Surgical Treatment of Axillary Hyperhidrosis Treatment & Management

Medical treatments seek to decrease sympathetic stimulation, block sweat gland openings, or produce neuro-acinar blockade.

Dietary restriction

Since hyperhidrosis of all kinds can be exacerbated by stimulant-containing foods, especially caffeine and theobromine, dietary restriction of coffee, tea, cola soft drinks, and chocolate may improve mild cases of hyperhidrosis.

Topical antiperspirants

Most patients with the clinical syndrome try a variety of topical antiperspirants and deodorants, but find no relief until they use 10% or 20% aluminum chloride (Drysol) applied daily. These treatments may be quite effective for mild-to-moderate cases of hyperhidrosis.

Shelley and Hurley suggest a more complicated regimen using aluminum chloride hexahydrate or zirconyl chloride in absolute alcohol applied at bedtime, with an occlusive plastic dressing applied until morning. ^[8] They hypothesize that the metallic antiperspirants enter the sweat gland duct and form an occlusive plug by combining with ductal keratin; this treatment may be effective for a week at a time.

Unfortunately, skin irritation is very common with these antiperspirants and often forces discontinuance of the treatment.

Iontophoresis

Tap water iontophoresis is a fairly effective therapy in palmoplantar hyperhidrosis but is more cumbersome and less effective in axillary hyperhidrosis. In this procedure, moistened sponges wrapped around metal electrodes are inserted into each axilla for 20 minutes, and a low-voltage current is applied to the skin several times a week; it produces a stinging sensation. Adverse effects of this method are discomfort with burning, tingling, and skin irritation, including erythema and vesicle formation. Incorrect use may induce iontophoretic burns at sites of minor skin injury. A commonly marketed brand name is the Drionics device. The mechanism by which iontophoresis produces its sweat-decreasing effect is unknown.

Sympatholytics, anxiolytics, and sedatives

Administration of anticholinergic agents and beta-blockers can be quite helpful in mild cases of hyperhidrosis. Glycopyrronium bromide (Robinul), atropine, propranolol, and a host of anxiolytic (eg, Klonopin) and psychotherapeutic (eg, Prozac) medications have been used in the treatment of this disorder, with varying degrees of success.

Based on two replicate randomized, double-blind, vehicle-controlled, 4-week, phase 3 trials (ATMOS-1 and ATMOS-2), Glaser et al reported that primary axillary hyperhidrosis can be effectively treated topically with glycopyrronium tosylate 3.75%, administered daily. The severity of sweating, as assessed using the Axillary Sweating Daily Diary, and the production of sweat, as evaluated gravimetrically, were both reduced over the 4-week period. ^[9] In June 2018, the US Food and Drug Administration approved glycopyrronium tosylate 2.4% (Qbrexza) for the topical management of primary axillary hyperhidrosis in adults and in children aged 9 years or older. ^[10]

Unfortunately, in all but the mildest cases of hyperhidrosis, the doses of medications required to truly control abnormal sweating often cause significant adverse effects, including drowsiness, dry mouth, dilated pupils, photophobia, blurred vision, acute glaucoma, impaired micturition, reduced bronchial secretions, constipation, confusion, nausea, vomiting, giddiness, tachycardia, palpitations, and arrhythmias. Thus, many patients are forced to discontinue this avenue of treatment.

Radiation

Low-dose topical radiation produces at least transient decrease in local sweat production and has been suggested as a treatment for hyperhidrosis. Postradiation atrophy of the axillary sweat gland layer can be demonstrated. However, potential serious late adverse effects of radiotherapy make this a modality that cannot be recommended as a realistic solution for hyperhidrosis.

BOTOX® injections

Botulinum toxin A blocks neuronal acetylcholine release at the neuromuscular junction and in cholinergic autonomic neurons; it thus disconnects axillary sweat glands from their innervation. In an elegant study, Heckmann et al demonstrated quantitatively the effective safe treatment of axillary hyperhidrosis by intradermal injection of botulinum toxin A. They also demonstrated the longevity of the relief produced: 24 weeks after the injection of 100 U, the rates of sweat production (in the 136 patients in whom the rates were measured at that time) were still lower than baseline values (67 ± 66 mg/min in the axilla that received 200 U and 65 ± 64 mg/min in the axilla that received placebo and 100 U of the toxin).

