Commentary (Abrams): Lumpectomy With and Without Radiation for Early-Stage Breast Cancer and DCIS

To irradiate or not to irradiate, that is the question posed by Marks and Prosnitz for women with early invasive breast cancer or ductal carcinoma in situ (DCIS) undergoing breast conservation therapy (BCT). Due in large part to mammography, there has been in the 1990s a significant increase in the percentage of women presenting with stage 0 (DCIS) and stage I breast cancer, as recorded by the Surveillance, Epidemiology and End-Results Program of the National Cancer Institute (NCI). The generally excellent outcome of these tumors with current therapy has focused research efforts on studies directed at improving quality of life and minimizing the side-effects and expense of local therapy. In this regard, Marks/Prosnitz present a concise and balanced summary of results from randomized and non-randomized clinical trials, and develop a rationale for current treatment recommendations and future studies regarding the role of breast irradiation (RT).

Invasive Cancers

For patients with early stage invasive cancer, Marks/Prosnitz provide a tabular review of the five, pivotal randomized trials testing local excision with and without RT that demonstrate that the average annual breast recurrence rates have increased fourfold (from about 1% to 4% when breast RT was omitted). Subset analyses of these trials have been unable to find prognostic factors that discriminate adequately to avoid the use of RT. For example, in the node-negative Ontario trial cited in the review, examination of women over 50 years old with tumors smaller than 2 cm revealed a local recurrence rate of 19% without RT compared to 8% with RT. Similarly, the rates for those with tumors smaller than 1 cm were 28% and 10%, respectively.[1] However, in the Milan trial, the 3 year local recurrence rate after quadrantectomy was 17.5% in patients aged 55 or younger, but only 3.8% in those over 55.[1]

Although no survival differences have as yet appeared in these trials, Marks/Prosnitz demonstrate correctly that the available studies are not adequately sized to eliminate the possibility of small, yet significant, survival differences. Even if survival is not ultimately affected, I agree with their conclusion that local recurrence rates are unacceptably high without breast irradiation. In view of the devastating psychological impact of a local recurrence, breast irradiation should remain an integral component of breast conservation therapy outside of a controlled clinical trial. If this argument is accepted, then the availability and feasibility of irradiation must be considered by physicians when offering women a choice between breast conservation therapy and mastectomy.[2]

An alternative approach to eliminating breast irradiation from breast conservation therapy is currently being tested by the NSABP (B-21) and the CALGB (9343). Both trials test whether tamoxifen (Nolvadex), with its proven ability to decrease local recurrences,[3] can substitute for irradiation in low-risk populations. It is noteworthy that the B-21 trial, an NCI high priority study open for about 6 years, has only accrued half of its targeted 1690 patients. It may be that a portion of the NSABP membership find the benefits of irradiation demonstrated in previous trials in early stage invasive cancer (B-06) and in ductal carcinoma in situ (B-17) so compelling that they are unwilling to randomize patients.

Fortunately however, other groups, namely the North Central Cancer Treatment Group (NCCTG) and the Australian-New Zealand Breast Cancer Trials Group find this question important enough that they have agreed to join this NSABP trial. The CALGB trial (9343) focuses on older women where randomized data is sorely lacking. Although not mentioned in the review, it is noteworthy that the European Organization for Research and Treatment of Cancer (EORTC) has also initiated a randomized trial comparing segmental surgery with and without breast RT. This trial (EORTC 10932) is limited to women 50 years of age or older with node-negative tumors that are grade 1, less than 2 cm and without evidence of vascular invasion or extensive DCIS.

Continued efforts to find a safe and acceptable alternative to breast irradiation as part of breast conservation therapy are justified. Despite the conclusion of the 1990 NIH Consensus Development Conference that BCT is the preferrable treatment option for women with early-stage invasive breast cancer, the rate of breast conservation therapy remains unacceptably low. The reasons for this are clearly multifactorial.[1] However, for many patients, young and old alike, the fear, cost and inconvenience of irradiation are important barriers to breast conservation therapy. If certain groups can be safely spared irradiation, then the use of breast conservation therapy is likely to increase and the quality of life of many women with breast cancer will improve substantially.

Ductal Carcinoma In Situ

As described by Marks/Prosnitz, the seminal NSABP (B-17) trial did resolve the paradox that developed in the 1980's whereby patients with invasive cancer had the option of lumpectomy and radiation therapy while the standard approach for ductal carcinoma in situ was mastectomy. B-17 demonstrated that both lumpectomy and radiation therapy and lumpectomy alone offered survival results that appear comparable to those of mastectomy for DCIS. Yet, this trial opened up a whole new series of questions.

