- Email: [email protected]
Resection of Neurogenic Tumors in Children: Is Thoracoscopy Superior to Thoracotomy?
Resection of Neurogenic Tumors in Children: Is Thoracoscopy Superior to Thoracotomy? John K Petty, MD, Denis D Bensard, MD, FACS, David A Partrick, MD, FACS, Richard J Hendrickson, MD, Edythe A Albano, MD, Frederick M Karrer, MD, FACS Minimally invasive resection of solid tumors is controversial because of concerns of inadequate resection and local recurrence. Thoracoscopy has been used in the diagnosis of mediastinal tumors in children, but its role in resection is unproved. The purpose of this study was to compare thoracoscopic and open approaches to the resection of thoracic neurogenic tumors in children. STUDY DESIGN: The tumor registry of a regional children’s hospital was queried to identify patients who underwent resection of neurogenic tumors over a 6-year period. Thoracoscopic and open groups were compared for demographic, operative, oncologic, and outcomes characteristics. RESULTS: Seventeen children underwent resection of mediastinal neurogenic tumors (10 thoracoscopic resections, 7 open resections). Mean age was 4.7 years (range 6 months to 12 years). The thoracoscopic and open groups showed no difference in operative time or blood loss. Tumors in the two groups were comparable in size (5.2 ⫾ 2.2 cm versus 5.7 ⫾ 2.6 cm), histology, surgical margin, and stage. Hospital stay was shorter after thoracoscopic resection (1.9 ⫾ 0.7 days versus 4.1 ⫾ 2.5 days, p ⬍ 0.05). There were no regional recurrences. Distant metastases developed in one patient in each group. Eight of 10 children with malignant tumors remain disease-free at an average of 25 months of followup (range 3 to 80 months). CONCLUSIONS: Thoracoscopic resection of neurogenic tumors achieved similar local control and disease-free survival when compared with open resection in this preliminary series. These results were accompanied by a shorter hospital stay. These findings suggest that thoracoscopic resection of neurogenic tumors in children may offer advantages to open resection and should be studied in the context of a large, cooperative trial. (J Am Coll Surg 2006;203:699–703. © 2006 by the American College of Surgeons) BACKGROUND:
Mediastinal tumors pose a unique set of challenges to the pediatric surgeon. Neurogenic tumors arise in the posterior mediastinum and comprise one-third of all mediastinal tumors in children.1 In children, 60% of posterior mediastinal tumors are malignant, with neuroblastoma being the most common tumor type.2 Thoracoscopic surgery is well established as the preferred approach for other diseases of the chest. The thoracoscopic approach to mediastinal disease is associated with decreased pain and quicker return to full function.3
A comparison of thoracoscopic and open resection of benign neurogenic tumors in adults found a notably shorter hospital stay and an earlier return to work in those who underwent thoracoscopic surgery.4,5 In children, thoracoscopic procedures are likewise well tolerated, with low rates of complication.6 Thoracoscopy is the preferred approach to neurogenic tumors of the mediastinum in adults. This approach is highly successful in resection of such tumors in adults.4,5,7-16 Rates of conversion to thoracotomy have ranged from 13% to 40% in larger adult series, and many of these are “mini-thoracotomies:” smaller than conventional posterolateral thoracotomies.5,13,14,16 Although this minimally invasive approach is accepted for adult patients, some authors caution that it may not be appropriate for tumors ⬎ 5 cm in size or for malignant tumors.14 In addition, complications of minimally inva-
Competing Interests Declared: None. Received April 24, 2006; Revised July 20, 2006; Accepted July 21, 2006. From the Department of Pediatric Surgery (Petty, Bensard, Partrick, Hendrickson, Karrer) and the Department of Hematology/Oncology (Albano), The Children’s Hospital/University of Colorado, Denver, CO. Correspondence address: David A Partrick, MD, Department of Pediatric Surgery, The Children’s Hospital, 1056 E 19th Ave B-323, Denver, CO 80218.
© 2006 by the American College of Surgeons Published by Elsevier Inc.