A study by Brehmer et al indicated that the duration of botulinum toxin A’s effects in primary axillary hyperhidrosis increase with successive treatments. The first injection, in 101 patients, had a median efficacy of 4.0 months, compared with 4.5 months for the second treatment and 5.0 months for the third injection. ^[11]

BOTOX® injections have become an established and, often, effective treatment for axillary hyperhidrosis. This treatment has also raised the profile of the condition with insurers. Although effective, the success of the treatment seems to be quite technique-dependent (ie, the treatment works when injections are done by one practitioner and not when done by another). Major drawbacks are the expense of the toxin and the need for repeated treatments.

Surgical treatment of axillary hyperhidrosis involves resection of the end organ (adenectomy) or interruption of sympathetic innervation to autonomically innervated structures, including sweat glands.

Open sympathectomy

Open sympathectomy (via thoracotomy), though still practiced, is rapidly losing ground to its endoscopic counterpart, which theoretically achieves the same therapeutic goal with much less patient morbidity. Kotzareff first used the technique in 1920, with removal of part of the sympathetic chain for hyperhidrosis.

Endoscopic sympathectomy

ETS interrupts or ablates the high thoracic sympathetic chain to decrease sympathetic tone to the upper extremity and/or face. It is useful for causalgia, unreconstructable vascular disease, and treatment of severe frostbite, Raynaud syndrome, and acrocyanosis, in addition to its usefulness in hyperhidrosis. ETS is carried out under general anesthesia through one or more small insertion incisions. Among other things, sympathectomy results in a relaxation of peripheral arteriovenous connections; this results in vasodilatation of the skin with increased warmth. It also produces a cessation of sweating. Beneficial effects reportedly last as long as 5-10 years.

ETS is now often accomplished on an outpatient basis.

Clipping

A prospective study by Fibla Alfara et al indicated that axillary hyperhidrosis can also be managed by inserting clips bilaterally into the thoracic sympathetic chain, a procedure that can be reversed if the patient develops intolerable compensatory sweating. The investigators reported that clipping was effective in 74.3% of patients with axillary hyperhidrosis. ^[12]

Skin resection (debulking of gland volume)

Wu et al and others advocate composite removal of skin and associated axillary glands to treat axillary hyperhidrosis, but such resection is likely either to leave behind significant functional gland tissue or to produce major scarring and limitation of range of motion. Such resections are more appropriately reserved for patients with hidradenitis.

Suction adenectomy

A liposuction technique has been used to attempt axillary adenectomy in hyperhidrosis and has been recommended by Lillis and Coleman and Shenaq et al. ^{[13, 14]} A small-bore liposuction cannula is used, with the suction opening turned toward the underside of the skin. As the cannula is raked across the underside of the skin bearing the hypertrophic sweat glands, the expectation is that a high proportion of the glands are either sliced off or disrupted. In the author's experience, a disappointingly high fraction of patients treated by this method experienced relapse of abnormal sweating. The early disappearance of excess sweating is conjectured to be the result of local denervation and gland disruption; however, as patients are observed over time, reinnervation and/or gland regeneration takes place (typically within the first 6 mo postoperatively), and the symptoms recur. The author has abandoned this method of treatment; this conclusion is echoed by Ellis, who used curettage. ^[15]

A variant method of suction adenectomy using ultrasonic liposuction is thought to produce a better long-term result by producing more ablation than resection of gland tissue (although this method is believed to be operator-dependent).

Open adenectomy

See Skoog procedure, below, and Preoperative Details and Intraoperative Details.

Skoog procedure

The Skoog procedure ^[16] is a well-tolerated, effective, and permanent treatment for axillary hyperhidrosis. It is only appropriate for patients with hyperhidrosis--solely or predominantly--of the axilla. ETS is the more appropriate choice for patients who also suffer from palmar or facial hyperhidrosis.

Endoscopic adenectomy

Arneja et al have developed an endoscopic method of gland resection using an arthroscopic shaving device. ^[17] While the gland layer is not visualized with this method, it promises to produce a thoroughness of resection that approaches open gland removal under magnification. If results from this approach match those of the Skoog adenectomy in terms of cure rate, reoperation rate, and freedom from complications, this approach may well supplant open adenectomy.

Endoscopic sympathectomy

Use history, physical, and preoperative chest radiographs to exclude patients with preexisting pulmonary pathology. Many surgeons prefer to have the anesthesia conducted with a double-lumen endotracheal tube. Anesthesia monitoring should be extensive and may include arterial blood pressure, ECG, pulse oximetry, end-tidal carbon dioxide, and peak airway pressures.

Skoog procedure

Very little preoperative preparation is required for this local-anesthetic procedure. A shower with regular soap or antibacterial soap (eg, chlorhexidine) the morning of surgery is an appropriate precaution. The surgery is slightly more convenient if the patient (or surgeon) shaves the axilla, but the operation is perfectly feasible in an unshaved field. In light of the relative hypoperfusion that flap dissection produces, a single dose of intravenous antibiotics at the commencement of the operation is probably a wise precaution. The patient should be scrupulously free of any anticoagulant medications (eg, aspirin, ibuprofen, high-dose vitamin E, ginkgo, ginger, garlic, dong quai).

