- Email: [email protected]
Thoracoscopic cervicodorsal sympathectomy: Preliminary results
Thoracoscopic cervicodorsal sympathectomy: Preliminary results Samuel S. Ahn, M D , H e r b e r t I. Machleder, M D , Blessie Concepcion, BS, and Wesley S. Moore, M D , Los Angeles and Sylmar, Calif.
Purpose: The purpose of this study was to determine the feasibility, safety, and efficacy of thoracoscopic cervicodorsal sympathectomy. Methods: From March 1990 to December 1993, we performed 21 thoracoscopic sympathectomies in 19 patients. There were 13 women and six men; age 17 to 64 years, mean 37 years. Thirteen procedures were performed on the left side and eight on the right. Indications for surgery were cansalgia/reflex sympathetic dystrophy in nine patients, Raynaud's/vasculitis in six, hyperhidrosis in five, and medically refractory cardiac arrhythmia in one. Results: The T1-4 sympathetic ganglia were readily identified, dissected free, and resected thoracoscopically in 19 cases, and the T3-7 ganglia were resected with thoracoscopy in one case. One case required conversion to an open thoracotomy because of dense scar from a previous first rib resection, which obscured the anatomy. Histologic confirmation of ganglia were obtained in all 21 cases. Operative duration ranged from 1.0 to 3.5 hours. Estimated blood loss was 5 to 300 co, mean 42 cc, median 10 co. No patient required transfusion. All 21 patients had an excellent immediate sympathectomy response. Transient Homer's syndrome developed in two patients. Postoperative residual pneumothorax (< 10%) occurred in three cases and resolved spontaneously without further treatment. In one patient plettral effusion and pneumothorax developed, which were treated with the reinserdon of the chest tube. Postoperative pain was well controlled with oral analgesics. Hospital stay was 1 to 4 days, mean 2 days, median 1 day. Follow-up at 1 to 42 months, mean 11 months, median 6 months, showed continued evidence of sympathectomy effect in all patients, except one who died of her underlying disease 1 month after operation. Conclusions: We conclude that thoracoscopic sympathectomy is feasible, safe, and effective. Further studies are indicated to confirm its long-term benefits and to determine optimal thoracoscopic techniques. (J VAse SURG 1994;20:511-9)
Numerous methods of cervicothoracic sympathectomy have been described in the literature, which attests to the fact that an optimal method of sympathectomy is still lacking. The supraclavicular approach is well tolerated by the patient but carries significant risk of complications such as Homer's syndrome, phrenic nerve injury, brachial plexus injury, chylous leak, pneumothorax, and bleeding. From the UCLACenterfor the Health Sciences,Los Angeles,and Olive View Medical Center, Sylmar. This study was fimdedin part by the JOASH Foundation, Santa Monica, CA. Presented at the Ninth AnnualMeeting of the WesternVascular Society, Santa Barbara, Calif.,Jan. 9-13, 1994. Reprint requests: SamuelS. Ahn, MD, 100 UCLAMedicalPlaza, Suite 510, Los Angeles, CA 90024 Copyright © 1994 by The Society for Vascular Surgery and InternationalSocietyfor CardiovascularSurgery,North American Chapter. 0741-5214/94/$3.00 + 0 24/6/58563
These complications are less likely to occur with the transthoracic approaches, but an open thoracotomy causes greater morbidity and mortality rates. The posterior approach in particular, which transgresses the posterior shoulder girdle muscles and requires partial rib resection, causes significant tissue trauma and postoperative pain. Not surprisingly, physicians and patients have not embraced these procedures except for very selected situations. In 1951, Kux ~ described a minimally invasive thoracoscopic approach to perform cervicodorsal sympathectomy. In 1978, 2 he reported his results in 63 patients who underwent thoracoscopic sympathectomy for hyperhidrosis. Although his results were excellent, his work went relatively unnoticed. In the 1970s and 1980s, only a few investigators pursued this less invasive technique. 3-7 Recently, advances in endoscopic surgery and patient demand for less invasive procedures have 511
512
JOURNAL OF VASCULAR SURGERY October 1994
A h n et al.
/
Fig. 1. Patient position. Note lateral decubitis position, arm elevation, and hyperextended anterior chest wall.
fueled an explosion of minimally invasive endoscopic approaches to many procedures, including cervicodorsal sympathectomy. 819 In most of these approaches, the surgeon ablated rather than resected the ganglia by various means such as thermocoagulation or chemical exposure. Although long-term results have been reported infrequently, recurrence rates have ranged from 2% to 50%. We reasoned that thoracoscopic resection of the sympathetic ganglia would have several advantages. It would be minimally invasive and thus minimize morbidity rates and length of hospital stay. At the same time, it should be effective and give good long-term results because the ganglia are disconnected and resected, thereby ensuring complete ablation. Accordingly, patients and physicians would readily accept this procedure to treat sympathetic hyperactivity. Furthermore, because of the decreased hospital stay, the health care cost would decrease. Thus in 1990 we embarked on a clinical trial of thoracoscopic cervicodorsal sympathectomy, whereby the sympathetic ganglia and chain are dissected and resected with endoscopic visualization through small thoracotomy incisions. The objectives of our study were to determine the feasibility, safety,
and efficacy of this procedure. We report herein our initial experience. MATERIAL A N D METHODS This study was performed between March 1990 and December 1993. We enrolled 19 patients, 13 women and 6 men, who underwent 21 procedures: 13 on the left side and 8 on the right. Their age ranged from 17 to 64 years, mean 37 years, median 36 years. Six patients underwent standard open cervicodorsal sympathectomies during the study period (three underwent supraclavicular procedures during a scalenectomy and three underwent transaxillary transthoracic procedures during resection of the first rib) and were excluded from this study. No patient during this study period was excluded because of poor pulmonary function, previous thoracotomy, previous rib resection, previous stellate ganglion blocks or ablation, or previous pulmonary sepsis. In fact, one of the patients in this study had undergone a previous thoracotomy; six patients had undergone a previous first rib transaxillary resection, and six patients had undergone previous stellate ganglion blocking. Indications for thoracoscopic sympathectomy
JOURNAL OF VASCULXR SURGERY Volume 20, Number 4
were reflex sympathetic dystrophy/causalgia in nine patients, Raynaud's syndrome/vasculitis in six patients, hyperhidrosis in five patients, and medically refractory cardiac arrhythmia caused by a prolonged QT syndrome in one patient. Before the study, we determined the proper equipment requirement and port placement by practicing the procedure on three fresh human cadavers. Furthermore, the first three patients in the study underwent the thoracoscopic procedure with the chest partially open in preparation for any surgical mishaps. No technical misadventures occurred in these three cases, and the subsequent 18 thoracoscopic cases were treated with a close technique as described below. All patients underwent the procedure under general endotracheal anesthesia with a double lumen endotracheal tube. The ipsilateral lung was deflated and nonventilated throughout the procedure. The patient was placed in a lateral decubitis position with the ipsilateral arm abducted on a mechanical arm holder. The surgical table was broken to hyperextend the ipsilateral chest to spread the intercostal spaces (Fig. 1). The patient was in position to undergo an open thoracotomy if necessary. All ports were inserted by making a 1 cm skin incision followed by blunt dissection with a curved clamp through the muscles and pleura. All patients had at least three ports (Fig. 2). Port placements were as follows. The scope port went through the fifth or sixth intercostal space in the midaxillary line. An instrument port went in this same intercostal space but in the posterior axillary line. A second instrmnent port was placed in the third or fourth intercostal space in the anterior axillary line. Two patients required an additional instrumentation port to retract the lung; this port was placed in the fifth intercostal space along the anterior axillary line. Instruments that were used include the following: video/monitor equipment, 8 to 10 mm diameter rigid laparoscope with 0- or 30-degree angled lens, a 5 mm diameter suction/irrigation/electrocautery hook apparatus, a 5 mm diameter, blunt-tipped scissors with electrocautery attachment, a 10 mm diameter curved grasper, a 10 mm diameter right angled clamp, a 10 mm diameter laparoscopic clip applier, and, if needed, a 10 mm diameter laparoscopic fhnshaped retractor. The steps of the procedure were as follows. (1) Insert scope and evaluate the intrathoracic anatomy. (2) Insert two instrument ports as described above with scope assistance. (3) Take down adhesions if necessary (this was required in four cases). (4)
A h n et al.
