•  
  •  
 

Article Title

Survival After Recurrence of Stage I–III Breast, Colorectal, and Lung Cancer

Publication Date

8-15-2016

Keywords

cancer recurrence, cancer outcomes

Abstract

Background/Aims: Cancer recurrence is common, costly and lethal. Little is known about it in community-based populations, in part because electronic health records and tumor registries usually lack recurrence status. We examined determinants of survival duration after recurrence of breast, colorectal and lung cancers using two databases with excellent recurrence and death data.

Methods: Recurrence data were abstracted from medical records in the Cancer Care Outcomes Research and Surveillance (CanCORS) study and by two Cancer Research Network (CRN) sites — Kaiser Permanente Colorado and Kaiser Permanente Northwest. Death data were extracted from death certificates collected from the National Death Index or health plans. Claims and utilization data provided information on comorbidities and treatments. CanCORS enrolled only those patients who consented, whereas the HCSRN-CRN sample included all eligible patients who received care. This analysis included 1,653 patients who developed recurrence after having had definitive therapy for stage I–III cancer and were followed for at least one day after recurrence.

Results: Restricted mean survival times (RMST) after recurrence for breast, colorectal and lung cancer patients were 34.0, 27.7 and 16.9 months, respectively. Multivariable models revealed lower RMST in breast cancer was associated with: age ≥ 70 (vs. < 55), higher comorbidity burden (2+ vs. 0), stage III (vs. stage 1), chemotherapy for primary tumor (vs. no chemotherapy), and distant/regional recurrence vs. local. Higher RMST in breast cancer was associated with: longer time to recurrence, and stage III and radiation for primary tumor (vs. stage I or no radiation for primary tumor). Lower RMST in colorectal cancer was associated with: age ≥ 70 (vs. < 55), comorbidity burden ≥ 1 (vs. 0), distant/regional recurrence (vs. local), and radiation therapy for recurrence (vs. no radiation). Higher RMST in colorectal cancer was associated with: chemotherapy for recurrence but not for primary (vs. no chemotherapy for primary or recurrence). Higher RMST in lung cancer was associated with: income ≥ $60,000 (vs. < $20,000). Lower RMST in lung cancer was associated with: stage III (vs. stage I), and interaction of income ≥ $60,000 and comorbidity burden ≥ 2 (vs. income < $60,000 or comorbidity of 0).

Conclusion: Breast and colorectal, but not lung, cancer patients with local recurrences had longer survival than those with regional/distant recurrences. Shorter time to recurrence was associated with inferior survival after recurrence among breast cancer patients only. Original cancer stage was a predictor of survival after recurrence; original cancer grade was not.

Share

COinS