699
ISSN 1072-7515/06/$32.00 doi:10.1016/j.jamcollsurg.2006.07.022
700
Petty et al
Pediatric Tumors: Open Versus Thoracoscopy
sive resections include hemorrhage, Horner’s syndrome, chylothorax, cerebrospinal fluid leak, and port-site recurrence.4,5,17,18 Minimally invasive surgery for pediatric malignancy continues to be controversial. Minimally invasive surgery is highly effective for establishing a tissue diagnosis, disease staging, and assessment of resectability in children with cancer.19-21 A recent survey of experts in pediatric surgical oncology and pediatric minimally invasive surgery revealed unanimous endorsement of minimally invasive surgery for diagnosis, biopsy, and staging.22 Case series have described the use of minimally invasive surgery for resection of pediatric neuroblastoma,17,23 and a comparison of laparoscopic and open resection for abdominal neuroblastoma revealed shorter times to full feeding, hospital discharge, and initiation of chemotherapy in patients with minimally invasive resections.23 But the minimally invasive approach to the resection of pediatric tumors has not been studied thoroughly, so it is not well supported. In fact, a recent National Institutes of Health-funded initiative to study the application of minimally invasive surgery for pediatric cancer failed from lack of patient accrual.24 To date, no series of thoracoscopic resection of neurogenic tumors has examined outcomes in extended followup, particularly with regard to tumor recurrence. The purpose of this study was to compare the effectiveness of thoracoscopic and open approaches to the resection of neurogenic tumors, with attention to outcomes in the perioperative period and in extended followup. METHODS The tumor registry of a major children’s hospital was queried to identify all children who underwent resection of neurogenic tumors of the posterior mediastinum from 1998 to 2005. Seventeen patients were identified. Patient charts were reviewed for demographic, operative, tumor, and followup data. The decision for open or thoracoscopic resection was entirely at the discretion of the surgeon and the patient’s family. Both approaches were reviewed as part of the informed consent discussion with the family. Patients in both groups received single-lung ventilation for the operation. They were positioned in the lateral decubitus position and rotated partially prone to improve exposure of the posterior mediastinum. Thoracoscopic resection typically used four 5-mm ports. The tumor was mobi-
J Am Coll Surg
lized using a combination of blunt dissection and electrocautery. Once the tumor was dissected free, it was placed in a specimen bag and removed by enlarging one of the port sites as necessary. Posterolateral thoracotomy generally proceeded through the fifth intercostal space. In both groups, tumors that abutted or involved the neural foramina were divided at the level of the neural foramina. No attempt was made to resect the intraforaminal disease. In such cases, a positive pathologic margin was expected. One patient underwent attempted resection with the thoracoscopic approach. The tumor was partially mobilized thoracoscopically, but because of difficulties with maintaining single-lung ventilation and subsequent impairment of visualization, the operation was converted to a small thoracotomy to complete the resection. For the sake of analysis, this patient was included in the “open resection” group. Operative times were calculated from the time of incision to the time of leaving the room. Characteristics of resected tumors were determined by review of pathology reports. Tumor sizes were compared according to the greatest dimension recorded in these reports because complete information about tumor size was not available from radiology reports or surgeons’ operative descriptions in all cases. Malignant tumors were staged according to International Neuroblastoma Staging System criteria. The length of hospital stay was at the discretion of the surgeon. The patients were not treated according to a clinical pathway during the study period. Factors such as need for tube thoracostomy, postoperative pain, tolerance of a diet, and respiratory stability were all included in this decision. Some patients with malignant tumors were kept in the hospital for additional postoperative days to undergo chemotherapy or additional testing. For the sake of this analysis, hospital stays were calculated on the basis of their surgical recovery (removal of the chest tube, pain controlled with oral analgesics, and no supplemental oxygen requirement). Additional hospital days for adjuvant therapy or additional workup of malignant tumors were not included in the hospital stay. Followup times were calculated from the day of operation to the patient’s latest recorded followup. Statistical comparisons between the thoracoscopic and open groups were performed using Student’s unpaired t-test. Microsoft Excel software was used for these computations. A p value ⬍ 0.05 was considered statistically significant.
Vol. 203, No. 5, November 2006
Petty et al
Table 1. Summary of Patient and Operation Characteristics Characteristic
Age, y (range) Male, n (%) Operative time, h (⫾ SD) Blood loss, mL (⫾ SD)
Thoracoscopic (n ⴝ 10)
Open (n ⴝ 7)
5.3 (0.5–12) 5 (50) 1.91 ⫾ 0.87 24 ⫾ 45
3.8 (1.2–8) 5 (71) 2.36 ⫾ 0.81 96 ⫾ 180
RESULTS Seventeen patients underwent resection of thoracic neurogenic tumors during the study period, 10 in the thoracoscopic resection group and 7 in the open resection group (Table 1). Nearly all patients received their operations from surgeons who were capable of performing both thoracoscopic and open resections. In fact, only one patient received an open resection from a surgeon who did not also perform a thoracoscopic resection in this series. The thoracoscopic and open groups were similar in age and gender, with a slightly higher mean age in the thoracoscopic group and a slight male predominance in the open group. Although the operative technique in each of the two groups was different, overall operative times were not notably different. Blood loss for each group was generally low, but one patient in each group had a more complicated resection, resulting in higher blood loss. (One patient in the thoracoscopic group initially presented with metastatic neuroblastoma and was treated with preoperative chemotherapy, promoting fibrosis around the tumor as it regressed. One patient in the open group had a very large tumor [7⫻7⫻9 cm] located very high in the chest). Tumor margins were grossly positive at the neural foramina in three thoracoscopic patients and three open patients. Tumor specimens had microscopically positive margins in two thoracoscopic patients and one open patient. No positive lymph nodes were identified with either approach. Both benign and malignant tumors were resected with the two operative approaches (Table 2). Neuroblastoma was the most common tumor type in both the thoracoscopic and open groups. The two groups were comparable with regard to International Neuroblastoma Staging System (INSS) classification; one patient in each group presented with stage 4 disease, and the remaining patients presented with stage 1 or stage 2 disease at diagnosis. The Shimada pathologic classification gives prognostic information for neuroblastoma, independent of stage, and the two groups were similar with regard to this tumor factor.
