Early complications of thoracic endoscopic sympathectomy: a prospective study of 940 procedures

Ладонный гипергидроз (повышенная потливость ладоней) - лечение повышенной потливости Блашинг-синдром (стрессовое покраснение лица, эритрофобия) Рефлекторная симпатическая дистрофия Комплексный регионарный болевой синдром как вариант хронической нейропатической боли Комплексный регионарный болевой синдром в травматологической практике Severe pompholyx following endoscopic thoracic sympathectomy for palmar hyperhidrosis Early complications of thoracic endoscopic sympathectomy: a prospective study of 940 procedures

Dominique Gossot, MD^a, Hassan Kabiri, MD^a, Raffaele Caliandro, MD^a, Denis Debrosse, MD^a, Philippe Girard, MD^a, Dominique Grunenwald, MD^a

^a Thoracic Department, Institut Mutualiste Montsouris, Paris, France

Accepted for publication December 18, 2000.

Address reprint requests to Dr Gossot, Thoracic Department, Institut Mutualiste Montsouris, 42 Bd Jourdan, F? Paris, France
e-mail: dominique.gossot@imm.fr

Background. Thoracic endoscopic sympathectomy (TES) has becomethe surgical technique of choice for treating intractable palmarhyperhidrosis and is usually considered as a simple and safeprocedure. To evaluate the complication rate of TES, we conducteda prospective study of peri- and postoperative complications.

Methods. From 1995 to 1999, 467 consecutive patients were operatedon for upper limb hyperhidrosis. There were 164 men and 303women, ranging in age from 15 to 59 years (mean 31 years). Inall but 5 cases, the procedure was bilateral. Eleven patientsunderwent a reoperation for failure; thus the total number ofsympathectomies was 940. The procedure was performed in twostages in 182 patients and in one stage in 267 patients. Allpatients were seen 1 month after the operation.

Results. There was no mortality. The mean postoperative hospitalstay was 2.3 days in the group of patients who were operatedon in two stages and 1.1 day in patients who were operated onin one stage. There were three major complications: one tearof the right subclavian artery and two chylothoraces. Therewere 25 cases (5.3%) of bleeding (300 to 600 mL) during dissectionof the sympathetic trunk due to injury to an intercostal vein;in all cases it was controlled thoracoscopically. There were12 pneumothoraces (1.3%) after removal of chest tubes. All ofthese were unilateral. Four required chest drainage for a periodof less than 24 hours. One patient had a mild pleural effusion.Four patients had a unilateral partial Horner Syndrome (0.4%)that disappeared within 3 months in 2 patients. The other 2patients were lost to follow-up. One patient complained of rhinitis.

Conclusions. Although morbidity was low, significant complicationsof TES occurred. Patients should be clearly warned that TESis not as minor a procedure as usually asserted. Complicationsas well as adverse effects should be considered when discussingthis surgical indication.

Thoracic endoscopic sympathectomy (TES) has become the preferredmethod for treating intractable palmar hyperhidrosis. It hasbeen popularized by European [1 — 3] and Asian teams [4]reporting a 95% to 99% success rate. Although some recent papershave called for caution with respect to adverse effects suchas compensatory sweating [5 — 7], TES is usually seen asa simple and safe procedure with a short postoperative hospitalstay. It has even been proposed as an outpatient procedure [8].This reputation of safe and benign procedure leads patientsand/or physicians to ask for TES without attempting medicaltherapy. To evaluate the complication rate of TES, we prospectivelyrecorded our results in a series of 940 consecutive sympathectomies.

From 1995 to 1999, 467 patients were operated on for upper limbhyperhidrosis. All were operated on by a single surgeon. Therewere 164 men and 303 women, ranging in age from 15 to 59 years(mean 31 years). In all but 5 cases, the procedure was bilateral.Eleven patients underwent a reoperation for failure; thus thetotal number of sympathectomies was 940.

The technique has been previously described [9]. The procedurewas performed under general anesthesia and selective trachealintubation. No CO₂ insufflation was used. We used a 5-mm, 0degree telescope and two additional 3-mm ports for microinstruments.The mediastinal pleura was opened and the sympathetic chainwas dissected, severed, and removed from the second thoracicganglion (T2) to the fourth (T4) or fifth ganglion (T5), dependingon whether the axillae were involved. Dissection was carriedout with high-frequency cautery except at the level of T2, whereno coagulation was used to prevent current diffusion to thestellate ganglion.

In 69 patients, a technique of selective sympathectomy was performed.Hence, the main trunk was preserved and only the Rami Comunicanteswere divided, according to the description of R. Wittmoser [10].This technique aimed at reducing the rate of compensatory sweating.Because of the high recurrence rate [9], it was abandoned andwe went back to a complete division of the sympathetic chain.All 11 reoperations were in this group of patients.