Endoscopic sympathectomy

For unilateral sympathectomy (or if the patient is to be turned during surgery), the lateral decubitus approach allows good visualization of the sympathetic chain in the upper thorax. Access to both sympathetic chains also can be achieved easily with the patient supine and both arms abducted. ^[18]

Make a small first incision in the anterior axillary line over the third rib and insert a needle into the pleural space to deflate the ipsilateral lung (or stop ipsilateral ventilation if using a double-lumen endotracheal tube).

Remove the needle and replace it with a 5.0 or smaller trocar matching the endoscope being used. Introduce the endoscope; the upper sympathetic chain and ganglia then are visualized easily lying over the costovertebral junctions. Pleural adhesions need to be taken down, avoiding (if possible) production of a parenchymal pleural leak. Many surgeons want a second port. This can be placed using a trocar at the fourth rib, also in the anterior axillary line. ^[19]

Identify the second rib within the chest cavity and open the pleura over it, starting just medial to the sympathetic chain and continuing laterally for approximately 1-2 inches.

Search for the nerves of Kuntz, a neural communication bypassing the sympathetic chain from the second or third thoracic segments and innervating the upper extremity. Although this anatomic variant is present in only approximately 10% of patients, it should be divided (if present) to avoid a failure of sympathectomy.

Make a longitudinal incision following the medial border of the sympathetic chain down to the fourth or fifth rib. Dissect thoracic ganglia T2, T3, and T4 free, and interrupt the sympathetic chain at the appropriate level. Small blood vessels are often encountered adjacent to the sympathetic chain and can be controlled by clips or cautery. Methods of interruption include cautery, harmonic scalpel, clip, or resection.

Individualize the sympathetic resection somewhat according to patient symptomatology. Since innervation of the head and neck is from T1 to T5, while that of the upper extremities is from T2 through T9, treatment for facial hyperhidrosis probably should include interruption above the T2 ganglion, while ablation of T5 probably is required to treat axillary hyperhidrosis. Extensive proximal dissection of the sympathetic chain (above the T1 ganglia) may lead to injury of the stellate ganglion, and injury to the cervical C7 component of this ganglion may result in permanent postoperative Horner syndrome. ^[20]

Gradually expand the lung by positive pressure ventilation while air in the pleural cavity exits via one of the cannulas.

Suture closed the trocar sites in layers. If clinically indicated and if the patient is doing well, a sympathectomy then can be performed on the opposite side.

Obtain a chest radiograph in the recovery room. A chest tube is not necessary unless a parenchymal air leak is noted. Flooding the operative field with saline while ventilating the lung may help demonstrate an air leak. If an air leak is present, a small drainage catheter can be introduced via one of the cannula sites and aimed toward the apex.

Thanks to Dr. Cliff Connery for editing and revising the above description of ETS. He is chief of the Thoracic Surgery Division at St. Luke's/Roosevelt Hospital in New York and Professor of Clinical Surgery at Columbia College of Physicians and Surgeons and has extensive experience in ETS.

Skoog procedure

Position the patient supine with the arm abducted approximately 120°. Use routine sterile skin preparation and draping. Consider profuse sweating, which may wet the prepared field, sterile; it ceases when local anesthetic is infiltrated.

Infiltrate the entire hair-bearing area of the axilla with local anesthetic; the author routinely uses a mixture of 0.5% lidocaine with epinephrine and 0.25% bupivacaine and a 25-gauge spinal needle. See the image below.

Plan the incision with a transverse line following a skin crease across the center of the hair-bearing axillary skin, from the anterior axillary fold to the posterior extent of the hair. In the classic Skoog procedure, staggered-cross incisions are added roughly perpendicular to the transverse limb and extending to the proximal and distal extremes of the hair-bearing area, so that the entire hair-bearing axilla is divided into 4 quadrants. See the image below.

In the author's variation of the Skoog procedure, only the single transverse incision is used. Although this makes the dissection substantially more difficult, it produces a less visible scar, a lower risk of flap necrosis, and an easier closure. Wang and others have recommended multiple parallel longitudinal incisions. See the image below.

Carry the incisions through the skin and the subcutaneous layer, which in these patients is formed almost entirely by tan-pink sweat glands (color very similar to pancreatic or parotid gland). In thin patients, a fascia layer is usually discernible covering the deep axilla; a cleavage plane just superficial to this facilitates dissection. See the image below.

