Geng Report: On methodological and logical red flags in 'Oncolytic peptide LTX-315 plus an anti-CTLA-4 antibody induces a synergistic anti-cancer immune response in residual tumors after radiofrequency ablation of hepatocellular carcinoma' (Cell Death and Disease, 2025)

Verdict

🟠 Highly suspicious. The paper contains what the reporter characterizes as fatal, non-recoverable methodological contradictions and strong signs of template-based ('cookie-cutter') production. Image-level forensic analysis (image reuse, splicing) could not be performed because only text was available; conclusions on figures are limited accordingly.

Key findings

• 🔴 Cross-species reagent misuse (Methodology). In *MATERIALS AND METHODS → Hepa1-6 cells purification*, mouse CD8+ T cells from a C57BL/6 / Hepa1-6 system were reportedly isolated using the "CD8 T Cell Isolation Kit, Human (Miltenyi Biotec, Cat. No. 130-096-495)". Human anti-CD8 magnetic beads are not expected to cross-react with mouse CD8, making the step biologically implausible and suggesting copy-paste of a human-cell protocol or fabrication of the procedure.
• 🔴 Text–figure legend inconsistency on ICD markers. Main text states ICD proteins detected are (Annexin A1, CALR, HMGB); Figure 2 (A, B) and Figure 4 (A, B) legends list (ATP, CALR, HMGB1). ATP is a small nucleotide, not a protein, and is not detectable by Western blot (typically measured by luciferase-based assays). The discrepancy and the choice of ATP as a WB target are strong indicators of fabricated or pasted-in content.
• 🟡 'Template-style' publication pattern. The corresponding author group appears to systematically vary the RFA-immunomodulator-checkpoint inhibitor combination while reusing similar animal models. Cited references include: Ref. 14 (Kan X… 2022, RFA + oncolytic immunotherapy); Ref. 15 (Zhou G, Kan X… 2022, RFA + LTX-315); Ref. 19 (Liu J… 2024, sublethal RF hyperthermia); Ref. 27 (Sun T, Sun B… 2023, RFA + OK-432 + anti-PD-1); Ref. 28 (Cao Y… 2023, RFA + OK-432 + doxorubicin hydrogel). The 2025 paper follows the same template with RFA + LTX-315 + anti-CTLA-4.
• 🟠 Implausibly 'clean' statistics. Across Figures 1–5, the combination treatment (iRFA + LTX-315 + anti-CTLA-4) reportedly shows p < 0.001 for virtually every endpoint, with extremely tight error bars on IFN-β, TNF-α, CXCL9 and CXCL10 measured by qRT-PCR and ELISA at small sample sizes (in vitro n=3; in vivo n=6). Such uniformity in mouse tumor models is described as atypical.

Evidence highlights

• Direct quote of method: "CD8 T Cell Isolation Kit, Human (Miltenyi Biotec, Cat. No. 130-096-495)" applied to a mouse Hepa1-6 / C57BL/6 system.
• Direct mismatch: main text ICD list — "Annexin A1, CALR, HMGB" vs. Figure 2A/B and Figure 4A/B legend — "ICD proteins (ATP, CALR, HMGB1)".
• Sample-size flag: n=3 (in vitro), n=6 (in vivo) producing p<0.001 across nearly all reported panels.
• DOI: 10.1038/s41419-025-07622-z preserved verbatim.
• Catalog number preserved: 130-096-495 (Miltenyi Biotec, Human kit).

Notes

• Image-level analyses (image reuse, PS splicing) were not performed; the report explicitly states that only text was available, so no pixel-level judgments are made.
• Confidence is moderate-to-high on the textual/methodological issues (they are directly verifiable from the manuscript) and low on any image-based conclusions (untested).
• The reporter recommends requesting raw FCS files from the authors for the CD8+ sort, posting on PubPeer focused on the cross-species kit and the ATP-as-WB-target inconsistency, and notifying the journal editorial office. They further suggest auditing the group's other 'template' papers (e.g., refs 14, 15, 19, 27, 28) for the same reagent mislabeling and text reuse.
• Disclaimer: This is an AI-assisted report for academic discussion only; final determination of misconduct requires institutional investigation.