Our NECTIN4 scoring system for predicting enfortumab vedotin response in urothelial carcinoma is just published. 🔬
The short version: NECTIN4 is not a failed biomarker. It has been measured in the wrong compartment.
📊 179 EV-treated mUC patients, multicenter, FISH plus IHC, benchmarked against EV-301 EPAR data.
Why the field thought NECTIN4 was ubiquitous 🧠
In EV-301 the median H-score was 250, with 82.6% ≥150. That near-universal expression underpinned the biomarker-unselected label.
In our cohort, membranous H-score alone was 160. Cytoplasmic was 200. Combined, 260, with 78.2% ≥150, almost exactly mirroring EV-301.
So the trial numbers most likely reflect membranous plus cytoplasmic staining. That single methodological choice is why NECTIN4 looked like it was everywhere. 🎯
And it matters, because only one compartment predicts anything:
✅ Membranous 2+/3+ vs 0/1+: ORR 55.1% vs 25.5% (p<0.001), mPFS 7.1 vs 2.9 months (HR 0.45), mOS 12.3 vs 6.9 months (HR 0.57).
❌ Cytoplasmic: ORR 49.6% vs 39.5% (p=0.4). No PFS or OS signal.
❌ Bulk NECTIN4 mRNA, which cannot resolve localisation: no association either.
Adding cytoplasmic staining inflates positivity and dilutes prediction. Negativity rate drops from 12% to 3.4%. More positives, less information. 📉
The scoring system 🔬
Four tiers (0, 1+, 2+, 3+) adapted from the ASCO/CAP HER2 gastric algorithm, chosen deliberately because it accounts for incomplete membranous staining, the same framework that carried T-DXd to tumour-agnostic approval.
Blinded interobserver testing, 44 cases, three raters, two with no prior NECTIN4 experience:
📈 Fleiss' κ 0.749 across all four tiers.
📈 Fleiss' κ 0.874 for the clinically relevant 0/1+ vs 2+/3+ cutoff. Almost perfect agreement.
Reproducible, teachable in a short training session, deployable in routine practice.
Genomics on top 🧬
NECTIN4 amplification in 25.1% (45/179). ORR 76.2% vs 36.6%. mPFS 12.2 vs 3.9 months (HR 0.40). mOS 30.1 vs 8.5 months (HR 0.33). At 24 months, 58.8% of amplified patients were alive versus 19.0%.
And IHC prescreens for it: 80% of amplified tumours are 3+, under 5% of 0/1+ tumours carry an amplification. Restricting FISH to 2+/3+ captures 97.8% of amplified cases while sparing 29.1% of patients any molecular testing. ⚡
The integrated three-tier model:
1️⃣ Amplified: ORR 76.2%, mOS 30.1 mo
2️⃣ Non-amplified, membranous high: ORR 42.9%, mOS 8.8 mo
3️⃣ Non-amplified, membranous low: ORR 26.1%, mOS 6.9 mo
Both remained independent in multivariable models. C-index for OS improves from 0.557 to 0.621 when amplification and membranous expression are combined. Neither marker alone does what the two do together.
Where I think this lands 🎯
I do not (!) read this as an argument for testing every patient . With EV+P as first-line standard and few alternatives, a negative result rarely changes what you do today and shouldn't change it right now given the high EVP efficacy.
But that is an argument about the therapeutic landscape, not about the biology. And the landscape is changing fast: HER2 ADCs, TROP2, HER3/EGFR, NECTIN4 radioligands and T-cell engagers. The moment two viable options compete for the same patient, the question stops being "does the target matter" and becomes "which target first."
Notably, unlike the HER2-low paradigm, EV-301 showed an attenuated treatment effect in NECTIN4-low disease, PFS HR around 0.9 with confidence intervals crossing 1 (check out Figure 1 of the paper).
⚠️Prospective validation is running: EVOKE (DRKS00034745).
Immense thanks to
@niklas_kluemper and Jonas Saal, to Thomas Büttner, Sebastian Rauch and Fabienne Lange, and to every centre and biobank that contributed. This was a genuinely collaborative effort within the fantastic GUARDIANS and BRIDGE Consortium. 🙏
🔗
doi.org/10.1158/1078-0432.CC…
@Uroweb @AmerUrological @ASCO @PathSoc @OncoAlert @Histo_Journal @andreanecchi @emanuele_crupi @danieleraggi83 @raffcolo @Dr_Aggen @DrRosenbergMSK @h_alahmadie @PGrivasMDPhD @shilpaonc @drenriquegrande @amerseburger @brookmans76 @DrYukselUrun @onkowissen @imedverse @UroDocAsh @urotoday @EUplatinum @JCO_ASCO @CCR_AACR @niklas_kluemper @DrRosenbergMSK @Dr_Aggen @Uromigos @tompowles1 @shilpaonc @PGrivasMDPhD
#Uropathology #urothelial #BladderCancer #NECTIN4 #pathtwitter