Ductal Carcinoma In Situ: When is Conservative Surgery Alone Sufficient?

abstracts & commentary

Synopsis: Ductal carcinoma in situ is highly curable by mastectomy. More conservative surgical approaches have become commonly employed, but local recurrences are considerably more frequent (approximately 20%). The addition of radiation therapy effectively reduces local recurrence by about 50%. Yet, it is still unclear which patients are likely to benefit from more radical surgery or additional radiation. Two recent reports have addressed this issue. On the basis of these studies, patients best suited for conservative therapy alone (outside of a clinical trial) are those with low grade, small lesions with no necrosis and clear surgical margins.

Sources: Hetelekidis S, et al. Cancer1999;85:427-431; Boyages J, et al. Cancer 1999;85: 616-628.

The management of ductal carcinoma in situ (DCIS) involves initial treatment decisions that are difficult. The difficulty emerges because mastectomy is nearly 100% curative1 but is more aggressive treatment than is given to patients with invasive cancer and is perhaps more therapy than most patients need. Indeed, conservative surgical approaches have provided excellent results with low local recurrence rates and more satisfactory cosmetic and psychological outcomes.2,3 When radiation therapy is added to a conservative surgical procedure, local recurrence rates approximate those observed after mastectomy in some series.

Recently, two papers published in Cancer provide useful data for clinicians advising patients in this decision-making process. In the first report, Hetelekidis and collegues from Harvard examined outcomes in 59 patients with DCIS who presented from 1985 to 1990 and were treated with excision alone. Study pathologists examined histologic slides and all had negative margins at the time of initial review. The median age at diagnosis was 54 years and the median follow-up time was 95.5 months. All but two presented with mammographic findings only and all of the patients who recurred had reexcision.

One factor considered important in predicting the chance for local recurrence has been the lesion size.4,5 However, size has been difficult to assess because DCIS is usually not grossly apparent. Thus, histologic specimens have been used to assess size, and these methods have proven imprecise.6 In this series, a novel approach was developed. The number of low power fields (40X) in which DCIS was identified were summed and used as a surrogate measure of overall size.

Similarly, margin status has proven to be a difficult assessment, primarily because DCIS lesions can have gaps of several millimeters separating them, suggesting that margins of just a few millimeters may be inadequate to assure total excision. In this series, patients with involved margins were not included. Those with margins of 1 mm or less were considered "close," whereas those with margins greater than1mm were considered "negative."

Although local recurrence developed in 10 of the 59 patients, no patients developed metastatic disease or died of breast cancer. The actuarial five-year local recurrence rate with a conservative surgical procedure alone was 10%. Four of the 10 recurrences were invasive carcinomas and six were DCIS. All recurrences occurred in close proximity to the originally resected lesion.

In this series, the single most important factor that predicted the development of local recurrence was lesion size. Those with low power field (LPF) scores of greater than 5 had 17% actuarial five-year local recurrence rates compared to 3% five-year local recurrence rates for those with smaller (<5 LPF) lesions. In fact, size was the only factor that reached a level of statistical significance in univariate analysis. Although patients with nuclear Grade 3 lesions had a higher local recurrence rate than those with nuclear Grade 1 and 2 lesions (18% vs 6% and 5% respectively), and patients with close margins (< 1mm) had a higher local recurrence rate than those with negative margins (> 1mm 25% vs 8%), these differences did not reach statistical significance. Yet, Hetelekidis et al point out, of the 19 cases with margins of less than 1 mm, lesion size of less than 5LPF and nuclear grade of 1 or 2, there were no local recurrences; by contrast, the remaining 40 patients had a five-year actuarial local recurrence rate of 15%.

Thus, in this single series, it appears that lesion size, nuclear grade,and margin status were prognostic factors that would help predict local recurrence and, thereby, be useful in advising patients regarding the use of adjunctive radiation therapy or mastectomy. Interestingly, in this series, factors such as patient age, family history, predominant architectural pattern (comedo type or other), or the presence of necrosis did not offer additional predictive value.