Pediatric Tumors: Open Versus Thoracoscopy
701
Table 2. Summary of Tumor Characteristics Characteristic
Thoracoscopic (n ⴝ 10)
Open (n ⴝ 7)
5.2 ⫾ 2.2
5.7 ⫾ 2.6
5 (50)
5 (71)
2 2 0 1
3 1 0 1
2 3 4 (40)
3 2 2 (29)
Tumor size, cm (⫾ SD) Tumor type, n (% of group total) Neuroblastoma INSS Stage Stage 1 Stage 2 Stage3 Stage 4 Shimada classification Favorable Unfavorable Ganglioneuroma Malignant peripheral nerve sheath tumor
1 (10)
INSS, International Neuroblastoma Staging System.
Perioperative outcomes revealed a shorter hospital stay for the group managed with thoracoscopic resection (Table 3). On average, hospital stay was 54% shorter in these patients, who did not undergo a thoracotomy for resection of their tumor. A Horner’s syndrome developed in five patients postoperatively. These patients had tumors in the upper aspects of the posterior mediastinum. In followup, at least two of these patients had complete resolution of these symptoms. In extended followup, outcomes were comparable in both groups, with an average followup of more than 2 years. There were no regional recurrences, there were no port site recurrences in the thoracoscopic group, and none of the benign tumors recurred in either group. With regard to the neuroblastomas, two patients presented with stage 4 disease and were treated with chemotherapy, resulting in a good response before resection. The stage 4 patient in the thoracoscopically resected group recovered well from his operation, but ultimately died of neuroblastoma carciTable 3. Comparison of Short Term and Longterm Outcomes by Type of Resection Variable
Hospital stay, d (⫾ SD) Complications, n Horner’s syndrome Pleural effusion Followup, mo Tumor progression, n Death, n *p ⬍ 0.05.
Thoracoscopic
Open
1.9 ⫾ 0.7
4.1 ⫾ 2.5*
2
3 1 33 1 0
19 1 1
702
Petty et al
Pediatric Tumors: Open Versus Thoracoscopy
nomatous meningitis 16 months later. The stage 4 patient who was treated with thoracotomy has survived to 80 months after his resection. One patient in the thoracotomy group initially had a completely resected stage 1 neuroblastoma, but a cervical mass developed 5 months later, which proved to be neuroblastoma on resection. He subsequently proceeded to chemotherapy followed by bone marrow transplantation. DISCUSSION Thoracoscopic resection of neurogenic tumors in children provided equally good regional tumor control, with an apparently lower hospital stay, when compared with open resection in this preliminary retrospective series. Although other authors have examined the technical feasibility of this approach in children,17,25 this is the first study to compare minimally invasive resection to open resection in the context of extended followup. The strength of these conclusions must be qualified by the limited number of patients, the mix of tumor types, and the nonrandomized, retrospective nature of this study. It is difficult to account for the variability of a particular surgeon’s management of a particular patient in a limited, retrospective series. But 94% of the patients in this series were managed by surgeons who had expertise in both thoracoscopic and open resection, so this variability should be less than if this series compared different institutions or different groups of surgeons within an institution. These limitations preclude a definitive conclusion about the superiority of one approach over another. But within the framework of these limitations, these initial results are very promising for both immediate and extended outcomes. Children with thoracoscopic resections received the benefits of a minimally invasive operation, but these benefits did not come at the expense of a higher complication rate. The findings of Horner’s syndrome in both groups were believed to be a function of the location of these tumors in the upper mediastinum, and not a result of the approach to resection. Pleural drainage of the child with a recurrent pleural effusion revealed neither blood nor chyle, and this child ultimately recovered well. Concerns about port site recurrences were not supported by this study because no such recurrences were seen. This is in keeping with recent data that suggest that port site recurrences are extremely rare in pediatric minimally invasive cancer surgery.26 Hypothetical concerns about port site recurrences should not be a deterrent to the use
J Am Coll Surg
of minimally invasive surgery, given the important benefits demonstrated. Our results parallel adult data that suggest equivalent tumor resection with thoracoscopic resection or open resection, but with a considerably shorter hospital stay and return to work among those in the thoracoscopic group.4 In contrast, we did not find a considerably longer operative time in the thoracoscopic group, although the use of thoracoscopic resection was not uniformly applied in this study. In adults, thoracoscopic resection has become the preferred approach to resection of neurogenic tumors. But this approach has not been widely recommended for children, in part, because of the comparatively higher incidence of malignancy and concerns of the proportionately larger size of many of these tumors in children. Our findings do not support malignancy as a contraindication to thoracoscopic resection. There were no regional recurrences in extended followup among those who underwent thoracoscopic resection. Similarly, tumor size did not preclude this approach. The largest tumor removed thoracoscopically was 8.5 cm in its greatest dimension, in a 2-year-old patient. Although a port site needs to be extended to permit removal of the specimen, this incision can be extended to the smallest extent necessary and does not need to be subjected to a long period of rib retraction. One patient required conversion from