513
Collapse and deflate the ipsilateral lung mechanically. The lung is simply compressed by blunt instruments with the endotracheal tube opened to that lung. (5) Identify the ribs with its intercostal nerves and follow these ribs to the vertebral body. (6) Dorsal to the vertebral body, identify the sympathetic ganglia and chain that will appear as a white, glistening, raised, longitudinal structure. (7) Identify the subclavian artery, which marks the superior aspect of the sympathetic dissection. Also, identify the azygous and subclavian veins and the highest intercostal artery and veins. These structures should be avoided during the subsequent dissection. (8) Incise the pleura over the sympathetic chain. (9) Dissect free and lift the chain from its bed. (10) Clip and then sharply cut the rami communicantes of sympathetic ganglia T2-4. (11) Clip and divide just below the T4 ganglion. (12) Dissect free, clip, and sharply divide the lower third of the stellate ganglion along with the rami communicante coursing caudally from the stellate ganglion. The rami coursing in a rostral direction should be left intact, and the clip should be applied caudal to these rostral-directed rami. (13) Send the ganglion to pathology for frozen section and confirm the tissue histologically. (14) Irrigate and control any bleeding points with electrocautery or clips. (15) Remove the instrument ports and inspect the intercostal defects for any bleeding points. (16) Insert the scope through the anterior instrument port and inspect the thoracoscope port for any bleeding points. (17 I) Insert a small chest tube through the thoracoscope port and position the tube under direct visualization. (18) Reinflate the lung using 30 to 40 mm Hg positive pressure and confirm adequate reinflafion with scope visualization. (19) Close the incision in two layers using absorbable sutures. After operation, the chest tube was placed to waterseal and negative 15 cm water suction. A chest radiograph was obtained in the recovery room to check for any pneumothorax or pleural effusions. The chest tube was removed in the recovery room or by the next day if there was no air leak. If there was an air leak, which occurred in one patient, then the chest tube was left in place until the air leak resolved. Oral analgesia was adequate in all patients. All patients were also given an incentive spirometer to prevent atelectasis. RESULTS
Technical success was achieved in 20 of 21 cases (95%). In 19 cases, T1-4 ganglia were dissected and resected as described above. In one case, the T3-7 sympathetic ganglia were resected because the patient
JOURNAL OF VASCULAR SURGERY October 1994
514 A h n et al.
i
Fig. 2. Illustration of chest wall. Note 2 instrument ports in anterior and posterior axillaryline, and scope port in midaxillary line.
had a previous open supraclavicular resection of T1-2. This patient required resection of the lower ganglia because of causalgia-like pain of the fourth, fifth, and sixth intercostal nerve distribution caused by herpes zoster. We had a technical failure in one patient. We were unable to adequately visualize the sympathetic chain because of dense adhesions from a previous first rib resection, which had been complicated by pneumothorax and hemothorax. This case was converted to an open procedure for safety reasons. This case also occurred early in our series when our confidence level was stiU on the rise. Now, we probably would proceed with a closed procedure in this patient. The operative duration was 1.0 to 3.5 hours, mean 2.0 hours. Estimated blood loss was 5 to 300 cc, mean 42 cc, median 10 cc. No patient required transfusion. All patients had histologic confirmation of ganglia. After operation, all patients demonstrated a satisfactory clinical sympathectomy response manifested by a warm, dry hand and upper extremity. Two patients had a transient Homer's syndrome, which resolved within 1 week. Three patients showed a residual pneumothorax but did not require further treatment. In one patient a pneumothorax and a pleural effusion developed, which were treated with the reinsertion of the chest tube. Transient intercostal
neuralgia that resolved within weeks developed in two patients. The patients stayed in the hospital 1 to 4 days after operation, mean 2 days, median 1 day. All patients, except one, were also admitted on the same day as the surgery. All patients had a complete follow-up with a direct examination and report by a physician. Followup ranged from 1 to 42 months, mean 11 months, median 6 months. According to indications for the surgery, follow-up were as follows: reflex sympathetic dystrophy, 2 to 18 months; Raynaud's/vasculitis, 1 to 41 months; hyperhidrosis, 1 to 15 months; heart related, 42 months. All patients demonstrated a clinical benefit as follows. The patients with reflex sympathetic dystrophy all showed improvement of their symptoms and required less or no pain medication. All six patients with Raynaud's/vasculitis had improvement of their ulcer-healing or improvement of their gangrene, as well as relief of pain. All five patients with hyperhidrosis were satisfied and had resolution of their hyperhidrosis symptoms. The one patient, who underwent the procedure for a medica]ly refractory cardiac arrhythmias caused by prolonged QT syndrome, demonstrated shortening of the QT interval on electrocardiography and required significandy less heart medication. There were no intraoperative deaths, and only one patient died during the follow-up. This patient
JOURNAL OF VASCULAR SURGERY Volume 20, Number 4
died of a myocardial infarction caused by a severe calcinosis of the coronary arteries. DISCUSSION
This study demonstrates that thoracoscopic cervicodorsal sympathectomy is feasible, safe, and effective. We achieved technical success in 20 of 21 (95%) patients. Complications were few, and most were transient and minor. Hospital stay was relatively short, often comprising only 1 day. Oral analgesia was adequate in all patients. Furthermore, patient acceptance of the procedure was uniformly excellent. Although we did not study a comparable patient group that underwent open procedures, we befieve from our previous experience that the thoracoscopic approach is preferable. Open thoracotomy and sympathectomy have significant morbidity, such as postoperative bleeding, wound infection, chest pain, pneumothorax, Homer's syndrome, pleura] and dural tears, thoracic duct and nerve injury, and intercostal neuralgia. 14-16 However, because our follow-up is relatively short and our experience is small, we cannot recommend broad application of this procedure at this time. Furthermore, our techniques are still being refined and other techniques that have been described need to be considered as well. The most intriguing technique was recently reported by Claes et al.14,1s and Drott et al.16 These investigators have performed more than 700 thoracoscopic sympathectomies using an urologic electroresectoscope through a single puncture in the second intercostal space midclavicular fine to electtocoagulate ganglia T2-3. Using this technique, they have performed bilateral thoracoscopic sympathectomy in 30 minutes with a 99% success rate and a 2% recurrence rate in patients with hyperhidrosis. Their work clearly deserves confirmation by other investigators. The issue of whether the ganglia should be ablated versus resected still needs to be determined by further studies. Other investigators who have relied on thermal ablation have not had such excellent late results and in fact, most investigators have not reported late results at all. 4"13,17"19 Poor results have been attributed to inadequate ablation or regrowth of the nerve tissue. Furthermore, intercostal neuralgia has been reported in patients treated by electroablation, suggesting possible thermal damage to the adjacent intercostal nerves. Few investigators have actually performed a nonelectrocoagulation resection of the ganglia, as we reported in our present study. It is possible that resection of the ganglia with minimal electrocoagulation may prevent postoperative inter-
A h n et al.
515
costal neuralgia, as well as recurrence of sympathetic: activity. The extent of sympathectomy also needs to be addressed. Some investigators have reported that di.. vision or ablation of T-2 ganglion is adequate, whereas others advocate ablation of T2-3, and still[ others advocate ablation of T1-4.131 Interestingly:. Claes et al.14,~s reported excellent results in patients who underwent ablation of T2-3 for hyperhidrosis but poor results for patients with Raynaud's. They also found that T2-3 ablation provides excellent relief ofpalmar hyperhidrosis, but that resection ofT-4 was necessary to block axillary hyperhidrosis. In our se-. ries, we uniformly resected T1-4 ganglia and ob-. tained excellent results in all patients with Ray-. naud's/vasculitis and causalgia/reflex sympathetic: dystrophy, as welllas those with hyperhidrosis. Until[ the required level of ganglia resection is better de-. fined, we recommend resection ofT1-4 as advocated[ byKux2 and supported by our study. The issues of Homer's syndrome and postoper-. ative intercostal neuralgia also deserve f m ~ e r con-. sideration. We were able to avoid permanent Horner's syndrome in all 21 of our cases, despite the fact that we resected the lower third of the stellate ganglion. We attribute this to our ability to accurately' visualize and identify the multiple rami of the stellate ganglion and clearly visualized the direction of their course. Because the endoscope greatly magnifies the surgical field, such visualization is readily available. Hence, we can divide the stellate ganglion precisely' between the fibers that course rostrally and the fibers; that course caudally. Our results suggest that the fibers that course rostrally innervate the ocular pupillary muscles, and that the resection of the rami coursing caudally does not cause a Homer's syndrome. If this is true and confirmed, then thoracoscopic visualization during sympathectomy is clearly' advantageous. The cause of transient Homer's syndrome in our two patients is not clear; we attribute it to the traction that we placed on the stellate ganglion during its dissection. The postoperative intercostal neuralgia seen in two of our patients occurred early in our series when we used rigid 10 to 12 mm diameter thoracoports. Later in our series, we switched to soft rubber thoracoports; and since that time, we have not encountered this side effect. This suggests that the thoracoports applied direct pressure on the intercostal nerves, thereby causing neuropraxia and hypersensitivity of these nerves. Other possible explanations for the neuralgia include electtocautery injury or inadvertent clipping of the intercostal nerves.
516
A h n et al.
The one patient who underwent sympathectomy for a medically refractory cardiac arrhythmia deserves further discussion. This patient had an exceUent response to surgery, and there are numerous reports in the literature that have documented the therapeutic benefits of sympathectomy for cardiac arrhythnlias and even angina. 2°-31 Despite these reports, sympathectomy has not been widely advocated for heart disease because of the significant morbidity and mortality rates associated with this procedure in heart patients at high risk. However, the less invasive thoracoscopic approach with its minimal morbidity and mortality rates could be better accepted and tolerated. These patients deserve further consideration for thoracoscopic sympathectomy. Further studies in this regard are clearly indicated. We conclude that thoracoscopic sympathectomy is feasible, safe, and effective. It appears to be preferable to open sympathectomy. However, our patient follow-up is short, and there are numerous other thoracoscopic sympathectomy techniques described in the literature. Furthermore, this study does not adequately address the larger issue of the specific value of sympathectomy in individual patients given the small number of patients within each subgroup. Accordingly, further studies are needed to confirm its long-term benefits and to establish the optimal thoracoscopic techniques. REFERENCES
1. Kux E. The endoscopic approach to the vegetative nervous system and its therapeutic possibihties. Dis Chest 1951;20: 139-47. 2. Kux M. Thoracic endoscopic sympathectomy in pahnar and axillary hyperhidrosis. Arch Surg 1978;113:264-6. 3. Fritsch A, Kokoschka R, Mach K. Ergebnisse der thorakoskopischer sympathectomie bei hyperhidrosis der oberen Extremitat. Wien Klin Wochenschr 1975;87:548-50. 4. Weale FE. Upper thoracic sympathectomy by transthoracic electrocoagulation. Br J Surg 1980;67:71-2. 5. Malone PS, Duignan )'P, Hederman WP. Transthoracic electrocoagulation (TTEC): a new and simple approach to upper hmb sympathectomy. Ir Med J 1982;75:20-1. 6. Horgan K, O'Flanagan S, Duignan PJ, Hederman W. Pahnar and axillary hyperhidrosis treated with sympathectomy by transthoracic endoscopic electrocoagulation [Abstract]. Br J Surg 1984;71:1002. 7. Malone PS, Cameron AEP, Rennie JA. The surgical treatment of upper limb hyperhidrosis. Br J Dermatol 1986;115:81-4. 8. Banerjee AK, Edmonson R, Rennie JA. Endoscopic transthoracic electrocautery of the sympathetic chain for palmar and axillary hyperhidrosis. Br J Surg 1990;77:1435-6. 9. Lin CC. A new method of thoracoscopic sympathectomy in hyperhidrosis palmaris. Surg Endosc 1990;4:224-6.
