Patterns of relapse in breast cancer: changes over time.
ABSTRACT Adjuvant systemic treatment for breast cancer has evolved resulting in improved outcomes. A relevant question is whether these advances have changed the pattern of distant relapse. Women diagnosed with stage I-III breast cancer were divided into three time cohorts according to changes in adjuvant therapy; A: 1989-1991-CMF chemotherapy in premenopausal and tamoxifen for postmenopausal women; B: 1992-1997-anthracycline chemotherapy and tamoxifen for pre/postmenopausal women; C: 1998-2001-broader use of anthracyclines. The primary endpoint was 5-year cumulative incidence of bone metastasis (BM) as first site of metastasis (FSOM) versus non-bone metastasis (NBM). The ratios NBM/BM in each period were calculated. The eligibility criteria were met by 21,415 cases; Cohorts A: 1989-1991 (n = 3,915), B: 1992-1997 (n = 9,229) and C: 1998-2001 (n = 8,271). Between 1989 and 2001, the percentage of patients receiving adjuvant chemotherapy increased from 23.1 to 34.4%. A decline in cumulative 5-year incidence rates for BM and NBM as FSOM was seen comparing cohort A to C, P < 0.0001. The ratio NBM/BM was significantly increased from 1.53 in the early cohort to 2.00 in the later one, P = 0.0083. The most prominent increase (84%) was in the ER-negative group, chemotherapy treated, P = 0.0272. A significant decline in 5-year cumulative incidence of metastases and an increase in the proportion of NBM as first site of metastasis were observed between earlier and later cohorts. This may reflect the need for more successful adjuvant treatment options for aggressive breast cancer subtypes which are more likely to present with early spread to visceral organs. Understanding patterns of relapse may help design new adjuvant strategies.
- SourceAvailable from: Corina van den Hurk[Show abstract] [Hide abstract]
ABSTRACT: Little is known about time trends in metastases in the patients treated in routine health care facilities without metastases at diagnosis (M0) and about survival after these metastases. Data on 33,771 M0 patients with primary breast cancer diagnosed between 1978 and 2003 were obtained from the Munich Cancer Registry. Survival analyses were restricted to the patients with metastases within 5 years of the initial diagnosis. The incident number of the patients approximately doubled each period and 5-year overall survival increased from 77% in the first to 82% percent in the last period. 5490 (16%) M0 patients developed metastases within 5 years after the initial diagnosis. The hazard of developing metastases was lowest in the most recent period compared to the first period (HR = 0.50, P < 0.001). The hazard of dying after metastases was equal for patients diagnosed between 1978-1984 and 1995-2003 (HR 1.08, P = 0.3). The percentage of the patients that developed bone metastases decreased each time period, but the percentage primary liver and CNS metastases increased. Exclusion of site of metastases in the multivariate analysis led to a 20% (P = 0.02) higher hazard of dying following metastases in the last versus the first period. In the period 1978-2008, unfavourable changes in the pattern of metastases were exhibited and no improvement was observed in survival of the patients after occurrence of metastases. An explanation might be the increased use of adjuvant systemic treatment, which has less effect on the highly lethal liver and CNS metastases than on bone metastases. The increased use also appeared to contribute to the overall prevention of metastases in breast cancer and therefore to improve overall survival.Breast Cancer Research and Treatment 02/2011; 128(3):795-805. · 4.47 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: We aimed to define the clinicopathologic characteristics of breast cancer (BC) patients with brain metastasis (BM) and to investigate the effect of these parameters on survival. Seventy-nine patients diagnosed with BC and symptomatic BM between 1995 and 2011 were retrospectively evaluated. The relationship between clinicopathological features and outcome was investigated. Triple negative patients had the shortest overall survival (OS) while HR(+)HER2(-) patients had the longest (48.2 vs 88.2 months, 95 % CI; p = 0.33). Multivariate analysis demonstrated that luminal A subtype was the strongest positive predictor of prolonged OS (HR 0.48, 95 % CI 0.28-0.84; p = 0.01), while poor performance status (PS) (ECOG 3-4) at BM was the strongest predictor of shortened OS (HR 1.92, 95 % CI 1.21-3.06; p = 0.006). The patients with early-stage BC at diagnosis had BM later than the advanced-staged patients (47 months for Stage I-II disease, 23.2 months for Stage III-IV disease, 95 % CI; p = 0.002). Median survival after BM was 10.2 months (6.4-14 months, 95 % CI). The patients with liver or skin metastases had significantly shorter survival than the patients with only BM (4.8 vs 17 months, p < 0.001 for liver and 4.8 vs 11.1 months, p = 0.04 for skin). Multivariate analysis demonstrated that regardless of the BC subtype, lack of systemic therapy, and liver involvement were independent factors associated with increased risk of death (HR 4, 95 % CI 1.7-9.1; p = 0.001 and HR 2.2, 95 % CI 1.05-4.9; p = 0.036 respectively). Clinical outcome after BM mostly depends on the ECOG PS and the fact that whether the patient received systemic therapy or not. Systemic therapy prolongs survival especially in HER2 positive patients.Clinical and Experimental Metastasis 08/2012; · 3.46 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: We investigate the potential of mammographic parenchymal texture as a surrogate marker of the risk to develop Estrogen Receptor (ER) sub-type specific breast cancer. A case-control study was performed, including 118 cancer cases stratified by ER receptor status and 354 age-matched controls. Digital mammographic (DM) images were retrospectively collected and analyzed under HIPAA and IRB approval. The performance of the texture features was compared to that of the standard mammographic density measures. We observed that breast percent density PD% and parenchymal texture features can both distinguish between cancer cases and controls (A z > 0.70). However, for ER subtype-specific classification, PD% alone does not provide sufficient classification (A z = 0.60), while texture features have significant classification performance (A z = 0.70). Combining breast density with texture features achieves the best performance (A z = 0.71). These findings suggest that mammographic texture analysis may have value for sub-type specific breast cancer risk assessment.Breast Imaging; 06/2012