Do Tumors Detected by Mammography Screening Have a Favorable Prognosis?

Considerable progress has been made in the last 20 years in improving outcomes for patients with breast cancer because of earlier detection and application of more effective therapies. The use of tamoxifen reduces the risk of recurrence and improves survival in patients with early stage hormone receptor–positive breast cancer.¹ Recent evidence suggests that adjuvant use of aromatase inhibitors may further add to the benefits seen with tamoxifen.^{2 - 4} Likewise, the use of adjuvant chemotherapy has been demonstrated to reduce recurrence and mortality in patients with early stage breast cancer, particularly in younger patients and in those with tumors that do not express hormone receptors.⁵ Nonetheless, many patients would remain free of cancer following local therapy (surgery with or without radiotherapy) with no adjuvant systemic hormonal or chemotherapy. Much research is currently focused on tailoring the use of adjuvant systemic therapies and on identifying patients with tumor characteristics associated with a favorable prognosis.

Currently, the most common prognostic factors used in making decisions regarding adjuvant systemic therapy include patient age, axillary lymph node involvement, and tumor size. In general, at least in the United States, most patients with involved axillary lymph nodes or with tumors larger than 2 cm are offered systemic therapy. Therefore, the dilemma regarding the use of adjuvant systemic therapies primarily involves patients with small, node-negative tumors (≤2 cm), who may have such favorable prognoses that they will not benefit from further therapy. Recommendations have been developed based on the prognostic markers and other factors, particularly hormone receptor status and grade of the tumor, to predict the risk of recurrence for patients with small tumors.^{6 - 7}

However, these methods are not always accurate and may lead to undertreatment of some patients and overtreatment of others. A recent study⁸ examined the use of gene-expression profiling of tumors as a method of predicting outcome and noted that a relatively large proportion of patients were misclassified when either St Gallen⁶ or National Institutes of Health criteria were used.⁷ Based on these observations, it is essential to determine more accurate ways of identifying patients who will not benefit from adjuvant systemic therapies. To date, the method of tumor detection has not been considered an independent prognostic factor, although it is clear that cancerous tumors detected by mammography screening are generally smaller and may have a less aggressive phenotype than those detected outside of screening programs.^{8 - 10}

In this issue of JAMA, Joensuu and colleagues¹¹ propose a new prognostic factor in early stage breast cancer. By using a relatively large database from Finland, the authors retrospectively compared the outcomes of patients with tumors diagnosed by mammography screening with those whose tumors were detected outside of screening programs (including tumors detected between screening rounds). As expected, the tumors detected during mammography screening had more favorable classic prognostic characteristics (smaller size, uninvolved axillary lymph nodes, lower grade) compared with tumors detected outside of screening. However, taking into account these classic prognostic factors, the method of detection was noted to be an independent predictor of both distant disease-free survival, which includes distant recurrences and death due to breast cancer, and of overall survival, with patients with tumors detected by mammography screening having a statistically better outcome than patients whose tumors were detected outside of screening. It is not clear why patients whose tumors were detected by mammography screening had a better outcome, and it would be interesting to examine these tumors molecularly to determine if the differences between the tumor detection modes can be delineated at a genetic level.

Tumor size is an important prognostic factor, particularly when axillary lymph nodes are not involved. Cancerous tumors detected by mammography screening tend to be small, many are 2 cm in diameter or smaller (T1 by tumor node metastasis classification). Therefore, a major strength of this study is that Joensuu et al divided patients into groups based on actual tumor size, not just by T classification. Looking specifically at node-negative tumors, patients aged 50 to 69 years (the largest age group) with tumors measuring 2 cm or smaller in the mammography screening group had a better distant disease-free survival at 10 years compared with the patients with tumors found outside of screening (93% vs 87% for tumors ≤1 cm and 91% vs 79% for tumors 1.1-2 cm).

Tumors detected by mammography screening resulted in a longer 10-year distant disease-free survival compared with tumors detected outside of screening, regardless of age at detection. The 10-year distant disease-free survival was 92% in patients aged 40 to 49 years with tumors detected by screening compared with 72% in the group with tumors detected outside of screening. Mammography screening in this age group remains controversial, but these data may provide another reason to consider screening these women.

How do the outcomes for patients with small, node-negative tumors in the current study compare with historical databases looking at patients with similar sized tumors? In general, it is extremely difficult to make comparisons across retrospective early breast cancer trials. For instance, the end points and length of follow-up may be quite different in different studies. In addition, data from other trials^{12 - 13} often use estrogen-receptor status and tumor grade to determine prognosis, but this data is not specifically given for tumors 2 cm or smaller by Joensuu et al.

The National Surgical Adjuvant Breast and Bowel Project recently performed an analysis of relevant trials looking at the outcome of patients with node-negative breast tumors measuring 1 cm or smaller.¹² The authors noted a relapse-free survival, which includes local and regional metastases as well as distant metastases, of 81% and 86% at 8 years for patients with tumors positive and negative for estrogen receptors, respectively.¹² Considering only distant recurrences, 10% of patients with estrogen receptor–negative tumors and 7% with estrogen receptor–positive tumors developed distant metastases by 8 years.¹² Ravdin et al¹³ described a computer model that uses classic prognostic factors to estimate an individual's risk of relapse (including local recurrences and second primaries as well as distant recurrences) and survival from early stage breast cancer. Using this computer model, a patient aged 60 years with a node-negative tumor measuring 1.1 to 2 cm would have an estimated 18% to 30% chance of relapse at 10 years depending on tumor status and grade.¹⁴ In the current study, among patients with tumors detected by mammography screening, 7% of patients who had tumors measuring 1 cm or smaller and 9% who had tumors measuring 1.1 to 2 cm would have developed distant disease or died at 10 years of follow-up. Therefore, it would appear that the 10-year outcome for patients with 2 cm or smaller tumors detected by mammography screening is at least as good, if not better, than what would be predicted by historical studies.

A common criticism of retrospective studies of early stage breast cancer is recognition of the increasingly difficult task of identifying pure prognostic factors because the majority of patients receive some form of adjuvant treatment. However, in this study, as the authors pointed out,¹¹ the use of adjuvant treatment cannot be a confounding factor because more patients with tumors detected outside of screening received adjuvant therapy. Because of the wide use of adjuvant therapies, current research seeks to identify predictive factors for specific therapies in an attempt to delineate which patients will benefit from these therapies. New techniques, for example multigene reverse transcriptase polymerase chain reaction assays,¹⁴ have recently provided some promising data on tailoring adjuvant therapies for early stage breast cancer. Until further data are available, clinicians and researchers must use available prognostic and predictive factors to determine the best adjuvant treatment, if any, for patients with breast cancer.

Joensuu et al¹¹ identified a group of patients with small, node-negative breast tumors that are detected by mammography screening and who have a risk of distant recurrence or death due to breast cancer of less than 10% at 10 years. In the United States, patients younger than 70 years with tumors measuring between 1 and 2 cm likely would be offered systemic therapy. In addition, some data suggest that adjuvant therapy can improve outcome in patients with even smaller tumors.¹² Therefore, these results, if confirmed, could help determine which patients would truly benefit from systemic therapy. Although newer screening techniques, such as breast magnetic resonance imaging,¹⁵ are being actively investigated, these data suggest that mammography screening may be more than a useful, acceptable screening tool—it may actually select for patients with a favorable prognosis.