Two cases of debulking surgery for lower limb diffuse plexiform neurofibroma with transcatheter arterial embolisation

Highlights • Debulking surgery was performed for DPN in two NF1 patients.• Preoperative TAE with gelatin particles to tumour feeder vessels was conducted.• TAE delayed wound healing; debridement and wound closure were required.• Motor function improvement was confirmed in both patients.


Introduction
Type 1 neurofibromatosis (NF1) is an autosomal dominant genetic disorder characterised by various fibrous lesions in systemic organs [1]. Diffuse plexiform neurofibroma (DPN) is a cutaneous lesion of NF1 that poses cosmetic and functional problems [2]. Critical intratumoural bleeding secondary to trauma [3] and malignant transformation [4] have been reported in DPN patients. Some surgeons recommend complete resection in an early stage [5]. However, massive haemorrhage can occur during surgery due to vascularity and vulnerability of the tumour and surrounding tissues [6].
When large feeding vessels are evident, preoperative transcatheter arterial embolisation (TAE), aiming the reduction of the blood loss during the surgical procedure, is effective. This was firstly reported in 1983 [7]. In this study, we report two cases of DPN resection managed with preoperative TAE, and discuss the pros and cons of TAE in terms of haemostasis and wound healing.
This case report has been reported in line with the SCARE criteria [8].

Presentation of cases
Case 1: A 20-year-old man with NF1 visited our clinic hoping to reduce the enlargement of his left lower limb. Examination confirmed a soft and drooping tumour encircling the thigh (Fig. 1A). DPN    (Astellas Pharma Inc., Tokyo, Japan) by interventional radiologists.
Under general anaesthesia, debulking surgery was performed. The skin of the lateral thigh turned purple, probably because of ischaemia following TAE. Enclosing the discolouration area, a 3-kg tumour was resected above the fascia level. The tumour's pedunculation and flexibility enabled us to temporarily ligate at the tumour's base (Fig. 1B). The total amount of bleeding was 500 mL; transfusion was not needed. Postoperative course was satisfactory, and the patient was discharged 17 days after surgery. One year after the surgery, he attends the hospital on foot (Fig. 1C).
Case 2: A 40-year-old man, diagnosed with NF1 in his childhood, was referred to our department with complaints of walking difficulty. His right lower limb had begun to enlarge since his 30 s. Eventually, he had become unable to walk by himself because of his huge and heavy leg ( Fig. 2A). We established that his symptom was caused by DPN in his leg, characterised by an elephant-like pigmented cutaneous tumour.
We planned serial excision with the resection area being from the buttock to the posterior thigh. Before the procedure, interventional radiologists performed TAE with Serescue TM to the feeding vessels originated from the superior and inferior gluteal arteries and the deep femoral artery of the affected side. An occlusion balloon was placed within the right internal iliac artery (IIA) to prevent uncontrollable critical haemorrhage.
Under general anaesthesia, we resected the tumour in the posterior thigh towards the head (Fig. 3A). In the buttock area, torrential bleeding from the wound occurred. The occlusion balloon in the IIA was inflated, and the bleeding was reduced temporarily. The total intraoperative blood loss was 4970 mL. We transfused 1960 mL of concentrated red cells and 1680 mL of fresh frozen plasma. The resected tumour weighed 5 kg with confirmed porous particles occluded in the arteries (Fig. 3C).
To prevent further bleeding from the wound edge, additional TAE was performed to the superior and inferior gluteal arteries immediately after surgery. The wound edge became gradually necrotic and dehiscent (Fig. 3B). Finally, we performed debridement and wound closure operation 4 weeks after the first surgery.
The second debulking surgery in the lower leg was conducted 4 months after the first debulking surgery. At this time, we could apply a tourniquet to the thigh area. The total blood loss during the second surgery was 420 mL, and TAE was not conducted. Wound healing was not delayed. The patient left the hospital walking by himself (Fig. 2C).

