Surgical Risk Factors in Primary Surgery for Localized Neuroblastoma: The LNESG1 Study of the European International Society of Pediatric On
http://www.100md.com
《临床肿瘤学》
the Department of Pediatrics, Division of Pediatric Surgery, University of Padova
Departments of Pediatric Hematology-Oncology and Surgery, and Service of Diagnostic Radiology, Giannina Gaslini Children's Hospital, Genova, Italy
Department of Pediatric Oncology, and Services of Statistics and Diagnostic Imaging, Institut Curie, and Department of Pediatric Surgery, Armand Trousseau Children's Hospital, Paris, France
Department of Surgery, The National Hospital, Rikshospitalet, Oslo, Norway
Department of Pediatric Surgery, Hospital La Fe, Valencia, Spain
Department of Surgery, Division of Child Surgery, University Hospital, Vienna, Austria
Department of Pediatric Surgery, St Georges Hospital, London, United Kingdom
ABSTRACT
PURPOSE: Although tumor resection is the mainstay of treatment for localized neuroblastoma, there are no established guidelines indicating which patients should be operated on immediately and which should undergo surgery after tumor reduction with chemotherapy. In an effort to develop such guidelines, the LNESG1 study defined surgical risk factors (SRFs) based on the imaging characteristics.
PATIENTS AND METHODS: A total of 905 patients with suspected localized neuroblastoma were registered by 10 European countries between January 1995 and October 1999; 811 of 905 patients were eligible for this analysis.
RESULTS: Information on SRFs was obtained for 719 of 811 patients; 367 without and 352 with SRFs. Of these 719 patients, 201 patients (four without and 197 with SRFs) underwent biopsy only. An attempt at tumor excision was made in 518 patients: 363 of 367 patients without and 155 of 352 patients with SRFs (98.9% v 44.0%). Complete excision was achieved in 271 of 363 patients without and in 72 of 155 patients with SRF (74.6% v 46.4%), near-complete excision was achieved in 81 and 61 patients (22.3% v 39.3%), and incomplete excision was achieved in 11 and 22 patients (3.0% v 14.2%), respectively. There were two surgery-related deaths. Nonfatal surgery-related complications occurred in 45 of 518 patients (8.7%) and were less frequent in patients without SRFs (5.0% v 17.4%). Associated surgical procedures were also less frequent in patients without SRFs (1.6% v 9.7%).
CONCLUSION: The adoption of SRFs as predictors of adverse surgical outcome was validated because their presence was associated with lower complete resection rate and greater risk of surgery-related complications. Additional studies aiming to better define the surgical approach to localized neuroblastoma are warranted.
INTRODUCTION
Neuroblastoma is the most common extracranial solid malignancy in childhood, accounting for 8% to 10% of all cancers in this age range.1 Approximately half of the patients have disseminated disease at diagnosis and, with the exception of infants,2 have a modest chance of cure.3 The other half of the patients present with localized disease and have a much better prognosis, although some will develop either local or disseminated relapse. Factors that influence the occurrence of relapse have not been clarified but include tumor resectability,4-7 histologic features,8 and biologic characteristics.9 Initial approaches to localized neuroblastoma include immediate complete tumor excision with no need for subsequent therapy, near-complete excision followed by adjuvant therapy in selected patients, or no attempt at excision (if the tumor is considered unresectable). The latter option suggests the administration of chemotherapy in most patients10-13 to shrink the tumor and facilitate subsequent excision.
The percentage of patients constituting these three subsets varies considerably in the published series in relation to the evolving imaging techniques14 and their interpretation, and to physicians' attitudes.4-7 This has relevant implications for the individual patient, given that the burden of the three treatment modalities is quite different. A definition of commonly agreed criteria of primary tumor operability in localized neuroblastoma would be highly desirable to spare chemotherapy to patients who do not benefit of it, to limit the surgery-related complications, to discourage surgical aggression with a greater risk of organ damage and morbidity, and to allow comparison of results derived from different clinical studies.
On the basis of these premises, in 1995 the European International Society of Pediatric Oncology Neuroblastoma Group activated the LNESG1 study on localized neuroblastoma. In an attempt to increase the possibility of successful and safe surgery, radiologic characteristics of the tumor were identified as possible predictors of surgical risks (surgical risk factors [SRFs]). This article focuses on the application and possible validation of these SRFs as useful guidelines for a more uniform surgical approach to localized neuroblastoma.
PATIENTS AND METHODS
The LNESG1 study was opened in January 1995 to registration of children with suspected localized neuroblastoma, and was closed in October 1999. One hundred seven institutions in 10 European countries participated in the study (Austria, Belgium, France, Italy, the Netherlands, Nordic countries, Portugal, Spain, Switzerland, and the United Kingdom; Appendix). National pediatric oncology organizations participated in the development and distribution of the management protocol to their pediatric oncology centers, where patients were managed by a multidisciplinary team, which included a specialist pediatric surgeon. Eligibility criteria included age at diagnosis younger than 20 years, histologic diagnosis of a neuroblastic tumor,8 absence of distant metastases, and no previous chemotherapy. The diagnostic imaging of the tumor included ultrasound scan plus computed tomography and/or magnetic resonance imaging. These investigations were carried out in an effort to detect features (SRFs) thought to reduce the probability of complete tumor resection and/or lead to an increased risk of surgery-related complications.
A number of SRFs were defined based on the previous experience of the principal study investigators. Objective and subjective SRFs were identified (Table 1). The objective SRFs mainly described an intimate relationship between the tumor and adjacent vital structures, in particular major blood vessels, infiltration of intervertebral foramina, and crossing of the midline. SRFs specific to the tumor sites were also identified. For cervical tumors, extension into thorax, and encasement of nervous structures were included. For thoracic tumors, encasement of trachea and/or principal bronchi, extension into abdomen, and involvement of lower mediastinum (origin of medullary artery), were included. For abdominal tumors, infiltration of renal pedicles or porta hepatis, and compromise of kidney and/or ureter were included (Table 1). The subjective SRFs included tumor size related to child size and tumor fragility, the latter suggesting an extensive necrotic component. Both were believed to predispose to tumor rupture during surgery.11 In the presence of any SRF, the surgeon was encouraged to limit primary operation to an open or needle-core biopsy in the expectation of performing complete and safe surgery after chemotherapy. Guidelines related to the operations were also prepared in an attempt to standardize the operating procedures (see Surgical Definitions). The study recommendation for patients without SRFs was immediate operation. All of the remaining patients would receive chemotherapy before surgery.
