Severe complication after microwave thermoablation of a residual limb: multifocal full thickness cutaneous necrosis and deep venous thrombosis

Abstract

Patients with limb loss face the challenge of having an increased risk of skin disease at residual limb sites. Hyperhidrosis is a common concern for persons with amputation and excessive sweat can cause various skin pathology. Recently, microwave thermoablation (MT) was reported as an effective off-label treatment for hyperhidrosis in patients with limb loss. We present a case in which a patient following MT procedure for hyperhidrosis of a transtibial amputation developed multifocal full thickness cutaneous necrosis and deep venous thrombosis. The possible aetiologies of these complications are discussed including vascular congestion and external thermal injury.

Such a case warrants the attention of individuals and providers seeking to use MT for off-label purposes, particularly for patients with limb loss, due to the large treatment surface area and potential for temporary or permanent functional loss of the amputated limb.

Background

Microwave thermoablation (MT) of eccrine glands (Miradry, Miramar Labs, Santa Clara, California, USA) is an U.S. Food and Drug Administration (FDA)-approved procedure to permanently reduce primary hyperhidrosis of the axillae. This non-invasive device is applied to the skin, after anaesthesia, typically local tumescent, and works by selectively heating layers where eccrine and apocrine sweat glands are located. By emitting a microwave frequency of 5.8 GHz that allows for large amounts of absorption at the dermal/hypodermal interface, the device causes thermolysis of the glands.¹ Low amounts of absorption at the epidermis, as well as the use of epidermal contact cooling, allows the device to target the glandular zone relatively independent of an individual’s skin thickness.¹ One clinical trial of 31 individuals with primary axillary hyperhidrosis reported that over 90% of patients experienced at least a 50% reduction in axillary sweat from baseline at the 12-month follow-up.² All subjects experienced relatively mild side effects, to include transient numbness, swelling and discomfort.² Twelve subjects experienced altered sensation of the skin, but this fully resolved in all cases.² Commonly reported complications associated with Miradry include transient skin irritation and numbness; however, uncommon yet severe complications have been reported. Such complications include brachial plexus injuries or, in one case, death.^3–5

Recently, off-label use of MT was performed in patients with limb loss, as hyperhidrosis was the most reported skin issue in a survey of lower limb amputees.⁶ Excessive sweat can lead to skin pathology such as inflammatory dermatoses, fungal and bacterial infections, maceration, ulceration and verrucous hyperplasia.⁷ Moreover, sweat accumulating within the liner can cause the prosthesis to slip and can contribute to discomfort, poor terminal device control and falls.⁸ Management of hyperhidrosis is important to ensure optimal skin health, mobility and quality of life in patients with limb loss.

Here, we present the case of a medically retired Air Force woman in her 30s who underwent Miradry microwave-based treatment for primary focal hyperhidrosis of a left transtibial amputation. She developed many scattered areas of retiform purpura within the treatment area 6 days postprocedure (figure 1) that developed into several partial and full-thickness wounds (figure 2) and concurrent popliteal deep vein thrombosis.

Figure 1.

Skin changes 6 days following treatment. (A) Left medial distal leg. Scattered many areas of retiform purpura (black circles) and haemorrhagic bullae (red arrows) confined to within the treatment area. (B) Left medial proximal leg. Scattered many patches of livedo racemosa (black circles), some with purple central purpura (red arrows).

Figure 2.

Skin changes 3 weeks following treatment. (A) Left lateral leg. (B) Left distal thigh. (C) Left posterior leg. (D) Left posterior thigh. Scattered Full thickness ulcers (black arrows), some with overlying eschar (red arrows) noted 3 weeks following treatment. Postinflammatory hyperpigmentation throughout treatment area.

Case presentation

A woman in her 30s with a medical history significant for elective left transtibial amputation for refractory complex regional pain syndrome of the ankle presented with complaints of hyperhidrosis of the posterior left lower residual limb. She previously underwent MT treatment 3 years prior for similar complaints of hyperhidrosis of the anterior residual limb without complication. Of note, due to her active lifestyle and the large volume sweat burden affecting ambulation during exercise as well as failure of topical aluminium chloride, she elected to forgo other intermediate treatments for hyperhidrosis prior to her first Miradry treatment.

