KEYNOTE-689 and CAMORAL agree: surgery 88–92% in all 4 arms, bigger stage III effect, benefit via distant control. Yet only 4/63 (6%) completed adjuvant camrelizumab and a third got none. If CAMORAL is real, it’s a neoadjuvant effect. Do we need maintenance IO? #HNCSM
Few considerations regarding CAMORAL (now in @JAMAOnc) neoadjuvant chemoIO for resectable #hncsm:View quoted post on X
1) surgery not response adapted
2) not meant to convert borderline resectable to resectable disease. 6% on both arms clinical stage 4B (presumably on basis of large neck nodes)?
3) surgery window up to 6 weeks post chemoIO. In KN-689 this was ~3 weeks. With NPC induction chemo prior to definitive chemoRT, the interval is 3-4 weeks. 6 weeks may be too permissive a benchmark for future real world translation of this approach. 17% delayed past 6 weeks post-chemoIO reportedly due to Covid.
4) per supplement, progression during neoadjuvant phase was NOT counted as progression event as long as could proceed to surgery. Similar to KN-689. Underscores importance of not selecting borderline inoperable patients for these paradigms and closely monitoring during neoadjuvant phase.
5) radiation #radonc omitted for pCR. I am curious to know what percentage/number of the failures were among these patients which is critical. While there is enthusiasm about de-escalation implications for adjuvant therapy as a result of neoadjuvant approaches, radiation omission should not be considered off trial IMO. KN-689 did NOT included omission of RT; while tempting if pathologic stage drops below stage 3 and margins adequate, don’t have pathologic sufficient response rates with IO alone and presently insufficient evidence. Zinner PMID 41385730 chemoIO trial has shown high rates of recurrence (5/8 pts) where radiation was not completed among patients with mPR/pCR. Perhaps pCR is a better selected cohort for exploring an omission approach on study, but would not consider this off trial. Stage 3-4 HPV negative disease is not the same major de-escalation scenario as early-stage HPV+ OPC: surgical salvage is more morbid and functionally impactful at the primary site and “definitive” RT has suboptimal outcomes.
6) MPR of 48% lower than on other chemoIO regimens where closer to 60ish percent. Is Carbo paclitaxel less cytoreductive than cisplatin from path response perspective, in absence of synergy with something like radiation? I am not sure the IO choice is the reason for the lower path response here and again numbers somewhat small in CAMORAL.
7) Benefit like in KN-689 seems to be in distant control. This is probably the early priming of an intact immune system. I am not sure how much the adjuvant IO does in light of what we know about IO efficacy across disease sites after irradiation of cervical nodes, let alone after both dissection and radiation. NIVOPOST-OP adjuvant IO reportedly improved locoregional and not distant control, but 45% were pathologically stage 4B so perhaps already too advanced to mitigate distant failure risk (assuming comprehensive staging upfront to exclude early Mets) and the outcomes were not positive upon BICR.
Exciting times as we find better strategies for treating stage 3-4 resectable disease where we need to do better!