functional significance of the 2 SNPsto provide a molecular rationale andvalidity for their epidemiologicobservations. However, manyquestions remain to be addressed. Forexample, did the different doses ofanastrozole (1 mg/day or 10 mg/day)or letrozole (2.5 mg/day or 10 mg/day)in the neoadjuvant study affect thedegree of suppression of the tumoraromatase activity post-AI therapy?Did body mass index or obesity affectthe estrogen level in the follow-upstudy? What if the electrophoreticmobility shift and reporter gene assayswere performed using the morerelevant estrogen receptor–positivebreast cancer cell lines or fibroblasts?Thus far, none of the coding SNPs ofCYP19 have been shown to alter theinhibitory effect of the third-generation AIs.1 Is it possible that thenon-coding SNPs of CYP19, such asthe 2 SNPs identified in this study, thatlead to an increase in aromataseproduction may overwhelm theseagents? CYP19 is subjected to tissue-specific transcriptional regulation

through the various untranslated exon1s. How are these SNPs involved inthis process in various normal andtumor tissues?

Unfortunately, the SNPs did notcorrelate with tumor response to AIsin this study by Wang and colleagues.One could attribute the finding to thesmall sample size of the study. How-ever, it may not be too surprising,considering the absence of a demon-strated correlation between baselinearomatase expression and clinicalresponse in the large phase IIIneoadjuvant trial of tamoxifen versusletrozole (P024 trial).2 What aboutlong-term clinical outcome? In theP024 study, the presence of tumoraromatase expression at baseline wasa favorable independent prognosticfactor for both relapse-free survivaland breast cancer–specific survival.2

Based on this observation, one wouldimagine that the variant alleles of the2 SNPs would predict a favorableclinical outcome; however, how do wereconcile the finding by Wang andcolleagues that these variants were

Breast D

actually associated with a higher post-AI estradiol level? Regardless, alarger study to validate their findingsis needed, and the relationship ofthese 2 SNPs with the long-termoutcomes of patients on these highlypotent AIs remains to be seen.Furthermore, the considerableheterogeneity of estrogen receptor–positive breast cancer adds anotherlayer of complexity to the problem.

C. X. Ma, MD, PhDM. J. Ellis, MB, BChir, PhD, FRCP

References1. Ma CX, Adjei AA, Salavaggione OE,

et al. Human aromatase: generesequencing and functionalgenomics. Cancer Res. 2005;65:11071-11082.

2. Ellis MJ, Miller WR, Tao Y, et al.Aromatase expression and outcomesin the P024 neoadjuvant endocrinetherapy trial. Breast Cancer ResTreat. 2009;116:371-378.

RADIATION THERAPY

Long-Term Results ofHypofractionated RadiationTherapy for Breast Cancer

Whelan TJ, Pignol J-P, Levine MN, et al(McMaster Univ, Hamilton, Ontario,Canada; Odette Cancer Centre, Toronto,Ontario, Canada; et al)

N Engl J Med 362:513-520, 2010

Background.—The optimal frac-tionation schedule for whole-breast

irradiation after breast-conservingsurgery is unknown.

Methods.—We conducted a study todetermine whether a hypofractionated3-week schedule of whole-breast irradi-ation is as effective as a 5-weekschedule. Women with invasive breastcancer who had undergone breast-conserving surgery and in whom resec-tion margins were clear and axillarylymph nodes were negative wererandomly assigned to receive whole-

breast irradiation either at a standarddose of 50.0 Gy in 25 fractions overa period of 35 days (the control group)or at a dose of 42.5 Gy in 16 fractionsover a period of 22 days (the hypofrac-tionated-radiation group).

Results.—The risk of local recur-rence at 10 years was 6.7% among the612 women assigned to standard irradia-tion as compared with 6.2% among the622 women assigned to the hypofrac-tionated regimen (absolute difference,

iseases: A Year Book� Quarterly 267Vol 21 No 3 2010

0.5 percentage points; 95% confidenceinterval [CI], �2.5 to 3.5). At 10 years,71.3% of women in the control groupas compared with 69.8% of the womenin the hypofractionated-radiation grouphad a good or excellent cosmeticoutcome (absolute difference, 1.5percentage points; 95% CI,�6.9 to 9.8).

Conclusions.—Ten years aftertreatment, accelerated, hypofraction-ated whole-breast irradiation was notinferior to standard radiation treat-ment in women who had undergonebreast-conserving surgery for invasivebreast cancer with clear surgicalmargins and negative axillarynodes. (ClinicalTrials.gov number,NCT00156052.)

This randomized trial of theOntario Clinical Oncology Group, theresults of which were recentlypublished in the New England Journalof Medicine, is one of the largestrandomized trials with long-termfollow-up (10 years) and demon-strated equivalent long-term resultswith a modestly hypofractionatedcourse of radiation therapy that canbe delivered in 3 weeks as opposed tothe standard 5-7 weeks of radiationtherapy for early-stage invasive breastcancer. This landmark article byWhelan and colleagues is potentiallypractice changing, given its relativelylarge size and excellent long-termfollow-up. The trial showed equivalentlocal control and cosmetic outcomesbetween the standard regimen and this

268 Breast Diseases: A Year Book� Quar

Vol 21 No 3 2010

hypofractionated regimen of 42.5 Gyin 16 fractions. While the results ofthis study clearly support theequivalence of this regimen tostandard fractionation in terms ofoutcomes for selected patients, thetrial raises several questionsregarding its potential clinicalimplications. Since the patients in thistrial had early-stage invasive cancersthat were all node-negative and weretreated without a boost to the tumorbed, and the majority did not receivechemotherapy, questions arise as towhether these results can be extrap-olated to patients who have receivedcurrent forms of chemotherapy, maybenefit from a boost to the tumor bed,have ductal carcinoma in situ asopposed to invasive cancer, and, mostimportantly, have high-grade tumors,which were shown in this trial to haveinferior local control.

Based on this study, given itsexcellent design, execution, andfollow-up, most clinicians wouldagree that this hypofractionatedregimen appears to be safe andeffective in patients who met theeligibility criteria for the trial, are notgoing to receive a boost, are notreceiving systemic chemotherapy, andwho do not have high-grade tumors.For other patients, who either werenot represented at all in this trial orwho composed a minority of patientsin the trial, these issues remainunsettled and clearly requireadditional data and follow-up. While

terly

some clinicians and patients mightsupport either using a boost in thepatients in this study population orusing this regimen in patients who hadreceived cytotoxic chemotherapy,others may not feel comfortable doingso. The issue of why patients withhigh-grade tumors had poorer localcontrol, and whether this would beovercome with the use of a boostand/or systemic therapy, remainsunresolved. These issues are discussedat length in an American Society ofTherapeutic Radiation Oncologyconsensus panel document that iscurrently in press.1 However, this timelyrandomized trial from Canada clearlyestablishes hypofractionated whole-breast irradiation as an acceptablestandard of care for selected patientswith node-negative breast cancer.Hopefully, future studies will addresssome of the unanswered questionsregarding the appropriateness ofhypofractionated radiation therapy ina broader group of patients.

B. G. Haffty, MD

Reference1. Smith BD, Bentzen S, Correa C, et al.

Fractionation for whole-breastirradiation: an American Society forRadiation Oncology (ASTRO)evidence-based guideline [publishedonline ahead of print July 15, 2010].Int J Radiat Oncol Biol Phys. doi:10.1016/j.ijrobp.2010.04.042.