Details

Autor(en) / Beteiligte

• Dean-Colomb, Windy Marie
• Hess, Kenneth R.
• Young, Elliana J.
• Gornet, Terrie
• Handy, Beverly Carol
• Moulder, Stacy L.
• Ibrahim, Nuhad K.
• Pusztai, Lajos
• Booser, Daniel J.
• Valero, Vicente
• Hortobagyi, Gabriel N.
• Esteva, Francisco J.

Titel

Evaluation of circulating biomarkers of bone metastasis in early-stage breast cancer

Ist Teil von

• Journal of clinical oncology, 2012-05, Vol.30 (15_suppl), p.10613-10613

Erscheinungsjahr

2012

Link zum Volltext

Quelle

Elektronische Zeitschriftenbibliothek - Frei zugängliche E-Journals

Beschreibungen/Notizen

• Abstract only 10613 Background: Bone is the preferred site for metastasis of breast cancer, affecting approximately 70% of women with advanced disease. N-terminal of procollagen type 1 (P1NP), c-terminal peptide crosslinks (CTX), Osteocalcin (OC) and interleukin-6 (IL-6) are markers of bone turnover that may have clinical utility as predictors of breast cancer recurrence in the bone. Methods: Serum was collected from 168 patients with stage I/II/III breast cancer prior to treatment from 09/2001 to 12/2008 and stored at -80 C. Serum levels of P1NP, CTX, OC and IL-6 were determined using the Roche’s Elecsys 2010 automated immunoassay system. Correlations of biomarker levels with time to bone metastasis (BM) development were assessed with Cox proportional hazards regression analysis and the Kaplan-Meier method. Results: Among the 168 patients analyzed, 60 patients subsequently developed BM. The biomarkers all had skewed distributions with long right tails and thus were all analyzed on the log scale. Residual analysis suggested non-linear relationships between each biomarker and risk of developing BM during follow-up. Thus, we fit Cox proportional hazards regression models for each biomarker with quadratic polynomials (on the log scale). On univariate analysis, these analyses generated p = 0.33 for IL-6, 0.26 for osteocalcin, 0.40 for CTX, and 0.032 for P1NP. Adjusting for clinical factors (stage, age, race, post menopausal, ER/PR status, HER2 status, nuclear grade) yielded p = 0.0035 for the quadratic polynomial for log P1NP. A cut-point of 75 ng/mL identified patients with a short time to development of BM. The 1, 3, and 5-year freedom-from-BM probabilities were 96%, 77% and 66% for the 150 patients with P1NP values ≤ 75 ng/mL and 88%, 45%, and 36% for the 16 patients with P1NP values > 75 ng/mL. The hazard ratio comparing patients with P1NP values ≤ 75 ng/mL to patients with P1NP values > 75 ng/mL was 3.0 (95% CI, 1.5 - 6.2) and p = 0.0075 ng/mL. After adjustment for clinical factors, the hazard ratio was 3.4 (1.5, 7.6) with p = 0.0026. Conclusions: Serum P1NP levels >75 ng/mL correlate with a shorter time to development of BM in patients with stage I-III breast cancer.