Today at #ESMO26, the 6-year results of NATALEE will be presented. Ahead of the presentation, my personal view on the survival gains with adjuvant CDK4/6 inhibitors…
The 6-year OS rates in NATALEE were 91.86% vs 89.73%: an absolute gain of 2.1 percentage points (HR 0.809; 95% CI 0.661–0.989; nominal one-sided p=0.019).
The absolute 7-year OS gain in monarchE was 1.8 percentage points (HR 0.842; p=0.027). I do not think these gains should be dismissed as “only 2%.”
In the EBCTCG analysis, adding taxanes to anthracycline-based chemotherapy—an approach we have accepted as standard for years—provided an absolute 5-year OS gain of approximately 1.9 percentage points. The populations and follow-up periods differ, so we cannot directly compare treatment efficacy. However, the standards we use to judge a few percentage points of survival benefit should be consistent.
Moreover, 95.4% of patients in monarchE had already received chemotherapy. This survival gain represents an additional benefit on top of chemotherapy and endocrine therapy.
Patient selection, toxicity, treatment duration, and cost certainly deserve discussion. However, the survival benefit of adjuvant CDK4/6 inhibitors in appropriately selected patients should not be underestimated.
7508 浏览 · 57 赞 · 16 转发 · 5 回复
Adjuvant ribociclib (RIB) + nonsteroidal aromatase inhibitor (NSAI) in patients (pts) with HR+/HER2− early breast cancer (EBC): NATALEE 6-y overall survival (OS) and efficacy outcomes
https://esmocongress.esmo.org#ESMO26 | NATALEE – 6-year results (Pre-presentation abstract) 6-year OS: 92.73% vs 89.38% (HR 0.809; nominal p=0.019). iDFS: HR 0.741; absolute benefit of 4.1% at 6 years. In monarchE, a 7-year OS benefit was demonstrated (HR 0.842; p=0.027). NATALEE also shows a promising OS signal; however, formal statistical significance has not yet been established. Looking forward to the presentation....在 X 查看被引用的帖子
- Dr Amol Akhade@SuyogCancer
@dr_yakupergun As expected neither Presenter nor disscusaant touched upon nominal p value 😀 . May be it is too trivial against MAGNITUDE of benifit .
- Bijoy Telivala@BijoyTelivala
@dr_yakupergun @SuyogCancer This 2 % comes at an astronomical financial toxicity
- Dr Amol Akhade@SuyogCancer
@dr_yakupergun Let's wait and see if Presenter or disscusant takes up this point . 🙂👍
- Yakup Ergün@dr_yakupergun
We can wait an hour or two to hear the study authors’ answer, but my personal view is as follows: Absolute benefits in NATALEE become more apparent over time: within the same data cut, the OS difference increases from 1.2 percentage points at 5 years to 2.1 points at 6 years, while the iDFS difference increases from 4.1 to 5.3 points. Breast cancer–specific survival findings also support this pattern. Given that recurrences and subsequent deaths in HR-positive disease extend over many years, early OS differences should not be viewed as the limit of a treatment’s overall contribution. The statistical distinction needs to be clear. OS is a secondary endpoint in NATALEE, and a confirmatory testing strategy controlling the overall type I error across multiple endpoints and repeated OS analyses was not prespecified. In monarchE, OS was assessed against a prespecified significance boundary accounting for repeated analyses. I wish NATALEE had included a similar framework for OS. However, saying “the p value is nominal, therefore the OS result is not significant” is also an oversimplification. The 6-year OS analysis shows nominal statistical significance: HR 0.809; 95% CI 0.661–0.989; one-sided nominal p=0.019. This does not carry the same evidentiary weight as a multiplicity-adjusted, confirmatory OS result, but it does not invalidate the observed survival difference either. The emergence of an OS benefit with longer follow-up in monarchE illustrates why time matters in this disease. I consider a 2.1-percentage-point absolute OS benefit at 6 years clinically meaningful in this study population. Although there is no head-to-head comparison, I believe abemaciclib and ribociclib may provide long-term benefits of a similar magnitude in appropriately selected populations. This is my interpretation of the available evidence, rather than a demonstrated equivalence. We can certainly discuss NNT, censoring, numbers at risk, uncertainty around the estimates, toxicity, and treatment burden. But if we reduce our assessment to whether a p value crosses a threshold, we risk overlooking clinically valuable gains—particularly in HR-positive breast cancer. Sorry, this turned into quite a long comment, almost like a letter. I just wanted to go into a bit more detail🙏
- Dr Amol Akhade@SuyogCancer
@dr_yakupergun Is this OS statistically significant? 🤔
来源：Yakup Ergün · x.com