Triapine

Triapine is a small molecule that blocks ribonucleotide reductase - the enzyme that builds the DNA building blocks cells need to copy and repair their genomes. The National Cancer Institute is running a Phase 1 trial (NCT04234568) adding triapine to lutetium Lu-177 dotatate, the targeted radiopharmaceutical Novartis sells as Lutathera, in patients with neuroendocrine tumors [1][2]. The bet: starve tumor cells of DNA repair tools while bombarding them with internal radiation, and you get deeper, longer responses than Lutathera alone delivers. Context that matters: the Phase 3 NRG-GY006 trial of triapine + cisplatin chemoradiation in cervical/vaginal cancer (n=448) read out NEGATIVE in 2025 - adding triapine did not improve overall survival [3][4]. That is the highest-stakes test of the radiosensitization hypothesis to date, and it failed. The NET Phase 1 is dose-finding work in 31 patients with no commercial sponsor - pure NCI mechanism translation, not a registration program, and now running into a much weaker prior.

Triapine is not new. It's been kicking around clinical oncology since Vion Pharmaceuticals developed it as 3-AP in the late 1990s, has cycled through names (PAN-811, OCX-191), and has failed to win an approval in any indication despite dozens of trials. There are no FDA designations on this NET program. The most consequential recent event for the molecule is unambiguously negative: the Phase 3 NRG-GY006 readout in locally advanced cervical and vaginal cancer (NCT02466971, n=448), published in Gynecologic Oncology in March 2025, showed that triapine added to cisplatin chemoradiation did NOT improve 3-year overall survival versus chemoradiation alone [3][4]. That Phase 3 miss is the referendum on whether triapine has a commercial future in radiosensitization, and the answer was no. NCI is also running an early Phase 1 cervical/vaginal program (NCT02595879), a hysterectomy biomarker study (NCT04494113), and - most importantly for this node - a follow-on randomized Phase 2 in metastatic NETs (NCT05724108, n=94, opened 2023 based on Phase 1 safety + preliminary efficacy) comparing Lu-177 + triapine to Lu-177 monotherapy, with overall response rate as primary [5]. Expected readouts on the Phase 1 NET trial: likely 2026. Phase 2 readout: estimated 2027-2028 given enrollment timeline.

Cells need four DNA building blocks (dNTPs) to copy their genome and to patch up damage. Ribonucleotide reductase, or RR, is the factory that makes those building blocks. Block RR and the tumor runs out of raw material - DNA replication stalls and, more importantly here, DNA damage repair stalls. That's why triapine is paired with things that damage DNA: cisplatin in cervical cancer, and now beta-particle radiation from Lu-177 dotatate in NETs [1]. The principle is validated. Hydroxyurea, an older RR inhibitor, is FDA-approved and standard of care in sickle cell disease and certain leukemias, and the synergy with radiation is well-established preclinically. Triapine binds the iron cofactor at RR's active site, which makes it more potent than hydroxyurea but also creates the molecule's signature toxicity - it can oxidize hemoglobin into methemoglobin, a form that can't carry oxygen [6]. At high levels this causes symptomatic oxygen deprivation and requires emergency treatment with methylene blue, which is why it caps how much triapine can be given. Historical incidence of symptomatic methemoglobinemia in triapine trials runs 5-25% depending on dose and population [6]. Recent work has also flagged an immune angle: triapine triggers immunogenic cell death and upregulates FAS on tumor cells, potentially making them more visible to T cells [7]. Whether that matters clinically in NETs - a tumor type with poor immunotherapy response - is unknown.