thoracoscopy to thoracotomy, but this had more to do with difficulties with single lung ventilation in a patient with a very high takeoff of the right upper lobe bronchus. In this specific patient, thoracoscopy helped direct the interspace for the thoracotomy and also gave a visual guide to help limit the size of the thoracotomy. The results were unchanged if this patient was analyzed with either the thoracoscopic or the open group. Our findings support the sentiments expressed in a recent survey of pediatric surgical oncology and minimally invasive surgery experts, in which thoracoscopic resection of neurogenic tumors was believed to be the best suited application of minimally invasive surgery to resection of pediatric malignancies.22 The question of longterm tumor control would be best answered in a well-powered multicenter trial. Although a recent initiative for minimally invasive surgery in pediatric cancer was not well supported, we suggest that neurogenic tumors of the chest serve as the specific “test case” to reinvigorate such a collaborative study. It is unclear why this previous effort was not widely embraced, but the suggestion of equivalent tumor resection
Vol. 203, No. 5, November 2006
with notably lower morbidity should serve as a patientcentered incentive to pursue a more definitive answer to this question. If the current results are borne out in a cooperative multicenter trial, this would validate thoracoscopy as a superior approach to thoracotomy for resection of neurogenic tumors in children. Author Contributions
Study conception and design: Petty, Bensard, Albano Acquisition of data: Petty, Bensard, Albano Analysis and interpretation of data: Petty, Bensard, Partrick, Hendrickson, Karrer Drafting of manuscript: Petty, Bensard, Partrick, Hendrickson, Albano, Karrer Critical revision: Petty, Bensard, Partrick, Hendrickson, Albano, Karrer REFERENCES 1. Azarow KS, Pearl RH, Zurcher R, et al. Primary mediastinal masses. J Thorac Cardiovasc Surg 1993;106:67–72. 2. King DR, Groner JI, Teich S. Mediastinal cysts and tumors. In: Ziegler MZ, Azizkhan RG, Weber TR, eds. Operative pediatric surgery. New York: McGraw-Hill; 2003:407–419. 3. Lin JC, Hazelrigg SR, Landrenau RJ. Video-assisted thoracic surgery for diseases within the mediastinum. Surg Clin North Am 2000;80:1511–1533. 4. Bousamara M, Haasler GB, Patterson GA, et al. A comparative study of thoracoscopic vs. open removal of benign neurogenic mediastinal tumors. Chest 1996;109:1461–1465. 5. Demmy TL, Krasna MJ, Detterbeck FC, et al. Multicenter VATS experience with mediastinal tumors. Ann Thorac Surg 1998;66:187–192. 6. Rescorla FJ, West KW, Gingalewski CA, et al. Efficacy of primary and secondary video-assisted thoracic surgery in children. J Pediatr Surg 2000;35:134–138. 7. Arapis C, Gossot D, Debrosse D, et al. Thoracoscopic removal of neurogenic mediastinal tumors. Surg Endosc 2004;18:1380– 1383. 8. Boons P, Van Hee R, Hendrickx L. Videothoracoscopic resection of intrathoracic neurogenic tumors: report of two cases. Surg Endosc 2003;17:2028–2031. 9. Kitami A, Suzuki T, Usuda R, et al. Diagnostic and therapeutic thoracoscopy for mediastinal disease. Ann Thorac Cardiovasc Surg 2004;10:14–18.
Petty et al
Pediatric Tumors: Open Versus Thoracoscopy
703
10. Pons F, Lang-Lazdunski L, Bonnet PM, et al. Videothoracoscopic resection of neurogenic tumors of the superior sulcus using the harmonic scalpel. Ann Thorac Surg 2003;75:602– 604. 11. Kumar A, Kumar S, Aggarwal S, et al. Thoracoscopy: the preferred approach for the resection of selected posterior mediastinal tumors. J Laproendosc Adv Surg Tech A 2002;12: 345–353. 12. Han PP, Dickman CA. Thoracoscopic resection of thoracic neurogenic tumors. J Neurosurg Spine 2002;96:304–308. 13. Sakumoto N, Inafuku S, Shimoji H, et al. Videothoracoscopic surgery for thoracic neurogenic tumors: a 7-year experience. Surg Today 2000;30:974–977. 14. Zierold D, Halow KD. Thoracoscopic resection as the preferred approach to posterior mediastinal neurogenic tumors. Surg Laparosc Endosc Percutan Tech 2000;10:222–225. 15. Hazelrigg SR, Boley TM, Krasna MJ, et al. Thoracoscopic resection of posterior neurogenic tumors. Am Surg 1999;65: 1129–1133. 16. Riquet M, Mouroux J, Pons F, et al. Videothoracoscopic excision of thoracic neurogenic tumors. Ann Thorac Surg 1995;60: 943–946. 17. Iwanaka T, Arai M, Kawashima H, et al. Endosurgical procedures for pediatric solid tumors. Pediatr Surg Int 2004;20:39–42. 18. Iwanaka T, Arya G, Ziegler MM. Mechanism and prevention of port-site tumor recurrence after laparoscopy in a murine model. J Pediatr Surg 1998;33:457–461. 19. Ryckman FC, Rodgers BM. Thoracoscopy for intrathoracic neoplasia in children. J Pediatr Surg 1982;17:521–524. 20. Waldhausen JH, Tapper D, Sawin RS. Minimally invasive surgery and clinical decision-making for pediatric malignancy. Surg Endosc 2000;14:250–253. 21. Holcomb GW III, Tomita SS, Haase GM, et al. Minimally invasive surgery in children with cancer. Cancer 1995;76:121– 128. 22. Holcomb GW III. Indications for minimally invasive surgery in pediatric oncology. Pediatr Endos Innov Tech 2001;5:299–303. 23. Iwanaka T, Arai M, Ito M, et al. Surgical treatment for abdominal neuroblastoma in the laparoscopic era. Surg Endosc 2001; 15:751–754. 24. Ehrlich PF, Newman KD, Haase GM, et al. Lessons learned from a failed multi-institutional randomized controlled study. J Pediatr Surg 2002;37:431–436. 25. Partrick DA, Rothenberg SS. Thoracoscopic resection of mediastinal masses in infants and children: an evaluation of technique and results. J Pediatr Surg 2001;36:1165–1167. 26. Iwanaka T, Arai M, Yamomoto H, et al. No incidence of portsite recurrence after endosurgical procedure for pediatric malignancies. Pediatr Surg Int 2003;19:200–203.
