A response-adapted treatment strategy guided by positron emission tomography (PET) imaging and pathologic response may allow selected patients with HER2-positive early breast cancer to avoid chemotherapy while maintaining favorable long-term outcomes.
In the multicenter, open-label, noncomparative PHERGain trial, researchers enrolled 376 patients with stage I to IIIA operable HER2-positive breast cancer measuring at least 1.5 cm across 45 hospitals in 7 European countries between June 2017 and April 2019. Among the patients, 71 were assigned to standard chemotherapy with docetaxel, carboplatin, trastuzumab, and pertuzumab, whereas 285 received an adaptive strategy beginning with trastuzumab and pertuzumab, with endocrine therapy according to hormone receptor status. Another 20 patients were assigned to a separate exploratory group. In the adaptive group, chemotherapy decisions were guided by fluorine (F)-18–fluorodeoxyglucose (FDG)-PET response following 2 treatment cycles and pathologic complete response (pCR) at surgery.
Previously reported primary endpoints were pCR and 3-year invasive disease-free survival (iDFS). The current analysis assessed 5-year iDFS as a prespecified secondary endpoint, alongside distant DFS, event-free survival, and overall survival. Exploratory analyses evaluated associations between circulating tumor DNA (ctDNA) detection and clearance, pCR, and long-term outcomes. Median follow-up was 5.6 years.
Among 267 patients in the adaptive group who underwent protocol-defined surgery, estimated 5-year iDFS was 90%, with 26 patients experiencing an invasive disease event or mortality. Nearly one-third of patients assigned to the adaptive strategy avoided chemotherapy. Distant recurrence occurred in 14 patients (5%), while estimated 5-year distant DFS was 93%.
Among 86 patients who demonstrated an early FDG-PET response and achieved pCR without chemotherapy, estimated 5-year iDFS was 92%. Six patients experienced an invasive disease event, including 1 distant recurrence. Estimated 5-year distant DFS was 99%.
In the standard-treatment group, estimated 5-year iDFS was 97% among 63 patients who underwent surgery. Distant recurrence occurred in 2 patients (3%). The trial was not designed for formal comparisons between treatment strategies, and the standard-treatment group served as an internal benchmark rather than a basis for establishing treatment equivalence.
Overall survival remained high in both randomized groups, with estimated 5-year survival of 98% in each group. There was 1 mortality among 71 patients assigned to standard treatment and 5 among 285 patients assigned to adaptive treatment.
Exploratory ctDNA analyses included 932 plasma samples collected at baseline, following 2 treatment cycles, and prior to surgery, 801 of which were successfully sequenced. Baseline ctDNA was detected in 71% of evaluable patients. Detection increased with disease stage, from 33% among patients with stage I disease to 93% among those with stage III disease, and was more frequent in patients with lymph node involvement.
Baseline ctDNA detection was associated with less favorable long-term outcomes. Estimated 5-year iDFS was 87% among patients with detectable ctDNA vs. 97% among those without detectable ctDNA. Detection following 2 treatment cycles was similarly associated with lower 5-year iDFS (78% vs. 94%).
Early ctDNA clearance was associated with higher pCR rates compared with persistent detection (48% vs. 22%). Patients whose ctDNA cleared early had higher estimated 5-year iDFS than those with persistent detection (92% vs. 76%). However, the association between early clearance and survival was not statistically significant following adjustment.
Presurgical ctDNA clearance was associated with pathologic response. Among patients with detectable ctDNA at baseline, 49% of those who cleared ctDNA prior to surgery achieved pCR vs. 0% of the patients with persistent detection. Although presurgical ctDNA detection was associated with inferior iDFS in unadjusted analyses, the association was not statistically significant following adjustment.
The findings were limited by the trial's noncomparative phase 2 design, modest sample size, and small standard-treatment group. The chemotherapy-free subgroup was defined by treatment response, preventing conclusions about the comparative effectiveness of chemotherapy omission. Incomplete reconsent for extended follow-up may have introduced attrition bias. Additionally, patients with residual disease did not receive adjuvant trastuzumab emtansine, which subsequently became an established treatment option. The exploratory ctDNA analyses were limited by missing samples, use of a tumor-uninformed assay, and lack of adjustment for multiple comparisons.
The findings suggest that PET-guided, response-adapted treatment may offer a chemotherapy-sparing approach for selected patients with HER2-positive early breast cancer.
“Further prospective validation will be required before ctDNA-guided therapeutic decisions can be implemented in routine clinical practice,” wrote lead study author José Manuel Pérez-García, of the International Breast Cancer Center at Pangaea Oncology of the Quiron Group in Spain, and colleagues.
Full disclosures of the study authors can be found in the study.
Source: Nature Communications
