Factors associated with negative margins of lumpectomy specimen: potential use in selecting patients for intraoperative radiotherapy.

Factors associated with negative margins of lumpectomy specimen: potential use in selecting patients for intraoperative radiotherapy

Schiller DE, Le LW, Cho BCJ et al.

Annals of Surgical Oncology, 2007, DOI: 10.1245/s10434-007-9711-2

This study is a retrospective analysis of a prospectively maintained database to define an algorithm to potentially select patients for intra-operative radiotherapy with partial breast irradiation by intra-operative radiotherapy or balloon catheters (MammoSite). Modern developments have rekindled an interest in predicting prior to surgery the patients who will have negative margins. With intra-operative radiotherapy, the treatment would already have been delivered before the margin status is known. If the MammoSite technique is used, the balloon catheter would usually have been sited intra-operatively. Subsequent identification of a close margin is not necessarily a catastrophe. With the intra-operative radiotherapy, either whole-breast irradiation can be delivered or the margins can be re-excised. With MammoSite technology, the catheter can be removed and the site re-excised. It is also possible with the MammoSite approach to have the catheter placed by ultrasound guidance after surgery. Nevertheless, an algorithm that reduces the need for margin re-excision rates would be of clinical value.

The study cohort consisted of 730 lumpectomies performed in 708 patients. The median age of patients was 56 (range, 25-93) years. Unlike in UK practice, the pre-operative diagnosis of malignancy was made in only 80% of lesions. Of all cancers, 72% were invasive ductal cancers, 6% invasive lobular cancers and 6% of mixed ductal and lobular histology; 17% of cases were ductal carcinoma in situ alone. Surgery was performed without wire localisation in 24% of patients and this study therefore is likely to have consisted of predominately screen-detected lesions. The overall positive margin rate was 17%. Multivariate analyses revealed older age (P=0.0006), smaller tumour size (P<0.0025), wire-guided surgery (P=0.003) and obtaining a pre-operative diagnosis (P<0.0001) to be more associated with a negative margin.

The authors concluded that patients older than 50 years with a pre-operative diagnosis (particularly with a core needle biopsy) and with an invasive cancer <3 cm (localised for excision, especially suitable by ultrasound wire) provided the optimum subgroup suitable for achieving negative margins at the first resectional operation. Within this training set, application of this algorithm to the 731 lumpectomies would have selected a group of 178 lumpectomies to achieve a negative margin rate of 97.8%. A prospective validation set will be required to evaluate this conclusion before these data can be considered applicable.

With increasing uptake of intra-operative radiotherapy and modern brachytherapy, low rates of re-excision surgery for incompletely excised margins need to be achieved in order to make the process clinically and cost effective. Potential pathological predictors previously suggested to be associated with margin positivity include the presence of vascular invasion, extensive intraductal component and node positivity; however, these parameters are difficult to incorporate into an algorithm to select patients preoperatively as they may not be known until the surgical resection specimen is obtained for formal pathology assessment.

This study is limited in not offering a validation data set to test the predictive parameters identified. While biomarkers and epigenetics may be useful for determining tumour behaviour and systemic therapy options, factors that determine surgical resection margins are at best simplistic. These factors have to be readily assessable on pre-operative imaging or on tissue cores. Developments in imaging technology may well provide the way forward rather than yet more studies that evaluate current conventional parameters.