When standard cancer therapies fail, patients enter a bewildering gray zone—experimental trials with impossibly narrow criteria, off-label drug combinations, compassionate-use programs buried in paperwork. Most oncologists, stretched thin by caseloads, lack the bandwidth to navigate this maze on behalf of every patient. Now a wave of venture funding, regulatory approvals, and platform launches promises to automate the search. Whether these tools deliver on that promise, or simply add another layer of complexity, is the question doctors and families are starting to ask.
Money Chases the Matching Problem
xCures announced a $46 million Series B round on June 24, pushing the company's total raise past $76 million. The pitch: extract clinical data from patient charts, then deploy AI to surface trial matches and expanded-access pathways for late-stage patients who've exhausted FDA-approved options. Earlier this year xCures inked partnerships with careMESH in March and the Colorectal Cancer Alliance in February, betting that smoother data-sharing and advocacy networks will accelerate enrollment. It's a theory that looks tidy on a pitch deck; execution in fragmented health systems is messier.
Massive Bio has moved just as aggressively. February saw the launch of TrialRelay, a logistics tool meant to untangle enrollment red tape across multiple sites. By April the company published findings from a prospective study—3,804 cancer patients tracked through its platform—in ESMO Real-World Data & Digital Oncology. Later that month came a collaboration with OpenAI aimed at automating pre-screening at scale, followed by partnerships in May with TOUCH BBCA and in June with FOMAT. The flurry of announcements reflects a recognition that the bottleneck isn't just matching patients to trials; it's getting them enrolled before their disease outpaces eligibility windows.
Mount Sinai jumped in as well. The Tisch Cancer Center went live in January with a systemwide AI matching platform, billing itself as the first NCI-designated Comprehensive Cancer Center in New York to deploy an oncology-specific AI tool across all locations. The emphasis was on speed—collapsing the weeks or months a navigator might spend manually sifting through ClinicalTrials.gov into something closer to real time. Whether that speed advantage holds up under the weight of real-world coordination—insurance pre-approvals, tissue shipment, scheduling conflicts—is another matter.
An FDA Nod Opens Reimbursement Doors
On May 29, 2026, Tempus AI secured FDA approval for its xT CDx assay, a 648-gene tissue test for solid tumors. The move matters less for the science—Tempus had been running versions of the test under lab-developed protocols for years—than for the business model. FDA clearance lets Tempus migrate its entire solid-tumor DNA portfolio to Advanced Diagnostic Laboratory Test pricing, a designation that typically unlocks broader payer coverage and makes community oncologists more comfortable ordering the test.
Tempus frames its platform as decision support, stitching together tumor board insights and molecular data. The FDA approval signals that regulators see value in standardizing genomic profiling, even if the therapeutic options those profiles reveal remain scattered across hundreds of trials with maddeningly specific entry criteria. It's a half-step forward: better diagnostics, same fragmented treatment landscape.
Nonprofits Step Into the Off-Label Gap
Cancer Commons, a nonprofit that pairs PhD-level oncology specialists with patients who've run out of standard options, opened enrollment on February 18 for its N-of-1 Observational Clinical Study (NCT07343024). The study formalizes what has historically been an ad hoc scramble—off-guideline treatments pursued under physician oversight, with data collection tacked on. It's an acknowledgment that FDA-approved Expanded Access, the official "compassionate use" pathway, remains cumbersome enough that few doctors outside major academic centers bother with it.
The NIH's Undiagnosed Diseases Network continues accepting applications across multiple sites, though review timelines stretch for months. For oncology patients racing against metastatic progression, that pace is often a non-starter. The network's research-oriented diagnostic process doesn't align well with the urgency most cancer cases demand.
Second Opinions, Now Virtualized
Established health systems have quietly scaled remote second-opinion services. Memorial Sloan Kettering updated its process documentation in January, citing research linking second opinions to lower costs and improved survival. Cleveland Clinic's virtual multidisciplinary program, launched in 2021, remains operational with sample reports dated through 2024. Included Health, which bundles expert medical opinions into employer benefit plans, announced an expansion of its integrated model in late January.
These offerings serve a different cohort than the precision oncology startups—patients seeking validation or refinement of an existing plan rather than a molecular deep dive. Still, they reflect the same friction: community oncologists can't specialize in everything, and patients increasingly expect someone to verify they're not missing a better path. Perhaps the real insight is that confidence, not just data, has become a product.
The Hard Parts Remain
For all the venture capital and AI hype, the practical obstacles haven't budged much. Trial eligibility criteria are often written so narrowly that even a sophisticated algorithm returns zero matches. Genomic profiling reimbursement varies wildly by payer. The logistical slog of actually enrolling someone—scheduling biopsies, coordinating tissue shipment, wrestling with insurance pre-approvals—still lands on understaffed navigator teams who didn't sign up to be project managers.
The funding rounds and regulatory wins signal investor conviction that technology can streamline at least part of the workflow. But conviction and execution are different animals. Whether these platforms measurably improve outcomes for the patients who need them most—those with weeks, not months—will depend on what happens after the press releases go out and the deal announcements fade. Right now, that story is still being written.