Mediastinal tumors pose a unique set of challenges to the pediatric surgeon. Neurogenic tumors arise in the posterior mediastinum and comprise one-third of all mediastinal tumors in children.1 In children, 60% of posterior mediastinal tumors are malignant, with neuroblastoma being the most common tumor type.2 Thoracoscopic surgery is well established as the preferred approach for other diseases of the chest. The thoracoscopic approach to mediastinal disease is associated with decreased pain and quicker return to full function.3
A comparison of thoracoscopic and open resection of benign neurogenic tumors in adults found a notably shorter hospital stay and an earlier return to work in those who underwent thoracoscopic surgery.4,5 In children, thoracoscopic procedures are likewise well tolerated, with low rates of complication.6 Thoracoscopy is the preferred approach to neurogenic tumors of the mediastinum in adults. This approach is highly successful in resection of such tumors in adults.4,5,7-16 Rates of conversion to thoracotomy have ranged from 13% to 40% in larger adult series, and many of these are “mini-thoracotomies:” smaller than conventional posterolateral thoracotomies.5,13,14,16 Although this minimally invasive approach is accepted for adult patients, some authors caution that it may not be appropriate for tumors ⬎ 5 cm in size or for malignant tumors.14 In addition, complications of minimally inva-
Competing Interests Declared: None. Received April 24, 2006; Revised July 20, 2006; Accepted July 21, 2006. From the Department of Pediatric Surgery (Petty, Bensard, Partrick, Hendrickson, Karrer) and the Department of Hematology/Oncology (Albano), The Children’s Hospital/University of Colorado, Denver, CO. Correspondence address: David A Partrick, MD, Department of Pediatric Surgery, The Children’s Hospital, 1056 E 19th Ave B-323, Denver, CO 80218.
© 2006 by the American College of Surgeons Published by Elsevier Inc.
699
ISSN 1072-7515/06/$32.00 doi:10.1016/j.jamcollsurg.2006.07.022
700
Petty et al
Pediatric Tumors: Open Versus Thoracoscopy
sive resections include hemorrhage, Horner’s syndrome, chylothorax, cerebrospinal fluid leak, and port-site recurrence.4,5,17,18 Minimally invasive surgery for pediatric malignancy continues to be controversial. Minimally invasive surgery is highly effective for establishing a tissue diagnosis, disease staging, and assessment of resectability in children with cancer.19-21 A recent survey of experts in pediatric surgical oncology and pediatric minimally invasive surgery revealed unanimous endorsement of minimally invasive surgery for diagnosis, biopsy, and staging.22 Case series have described the use of minimally invasive surgery for resection of pediatric neuroblastoma,17,23 and a comparison of laparoscopic and open resection for abdominal neuroblastoma revealed shorter times to full feeding, hospital discharge, and initiation of chemotherapy in patients with minimally invasive resections.23 But the minimally invasive approach to the resection of pediatric tumors has not been studied thoroughly, so it is not well supported. In fact, a recent National Institutes of Health-funded initiative to study the application of minimally invasive surgery for pediatric cancer failed from lack of patient accrual.24 To date, no series of thoracoscopic resection of neurogenic tumors has examined outcomes in extended followup, particularly with regard to tumor recurrence. The purpose of this study was to compare the effectiveness of thoracoscopic and open approaches to the resection of neurogenic tumors, with attention to outcomes in the perioperative period and in extended followup. METHODS The tumor registry of a major children’s hospital was queried to identify all children who underwent resection of neurogenic tumors of the posterior mediastinum from 1998 to 2005. Seventeen patients were identified. Patient charts were reviewed for demographic, operative, tumor, and followup data. The decision for open or thoracoscopic resection was entirely at the discretion of the surgeon and the patient’s family. Both approaches were reviewed as part of the informed consent discussion with the family. Patients in both groups received single-lung ventilation for the operation. They were positioned in the lateral decubitus position and rotated partially prone to improve exposure of the posterior mediastinum. Thoracoscopic resection typically used four 5-mm ports. The tumor was mobi-
J Am Coll Surg
