NIH pulls the trial step from its physician-scientist training grant

NIH has just pulled the independent clinical trial step out of the K-award pipeline that has launched many American physician-scientists for the last three decades. The grant is not being ended, but the trial that turned trainees into independent investigators no longer comes with it.

The policy change, issued last week by NIH, affects the K series, the federal training grant for early-career clinical researchers. The traditional K trajectory runs formative research, then intervention development, then a pilot or feasibility trial, then independent R-level funding. The trial was the launchpad. Without it, the pipeline stops one rung short.

Three things change. NIH is consolidating individual K NOFOs into a streamlined set of parent NOFOs. New activity codes: KR1 (covering K01, K21, K25, K32), KC1 (K08, K23, K38), KS1 (K02, K07, K18, K24, K26), and KT1/R00 (K99, K22). K00, K12, K43, and K76 are unaffected. K applications no longer require a Data Management and Sharing Plan, on the rationale that K awards are training awards, not research awards. Third, and this is the structural change, K funds may not be used to support the conduct of a clinical trial, a trial feasibility study, or an ancillary study to an existing trial. K awardees may gain trial experience only by joining a mentor or co-mentor-led trial. Anyone who wants to lead a trial as principal investigator must secure separate funding.

NIH's stated rationale is that K awards should "better align" with their "primary purpose of career development." The agency says it is "actively working on additional strategies" to support the development of scientists who can lead rigorous clinical trials, but has not named them. Current NOFOs using the old activity codes expire July 13, 2027. New rules take effect for due dates on or after October 12, 2027.

The trial cut is structural. A K awardee at a research-intensive medical center can still hop onto a senior mentor's funded trial to get the experience. A K awardee at a community-based or under-resourced institution, where mentor-led trials are scarcer, cannot, and is now expected to bring her own R01 or equivalent to the launchpad. That is the equity concern Julie Locher, Professor Emeritus at the University of Alabama at Birmingham, raised in a LinkedIn post that OncoDaily published last week. Requiring access to a mentor's funded trial or to separate institutional or philanthropic money, she wrote, "could unintentionally advantage investigators at large, research-intensive institutions with extensive clinical trial infrastructure and internal funding."

Locher also points out the policy's deeper logic. Under the new restriction, the K is functionally equivalent to the F32 postdoctoral fellowship in this respect: F32 fellows already cannot lead independent trials. The F32 is a training fellowship. The K was, historically, a launchpad. The notice treats them as the same thing.

For current K applicants, the clock is already running. Anyone planning a submission for the upcoming 2027–2028 cycle is reading the new activity codes now, and any institution that built a K training program around independent trial experience has until July 2027 to redesign it. The next cohort of MD-PhDs finishing residency is the first to face the new arithmetic. The K still exists. The question, until NIH names those "additional strategies," is who pays for the trial that used to come with it.