For this study, 61 women treated with PLA from June 2012 to May 2015 (Table 1) met the primary end point as the intention-to-treat (ITT) cohort determining PLA efficacy. Post-ablation MRI imaging of one subject violated protocol, which left 60 patients for MRI pathology correlation.
All 61 enrolled subjects were treated with PLA. Of these 61 patients, 60 underwent PLA using ultrasound guidance, with 1 patient using stereotaxis. The mean laser time was 15.8 min (range, 14.5–36.5 min). The mean energy expenditure was 4560 J (range, 864–11,231 J). The mean saline infusion was 8.7 ml (range, 2.5–21.0 ml), and the mean local anesthesia injection was 12.3 ml (range, 2–50 ml). The total procedure time averaged less than 1 h. All 61 patients underwent surgical excision of the ablated areas, with 55 patients undergoing lumpectomy and 6 patients undergoing mastectomy.
No serious adverse events occurred. Eight adverse events considered mild by the investigators included lump (1 patient), blister (2 patients), hematoma (1 patient), erythema (1 patient), and fat necrosis (3 patients). Six adverse events considered moderate by the investigators included pain (4 patients), lump (1 patient), and seroma (1 patient). All adverse events resolved.
A total of 19 protocol deviations were documented in 17 of the 61 PLA cases. Documentation for 12 procedural protocol deviations showed that target temperatures were not reached, whereas documentation for 4 procedural protocol deviations showed tumor mistargeting. Three protocol deviations documented screening failures as PLA was performed on tumors larger than protocol specifications. All protocol guidelines were fulfilled in the PLA procedures performed for 44 of the 61 subjects. Of these 44 patients, 40 (91%) had complete pathologic ablation.
A steep first-case learning curve was encountered at each site. Only two of the first cases at the nine sites achieved technical successes. After these 9 first cases, 42 (81%) of the remaining 52 ITT cases were deemed technical successes, with 5 (9.6%) of these 52 cases found to have post-PLA macroscopic residual tumor.
The patients reported qualitative satisfaction with the procedure. The average maximum pain reported by the 61 patients during treatment was 4.2 ± 2.9 (range 0–10). Return to activities of daily living averaged less than 8 h and was immediate for 45 (78%) of 58 subjects. At the 28-day follow-up evaluation, 97% of the 58 subjects who responded to cosmesis questions rated their cosmetic satisfaction as excellent (64%) or good (33%). One subject reported a significant scar at the 28-day follow-up evaluation.
Health-related quality-of-life assessments using the EORTC QLQ-BR23 and QLQ-C30 surveys were completed by all 61 subjects during their baseline visit and at their 28-day follow-up evaluation. Compared with the EORTC reference mean for early-stage breast cancer treatment, a positive meaningful change of at least 5 points was reported by the 61 subjects in this trial 28 days after ablation for all functional scales. This indicated an improved outlook and function after PLA treatment. The scores reflecting severity of treatment-associated side effects reported in this trial were less than published post-lumpectomy EORTC data or the EORTC reference mean.9 In particular, scores were lower than the reference mean for published lumpectomy data9 for fatigue (− 17.49), pain (− 13.40), and insomnia (− 15.41). This indicated that PLA treatment-associated side effects were less severe.
All 61 ITT cases demonstrated characteristic post-ablation gross pathology changes. A series of concentric rings surrounded the central area of char, which housed the laser tip. This was immediately surrounded by an area of gray coagulation followed by a ring of hyperemia. For evaluation of post-ablation changes, CK8/18 and ER microscopic stains were superior to H&E and Ki67 in documenting post-ablation cell viability. The CK8/18 stain demonstrated post-ablation tumor destruction by staining either totally negative or by having a granular, interrupted membranous staining pattern. In addition, CK8/18 staining changed from a cellular to a “stringy” pattern, which documented cellular disruption.
Of the 61 cancers in this study, 58 (95%) were ER-positive. Lack of post-ablation ER staining in these specimens confirmed non-viability of residual tumor.
Pathologic complete primary tumor ablation was achieved in 51 (84%) of the 61 ITT cases. The volume of the primary tumor was reduced 96%. Three procedures had less than 0.6% cancer cells in any cm2 section of the target tumors. These cells, which represented a minute amount of microscopic residual tumor, were scattered, noncontiguous, and too small to be seen by post-ablation MRI.
Residual tumor was present at the index targeted tumor site in 10 (16%) of the 61 ITT lesions. The overall success rate for ablation of the index tumors was 84%. One patient was mistargeted, which represented a 100% post-PLA residual volume. Residual tumor averaged 12.1% (range, 0.3–40%) in the remaining nine ITT subjects. Two additional subjects had residual tumor represented as previously undiagnosed secondary tumors, which denoted multifocal disease.
Post-Ablation Imaging and Pathology Correlation
Radiology-pathology correlations were performed for the 60 cases in which MRI images were performed per protocol. These were categorized into four groups. “True-positive” described a group in which both imaging and post-excision pathology showed macroscopic residual tumor, whereas “true-negative” described a group in which neither imaging nor post-excision pathology showed macroscopic residual tumor.
“False-negative” constituted a group in which imaging showed no macroscopic residual tumor whereas post-excision pathology demonstrated this. “False-positive” represented a group in which imaging showed macroscopic residual tumor whereas post-excision pathology did not. The results were summarized for the 60 patients with images performed per protocol (Table 2). The MRI-negative predictive value (NPV), specificity, sensitivity, and accuracy were superior to mammography and ultrasound in this study.
Correlations with pathology were performed for the 60 patients whose MRI images were obtained per protocol (Table 3). The NPV of MRI for these 60 patients was 92.2% (95% confidence interval [CI], 71.9–91.9%). The NPV of MRI for the 47 patients with cancers 15 mm or smaller was 97.7% (95% CI, 86.2–99.9%).
Patient Follow-Up Evaluation
Two recurrences were reported during a mean follow-up period of 43 months (range, 34–65 months). Both patients had complete ablations documented by post-ablation MRI and pathology. One patient, with an 8.5-mm ER-positive tumor, declined any additional treatment after PLA and excision. At 3 years, a recurrence was documented 25 mm from the PLA site. A second patient, with an 11.5-mm triple-negative tumor, was treated with chemotherapy plus whole-breast radiation after PLA and resection. A skin recurrence was noted in the lumpectomy incision at the 3-year follow-up assessment. No patient experienced PLA-associated sequelae.