JOURNAL OF VASCULARSURGERY October 1994
10. Byrne J, WalshTN, HedermanWP. Endoscopictransthoracic electrocautery of the sympathetic chain for palmar and axillary hypethidrosis. Br J Surg 1990;77:1046-9. 11. Adams DCR, Poskitt KR. Surgical management of primary hypethidrosis. Br J Surg 1991;78:1019-20. 12. Norback B, Svartholm E. Endoskopisk torakal sympatektomi enkel och effektiv metod mot handsvett [Abstract]. Lakartidningen 1992;89:1478. 13. Edmonson RA, Banerjee AK, Rennie JA. Endoscopic transthoracic sympathectomy in the treatment of hypethidrosis. Ann Surg 1992;215:289-93. 14. Claes G, Gothberg G, Drott C. Endoscopic electrocautery of the thoracic sympathetic chain: a minimally invasive method to treat palmar hypethidrosis. Scand J Hast Reconstr Surg Hand Surg 1993;27:29-33. 15. Claes G, Gothberg G. Endoscopic transthoracic electrocautery of the sympathetic chain for palmar and axiUaryhyperhidrosis [Letter]. Br J Surg 1991;78:760. 16. Drott C, Gothberg G, Claes G. Endoscopic procedures of the upper thoracic sympathetic chain: a review. Arch Surg 1993; 128:237-41. 17. Chuang KS, Liou NH, Liu JC. New stereotactic technique for percutaneous thermocoagulation of upper thoracic ganghonectomy in cases of palmar hyperhidrosis. Neurosurgery 1988;22:600-4. 18. Adler OB, Engel A, Rosenberger A, Dondehnger R. Palmar hyperhidrosis CT guided chemical percutaneous thoracic sympathectomy. Fortschr Rontgenstr. 1990;153:400-3. 19. Wilkinsson HA. Percutaneous radiofrequency upper thoracic sympathectomy: a new technique. Neurosurgery 1984;15: 811-4. 20. Crampton R. Preeminence of the left stellate ganglion in the long Q-T syndrome. Circulation 1979;59:769-78. 21. Schwartz H, Snebold NG, Brown AM. Effects of unilateral cardiac sympathetic denervation on the ventricular fibrillation threshold. Am J Cardiol 1976;37:1034-40. 22. Zaza A, Malfatto G, Schwartz P[L Sympathetic modulation of the relation between ventricular repolarization and cycle length. Circ Res 1991;68:1191-203. 23. Austoni P, Rosati R, Gregorini L, Bianchi E, Bortolani E, Schwartz PJ. SteUectomy and exercise in man [Abstract]. Am J Cardiol 1979;43:399. 24. Schwartz PJ, Stone HL. Left steUectomy in the prevention of ventricfllar fibrillation caused by acute myocardial ischaemia in conscious dogs with anterior myocardial infarction. Circulation 1980;62:1256-65. 25. White JC, White PD. Angina pectoris: treatment with paravertebral alcohol injections. JAMA 1990;28:1099-1103. 26. Lindgren I. Angina pectoris: a clinical study with special reference to neurosurgical treatment. Acta Med Scand 1950(suppl):1-141. 27. Papa MZ, Schneiderman J, Tucker E, Bass A, Drori Y, Adar R. Cardiovas~lar changes after bilateral upper dorsal sympathectomy. Ann Surg 1986;204:715-8. 28. Blomberg S, Curelaru I, Emanuelsson H, Herhtz J, Ponten J, Ricksten SE. Thoracic epidural anaesthesia in patients with unstable angina pectoris. Eur Heart J 1989;10:437-44. 29. Blomberg S, Emannelsson H, Ricksten SE. Thoracic epidural anaesthesia and central hemodynamics in patients with unstable angina pectoris. Anesth Analg 1989;69: 558-62. 30. Schwartz H, Locati EH, Moss AJ, Crampton RS, Trazzi R, Ruberti U. Left cardiac sympathetic denervation in the
JOURNAL OF VASCULARSURGERY Volume 20, Number 4
therapy of congenital long QT syndrome: a worldwide report. Circulation 1991;84:503-11. 31. Moss AJ, McDonald J. Unilateral cervicothoracic sympathetic ganglionectomy for the treatment of long QT interval syndrome. N Engl J Med 1971;285:903-4.
A h n et al.
517
Submitted Feb. 15, 1994; accepted June 18, 1994.
DISCUSSION Dr. T. J. Bunt (Loma Linda, Calif.). The basic premise to surgery's newest rage, "minimally invasive surgery," is reflected in its very designation- the expectation that it will allow performance of a surgical procedure with less pain and disability for the patient. However, in the flush of being pioneers in new surgical technology, many noninvasive surgeons seem to forget that there is more to the health care delivery equation than the concept that the patient and his surgeon believe that the new technology is better. Appropriate complete evaluation of newly introduced procedures needs also to include fair analyses of the economic impact, including the efficient use of resources, the cost of the technology not only per patient or per procedure, but to the hospital, the costs of future technology upgrades, the costs of retraining personnel, and so on. Before we abandon a tried-and-true standard surgical approach, we should be certain that the new method is significantly better in all respects. Just showing that it is feasible is not adequate justification for either prosetylization nor promulgation. The overriding question that therefore needs to be addressed by the authors is not can we do a cervical sympathectomy via thoracoscopy, but rather do we need to? The other major philosophic problem with "less invasive" procedures is the tendency for surgeons to associate a reduced complication rate (whether proven or not) with the nomenclature of "minimally invasive," and to then extend the indications for the procedure to patients with less symptoms on the possibly false premise that the risk is less. This has been the case for endovascular surgery; will it now be the same for thoracoscopy? The authors made no direct comparisons to standard procedures at their institution. Their thoracoscopic approach requires the same anesthetics and same perioperatire care, but a much more expensive technology, and may result in fewer hospital days. Was the cost saving on days used worth the capital investment for technology? What were the comparative costs? The morbidity rate associated with this procedure seems low; how did it compare with standard operative approaches at your institution? Was the perception of low risk a factor in patient selection, because more than half of the patients were treated for clinical situations that are not universally accepted as compelling reasons for surgical sympathectomy-such as hyperhidrosis and Raynaud's disease. On the other hand, you have presented initial data
that suggest that stellate ganglion resection can be redone in a more refined fashion because of better visualization of the rami, thereby avoiding Homer's s y n d r o m e - certainly a positive attribute of the procedure. In summary, the authors have presented an initial series of thoracoscopic cervical sympathectomy, showing that it can indeed be technically performed. Their short-term results are good, although follow-up is short. They are to be congratulated for their pioneering effort. They and others must demonstrate that it's better than what we already have, and that the cost of the technology is justified. Dr. Herbert I. Machleder. Clearly we haven't demonstrated that we need to do a sympathectomy this way, but I think it is important to recognize that we can and that we can do it safely. In terms of the complications, in our own institution 20 years ago Dr. Barker and I published our series of thoracocervical sympathectomy for upper extremity disorders, and we reported 53 consecutive cases. Twenty-nine were transcervical procedures and 24 were transthoracic procedures. We were trying to decide which was more effective and which approach had less complications, and they were entirely comparable. The complications are no different than the ones we had with thoracoscopic sympathectomy. Occasionally a pneumothorax requires a tube and occasional retropleural hematoma. So clearly, the complication rate isn't reduced by the less invasive technique. On the other hand, the incidence of Homer's syndrome is less, probably because we don't dissect the stellate ganglion quite as far as one would in a transcervical approach. What about the cost? We purchased no new equipment for this procedure. The scopes are used routinely for laparoscopy and thoracoscopy. The video monitors are exactly the same. The thoracoports are all the same. It requires a minimal amount of disposable equipment, and it should be remarked that some of this can be done with cystoscopes or ureteroscopes that are available and resterilizable. So I think the cost is relatively comparable in terms of operating time and equipment. The real saving is that this can probably be done on an outpatient basis or with a 1-day operation, whereas a transthoracic approach really can't:. I think Dr. Bunt is correct about increased utilization,, and we've seen this with laparoscopic cholecystectomy. There's been no saving with doing this with laparoscopy
518
A h n et al.