lized using a combination of blunt dissection and electrocautery. Once the tumor was dissected free, it was placed in a specimen bag and removed by enlarging one of the port sites as necessary. Posterolateral thoracotomy generally proceeded through the fifth intercostal space. In both groups, tumors that abutted or involved the neural foramina were divided at the level of the neural foramina. No attempt was made to resect the intraforaminal disease. In such cases, a positive pathologic margin was expected. One patient underwent attempted resection with the thoracoscopic approach. The tumor was partially mobilized thoracoscopically, but because of difficulties with maintaining single-lung ventilation and subsequent impairment of visualization, the operation was converted to a small thoracotomy to complete the resection. For the sake of analysis, this patient was included in the “open resection” group. Operative times were calculated from the time of incision to the time of leaving the room. Characteristics of resected tumors were determined by review of pathology reports. Tumor sizes were compared according to the greatest dimension recorded in these reports because complete information about tumor size was not available from radiology reports or surgeons’ operative descriptions in all cases. Malignant tumors were staged according to International Neuroblastoma Staging System criteria. The length of hospital stay was at the discretion of the surgeon. The patients were not treated according to a clinical pathway during the study period. Factors such as need for tube thoracostomy, postoperative pain, tolerance of a diet, and respiratory stability were all included in this decision. Some patients with malignant tumors were kept in the hospital for additional postoperative days to undergo chemotherapy or additional testing. For the sake of this analysis, hospital stays were calculated on the basis of their surgical recovery (removal of the chest tube, pain controlled with oral analgesics, and no supplemental oxygen requirement). Additional hospital days for adjuvant therapy or additional workup of malignant tumors were not included in the hospital stay. Followup times were calculated from the day of operation to the patient’s latest recorded followup. Statistical comparisons between the thoracoscopic and open groups were performed using Student’s unpaired t-test. Microsoft Excel software was used for these computations. A p value ⬍ 0.05 was considered statistically significant.
Vol. 203, No. 5, November 2006
Petty et al
Table 1. Summary of Patient and Operation Characteristics Characteristic
Age, y (range) Male, n (%) Operative time, h (⫾ SD) Blood loss, mL (⫾ SD)
Thoracoscopic (n ⴝ 10)
Open (n ⴝ 7)
5.3 (0.5–12) 5 (50) 1.91 ⫾ 0.87 24 ⫾ 45
3.8 (1.2–8) 5 (71) 2.36 ⫾ 0.81 96 ⫾ 180
RESULTS Seventeen patients underwent resection of thoracic neurogenic tumors during the study period, 10 in the thoracoscopic resection group and 7 in the open resection group (Table 1). Nearly all patients received their operations from surgeons who were capable of performing both thoracoscopic and open resections. In fact, only one patient received an open resection from a surgeon who did not also perform a thoracoscopic resection in this series. The thoracoscopic and open groups were similar in age and gender, with a slightly higher mean age in the thoracoscopic group and a slight male predominance in the open group. Although the operative technique in each of the two groups was different, overall operative times were not notably different. Blood loss for each group was generally low, but one patient in each group had a more complicated resection, resulting in higher blood loss. (One patient in the thoracoscopic group initially presented with metastatic neuroblastoma and was treated with preoperative chemotherapy, promoting fibrosis around the tumor as it regressed. One patient in the open group had a very large tumor [7⫻7⫻9 cm] located very high in the chest). Tumor margins were grossly positive at the neural foramina in three thoracoscopic patients and three open patients. Tumor specimens had microscopically positive margins in two thoracoscopic patients and one open patient. No positive lymph nodes were identified with either approach. Both benign and malignant tumors were resected with the two operative approaches (Table 2). Neuroblastoma was the most common tumor type in both the thoracoscopic and open groups. The two groups were comparable with regard to International Neuroblastoma Staging System (INSS) classification; one patient in each group presented with stage 4 disease, and the remaining patients presented with stage 1 or stage 2 disease at diagnosis. The Shimada pathologic classification gives prognostic information for neuroblastoma, independent of stage, and the two groups were similar with regard to this tumor factor.
Pediatric Tumors: Open Versus Thoracoscopy
701
Table 2. Summary of Tumor Characteristics Characteristic
Thoracoscopic (n ⴝ 10)
Open (n ⴝ 7)
5.2 ⫾ 2.2
5.7 ⫾ 2.6
5 (50)
5 (71)
2 2 0 1
3 1 0 1
2 3 4 (40)
3 2 2 (29)
Tumor size, cm (⫾ SD) Tumor type, n (% of group total) Neuroblastoma INSS Stage Stage 1 Stage 2 Stage3 Stage 4 Shimada classification Favorable Unfavorable Ganglioneuroma Malignant peripheral nerve sheath tumor
1 (10)
INSS, International Neuroblastoma Staging System.
