CLINICAL INVESTIGATION Breast MOLECULAR EVIDENCE DEMONSTRATING LOCALTREATMENT FAILURE IS THE SOURCE OF DISTANT METASTASES IN SOME PATIENTS TREATED FOR BREAST CANCER FRANK A. VICINI, M.D.,* NEAL S. GOLDSTEIN, M.D., y MICHELLE WALLACE, R.N.,* AND LARRY KESTIN, M.D.* Department of *Radiation Oncology and y Pathology, William Beaumont Hospital, Royal Oak, MI Purpose: To examine the clonality relationships among initial invasive breast carcinoma (IBC), ipsilateral breast failure (IBF), and distant metastasis (DM) to determine the effect of local tumor recurrence on the development of DMs. Methods and Materials: A total of 18 patients treated with breast-conserving therapy who developed an IBF fol- lowed by DMs were studied using a 20 informative-marker, polymerase chain reaction-based allelic imbalance clonality assay. Results: Four relationships were identified. First, in 7 cases, the IBF and DMs were clonally related to the initial IBC as one progressively genetic unstable process. Second, in 3 cases, the IBF and DMs were each clonally related to the IBC but clonally distinct from each other. Third, in 3 cases, the IBC and the IBF were clonally related and the DMs were clonally related to the IBFs, with a weak relationship to the initial IBC. Finally, in 5 cases, the IBF was clonally distinct from the initial IBC (new second primary) and the DMs were clonally related to the IBF and clonally distinct from the initial IBC. Conclusion: These findings provide molecular evidence demonstrating that some DMs can directly develop from IBFs and support the importance of local tumor control in the overall treatment of breast cancer patients. Ó 2008 Elsevier Inc. Breast-conserving therapy, Local recurrence, Loss of heterozygosity, Radiotherapy, Lumpectomy, Survival. INTRODUCTION The impact of local failure on survival in patients with early- stage breast cancer treated with breast-conserving therapy (BCT) remains controversial. Although it has consistently been demonstrated that patients who develop local failure af- ter BCT have an increased risk of developing distant metas- tases (DMs), it is uncertain whether local failure signals a more biologically aggressive tumor or is the nidus for future dissemination (1–3). For many clinicians, local failure after BCT is presumed to have no detrimental effect on survival owing to the belief that breast cancer is a systemic disease at inception (4, 5). As a result, local failure is considered a marker for DMs, rather than a cause. In contrast, others be- lieve that preventing local failure could improve survival by avoiding ‘‘secondary’’ dissemination of the cancer cells di- rectly from the local failure (6). The latter hypothesis has been corroborated by data from three large, prospective, ran- domized trials in which survival was improved in high-risk pre- and postmenopausal breast cancer patients who under- went adjuvant postmastectomy locoregional radiotherapy and systemic therapy (7–9). It is possible that local failure af- ter BCT might have the same deleterious effect on survival as after mastectomy. In a recent meta-analysis of 78 breast can- cer trials with 42,000 patients (radiotherapy vs. no radiother- apy, 23,500; more vs. less surgery, 9,300; and more surgery vs. radiotherapy, 9,300), the importance of local control on survival was demonstrated (10). Differences in local treat- ment that substantially affected local failure rates were shown to avoid about one breast cancer death during the next 15 years for every four local failures avoided and were thought to reduce the 15-year overall mortality. The purpose of the present analysis was to evaluate the re- lationship of ipsilateral breast failure (IBF) in patients who subsequently developed DMs after BCT using a molecular clonality assay. A comparison of the polymerase chain reaction (PCR)–based allelic imbalance gene alteration pat- terns in carcinoma is a well-characterized and reliable method of evaluating the clonality among multiple breast Reprint requests to: Frank A. Vicini, M.D., F.A.C.R., Department of Radiation Oncology, Beaumont Cancer Institute, 3601 W. 13 Mile Rd., Royal Oak, MI 48072. Tel: (248) 551-1219; Fax: (248) 551-0089; E-mail: fvicini@beaumont.edu Funded, in part, by a generous gift from the Alfred Berkowitz Foundation and the William Beaumont Hospital Foundation. Conflict of interest: none. Received Jan 3, 2008, and in revised form Feb 14, 2008. Accepted for publication Feb 15, 2008. 689 Int. J. Radiation Oncology Biol. Phys., Vol. 71, No. 3, pp. 689–694, 2008 Copyright Ó 2008 Elsevier Inc. Printed in the USA. All rights reserved 0360-3016/08/$–see front matter doi:10.1016/j.ijrobp.2008.02.045