Surgical Definitions
Immediate or primary surgery was defined as any surgery performed before any other treatment. This may lead to complete, near-complete, or incomplete excision of the tumor, or be limited to a biopsy. Biopsy was defined as an operation aiming to obtain the amount of tumor tissue required for histologic diagnosis and biologic investigations. Complete excision was defined as the removal of all visible tumor, including abnormal lymph nodes. The excision is considered complete even if residual microscopic tumor is revealed by histologic examination. This corresponds to International Neuroblastoma Staging System (INSS15) stage 1. Near-complete excision was defined as excision of the tumor leaving a minimal macroscopical residue. Regional homolateral lymph nodes may be infiltrated by tumor. This corresponds to INSS stage 2. Incomplete excision was defined as excision of the tumor leaving a macroscopic gross residue. This corresponds to INSS stage 3. The term incomplete excision is also attributed to the rare cases in which the tumor is resected completely or nearly, but the contralateral lymph nodes are infiltrated by the tumor. Associated procedures were defined as any additional operations carried out during the main operation. Surgery-related death was defined as a death occurring within 60 days from the initial operation.
Statistical Analyses
National groups were responsible for collecting and registering patient data in their own databases and sending them at 6-month intervals to the international database located at the Institute Curie (Paris, France). The clinical records of all patients were reviewed retrospectively to analyze the compliance of the surgical approach, to determine the results of operation, and to describe the surgery-related complications. Surgical data were collected on the type of operation that was performed (excision or biopsy), the completeness of excision, and the surgery-related complications. Statistical analyses were performed by the 2 test, with the two-tailed Fisher's exact test, when necessary.
RESULTS
Between January 1995 and October 1999, a total of 905 patients with suspected localized neuroblastoma were registered in the study database. Of these 905 patients, 94 were excluded for a variety of reasons, including 28 with definitive diagnosis different from neuroblastoma. This left 811 patients suitable for evaluation (Table 2). Seven hundred nineteen (88.7%) of these 811 patients had information regarding the SRF status, of whom 367 (51.0%) had no SRF and 352 (49.0%) had one or more SRFs. The SRF rate was similar for the main primary tumor sites (P > .19) and ranged from a maximum of 55.0% for the neck to a minimum of 43.4% for the thorax (Table 3). The frequency of the objective and subjective SRFs, distinct for tumor sites are listed in Table 1.
SRFs and Type of Tumor Resection
In 201 of 719 patients with known SRF status (four of 367 without and 197 of 352 with SRFs), the initial surgical approach was limited to a biopsy. The 518 remaining patients underwent an attempt at immediate tumor excision, of whom 363 had no SRFs and 155 had one or more SRFs (Table 4). The type of excision varied significantly according to the presence or absence of SRFs (P < .0001). Complete tumor resection was achieved in 343 of 518 patients (66.2%), and was more frequent in patients without any SRF (74.7% v 46.5%). Near-complete tumor resection occurred in 142 patients (27.4%), and was less frequent in patients without SRFs (22.3% v 39.3%). Lastly, 33 patients (6.4%) underwent incomplete tumor resection with gross residue, of whom 11 had no SRF and 22 had one or more SRFs (3.0% v 14.2%; Table 4).
Complete tumor resection in relation to the primary tumor sites (neck, thorax, abdomen) was achieved in a comparable number of patients (70.4%, 59.7%, and 68.5%, respectively; P > .13; Table 5), and occurred significantly more often in patients without SRFs with thoracic (67.3% v 39.5%; P < .01) and abdominal locations (77.9% v 47.2%; P < .0001; Table 5). No difference was observed for cervical tumors (P = .99).
The subjective SRF relating to tumor size was detected in 180 of 719 patients but was the sole SRF in only 22 (3.1%), of whom 12 underwent primary surgery, which was complete in seven and nearly complete in five. The subjective SRF relating to tumor fragility was detected in 57 of 719 patients but was the sole SRF in only nine patients (1.3%), of whom seven underwent primary surgery with three complete resections; three were near complete and one was an incomplete resection.
Surgery-Related Complications
Of the 201 patients who underwent biopsy alone, four developed nonfatal complications related to this procedure (severe bleeding in two patients, one Bernard-Horner syndrome, and one venous thrombosis). Among the 518 patients who underwent a primary operation, two died as a result of surgery-related complications: one patient without SRFs who died as a result of renal failure, and one patient with SRFs (whose operation included partial pancreatectomy) who died as a result of bleeding and multiple-organ failure. Forty-five additional patients (8.7%) developed nonfatal complications, which were less frequent in patients with complete compared with near-complete and incomplete excision (5.8% v 12.7% v 21.2%, respectively; P < .002). The occurrence of complications was significantly lower in patients with no SRF compared with patients with SRFs (5.0% v 17.4%; P < .0001; Table 6). This occurred in all categories of tumor excision: complete (4.8% v 9.7%), near-complete (6.2% v 21.3%), and incomplete excision (0 v 31.8%; Table 6), but the difference was statistically significant only for patients who underwent near-complete excision.
Surgery-related complications occurred more often in the neck compared with the thorax and abdomen (33.3% v 10.8% v 6.0%; P < .0001; Table 7). The complication rate was lower for patients without SRFs in all tumor sites: 11.0% v 77.8% in the neck, 7.9% v 17.4% in the thorax, and 3.3% v 12.0% in the abdomen. However, the difference was statistically significant only for patients with cervical and abdominal tumor sites (Table 7). Details on complications related to primary tumor site are listed as a footnote to Table 7.