Large volume tumescent anaesthesia (ratio of 250 cc normal saline; 50 cc 1% lidocaine with 1:100 000 epinephrine; 2.5 cc of 2.4% sodium bicarbonate) was injected locally until the patient was comfortable intraprocedure. A total of 600 cc were used and the procedural area was cleaned and then marked with a custom grid pattern (figure 3). A thin layer of water-based lubricant was applied, and the procedure was performed without immediate complication. The patient was provided a compression cooling device to control swelling and instructed to use ibuprofen 800 mg alternating with 650 mg of acetaminophen every 6–8 hours for pain control. Six days postprocedure, the patient presented to the clinic with prominent oedema of the lower extremity, limited range of motion as well as scattered many red to purple retiform purpuric patches and palpable plaques as well as scattered few tense haemorrhagic bullae within the treatment area (figure 1A,B). Otherwise, she reported minimal pain and was well appearing with stable vitals and no signs of infection. The patient reported that she had been applying ice only to the treatment area because the compression device was not working.

Figure 3.

Intraprocedure photo. Note to limit treatment overlap, a custom grid pattern is used. Every effort was made to keep the system at the highest setting of level 5. The energy delivery was aborted 3–4 times during treatment due to patient discomfort. The popliteal fossa was not treated due to concern over the proximity to the deep neurovascular bundle.

Investigations

A left lower extremity Doppler ultrasound was ordered to evaluate for a superimposed deep venous thrombosis (DVT). Wound culture was obtained to assess for potential infection. A skin biopsy was performed to provide clinical pathological correlation. Due to the biopsy results and her findings of a popliteal deep vein thrombosis on ultrasound, a coagulopathy panel was initiated after cessation of her 90-day course of apixaban.

Differential diagnosis

The retiform purpura morphology within the treatment area suggested a vascular aetiology (vasculopathy vs vasculitis) thus prompting the skin biopsy. The haemorrhagic bullae and background erythema also led to a high suspicion for thermal injury, especially given the recent MT procedure. Lower on the differential was an infectious aetiology given the patient was asymptomatic, afebrile and with normal vital signs otherwise.

Treatment

A 15-day prednisone taper (1 mg/kg) was initiated along with an empiric 7-day oral cephalexin course. Weekly wound care follow-up was arranged, and the patient was instructed to be non-weight-bearing. As mentioned above, a left popliteal venous thrombosis was noted on ultrasound and a 3-month course of apixaban was initiated.

Outcome and follow-up

The patient returned to wear of her socket at three and a half months following the procedure and evaluation at 5.5 months showed that all her wounds were healed and she was left with several complex hypertrophic and atrophic hyperpigmented scars as well as scattered postinflammatory hyperpigmentation (figure 4). The patient declined treatment of her scars at that time as they were largely asymptomatic. She returned to her preprocedure limb volume, obviating the need for a new socket.

Figure 4.

Skin changes 5.5 months postprocedure. (A) Left posterior leg. (B) Left lateral leg. Scattered multifocal hypertrophic (black arrows) and atrophic scars (red arrows). Scattered postinflammatory hyperpigmentation (black circles) throughout treatment area.

Discussion

Severe complications of MT treatment have been reported, the most common being injury to the brachial plexus.^{1 2} It is believed that such injuries occur due to close proximity of the nerves to the skin in individuals with a low body mass index.⁴ Off-label use of MT outside the axillae has been reported, with one case report of death secondary to infection following treatment. Wen et al detailed the case of a woman in her 20s who underwent MT for concern of odour in the perineal, perianal and genital regions.⁵ Six days following the procedure, she died of necrotising fasciitis (Fournier’s gangrene) complicated by streptococcal toxic shock syndrome.⁵ While this is a rare example of a fatal outcome, it demonstrates that off-label use of MT is not a procedure without risks.

