# Adjuvant Chemotherapy +/- Chemoradiotherapy for Adenocarcinoma of The Pancreatic Head: Results of The Radiotherapy Randomization of NRG Oncology/RTOG 0848

- **DOI:** [10.1200/jco-25-02520](https://doi.org/10.1200/jco-25-02520)
- **Rating:** Strong (3.9 out of 5)
- **Verdict:** A rigorously quality-assured phase III trial that definitively shows no overall survival benefit from adjuvant chemoradiotherapy, constrained by its obsolete gemcitabine-era chemotherapy backbone.
- **Framework version:** v5.0
- **Pipeline version:** v0.4
- **Evaluation date:** 2026-06-17
- **Evaluation page:** https://www.nabu.science/paper/10.1200%2Fjco-25-02520

## Strengths

- **Rigorous real-time radiotherapy quality assurance** Participating centers were credentialed and every radiotherapy plan underwent real-time central review before treatment, with reported per-protocol compliance of 92% among treated patients. This directly addresses the quality-assurance weakness of prior negative trials.
  - Dimension: methodological rigour
  - Location: Methods—Interventions; Results—Patients and treatment; Data Supplement Table S2
- **Definitive null result on a contested question** The well-powered intention-to-treat analysis shows no overall survival benefit (HR 0.96, 90% CI 0.79-1.18), reinforced by a concordant as-treated sensitivity analysis (HR 0.98). This resolves a long-standing controversy for unselected patients.
  - Dimension: contribution
  - Location: Results—Outcomes; Figure 2A
- **Actionable, specific clinical decision context** The trial targets an explicit decision faced by multidisciplinary teams managing resected pancreatic head cancer, with clearly identifiable stakeholders and direct relevance to current guideline pathways.
  - Dimension: application pathway
  - Location: Introduction; Discussion; Context Summary

## Limitations

- **Outdated chemotherapy backbone limits relevance** All patients received gemcitabine-based chemotherapy and none received FOLFIRINOX/mFOLFIRINOX, which is now preferred for fit patients. This is the central constraint on the generalizability of the findings to current practice.
  - Dimension: contribution
  - Location: Table 1; Discussion
- **Affirmative message rests on small subgroup** The practice-supporting node-negative finding derives from a treatment-by-nodal-status interaction in only 91 node-negative patients within a trial whose overall primary endpoint was null. Some conclusion language presents this more strongly than the evidence supports.
  - Dimension: reporting
  - Location: Results—Multivariable; Conclusions; Figure 4
- **Reduced statistical power and treatment refusal** Analysis was triggered at five years of follow-up before reaching the planned 316 events, leaving 270 events and power of 72%, and 14% of the chemoradiotherapy arm did not receive the assigned treatment. These factors lower inferential strength for secondary and subgroup interpretations.
  - Dimension: methodological rigour
  - Location: Statistical Analysis; Results—Patients and treatment; Discussion

## Reliability checks

Concerns (4 of 4 checks).

- **Internal coherence:** Noted. The null overall primary endpoint and the node-negative subgroup interaction are reported consistently across abstract, results, and discussion, but some concluding statements assert the data 'do support the use' of adjuvant CXRT for node-negative patients, which is stronger than the paper's own repeated 'if confirmed' hedging given a null primary trial and a small subgroup.
- **Research conduct:** Noted. The trial is prospectively registered (NCT01013649), reports central randomization, written informed consent, IRB approval per NCTN procedures, CONSORT reporting, and central radiotherapy quality assurance; ethics approval is declared in general terms without an independently verifiable identifier, and no data or code availability statement is present in the provided manuscript text.
- **Reference integrity:** Noted. 1 of 7 cited references could not be verified as real publications, exceeding the threshold for reliable citation. See list.
  - References not verified automatically are resolved manually within 24 hours.

## Post-publication record

- **Status:** Clean
- **Summary:** No retraction or expression of concern found in Crossref (includes Retraction Watch).
- **Last checked:** 2026-10-17