because so many more cholecystectomies are now done for lesser indications, regardless of whether its perceptional. People are more apt to have a minimally symptomatic gallbladder removed when it can be done with laparoscopy. So there has been a change in indications. Our original procedures were done for patients with digital ulcerations and severe vasculitis, and that represents a very small percentage here. About 20% of our cases were done for hyperhidrosis. The other group has causalgia and reflex sympathetic dystrophy. These patients are generally referred from pain management clinics. The medical therapy is repetitive stellate ganglion blocks. Some of these patients have had them done for 1 and 2 years, usually every 2 weeks to a month: Being able to do a transthoracic sympathectomy through a thoracoscope lowers the threshold for when these patients are referred. There's no question about it. The results are dramatic. I mean, if they get a good result from a stellate ganglion block, they're going to be cured with sympathectomy. This procedure is dramatically different than open. The patients go home early and they go to work in about 3 to 4 days. I think the maximum recovery time was 10 days, and this is dramatically different than the days when we did it transthoracically. Dr. John M. Porter (Portland, Ore.). I believe you should withdraw your conclusion that this technique is superior or preferable to operative sympathectomy because these patients were not randomized. You simply cannot conclude this from historic controls. In the days when we used to do sympathectomy, we used a 3-inch transaxiUary incision. The operation took 45 minutes, and the patients went home the next day. I wonder if you have a technique in search of an indication. We have 1500 patients with Raynaud's syndrome at Oregon. We have 200 plus patients with ischemic finger ulcerations including all sorts ofvascaflitis, and we do a large number of upper extremity arterial bypasses including hand bypasses. We have not performed a thoracic sympathectomy of any sort in the past 4 years. I do not think there's any indication for this. Your concern about reflex sympathetic dystrophy requires some elaboration. We have an entire research unit in our department of neurology devoted to reflex sympathetic dystrophy. They find reflex sympathetic dystrophy indeed to be one of the most overdiagnosed and underappreciated conditions that we deal with. You can't rely on stellate ganglion blocks at all unless you do placebo control injections. I would like to be reassured that the pain clinics that send you all these patients are doing placebo control injections without informing the patients and making sure they have the proper response. Finally, I strongly disagree with your technique of sympathectomy. I feel there is no indication whatsoever to transect the lower one third of the stellate ganglion, and if you do so, you're going to invite Homer's syndrome. Dr. Machleder. On the open sympathectomies Hor-
JOURNAL OF VASCULAR SURGERY October 1994
ner's syndrome routinely occurs. So all of the patients I treat with transcervical sympathectomy or with first rib have Homer's syndrome, fortuantely this can be corrected by an outpatient procedure. If the patient uses two drops of Visine in the eye twice a day, the Homer's syndrome is completely reversed. If they do not want to use Visine eyedrops, which has about the same sympathomimetic effect as the normal ocular pupiUary nerve, they can undergo a minor operation under local anesthetic. It's called the reefing of Mueller's muscle, and it's done by everting the lid and taking a little tuck in that muscle. So I don't think the Homer's syndrome is a problem. In terms of reflex sympathetic dystrophy, it's a dramatic problem for patients who have it. Usually it's diagnosed by radioisotope studies, as well as the typical atrophy and the bony changes ofhyperostosis that you see with the chronic causalgia. Some of the patients have had placebo blocks, it depends on the pain management clinic, and some of them just have repetitive stellate ganglion blocks. If you have good medical therapy for it, then your clinic is going to be very busy. But patients who have true causalgia and get dramatic results from sympathetic block will have dramatic results from sympathectomy. We see very few patients with vasculitis or Raynand's syndrome who need a sympathectomy, and so we find that there's very few indications now compared with our previous series. But there are still individuals with very painful ulcers who will heal with sympathectomy. Without having the benefit ofintraarterial reserpine and some of the other effective medical therapy we had in the past, sympathectomy occasionally is very useful for-digital gangrene and digital ischemia. So I think those are the major indications. Reflex sympathetic dystrophy is responsive to stellate ganglion block. With hyperhidrosis, the response is dramatic. Dr. D. Eugene Strandness (Seattle, Wash.). One of the major problems that always has plagued this area is permanent denervation. I don't see that your operation has done anything more to ensure permanent denervation. Yet as I look at your results, you essentially said that you obtained permanent denervation in all cases. The recurrence rate has been reported as high as 75% after the standard methods have been used. It has been stated that you can't assure complete denervation unless you take the whole stellate, which you obviously are not doing here. The other point I wish you'd comment on are that the restflts of this operation vary greatly with the indications. I agree with John Porter in this with regard to the area of Raynaud's vasculitis. The problem of hyperhidrosis is clearly one in which you ought to be able to get a cure. However, it has been my impression that people have generally been able to do this operation through the neck with very satisfactory results. Dr. Machleder. One thing I failed to mention is that
JOURNAL OF VASCULAR SURGERY Volume 20, Number 4
I don't claim this is any better than doing a transcervical or transthoracic procedure. It's another approach, and I think once you do it through a thoracoscope you probably will try to do it that way every time. But it's no better, and the complications are no different, but the return to work is quite different. I think it's very comparable to gallbladder surgery through a thoracoscope or doing it open. You don't get permanent denervation in that after a while, after three or four months, there's regeneration of the vacuoles at the end of the post ganglionic sympathetic nerve that show that. They again have no epinephrine so that I think you do get sensitization, particularly in those patients who have vasospastic diseases, and it's only a temporizing measure to treating underlying collagen vascular disorder. Clearly, the best results occur in patients with causalgia who are treated early and patients who have hyperhidrosis. The patients with hyperhidrosis are cured and they don't come back with wet hands again. The patients with causalgia generally do very well. But I agree with you that in patients who have vasospastic disorders, particularly those who have collagen vascular diseases, eventually the vasal occlusive problem is going to recreate the problem, and the sympathectomy is probably only good in a transient period of time when there's vasospasm superimposed on the early vasal occlusive process with a collagen vascular disease. The indications have changed over the years, and the thoracoscopic sympathectomy still has the same encumbrances as the transthoracic procedure did, except it is a lesser invasive procedure. Dr. Jerry Goldstone (San Francisco, Calif.). I think that thoracoscopic sympathectomy is better. We've done just a few of these, and, at least for those of us who had done them all supraclavicularly, doing them with thoracoscopy is much simpler. You can see the ganglia much better. I don't think there's any question that it's a better procedure. In our experience it's taken longer to set up the equipment than to do the procedure. Regarding the issue about all this equipment; I think it's time for vascular surgeons to begin to look at this kind of equipment. There may be vascular applications that none of us have thought of or at least none of us have used. It may be time for us to get involved. And, in fact, one of my interests in doing these few procedures was because it was the only thing I could think of that would allow me to get some hands-on experience using this technology for which I think there probably are other applications in our field. Do you think it really is necessary to take even the bottom third of the stellate ganglion? In those few cases that we have done, we have not done that and have had very good clinical results. Now, admittedly the follow-up is fairly short, along the order of what Dr. Ahn has shown, but maybe it's not necessary to take it at all. Dr. Machleder. There are a lot of people who would agree with that, and it may well be true that just taking T-2 and T-3 is adequate. I insisted on taking the lower end of
A h n et al.
519'
the steUate ganglion because that's the way we had done it: in the past, and a number of people had pointed out that T-2 and T-3 may be perfectly adequate• The problem that we encountered was that there's a tremendous experience in the European literature, in the Scandinavian literature, in the English-language literature, and the German literature doing these types of procedures thoracoscopicaUy, and they have tremendous numbers of' cases.
What bothered us was the recurrence rate, which was much higher than the ones we had looked at in the past where we excised the ganglion. All of these sympathetic chains were destroyed either by electric cautery or by radio frequency. So we believed that we ought to try to do it comparably to the way we had done it in the past by excising the ganglion. That's why we took the lower part of the stellate. Dr. N o r m a n L. Browse (London, United Kingdom). The discussion is becoming confused between the indications for sympathectomy and the technique. We may practice that the indication for cervical sympathectomy is, in fact, palmar hyperhidrosis and almost nothing else. The advantage claimed by those who practice the thorascopic coagulation technique is that it is easy to do. You simply place the diathermy electrode over the ganglia, and the one thing you cannot do without dissection is get to the stellate ganglion. That is considered a positive point because first you do not need to touch the stellate ganglia to dry up the hands. You only need to destroy the second and third ganglia, and you avoid Horner's syndrome. Because hyperhidrosis is a relatively benign condition, I would be happy to accept a moderate recurrent rate. The procedure can easily be repeated. But, as far as hyperhidrosis is concerned, there is not a high recurrence rate. We talk about reinnervation, but that belief is based on the recurrence of the symptoms of Raynaud's phenomenon. The evidence of reinnervation in terms of the recurrence of sweating after sympathectomy is almost zero. It will happen occasionally if you do not coagulate adequately, but it's a very quick simple safe procedure and can be repeated. However, it is not entirely without risk. It is far simpler and quicker just to put in a thoracoscope. Dr. Machleder. In your clinics where the sympathectomies have been done either transthoracic and transcervical, where transthoracic thoracoscopic sympathectomy has been introduced, has that been the procedure of choice? Dr. Browse. Yes. The thoracoscopes were already available because the chest surgeons and physicians use them. Most times the thoracic surgeon taught the vascular surgeon how to use the thoracoscope. So it has not involved additional expense. Most vascular surgeons in the United Kingdom have converted to coagulation sympathectomy for hyperhidrosis because it is a very simple, quick, effective procedure• If you want to excise the axillary skin at the same time, you can do that first and use the same incision for the thoracoscopy, thus avoiding twO incisions.
