Impact of Prior Chemotherapy Administration on Intratumor Heterogeneity of CLDN18.2-Positive Gastroesophageal Cancers

At the 2023 World Congress on Gastrointestinal Cancers, Antonella Cammarota, MD, Humanitas University, Milan, Italy, presented results from a study which found an association between prior chemotherapy administration and intratumor heterogeneity of patients with CLDN18.2-positive, resectable gastroesophageal cancers.

Dr Cammarota stated “we … provided, for the first time, evidence that there's some degree of heterogeneity for claudin as well which is no surprise for gastric cancer, although claudin is a marker which is considered quite widely expressed in gastroesophageal cancer.”

Transcript:

I'm Antonella Cammarota, medical oncologist from Humanitas University, Milano and currently working as a research fellow at the Early Cancer Institute in Cambridge and at Sarah Cannon Research Institute in London, and I had the pleasure yesterday to present our work on claudin expression in operable gastroesophageal adenocarcinoma.

We all know by now that claudin 18.2 (CLDN18.2) is a hot target in gastroesophageal adenocarcinoma and that earlier this year 2 large phase 3 trials have reported improved survival in patients with claudin-positive disease when a first-in-class antibody, zolbetuximab is added to standard chemotherapy. We also know from these trials that the prevalence of this marker is pretty high, up to 40% of esophageal adenocarcinomas could have this marker in their makeup and very limited data about claudin expression in the operable stages.

Therefore, we assessed claudin and PD-L1 expression in a cohort of patients only with operable disease, with paired clinical and genomic and transcriptomic data as well to gain insights into claudin prevalence in these stages, as well as PD-L1 co-expression rates and claudin association with possible clinical, genomic, and [tumor microenvironment] TME features. Also, we try to dig a little bit more into metrics of intratumoral heterogeneity.

What we saw was that adopting the fast trial scoring criteria, which actually used a bit lower threshold for positivity compared to the just mentioned phase 3 trials, 18% of our cases were claudin-positive, A bit less than we might have expected. Also, we saw that despite in a prior work, claudin-positivity was associated with higher transcripts level mostly because there is evidence of hypomethylation of the promoter in claudin-positive cancers, even outside gastric cancer. We couldn’t really confirm this in our dataset, possibly due to the lower number of positive cases.

Moving forward, we haven't really identified substantially different clinical genomic or tumor microenvironment features, even though, again, in prior works, there have been suggestions that claudin has been reported more frequently in diffuse gastric cancers, but genomic instability was not substantially lower in our work or that it might have been associated with more immune suppressive features but again, the tumor microenvironment features in our work were not substantially different between claudin-positive and -negative cases. Similarly, PD-L1 co-expressions was actually not lower, and this again was quite different from what has been reported by pulling the data of the two phase 3 trials, which reported less than 20% of claudin-positive cases being PD-L1 positive. Definitely lower than what we have seen in non-biomarkers selected trials in gastroesophageal cancer such as CheckMate-649 and KEYNOTE-859.

Interestingly, when it came to evaluate which sample-related factors might have affected claudin-positivity in our work, we saw that exposure to prior chemotherapy was substantially correlated with lack of claudin detection, whereas sample age or residual tumor cellularity were not, and so performing a multi-region assessment, leveraging multiple tumor samples that were available for around 40% of our cohort, we could clearly see that in claudin-positive patients who had not received any neoadjuvant treatment concordance between tumor samples was pretty high whereas in patients with claudin-positive disease who had actually received prior chemotherapy concordance between tumor samples was below 30%. And so, we basically provide, for the first time, evidence that there's some degree of heterogeneity for claudin as well which is no surprise for gastric cancer, although claudin is a marker which is considered quite widely expressed in gastroesophageal cancer, as I was mentioning before, in up to 40% of the cases.

This is something that requires investigation in larger datasets for sure and if confirmed, might prompt the evaluation of multiple tumor samples to tackle intratumor heterogeneity as has been previously suggested for other markers. That is the end of my talk, I would like to thank you for the kind invitation to speak about the preliminary results of this work. Thank you.

Source:

Cammarota A, Devonshire G, Miremadi A, et al. CLDN18.2 expression in resectable gastroesophageal cancers: Associated clinical and molecular signatures and impact of neoadjuvant chemotherapy on intratumor heterogeneity. Presented at the 2023 World Congress on Gastrointestinal Cancers; June 28-July 1, 2023; Barcelona, Spain. Abstract SO-12