Perioperative outcomes revealed a shorter hospital stay for the group managed with thoracoscopic resection (Table 3). On average, hospital stay was 54% shorter in these patients, who did not undergo a thoracotomy for resection of their tumor. A Horner’s syndrome developed in five patients postoperatively. These patients had tumors in the upper aspects of the posterior mediastinum. In followup, at least two of these patients had complete resolution of these symptoms. In extended followup, outcomes were comparable in both groups, with an average followup of more than 2 years. There were no regional recurrences, there were no port site recurrences in the thoracoscopic group, and none of the benign tumors recurred in either group. With regard to the neuroblastomas, two patients presented with stage 4 disease and were treated with chemotherapy, resulting in a good response before resection. The stage 4 patient in the thoracoscopically resected group recovered well from his operation, but ultimately died of neuroblastoma carciTable 3. Comparison of Short Term and Longterm Outcomes by Type of Resection Variable
Hospital stay, d (⫾ SD) Complications, n Horner’s syndrome Pleural effusion Followup, mo Tumor progression, n Death, n *p ⬍ 0.05.
Thoracoscopic
Open
1.9 ⫾ 0.7
4.1 ⫾ 2.5*
2
3 1 33 1 0
19 1 1
702
Petty et al
Pediatric Tumors: Open Versus Thoracoscopy
nomatous meningitis 16 months later. The stage 4 patient who was treated with thoracotomy has survived to 80 months after his resection. One patient in the thoracotomy group initially had a completely resected stage 1 neuroblastoma, but a cervical mass developed 5 months later, which proved to be neuroblastoma on resection. He subsequently proceeded to chemotherapy followed by bone marrow transplantation. DISCUSSION Thoracoscopic resection of neurogenic tumors in children provided equally good regional tumor control, with an apparently lower hospital stay, when compared with open resection in this preliminary retrospective series. Although other authors have examined the technical feasibility of this approach in children,17,25 this is the first study to compare minimally invasive resection to open resection in the context of extended followup. The strength of these conclusions must be qualified by the limited number of patients, the mix of tumor types, and the nonrandomized, retrospective nature of this study. It is difficult to account for the variability of a particular surgeon’s management of a particular patient in a limited, retrospective series. But 94% of the patients in this series were managed by surgeons who had expertise in both thoracoscopic and open resection, so this variability should be less than if this series compared different institutions or different groups of surgeons within an institution. These limitations preclude a definitive conclusion about the superiority of one approach over another. But within the framework of these limitations, these initial results are very promising for both immediate and extended outcomes. Children with thoracoscopic resections received the benefits of a minimally invasive operation, but these benefits did not come at the expense of a higher complication rate. The findings of Horner’s syndrome in both groups were believed to be a function of the location of these tumors in the upper mediastinum, and not a result of the approach to resection. Pleural drainage of the child with a recurrent pleural effusion revealed neither blood nor chyle, and this child ultimately recovered well. Concerns about port site recurrences were not supported by this study because no such recurrences were seen. This is in keeping with recent data that suggest that port site recurrences are extremely rare in pediatric minimally invasive cancer surgery.26 Hypothetical concerns about port site recurrences should not be a deterrent to the use
J Am Coll Surg
of minimally invasive surgery, given the important benefits demonstrated. Our results parallel adult data that suggest equivalent tumor resection with thoracoscopic resection or open resection, but with a considerably shorter hospital stay and return to work among those in the thoracoscopic group.4 In contrast, we did not find a considerably longer operative time in the thoracoscopic group, although the use of thoracoscopic resection was not uniformly applied in this study. In adults, thoracoscopic resection has become the preferred approach to resection of neurogenic tumors. But this approach has not been widely recommended for children, in part, because of the comparatively higher incidence of malignancy and concerns of the proportionately larger size of many of these tumors in children. Our findings do not support malignancy as a contraindication to thoracoscopic resection. There were no regional recurrences in extended followup among those who underwent thoracoscopic resection. Similarly, tumor size did not preclude this approach. The largest tumor removed thoracoscopically was 8.5 cm in its greatest dimension, in a 2-year-old patient. Although a port site needs to be extended to permit removal of the specimen, this incision can be extended to the smallest extent necessary and does not need to be subjected to a long period of rib retraction. One patient required conversion from thoracoscopy to thoracotomy, but this had more to do with difficulties with single lung ventilation in a patient with a very high takeoff of the right upper lobe bronchus. In this specific patient, thoracoscopy helped direct the interspace for