The omission of radiation therapy resulted in an increase in the occurrence of ipsilateral breast tumors (IBTs), nearly 50% of which were invasive. As shown in Table 2 of the Mark/Prosnitz article, this finding has been confirmed in numerous, non-randomized trials. However, some investigators argue that the rate of ipsilateral breast tumors in B-17 can be markedly reduced by careful patient selection. They have demonstrated in non-randomized, single institution studies that the use of histologic features, tumor grading, and postoperative mammography can select patients with lower rates of ipsilateral breast tumors for whom lumpectomy alone appears sufficient. By contrast, the NSABP investigators could not detect a subset that did not benefit from RT, although the absolute benefit was small in patients with negative resection margins and without evidence of comedo necrosis.

As Marks/Prosnitz indicate, these results have not been uniformly accepted. Those advocating lumpectomy alone for selected subsets suggest that the discrepant results arise from differences in the assessment of surgical post-excision margins and histopathological interpretations, as well as the lack of postoperative magnification view mammograms in the B-17 trial. In view of the paucity of multi-institutional, prospective trials in women with ductal carcinoma in situ, the current debate will only be resolved through additional research.

Fortunately, additional data will be soon forthcoming from several large, prospective, multi-center national and international trials testing the validity of omitting breast radiation. Two European trials nearing completion include an EORTC study of 1,010 patients which completed accrual in 1996 and a Cancer Research Campaign (CRC) study in the United Kingdom that is ongoing and will accrue 1,000 patients.

To specifically target the most frequent presentation of this disease in patients today, the National Cancer Institute (NCI) is supporting two new trials in patients presenting with DCIS as detected by mammography. The first of these is a single arm, non-randomized trial performed by the Eastern Cooperative Oncology Group (ECOG) in which patients undergo lumpectomy and are then observed. This trial opened to accrual in April 1997. To be eligible, patients must have high-grade tumors less than or equal to 1 cm or low/intermediate grade tumors less than or equal to 2.5 cm. Importantly, these features must be determined prospectively by the local pathologist based on clear guidelines described in the protocol, and not by a central referee pathologist. In view of the lack of uniformity in the histopathological characterization, grading and margin assessment of DCIS, the prospective determination of eligibility was deemed critical if the results of this trial were to be widely generalized. This trial has early stopping rules in the event that the occurrence of ipsilateral breast tumors exceed an overall 5-year failure rate of 10% or an invasive cancer 5-year rate of 5%.

The second trial, being developed by the Radiation Therapy Oncology Group (RTOG), should open within the next 6 months. This is a randomized, controlled trial for patients with tumors less than 2.5 cm with no comedo elements or high nuclear grade. Patients will receive either lumpectomy with irradiation or lumpectomy alone. Administration of tamoxifen is optional but must be declared before randomization so that patients can be stratified accordingly.

Both the ECOG and RTOG trials plan to accrue approximately 1,000 patients so that there will be ample statistical power to provide tight error bars around the ipsilateral breast tumor rate in ECOG's case or to detect small differences in the RTOG study. The need for two distinct trials resulted from prevailing biases in the US medical community. The ECOG investigators found it unacceptable to irradiate every patient when only a small percentage benefit while the RTOG investigators believed that only a randomized trial would control for the subtle selection biases that could undermine the applicability of the results to the community at large. In addition to determining if RT can be safely omitted, these trials will provide an invaluable tumor tissue resource as both will bank tissue for future studies of molecular markers. The studies on this tissue should make important contributions to a better understanding of the biologic mechanisms that underlie DCIS, which hopefully will permit refinement of prognostication and treatment for individual patients.

References:

1. Abrams J, Phillips PH, Friedman MA:
Meeting highlights: A reappraisal of research results for the local treatment of early stage
breast cancer. J Natl Cancer Inst 87:1837-1845, 1995.

2. Farrow DC, Hunt WC, Samet JM: Geographic variation in the treatment of localized breast cancer. N Engl J Med 326:1097-2001, 1992.

3. Fisher B, Costantino J, Redmond C, et al: A randomized clinical trial evaluating tamoxifen in the treatment of patients with node-negative breast cancer who have estrogen-receptor-positive tumors. N Engl J Med 320: 479-484, 1989.