Lifyorli Approval: First FDA-Cleared Glucocorticoid Receptor Antagonist

On March 25, 2026, the U.S. Food and Drug Administration cleared a drug aimed at a target no previous oncology therapy had ever engaged. Relacorilant, branded Lifyorli by Corcept Therapeutics, became the first selective glucocorticoid receptor antagonist approved in cancer, paired with nab-paclitaxel for adults with platinum-resistant epithelial ovarian, fallopian tube, or primary peritoneal cancer who had already received one to three prior systemic regimens — at least one of which had to include bevacizumab. The approval rests on a single pivotal study, ROSELLA, whose final overall-survival data — published in The Lancet on April 10, 2026 and presented the same week at the Society of Gynecologic Oncology meeting in San Juan, Puerto Rico — moved median overall survival from 11.9 months on chemotherapy alone to 16.0 months in the combination arm.

For a disease where progression has usually been the only thing the field could reliably extend, that is a different kind of number.

A Cancer That Has Not Had a New Survival Standard in a Generation

Platinum-resistant ovarian cancer is defined more by what does not work than by what does. Patients meet the clinical threshold when their disease progresses or recurs within six months of the last platinum-based regimen. Once a tumor crosses that line, the next standard of care has historically been a single-agent chemotherapy — most often weekly paclitaxel, liposomal doxorubicin, topotecan, or nab-paclitaxel — with or without bevacizumab. Response rates are modest, durable remissions are rare, and the field's prior randomized trials have tended to show progression-free survival gains measured in weeks rather than months, often without a statistically significant extension of life.

That is why the headline comparison in ROSELLA — a four-month gain in median overall survival, from 11.9 months to 16.0 months, in a platinum-resistant population that had already been through bevacizumab — is being read as structurally different from the run of negative or marginal trials the field has absorbed over the past decade. Drug Discovery World summarized it as a "new standard of care" framing, with the NCCN Guidelines adding the regimen as a preferred option for platinum-resistant disease.

Domenica Lorusso of Humanitas Hospital, one of the trial's co-chairs, put the comparative context bluntly: patients need more good treatment options, and relacorilant "provides one. It significantly extends overall survival, with very little added toxicity."

Why a Cortisol-Receptor Antagonist Showed Up in an Oncology Trial

The biology here is older than the trial. The glucocorticoid receptor — the nuclear receptor that endogenous cortisol binds to and that anti-inflammatory steroids like dexamethasone activate — has been known for decades to suppress apoptosis in epithelial tumor cells. That matters, because most cytotoxic chemotherapy relies on induced apoptosis to kill dividing cancer cells. Chronic cortisol tone, or routine steroid pre-medication before chemotherapy, can therefore blunt the very mechanism a taxane is supposed to exploit. Preclinical work over the past decade showed that selectively blocking the glucocorticoid receptor — without touching the progesterone receptor, which close chemical relatives like mifepristone also bind — could restore the sensitivity of ovarian cancer cells to nab-paclitaxel in mouse models.

Relacorilant was engineered to sit in that lane. It is a selective antagonist of the glucocorticoid receptor with minimal cross-reactivity to other steroid-hormone receptors. As FiercePharma summarized the mechanism, the drug works by binding the glucocorticoid receptor to enhance chemotherapy sensitivity and lift cortisol's suppression of programmed cell death — a mechanism framed less as an independent cytotoxic and more as a way of letting the chemotherapy do what a taxane is supposed to do in the first place.

This is the first time that hypothesis has been vindicated by a positive phase 3 survival endpoint in any solid tumor. The approval therefore lands with two simultaneous implications: a new option for a specific ovarian-cancer population, and a proof of concept for a receptor class that had never been approved in oncology. How far that second implication travels — to other chemo-resistant settings, to taxane-sensitive combinations in other tumors — is exactly the kind of question the next wave of trials will have to answer.

Trial Design: ROSELLA Was Built to Survive Its Own Open-Label Problem

The pivotal ROSELLA trial enrolled 381 patients across 117 sites in 14 countries spanning Australia, Europe, Latin America, North America, and South Korea, with randomization running from January through April 2024. Patients had to have platinum-resistant epithelial ovarian, fallopian tube, or primary peritoneal cancer with disease progression or intolerance on their most recent therapy, and had received one to three prior lines of treatment. Prior bevacizumab was allowed at the trial level — it later became a requirement at the FDA-label level, a narrowing worth noting.

The combination arm received oral relacorilant at 150 mg on the day before, the day of, and the day after each nab-paclitaxel dose, together with nab-paclitaxel at 80 mg/m² IV on days 1, 8, and 15 of 28-day cycles. The monotherapy control received nab-paclitaxel at 100 mg/m² on the same schedule, which is the standard single-agent platinum-resistant regimen. The trial used two co-primary endpoints — progression-free survival assessed by blinded independent central review per RECIST v1.1, and overall survival — which matters because an open-label trial in a chemotherapy setting is exposed to investigator-assessment bias. Using central blinded review for PFS and an objective OS endpoint is the design response to that concern.