Associated Surgical Procedures
Of 518 patients who underwent an attempt at tumor resection, 21 underwent additional surgical procedures: six of 363 without SRFs (1.6%) versus 15 of 155 with SRFs (9.7%). Procedures for patients with SRFs included 12 nephrectomies (two due to intraoperative complications and 10 undertaken to achieve complete tumor resection), one double thoracotomy, one laparotomy after thoracotomy, and one partial pancreatectomy. Procedures for patients without SRFs included two parietal pleural resections, one tracheotomy, one nephrectomy, one heminephrectomy, and one splenectomy. Only one associated surgical procedure occurred in the 27 patients with cervical tumors (3.7%), compared with four in the 139 patients with thoracic tumors (2.9%) and 16 in the 352 patients with abdominal tumors (4.5%).
DISCUSSION
The role of surgery in the treatment of neuroblastoma is not yet clarified. For instance, there is no convincing evidence that tumor resection improves the outcome of high-risk disseminated neuroblastoma,16,17 nor of stage 4s.18 There is little doubt, however, that tumor resection is the mainstay of therapy for the majority of patients with localized neuroblastoma.4,6,7 If resection is complete, adjuvant therapy may be avoided, even in the presence of unfavorable biologic factors.8 In young patients19 and in certain tumor sites (eg, thorax20), adjuvant treatment may be omitted even after an incomplete resection, provided that the MYCN gene is not amplified.21 Tumors considered unresectable at presentation require neoadjuvant chemotherapy to facilitate effective and safe surgery.22,23
Given that tumor resection may spare patients toxic chemotherapy, surgeons are urged to do their best to perform an excision without tumor residue. However, the incidence of surgery-related complications when tumor resection is pursued in an inappropriate situation may be unacceptably high. Shamberger et al24 reported a 25% incidence of renal damage. The safety and completeness of operation depends on tumor location, and its relationship to major vessels and vital organs. The assessment of operability must take into account the feasibility of complete surgery and the likelihood of injury to related structures. The goal of the surgical guidelines in the LNESG1 study was to define patients in whom the tumor was amenable to initial complete excision without undue risk to the patient. SRFs were defined to avoid operations likely to result in complication or leave gross residual disease. An attempt at resection was recommended only if these factors were excluded by preoperative imaging.
The present data demonstrate a less than perfect compliance to protocol in that immediate operation was often undertaken when the presence SRFs indicated that complete excision was not predicted by the preoperative imaging. Perhaps this is not cause for surprise, given that this was the first study that offered specific surgical guidelines for localized neuroblastoma. Previously, the decision to attempt tumor resection was dependent mainly on the ability and attitude of the surgeon. In the group of 352 patients in whom SRFs were present, primary operation was attempted in 155 (44%). Complete tumor excision was achieved in less than half of this group. Conversely, among patients without SRFs, complete tumor excision was achieved in 271 (74.7%) of 363. Patients with thoracic and abdominal tumors had a statistically higher incidence of macroscopic residual disease when operated on in the presence of SRFs.
The operative complications were few. In the group of 719 patients there were only two surgery-related deaths (0.3%) and 45 nonfatal complications (8.7%). Complications occurred more frequently when an attempt at tumor resection was performed despite SRFs. The low operative death and complication rates recorded in this large multicenter trial that encompassed 107 institutions from nine European countries reflect the expertise of the surgeons involved with pediatric tumors, and could be the result of adopting prudent surgical guidelines that discourage aggressive interventions in presence of imaging documenting a high operative risk. Finally, it should be noted that the large majority of these patients, and in particular those with tumor SRFs, were managed in centers offering specific experience on the surgical treatment of childhood tumors.
The subjective SRFs related to tumor size and fragility were detected frequently (180 and 57 patients, respectively) but were present in isolation from SRFs only in 22 and nine patients, respectively. Operation was performed in 12 and seven of these patients, respectively, leading to complete tumor resection in seven and three. The rarity of these criteria existing independently of other risk factors and the imprecision inherent in their definition make it difficult to establish their significance.
Our data show that a large number of patients underwent initial operation despite SRFs. These operations were more likely to result in an incomplete excision. Operations in this situation led to a higher incidence of complications—in particular, nephrectomy. Surgery remains the mainstay of treatment but every effort should be made to plan a safe and effective operation. SRFs have been defined which predict the outcome of primary operation and in particular the risk of postoperative complication and residual disease. We hope that more stringent application of these criteria will improve the quality of surgical treatment in the future.
Appendix
The following investigators participated in this study. National Coordinators: R. Ladenstein (Austria); G. Laureys (Belgium); H. Rubie (France); B. De Bernardi (Italy); J. De Kraker (the Netherlands); P. Kogner (Nordic Countries); A. de Lacerda (Portugal); V. Castel (Spain); D. Beck (Switzerland); A. Foot (United Kingdom). Pediatric Oncologists: H. Gadner, R. Ladenstein, W. Emminger, I. Mutz, C. Urban, A. Zoubek (Austria); M.C. Baranzelli, P. Chastagner, C. Coze, D. Frappaz, O. Hartmann, J. Michon, D. Plantaz, H. Rubie (France); B. De Bernardi, F. Casale, M. Conte, G. Deb, A. Mancini, M. Tonello (Italy); H.N. Caron, J. de Kraker, P.A. Vo?te (the Netherlands); P. Kogner, I. Storm-Mathisen, H. Schroeder, G. Jonmunosson, M. M?tt?nen (Nordic Countries); V. Castel, P. Garcia de Miguel, C. Melero, A. Navajas, (Spain); D. Beck, A. Feldges (Switzerland); A. Foot, J. Kohler, A. Pearson (United Kingdom). Surgeons: E.P. Horcher (Austria); F. Gauthier, J.M. Guys, P. Helardot (France); C. Boglino, C. Granata, G. Cecchetto, L. Piva (Italy); H.A. Heij, A. Vos (the Netherlands); T. Monclair (Nordic Countries); J.I. Ruiz-Jimenez, E. Costa (Spain); B. Kehrer (Switzerland); K. Holmes (United Kingdom). Radiologists: R. P?tter (Austria); D. Couanet, S. Neuenschwander (France); C. Maleta, P. Tomà (Italy); D. Muro (Spain); F. Gudinchet (Switzerland); R.E.J. Lee (United Kingdom). Pathologists: G. Amman, I. Ambros (Austria); M. Peuchmaur (France); A. Cavazzana, E.G. D'Amore, G. Gambini, G. Pettinato (Italy); J. Brast, D. Troost (the Netherlands); K. Beiske, B. Roald (Nordic Countries); S. Navarro, A. Llombart-Bosch (Spain); R. Laurini (Switzerland); C.J. Cullinane, D. Variend (United Kingdom). Biologists: P. Ambros (Austria); J. Bénard, O. Delattre, M. Favrot (France); G. Basso, C. Dominici, A. Iolascon, A. Pession, G.P. Tonini (Italy); H.N. Caron, R. Versteeg (the Netherlands); T. Martinsson (Nordic Countries); A. Pellin (Spain); N. Gross, F. Niggli (Switzerland); N. Brown, J. Lunec, D. Sheer (United Kingdom).
