Phyllodes tumours: Wider surgical margins unlikely to reduce local recurrence

A wider surgical margin width is not associated with a reduced risk of local recurrence (LR) of Phyllodes tumours (PTs), a study in patients treated in 2007–2017 has shown.

The study, based on review of data from 550 patients who underwent definitive surgical management for an initial PT at 11 institutions in the US, also showed that many patients were managed outside of current National Comprehensive Cancer Network (NCCN) guidelines, which recommend wide local excision with the intent of obtaining a tumour-free margin of ≥1 cm for PTs of all histologic grades. [J Clin Oncol 2020, doi: 10.1200/JCO.20.02647]

The cohort included 379 patients (68.9 percent) with benign, 108 patients (19.6 percent) with borderline, and 58 patients (10.5 percent) with malignant PTs. The median tumour size was 3 cm.

About half (54.9 percent) of the patients were initially managed with excisional biopsy with no attempt at margins, while 38.2 percent underwent wide local excision with attention to margins and 6.2 percent underwent mastectomy.

“The initial surgical margin was positive in 42 percent of the patients, despite a suggestive core needle biopsy in 63 percent and a clinical suspicion of PT by the surgeon in approximately 70 percent of the patients,” the researchers noted.

More than a third (37.6 percent) of patients underwent a second operation, including 51 patients with an initial negative margin who underwent surgery to obtain a wider negative margin (82.4 percent with an initial margin <2 mm).

“Nearly 40 percent of women with PTs were subjected to a second operation, without clear benefit,” the researchers pointed out.

However, 32 percent of patients with an initial positive margin did not undergo re-excision.

Recurrence occurred in 3.3 percent of the patients (n=18), with the majority (n=15) being LRs (LR by grade: benign, n=5; borderline, n=6; malignant, n=4). LRs occurred at a median of 18.8 months, with histologic upgrade occurring in 20 percent of the cases (from benign to borderline, n=2; from borderline to malignant, n=1).

LR was not associated with final margin status (positive vs negative: odds ratio [OR], 0.96; 95 percent confidence interval [CI], 0.26 to 3.52; p=0.96) or a wider negative margin width (≥2 mm vs <2 mm: OR, 0.39; 95 percent CI, 0.07 to 2.10; p=0.27), or with the type of surgery, mitoses, histologic tumour border, or age at diagnosis.

Instead, LR was associated with malignant grade (OR, 3.19; 95 percent CI, 1.63 to 6.21; p=0.001), moderate or marked stromal atypia (OR, 8.58; 95 percent CI, 2.39 to 30.79; p=0.001), stromal overgrowth (OR, 3.78; 95 percent CI, 1.32 to 10.79; p=0.01), and increased pathologic tumour size (OR, 1.01; 95 percent CI, 1.00 to 1.01; p<0.001).

“The NCCN guidelines are reflective of broad principles based on lower-level evidence. With no prospective data in the literature, and conflicting large retrospective series on both ends of the margin spectrum, what constitutes an adequate margin for PTs remains uncertain,” the researchers wrote.

“While we endorse current margin recommendations for borderline and malignant PTs, guideline revision for benign PTs should be considered,” they continued. “Based on the totality of available data, we do not recommend re-excision of a negative margin for benign PTs, regardless of negative margin width, as progressively wider surgical margins do not reduce local or distant recurrences.”