NovoCure's Best Quarter Was Also Its Warning

NovoCure's Best Quarter Was Also Its Warning

On the numbers, this was the best quarter NovoCure has ever printed. Net revenue of $193 million, up 16% year over year. A record 4,730 GBM patients on therapy. Gross margin of 77%. Adjusted EBITDA of positive $16 million against a loss of $3 million a year earlier — the first genuine proof point on a profitability promise that has been restated to investors for at least five consecutive calls. Guidance went up, not down: full-year revenue to $736–743 million, adjusted EBITDA to $24–34 million.

And yet the stock walked into the print down 10.9% over ninety days and 34.4% from a July 23 peak of $19.99 — a peak struck, tellingly, on the exact day of the prior earnings call. The tape had already sniffed out the thing the headline numbers paper over: the pancreatic launch, the entire second-pillar thesis, broke in the third quarter.

The Pax Reversal

Three months ago the tone was unrecognizable. The company was describing pancreatic uptake as a repeat-prescriber flywheel, and in February it had boasted of "over 800 certified prescribers" in the opening weeks. “We are pleased with the launch in pancreatic cancer” — Frank Leonard, Chief Executive Officer · 2026-07-23, Frank Leonard said in July, adding that in the highest-volume sites, “we're seeing repeat prescribing in 50% of the population” — Frank Leonard, Chief Executive Officer · 2026-07-23.

Then in October, the flat admission: “We have seen momentum slow in the U.S. due to the introduction of diraxonrasib” — Frank Leonard, Chief Executive Officer (CEO) · 2026-10-29. Optune Pax finished the quarter with 276 active patients and $3.7 million of net revenue — a sequential step backwards from a launch that had been growing triple digits. The mechanism, as management describes it, is not patient defection but practice paralysis. Pancreata full of patients asking for the new drug, appeals paperwork, skin-irritation management, and a device company that needs training, certification, and face time to sell anything at all. As Leonard put it, “we are not seeing patients leaving Optune Pax in substantial numbers to start diraxonrasib” — Frank Leonard, Chief Executive Officer (CEO) · 2026-10-29 — the problem is a share of voice problem, not a share-of-patient problem. Yet.

The keyword set captures the change precisely. diraxonrasib did not exist in NovoCure's vocabulary a quarter ago; it now sits at the top of the momentum table alongside Verastem and KRAS inhibitor. The competitive threat arrived with a name, a prescriber base, and a paperwork burden. This is not a sector-wide theme — cross-sectionally, the KRAS wave shows up in oncology reporters like Chugai (KRAS G12C) and Merck, but NovoCure's exposure is uniquely defensive: it is the incumbent being displaced from a physician's calendar.

If You Can't Beat KRAS, Enroll With It

The strategic answer is the most interesting thing on the call, and it is a genuine pivot rather than a talking point. NovoCure is no longer positioning Tumor Treating Fields against KRAS inhibition; it is positioning it underneath. Two small safety-and-feasibility studies, PANOVA-5 with Verastem's VS-5 in the approved locally advanced setting and PANOVA-6 with diraxonrasib itself in the metastatic setting, both slated for the first half of 2027. concomitant use is the quarter's defining phrase — and the preclinical rationale, TTFields downregulating c-Myc downstream of RAS, has been a slow build across prior calls rather than a fresh improvisation.

But read Uri Weinberg's framing carefully: roughly 25 patients per study, single-arm, first and foremost "to give physicians the right level of confidence and experience to see the feasibility." These are relationship-maintenance trials, not registrational ones. That is a materially smaller ambition than the "second major revenue pillar" language of six months ago.

But we also know that the KRAS therapies are not cures, and we are gonna see a wave of recurrence later this year as well. If the data that we have seen in the clinical trials bears out, which we expect it to. And that is where we are really going to see, I think, the opportunity to, 1, get back into the practices, but really find the sweet spot for tumor treating fields in this emerging landscape.

That is an Executive Chairman conceding the near-term fight and betting on the relapse market. It may even be right. It is not the pitch investors bought in the spring.

The Profitability Math Actually Works

Strip out pancreatic and the core engine is doing what it promised. GBM patient growth of 11% is coming from both established and new markets — the U.S. growing again, Spain and Japan layering on. Gross margin of 77.6% in the June quarter, up 3.7 points year over year, is the number that makes the EBITDA story credible: as supply-chain work on the HFE array matured, the cost of treating each additional patient fell. R&D is coming down as KEYNOTE D58 finishes enrollment by year end, and management guided 2027 R&D to be flat to slightly declining. SG&A fell 9%.

The adjusted-EBITDA milestone — step one of the three-step path management laid out — has now been cleared a full year ahead of the 2027 target set in October 2025, when the CFO anchored breakeven to "a revenue level of around $700 million to $750 million." They will land inside that band in 2026. “Step 1 is to get adjusted EBITDA breakeven, step 2 is cash flow breakeven, and step 3 is net income breakeven” — Christoph Brackmann, Chief Financial Officer (CFO) · 2026-10-29, Christoph Brackmann reminded the line — and he explicitly refused to straight-line it. Q4 guidance of $185–192 million implies a sequential decline from Q3, absorbing the $4.4 million one-time benefit from improved U.S. and German approval rates. The progress is real but the quarterly net loss has not yet gone to zero, and the cash pile has thinned: effective net cash sat at $214 million in June, down sharply from the $826 million peak.

That guidance conservatism is itself a recurring pattern — Jason Bednar probed the identical issue in July — and at 2.5x price to revenue versus 42.2x at the 2021 peak, the market has already repriced this as a profitability story rather than a growth story. Which is why an EBITDA beat mattered more than the revenue beat, and why the Pax stall mattered more than both.

The Catalyst That Keeps Not Arriving

Watch what is missing. For four straight calls, NCCN guideline inclusion for pancreatic cancer has been framed as the unlock that triggers payer engagement. In October 2025 the CEO said the application was reviewed in early July and expected publication "in the upcoming months." In April they had "filed with the NCCN." In July it was "the next major milestone." This quarter: “We have submitted. We understand that we have been reviewed and we are, you know, eagerly waiting for those results to be published” — Frank Leonard, Chief Executive Officer (CEO) · 2026-10-29. A year of waiting on a document that is published annually, with one commercial payer win (Elevance) to show for it. The keyword's momentum has faded hard relative to where it stood earlier in the year — the market has stopped paying for the promise.

The fresher catalyst is Optune Maya — the brain-metastases PMA, decision expected by year end, launch early 2027, with management explicitly modeling a two-year reimbursement runway. And Japan remains the cleanest story in the portfolio: the launch in Japan delivered 110 Lua patients and $4 million in a single quarter, with a negotiated monthly price roughly in line with global Gio pricing, pancreatic reimbursement expected in Q1 2027, and — critically — a lung-and-pancreatic incidence profile that makes Japan structurally larger than GBM ever allowed it to be.

The honest summary: this is a company that finally became profitable-ish on the back of a durable GBM franchise, got unlucky in the sequencing of a competitive launch, and responded by repositioning its science rather than its guidance. The multiple already reflects the disappointment. What it does not yet reflect is whether the KRAS wave is a two-quarter disruption or a permanent tax on the pancreatic opportunity. That question gets answered by prescription data, not by PANOVA-5.