Pearl
Oncology is often really messy - sometimes itβs less about what βis bestβ and more about what βis not wrongβ
For example, take the CheckMate 649 study which looked at the addition of nivolumab to chemotherapy (FOLFOX or XELOX) in advanced or metastatic, HER2-, upper GI adenocarcinomas (gastric, esophageal, or gastroesophageal junction)
They were looking for overall survival (OS) and progression-free survival (PFS) and initially enrolled all CPS scores but later amended it to include only CPS 5+
The dual primary endpoints were both significant for those with a CPS of 5+
And because they had patients with CPS <5, they also looked at them, and found significant differences in those with CPS 1+ and all comers (no matter the CPS)
This gets a little dicey - groups with CPS <5 are being enriched by those in the 5+ group
Itβs like taking lemon concentrate, adding water, and making lemonade - would it taste like lemonade if you didnβt start with the concentrate? Or would it just taste like waterβ¦?
Now, if youβre the FDA, itβs a fairly black or white decision. There was a statistically significant benefit and reasonable safety profile for patients with CPS 5+, 1+, and even all comers. So the label can be broad and allow clinicians to use it pretty liberally because it worked in those populations and was reasonably tolerated.
If youβre NCCN or ASCO, you look at the data with a little more scrutiny. The intended population was CPS 5+ so that gets the highest grade recommendation (category 1, preferred). The panel recognized the dilution of the data in CPS <5 and graded it less strongly (2B).
If youβre the study sponsor, youβre feeling pretty good about yourself. Weβll never really know the benefit of those with CPS <5 because they wonβt go back and specifically look at it in a randomized controlled trial.
If youβre the front line pharmacist, you end up having to practice a bit like an attorney, itβs all about the interpretation
So, is it wrong to add nivolumab to FOLFOX for a patient with a CPS of 3? No, itβs not wrong. Is it the best choice? We donβt know π€·ββοΈ
Knowing background like this can help you understand and educate others about flaws in our data and decision making and why itβs so important for pharmacists to continue to deepen their knowledge over time. |