Rafeh Naqash: Happy to share these two papers out today in the same issue of JCOPO_ASCO

Rafeh Naqash, Assistant Professor at OU Health Stephenson Cancer Centre, shared an X post by Hassan Abushukair, a medical student at the Jordan University of Science and Technology:

“Happy to share these two papers out today in the same issue of JCO Precision Oncology – ASCO:

1) multi omics work on Stk11 NSCLC with Caris Life Sciences,
and
2) biomarker prediction in irAEs using machine learning and omics approaches (thread below). Lots of people to thank / Hassan Abushukair, Caris Life Sciences.”

Quoting Hassan Abushukair’s post:

“Excited to announce the publication of our paper on TIME features and irAEs in JCOPO_ASCO!
Can’t thank Dr. Rafeh Naqash enough for this opportunity and his mentorship. Huge shout out to the all-star team, learned a lot from you guys!
Tweetorial below:”

Contrary to ICI efficacy, there is limited research dissecting determinants of ICI toxicity.
We aimed to provide insights into potential TIME parameters correlating with ICI irAEs (anti-PD-1/L1/CTLA-4) across 33 cancer types using clinical, omics and digital pathology data.

Clinical data:
Using the US FDA FAERS database, 58,961 ICI-treated pts were identified, 25.6% had irAEs, of which 41.2% had serious irAEs.
Colitis highest incidence in melanoma (9.7%),
Pneumonitis highest incidence in NSCLC (4.5%).

Clinical Data:
Through 75 clinical trials encompassing 18,512 melanoma/NSCLC ICI-treated patients further confirmed these results.”

Transcriptomics analysis of TIME features:
Using bulk RNA-seq data from TCGA, reporting OR (RORs) were calculated for 33 TIME features
+ve correlations were reported between irAE RORs and median CD8+ T cells fraction, INDEL and SNV neoantigens.

Transcriptomics analysis of TIME features:
A composite model consisting of median number of INDEL neoantigens and CD8+ T cells fraction had superior irAE ROR predictive performance.
Notably, myocarditis and myositis both were associated with INDEL load.

Digital pathology:
Using a cohort of 156 ICI-treated NSCLC pts from DFCI who underwent multiplexed immunofluorescence staining:
There was a trend for higher baseline median CD8+ T cells and irAEs.
Yet, baseline PD-L1 status did not correlate with irAEs.

Digital pathology:
Through another cohort of 378 ICI-treated NSCLC patients, using a machine learning approach for H&E-based TILs detection:
We further confirmed the correlation between baseline median TILs and irAEs development.

Incredibly thankful and proud to be part of such important collaborative work that truly embodies the spirit of global alliance!
A huge shoutout to the entire team who made this possible.

View additional information.
Source: Rafeh Naqash/X and Hassan Abushukair/X