Precision medicine and clinical operations

All of Us Pharmacogenomics: What Diverse Drug-Response Data Means for Oncology Trials

What the All of Us drug-response genetics study means for representation, DPYD testing, clinical review and oncology trial operations.

By Lampgen Clinical Research

AI-generated conceptual illustration of two clinical professionals handing over a genetic test report in an oncology clinic. Text reads: A genetic result needs an owner. Who receives it? Who can act on it? Define the clinical review pathway before dosing.
AI-generated conceptual editorial illustration. The people and report do not depict an actual patient, clinical result or trial record.
16 genes24.4%
Drug-response genes studied; participants with an actionable predicted profile and prior exposure to a relevant medication. This is not a measured adverse-event rate. [1, 2]

What the source reported

The new All of Us pharmacogenomics analysis strengthens the evidence base for studying how inherited genetic differences influence medication response across populations. Its practical value depends on translating validated results into an accessible clinical review process. It does not show that every genetic variant requires a treatment change. [1, 2]

What the November 2026 study found

Haddad and colleagues analyzed whole-genome sequencing and linked health-record information from nearly 100,000 All of Us participants across 16 pharmacogenes. All participants had at least one pharmacogenomic variant. The authors reported that 24.4% had both an actionable predicted profile and prior exposure to a relevant medication with pharmacogenomic prescribing guidance. The institutional release also reports more than 17,000 previously undocumented variants. [1, 2]

These are descriptive findings from an observational dataset. A prior medication record and a genetic profile can identify a potential prescribing opportunity; together they do not establish that a patient experienced toxicity, received an inappropriate prescription or would have benefited from a different treatment. Newly discovered variants require additional evaluation before they can support clinical decisions.

Why broader representation matters

The research expands information on variation and population frequencies that earlier, less representative datasets did not fully capture. This matters because a discovery program can overlook clinically relevant biology when the people contributing data cover too narrow a range of backgrounds. The study highlights previously unrecognized variation and frequency differences, including in African and Americas ancestry groups. [1]

For a development team, the implication is to assess who contributes to the evidence and who can complete the testing pathway. Ancestry is useful for examining evidence gaps. It is not an individual genetic test and should not serve as a shortcut for assigning a drug-response phenotype.

An existing oncology example: DPYD and fluoropyrimidines

The DPYD gene encodes DPD, an enzyme that breaks down fluorouracil. Certain variants that reduce enzyme activity can increase the risk of severe, including fatal, toxicity with fluorouracil or capecitabine. FDA advises genetic testing before initiating these drugs unless immediate treatment is necessary. Its updated labeling advises avoiding use in patients with specified variants causing complete DPD deficiency and individualizing dosing for partial deficiency. [3]

This is existing drug-specific guidance. The All of Us study does not create a universal testing requirement for every medication or authorize a trial team to alter dosing independently. The protocol, applicable labeling and responsible medical review determine the study-specific pathway.

What remains to be demonstrated

This analysis is not a randomized comparison of genotype-guided prescribing against usual care. It does not quantify a reduction in adverse events, improved survival or cost savings from implementing a testing program. Those questions require separate clinical and implementation evidence.

The potential benefit is more informed drug selection and dosing for patients with validated, relevant findings. Access to an appropriate clinical assay, interpretation support and timely result delivery will determine whether that potential reaches patients. STAT’s reporting also describes constraints on the All of Us program, underscoring that building a dataset and sustaining its clinical translation are separate challenges. [4]

What this means for trial teams

The following are Lampgen’s proposed operational checks, not outcomes demonstrated by the study.

Map the result before activation. Define the assay, collection requirements, expected turnaround, delivery destination and backup contact. Specify which results require review before treatment and how urgent exceptions are handled under the protocol and responsible medical process.

Name the clinical reviewer. A report reaching an inbox is an intermediate milestone. Document who reviews it, who can authorize an action and how the decision becomes available to pharmacy and the site team. A CRA can check documentation and escalate gaps; the CRA does not independently interpret variants or prescribe treatment.

Review the pattern across sites. If required results arrive late, investigate whether the cause is collection, shipping, laboratory capacity, reporting or an unclear handoff. Assign the corrective action to the function able to change the process, then check subsequent cases. This is a focused application of SIGNAL™: investigate the observation, act on its cause and learn from follow-up.

Measure access as well as turnaround. Compare completion and delay reasons across sites, referral sources and relevant participant groups. Language support, travel, coverage and sample requirements are questions to investigate, not assumed explanations. Avoid introducing a research procedure that narrows participation without assessing its scientific necessity and burden.

Lampgen Clinical Research supports the definition of oncology CRA and clinical-operations staffing needs around these responsibilities. Start with the work requiring ownership and the escalation pathway the study needs.

Analysis by Tommy Lampley, Jr., M.S., Founder, Lampgen Clinical Research. Published November 13, 2026. The accompanying image is an AI-generated conceptual editorial illustration and does not depict an actual patient record, research participant or study result.

Sources

  1. Haddad et al. — Frequency of pharmacogenomic variation and medication exposures among participants in the All of Us Research Program. Science Translational Medicine, November 11, 2026
  2. University of Pittsburgh / UPMC — Everyone Carries Genes That Affect How Medications Work, November 11, 2026
  3. FDA — Safety labeling update for capecitabine and fluorouracil on DPD deficiency
  4. STAT, Anil Oza — All of Us study finds thousands of new genetic variants affecting drug responses, November 11, 2026
  5. Related Lampgen insight — biomarker screening and specimen workflows
  6. Lampgen SIGNAL™ — illustrative monitoring case

AVAILABLE FOR NEW PARTNERSHIPS

What does your clinical team need next?