The trial was run through an unusually broad cooperative-group consortium. Per the final OS publication record, ROSELLA carries the identifiers GOG-3073 (the U.S. GOG Foundation), ENGOT-ov72 (the European Network of Gynaecological Oncological Trial groups), APGOT-Ov10 (Asia-Pacific), LACOG-0223 (Latin American Cooperative Oncology Group), and ANZGOG-2221/2023 (Australia and New Zealand). Listing all five matters: they are not marketing badges but the practical reason the trial could finish enrollment in four months across five continents.

What the Numbers Actually Said

The FDA's decision rested on two comparisons, both of them statistically significant and both of them moving in the same direction.

Median progression-free survival was 6.5 months in the combination arm (95% CI, 5.6–7.4) against 5.5 months on nab-paclitaxel alone (95% CI, 3.9–5.9), yielding a hazard ratio of 0.70 (95% CI, 0.54–0.91) at a two-sided p-value of 0.0076. Median overall survival was 16.0 months (95% CI, 13.0–18.3) versus 11.9 months (95% CI, 10.0–13.8), with a hazard ratio of 0.65 (95% CI, 0.51–0.83) at p=0.0004.

Two things are worth pulling out of those figures rather than leaving them as standalone numbers.

First, the OS gain is roughly four months, and the PFS gain is roughly one. That ratio is noteworthy: the combination did not just delay progression, it appears to have translated the delay into life-extension. That is not guaranteed even when a trial hits both endpoints, and historically in platinum-resistant ovarian cancer the two endpoints have dissociated more often than not.

Second, the confidence intervals around the overall-survival hazard ratio do not cross 1.0 — and they do not come close, with the upper bound at 0.83. Statistically, this is the kind of signal that tends to hold up under further follow-up rather than contract toward the null as data mature. That pattern is already visible within ROSELLA itself: the earlier interim analysis published via PubMed reported OS at 15.97 versus 11.50 months with a hazard ratio of 0.69 (95% CI, 0.52–0.92) at p=0.0121, and the final data moved the hazard ratio lower rather than higher.

Landmark survival rates reinforced the same direction: roughly a ten-percentage-point gap at 12 months and a substantially wider gap by 18 months, favoring the combination. Patients in the relacorilant arm were simply more likely to still be alive at the follow-up windows where platinum-resistant ovarian cancer has historically done its worst attrition.

Safety: Chemotherapy-Dose Adjustment Was the Design, Not an Afterthought

The most common adverse reactions in the combination arm, each occurring in at least 20% of patients, were the familiar cytotoxic-chemotherapy profile: decreased hemoglobin, decreased neutrophils, fatigue, nausea, diarrhea, decreased platelets, rash, and decreased appetite. Label warnings include neutropenia, severe infections, adrenal insufficiency, exacerbation of conditions that themselves require glucocorticoids, and embryo-fetal toxicity. These are not minor; they are the kind of warnings every practicing oncologist reads carefully.

But the key design choice that shapes how those numbers read is that the combination arm used a lower nab-paclitaxel dose — 80 mg/m² rather than the 100 mg/m² used in the monotherapy control. That matters because it frames the comparison less as adding a drug on top of full-dose chemotherapy and more as substituting part of the cytotoxic exposure for an oral sensitizer. The Lancet investigators stated directly that adverse events were similar across study groups when adjusted for nab-paclitaxel exposure and that no new safety signals were observed. Drug Discovery World framed the combination as achieving its efficacy "without an increased safety burden," a phrasing that is worth being slightly careful with — the adverse-event inventory is not zero, but the reduced-chemotherapy design appears to have absorbed the added toxicity of the targeted agent rather than stacking it.

This is where the clinical-practice implication diverges from the trial-design implication. In practice, a physician is not adding a new toxicity to full-dose nab-paclitaxel; they are swapping into a slightly different chemotherapy schedule that includes an oral sensitizer.

Clinical Integration: NCCN, Pricing, and What Patients Will See

Two adjacent developments matter for how quickly this moves into clinics. The National Comprehensive Cancer Network added the relacorilant-plus-nab-paclitaxel combination as a preferred regimen for platinum-resistant ovarian cancer, a listing that substantially shapes U.S. reimbursement pathways and oncology practice. And Corcept has priced Lifyorli at a wholesale acquisition cost of $37,900 per 28-day treatment cycle, with the company expecting broad coverage across Medicare, Medicaid, and commercial plans and offering patient-assistance programs for uninsured or under-insured recipients.

That price — which does not include the nab-paclitaxel line item, which most U.S. cancer centers already administer — positions Lifyorli in the same general tier as other recent branded oncology oral agents. Whether payers absorb that price cleanly at scale will depend partly on the NCCN preferred listing (which tends to reduce prior-authorization friction), partly on real-world outcomes versus the trial, and partly on how quickly any Medicare Part D negotiation cycle catches up.

