ESMO Congress 2026

Prostate cancer is one of the most heavily discussed urologic oncology topics on social media, and this communication ecosystem is supported by published quantitative evidence: over one year, the hashtag #ProstateCancer generated 110,971 tweets on Twitter with more than 544 million impressions from 32,537 users ^[2]; an earlier analysis showed the tag averaged 6,603 tweets per month, the leading urologic oncology discussion, with activity up 122% in 2014 ^[3]. Channel quality is uneven, however: an assessment of Instagram prostate cancer content found only 9% created by physicians, 90% of low-to-moderate quality, and 40% of posts containing objective information had significant misinformation ^[6]. The ESMO26 prostate cancer data flowing out via conference posts therefore must be copied figure by figure and await confirmation in formal reports.

The heaviest storyline at the meeting was Niven Mehra's report on emerging immune therapies in mCRPC. International real-world data showed that checkpoint inhibition in dMMR/MSI-H mCRPC achieved an ORR of 46%, PSA50 of 60%, median PFS of 7.7 months and median OS of 27 months; the role of immunotherapy in TMB-high but microsatellite-stable disease remains unclear. On T-cell engagers, Xaluritamig (STEAP1×CD3) had any-grade CRS in 72% but grade ≥3 in only 2%, mostly confined to cycle 1 with step-up dosing; Pasritamig (KLK2×CD3, called first-in-class in the post) has treated 174 patients with PSA50 of 42.4% at RP2D. The same session also noted early signals with masked PSMA×CD3, traction for personalized neoantigen vaccines, CAR-T remaining proof-of-concept, and design strategies such as masking/conditional activation and costimulation to improve tolerability and outpatient feasibility.

The second storyline is ctDNA. New analyses from TALAPRO-3 in HRR-mutated mCSPC showed that baseline ctDNA positivity predicted poorer rPFS in both arms; ctDNA clearance rates were higher with TALA+ENZA than PBO+ENZA; clearance by week 5 predicted better rPFS in both arms; and persistently undetectable ctDNA was associated with the most favorable outcomes. The post stressed, however, that early ctDNA dynamics show prognostic promise but are not yet ready to guide routine treatment decisions. Another report noted that HRR mutations detected on liquid biopsy are not all tumor-derived—56% were predicted CH-only alterations, and these patients appeared to derive little benefit from PARi monotherapy. Also reported were AFT-47 (cognitive effects of darolutamide vs enzalutamide), wearable data from LIBERTAS, and the PEACE-6 trial in oligometastatic HSPC.

Against prior knowledge, the advance here is reframing prostate cancer immunotherapy from a yes-or-no question into one of biomarker- and platform-specific matching: checkpoint inhibition is anchored in the dMMR/MSI-H subset, while T-cell engagers may extend immune-based treatment beyond this small subset; ctDNA dynamics add an early prognostic dimension in mCSPC, and CH filtering sounds a caution for PARi patient selection. The limitations are equally clear: the posts are second-hand sources, sample sizes, follow-up and statistical details of each analysis were not disclosed, engager data remain early-stage, and the ctDNA conclusion itself concedes it is not decision-grade—all pending formal abstracts and full publications.