Cediranib

Cediranib (AZD2171) is AstraZeneca's oral pan-VEGFR tyrosine kinase inhibitor that the company essentially shelved after the 2016 ICON6 ovarian cancer Phase 3 produced a progression-free survival signal but no convincing overall survival benefit (final OS HR ≈0.85, p=NS), and which the National Cancer Institute has since adopted as a combination partner with the PARP inhibitor olaparib across recurrent ovarian and advanced prostate cancer programs [1][2][12]. The current Phase 3 readout that matters most - NRG-GY004 in platinum-sensitive recurrent ovarian cancer (cancer that responded to prior platinum chemotherapy and relapsed more than six months later; NCT02446600, n=579) - reported in 2022 that cediranib plus olaparib did not beat platinum-based chemotherapy on PFS (combination HR ≈0.86 vs. chemo, olaparib monotherapy HR ≈1.20, neither statistically significant), so the live question is whether NCI's smaller Phase 2 combinations (NRG-GY023 in platinum-resistant disease, the durvalumab triplet NCT04739800) can salvage a niche use case [3][13]. Commercially, AstraZeneca's $54.1B total 2024 revenue base has moved on to olaparib + bevacizumab (PAOLA-1, approved), so cediranib is now mostly an academic and translational story about whether VEGF blockade can manufacture synthetic lethality with PARP inhibition in tumors that aren't BRCA-mutant [9]. Cediranib's oral dosing remains its one underappreciated practical edge over IV-infused bevacizumab in any combination context.

Not a novel compound - cediranib has roughly two decades of clinical data, a clean ChEMBL profile (max phase 3), and an established VEGFR/PDGFR/KIT inhibition signature [4]. Composition-of-matter patents on AZD2171 trace to the early 2000s and are likely expired or near-expired; any commercial path would depend on combination-use or formulation IP, which AstraZeneca has not aggressively prosecuted. No current FDA designations apply because AstraZeneca never filed for approval; the company gave up the brand candidate name 'Recentin' after the regulator-relevant ICON6 OS endpoint missed. What's active today is NCI-sponsored. NCT02446600 (NRG-GY004) is listed as Phase 3, active-not-recruiting, with results already in the field showing the cediranib+olaparib arm failed against chemotherapy in platinum-sensitive recurrent ovarian cancer [3]. NCT02893917 is testing cediranib+olaparib vs. olaparib alone in advanced prostate cancer (Phase 2, n=90, primary completion currently listed in late 2026 per ClinicalTrials.gov). NCT01064648 is a Phase 1/2 cisplatin/pemetrexed combination in malignant pleural mesothelioma. NCT02345265 is the Phase 2 ovarian study that gave rise to the whole NCI program after Liu et al. reported a striking PFS benefit in Lancet Oncology 2014 (median PFS 17.7 vs. 9.0 months, HR ≈0.42, 95% CI 0.23-0.76) [14]. NRG-GY023 published in Clin Cancer Res in early 2025 in platinum-resistant ovarian after prior bevacizumab - modest signals (objective response rate in the ~10-15% range, no PFS HR meeting predefined success bars), not practice-changing [5]. There is no announced timeline for any regulatory submission because no commercial sponsor is driving one.

Cediranib blocks three families of receptor tyrosine kinases sitting on the surface of cells that build and maintain blood vessels: VEGFR-1/2/3 (the main growth factor receptors that tell endothelial cells to sprout new vessels), PDGFR-α/β (which recruits the pericyte support cells that wrap around capillaries), and KIT [6][7]. Tumors past a few millimeters need new vasculature or they suffocate, so starving them of VEGF signaling slows growth - the same logic that built bevacizumab into a multi-billion-dollar drug. The more interesting mechanism for cediranib's modern life is the PARP synergy hypothesis: blocking VEGF makes tumors hypoxic, hypoxia suppresses the BRCA1/RAD51 homologous recombination machinery cells use to fix DNA double-strand breaks, and that functional 'BRCA-ness' should sensitize tumors to PARP inhibitors even when they are HRD-proficient (homologous recombination deficiency-proficient, meaning tumors that have intact DNA-repair machinery and aren't already sensitive to PARP inhibition on their own) [1]. The biology is plausible and there are preclinical and Phase 2 data supporting it (Liu 2014: PFS HR ≈0.42 in unselected platinum-sensitive recurrent ovarian) [14]. The trouble is the Phase 3 confirmation in platinum-sensitive ovarian (GY004) didn't replicate the benefit against the right comparator - combination vs. chemo PFS HR ≈0.86, olaparib alone vs. chemo HR ≈1.20, neither significant - which is what you'd expect if the synergy is real but small, or real only in a subpopulation no one has identified yet [3].