The Appendix is included in the full-text version of this article, available online at www.JCO.org. It is not included in the PDF (via Adobe? Acrobat Reader?) version.
Authors' Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.
Acknowledgment
We thank the numerous surgeons, pediatricians, and radiologists who collaborated on this study by providing the information regarding their patients. The authors are also indebted to Sara Calmanti for the excellent editing work.
NOTES
Supported by the Italian Neuroblastoma Association.
Presented in part at the Advances in Neuroblastoma Research 11th Conference, June 16-19, 2004, Genova, Italy.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Castleberry RP: Neuroblastoma. Eur J Cancer 33:1430-1437, 1997
De Bernardi B, Pianca C, Boni L, et al: Disseminated neuroblastoma (stage IV and IV-S) in the first year of life: Outcome related to age and stage—Italian Cooperative Group on Neuroblastoma. Cancer 70:1625-1633, 1992
De Bernardi B, Nicolas B, Boni L, et al: Disseminated neuroblastoma in children older than one year at diagnosis: Comparable results with three consecutive high-dose protocols adopted by the Italian Co-Operative Group for Neuroblastoma. J Clin Oncol 21:1592-1601, 2003
O'Neill JA, Littman P, Blitzer P, et al: The role of surgery in localized neuroblastoma. J Pediatr Surg 20:708-712, 1985
Kiely EM: The surgical challenge of neuroblastoma. J Pediatr Surg 29:128-133, 1994
La Quaglia M: Surgical management of neuroblastoma. Semin Pediatr Surg 10:132-139, 2001
Von Schweinitz D, Hero B, Berthold F: The impact of surgical radicality on outcome in childhood neuroblastoma. Eur J Pediatr Surg 12:402-409, 2002
Perez CA, Matthay KK, Atkinson JB, et al: Biologic variables in the outcome of stages I and II neuroblastoma treated with surgery as primary therapy: A Children's Cancer Group study. J Clin Oncol 18:18-26, 2000
Shimada H, Chatten J, Newton WA Jr., et al: Histopathologic prognostic factors in neuroblastic tumors: Definition of subtypes of ganglioneuroblastoma and an age-linked classification of neuroblastomas. J Natl Cancer Inst 73:405-416, 1984
Nitschke R, Smith EI, Shochat S, et al: Localized neuroblastoma treated by surgery: A Pediatric Oncology Group Study. J Clin Oncol 6:1271-1279, 1988
De Bernardi B, Conte M, Mancini A, et al: Localized resectable neuroblastoma: Results of the second study of the Italian Cooperative Group for Neuroblastoma. J Clin Oncol 13:884-893, 1995
Rubie H, Michon J, Plantaz D, et al: Unresectable localized neuroblastoma: Improved survival after primary chemotherapy including carboplatin-etoposide—Neuroblastoma Study Group of the Société Fran?aise d'Oncologie Pédiatrique (SFOP). Br J Cancer 77:2310-2317, 1998
Simon T, Spitz R, Faldum A, et al: New definition of low-risk neuroblastoma using stage, age and 1p and MYCN status. J Pediatr Hematol Oncol 26:791-796, 2004
Kushner BH: Neuroblastoma: A disease requiring a multitude of imaging studies. J Nucl Med 45:1172-1188, 2004
Brodeur GM, Pritchard J, Berthold F, et al: Revisions of the international criteria for neuroblastoma diagnosis, staging, and response to treatment. J Clin Oncol 11:1466-1477, 1993
Castel V, Tovar JA, Costa E, et al: The role of surgery in stage IV neuroblastoma. J Pediatr Surg 37:1574-1578, 2002
Adkins ES, Sawin R, Gerbing RB, et al: Efficacy of complete resection for high-risk neuroblastoma: A Childrens Cancer Study Group study. J Pediatr Surg 39:931-936, 2004
Guglielmi M, De Bernardi B, Rizzo A, et al: Resection of primary tumor at diagnosis in stage IV-S neuroblastoma: Does it affect the clinical course? J Clin Oncol 14:1537-1544, 1996
Kaneko M, Iwagawa M, Ikebukuro K: Complete resection is not required in patients with neuroblastoma under one year of age. J Pediatr Surg 33:1690-1694, 1998
Horiuchi A, Muraji T, Tsugawa C, et al: Thoracic neuroblastoma: Outcome of incomplete resection. Pediatr Surg Int 20:714-718, 2004
Laprie A, Michon J, Hartmann O, et al: High-dose chemotherapy followed by locoregional irradiation improves the outcome of patients with international neuroblastoma staging system stage II and III neuroblastoma with MYCN amplification. Cancer 101:1081-1089, 2004
Castleberry RP, Shuster JJ, Altshuler G, et al: Infants with neuroblastoma and regional lymph node metastases have a favorable outlook after limited postoperative chemotherapy: A Pediatric Oncology Group study. J Clin Oncol 10:1299-1304, 1992
Garaventa A, De Bernardi B, Pianca C, et al: Localized but unresectable neuroblastoma: Treatment and outcome of 145 cases. J Clin Oncol 11:1170-1179, 1993
Shamberger RC, Smith EI, Joshi VV, et al: The risk of nephrectomy during local control in abdominal neuroblastoma. J Pediatr Surg 33:161-164, 1998(Giovanni Cecchetto, Veron)
Departments of Pediatric Hematology-Oncology and Surgery, and Service of Diagnostic Radiology, Giannina Gaslini Children's Hospital, Genova, Italy
Department of Pediatric Oncology, and Services of Statistics and Diagnostic Imaging, Institut Curie, and Department of Pediatric Surgery, Armand Trousseau Children's Hospital, Paris, France
Department of Surgery, The National Hospital, Rikshospitalet, Oslo, Norway
Department of Pediatric Surgery, Hospital La Fe, Valencia, Spain
Department of Surgery, Division of Child Surgery, University Hospital, Vienna, Austria
Department of Pediatric Surgery, St Georges Hospital, London, United Kingdom
ABSTRACT
PURPOSE: Although tumor resection is the mainstay of treatment for localized neuroblastoma, there are no established guidelines indicating which patients should be operated on immediately and which should undergo surgery after tumor reduction with chemotherapy. In an effort to develop such guidelines, the LNESG1 study defined surgical risk factors (SRFs) based on the imaging characteristics.