Discussion
DPN presents as a highly vascular tumour, and the number of feeder vessels increases with the proliferation of the tumour. Histological study has confirmed that the vascular walls become vulnerable because of thinning of the tunica media and fragmentation of the tunica elastic [9]. Hence, this tumour is prone to bleeding, and haemostasis is difficult. Especially in the lower limb DPN, massive haemorrhage is inevitable [10]. In our cases, the bleeding volumes were 500 and 4970 mL.
We assumed that the difference in the amount of bleeding between the two cases was caused by two reasons: location and invasiveness of the tumour. In case 2, the tumour resection area extended to the superior buttock. There is an abundant blood supply from the superior gluteal artery, a branch of IIA, in the superior buttock area [11]. The collateral pathway from the contralateral IIA via the iliolumbar artery has been reported [12]. We concluded that ipsilateral TAE and balloon occlusion are not enough to achieve haemostasis from the tumour in the buttock area, because of the nature of the numerous collateral pathways.
Second, the invasiveness of the tumour was markedly different in the two cases. The tumour invaded into the muscle and its surrounding soft tissues to a higher degree in case 2 than in case 1. As delineating the tumour margin in deeply infiltrative DPN is almost impossible, unintentional intralesional resection is not rare [13]. Therefore, we inferred that one of the reasons of high-volume blood loss in case 2 was cutting into the well-vascularised tumour.
The efficacy of TAE for haemorrhage control in surgical resection of lower limb DPN has been reported in several studies [7,14]. Among various materials used for TAE, gelatine particles made from skin and ligament of other animals were introduced as absorbable haemostatic agents in 1946 [15]. In this study, we adopted Serescue TM , porous gelatine particles available in Japan. The reported mean diameter of occluded arteries with gelatine particles was 0.5 mm [16]. Also, we confirmed the gelatine particles within the 0.5-mm-sized arteries of the resected tumour (Fig. 3C).
Although gelatine particles are reliable haemostatic materials, focal ischaemia and necrosis of normal tissues are expected side effects of embolisation [17]. Gelatine particles in occluded arteries are thought to be removed by phagocytosis. Local tissues are found to be ischaemic for 2 weeks until vessel recanalisation was achieved. In addition, the recanalised vessel lumens are stenosed as a result of transluminal inflammation and smooth muscle migration due to foreign body reaction [18]. Delayed wound healing in case 2 can be attributed to TAE with gelatine particles. In fact, the stenosis of recanalised feeder vessels in the specimens of debridement was microscopically confirmed (Fig. 3D).
We struggled to achieve both adequate haemostasis and wound healing. As mentioned above, IIA receives blood supply from the contralateral side. Therefore, effective haemostasis can be expected in the buttock area by the intervention of bilateral IIAs. However, TAE with gelatine particles to bilateral IIAs leads to extensive necrosis of skin and muscle tissue in the buttock [19]. This can be explained by the infarction induced by gelatine particles remaining within the vessels after a certain period of time. On the contrary, balloon occlusion does not cause prolonged ischaemia and delay in wound healing postoperatively, because the balloon can be deflated as needed. In case 2, we placed the occlusion balloon to the unilateral IIA. In the field of obstetrics and gynaecology, balloon occlusion of bilateral IIAs has been reported to be an alternative haemorrhage-control method [20]. Further studies about indications and effects of balloon occlusion of bilateral IIAs in the debulking surgery of lower limb DPN are desired.

Conclusions
We adopted TAE with gelatine particles and unilateral balloon occlusion as haemorrhage control for the debulking surgery of lower limb DPN. Development of the ideal method for reducing blood loss to ensure the balance of haemostasis and wound healing is necessary.

Conflicts of interest
The authors have no conflicts of interest to declare.

Funding
None.

Ethical approval
This study was exempt from ethical approval in Kobe University Hospital.

Consent
Written informed consent was obtained from the patients for publication of this article including all figures. A copy of the written consent is available for review by the Editor-in-Chief of this journal on request.

Author contribution
Daiki Kitano, Takeo Osaki, Mika Nakasone and Tadashi Nomura performed the operations. Daiki Kitano and Takeo Osaki drafted the manuscript. Tadashi Nomura, Kazunobu Hashikawa and Hiroto Terashi supervised the operations and this case report.

Registration of research studies
Research Registry. UIN: researchregistry4516.