Magnetic resonance-guided focused ultrasound treatment of extra-abdominal desmoid tumors: a retrospective multicenter study

Abstract

Objectives: To assess the feasibility, safety and preliminary efficacy of magnetic resonance-guided focused ultrasound (MRgFUS) for the treatment of extra-abdominal desmoid tumours.

Methods: Fifteen patients with desmoid fibromatosis (six males, nine females; age range, 7-66 years) were treated with MRgFUS, with seven patients requiring multiple treatments (25 total treatments). Changes in viable and total tumour volumes were measured after treatment. Efficacy was evaluated using an exact one-sided Wilcoxon test to determine if the median reduction in viable tumour measured immediately after initial treatment exceeded a threshold of 50 % of the targeted volume. Median decrease after treatment of at least two points in numerical rating scale (NRS) worst and average pain scores was tested with an exact one-sided Wilcoxon test. Adverse events were recorded.

Results: After initial MRgFUS treatment, median viable targeted tumour volume decreased 63 %, significantly beyond our efficacy threshold (P = 0.0013). Median viable total tumour volume decreased (105 mL [interquartile range {IQR}, 217 mL] to 54 mL [IQR, 92 mL]) and pain improved (worst scores, 7.5 ± 1.9 vs 2.7 ± 2.6, P = 0.027; average scores, 6 ± 2.3 vs 1.3 ± 2, P = 0.021). Skin burn was the most common complication.

Conclusions: MRgFUS significantly and durably reduced viable tumour volume and pain in this series of 15 patients with extra-abdominal desmoid fibromatosis.

Key points: • Retrospective four-centre study shows MRgFUS safely and effectively treats extra-abdominal desmoid tumours • This non-invasive procedure can eradicate viable tumour in some cases • Alternatively, MRgFUS can provide durable control of tumour growth through repeated treatments • Compared to surgery or radiation, MRgFUS has relatively mild side effects.

Keywords: Desmoid fibromatosis; High-intensity focused ultrasound (HIFU); MR-guided focused ultrasound; MRgFUS; Thermal ablation.

Figures

Figure 1

MR guided focused ultrasound treatment of an enlarging, painful desmoid tumour located in the posterior right ankle of a 66 year old man. He refused surgery, radiation and chemotherapy. In A, a sagittal post-contrast T1 weighted 2D FSPGR fat saturated image obtained 2 weeks before initial treatment shows an enhancing mass (white arrow) located between the Achilles tendon (+) and the flexor hallucis longus tendon (*). In B, an axial MR thermometry magnitude image acquired during sonication shows a representative sonication (green rectangle) and beam path (blue hourglass). The location of the tibial nerve and posterior tibial vessels has been demarcated (purple line); the system does not allow the beam path to cross this contour. The orange crosshair represents the location where temperature is being examined; this crosshair can be adjusted during sonication. Temperature is displayed in real-time during treatment (inset, red curve represents single hottest pixel, and green curve represents average of a 3 × 3 pixel area around the crosshair). In C, an axial post-contrast T1 weighted 3D SPGR image with fat suppression acquired immediately after treatment reveals the ablated area of the tumour (white arrow). Out of 42mL total tumour volume, 26mL (62%) no longer enhanced after ablation. The red arrow points to preserved enhancement between the posterior tibial vessels, which remain patent, and the flexor hallucis longus tendon (*). This area was not treated in order to avoid damaging the tibial nerve. In D, a sagittal post-contrast T1 weighted 2D FSPGR fat saturated image obtained 20 months after initial treatment and 11 months after a second treatment shows residual tumour, which is necrotic centrally, with peripheral enhancement. Total tumour volume declined to 29mL (31% decline) and the perfused viable volume was 17mL (60% decline). The patient's tumour-related pain resolved after MRgFUS treatment. Please see Supplemental Digital Content for additional images of this and other treatments.

Figure 2

Decline in tumour volumes after ablation. Total (A) and viable (B) tumour volumes are plotted relative to time. The vertical dashed line in each graph indicates the initial treatment, and black arrows indicate subsequent MRgFUS treatments. In A, after MRgFUS treatment, the total tumour volume stabilized or declined for varying amounts of time; when tumour regrowth was seen, another treatment was performed. In B, the red curves or data points show the enhancing tumour volume prior to treatment, and the blue lines are enhancing volumes after treatment. Figure 1 demonstrates results from patient #4. In patients #5 and #11, planned partial treatments were combined into single volumes for representation. In patient #3 tumour volumes declined with chemotherapy, but he did not tolerate the side effects. Prior to MRgFUS, chemotherapy was discontinued, and MRgFUS alone was subsequently used to control tumour growth. Asterisks indicate initiation of additional treatment modalities after MRgFUS. Patient #5 had cryoablation of residual viable tumour. Approximately 67% of the tumour was treated with MRgFUS, but portions close to the skin and the anterior abdominal wall and stomach were deemed safer to treat with cryoablation after hydrodissection. In patient #9, the inferior portion of the tumour encased the Achilles tendon, and the patient opted not to ablate near the tendon; chemotherapy was initiated 3 months after MRgFUS to treat residual, untreated tumour. Patient #11 had residual tumour treated with sorafenib beginning 8 months after MRgFUS. Patient #13 had two separate tumours treated. Note that post-contrast imaging was not obtained in patient #14 immediately after treatment.