NCDB Creasman

Background

National Cancer Data Base (NCDB) population-based analysis. N=4,885 primary vaginal carcinoma cases reported 1985-1994 (~62% of estimated US cases). Epidemiology, treatment patterns, and survival by histology, stage, and modality. Largest US population-based analysis of vaginal cancer outcomes.

Interventions and follow up

Treatment patterns: RT alone in majority (53%), surgery ± RT ~30%, observation ~8%
Primary endpoint: 5yr relative survival by stage and histology
mFollow up: Database-linked, variable

Results

5yr relative survival by stage: Stage I 96%, Stage II 60%, Stage III 44%, Stage IVA 16%, Stage IVB 15%
By histology: squamous 52%, adenocarcinoma 51%, melanoma 14%
Histology distribution: squamous 70%, adenocarcinoma 16%, melanoma 2%, sarcoma 1%
FIGO stage distribution: stage I ~26%, stage II ~37%, stage III ~24%, stage IV ~13%

Adverse events

Main adverse events: Not reported (administrative database)
Toxicity data: Not captured in NCDB

Conclusions

Primary vaginal cancer is rare with stage-dependent survival. Stage I has excellent relative survival (~96%); advanced stages carry poor prognosis. Squamous and adenocarcinoma have similar survival; melanoma is very poor. RT is the predominant treatment modality in practice.

Key Limitations

Administrative database: no detail on RT technique, dose, or brachytherapy use; no concurrent chemotherapy data; possible histology coding errors; mixed quality-of-care era (1985-1994) predating modern IMRT/image-guided brachytherapy; relative survival overestimates cause-specific survival; no recurrence or salvage data.

Clinical Context

Standard reference for baseline vaginal cancer epidemiology and stage-specific survival; confirms vaginal cancer is ~1-2% of gynecologic malignancies, making prospective trials infeasible. Stage I relative survival of 96% supports aggressive treatment of early disease. Stage-specific survival data are widely cited in ASCO/ESMO and SGO practice guidance.

References