PATIENTS AND METHODS: A total of 905 patients with suspected localized neuroblastoma were registered by 10 European countries between January 1995 and October 1999; 811 of 905 patients were eligible for this analysis.
RESULTS: Information on SRFs was obtained for 719 of 811 patients; 367 without and 352 with SRFs. Of these 719 patients, 201 patients (four without and 197 with SRFs) underwent biopsy only. An attempt at tumor excision was made in 518 patients: 363 of 367 patients without and 155 of 352 patients with SRFs (98.9% v 44.0%). Complete excision was achieved in 271 of 363 patients without and in 72 of 155 patients with SRF (74.6% v 46.4%), near-complete excision was achieved in 81 and 61 patients (22.3% v 39.3%), and incomplete excision was achieved in 11 and 22 patients (3.0% v 14.2%), respectively. There were two surgery-related deaths. Nonfatal surgery-related complications occurred in 45 of 518 patients (8.7%) and were less frequent in patients without SRFs (5.0% v 17.4%). Associated surgical procedures were also less frequent in patients without SRFs (1.6% v 9.7%).
CONCLUSION: The adoption of SRFs as predictors of adverse surgical outcome was validated because their presence was associated with lower complete resection rate and greater risk of surgery-related complications. Additional studies aiming to better define the surgical approach to localized neuroblastoma are warranted.
INTRODUCTION
Neuroblastoma is the most common extracranial solid malignancy in childhood, accounting for 8% to 10% of all cancers in this age range.1 Approximately half of the patients have disseminated disease at diagnosis and, with the exception of infants,2 have a modest chance of cure.3 The other half of the patients present with localized disease and have a much better prognosis, although some will develop either local or disseminated relapse. Factors that influence the occurrence of relapse have not been clarified but include tumor resectability,4-7 histologic features,8 and biologic characteristics.9 Initial approaches to localized neuroblastoma include immediate complete tumor excision with no need for subsequent therapy, near-complete excision followed by adjuvant therapy in selected patients, or no attempt at excision (if the tumor is considered unresectable). The latter option suggests the administration of chemotherapy in most patients10-13 to shrink the tumor and facilitate subsequent excision.
The percentage of patients constituting these three subsets varies considerably in the published series in relation to the evolving imaging techniques14 and their interpretation, and to physicians' attitudes.4-7 This has relevant implications for the individual patient, given that the burden of the three treatment modalities is quite different. A definition of commonly agreed criteria of primary tumor operability in localized neuroblastoma would be highly desirable to spare chemotherapy to patients who do not benefit of it, to limit the surgery-related complications, to discourage surgical aggression with a greater risk of organ damage and morbidity, and to allow comparison of results derived from different clinical studies.
On the basis of these premises, in 1995 the European International Society of Pediatric Oncology Neuroblastoma Group activated the LNESG1 study on localized neuroblastoma. In an attempt to increase the possibility of successful and safe surgery, radiologic characteristics of the tumor were identified as possible predictors of surgical risks (surgical risk factors [SRFs]). This article focuses on the application and possible validation of these SRFs as useful guidelines for a more uniform surgical approach to localized neuroblastoma.
PATIENTS AND METHODS
The LNESG1 study was opened in January 1995 to registration of children with suspected localized neuroblastoma, and was closed in October 1999. One hundred seven institutions in 10 European countries participated in the study (Austria, Belgium, France, Italy, the Netherlands, Nordic countries, Portugal, Spain, Switzerland, and the United Kingdom; Appendix). National pediatric oncology organizations participated in the development and distribution of the management protocol to their pediatric oncology centers, where patients were managed by a multidisciplinary team, which included a specialist pediatric surgeon. Eligibility criteria included age at diagnosis younger than 20 years, histologic diagnosis of a neuroblastic tumor,8 absence of distant metastases, and no previous chemotherapy. The diagnostic imaging of the tumor included ultrasound scan plus computed tomography and/or magnetic resonance imaging. These investigations were carried out in an effort to detect features (SRFs) thought to reduce the probability of complete tumor resection and/or lead to an increased risk of surgery-related complications.
A number of SRFs were defined based on the previous experience of the principal study investigators. Objective and subjective SRFs were identified (Table 1). The objective SRFs mainly described an intimate relationship between the tumor and adjacent vital structures, in particular major blood vessels, infiltration of intervertebral foramina, and crossing of the midline. SRFs specific to the tumor sites were also identified. For cervical tumors, extension into thorax, and encasement of nervous structures were included. For thoracic tumors, encasement of trachea and/or principal bronchi, extension into abdomen, and involvement of lower mediastinum (origin of medullary artery), were included. For abdominal tumors, infiltration of renal pedicles or porta hepatis, and compromise of kidney and/or ureter were included (Table 1). The subjective SRFs included tumor size related to child size and tumor fragility, the latter suggesting an extensive necrotic component. Both were believed to predispose to tumor rupture during surgery.11 In the presence of any SRF, the surgeon was encouraged to limit primary operation to an open or needle-core biopsy in the expectation of performing complete and safe surgery after chemotherapy. Guidelines related to the operations were also prepared in an attempt to standardize the operating procedures (see Surgical Definitions). The study recommendation for patients without SRFs was immediate operation. All of the remaining patients would receive chemotherapy before surgery.
