Background: The I-SPY 2 TRIAL randomizes patients with breast cancer to neoadjuvant chemotherapy (NAC) with or without experimental agents followed by anthracycline-cyclophosphamide (AC). As part of a de-escalation strategy to avoid overtreatment in good responders, combined information from serial MRI and mid-treatment core biopsy will be used to identify patients who may be candidates to skip AC. Presented are results of a pilot study to assess the location of biopsy clip in relation to tumor enhancement in MRI before and during treatment.
Methods: 40 patients enrolled in I-SPY 2 who underwent mid-treatment core biopsy were reviewed. Patients had MRIs at four time points: pre-treatment (T0), early treatment (T1, 3 weeks after the start of first regimen), inter-regimen (T2, 12 weeks after completing the first regimen), and pre-surgery (T3). Clip visibility and location were assessed by a trained researcher on T1 weighted, fat suppressed dynamic contrast enhanced MR images at three time points: T0, T1, and T2. If clip was visible, location was scored 1 (inside), 2 (edge), or 3 (outside) in relation to tumor enhancement seen on a signal enhancement ratio (SER) map. Clips inside (score 1) were fully embedded and surrounded by a clear margin of tumor enhancement. Clips on the edge (score 2) were not fully embedded, with a portion of the clip touching tumor enhancement. Clips outside (score 3) had no portion of clip touching tumor enhancement. For patients with a focal tumor but with multiple clips visible in MRI, the clip most embedded within tumor enhancement was designated as the primary clip for evaluation. In cases of multifocal disease, the clip visualized in the largest area of tumor enhancement was assessed. Clips touching tumor cavity edge and enhancement were scored as 3.
Results: Among 40 patients, two patients had no clips visualized in MRI at all three time points. One patient had no clip visualized at T0, but a clip was observed at T1 and T2. In addition, one T1 MRI was rejected due to incomplete exam and one T2 MRI was rejected due to different scanner from baseline. At T0, 51% (19/37) of clips were inside tumor enhancement and no clips were assessed outside tumor enhancement. 59% (22/37) of clips at T1 were on the edge or outside. At T2, 73% (27/37) of clips were on the edge and 19% (7/37) of clips were outside. While no clips were identified outside of tumor at baseline, tumor shrinkage with treatment resulted in clips outside of tumor in approximately 20 percent of cases at inter-regimen, and higher rates of clips identified at tumor edge.
Conclusions: Clips were visible and location could be assessed in MRI in a majority of cases. Clip evaluation can be challenging and attention to clip placement is essential for patients with multifocal disease or diffuse tumors. The results of this study may have implications for assisting the mid-treatment core biopsy and selecting candidates for de-escalation of neoadjuvant chemotherapy.