At the end of the procedure, a 15F chest tube was left in placefor a few hours. It was removed after checking the chest roentgenogramand the patients were usually discharged the morning after surgery.All patients received an analgesic prescription and a recommendationform. They were advised to see their general practitioner incase of severe pain to receive a prescription of opioid analgesics.In the beginning of our experience, the procedure was performedin two stages after an interval of 2 weeks (182 patients); thenboth sides were done as a one-stage procedure (267 patients).All patients were seen 1 month after the operation. Complicationswere recorded in a database (File-Maker Pro, Claris, Santa Clara,CA) at the time of discharge and on the day of consultation.

There was no mortality. The mean postoperative stay was 2.3days in the group of patients who were operated on in two stagesand 1.1 day for those operated on in one stage. The successrate was 100% for palmar hyperhidrosis. In the group of patientsoperated on according to the Wittmoser technique, there were11 unilateral relapses. Despite some intraoperative difficultiesdue to pleural adhesions, resecting the sympathetic trunk wasalways possible and all 11 reoperations were successful.

Intraoperative complications
In 1 patient, the endoscopic approach was impossible on oneside because of tight pleural adhesions. The procedure was notconverted to open surgery.

There were three major complications. One patient had a tearof the right subclavian artery in the beginning of our experience.It was caused by sliding of the hook during the opening of themediastinal pleura. An immediate axillary thoracotomy was performedand the tear was sutured. The postoperative course was uneventfuland the patient was discharged 4 days later. In 2 patients achylothorax occurred, one on the left side [11] and one on theright side. The left sided chylothorax was diagnosed only 3days after the patient had been discharged. It was treated bychest drainage and total parenteral nutrition. The effusionstopped totally after 6 days. The right chylothorax was diagnosedintraoperatively. Two clips were applied on the lymphatic duct.The patient was discharged after 2 days once it was confirmedthat the chest drainage was not productive after resumptionof oral intake.

There were 25 cases (5.3%) of significant bleeding (300 to 600mL) during dissection of the sympathetic trunk due to injuryto an intercostal vein. The amount of lost blood was not relatedto the vessel diameter but to the fact that the vessel usuallyretracted and was difficult to control. In all cases the hemorrhagewas controlled thoracoscopically. However, one 3-mm port hadto be changed to a 10-mm port because a 3-mm suction tube wasinefficient. In addition, in case of long lasting aspirationwith a 5-mm suction tube, we had to use a 10-mm trocar to avoidlung inflation. No patient needed blood transfusion.

No other intraoperative complication occurred. Eventually therate of major intraoperative complications was 3/940 (3/1000)and the total rate was 28/940 (2.9/100).

Postoperative complications
There were 12 pneumothoraces (1.3%) after removal of chest tubes.All of these were unilateral. Four required chest drainage forless than 24 hours. One patient had a mild pleural effusionthat resolved after physiotherapy.

Four patients had a unilateral partial Horner syndrome (0.4%).Two recovered within 3 months; the other 2 patients were lostto follow-up. One patient complained of rhinitis.

Although pain was not systematically recorded in our files,it was present in most patients for 2 to 4 weeks. Many patientscomplained of severe dorsal pain that required prescriptionof morphinics. Some patients reported pain along the internalaspect of one or both arms during several weeks. This pain alwaysdisappeared spontaneously. Three patients mentioned that theywould not have been operated on had they known that the postoperativecourse was so painful. In 1 patient who was scheduled to undergoa two-stage procedure, this was the reason for declining thesecond operation.

Excluding postoperative pain and the other above-mentioned problems,the total postoperative morbidity was 2/100.

Side effects
At 1 month, 236 patients (50.5%) complained of compensatorysweating. This was considered by patients as acceptable (67%),disturbing (25%), or debilitating (8%).

Of the patients, 33% experienced hand dryness. This conditionalways improved after some months; no patient found it disturbing.Improvement or even cure of the associated plantar hyperhidrosiswas observed in 29% of the patients.

Thoracic sympathectomy is an established operation. Althoughthe endoscopic approach was described many years ago [10], itbecame popular only at the end of the 1980s. TES is now consideredthe method of choice because of its low morbidity and shorthospital stay. The success rate for palmar hyperhidrosis ismore than 95% in most reports [1 — 3, 8, 9, 12 — 14]and the complication rate is significantly lower than with aconventional open approach [15]. It is therefore consideredby patients — seeking information on many dedicated Internetweb sites — as a minor and benign procedure. Recently someauthors have pointed out the underestimated incidence of adverseeffects such as troublesome compensatory sweating [4 — 6].The rate of intra- and perioperative complications, althoughlimited, should also be taken into consideration.

