Is Preoperative Staging with Breast MRI Causing Overtreatment with Mastectomies?

“The rate of cases upstaged by MRI as warranting larger resection/mastectomy … is substantially higher that breast recurrence expected after conservative treatment based on conventional staging.”

Preoperative staging with MRI has been proposed as an alternative to conventional staging as it upstages approximately 20% of cases eligible for breast conserving treatment, prompting larger resection/mastectomy. The rate of cases upstaged by MRI as warranting larger resection/mastectomy (~20%) is substantially higher that breast recurrence expected after conservative treatment based on conventional staging. This suggests that a substantial proportion of cancer foci are occult to conventional staging, and are thus left in site at conservative surgery: however they evidently are not bound to progress to breast recurrence, as they are probably controlled by adjuvant breast irradiation and chemo/ormonotherapy. Thus, it is likely that a substantial proportion of larger resections/mastectomies prompted by MRI staging correspond to overtreatment. As the only benefit of MRI staging is to anticipate mastectomy, avoiding delayed mastectomy for breast recurrence – indeed a psychological benefit with no prognostic implication – the cost of overtreatment by MRI prompting unnecessary mastectomies seems unjustified. The use of routine MRI preoperative staging is, therefore, not recommended.

The state of art

Approximately 35 years after the first randomized clinical trials on breast conservative treatment (BCT) were designed, BCT has become a standard of treatment for the majority of breast cancer (BC) cases given its prognostic equivalence with radical mastectomy and its better psychological and cosmetic impact.

Indeed, local control is suboptimal with BCT as compared to mastectomy, and breast recurrences are expected after BCT with a higher rate as compared with mastectomy. Nevertheless, such breast recurrences have no negative prognostic implication as delayed mastectomy will achieve local control and long-term survival will not be negatively affected.

Moreover, breast recurrences after BCT are not frequent. Based on more than 20–25 years of follow-up, breast recurrences occur at a rate of less than 1% per year (0.7% per year, on average, or up to 12–15% in a lifetime) and the large majority of women undergoing BCT will conserve their breast.

Based on all this long-term evidence, BCT is commonly considered a safe option and the frequency of mastectomy as a first surgical treatment has dramatically reduced in the last decades [1,2], an event that is commonly perceived as a net benefit: for example, the proportion of screen detected BC subjects undergoing BCT is measured as an indicator of quality according to European Community screening guidelines [3].

Looking for perfection

Preoperative staging to assess eligibility to BCT is conventionally based on palpation, mammography and ultrasonography. However, several studies based on careful surgical specimen analysis have shown that the real extension of BC may be underestimated at conventional staging: occult foci at the same or another quadrant may be missed in a relevant proportion of cases, ranging from 21 to 63% [4–7], a substantially higher rate as compared with observed breast recurrences.

“MRI prompts larger resection/mastectomy in approximately 20% of staged cases, while approximately 10% of breast recurrences are expected within 15 years.”

Delayed mastectomy due to breast recurrence is undoubtedly a psychologically negative event, often perceived by the woman as a generalized disease relapse associated with unfavorable prognosis (which is clearly not the case according to scientific evidence).

In order to improve the accuracy of preoperative staging, that is to identify cases underestimated at conventional staging as to their extension, MRI can be used as a staging modality for its high sensitivity, with the aim of achieving a better assessment of tumor volume/multifocality and to allow a more precise indication to BCT. In several experiences MRI has been shown to be more sensitive as compared with conventional staging [8–10], prompting larger resection/mastectomy in a substantial proportion of cases for which BCT had been planned according to conventional staging. It was assumed that these cases would have developed a breast recurrence if undergoing BCT. MRI staging would thus allow for avoiding the unpleasant and stressful experience of delayed mastectomy and has been proposed by many as the optimal preoperative staging procedure.

“…residual disease after BCT is controlled by routine breast irradiation and adjuvant hormone or chemotherapy…”

Looking for perfection has a cost

Although it is quite evident that MRI is more sensitive than conventional staging and allows for more precise definition of BC extension, it is still unclear whether its systematic use really translates into a benefit for the patient [9,10]. A recent meta-analysis of the literature suggests that magenetic resonance may cause overtreatment, as it may prompt immediate mastectomy in subjects otherwise eligible to BCT at a higher rate than that of expected breast recurrences [11]. MRI prompts larger resection/mastectomy in approximately 20% of staged cases, while approximately 10% of breast recurrences are expected within 15 years. In addition, the limited specificity of MRI may overestimate tumor extent as MRI positivity is not always confirmed at second-look ultrasound- or MRI-guided core biopsy; a false-positive MRI may cause unnecessary larger resection/mastectomy in approximately 5% of staged cases. Moreover, randomized studies of MRI in preoperative staging reported thus far still give controversial results [12,13].