Purpose: The purpose of this study was to determine the feasibility, safety, and efficacy of thoracoscopic cervicodorsal sympathectomy. Methods: From March 1990 to December 1993, we performed 21 thoracoscopic sympathectomies in 19 patients. There were 13 women and six men; age 17 to 64 years, mean 37 years. Thirteen procedures were performed on the left side and eight on the right. Indications for surgery were cansalgia/reflex sympathetic dystrophy in nine patients, Raynaud's/vasculitis in six, hyperhidrosis in five, and medically refractory cardiac arrhythmia in one. Results: The T1-4 sympathetic ganglia were readily identified, dissected free, and resected thoracoscopically in 19 cases, and the T3-7 ganglia were resected with thoracoscopy in one case. One case required conversion to an open thoracotomy because of dense scar from a previous first rib resection, which obscured the anatomy. Histologic confirmation of ganglia were obtained in all 21 cases. Operative duration ranged from 1.0 to 3.5 hours. Estimated blood loss was 5 to 300 co, mean 42 cc, median 10 co. No patient required transfusion. All 21 patients had an excellent immediate sympathectomy response. Transient Homer's syndrome developed in two patients. Postoperative residual pneumothorax (< 10%) occurred in three cases and resolved spontaneously without further treatment. In one patient plettral effusion and pneumothorax developed, which were treated with the reinserdon of the chest tube. Postoperative pain was well controlled with oral analgesics. Hospital stay was 1 to 4 days, mean 2 days, median 1 day. Follow-up at 1 to 42 months, mean 11 months, median 6 months, showed continued evidence of sympathectomy effect in all patients, except one who died of her underlying disease 1 month after operation. Conclusions: We conclude that thoracoscopic sympathectomy is feasible, safe, and effective. Further studies are indicated to confirm its long-term benefits and to determine optimal thoracoscopic techniques. (J VAse SURG 1994;20:511-9)
Numerous methods of cervicothoracic sympathectomy have been described in the literature, which attests to the fact that an optimal method of sympathectomy is still lacking. The supraclavicular approach is well tolerated by the patient but carries significant risk of complications such as Homer's syndrome, phrenic nerve injury, brachial plexus injury, chylous leak, pneumothorax, and bleeding. From the UCLACenterfor the Health Sciences,Los Angeles,and Olive View Medical Center, Sylmar. This study was fimdedin part by the JOASH Foundation, Santa Monica, CA. Presented at the Ninth AnnualMeeting of the WesternVascular Society, Santa Barbara, Calif.,Jan. 9-13, 1994. Reprint requests: SamuelS. Ahn, MD, 100 UCLAMedicalPlaza, Suite 510, Los Angeles, CA 90024 Copyright © 1994 by The Society for Vascular Surgery and InternationalSocietyfor CardiovascularSurgery,North American Chapter. 0741-5214/94/$3.00 + 0 24/6/58563
These complications are less likely to occur with the transthoracic approaches, but an open thoracotomy causes greater morbidity and mortality rates. The posterior approach in particular, which transgresses the posterior shoulder girdle muscles and requires partial rib resection, causes significant tissue trauma and postoperative pain. Not surprisingly, physicians and patients have not embraced these procedures except for very selected situations. In 1951, Kux ~ described a minimally invasive thoracoscopic approach to perform cervicodorsal sympathectomy. In 1978, 2 he reported his results in 63 patients who underwent thoracoscopic sympathectomy for hyperhidrosis. Although his results were excellent, his work went relatively unnoticed. In the 1970s and 1980s, only a few investigators pursued this less invasive technique. 3-7 Recently, advances in endoscopic surgery and patient demand for less invasive procedures have 511
512
JOURNAL OF VASCULAR SURGERY October 1994
A h n et al.
/
Fig. 1. Patient position. Note lateral decubitis position, arm elevation, and hyperextended anterior chest wall.
fueled an explosion of minimally invasive endoscopic approaches to many procedures, including cervicodorsal sympathectomy. 819 In most of these approaches, the surgeon ablated rather than resected the ganglia by various means such as thermocoagulation or chemical exposure. Although long-term results have been reported infrequently, recurrence rates have ranged from 2% to 50%. We reasoned that thoracoscopic resection of the sympathetic ganglia would have several advantages. It would be minimally invasive and thus minimize morbidity rates and length of hospital stay. At the same time, it should be effective and give good long-term results because the ganglia are disconnected and resected, thereby ensuring complete ablation. Accordingly, patients and physicians would readily accept this procedure to treat sympathetic hyperactivity. Furthermore, because of the decreased hospital stay, the health care cost would decrease. Thus in 1990 we embarked on a clinical trial of thoracoscopic cervicodorsal sympathectomy, whereby the sympathetic ganglia and chain are dissected and resected with endoscopic visualization through small thoracotomy incisions. The objectives of our study were to determine the feasibility, safety,
and efficacy of this procedure. We report herein our initial experience. MATERIAL A N D METHODS This study was performed between March 1990 and December 1993. We enrolled 19 patients, 13 women and 6 men, who underwent 21 procedures: 13 on the left side and 8 on the right. Their age ranged from 17 to 64 years, mean 37 years, median 36 years. Six patients underwent standard open cervicodorsal sympathectomies during the study period (three underwent supraclavicular procedures during a scalenectomy and three underwent transaxillary transthoracic procedures during resection of the first rib) and were excluded from this study. No patient during this study period was excluded because of poor pulmonary function, previous thoracotomy, previous rib resection, previous stellate ganglion blocks or ablation, or previous pulmonary sepsis. In fact, one of the patients in this study had undergone a previous thoracotomy; six patients had undergone a previous first rib transaxillary resection, and six patients had undergone previous stellate ganglion blocking. Indications for thoracoscopic sympathectomy
JOURNAL OF VASCULXR SURGERY Volume 20, Number 4
were reflex sympathetic dystrophy/causalgia in nine patients, Raynaud's syndrome/vasculitis in six patients, hyperhidrosis in five patients, and medically refractory cardiac arrhythmia caused by a prolonged QT syndrome in one patient. Before the study, we determined the proper equipment requirement and port placement by practicing the procedure on three fresh human cadavers. Furthermore, the first three patients in the study underwent the thoracoscopic procedure with the chest partially open in preparation for any surgical mishaps. No technical misadventures occurred in these three cases, and the subsequent 18 thoracoscopic cases were treated with a close technique as described below. All patients underwent the procedure under general endotracheal anesthesia with a double lumen endotracheal tube. The ipsilateral lung was deflated and nonventilated throughout the procedure. The patient was placed in a lateral decubitis position with the ipsilateral arm abducted on a mechanical arm holder. The surgical table was broken to hyperextend the ipsilateral chest to spread the intercostal spaces (Fig. 1). The patient was in position to undergo an open thoracotomy if necessary. All ports were inserted by making a 1 cm skin incision followed by blunt dissection with a curved clamp through the muscles and pleura. All patients had at least three ports (Fig. 2). Port placements were as follows. The scope port went through the fifth or sixth intercostal space in the midaxillary line. An instrument port went in this same intercostal space but in the posterior axillary line. A second instrmnent port was placed in the third or fourth intercostal space in the anterior axillary line. Two patients required an additional instrumentation port to retract the lung; this port was placed in the fifth intercostal space along the anterior axillary line. Instruments that were used include the following: video/monitor equipment, 8 to 10 mm diameter rigid laparoscope with 0- or 30-degree angled lens, a 5 mm diameter suction/irrigation/electrocautery hook apparatus, a 5 mm diameter, blunt-tipped scissors with electrocautery attachment, a 10 mm diameter curved grasper, a 10 mm diameter right angled clamp, a 10 mm diameter laparoscopic clip applier, and, if needed, a 10 mm diameter laparoscopic fhnshaped retractor. The steps of the procedure were as follows. (1) Insert scope and evaluate the intrathoracic anatomy. (2) Insert two instrument ports as described above with scope assistance. (3) Take down adhesions if necessary (this was required in four cases). (4)
A h n et al.