None of the dissection should be deep to the fascia layer. The extent of dissection corresponds roughly to the hair-bearing axilla or to the mapped area if preoperative mapping is used. Reaching the limits of dissection farthest away from the skin incision usually requires use of curved scissors (eg, curved Mayo scissors) and retraction with skin hooks or narrow/deep retractors (eg, mini Deaver). The surgeon also may need to fold or roll the dissected flap on itself as dissection proceeds and may need to change position with relation to the abducted arm (eg, working across the patient's chest or from above the abducted arm). See the images below.

Meticulous hemostasis with a cautery device is appropriate at this stage, but needs to be limited once gland dissection begins to preserve flap vascularity.

Gland resection is started most easily at the cut edge of the incision and proceeds toward the axillary periphery. Using a fine skin hook, evert the flap over the side or pad of the surgeon's finger and use the convex side of a fine curved scissor (eg, Stevens or Littler scissors) to trim away the glands; the technique is similar to that used for defatting full-thickness skin grafts. With a little practice, removing essentially all of the gland tissue in a sheet without damaging the fine vessels revealed on the underside of the dermis by the resection is possible. The gland color and texture differentiate them from other structures in the operative field and are the most useful determinants of the adequacy of extirpation. See the images below.

The endpoint of gland resection is usually apparent when the carpet of tan gland tissue runs out. However, peripheral "islands" or "peninsulas" of glandular tissue covered by a thin layer of fat may fool the surgeon; missing these small collections of hypertrophic gland tissue is the probable cause of most sweating "recurrences" once the missed tissue becomes reinnervated. To prevent this failure, check the apparent end of hypertrophied gland mass against the axillary hair pattern or against an outline determined by preoperative mapping. Dissecting approximately 1 cm beyond the apparent end of the glandular tissue carpet also may help prevent recurrence without unduly compromising flap circulation. See the image below.

Close the incisions with inverted deep dermal absorbable sutures and running skin sutures. Place a drain through a small separate incision at the most inferior extreme of dissection, since without this step a hematoma or seroma often accumulates. See the images below.

Place a padded compression dressing over the entire dissected area, hinged at the depth of the axilla: this can be made with fluffed gauze, layers of Kerlix, or other resilient pad, bent into a V. The brachial limb of the V is taped to the upper arm or held in place with an elastic bandage, while the thoracic limb of the V is taped to the chest and may also be held with a halter of Kerlix gauze wrapped across the contralateral trapezius. Commercially-available elastic compression garments can also hold the dressings in place. See the images below.

Postoperative considerations and follow-up care: The drain may be removed in 72 hours. Some kind of padded compression dressing should remain in place for a week. With absorbable deep sutures in place, skin sutures can be removed safely at approximately 1 week. Most patients resume normal nonabducted arm activity as soon as the dressings are removed; exercises that include overhead stretches can be resumed gradually starting 2 weeks after surgery.

Endoscopic sympathectomy

In addition to the problems seen after chest procedures such as bleeding, infection, incisional problems, and pneumothorax, patients who have undergone ETS can also experience compensatory sweating, gustatory sweating, Horner syndrome, eyelid ptosis, and cardiac effects similar to beta-adrenergic blockade.

However, compensatory sweating is the main limitation of sympathetic surgery. In compensatory sweating, patients note increased perspiration over the trunk and upper thighs, especially in hot weather. Compensatory sweating may decrease with time, and most patients find it less objectionable than the hyperhidrosis for which they sought treatment. If patients are questioned carefully, compensatory sweating can be noted in as many as 75% of patients undergoing sympathectomy.

It has been believed that the number of levels interrupted and inclusion of the T2 ganglion in the resection increase the likelihood of severe compensatory sweating. (Current practice would usually employ T2 resection for facial symptoms, T3 for palmar, and T4-T5 for axillary.)

However, a retrospective study by Gunn et al indicated that in patients who undergo endoscopic thoracic sympathectomy for primary hyperhidrosis, the risk of developing compensatory hyperhidrosis may not be influenced by the extent of the surgery. The study, of 97 patients who underwent the procedure for palmar or axillary hyperhidrosis, found that the incidence of compensatory hyperhidrosis did not differ between patients based on whether they underwent resection at the T2-T3, T2-T4, T2-T5, or T2-T6 levels. Only 4 of the study’s patients reported severe compensatory hyperhidrosis. ^[21]

Gustatory sweating is a rare sequela of ETS. These patients experience the sensation of sweating when they eat, although no excessive sweat is actually produced.

