An Argument Against Routine Use of Radiotherapy for Ductal Carcinoma In Situ

Why does the debate over theappropriate treatment ofductal carcinoma in situ(DCIS) continue? Three widely publicizedmulti-institutional randomizedtrials have addressed this question,[1-4]and all have reached largely the sameconclusion. Radiation therapy reducesthe risk of local recurrence of DCISby approximately 50%. Despite thisfact, a significant percentage of DCISpatients (50% or more in many settings)in consultation with their cliniciansopt to undergo excision aloneand forgo radiotherapy.The paper by Silverstein summarizesthe argument for this less aggressiveapproach in a clear and straightforwardmanner. The major rationalefor this position is that several studiesfrom single institutions have clearlyshown that the clinical behavior ofDCIS is not the same in all cases andthat, in many situations, the expected benefit from radiotherapy is quite low.In fact, one of the more remarkableaspects of DCIS is its profound heterogeneity,[5] seen in its pathology,mammographic appearance, and clinicalmanifestations.Margin Status
The contention of Dr. Silversteinthat many cases of DCIS do not benefitfrom radiation therapy after excisionis undoubtedly true and well supportedby a large body of evidence.First, the factors involved in localrecurrence after excision of DCIS arewell known and accepted. These includemargin status, grade and pattern ofspread, and overall extent of the individualDCIS lesion. Of these, marginstatus is certainly the most important,[6]but it is also clear from abundant experiencethat larger, higher-grade lesionswith more irregular margins may havespread in ducts that go unnoticed unlessexcision margins are extended toapproximately 1 cm. This means thatthe lower-grade, smaller lesions thatoften tend to be remarkably rounded andconsist of clustered lobular units andducts may require less effort at margindetermination for excision because thespread of these tumors is more confinedand easily detected.Case Definition
Dr. Silverstein raises an importantpoint that has not been addressed specificallyin many studies but is borneout by excellent supportive evidence.His point is that larger, high-gradeDCIS lesions must be approachedquite differently from the smaller varieties.[3] Thus, the major misleadingmythology in the treatment of DCISis that these lesions are all the same.Indeed, it is case definition and its precisionthat has led some therapeuticstudies somewhat astray.For example, the final summary ofthe important National Surgical AdjuvantBreast and Bowel Project (NSABP)B-17 study used mammographic measuresto indicate the size of the lesions.[7] At least 30% of the cases werenot centrally reviewed in this area ofdifficult histopathology. Also in thisstudy, the number of local recurrencesamong cases that had not been irradiatedwas extremely large-more than10% at 5 years. This is to be understoodagainst the backdrop of studies withcareful case definition and a small numberof cases, particularly of low-gradelesions at 5 years without radiationtherapy. Thus, it is clear that NSABPB-17 is a study of women who had morethan a 10% local recurrence rate withor without radiation therapy, and thatexperience cannot be compared withstudies of carefully defined, smaller lesionsin which virtually no local recurrencesoccurred at 5 and even 10 yearsafter adequate determination of marginstatus at surgical excision without radiationtherapy.We seem to be discussing two kindsof studies: those with careful case definitionand individual case follow-up,and those with poor case definition andevaluation of overall therapeutic efficacy.Careful case definition wouldseem to be more important for largestudies that are evaluating therapy, inorder to identify patients who do notneed treatment. This would seem preferableto showing that the treatment iseffective in a certain percentage of casesand then averaging all the cases, includingmany that did not need the therapy.The design of the other importantlarge, multicenter trial performed inEurope had similar drawbacks, and thistrial found some recurrences of highgradeinvasive cancers-even somecases with lymph node metastases. Mostimportantly, the majority, if not all, ofthese life-threatening local recurrencesdeveloped among patients who initiallyhad high-grade DCIS.[3]At this time, we must recognize thatcase definition and the precise histologiccharacter of excised lesions mustbe documented.[8] It is also clear fromseveral studies that low-grade lesionsmay have smaller margins than 1 cm,but the proof of principle for Dr.Silverstein, Dr. Lagios, and their colleagues-that any DCIS excised toa 1-cm margin should have no untowardconsequences without radiationtherapy-represented a banner momentin this controversy.Conclusions
In summary, thousands of womenhave been treated for DCIS with surgicalexcision to negative margins andcareful, continued mammographic follow-up. The rates of local recurrence aresmall (often 0% at 5 to 8 years), particularlyamong patients with low-gradeand smaller lesions (less than 1 to 1.5 cmin size). Also, the rare local recurrencesare regularly similar in grade, unlessresidual disease remains untended fora prolonged number of years.[9] Manycenters in North America and Europeare treating one-half or more of theirDCIS patients in this conservativemanner. The nature and rate of recurrenceselsewhere in the treated breast(not in the original segment or quadrant)or in the contralateral breast areless well understood, are analogous todisease in the contralateral breast aftera mastectomy, and should be regardedas separate events.In stark contrast, the cases of excisedDCIS not treated further by radiation inthe large NSABP and European Organizationfor Research and Treatment ofCancer (EORTC 10853) trials have hadlocal recurrence rates of 10% or more.Although the local recurrences in patientswith low-grade lesions probablypose little threat to life, it is also clearthat, even with radiation, recurrences ofhigh-grade lesions may be associatedwith invasive cancer of a life-threateningnature. Clearly, precise case definitionincluding evaluation of marginsshould be mandatory in individuallytreated women (as well as in any furthertrials conducted in this disease). Itis also important to detail the size, grade,and invasive nature of local recurrencesrather than to simply regard a "localrecurrence" as an event not mandatingfurther definition.

Disclosures:

The author(s) have no significant financial interest or other relationship with the manufacturers of any products or providers of any service mentioned in this article.

References:

1.

Fisher ER, Dignam J, Tan-Chiu E, et al:Pathologic findings from the National SurgicalAdjuvant Breast Project (NSABP) eightyearupdate of Protocol B-17: Intraductal carcinoma.Cancer 86:429-438, 1999.

2.

Bijker N, Peterse JL, Fentiman IS, et al:Effects of patient selection on the applicabilityof results from a randomised clinical trial(EORTC 10853) investigating breast-conservingtherapy for DCIS. Br J Cancer 87:615-620,2002.