There was no mortality in our series as in other recent largeseries. However, lethal complications have been reported inthe literature. Cameron [16] has reported two cases of cerebraledema related to the use of intrapleural CO₂ insufflation. Onepatient died and the other suffered major neurologic sequellae.Gas insufflation is frequently used by some laparoscopist surgeons,although excessive caution should be used to avoid mediastinalhyperpressure and its consequences. Some surgeons perform thoracoscopywith airtight laparoscopic cannulae, which may worsen the situationbecause these do not allow air to be released if an intrapleuralhyperpressure occurs.

In our series three potentially serious complications were noticed:one tear of the subclavian artery and two chylothoraces. Allresolved without sequelae and after a short stay. Cameron has reported another case of subclavian artery injury that required34 units of blood and that led to secondary graft interposition.Lange [17] reported a case of serious damage to the brachialplexus early in his experience with TES. He stressed the factthat, despite his extensive experience with conventional thoracicsympathectomy, the endoscopic approach made him «lose his way»by misjudging the direction of the nerve fibers. A chylothoraxis rare and is more related to anatomical features than to surgicalexperience. Because it is due to the tear of an accessory duct,it usually remains moderate and can be cured by simple chestdrainage and medical therapy [11, 18].

Postoperative Horner’s syndrome (HS) is rare but is foundin almost all series. It can be total or partial (without miosis).It is caused by a direct or indirect damage to T1, ie, currentdiffusion or excessive traction on the nerve during dissection.Postoperative rhinitis is another symptom associated with aT1 lesion [15]. In our series we had four cases of partial HS(0.4%) and one of rhinitis. The reported rate of HS ranges from0% to more than 3% [2, 15]. Most authors agree than the endoscopicapproach reduces the rate of HS because of better visualization.Zacherl and colleagues [15] found a significant difference betweenthe open approach (4.6%) and the endoscopic approach (2.2%).As the magnification of the telescope allows for a much betterview of the sympathetic chain and ganglion, one may wonder whyHS is still observed. The cause may be the following: (1) diffusionof monopolar HF current to the stellate ganglion; (2) excessivetraction on the nerve during dissection, temporarily stretchingit; or (3) inadequate localization of the second rib.

The first cause (diffusion of current) can theoretically beavoided by using bipolar or ultrasonic technology. Unfortunately,none of these devices are available with a 3-mm diameter. Ifoperating with microinstruments, we therefore recommend notto use HF when coming close to the apex. In our experience,the mild oozing that is observed when severing the nerve withoutthe help of diathermy has never become a concern. In case ofpersistent oozing we sometimes introduce a microhemostatic gauze,which is applied on the sympathectomy bed. This maneuver hasalways been sufficient. The second cause (excessive traction)is the most difficult to prevent. One has to pay attention notto pull on the sympathetic chain or to leave the grasping forcepson the nerve and to release the sympathetic trunk during dissection.The third cause (misdetermination of the second rib) is alsoan issue. Theoretically, the first rib is not identified fromwithin the thorax. However, some authors claim that it can beseen or at least palpated [19]. We have found it sometimes difficultto determine whether the uppermost visible rib was the firstor the second, especially in thin and slim patients. The fatpad that usually covers the inferior part of the stellate ganglionmay be lacking. In these cases, it seems preferable to stayaway from the apex. Kopelman and colleagues [19] have shownthat the rate of HS decreases with experience. However, thefour cases of HS that we observed in our series occurred respectivelyin patients 114, 117, 190, and 389, which is not in favor ofa learning curve. We have observed a 1.3% rate of pneumothorax,as in all series (Table 1) Four patients needed redrainage.

Table 1. Intraoperative and Perioperative Morbidity of Thoracic Endoscopic Sympathectomy

P = number of patients; S = number of sympathectomies; NP = not precised.

Many other minor problems have been observed in our patientsthat have not been reported: shortness of breath, chest pain,severe dorsal pain, and neuralgia along the ulnar aspect ofthe upper limbs. These symptoms have always disappeared after2 to 6 weeks, but they can cause major discomfort during theearly postoperative period and may require prescription of opioidanalgesics. We did not observe persisting symptoms as reportedby other authors [19].

At 1 month, 236 patients (50.5%) complained of compensatorysweating. However, definite conclusions cannot be drawn fromthese data. It has been our experience that some patients seen1 or 2 months after the operation are satisfied and do not complainof excessive sweating, but do so 1 year later. Conversely, somepatients initially complain of disturbing compensatory sweatingand do much better after some months. Therefore, estimatingprecisely the rate of compensatory sweating would require athorough follow-up of at least 3 years after surgery [2]. Thissurvey is in process. It is expected that the true rate of compensatorysweating will be somewhat higher than the one found at earlyconsultation, as pointed out by Andrews and Rennie [5] who havefound this side effect in 85% of their patients.

