And our supporting institutions: @stanford-cancer.bsky.social, @yalecancer.bsky.social, @penncancer.bsky.social, @georgetownlombardi.bsky.social, and the Tisch Cancer Institute at Mount Sinai.

TY to an incredible team for their work on this labor of love:
Fauzia Riaz, John Vaughn, Huili Zhu, Jimbo Dickerson, Hoda Sayegh, Samantha Brongiel, Elena Baldwin, Melanie Kier, Jacob Zaemes, Caleb Hearn, Sam Abdelghany, Roger Cohen, Ravi Parikh, Josh Reuss, and Liz Prsic.

While we often hope that our patient might be the exception, our data suggests that clinical benefit from IP ICIs is very low. This may be due to the limited OS of hospitalized pts with cancer compared to the time it takes for ICIs to work.

But what about responders? 12% had a response on imaging according to treating MD. We found no clinical or demographic factors associated with OS - including pall care consult prior to rx. 6% of pts had G3 or higher irAEs including 2 deaths.

We looked at 215 patients treated with IP ICIs from 2012-2021. Three-quarters had advanced solid cancers and the same proportion were ICI naive. Median OS from the first dose of IP ICI was 1.55 mo, and among pts with adv solid tumors, 1.28 mo. 25% died in the hospital.