Surgical Definitions
Immediate or primary surgery was defined as any surgery performed before any other treatment. This may lead to complete, near-complete, or incomplete excision of the tumor, or be limited to a biopsy. Biopsy was defined as an operation aiming to obtain the amount of tumor tissue required for histologic diagnosis and biologic investigations. Complete excision was defined as the removal of all visible tumor, including abnormal lymph nodes. The excision is considered complete even if residual microscopic tumor is revealed by histologic examination. This corresponds to International Neuroblastoma Staging System (INSS15) stage 1. Near-complete excision was defined as excision of the tumor leaving a minimal macroscopical residue. Regional homolateral lymph nodes may be infiltrated by tumor. This corresponds to INSS stage 2. Incomplete excision was defined as excision of the tumor leaving a macroscopic gross residue. This corresponds to INSS stage 3. The term incomplete excision is also attributed to the rare cases in which the tumor is resected completely or nearly, but the contralateral lymph nodes are infiltrated by the tumor. Associated procedures were defined as any additional operations carried out during the main operation. Surgery-related death was defined as a death occurring within 60 days from the initial operation.
Statistical Analyses
National groups were responsible for collecting and registering patient data in their own databases and sending them at 6-month intervals to the international database located at the Institute Curie (Paris, France). The clinical records of all patients were reviewed retrospectively to analyze the compliance of the surgical approach, to determine the results of operation, and to describe the surgery-related complications. Surgical data were collected on the type of operation that was performed (excision or biopsy), the completeness of excision, and the surgery-related complications. Statistical analyses were performed by the 2 test, with the two-tailed Fisher's exact test, when necessary.
RESULTS
Between January 1995 and October 1999, a total of 905 patients with suspected localized neuroblastoma were registered in the study database. Of these 905 patients, 94 were excluded for a variety of reasons, including 28 with definitive diagnosis different from neuroblastoma. This left 811 patients suitable for evaluation (Table 2). Seven hundred nineteen (88.7%) of these 811 patients had information regarding the SRF status, of whom 367 (51.0%) had no SRF and 352 (49.0%) had one or more SRFs. The SRF rate was similar for the main primary tumor sites (P > .19) and ranged from a maximum of 55.0% for the neck to a minimum of 43.4% for the thorax (Table 3). The frequency of the objective and subjective SRFs, distinct for tumor sites are listed in Table 1.
SRFs and Type of Tumor Resection
In 201 of 719 patients with known SRF status (four of 367 without and 197 of 352 with SRFs), the initial surgical approach was limited to a biopsy. The 518 remaining patients underwent an attempt at immediate tumor excision, of whom 363 had no SRFs and 155 had one or more SRFs (Table 4). The type of excision varied significantly according to the presence or absence of SRFs (P < .0001). Complete tumor resection was achieved in 343 of 518 patients (66.2%), and was more frequent in patients without any SRF (74.7% v 46.5%). Near-complete tumor resection occurred in 142 patients (27.4%), and was less frequent in patients without SRFs (22.3% v 39.3%). Lastly, 33 patients (6.4%) underwent incomplete tumor resection with gross residue, of whom 11 had no SRF and 22 had one or more SRFs (3.0% v 14.2%; Table 4).
Complete tumor resection in relation to the primary tumor sites (neck, thorax, abdomen) was achieved in a comparable number of patients (70.4%, 59.7%, and 68.5%, respectively; P > .13; Table 5), and occurred significantly more often in patients without SRFs with thoracic (67.3% v 39.5%; P < .01) and abdominal locations (77.9% v 47.2%; P < .0001; Table 5). No difference was observed for cervical tumors (P = .99).
The subjective SRF relating to tumor size was detected in 180 of 719 patients but was the sole SRF in only 22 (3.1%), of whom 12 underwent primary surgery, which was complete in seven and nearly complete in five. The subjective SRF relating to tumor fragility was detected in 57 of 719 patients but was the sole SRF in only nine patients (1.3%), of whom seven underwent primary surgery with three complete resections; three were near complete and one was an incomplete resection.
Surgery-Related Complications
Of the 201 patients who underwent biopsy alone, four developed nonfatal complications related to this procedure (severe bleeding in two patients, one Bernard-Horner syndrome, and one venous thrombosis). Among the 518 patients who underwent a primary operation, two died as a result of surgery-related complications: one patient without SRFs who died as a result of renal failure, and one patient with SRFs (whose operation included partial pancreatectomy) who died as a result of bleeding and multiple-organ failure. Forty-five additional patients (8.7%) developed nonfatal complications, which were less frequent in patients with complete compared with near-complete and incomplete excision (5.8% v 12.7% v 21.2%, respectively; P < .002). The occurrence of complications was significantly lower in patients with no SRF compared with patients with SRFs (5.0% v 17.4%; P < .0001; Table 6). This occurred in all categories of tumor excision: complete (4.8% v 9.7%), near-complete (6.2% v 21.3%), and incomplete excision (0 v 31.8%; Table 6), but the difference was statistically significant only for patients who underwent near-complete excision.
Surgery-related complications occurred more often in the neck compared with the thorax and abdomen (33.3% v 10.8% v 6.0%; P < .0001; Table 7). The complication rate was lower for patients without SRFs in all tumor sites: 11.0% v 77.8% in the neck, 7.9% v 17.4% in the thorax, and 3.3% v 12.0% in the abdomen. However, the difference was statistically significant only for patients with cervical and abdominal tumor sites (Table 7). Details on complications related to primary tumor site are listed as a footnote to Table 7.