Postoperative Horner syndrome may occur if portions of the stellate ganglion are removed or coagulated. This is fairly uncommon after ETS since the stellate ganglion is well protected by the dome of the pleura. Eyelid ptosis also may occur following ETS.

Following an extensive thoracic sympathectomy (especially one extending down to T5), cardiac effects may occur similar to those produced by beta-adrenergic blockade.

An incomplete sympathectomy may fail to produce symptomatic relief. Failure to find nerves of Kuntz, if present, also may lead to a suboptimal result; these extraganglionic sympathetic pathways are present in at least 10% of patients and must be sought carefully during the resection procedure.

Skoog procedure

The most common complications of axillary sweat gland resection are related to vascularity of the flaps. Overly aggressive dissection damages the fine vessels that supply the dissected flaps; hematoma and infection can also act to compromise circulation. The marginal vascularity of the flap edges makes them susceptible to crushing damage that less-traumatized tissues survive without a mark; noting spots of flap edge necrosis wherever toothed or crushing forceps have been used is not uncommon. Overly helpful (aggressive) retraction by an assistant, especially during an attempt to reach the extremities of the undersides of the flaps, also can easily lead to flap edge damage.

Any but the most minor length of necrosis along the suture line will lead to dehiscence with consequent discomfort, drainage, and eventual hypertrophic scar.

Hematoma, seroma, and infection can occur and increase the risk of flap devascularization despite careful hemostasis, aseptic technique, preoperative preparation, drain placement, and compressive dressings.

Incomplete excision of a portion of the hypertrophic gland mass leads to recurrent sweating once the acinar tissue becomes reinnervated, typically several months after surgery. This is the most common reason for reoperation.

Although hypertrophic and keloid scars are uncommon in the axilla with primary healing, they CAN occur and may require triamcinolone injection, topical silicone gel sheeting, or even revision.

Although compensatory sweating can occur with transthoracic sympathectomy, it is never observed in axillary adenectomy.

Endoscopic sympathectomy

Herbst et al studied 323 patients nearly 15 years after endoscopic thoracic sympathectomy and summarized their results as follows:

There was no postoperative mortality and no major complications requiring surgical reintervention. A majority of the patients (98.1%) were relieved, and 95.5% were satisfied initially. Permanent side effects included compensatory sweating in 67.4%, gustatory sweating in 50.7% and Horner's triad in 2.5%. However, patient satisfaction declined over time, although only 1.5% recurred. This left only 66.7% satisfied, and 26.7% partially satisfied. Compensatory and gustatory sweating were the most frequently stated reasons for dissatisfaction. Individuals operated for axillary hyperhidrosis without palmar involvement were significantly less satisfied (33.3% and 46.2%, respectively). ^[22]

Sympathetic regeneration has been documented in animals and humans and may result in late recurrence of hyperhidrotic symptoms. In animals, sympathetic fibers have remarkable regenerative ability, with regrowth and reinnervation through muscle and scar over time.

To achieve axillary anhidrosis, more extensive and/or more caudal resection of the sympathetic chain has been advocated, but this also seems to increase the chances of compensatory sweating and cardiovascular dysautonomia.

Skoog procedure

Every dissected (thus denervated) axilla is dry, even if no sweat glands are resected. Therefore, long-term outcome of sweat gland resection can only be judged after 6 months, when reinnervation should already have taken place. Most patients have a dry axilla for the few months following surgery, then develop some sweating, which the patients characterize as "normal." With meticulous dissection, long-term relief of hyperhidrosis approaches 95%. In a study of use of a mechanical gland-shaving device, Park et al reported an overall satisfaction rate of 94.7% with a complication rate of 13.2%. ^[23]

See the image below.

Axillary hyperhidrosis is a condition that is little recognized by patients and physicians outside Scandinavia and Great Britain. Patients who suffer from this condition often are misdiagnosed or dismissed as having psychosomatic complaints. Worse, various ineffective therapies are often recommended by "experts."

In Britain and Sweden, treatment of axillary hyperhidrosis is generally covered under the respective National Health Service. In the United States, most insurers deny coverage because surgery is not considered "medically necessary" or because the insurer is either unaware of the procedures or considers them "experimental." No CPT code is defined for the Skoog procedure, but hyperhidrosis has ICD-9 codes as follows:

• 705.21 - Primary focal hyperhidrosis (axilla, face, palms, soles)

• 705.22 - Secondary focal hyperhidrosis

Increased popularity and promotion by the manufacturer of BOTOX ® treatments for hyperhidrosis have raised the profile of the condition and made some insurers a more aware of the diagnosis. However, coverage of this treatment for hyperhidrosis remains inconsistent and uncommon.