513
Collapse and deflate the ipsilateral lung mechanically. The lung is simply compressed by blunt instruments with the endotracheal tube opened to that lung. (5) Identify the ribs with its intercostal nerves and follow these ribs to the vertebral body. (6) Dorsal to the vertebral body, identify the sympathetic ganglia and chain that will appear as a white, glistening, raised, longitudinal structure. (7) Identify the subclavian artery, which marks the superior aspect of the sympathetic dissection. Also, identify the azygous and subclavian veins and the highest intercostal artery and veins. These structures should be avoided during the subsequent dissection. (8) Incise the pleura over the sympathetic chain. (9) Dissect free and lift the chain from its bed. (10) Clip and then sharply cut the rami communicantes of sympathetic ganglia T2-4. (11) Clip and divide just below the T4 ganglion. (12) Dissect free, clip, and sharply divide the lower third of the stellate ganglion along with the rami communicante coursing caudally from the stellate ganglion. The rami coursing in a rostral direction should be left intact, and the clip should be applied caudal to these rostral-directed rami. (13) Send the ganglion to pathology for frozen section and confirm the tissue histologically. (14) Irrigate and control any bleeding points with electrocautery or clips. (15) Remove the instrument ports and inspect the intercostal defects for any bleeding points. (16) Insert the scope through the anterior instrument port and inspect the thoracoscope port for any bleeding points. (17 I) Insert a small chest tube through the thoracoscope port and position the tube under direct visualization. (18) Reinflate the lung using 30 to 40 mm Hg positive pressure and confirm adequate reinflafion with scope visualization. (19) Close the incision in two layers using absorbable sutures. After operation, the chest tube was placed to waterseal and negative 15 cm water suction. A chest radiograph was obtained in the recovery room to check for any pneumothorax or pleural effusions. The chest tube was removed in the recovery room or by the next day if there was no air leak. If there was an air leak, which occurred in one patient, then the chest tube was left in place until the air leak resolved. Oral analgesia was adequate in all patients. All patients were also given an incentive spirometer to prevent atelectasis. RESULTS
Technical success was achieved in 20 of 21 cases (95%). In 19 cases, T1-4 ganglia were dissected and resected as described above. In one case, the T3-7 sympathetic ganglia were resected because the patient
JOURNAL OF VASCULAR SURGERY October 1994
514 A h n et al.
i
Fig. 2. Illustration of chest wall. Note 2 instrument ports in anterior and posterior axillaryline, and scope port in midaxillary line.
had a previous open supraclavicular resection of T1-2. This patient required resection of the lower ganglia because of causalgia-like pain of the fourth, fifth, and sixth intercostal nerve distribution caused by herpes zoster. We had a technical failure in one patient. We were unable to adequately visualize the sympathetic chain because of dense adhesions from a previous first rib resection, which had been complicated by pneumothorax and hemothorax. This case was converted to an open procedure for safety reasons. This case also occurred early in our series when our confidence level was stiU on the rise. Now, we probably would proceed with a closed procedure in this patient. The operative duration was 1.0 to 3.5 hours, mean 2.0 hours. Estimated blood loss was 5 to 300 cc, mean 42 cc, median 10 cc. No patient required transfusion. All patients had histologic confirmation of ganglia. After operation, all patients demonstrated a satisfactory clinical sympathectomy response manifested by a warm, dry hand and upper extremity. Two patients had a transient Homer's syndrome, which resolved within 1 week. Three patients showed a residual pneumothorax but did not require further treatment. In one patient a pneumothorax and a pleural effusion developed, which were treated with the reinsertion of the chest tube. Transient intercostal
neuralgia that resolved within weeks developed in two patients. The patients stayed in the hospital 1 to 4 days after operation, mean 2 days, median 1 day. All patients, except one, were also admitted on the same day as the surgery. All patients had a complete follow-up with a direct examination and report by a physician. Followup ranged from 1 to 42 months, mean 11 months, median 6 months. According to indications for the surgery, follow-up were as follows: reflex sympathetic dystrophy, 2 to 18 months; Raynaud's/vasculitis, 1 to 41 months; hyperhidrosis, 1 to 15 months; heart related, 42 months. All patients demonstrated a clinical benefit as follows. The patients with reflex sympathetic dystrophy all showed improvement of their symptoms and required less or no pain medication. All six patients with Raynaud's/vasculitis had improvement of their ulcer-healing or improvement of their gangrene, as well as relief of pain. All five patients with hyperhidrosis were satisfied and had resolution of their hyperhidrosis symptoms. The one patient, who underwent the procedure for a medica]ly refractory cardiac arrhythmias caused by prolonged QT syndrome, demonstrated shortening of the QT interval on electrocardiography and required significandy less heart medication. There were no intraoperative deaths, and only one patient died during the follow-up. This patient
JOURNAL OF VASCULAR SURGERY Volume 20, Number 4
died of a myocardial infarction caused by a severe calcinosis of the coronary arteries. DISCUSSION
This study demonstrates that thoracoscopic cervicodorsal sympathectomy is feasible, safe, and effective. We achieved technical success in 20 of 21 (95%) patients. Complications were few, and most were transient and minor. Hospital stay was relatively short, often comprising only 1 day. Oral analgesia was adequate in all patients. Furthermore, patient acceptance of the procedure was uniformly excellent. Although we did not study a comparable patient group that underwent open procedures, we befieve from our previous experience that the thoracoscopic approach is preferable. Open thoracotomy and sympathectomy have significant morbidity, such as postoperative bleeding, wound infection, chest pain, pneumothorax, Homer's syndrome, pleura] and dural tears, thoracic duct and nerve injury, and intercostal neuralgia. 14-16 However, because our follow-up is relatively short and our experience is small, we cannot recommend broad application of this procedure at this time. Furthermore, our techniques are still being refined and other techniques that have been described need to be considered as well. The most intriguing technique was recently reported by Claes et al.14,1s and Drott et al.16 These investigators have performed more than 700 thoracoscopic sympathectomies using an urologic electroresectoscope through a single puncture in the second intercostal space midclavicular fine to electtocoagulate ganglia T2-3. Using this technique, they have performed bilateral thoracoscopic sympathectomy in 30 minutes with a 99% success rate and a 2% recurrence rate in patients with hyperhidrosis. Their work clearly deserves confirmation by other investigators. The issue of whether the ganglia should be ablated versus resected still needs to be determined by further studies. Other investigators who have relied on thermal ablation have not had such excellent late results and in fact, most investigators have not reported late results at all. 4"13,17"19 Poor results have been attributed to inadequate ablation or regrowth of the nerve tissue. Furthermore, intercostal neuralgia has been reported in patients treated by electroablation, suggesting possible thermal damage to the adjacent intercostal nerves. Few investigators have actually performed a nonelectrocoagulation resection of the ganglia, as we reported in our present study. It is possible that resection of the ganglia with minimal electrocoagulation may prevent postoperative inter-
A h n et al.
515
costal neuralgia, as well as recurrence of sympathetic: activity. The extent of sympathectomy also needs to be addressed. Some investigators have reported that di.. vision or ablation of T-2 ganglion is adequate, whereas others advocate ablation of T2-3, and still[ others advocate ablation of T1-4.131 Interestingly:. Claes et al.14,~s reported excellent results in patients who underwent ablation of T2-3 for hyperhidrosis but poor results for patients with Raynaud's. They also found that T2-3 ablation provides excellent relief ofpalmar hyperhidrosis, but that resection ofT-4 was necessary to block axillary hyperhidrosis. In our se-. ries, we uniformly resected T1-4 ganglia and ob-. tained excellent results in all patients with Ray-. naud's/vasculitis and causalgia/reflex sympathetic: dystrophy, as welllas those with hyperhidrosis. Until[ the required level of ganglia resection is better de-. fined, we recommend resection ofT1-4 as advocated[ byKux2 and supported by our study. The issues of Homer's syndrome and postoper-. ative intercostal neuralgia also deserve f m ~ e r con-. sideration. We were able to avoid permanent Horner's syndrome in all 21 of our cases, despite the fact that we resected the lower third of the stellate ganglion. We attribute this to our ability to accurately' visualize and identify the multiple rami of the stellate ganglion and clearly visualized the direction of their course. Because the endoscope greatly magnifies the surgical field, such visualization is readily available. Hence, we can divide the stellate ganglion precisely' between the fibers that course rostrally and the fibers; that course caudally. Our results suggest that the fibers that course rostrally innervate the ocular pupillary muscles, and that the resection of the rami coursing caudally does not cause a Homer's syndrome. If this is true and confirmed, then thoracoscopic visualization during sympathectomy is clearly' advantageous. The cause of transient Homer's syndrome in our two patients is not clear; we attribute it to the traction that we placed on the stellate ganglion during its dissection. The postoperative intercostal neuralgia seen in two of our patients occurred early in our series when we used rigid 10 to 12 mm diameter thoracoports. Later in our series, we switched to soft rubber thoracoports; and since that time, we have not encountered this side effect. This suggests that the thoracoports applied direct pressure on the intercostal nerves, thereby causing neuropraxia and hypersensitivity of these nerves. Other possible explanations for the neuralgia include electtocautery injury or inadvertent clipping of the intercostal nerves.
516
A h n et al.
The one patient who underwent sympathectomy for a medically refractory cardiac arrhythmia deserves further discussion. This patient had an exceUent response to surgery, and there are numerous reports in the literature that have documented the therapeutic benefits of sympathectomy for cardiac arrhythnlias and even angina. 2°-31 Despite these reports, sympathectomy has not been widely advocated for heart disease because of the significant morbidity and mortality rates associated with this procedure in heart patients at high risk. However, the less invasive thoracoscopic approach with its minimal morbidity and mortality rates could be better accepted and tolerated. These patients deserve further consideration for thoracoscopic sympathectomy. Further studies in this regard are clearly indicated. We conclude that thoracoscopic sympathectomy is feasible, safe, and effective. It appears to be preferable to open sympathectomy. However, our patient follow-up is short, and there are numerous other thoracoscopic sympathectomy techniques described in the literature. Furthermore, this study does not adequately address the larger issue of the specific value of sympathectomy in individual patients given the small number of patients within each subgroup. Accordingly, further studies are needed to confirm its long-term benefits and to establish the optimal thoracoscopic techniques. REFERENCES
1. Kux E. The endoscopic approach to the vegetative nervous system and its therapeutic possibihties. Dis Chest 1951;20: 139-47. 2. Kux M. Thoracic endoscopic sympathectomy in pahnar and axillary hyperhidrosis. Arch Surg 1978;113:264-6. 3. Fritsch A, Kokoschka R, Mach K. Ergebnisse der thorakoskopischer sympathectomie bei hyperhidrosis der oberen Extremitat. Wien Klin Wochenschr 1975;87:548-50. 4. Weale FE. Upper thoracic sympathectomy by transthoracic electrocoagulation. Br J Surg 1980;67:71-2. 5. Malone PS, Duignan )'P, Hederman WP. Transthoracic electrocoagulation (TTEC): a new and simple approach to upper hmb sympathectomy. Ir Med J 1982;75:20-1. 6. Horgan K, O'Flanagan S, Duignan PJ, Hederman W. Pahnar and axillary hyperhidrosis treated with sympathectomy by transthoracic endoscopic electrocoagulation [Abstract]. Br J Surg 1984;71:1002. 7. Malone PS, Cameron AEP, Rennie JA. The surgical treatment of upper limb hyperhidrosis. Br J Dermatol 1986;115:81-4. 8. Banerjee AK, Edmonson R, Rennie JA. Endoscopic transthoracic electrocautery of the sympathetic chain for palmar and axillary hyperhidrosis. Br J Surg 1990;77:1435-6. 9. Lin CC. A new method of thoracoscopic sympathectomy in hyperhidrosis palmaris. Surg Endosc 1990;4:224-6.
