pCR when undergone neoadjuvant chemotherapy with dual anti-HER2 therapy.
Consistent with previous study8, our present study also showed that pCR rates vary according to ER/PR status, with the highest rates observed in ER-PR-HER2-positive breast cancer patients and the lowest in the ER + PR + HER2-positive breast cancer patients. Patients with ER + PR + HER2 positive (triple positive breast cancer) who received the standard neoadjuvant therapy still had the lowest pCR rates. To address this issue, we must first identify the reasons for the difference in pCR rates between HR-HER2-positive and HR + HER2-positive patients.
Firstly, HER2 positive disease is clinically and biologically heterogenous and not all patients benefit equally from the current therapies. Previous studies showed that HER2 disease was biologically heterogeneous and encompassed a spectrum of distinct molecular subtypes (Luminal A, Luminal B, HER2-enriched and basal-like)9–12. A recently published paper showed that HER2 heterogeneity is the most frequent in HR + HER2-positive disease with an incidence of 10% and is associated with lower pCR rates13. However, the current definition of HER2 positive does not sufficiently consider the heterogeneity of HER2-positive disease.
Secondly, as well known, the HER2-enriched subtype is more likely to activate the HER2 pathway and benefit the most from dual anti-HER2 therapy. However, there is a significantly difference in the distribution of HER2-enriched subtype between HR + HER2-positive and HR-HER2-positive (54% vs 81%). Although there was a clear relationship between HER2-enriched subtype and ERBB2 levels, they still provide additional information from each other. So, HER2-enriched subtype and ERBB2 levels should not be considered the same, and their combination into a single variable has a better predictive value for the rate of pCR and the pCR rate was highest in the HER2-enriched/ERBB2-high subtype, which represented 68.1% of HR-HER2-positive disease. However, this proportion significantly decreased to 31.7% in HR + HER2-positive disease14. Thus, the difference in distribution between HR-HER2-positive and HR + HER2-positive disease may be one reason why the pCR rate was significantly lower in HR + HER2-positive (especially in triple positive) than HR-HER2-positive breast cancer patients.
Thirdly, the drivers of HR + HER2-positive may differ from those HR-HER2-positive cancers. Wang et al. reported that some HR + HER2-positive breast cancer cells might be primarily driven by the ER pathway and weakly driven by the HER2 pathway, making them intrinsically less sensitive to anti-HER2 treatment and possible sensitive to endocrine therapy15. This may be another reason why pCR rates were lower in HR + HER2-positive than HR- HER2-positive breast cancer patients. Additionally, there is a complex molecular signaling crosstalk between the ER/PR and HER2 pathways, which may contribute to the low sensitivity to neoadjuvant chemotherapy with dual anti-HER2 therapy in HR + HER2-positive patients16.
Therefore, to improve the pCR rates of the triple positive subgroup, we should take HER2 heterogeneity and ER/PR status into consideration and not merely rely on the routine use of HER2 definition13. One strategy to improve the pCR rates is to combine endocrine therapy with anti-HER2 therapy simultaneously block both ER and HER2 signaling pathways. The MUKDEN 01 trail showed that triple positive breast cancer patients who received CDK4/6i + AI combined with anti HER2 therapy achieved a pCR rate of 30.4%, suggesting that a selected subgroup of HR + HER2-positive patients may benefit from this combined therapy without chemotherapy in the neoadjuvant setting17.
The established neoadjuvant therapies in HER2 positive breast cancer are the AC-THP or TCbHP regimens. However, a recently published study showed that anthracycline combinations do not improve the pCR rates nor survival outcomes18, 19. The TRYPHAENA trail also demonstrated similar efficacy for anthracycline-free versus anthracycline-containing regimens20, while anthracycline-containing regimens had significantly reduced cardiac safety21. Therefore, in clinical practice, the TCbHP regimen, which include doxcetaxel and carboplatin, has gained popularity in recent years and become the standard treatment. However, carboplatin causes hematological adverse events, such as anemia and thrombocytopenia2, 22, 23, which can delay neoadjuvant therapy or surgery and affect therapeutic outcomes24, 25. A recently publish paper mentioned that as healthcare provider, we must weigh the therapeutic benefits against short-term and long-term risks26. Therefore, while pursuing the pCR rate, we should also pay attention to patient treatment compliance and tolerance in the neoadjuvant therapy setting.