Associated Surgical Procedures
Of 518 patients who underwent an attempt at tumor resection, 21 underwent additional surgical procedures: six of 363 without SRFs (1.6%) versus 15 of 155 with SRFs (9.7%). Procedures for patients with SRFs included 12 nephrectomies (two due to intraoperative complications and 10 undertaken to achieve complete tumor resection), one double thoracotomy, one laparotomy after thoracotomy, and one partial pancreatectomy. Procedures for patients without SRFs included two parietal pleural resections, one tracheotomy, one nephrectomy, one heminephrectomy, and one splenectomy. Only one associated surgical procedure occurred in the 27 patients with cervical tumors (3.7%), compared with four in the 139 patients with thoracic tumors (2.9%) and 16 in the 352 patients with abdominal tumors (4.5%).
DISCUSSION
The role of surgery in the treatment of neuroblastoma is not yet clarified. For instance, there is no convincing evidence that tumor resection improves the outcome of high-risk disseminated neuroblastoma,16,17 nor of stage 4s.18 There is little doubt, however, that tumor resection is the mainstay of therapy for the majority of patients with localized neuroblastoma.4,6,7 If resection is complete, adjuvant therapy may be avoided, even in the presence of unfavorable biologic factors.8 In young patients19 and in certain tumor sites (eg, thorax20), adjuvant treatment may be omitted even after an incomplete resection, provided that the MYCN gene is not amplified.21 Tumors considered unresectable at presentation require neoadjuvant chemotherapy to facilitate effective and safe surgery.22,23
Given that tumor resection may spare patients toxic chemotherapy, surgeons are urged to do their best to perform an excision without tumor residue. However, the incidence of surgery-related complications when tumor resection is pursued in an inappropriate situation may be unacceptably high. Shamberger et al24 reported a 25% incidence of renal damage. The safety and completeness of operation depends on tumor location, and its relationship to major vessels and vital organs. The assessment of operability must take into account the feasibility of complete surgery and the likelihood of injury to related structures. The goal of the surgical guidelines in the LNESG1 study was to define patients in whom the tumor was amenable to initial complete excision without undue risk to the patient. SRFs were defined to avoid operations likely to result in complication or leave gross residual disease. An attempt at resection was recommended only if these factors were excluded by preoperative imaging.
The present data demonstrate a less than perfect compliance to protocol in that immediate operation was often undertaken when the presence SRFs indicated that complete excision was not predicted by the preoperative imaging. Perhaps this is not cause for surprise, given that this was the first study that offered specific surgical guidelines for localized neuroblastoma. Previously, the decision to attempt tumor resection was dependent mainly on the ability and attitude of the surgeon. In the group of 352 patients in whom SRFs were present, primary operation was attempted in 155 (44%). Complete tumor excision was achieved in less than half of this group. Conversely, among patients without SRFs, complete tumor excision was achieved in 271 (74.7%) of 363. Patients with thoracic and abdominal tumors had a statistically higher incidence of macroscopic residual disease when operated on in the presence of SRFs.
The operative complications were few. In the group of 719 patients there were only two surgery-related deaths (0.3%) and 45 nonfatal complications (8.7%). Complications occurred more frequently when an attempt at tumor resection was performed despite SRFs. The low operative death and complication rates recorded in this large multicenter trial that encompassed 107 institutions from nine European countries reflect the expertise of the surgeons involved with pediatric tumors, and could be the result of adopting prudent surgical guidelines that discourage aggressive interventions in presence of imaging documenting a high operative risk. Finally, it should be noted that the large majority of these patients, and in particular those with tumor SRFs, were managed in centers offering specific experience on the surgical treatment of childhood tumors.
The subjective SRFs related to tumor size and fragility were detected frequently (180 and 57 patients, respectively) but were present in isolation from SRFs only in 22 and nine patients, respectively. Operation was performed in 12 and seven of these patients, respectively, leading to complete tumor resection in seven and three. The rarity of these criteria existing independently of other risk factors and the imprecision inherent in their definition make it difficult to establish their significance.
Our data show that a large number of patients underwent initial operation despite SRFs. These operations were more likely to result in an incomplete excision. Operations in this situation led to a higher incidence of complications—in particular, nephrectomy. Surgery remains the mainstay of treatment but every effort should be made to plan a safe and effective operation. SRFs have been defined which predict the outcome of primary operation and in particular the risk of postoperative complication and residual disease. We hope that more stringent application of these criteria will improve the quality of surgical treatment in the future.
Appendix
The following investigators participated in this study. National Coordinators: R. Ladenstein (Austria); G. Laureys (Belgium); H. Rubie (France); B. De Bernardi (Italy); J. De Kraker (the Netherlands); P. Kogner (Nordic Countries); A. de Lacerda (Portugal); V. Castel (Spain); D. Beck (Switzerland); A. Foot (United Kingdom). Pediatric Oncologists: H. Gadner, R. Ladenstein, W. Emminger, I. Mutz, C. Urban, A. Zoubek (Austria); M.C. Baranzelli, P. Chastagner, C. Coze, D. Frappaz, O. Hartmann, J. Michon, D. Plantaz, H. Rubie (France); B. De Bernardi, F. Casale, M. Conte, G. Deb, A. Mancini, M. Tonello (Italy); H.N. Caron, J. de Kraker, P.A. Vo?te (the Netherlands); P. Kogner, I. Storm-Mathisen, H. Schroeder, G. Jonmunosson, M. M?tt?nen (Nordic Countries); V. Castel, P. Garcia de Miguel, C. Melero, A. Navajas, (Spain); D. Beck, A. Feldges (Switzerland); A. Foot, J. Kohler, A. Pearson (United Kingdom). Surgeons: E.P. Horcher (Austria); F. Gauthier, J.M. Guys, P. Helardot (France); C. Boglino, C. Granata, G. Cecchetto, L. Piva (Italy); H.A. Heij, A. Vos (the Netherlands); T. Monclair (Nordic Countries); J.I. Ruiz-Jimenez, E. Costa (Spain); B. Kehrer (Switzerland); K. Holmes (United Kingdom). Radiologists: R. P?tter (Austria); D. Couanet, S. Neuenschwander (France); C. Maleta, P. Tomà (Italy); D. Muro (Spain); F. Gudinchet (Switzerland); R.E.J. Lee (United Kingdom). Pathologists: G. Amman, I. Ambros (Austria); M. Peuchmaur (France); A. Cavazzana, E.G. D'Amore, G. Gambini, G. Pettinato (Italy); J. Brast, D. Troost (the Netherlands); K. Beiske, B. Roald (Nordic Countries); S. Navarro, A. Llombart-Bosch (Spain); R. Laurini (Switzerland); C.J. Cullinane, D. Variend (United Kingdom). Biologists: P. Ambros (Austria); J. Bénard, O. Delattre, M. Favrot (France); G. Basso, C. Dominici, A. Iolascon, A. Pession, G.P. Tonini (Italy); H.N. Caron, R. Versteeg (the Netherlands); T. Martinsson (Nordic Countries); A. Pellin (Spain); N. Gross, F. Niggli (Switzerland); N. Brown, J. Lunec, D. Sheer (United Kingdom).