Reconsidering the whole topic

The whole question probably needs to be addressed much more practically and more on established evidence than it has been thus far by considering some simple points:

• The advantage of MRI staging detecting more extensive BC is to prompt immediate larger resection/mastectomy in cases that otherwise would have undergone BCT with a high risk of residual disease and, possibly, of delayed breast recurrence and reintervention;

• Such an advantage may translate into psychological benefit (the sad experience of breast recurrence and of delayed mastectomy is avoided) but not into prognostic improvement, as breast recurrence is clearly not associated with worse prognosis;

• There is no evidence that all cases in which MRI prompts larger resection/mastectomy would have recurred in the breast if they had undergone BCT planned according to conventional staging, and that they would require delayed mastectomy. In fact, it is likely that a substantial proportion of residual disease after BCT is controlled by routine breast irradiation and adjuvant hormone or chemotherapy, which are commonly performed in these subjects. This might explain the discrepancy between the rate of larger resection/mastectomy generated by MRI and the rate of observed breast recurrences after conventional staging and BCT (~2:1 ratio). In other words, residual disease may be present and is likely to be detected by highly sensitive MRI, but is not bound to surface as a breast recurrence in a substantial proportion of cases;

• The issue of whether a woman prefers to run the risk of delayed mastectomy or the alternative of immediate mastectomy has not been explored. Considering that the advantage of the latter option is only psychological, it would be up to the woman to decide what she prefers as nobody but the woman can be a better judge of quality of life implications, after being correctly informed on the pros and cons of the two alternatives.

A philosophical controversy

There is a strong controversy around the risk of over treatment associated with MRI preoperative staging.

On one side, there are clinicians who are enthusiastic about the actual power of MRI to detect cancer foci in addition to conventional (palpation, mammography and ultrasonography) staging, and have a very strong resistance to accept the idea that detecting such foci may not be of worth. Radiologists often assume that detecting a disease is a positive event per se, and that there is no need to confirm that such a detection is associated with a real benefit to the patients. Such a pro detectione philosophy is quite common; leaving behind portions of undetected BC is regarded as a deadly sin, even if pushing for its detection will translate in a negative cost–benefit balance for the patient.

On the other side there are those who believe that medicine is aimed at being beneficial for the patient based on a final cost–benefit balance and tend to be less impressed by a good diagnosis unless it implies a real advantage for the patient. It took decades to limit extreme mutilation such as mastectomy to a minority of cases for which it was unavoidable. The idea that introducing a new preoperative staging protocol will finally generate an excess of mastectomies with no major benefit looks hardly acceptable. Recent contributions by pioneers of BCT such as Solin [14] and Fisher [15] comment negatively on MRI staging practice, claiming that no real benefit has been proven for such a procedure, but for a ‘resurgence’ of mastectomy.

What to do?

The risk of overtreatment subsequent to MRI staging practice has been acknowledged by the European Society of Mastology (EUSOMA), recommending specific criteria to select subjects who may benefit from staging, being at higher risk of conventional understaging and of multifocality [16]. The intention may be good, but the real effectiveness in reducing overtreatment is questionable. In fact, if these selection criteria are efficient in selecting those subjects bound to be upstaged by MRI, they will largely reduce the number of MRI staging procedures but will not substantially affect overtreatment; they will only concentrate in a smaller subset, those cases for which larger resection or mastectomy will be prompted by MRI.

It is quite difficult to say how this controversy will end. From a strict scientific point of view, considering that existing evidence is uncertain, there is no doubt that the question should be solved by another large randomized clinical trial, comparing conventional and nonconventional MRI staging on the basis of long-term follow-up, measuring survival, breast recurrences and mastectomies – immediate or delayed – over time. In the meantime, considering that substantial evidence exists suggesting that MRI staging may induce overtreatment and that the only benefit of MRI staging would be psychological, by replacing delayed with immediate larger resection/mastectomy the decision about what to do should be reasonably left to the woman, after accurate and impartial information is given on the possible implications of different choices.