背景：在HR+/HER2-N0-1早期乳腺癌中，RSTM＞25的絕經後患者和相當大比例的絕經前患者似乎受益於輔助化療（CT）和內分泌治療（ET）。然而，這種治療強度的絕對獲益程度似乎取決於臨床病理和生物學預後因素。首次報告前瞻性III期WSG-ADAPTHR/HER試驗中CT結合了基線RS評分及動態（Ki67反應）生物標誌物的結果，以優化管腔型乳腺癌的輔助治療決策。
Background: In HR+/HER2- N0-1 early BC, postmenopausal patients (pts) with RS™> 25 and a substantial proportion of premenopausal pts seem to benefit from addition of adjuvant chemotherapy (CT) to endocrine therapy (ET). However, the magnitude of absolute benefit from this treatment intensification seems to depend on clinical-pathological and biological prognostic factors. For the first time, we present outcome from the CT part of the prospective phase III WSG-ADAPT HR+/HER- trial combining both static (RS in baseline core biopsy (CB) and dynamic (Ki67 response) biomarkers to optimize adjuvant therapy in luminal EBC.
方法：對臨床高危型 HR/HER2-EBC（cT2-4 或臨床 N+或 G3 或 Ki67＞15%）的乳腺癌患者，在手術或後續治療前先給予標準內分泌治療治療 3 （+/-1） 周（絕經後 AI；絕經前：TAM）。然後 cN2-3 或 G3/Ki67＞40%的患者直接分配進一步化療組。RS 評分 0-11 的 pN0-1 患者或 RS 評分 12-25 患者中內分泌反應可（內分泌後腫 中心區 Ki67＜10%）者僅接受單純內分泌治療；剩下的高危患者在初始內分泌治療後隨機進入化療組接受輔助劑量密集化療（4xPaclitaxela√4xEC q2w vs. 8xNab-Paclitaxel q1wa√4xEC q2w）。主要研究終點是 CT 生存期有效比較；次要終點包括關鍵預後因素對生存率的影響。Kaplan-Meier 和 Cox 比例風險模型用於估計生存曲線和風險比。在這一分析中，通過 RS/ET 反應確定了無選擇偏差的亞組。
Methods: Pts with clinically high-risk HR+/HER2- EBC (cT2-4 OR clinically N+ OR G3 OR Ki67>15%) were initially treated by 3 (+/-1) weeks of standard ET (postmenopausal: mostly AI; premenopausal: TAM) before surgery or sequential CB. Pts with cN2-3 or G3/Ki67>40% were randomized directly to the CT trial. pN0-1 pts with RS0-11 OR RS12-25/ET-response (central Ki67postendocrine<10%) received ET alone; the remaining high-risk cohort was randomized to the CT trial: (neo)adjuvant dose-dense CT (4xPaclitaxelà4xEC q2w vs. 8xNab-Paclitaxel q1wà4xEC q2w) followed by ET. Primary endpoint is efficacy comparison of CT schedules for survival; secondary endpoints reported here involve impacts of key prognostic factors on survival. Kaplan-Meier and Cox proportional hazard models were used to estimate survival curves and hazard ratios. For this analysis, subgroups free of selection bias by RS/ET-response were defined.
結果：篩查 5625 例患者，4621 例（ITT）意向性人群入組。中位隨訪 4.9 年後，內分泌後較高基線的 Ki-67 水平與較差的 iDFS 相關（均 p＜0.001）。在化療組（n=2331）中，高 RS 評分、淋巴結狀態和腫瘤大小通常與較差的 IDF 相關。然而，iDFS 在 N1 and N0 患者中的差異僅存在年齡＜50 歲患者中。在 LN＞4 陽性的患者中（n=390），低 RS 與 iDFS 改善相關（RS0-11vsRS＞25：plogrank=0.016，5y-iDFS90%vs.64%）。在 RS＞25（n=965）的患者中，內分泌後低 Ki67、N0 狀態和 c/pT1 狀態與 iDFS 改善相關。尤其是內分泌治療應答有反應者的 5y-iDFS（84%）高於 ET 無應答者（77%）；plog-rank=0.040）。年齡小於 50 歲的 N0-1 RS 12-25 的患者中， 無內分泌治療應答反應的患者後續接受化療的 5 年 iDFS（89%）與有內分泌應答反應僅接受內分泌治療的 5 年 iDFS（92%）相比無顯著性差異（plog-rank=0.249）。
Results: 5625 pts were screened and 4621 (ITT) entered the trial. After 4.9y median follow-up, higher baseline and post-endocrine Ki-67 levels were associated with poorer iDFS (both p < 0.001). In the CT cohort (n = 2331), higher RS, nodal status, and tumor size were generally associated with poorer iDFS. However, iDFS differed between N1 and N0 status only among younger pts (<50 years). In pts with >4 positive LN (n = 390), lower RS was associated with improved iDFS (RS0-11 vs RS > 25: plog-rank= 0.016, 5y-iDFS 90% vs. 64%). In pts with RS > 25 (n = 965), low Ki67postendocrine, N0 status, and c/pT1 status were associated with improved iDFS. In particular, ET-responders had higher 5y-iDFS (84%) than ET-non-responders (77%; plog-rank= 0.040). Younger patients (<50 years old) with N0-1 RS 12-25/ ET-non-responders treated by CT had non-significantly poorer 5-year iDFS (89%) compared to those with ET-response treated by ET only (92%) (plog-rank= 0.249).
結論：大型前瞻性 III 期臨床研究 ADAPT 試驗的前瞻性高危隊列研究結果首次為部分 LN＞4 陽性和低 RS 評分的乳腺癌患者的良好預後提高研究證據。此外，內分泌治療後較低的 Ki-67 和有限的腫瘤負荷可能是 CT 降階治療策略選擇的一個有價值指標。
Conclusion: First results from the prospective high risk cohort from a large prospective phase III ADAPT trial provide evidence for good prognosis in some pts with >4 positive LN and e.g. low RS. Moreover combination of lower post-endocrine Ki-67 and limited tumor burden may be a promising criterion for CT de-escalation strategies even in patients with high RS.