NCT04234568 is a single-arm Phase 1 in 31 patients with progressive, well-differentiated NETs that overexpress somatostatin receptors. Standard 3+3 dose escalation, primary endpoints are MTD and RP2D, with PK and preliminary response as secondaries [2]. No comparator. Patients receive standard Lutathera dosing (4 cycles of 200 mCi every 8 weeks) plus oral triapine on days 1-14 of each cycle at escalating doses. Design is appropriate for a Phase 1 - you can't ask for a randomized comparator when you don't yet know the safe dose with the radiopharmaceutical. The status is ACTIVE_NOT_RECRUITING, meaning enrollment closed and follow-up is ongoing. What's more interesting is that NCI has already moved to NCT05724108, a randomized Phase 2 (n=94) comparing Lu-177 + triapine versus Lu-177 alone, with ORR as primary endpoint [5]. That trial is also ACTIVE_NOT_RECRUITING - they enrolled before the Phase 1 finished, based on interim Phase 1 safety and preliminary efficacy signals. NCI is moving fast on the hypothesis, which suggests either confidence in the mechanism or, more cynically, an academic program with low bar for advancement. The Phase 2 is the trial that matters for assessing whether this combination clears the noise of Lutathera monotherapy (18% ORR vs. 3% for high-dose octreotide in NETTER-1, midgut-specific [9]; the NET combo trial enrolls broader SSTR-positive disease so population differences will apply).

The model gives this drug a 23% chance of eventually being approved. That starts from a historical approval rate of about 48% for Phase 3 drugs in this area, then adjusts based on ten facts about the trial and the sponsor. The biggest drags on the estimate are a thin or weak sponsor approval record, smaller-than-typical enrollment for this phase, and weak or limited earlier-phase results. A non-randomized design works in the drug's favor, while the remaining factors are close enough to average that they leave the number roughly where it started.

Three concrete failure modes. First, methemoglobinemia. Triapine reliably oxidizes hemoglobin; historical trials report symptomatic methemoglobinemia in 5-25% of patients at standard doses, and a 2026 dose-optimization paper specifically targets mitigation strategies because the toxicity is dose-limiting and predictable [6]. Combine with Lu-177 dotatate, which itself causes anemia and thrombocytopenia, and the hematologic stacking is the most likely Phase 1 dose-limiter. Second, on-target combination toxicity. A 2026 case report (n=1) documented acute carcinoid heart failure during Lu-177 + triapine treatment - mechanistically plausible but a single case, not a population signal [8]. NETs secrete vasoactive substances, radiation-induced tumor lysis can trigger crisis, and adding a second agent may amplify that. Worth watching, not yet a confirmed risk. Third, the efficacy hurdle. Lutathera's NETTER-1 benchmark is 18% ORR vs 3% for high-dose octreotide in progressive midgut NETs [9]; the broader SSTR-positive population in the Phase 2 may show different baseline response. Triapine has to add a clinically meaningful response delta on top of an already-effective radiopharmaceutical in a slow-growing tumor type where measuring response takes years. The Phase 2 (n=94) is underpowered for OS and will live or die on ORR - a soft endpoint in indolent disease. Commercial risk on the back end: no pharma sponsor is currently developing triapine. After the NRG-GY006 Phase 3 miss and 25 years of academic development without an approval, the field has voted with its wallet.

Noise for the commercial pipeline, and the signal for the radiosensitization synergy thesis just got weaker. The Phase 1 NET trial itself won't move markets - no public sponsor, no commercial program, and the data will read out into the academic literature, not an 8-K. The bigger picture: the Phase 3 NRG-GY006 readout - triapine + cisplatin chemoradiation in cervical/vaginal cancer - was NEGATIVE in 2025 [3]. That was the highest-stakes test of triapine as a radiosensitizer, and it failed on OS. The NET combination is a mechanism extension running against a fresh prior of failure, not riding a wave of validation. The competitive landscape also pre-empts triapine as the template: PRRT + PARP inhibitor combinations are already in active clinical trials, including NCT04375267 (olaparib + Lu-177 DOTATATE in SSTR-positive tumors) and NCT05870423 (PARP inhibitor + PRRT) [10][11]. PARP inhibitors are better-tolerated than triapine - no methemoglobinemia, established safety profile across multiple oncology indications. If any DNA-damage-repair amplifier is going to make it as a PRRT companion, the smart money is on PARP, not RR inhibition. We could not find an active trial of a better-tolerated RR inhibitor + PRRT; the RR inhibitor + PRRT lane is essentially triapine alone, which is itself a tell. Check back when NCT05724108 reports ORR - if it meaningfully exceeds Lutathera monotherapy benchmarks in a comparable population, expect renewed interest. If it doesn't, triapine stays a 25-year mechanism story without a drug story, and PARP combos take the PRRT-radiosensitization franchise.