Background
Retrospective SEER database analysis. 901 patients with surgically resected localized (LOC, N=275) or regional (REG, N=626) thymoma or thymic carcinoma (T/TC) from 1973–2005. Analyzed impact of PORT vs no PORT on OS and cause-specific survival (CSS). Excluded patients dying within 3 months of surgery to reduce immortal time bias.
Interventions and follow up
Arm A: PORT (dose not specified — SEER limitation)
Arm B: Surgery alone
Primary endpoint: OS and CSS (5-year)
mFollow up: SEER registry, up to 32 year
Arm B: Surgery alone
Primary endpoint: OS and CSS (5-year)
mFollow up: SEER registry, up to 32 year
Results
Localized disease PORT vs no PORT (5-yr CSS): 91% vs 98%, P=.03 — PORT possibly harmful in stage I
Regional disease PORT vs no PORT (5-yr OS): 76% vs 66%, P=.01 — PORT associated with improved OS
Regional disease (5-yr CSS): 91% vs 86%, P=.12 — not significant
PORT subgroup: after extirpative surgery for REG disease: No benefit
Regional disease PORT vs no PORT (5-yr OS): 76% vs 66%, P=.01 — PORT associated with improved OS
Regional disease (5-yr CSS): 91% vs 86%, P=.12 — not significant
PORT subgroup: after extirpative surgery for REG disease: No benefit
Adverse events
Main adverse events: Retrospective registry — toxicity data not captured.
Conclusions
PORT appears to have no benefit and may be harmful for localized (stage I) thymoma/TC, while possibly improving OS for regional disease (stage II–III). However, PORT benefit in regional disease did not reach significance for CSS, and no benefit was seen after extirpative surgery. PORT decisions should be individualized based on stage and resection quality.
Key Limitations
Key Limitations: Major selection bias inherent to SEER registry data — PORT use reflects clinician judgment and unmeasured confounders. RT dose, technique, field, and fractionation are not captured in SEER. "Localized" and "regional" categories may be imprecise compared to Masaoka staging. The 1973–2005 data span includes very old RT techniques that may not reflect modern IMRT-based PORT. Histologic classification was imprecise — WHO classification not available for older cases. CSS analysis limited by SEER's cause-of-death coding accuracy.
Clinical Context
The Forquer SEER analysis is widely cited as evidence against PORT for stage I (Masaoka I) thymoma — the finding that PORT was associated with worse CSS in localized disease aligns with the biological rationale (complete resection alone is curative for stage I). For stage II–IV thymoma, PORT after complete resection is considered based on margin status and histology. Jackson NCDB 2017 found PORT was associated with OS benefit (HR 0.72) overall, with greatest benefit for stage IIB, III, and R1 disease. For thymic carcinoma, PORT is generally recommended regardless of stage due to high recurrence risk.
References