Our patient underwent an elective transtibial amputation for refractory complex regional pain syndrome associated with an injury to the left ankle. The patient received anterior thigh microwave thermal ablation therapy prior to current presentation for complaints of focal hyperhidrosis while wearing the prosthesis. The patient did not experience any complications at that time and was happy with the results. She continued to experience focal hyperhidrosis in the posterior aspect of her limb and thus desired additional therapy to eliminate skin pathology and prevent slippage of her socket-based prosthesis.

Although there are other available options for the treatment of primary hyperhidrosis,⁸ our patient elected to undergo MT as a second-line treatment due to large volume sweat burden affecting her active lifestyle and importantly a desire for a permanent solution. Typically, in our practice, topical aluminium chloride is first line, oral or topical glycopyrrolate or injection with onabotulinum toxin are second line and MT is reserved for refractory cases. However, patient preference, morbidity and clinical circumstances will sometimes dictate early adoption of MT. Similarly, Stuart et al found in their systematic evidenced-based review of treatments for primary hyperhidrosis that optimal treatment choice depends on many factors and should be patient centred.⁹ In the case of our patient, Miradry was selected as second line therapy in alignment with her lifestyle and goals of care.

It was unclear what caused her complication from the MT treatment and was likely multifactorial. The retiform pupura morphology suggested a vascular aetiology and biopsy did, in fact, show a vasculopathy with thrombi in multiple reticular dermal and subcutaneous vessels (figure 5A,C). We hypothesise that her prolonged and severe postprocedural oedema and inflammation caused by the procedure led to a hypercoagulable state secondary to congestion. This is reinforced by the fact that a popliteal vein DVT was found on imaging and her coagulopathy workup was negative. It is unclear whether large volumes of tumescent anaesthesia or vasoconstrictive effects from the component epinephrine were contributory. On our review of the literature, there are no similar reports, though this modality of anaesthesia is typically used for liposuction of the trunk or proximal extremities.

Figure 5.

Histology of 4 mm skin punch biopsy taken perilesional to haemorrhagic bullae on the lateral lower thigh within treatment area. (A) ×40 magnification, H&E stain. Partial and full thickness skin necrosis (black circles) with areas of subepidermal separation (black arrow). (B) ×100 magnification, H&E stain. Scattered many elongated and dyskeratotic keratinocytes (black arrows). Extravasated red blood cells notable throughout the papillary dermis (black circles). Homogenisation of the papillary dermal collagen (black dashed arrows). (C) ×100 magnification, H&E stain. Example of subcutaneous vessel vasculopathy and associated intraluminal thrombus (black arrows).

Biopsy also showed partial and full thickness necrosis of the epidermis and scattered necrotic keratinocytes suggesting contribution from external thermal injury, (figure 5A,B) though one would have expected cutaneous evidence of this immediately postprocedure if the primary mode of injury were heat-induced (figure 3). The possibility of a delayed postburn blister was entertained, though haemorrhagic bullae are seen more commonly with vasculitis or vasculopathy. When questioned about postprocedural cold therapy the patient reports that she did not use bare ice on the limb and did not keep the ice in place for more than 15–20 min at a time. Inspection of the MT system by the manufacturer yielded no error codes during treatment suggesting that epidermal cooling was intact during the procedure.

In summary, we present a case of multifocal full thickness cutaneous necrosis and DVT sustained in a young, otherwise healthy female patient following an MT procedure for hyperhidrosis of a transtibial amputation. Fortunately, the patient returned to preprocedure function with residual scarring in areas of the skin necrosis. While MT has been demonstrated to be an option for residual limb hyperhidrosis with measurable effectiveness and long-lasting results, it is important that patients be counselled that this treatment is off-label and that more serious side effects such as skin necrosis and DVT are possible.

Learning points.

• Patients with limb loss face increased risk for skin disease, often exacerbated by excess sweating associated with prosthetic use.

• Microwave thermoablation (MT) and its off-label use in patients with limb loss is considered an effective and long-lasting therapy for hyperhidrosis.

• Rare complications of MT now include full thickness skin necrosis and deep venous thrombosis, which should be discussed with the patient before the procedure.

Ethics statements

Patient consent for publication

Consent obtained directly from patient(s).