JOURNAL OF VASCULARSURGERY October 1994
10. Byrne J, WalshTN, HedermanWP. Endoscopictransthoracic electrocautery of the sympathetic chain for palmar and axillary hypethidrosis. Br J Surg 1990;77:1046-9. 11. Adams DCR, Poskitt KR. Surgical management of primary hypethidrosis. Br J Surg 1991;78:1019-20. 12. Norback B, Svartholm E. Endoskopisk torakal sympatektomi enkel och effektiv metod mot handsvett [Abstract]. Lakartidningen 1992;89:1478. 13. Edmonson RA, Banerjee AK, Rennie JA. Endoscopic transthoracic sympathectomy in the treatment of hypethidrosis. Ann Surg 1992;215:289-93. 14. Claes G, Gothberg G, Drott C. Endoscopic electrocautery of the thoracic sympathetic chain: a minimally invasive method to treat palmar hypethidrosis. Scand J Hast Reconstr Surg Hand Surg 1993;27:29-33. 15. Claes G, Gothberg G. Endoscopic transthoracic electrocautery of the sympathetic chain for palmar and axiUaryhyperhidrosis [Letter]. Br J Surg 1991;78:760. 16. Drott C, Gothberg G, Claes G. Endoscopic procedures of the upper thoracic sympathetic chain: a review. Arch Surg 1993; 128:237-41. 17. Chuang KS, Liou NH, Liu JC. New stereotactic technique for percutaneous thermocoagulation of upper thoracic ganghonectomy in cases of palmar hyperhidrosis. Neurosurgery 1988;22:600-4. 18. Adler OB, Engel A, Rosenberger A, Dondehnger R. Palmar hyperhidrosis CT guided chemical percutaneous thoracic sympathectomy. Fortschr Rontgenstr. 1990;153:400-3. 19. Wilkinsson HA. Percutaneous radiofrequency upper thoracic sympathectomy: a new technique. Neurosurgery 1984;15: 811-4. 20. Crampton R. Preeminence of the left stellate ganglion in the long Q-T syndrome. Circulation 1979;59:769-78. 21. Schwartz H, Snebold NG, Brown AM. Effects of unilateral cardiac sympathetic denervation on the ventricular fibrillation threshold. Am J Cardiol 1976;37:1034-40. 22. Zaza A, Malfatto G, Schwartz P[L Sympathetic modulation of the relation between ventricular repolarization and cycle length. Circ Res 1991;68:1191-203. 23. Austoni P, Rosati R, Gregorini L, Bianchi E, Bortolani E, Schwartz PJ. SteUectomy and exercise in man [Abstract]. Am J Cardiol 1979;43:399. 24. Schwartz PJ, Stone HL. Left steUectomy in the prevention of ventricfllar fibrillation caused by acute myocardial ischaemia in conscious dogs with anterior myocardial infarction. Circulation 1980;62:1256-65. 25. White JC, White PD. Angina pectoris: treatment with paravertebral alcohol injections. JAMA 1990;28:1099-1103. 26. Lindgren I. Angina pectoris: a clinical study with special reference to neurosurgical treatment. Acta Med Scand 1950(suppl):1-141. 27. Papa MZ, Schneiderman J, Tucker E, Bass A, Drori Y, Adar R. Cardiovas~lar changes after bilateral upper dorsal sympathectomy. Ann Surg 1986;204:715-8. 28. Blomberg S, Curelaru I, Emanuelsson H, Herhtz J, Ponten J, Ricksten SE. Thoracic epidural anaesthesia in patients with unstable angina pectoris. Eur Heart J 1989;10:437-44. 29. Blomberg S, Emannelsson H, Ricksten SE. Thoracic epidural anaesthesia and central hemodynamics in patients with unstable angina pectoris. Anesth Analg 1989;69: 558-62. 30. Schwartz H, Locati EH, Moss AJ, Crampton RS, Trazzi R, Ruberti U. Left cardiac sympathetic denervation in the
JOURNAL OF VASCULARSURGERY Volume 20, Number 4
therapy of congenital long QT syndrome: a worldwide report. Circulation 1991;84:503-11. 31. Moss AJ, McDonald J. Unilateral cervicothoracic sympathetic ganglionectomy for the treatment of long QT interval syndrome. N Engl J Med 1971;285:903-4.
A h n et al.
517
Submitted Feb. 15, 1994; accepted June 18, 1994.
DISCUSSION Dr. T. J. Bunt (Loma Linda, Calif.). The basic premise to surgery's newest rage, "minimally invasive surgery," is reflected in its very designation- the expectation that it will allow performance of a surgical procedure with less pain and disability for the patient. However, in the flush of being pioneers in new surgical technology, many noninvasive surgeons seem to forget that there is more to the health care delivery equation than the concept that the patient and his surgeon believe that the new technology is better. Appropriate complete evaluation of newly introduced procedures needs also to include fair analyses of the economic impact, including the efficient use of resources, the cost of the technology not only per patient or per procedure, but to the hospital, the costs of future technology upgrades, the costs of retraining personnel, and so on. Before we abandon a tried-and-true standard surgical approach, we should be certain that the new method is significantly better in all respects. Just showing that it is feasible is not adequate justification for either prosetylization nor promulgation. The overriding question that therefore needs to be addressed by the authors is not can we do a cervical sympathectomy via thoracoscopy, but rather do we need to? The other major philosophic problem with "less invasive" procedures is the tendency for surgeons to associate a reduced complication rate (whether proven or not) with the nomenclature of "minimally invasive," and to then extend the indications for the procedure to patients with less symptoms on the possibly false premise that the risk is less. This has been the case for endovascular surgery; will it now be the same for thoracoscopy? The authors made no direct comparisons to standard procedures at their institution. Their thoracoscopic approach requires the same anesthetics and same perioperatire care, but a much more expensive technology, and may result in fewer hospital days. Was the cost saving on days used worth the capital investment for technology? What were the comparative costs? The morbidity rate associated with this procedure seems low; how did it compare with standard operative approaches at your institution? Was the perception of low risk a factor in patient selection, because more than half of the patients were treated for clinical situations that are not universally accepted as compelling reasons for surgical sympathectomy-such as hyperhidrosis and Raynaud's disease. On the other hand, you have presented initial data
that suggest that stellate ganglion resection can be redone in a more refined fashion because of better visualization of the rami, thereby avoiding Homer's s y n d r o m e - certainly a positive attribute of the procedure. In summary, the authors have presented an initial series of thoracoscopic cervical sympathectomy, showing that it can indeed be technically performed. Their short-term results are good, although follow-up is short. They are to be congratulated for their pioneering effort. They and others must demonstrate that it's better than what we already have, and that the cost of the technology is justified. Dr. Herbert I. Machleder. Clearly we haven't demonstrated that we need to do a sympathectomy this way, but I think it is important to recognize that we can and that we can do it safely. In terms of the complications, in our own institution 20 years ago Dr. Barker and I published our series of thoracocervical sympathectomy for upper extremity disorders, and we reported 53 consecutive cases. Twenty-nine were transcervical procedures and 24 were transthoracic procedures. We were trying to decide which was more effective and which approach had less complications, and they were entirely comparable. The complications are no different than the ones we had with thoracoscopic sympathectomy. Occasionally a pneumothorax requires a tube and occasional retropleural hematoma. So clearly, the complication rate isn't reduced by the less invasive technique. On the other hand, the incidence of Homer's syndrome is less, probably because we don't dissect the stellate ganglion quite as far as one would in a transcervical approach. What about the cost? We purchased no new equipment for this procedure. The scopes are used routinely for laparoscopy and thoracoscopy. The video monitors are exactly the same. The thoracoports are all the same. It requires a minimal amount of disposable equipment, and it should be remarked that some of this can be done with cystoscopes or ureteroscopes that are available and resterilizable. So I think the cost is relatively comparable in terms of operating time and equipment. The real saving is that this can probably be done on an outpatient basis or with a 1-day operation, whereas a transthoracic approach really can't:. I think Dr. Bunt is correct about increased utilization,, and we've seen this with laparoscopic cholecystectomy. There's been no saving with doing this with laparoscopy
518
A h n et al.