To achieve this goal, some clinical trials in HER2 positive early breast cancer are evaluating further therapy de-escalation in selected patients. The WGS ADAPT trail demonstrated feasibility of avoiding overtreatment and individualizing neoadjuvant therapy. In the HR- HER2-positive breast cancer patients, a pCR rate of about 90% with 12 weeks of paclitaxel weekly plus trastuzumab and pertuzumab (THP)27, 28. These results suggest that we may be able to omit the carboplatin and still achieve a good pathological complete response. To further assess whether carboplatin can be de-escalated, the ongoing COMPASS HER2 trail is enrolling patients with stage II and III HER2-positive breast cancer being treated with THP neoadjuvant therapy29. Additionally, results from the phase III KATHERINE trial showed that using TDM1 for non-pCR patients after neoadjuvant therapy can provide a guarantee for prognosis30. The good news is that TDM1 is now available in China and can be covered by medical insurance. Therefore, a de-escalation treatment plan such as omitting carboplatin, may be feasible.
In our present study, we investigated whether there were significant differences in pCR rates among three neoadjuvant chemotherapy regimens: TCbHP, AC-THP and THP. Subgroup analyses demonstrated that there was no significant difference in pCR rate among the neoadjuvant chemotherapy regimens in the population with ER+, PR+, HER2 IHC 2+, and Ki67 index < 30%. This indicates that for patients with these characteristics, THP can be used as an alternative to the TCbHP regimen, when taking the high incidence of grade 3–4 thrombocytopenia observed with TCbHP compared to THP into account 31. However, for ER-PR-HER2-positive patients, the neoadjuvant chemotherapy regimen has a great influence on the pCR rates. Table 6.1 showed that the pCR rate of patients who received TCbHP regimen was 72% (78/108), but the pCR rate significantly decreased to 52% (43/83) using THP therapy.
This means although more than 50% of the patients who using THP regimen achieved pCR, there is still a 20% gap compared with patients who received with TCbHP regimen. The low pCR rate of the THP regimen may be related to the short chemotherapy cycle. The application of six cycles of THP may improve the pCR rate. This year’s updated guidelines also recommend six cycles of THP as one of the neoadjuvant treatment options. Therefore, HER2-positive breast cancers cannot be treated as a homogeneous group using the same neoadjuvant therapy.
To our knowledge, this is the first study in which the pCR of patients with invasive micropapillary carcinoma is explored in a HER2-positive cohort. A previous study showed that invasive micropapillary carcinoma in invasive ductal breast cancer may benefit less from standard adjuvant trastuzumab and chemotherapy than non-invasive micropapillary carcinoma32. Consistent with this, our present study showed that patients with invasive micropapillary carcinoma were hardly achieved pCR. This indicates that HER2-positive patients with invasive micropapillary carcinoma may rarely benefit from neoadjuvant chemotherapy with anti-dual HER2 therapy. Currently, there are still no specific treatment for invasive micropapillary carcinoma. Therefore, the results of this study suggest that oncologist should subject these patients to more intensive monitoring of treatment response and consider other anti-HER2 treatments to improve the pCR rates for HER2 positive invasive micropapillary breast cancer patients.
Some limitations should be taken into account when applying our results. Although the predictive ability of this pCR prediction nomogram constructed based on the combined IHC biomarker was acceptable, it was still not sufficiently refined. First, the IHC based subtype may not be reliable enough to reflect the intrinsic subtype. Second, the routine IHC based HER2 definition cannot precisely represent the amplification level of the HER2 gene33. Third, the AUC of the nomogram was 0.73. Therefore, it is necessary to incorporate more effective biological indicators screened by whole exome sequencing or RNA-seq to further improve and enhance predictive ability and provide individual precision treatment for HER2-positibve breast cancer.