The lead Phase 3 is NCT02446600 (NRG-GY004), platinum-sensitive recurrent ovarian or fallopian tube cancer, three arms: standard platinum doublet chemotherapy, olaparib monotherapy, or cediranib+olaparib. Enrollment target 579, PFS by RECIST 1.1 (the standardized imaging-based method oncology trials use to measure tumor shrinkage and growth) as primary endpoint, NCI-sponsored [3]. The design is clean - randomized, adequately powered, head-to-head against the chemotherapy backbone that actually defines current standard of care - and that's precisely why the negative result is hard to argue with: when Liu et al. published the GY004 readout, neither the combination nor olaparib alone met the bar against chemotherapy (combination HR ≈0.86, olaparib HR ≈1.20). The complementary Phase 2s - NCT02345265 in recurrent ovarian stratified by HRR (homologous recombination repair) gene status, NCT02893917 in mCRPC (metastatic castration-resistant prostate cancer - prostate cancer that has progressed despite androgen-deprivation therapy), NCT04739800 immunotherapy triplet (durvalumab + cediranib ± olaparib), NCT01064648 mesothelioma - are smaller signal-seeking studies, all NCI-sponsored, all active-not-recruiting, generally with PFS or MTD (maximum tolerated dose, the Phase 1/2 endpoint for finding a safe-but-effective dose level) endpoints [8][11][13]. None is sized for accelerated approval. The deeper concern is patient selection: the BRCA/HRD-mutant ovarian population is already well-served by olaparib + bevacizumab from PAOLA-1, so cediranib needs to find value in HRD-proficient tumors, and the trials haven't been built around that hypothesis with the kind of biomarker stratification that would make a regulator believe it.

Our model gives this drug an 18% chance of eventual approval. That number starts from the historical approval rate for Phase 3 drugs in this area - about 48% - then gets adjusted based on ten facts about the trial and the sponsor. The sponsor's strong track record of getting drugs approved pushes the estimate up, while weak or limited earlier-phase results, a randomized design, and the presence of a comparator arm pull it back down. The remaining factors fall close to average for this stage, so they leave the final estimate near where the base rate set it.

Efficacy risk is the dominant one. The pan-VEGFR mechanism is validated for tumor anti-angiogenesis, but cediranib has lost direct comparisons that matter - ICON6 didn't deliver OS (HR ≈0.85), GY004 didn't beat chemotherapy (combination HR ≈0.86). The PARP synergy hypothesis remains unproven at Phase 3. There is no biomarker that selects responders, so any signal-seeking trial is fishing in an unstratified population [1][3]. Safety risk is mechanism-class: hypertension, fatigue, diarrhea, hand-foot syndrome, proteinuria, occasional GI perforation and bleeding - the standard VEGFR TKI profile, manageable but a real obstacle to combination dosing. Olaparib stacks fatigue and cytopenias on top, which forces dose reductions that erode the combination's theoretical advantage. The one practical advantage worth naming: cediranib is oral, where bevacizumab requires IV infusion every two-to-three weeks, so a positive cediranib+olaparib readout would offer a fully oral regimen in a setting now dominated by an infusion-based combination. Execution risk: NCI cooperative-group trials enroll slowly and the active studies are all active-not-recruiting, so the readouts will come when they come (NCT02893917 prostate is currently slated for primary completion in late 2026, NRG-GY023 long-term analysis has no firm date) and there's no commercial pressure to accelerate. Commercial risk is the killer. Even if a Phase 2 produces a positive PFS signal in a niche population, no large sponsor is positioned to file - AstraZeneca's $54.1B 2024 total revenue is built around its existing oncology assets (olaparib, durvalumab, Enhertu, Tagrisso), and resurrecting a 20-year-old TKI with a complicated tox profile, likely-expired composition-of-matter IP, and no novel exclusivity hook isn't a rational capital allocation [9].

Mostly noise from a commercial standpoint, signal from a scientific standpoint. The interesting question cediranib is still answering is whether VEGF blockade can functionally phenocopy BRCA deficiency well enough to make PARP inhibitors work in HRD-proficient tumors - that's a real biology question with implications well beyond this specific molecule. The cleaner test of the same hypothesis is bevacizumab+olaparib (PAOLA-1), which already showed benefit but only in BRCA/HRD-positive ovarian. Watch for: a positive readout from the NRG-GY023 platinum-resistant analysis stratified by HRR status (Lee et al. 2025 reported only modest activity in Clin Cancer Res - read that one carefully for any HRR subgroup signal) [5], and the prostate combo (NCT02893917) where olaparib has a clearer monotherapy benefit in HRR-mutated mCRPC (prostate cancer carrying mutations in homologous-recombination repair genes like BRCA1/2 or ATM) and the cediranib add-on hypothesis is sharper. For an investor following AstraZeneca, cediranib is not a P&L item - the AZ pipeline that matters is camizestrant, datopotamab, volrustomig. Check back when NRG-GY004 long-term OS reports or when NCI publishes a biomarker-defined responder subgroup. Until then, this is academic ground.