Behind the Blind: The Committees That Decide Whether a Joint Event Is a Drug Injury

Somewhere between the patient's MRI and the FDA's review division sits a committee most clinical trial participants never learn exists. Its members are physicians — orthopedic surgeons, radiologists, rheumatologists — who are unblinded to treatment assignment and independent of the sponsor. Their task is narrow and consequential: to review individual cases of joint deterioration, fracture, or surgical intervention and issue a standardized verdict on what actually happened.

In the development programs for nerve growth factor (NGF) inhibitors such as tanezumab, these adjudication committees became the most consequential — and least visible — actors in the story.

What Clinical Event Adjudication Actually Is

Large outcome trials generate adverse event reports from hundreds of investigators across dozens of countries, each with different imaging equipment, surgical practices, and diagnostic habits. Left unstandardized, this heterogeneity can manufacture or mask safety signals entirely.

The solution is a Clinical Event Committee (CEC), sometimes called an adjudication committee, governed by a written charter. The charter specifies which events require review, what source documents are collected — operative notes, radiographs, MRI reads, pathology reports — and, critically, which definitions committee members must apply. Committee members are typically not employed by the sponsor, do not treat trial patients, and review cases without knowing whether the patient received drug or placebo.

The Definition That Changed Everything: RPOA

For NGF inhibitors, the pivotal adjudication question was not whether joint events occurred — they did, in both treatment and control arms — but how to classify them. The charter's central distinction separated routine osteoarthritis progression from a pattern called rapidly progressive osteoarthritis, or RPOA.

Under the definitions used in the tanezumab program, RPOA was characterized by marked joint space loss occurring over a short interval — commonly defined as more than 1 millimeter of narrowing within twelve months of drug exposure — often accompanied by subchondral bone changes, and typically in joints that were not yet severely diseased at baseline. 'Normal joint replacement,' by contrast, referred to arthroplasty performed for predictable, advanced osteoarthritis consistent with the patient's pre-existing disease trajectory.

The distinction sounds technical. It is not. A case labeled 'joint replacement due to osteoarthritis progression' reads very differently in a regulatory filing than one labeled 'RPOA type I with subchondral insufficiency.' The first is an expected event in an elderly, osteoarthritic population. The second suggests a mechanism-based injury — and, in the NGF program, the adjudicated RPOA rates became the primary reason the FDA issued a refusal-to-approve and later requested additional comparative data against NSAIDs.

Why Definitions Vary and Why That Matters

Adjudication charters are not standardized across the industry. Different programs have used different millimeter thresholds, different time windows, and different requirements for radiographic confirmation. A case meeting one committee's RPOA definition might fail another's. This variability means that cross-trial comparisons of 'adjudicated RPOA rates' should be made with caution — a point regulators themselves have acknowledged.

There is also an inherent tension in the process. Committee members apply definitions drafted before the data are fully known. If a novel pattern of injury emerges that the charter did not anticipate, early cases may be classified under imperfect categories, prompting charter amendments mid-program — a legitimate scientific refinement that critics can fairly describe as moving the goalposts. Both framings have appeared in public advisory committee discussions.

From Adjudication Verdict to FDA Decision

Once adjudicated, these classifications flow into the safety database that regulators actually review. At the FDA, adjudicated event rates feed directly into benefit-risk calculations: how many adjudicated RPOA cases occurred per thousand patient-years on drug versus comparator, whether the signal increased with dose and duration, and whether concomitant NSAID use modified the risk.

The tanezumab experience illustrates the stakes. Early trials that prohibited NSAID co-use showed lower joint-event rates than later trials that permitted them, and adjudicated analyses helped delineate that interaction. Without committee-level classification, the raw adverse event data would have been a tangle of 'arthralgia,' 'osteoarthritis,' and 'joint replacement' codes of uneven meaning.

What Patients Should Know

Participants in pain trials are entitled to understand that their joint imaging and surgical records may travel to an independent committee, and that a panel of physicians they will never meet may determine how their experience is counted. Asking whether a trial uses blinded adjudication — and whether the charter definitions are publicly available — is a reasonable question before enrollment.

For researchers, the lesson is broader: safety signals in subjective, procedure-heavy therapeutic areas are only as reliable as the classification infrastructure beneath them. The committee no one hears about may be the difference between a marketed medicine and a program that ends in a complete response letter.