the thoracotomy and also gave a visual guide to help limit the size of the thoracotomy. The results were unchanged if this patient was analyzed with either the thoracoscopic or the open group. Our findings support the sentiments expressed in a recent survey of pediatric surgical oncology and minimally invasive surgery experts, in which thoracoscopic resection of neurogenic tumors was believed to be the best suited application of minimally invasive surgery to resection of pediatric malignancies.22 The question of longterm tumor control would be best answered in a well-powered multicenter trial. Although a recent initiative for minimally invasive surgery in pediatric cancer was not well supported, we suggest that neurogenic tumors of the chest serve as the specific “test case” to reinvigorate such a collaborative study. It is unclear why this previous effort was not widely embraced, but the suggestion of equivalent tumor resection
Vol. 203, No. 5, November 2006
with notably lower morbidity should serve as a patientcentered incentive to pursue a more definitive answer to this question. If the current results are borne out in a cooperative multicenter trial, this would validate thoracoscopy as a superior approach to thoracotomy for resection of neurogenic tumors in children. Author Contributions
Study conception and design: Petty, Bensard, Albano Acquisition of data: Petty, Bensard, Albano Analysis and interpretation of data: Petty, Bensard, Partrick, Hendrickson, Karrer Drafting of manuscript: Petty, Bensard, Partrick, Hendrickson, Albano, Karrer Critical revision: Petty, Bensard, Partrick, Hendrickson, Albano, Karrer REFERENCES 1. Azarow KS, Pearl RH, Zurcher R, et al. Primary mediastinal masses. J Thorac Cardiovasc Surg 1993;106:67–72. 2. King DR, Groner JI, Teich S. Mediastinal cysts and tumors. In: Ziegler MZ, Azizkhan RG, Weber TR, eds. Operative pediatric surgery. New York: McGraw-Hill; 2003:407–419. 3. Lin JC, Hazelrigg SR, Landrenau RJ. Video-assisted thoracic surgery for diseases within the mediastinum. Surg Clin North Am 2000;80:1511–1533. 4. Bousamara M, Haasler GB, Patterson GA, et al. A comparative study of thoracoscopic vs. open removal of benign neurogenic mediastinal tumors. Chest 1996;109:1461–1465. 5. Demmy TL, Krasna MJ, Detterbeck FC, et al. Multicenter VATS experience with mediastinal tumors. Ann Thorac Surg 1998;66:187–192. 6. Rescorla FJ, West KW, Gingalewski CA, et al. Efficacy of primary and secondary video-assisted thoracic surgery in children. J Pediatr Surg 2000;35:134–138. 7. Arapis C, Gossot D, Debrosse D, et al. Thoracoscopic removal of neurogenic mediastinal tumors. Surg Endosc 2004;18:1380– 1383. 8. Boons P, Van Hee R, Hendrickx L. Videothoracoscopic resection of intrathoracic neurogenic tumors: report of two cases. Surg Endosc 2003;17:2028–2031. 9. Kitami A, Suzuki T, Usuda R, et al. Diagnostic and therapeutic thoracoscopy for mediastinal disease. Ann Thorac Cardiovasc Surg 2004;10:14–18.
Petty et al
Pediatric Tumors: Open Versus Thoracoscopy
703
10. Pons F, Lang-Lazdunski L, Bonnet PM, et al. Videothoracoscopic resection of neurogenic tumors of the superior sulcus using the harmonic scalpel. Ann Thorac Surg 2003;75:602– 604. 11. Kumar A, Kumar S, Aggarwal S, et al. Thoracoscopy: the preferred approach for the resection of selected posterior mediastinal tumors. J Laproendosc Adv Surg Tech A 2002;12: 345–353. 12. Han PP, Dickman CA. Thoracoscopic resection of thoracic neurogenic tumors. J Neurosurg Spine 2002;96:304–308. 13. Sakumoto N, Inafuku S, Shimoji H, et al. Videothoracoscopic surgery for thoracic neurogenic tumors: a 7-year experience. Surg Today 2000;30:974–977. 14. Zierold D, Halow KD. Thoracoscopic resection as the preferred approach to posterior mediastinal neurogenic tumors. Surg Laparosc Endosc Percutan Tech 2000;10:222–225. 15. Hazelrigg SR, Boley TM, Krasna MJ, et al. Thoracoscopic resection of posterior neurogenic tumors. Am Surg 1999;65: 1129–1133. 16. Riquet M, Mouroux J, Pons F, et al. Videothoracoscopic excision of thoracic neurogenic tumors. Ann Thorac Surg 1995;60: 943–946. 17. Iwanaka T, Arai M, Kawashima H, et al. Endosurgical procedures for pediatric solid tumors. Pediatr Surg Int 2004;20:39–42. 18. Iwanaka T, Arya G, Ziegler MM. Mechanism and prevention of port-site tumor recurrence after laparoscopy in a murine model. J Pediatr Surg 1998;33:457–461. 19. Ryckman FC, Rodgers BM. Thoracoscopy for intrathoracic neoplasia in children. J Pediatr Surg 1982;17:521–524. 20. Waldhausen JH, Tapper D, Sawin RS. Minimally invasive surgery and clinical decision-making for pediatric malignancy. Surg Endosc 2000;14:250–253. 21. Holcomb GW III, Tomita SS, Haase GM, et al. Minimally invasive surgery in children with cancer. Cancer 1995;76:121– 128. 22. Holcomb GW III. Indications for minimally invasive surgery in pediatric oncology. Pediatr Endos Innov Tech 2001;5:299–303. 23. Iwanaka T, Arai M, Ito M, et al. Surgical treatment for abdominal neuroblastoma in the laparoscopic era. Surg Endosc 2001; 15:751–754. 24. Ehrlich PF, Newman KD, Haase GM, et al. Lessons learned from a failed multi-institutional randomized controlled study. J Pediatr Surg 2002;37:431–436. 25. Partrick DA, Rothenberg SS. Thoracoscopic resection of mediastinal masses in infants and children: an evaluation of technique and results. J Pediatr Surg 2001;36:1165–1167. 26. Iwanaka T, Arai M, Yamomoto H, et al. No incidence of portsite recurrence after endosurgical procedure for pediatric malignancies. Pediatr Surg Int 2003;19:200–203.