Translation of a Pathology-Based Prognostic Marker Profile into a Validated Clinical Tool To Guide Adjuvant Treatment Decisions.

Abstract

Abstract Background: We previously developed a marker profile with a risk of recurrence algorithm for patients with operable, hormone receptor-positive breast cancer to help guide the appropriate level of adjuvant treatment. Development was done using robust cross-validation methods in two patient sets: University Hospital Basel (UHB) and Institut Paoli-Calmettes (IPC). Unlike most similar tests, which assess RNA transcript levels in homogenized mixtures of tumor and surrounding normal cells, our profile is based predominantly on functional protein markers assessed specifically in tumor cells by certified pathologists. In addition, standard clinicopathologic risk factors (tumor size, tumor grade, and nodal status) are incorporated into the algorithm to the extent that they are not replaced by the molecular markers, enabling a global assessment of risk. The goal of this study was to translate the profile into a validated laboratory test available for clinical use and to assess its performance in all available patients.Methods: ER, PR, HER2, EGFR, BCL2, and p53 (IHC) and MYC/8q24 (FISH) assays were validated in a CLIA-certified central reference laboratory. The assays were conducted on TMAs of patients from Royal Perth Hospital (RPH). An algorithm was used to assign a risk score to all stage I-IIIA patients in all three patient sets (UHB, IPC, and RPH) on a scale of 0 to 10+. A pre-determined risk score threshold of 3.8 was used to separate patients into low and high risk groups.Results: In an independent validation of the algorithm in the hormone therapy (HT) only-treated RPH patients (n=144), the low-risk group had a 10-year recurrence rate <5%, and the high-risk group had a recurrence rate more than ten times higher (p<0.0001). When all HT only-treated patients from the UHB, IPC, and RPH patient sets were combined (n=444), the profile separated patients into low and high risk groups with 10-year recurrence rates of 4% and 40%, respectively (HR=10, 95%CI=5 to 20, p<0.0001). In multivariate analyses, the profile was independent and fully replaced the significance of all individual prognostic factors and clinical treatment guideline combinations. A continuous risk curve demonstrated that risk of recurrence increased steadily with higher risk scores (p<0.0001). Within the group of patients with Adjuvant! Online scores >15% (n=228), the profile reclassified as low risk 85% (n=77/92) of pN0 patients and 35% (n=33/90) of pN1 patients, and both of these low-risk subgroups had only a 6% 10-year recurrence rate.Conclusion: The marker profile was validated both analytically and clinically in a CLIA-certified reference laboratory. It replaced and exceeded the prognostic value of all available individual factors and current multivariate treatment guidelines. This justified release to clinicians for diagnostic use and indicates that it is a strong candidate for future comparative studies with other treatment decision-making tools. Citation Information: Cancer Res 2009;69(24 Suppl):Abstract nr 4055.