Competing risks analyses for recurrence from primary breast cancer

Citation
Jw. Chapman et al., Competing risks analyses for recurrence from primary breast cancer, BR J CANC, 79(9-10), 1999, pp. 1508-1513
Citations number
17
Categorie Soggetti
Oncology,"Onconogenesis & Cancer Research
Journal title
BRITISH JOURNAL OF CANCER
ISSN journal
00070920 → ACNP
Volume
79
Issue
9-10
Year of publication
1999
Pages
1508 - 1513
Database
ISI
SICI code
0007-0920(199903)79:9-10<1508:CRAFRF>2.0.ZU;2-E
Abstract
The effects of prognostic factors on local, regional or distant metastasis are standardly assessed separately. Competing risks analyses may be used to assess simultaneously the effects of factors on different types of first r ecurrence. Data for a cohort of 678 primary invasive breast cancer patients accrued between 1971 and 1990, updated to 1995, included type of first rec urrence (local, regional, distant). We investigated the effects of the trad itional factors of age, tumour size, nodal status, ER, PgR, adjuvant therap y (hormones, chemotherapy, radiotherapy) on type of recurrence and time to recurrence for all patients and for those aged greater than or equal to 65. For all ages of patients, there were five factors with significant associa tions with type or time to first recurrence. Adjuvant radiation was the onl y factor which had an effect (P less than or equal to 0.05) on the type of first recurrence: being associated with a reduction in local recurrence. Ag e, nodal status, tumour size and adjuvant chemotherapy all had significant associations across ail types of first recurrence, and in particular with t ime to recurrence for both local and distant metastasis. This indicates a p otential tack of independence in these end-points. For patients greater tha n or equal to 65 years of age, there were no factors which differentially a ffected type of recurrence, while only nodal status and tumour size had sig nificant associations with time to recurrence. Analyses were used to assess simultaneously the effects of traditional prognostic factors and treatment options on type of first recurrence and time to first recurrence. The exte nsion to evaluations with newer prognostic factors would expedite the deter mination and mode of biologic activity for such factors.