because so many more cholecystectomies are now done for lesser indications, regardless of whether its perceptional. People are more apt to have a minimally symptomatic gallbladder removed when it can be done with laparoscopy. So there has been a change in indications. Our original procedures were done for patients with digital ulcerations and severe vasculitis, and that represents a very small percentage here. About 20% of our cases were done for hyperhidrosis. The other group has causalgia and reflex sympathetic dystrophy. These patients are generally referred from pain management clinics. The medical therapy is repetitive stellate ganglion blocks. Some of these patients have had them done for 1 and 2 years, usually every 2 weeks to a month: Being able to do a transthoracic sympathectomy through a thoracoscope lowers the threshold for when these patients are referred. There's no question about it. The results are dramatic. I mean, if they get a good result from a stellate ganglion block, they're going to be cured with sympathectomy. This procedure is dramatically different than open. The patients go home early and they go to work in about 3 to 4 days. I think the maximum recovery time was 10 days, and this is dramatically different than the days when we did it transthoracically. Dr. John M. Porter (Portland, Ore.). I believe you should withdraw your conclusion that this technique is superior or preferable to operative sympathectomy because these patients were not randomized. You simply cannot conclude this from historic controls. In the days when we used to do sympathectomy, we used a 3-inch transaxiUary incision. The operation took 45 minutes, and the patients went home the next day. I wonder if you have a technique in search of an indication. We have 1500 patients with Raynaud's syndrome at Oregon. We have 200 plus patients with ischemic finger ulcerations including all sorts ofvascaflitis, and we do a large number of upper extremity arterial bypasses including hand bypasses. We have not performed a thoracic sympathectomy of any sort in the past 4 years. I do not think there's any indication for this. Your concern about reflex sympathetic dystrophy requires some elaboration. We have an entire research unit in our department of neurology devoted to reflex sympathetic dystrophy. They find reflex sympathetic dystrophy indeed to be one of the most overdiagnosed and underappreciated conditions that we deal with. You can't rely on stellate ganglion blocks at all unless you do placebo control injections. I would like to be reassured that the pain clinics that send you all these patients are doing placebo control injections without informing the patients and making sure they have the proper response. Finally, I strongly disagree with your technique of sympathectomy. I feel there is no indication whatsoever to transect the lower one third of the stellate ganglion, and if you do so, you're going to invite Homer's syndrome. Dr. Machleder. On the open sympathectomies Hor-
JOURNAL OF VASCULAR SURGERY October 1994
ner's syndrome routinely occurs. So all of the patients I treat with transcervical sympathectomy or with first rib have Homer's syndrome, fortuantely this can be corrected by an outpatient procedure. If the patient uses two drops of Visine in the eye twice a day, the Homer's syndrome is completely reversed. If they do not want to use Visine eyedrops, which has about the same sympathomimetic effect as the normal ocular pupiUary nerve, they can undergo a minor operation under local anesthetic. It's called the reefing of Mueller's muscle, and it's done by everting the lid and taking a little tuck in that muscle. So I don't think the Homer's syndrome is a problem. In terms of reflex sympathetic dystrophy, it's a dramatic problem for patients who have it. Usually it's diagnosed by radioisotope studies, as well as the typical atrophy and the bony changes ofhyperostosis that you see with the chronic causalgia. Some of the patients have had placebo blocks, it depends on the pain management clinic, and some of them just have repetitive stellate ganglion blocks. If you have good medical therapy for it, then your clinic is going to be very busy. But patients who have true causalgia and get dramatic results from sympathetic block will have dramatic results from sympathectomy. We see very few patients with vasculitis or Raynand's syndrome who need a sympathectomy, and so we find that there's very few indications now compared with our previous series. But there are still individuals with very painful ulcers who will heal with sympathectomy. Without having the benefit ofintraarterial reserpine and some of the other effective medical therapy we had in the past, sympathectomy occasionally is very useful for-digital gangrene and digital ischemia. So I think those are the major indications. Reflex sympathetic dystrophy is responsive to stellate ganglion block. With hyperhidrosis, the response is dramatic. Dr. D. Eugene Strandness (Seattle, Wash.). One of the major problems that always has plagued this area is permanent denervation. I don't see that your operation has done anything more to ensure permanent denervation. Yet as I look at your results, you essentially said that you obtained permanent denervation in all cases. The recurrence rate has been reported as high as 75% after the standard methods have been used. It has been stated that you can't assure complete denervation unless you take the whole stellate, which you obviously are not doing here. The other point I wish you'd comment on are that the restflts of this operation vary greatly with the indications. I agree with John Porter in this with regard to the area of Raynaud's vasculitis. The problem of hyperhidrosis is clearly one in which you ought to be able to get a cure. However, it has been my impression that people have generally been able to do this operation through the neck with very satisfactory results. Dr. Machleder. One thing I failed to mention is that
JOURNAL OF VASCULAR SURGERY Volume 20, Number 4
I don't claim this is any better than doing a transcervical or transthoracic procedure. It's another approach, and I think once you do it through a thoracoscope you probably will try to do it that way every time. But it's no better, and the complications are no different, but the return to work is quite different. I think it's very comparable to gallbladder surgery through a thoracoscope or doing it open. You don't get permanent denervation in that after a while, after three or four months, there's regeneration of the vacuoles at the end of the post ganglionic sympathetic nerve that show that. They again have no epinephrine so that I think you do get sensitization, particularly in those patients who have vasospastic diseases, and it's only a temporizing measure to treating underlying collagen vascular disorder. Clearly, the best results occur in patients with causalgia who are treated early and patients who have hyperhidrosis. The patients with hyperhidrosis are cured and they don't come back with wet hands again. The patients with causalgia generally do very well. But I agree with you that in patients who have vasospastic disorders, particularly those who have collagen vascular diseases, eventually the vasal occlusive problem is going to recreate the problem, and the sympathectomy is probably only good in a transient period of time when there's vasospasm superimposed on the early vasal occlusive process with a collagen vascular disease. The indications have changed over the years, and the thoracoscopic sympathectomy still has the same encumbrances as the transthoracic procedure did, except it is a lesser invasive procedure. Dr. Jerry Goldstone (San Francisco, Calif.). I think that thoracoscopic sympathectomy is better. We've done just a few of these, and, at least for those of us who had done them all supraclavicularly, doing them with thoracoscopy is much simpler. You can see the ganglia much better. I don't think there's any question that it's a better procedure. In our experience it's taken longer to set up the equipment than to do the procedure. Regarding the issue about all this equipment; I think it's time for vascular surgeons to begin to look at this kind of equipment. There may be vascular applications that none of us have thought of or at least none of us have used. It may be time for us to get involved. And, in fact, one of my interests in doing these few procedures was because it was the only thing I could think of that would allow me to get some hands-on experience using this technology for which I think there probably are other applications in our field. Do you think it really is necessary to take even the bottom third of the stellate ganglion? In those few cases that we have done, we have not done that and have had very good clinical results. Now, admittedly the follow-up is fairly short, along the order of what Dr. Ahn has shown, but maybe it's not necessary to take it at all. Dr. Machleder. There are a lot of people who would agree with that, and it may well be true that just taking T-2 and T-3 is adequate. I insisted on taking the lower end of
A h n et al.
519'
the steUate ganglion because that's the way we had done it: in the past, and a number of people had pointed out that T-2 and T-3 may be perfectly adequate• The problem that we encountered was that there's a tremendous experience in the European literature, in the Scandinavian literature, in the English-language literature, and the German literature doing these types of procedures thoracoscopicaUy, and they have tremendous numbers of' cases.
What bothered us was the recurrence rate, which was much higher than the ones we had looked at in the past where we excised the ganglion. All of these sympathetic chains were destroyed either by electric cautery or by radio frequency. So we believed that we ought to try to do it comparably to the way we had done it in the past by excising the ganglion. That's why we took the lower part of the stellate. Dr. N o r m a n L. Browse (London, United Kingdom). The discussion is becoming confused between the indications for sympathectomy and the technique. We may practice that the indication for cervical sympathectomy is, in fact, palmar hyperhidrosis and almost nothing else. The advantage claimed by those who practice the thorascopic coagulation technique is that it is easy to do. You simply place the diathermy electrode over the ganglia, and the one thing you cannot do without dissection is get to the stellate ganglion. That is considered a positive point because first you do not need to touch the stellate ganglia to dry up the hands. You only need to destroy the second and third ganglia, and you avoid Horner's syndrome. Because hyperhidrosis is a relatively benign condition, I would be happy to accept a moderate recurrent rate. The procedure can easily be repeated. But, as far as hyperhidrosis is concerned, there is not a high recurrence rate. We talk about reinnervation, but that belief is based on the recurrence of the symptoms of Raynaud's phenomenon. The evidence of reinnervation in terms of the recurrence of sweating after sympathectomy is almost zero. It will happen occasionally if you do not coagulate adequately, but it's a very quick simple safe procedure and can be repeated. However, it is not entirely without risk. It is far simpler and quicker just to put in a thoracoscope. Dr. Machleder. In your clinics where the sympathectomies have been done either transthoracic and transcervical, where transthoracic thoracoscopic sympathectomy has been introduced, has that been the procedure of choice? Dr. Browse. Yes. The thoracoscopes were already available because the chest surgeons and physicians use them. Most times the thoracic surgeon taught the vascular surgeon how to use the thoracoscope. So it has not involved additional expense. Most vascular surgeons in the United Kingdom have converted to coagulation sympathectomy for hyperhidrosis because it is a very simple, quick, effective procedure• If you want to excise the axillary skin at the same time, you can do that first and use the same incision for the thoracoscopy, thus avoiding twO incisions.