Alexander Olawaiye of the University of Pittsburgh, who presented the final data at SGO, framed the clinical-integration implication directly: the results "give hope to patients and oncologists now that we have a new preferred regimen in this devastating and difficult-to-treat cancer."

Corcept's Second Act: From Cushing's Rejection to Oncology Approval

The institutional story here is also unusual. Corcept Therapeutics has been known for nearly two decades mostly as a single-drug company — Korlym, a mifepristone formulation approved for the hyperglycemia of Cushing's syndrome. Relacorilant was the company's next-generation, more selective glucocorticoid receptor antagonist, and its lead indication for most of the past decade was also Cushing's. In late 2025, the FDA rejected that application — the agency did not dispute the phase 3 results, but cited a need for "additional evidence of effectiveness" — and Corcept's stock absorbed a steep drop.

The ROSELLA data arrived six months later at a different agency review division, for a different disease, and with a radically different clinical rationale. CEO Joseph Belanoff, speaking to trade press after the March approval, characterized the approval as an "important first step" — language that is not accidentally positioned. The company's pipeline includes several additional oncology combinations with cortisol-receptor antagonism as the mechanistic spine, and whether any of them replicate the ROSELLA survival signal in other solid tumors will determine whether relacorilant becomes a single-indication drug or the beginning of a class.

The structural lesson for drug developers is narrower than the cancer story but worth naming: a target that fails its expected indication on regulatory grounds can still find its first approval in a setting the original hypothesis did not prioritize. In Corcept's case, the pivot was from endogenous-hormone-disease to chemotherapy-sensitization — two nearly unrelated clinical problems tied together by the same receptor.

What This Does Not Tell Us — Yet

For all of ROSELLA's clean endpoints, several questions remain open, and any honest read of the approval should name them.

1. Biomarker selection is unresolved. The publicly verified data have not yet stratified outcomes by BRCA1/2 status, homologous-recombination deficiency, glucocorticoid-receptor expression, or prior PARP inhibitor exposure. For a drug whose mechanism is explicitly about modulating a nuclear receptor that varies in expression across tumors, the absence of a prospectively validated response predictor means all comers currently benefit on average, but the individual-level calculus is not yet worked out.
2. The bevacizumab requirement is a label-level narrowing. At the trial level, prior bevacizumab was allowed rather than required; at the FDA-label level it is required. That creates a real-world gap between the population that benefited inside the trial and the narrower population the FDA label actually covers. Some bevacizumab-naïve patients may pursue access through off-label pathways or coverage appeals.
3. Mechanism-to-clinic causality is indirect. The glucocorticoid-receptor-antagonism hypothesis is biologically plausible and pre-clinically supported, but a positive phase 3 combination trial does not by itself prove that the mechanism caused the benefit. A reduced-dose nab-paclitaxel schedule could contribute independently to tolerability, and subgroup analyses by cortisol exposure have not been published.
4. Durability beyond 18 months is not yet publicly matured. Landmark OS at 12 and 18 months favors the combination by roughly ten and nearly twenty percentage points respectively, but the tail of the survival curves past two years will shape whether this is a delayed-progression story or a sub-population of true long responders.
5. Cross-tumor generalizability is a hypothesis, not a result. The SGRA class could in principle extend to other taxane-treated solid tumors, but ROSELLA is a single-disease trial. Extrapolation is plausible; evidence is absent.

None of these caveats subtracts from the approval itself. They are what the next five years of Corcept and cooperative-group trials will have to answer.

Key Takeaways

• First-in-class at the receptor level. Lifyorli (relacorilant) is the first selective glucocorticoid receptor antagonist cleared by the FDA in oncology, approved March 25, 2026, in combination with nab-paclitaxel for platinum-resistant ovarian, fallopian-tube, or primary-peritoneal cancer after 1–3 prior lines including bevacizumab.
• Survival, not just progression. In ROSELLA, median overall survival moved from 11.9 to 16.0 months, a hazard ratio of 0.65 at p=0.0004 — a rare positive OS readout in platinum-resistant disease.
• Design absorbed the added toxicity. The combination used lower nab-paclitaxel dosing (80 vs 100 mg/m²), and the trial's investigators reported no new safety signals once exposure was adjusted.
• Clinical integration is already moving. NCCN has added the regimen as a preferred option for platinum-resistant ovarian cancer, with Corcept positioning Lifyorli at a $37,900 28-day wholesale acquisition cost and expecting broad payer coverage.
• Corcept's second act, the class's first chapter. After a late-2025 Cushing's rejection, the ROSELLA-driven oncology approval is both the company's first cancer drug and the first regulatory validation of a receptor class whose next trials will decide whether this is a single-indication win or the start of a new oncology toolkit.