Will a third-party ANDA on mifepristone be filed by year-end 2026?

The economic case for a third-party ANDA against mifepristone is weak: AG is already at 78% volume share at 30% WAC discount, leaving only ~$170M of branded revenue to compete for, which a third-party generic would split with Corcept's AG at thin margins. REMS-like specialty dispensing requirements add ~$20-40M of operational setup. Orphan market (~50,000 US patients) limits share capture. The 10-month window (April-December 2026) is short for a new entrant to both develop, submit, and make the filing observable via Notice Letter or Orange Book. However, some pre-developed ANDAs may exist from other generic manufacturers who were watching Teva's litigation. Base rate for third-party ANDA filings within 10 months of Fed Cir patent loss on small-market orphan drugs is ~20-30%. Net: 0.25.

Key fact: Teva itself did NOT launch a generic post-Fed-Cir loss despite being the litigator, which is a strong revealed-preference signal that the economics are unfavorable. If Teva (the entity with the most information about Corcept's market) declined, third parties with less information would be even more cautious. Additional considerations: (a) Orange Book lag - even if ANDA is filed, Orange Book update to Paragraph IV cert may take 3-6 months, pushing observable date to 2027; (b) Notice Letter disclosure via 8-K is required within 45 days but only for NDA holder - a third-party ANDA filer doesn't broadcast their filing; (c) relacorilant Cushing's status (April 2026 FDA meeting) may further deter ANDAs who would want to wait for clearer market size picture. Probability 0.22.

Partial upside risk: some generic manufacturers may have been preparing ANDA submissions in parallel with Teva's litigation, anticipating a possible patent loss outcome. Large generics (Viatris/Mylan, Amneal, Sun Pharma) have specialty orphan capabilities. If any pre-developed ANDA exists and Fed Cir loss triggers submission, observable via Orange Book or 8-K by Dec 31 2026 is plausible. However, the $761M revenue pool is split across AG (78% volume at 30% discount = ~$412M revenue) and branded (22% volume at list = ~$350M revenue). A third-party generic entering at 80% of AG price would undercut the AG but face paper-thin margins (~10-15% vs list). Only large-volume low-margin generic players find this attractive. Net: 0.28.

Economics strongly deter third-party entry: (1) $761M revenue but 78% is already at 30% AG discount, so branded pool is ~$170M, (2) REMS dispensing adds operational friction, (3) orphan market is small. Teva's non-entry despite winning patent litigation is the clearest signal. Probability of observable ANDA filing via any of the three resolution channels (Orange Book, 8-K Notice Letter, CourtListener litigation) by year-end 2026 is ~20-25%.

Third-party ANDA filings within 10 months of patent loss happen in specialty drugs with >$500M addressable markets where generic economics are viable. Korlym's $761M revenue is at the bottom of this range but not below. Some generic manufacturer may have been developing in parallel and file opportunistically. The resolution criteria are broad (any of three channels), so observability is reasonable if a filing occurs. Probability 0.30.

Combined considerations all point toward low probability: (1) Teva's non-entry, (2) thin margins after AG compression, (3) REMS friction, (4) small orphan market, (5) short 10-month window, (6) resolution ambiguity via Notice Letter disclosure lag. Lean toward 0.20.

AG at 78% volume at 30% discount = thin margins for new entrant. Teva's non-entry is strong signal. Small orphan market + REMS adds friction. 10-month observable window is short. Probability 0.25.

Weight Teva non-entry heavily (strongest market signal). Economics deter. Observability via 3 channels is broad but Notice Letter disclosure delays. Probability 0.22.

Some generic manufacturers may have been developing in parallel with Teva litigation. Pre-developed ANDAs could surface. Economics thin but not impossible. Probability 0.28.