The Appendix is included in the full-text version of this article, available online at www.JCO.org. It is not included in the PDF (via Adobe? Acrobat Reader?) version.
Authors' Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.
Acknowledgment
We thank the numerous surgeons, pediatricians, and radiologists who collaborated on this study by providing the information regarding their patients. The authors are also indebted to Sara Calmanti for the excellent editing work.
NOTES
Supported by the Italian Neuroblastoma Association.
Presented in part at the Advances in Neuroblastoma Research 11th Conference, June 16-19, 2004, Genova, Italy.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Castleberry RP: Neuroblastoma. Eur J Cancer 33:1430-1437, 1997
De Bernardi B, Pianca C, Boni L, et al: Disseminated neuroblastoma (stage IV and IV-S) in the first year of life: Outcome related to age and stage—Italian Cooperative Group on Neuroblastoma. Cancer 70:1625-1633, 1992
De Bernardi B, Nicolas B, Boni L, et al: Disseminated neuroblastoma in children older than one year at diagnosis: Comparable results with three consecutive high-dose protocols adopted by the Italian Co-Operative Group for Neuroblastoma. J Clin Oncol 21:1592-1601, 2003
O'Neill JA, Littman P, Blitzer P, et al: The role of surgery in localized neuroblastoma. J Pediatr Surg 20:708-712, 1985
Kiely EM: The surgical challenge of neuroblastoma. J Pediatr Surg 29:128-133, 1994
La Quaglia M: Surgical management of neuroblastoma. Semin Pediatr Surg 10:132-139, 2001
Von Schweinitz D, Hero B, Berthold F: The impact of surgical radicality on outcome in childhood neuroblastoma. Eur J Pediatr Surg 12:402-409, 2002
Perez CA, Matthay KK, Atkinson JB, et al: Biologic variables in the outcome of stages I and II neuroblastoma treated with surgery as primary therapy: A Children's Cancer Group study. J Clin Oncol 18:18-26, 2000
Shimada H, Chatten J, Newton WA Jr., et al: Histopathologic prognostic factors in neuroblastic tumors: Definition of subtypes of ganglioneuroblastoma and an age-linked classification of neuroblastomas. J Natl Cancer Inst 73:405-416, 1984
Nitschke R, Smith EI, Shochat S, et al: Localized neuroblastoma treated by surgery: A Pediatric Oncology Group Study. J Clin Oncol 6:1271-1279, 1988
De Bernardi B, Conte M, Mancini A, et al: Localized resectable neuroblastoma: Results of the second study of the Italian Cooperative Group for Neuroblastoma. J Clin Oncol 13:884-893, 1995
Rubie H, Michon J, Plantaz D, et al: Unresectable localized neuroblastoma: Improved survival after primary chemotherapy including carboplatin-etoposide—Neuroblastoma Study Group of the Société Fran?aise d'Oncologie Pédiatrique (SFOP). Br J Cancer 77:2310-2317, 1998
Simon T, Spitz R, Faldum A, et al: New definition of low-risk neuroblastoma using stage, age and 1p and MYCN status. J Pediatr Hematol Oncol 26:791-796, 2004
Kushner BH: Neuroblastoma: A disease requiring a multitude of imaging studies. J Nucl Med 45:1172-1188, 2004
Brodeur GM, Pritchard J, Berthold F, et al: Revisions of the international criteria for neuroblastoma diagnosis, staging, and response to treatment. J Clin Oncol 11:1466-1477, 1993
Castel V, Tovar JA, Costa E, et al: The role of surgery in stage IV neuroblastoma. J Pediatr Surg 37:1574-1578, 2002
Adkins ES, Sawin R, Gerbing RB, et al: Efficacy of complete resection for high-risk neuroblastoma: A Childrens Cancer Study Group study. J Pediatr Surg 39:931-936, 2004
Guglielmi M, De Bernardi B, Rizzo A, et al: Resection of primary tumor at diagnosis in stage IV-S neuroblastoma: Does it affect the clinical course? J Clin Oncol 14:1537-1544, 1996
Kaneko M, Iwagawa M, Ikebukuro K: Complete resection is not required in patients with neuroblastoma under one year of age. J Pediatr Surg 33:1690-1694, 1998
Horiuchi A, Muraji T, Tsugawa C, et al: Thoracic neuroblastoma: Outcome of incomplete resection. Pediatr Surg Int 20:714-718, 2004
Laprie A, Michon J, Hartmann O, et al: High-dose chemotherapy followed by locoregional irradiation improves the outcome of patients with international neuroblastoma staging system stage II and III neuroblastoma with MYCN amplification. Cancer 101:1081-1089, 2004
Castleberry RP, Shuster JJ, Altshuler G, et al: Infants with neuroblastoma and regional lymph node metastases have a favorable outlook after limited postoperative chemotherapy: A Pediatric Oncology Group study. J Clin Oncol 10:1299-1304, 1992
Garaventa A, De Bernardi B, Pianca C, et al: Localized but unresectable neuroblastoma: Treatment and outcome of 145 cases. J Clin Oncol 11:1170-1179, 1993
Shamberger RC, Smith EI, Joshi VV, et al: The risk of nephrectomy during local control in abdominal neuroblastoma. J Pediatr Surg 33:161-164, 1998(Giovanni Cecchetto, Veron)