Currently TES remains the most efficient treatment for upperlimb hyperhidrosis. However, considering possible side effectsas well as rare but significant complications, patients shouldbe clearly warned that TES is not a minor procedure. From thesurgeon’s standpoint, TES is usually considered to bea straightforward procedure; however, experience in endoscopicsurgery is essential to cope with a possible intraoperativedifficulty [17, 19]. Anesthesia must be done by an anesthesiologistexperienced in thoracic surgery and CO₂ insufflation must beproscribed. Adequate equipment is another important componentof a safe operation: (1) laparoscopic trocars must be abandoned(two cases of subclavian artery injury reported in the literatureare related to trocar insertion [13, 16); (2) microinstrumentsshould be preferred to conventional 5-mm laparoscopic or thoracoscopicinstruments because they are better adapted to the size of nervousstructures; and (3) optimal visualization must be obtained (Table 2).

Table 2. Case Reports of Major Intra- and Perioperative Complications During Thoracic Endoscopic Sympathectomy

1. Drott C., Göthberg G., Claes G. Endoscopic transthoracic sympathectomy: an efficient and safe method for the treatment of hyperhidrosis. J Am Acad Dermatol 1995;33:78㫩.
2. Herbst F., Plas E.G., Függer R., Fritsch A. Endoscopic thoracic sympathectomy for primary hyperhidrosis of the upper limbs: a critical analysis and long-term results of 480 operations. Ann Surg 1994;220:86㫲
3. Rex L., Drott C., Claes G., Göthberg G., Dalman P. The Boras experience of endoscopic thoracic sympathectomy for palmar, axillary, facial hyperhidrosis and facial blushing. Eur J Surg 1998;164:23㪲
4. Lin C.C., Mo L.R., Lee S.M., Ng S.M., Hwang M.H. Thoracoscopic T2-sympathectomy block by clipping: a better and reversible operation for treatment of hyperhidrosis palmaris. Experience in 326 cases. Eur J Surg 1998;164:13㪨
5. Andrews B.T., Rennie J.A. Predicting changes in the distribution of sweating following thoracoscopic sympathectomy. Br J Surg 1997;84:1702
6. Bonjer H.J., Hamming J.F., du Bois N.A. J.J., van Urk H. Advantages of limited thoracoscopic sympathectomy. Surg Endosc 1996;10:721��
7. Dohayan A. Transaxillary thoracoscopic sympathectomy experience in a hot climate: management of the dominant hand. Surg Laparosc Endosc 1999;5:317��.
8. Hsia J., Chen C., Hsu C., Shai S., Yang S. Outpatient thoracoscopic limited sympathectomy for hyperhidrosis palmaris. Ann Thorac Surg 1999;67:258��
9. Gossot D., Toledo L., Fritsch S., Célérier M. Thoracoscopic sympathectomy for upper limb hyperhidrosis: looking for the right operation. Ann Thorac Surg 1997;64:975��
10. Wittmoser R. Thoracoscopic sympathectomy and vagotomy. In: Cuschieri A., Buess G., Perissat J., eds. Operative manual of endoscopic surgery. New York: Springer Verlag, 1992:110��.
11. Gossot D. Chylothorax after thoracic endoscopic sympathectomy. Surg Endosc 1996;10:949
12. Byrne J., Walsh T.N., Hederman W.P. Endoscopic transthoracic electrocautery of the sympathetic chain for palmar and axillary hyperhidrosis. Br J Surg 1990;77:1046
13. Duarte J., Kux P. Improvements in video-endoscopic sympathectomy for the treatment of palmar, axillary, facial and palmar hyperhidrosis. Eur J Surg 1998;164:9㪣.
14. Shachor D., Jedeikin R., Olsfanger D., Bendethan J., Sivak G., Freund U. Endoscopic transthoracic sympathectomy in the treatment of primary hyperhidrosis. Arch Surg 1994;129:241��
15. Zacherl J., Imhof M., Huber E., et al. Video assistance reduces complication rate of thoracoscopic sympathectomy for hyperhidrosis. Ann Thorac Surg 1999;68:1177
16. Cameron A.E. Complications of endoscopic sympathectomy. Eur J Surg 1998;164:33㪻.
17. Lange J.F. Inferior brachial plexus injury during thoracoscopic sympathectomy. Surg Endosc 1995;9:830
18. Cheng W., Chang C., Lin T. Chylothorax after endoscopic sympathectomy: case report. Neurosurgery 1994;35:330��
19. Kopelman D., Hashmonai M., Ehrenreich M., Assalia A. Thoracoscopic sympathectomy for hyperhidrosis: is there a learning curve?. Surg Laparosc Endosc 1998;8:370��