In the second report by Boyages and collegues, a meta-analysis of the patient, tumor, and treatment factors relevant to local recurrence after DCIS resection was undertaken. Boyages et al included the effects of such factors as patient age, presence or absence of family history, DCIS subtype, presence or absence of necrosis, and margin status (among others) to calculate risks of local recurrence for each treatment approach (mastectomy or conservative surgery plus radiation). The overall local recurrence rate was 22.5% for studies examining conservative surgical procedures alone, 8.9% for conservative surgery plus radiation therapy, and 1.4% for studies involving mastectomy alone. These summary figures are not exactly comparable because of the greater likelihood that those with smaller lesions and clear margins would be treated by conservative surgical procedures. Nonetheless, the patients with risk factors of presence of necrosis, high grade cytological features, or comedo subtype were found to derive the greatest improvement when radiation therapy was added to a conservative surgical procedure.

Local recurrence among patients treated with conservative surgery alone was about 20% and radiation therapy would appear to reduce this risk by half. The differences in local recurrence rates between conservative surgery alone and conservative surgery plus radiation therapy were found to be most apparent for those patients with high grade tumors, DCIS with necrosis, or of the comedo subtype, and for those with close or positive surgical margins.


DCIS now accounts for about half of the malignancies detected by mammography. Although DCIS can present as a palpable tumor, or as a nipple discharge, or as Paget’s disease of the nipple, most cases are found on mammography. There are several histologic patterns that DCIS can take; they are mainly divided into comedo and noncomedo forms. The latter include cribriform, micropapillary, solid, and clinging histologic patterns. The comedo form is more likely to be comprised of cells with greater cellular atypia, a higher proliferative rate, and greater expression of oncogenes such as HER-2/neu and mutant p53. The comedo and noncomedo forms can often be distinguished on the basis of the mammographic appearance: comedo forms are associated with coarse granular microcalcifications and noncomedo types are more often associated with fine granular microcalcifications.

Initial decisions in the management of DCIS are complex and involve a range of individual patient preferences. For some, conservative surgery is favored and it has been proposed that this approach may be equivalent in outcome to mastectomy, especially if adjunctive radiation therapy is included. However, many questions remain and are the subject of ongoing, randomized clinical trials (such as NSABP B17).7 Until the completion of these trials, the clinical series by Hetelekidis et al and the meta-analysis by Boyages et al highlighted here provide useful insights.

The data from both reports would suggest that radiation therapy be added to conservative surgery in those patients with increased likelihood of local recurrence. Included among these patients who should receive radiation therapy would be those with larger lesions, although this is difficult to assess with quantitative certainty. The size of comedo lesions can be corroborated by amplified views of the mammogram in many cases; the mammagraphic appearance of the noncomedo forms is less reliable as a gauge of lesion size. Other factors that would favor the addition of radiation therapy include the absence of a reasonably wide surgical margin (optimally 10 mm), the presence of necrosis, comedo subtype, and, high cytological grade. Of course, for patients who seek the lowest risk for recurrence, the meta-analysis confirms that mastectomy reduces the risk of recurrence to close to 1%. As the Harvard series indicated, recurrences were local, and cures after recurrence remain likely.

Until more data is available, it would be inappropriate to be dogmatic in this decision-making process. Instead, it seems prudent to review those recurrence risk factors on a case-by-case basis and allow the patient to incorporate her own preferences for initial treatment. Women with small noncomedo lesions with an adequate margin around the resected tumor specimen have a low risk of recurrence with surgery alone. Most of the remaining patients may be cured with a combination of resection of the lesion and radiation therapy to the breast.


1. Schuh ME, et al. Arch Surg 1986;121:1303-1307.

2. Fisher B, et al. N Engl J Med 1993;328:1581-1586.

3. Delaney G, et al. Aust N Z J Surg 1997;67:157-165.

4. Cataliotti L, et al. Eur J Cancer 1992;28A:917-920.

5. Silverstein MJ, et al. Eur J Cancer 1992;28:630-634.

6. Lagios MD. Surg Clin North Am 1990:70:853-871.

7. Fisher B, et al. J Clin Oncol 1998;16:441-452.