- Email: [email protected]
Analysis of risk factors for recurrence after the resection of sacral chordoma combined with embolization
The Spine Journal 9 (2009) 972–980
Clinical Study
Analysis of risk factors for recurrence after the resection of sacral chordoma combined with embolization Huilin Yang, MD, PhDa, Lifan Zhu, MDa, Nabil A. Ebraheim, MDb, Xiaochen Liu, BSb, Sharmaine Castillo, BSb, Tiansi Tang, MDa, Jiayong Liu, MDb,*, Hongjuan Cui, PhDc a
Department of Orthopaedic Surgery, First Hospital Affiliated to Suzhou University, 188 Shizi Street, Suzhou 215006, China b Department of Orthopaedic Surgery, University of Toledo, 3065 Arlington Avenue, Toledo, OH 43614, USA c Department of Stem Cell Biology and Regenerative Medicine, Cleveland Clinic, 9500 Euclid Avenue, Cleveland, OH 44195, USA Received 27 May 2009; revised 28 July 2009; accepted 21 August 2009
Abstract
BACKGROUND CONTEXT: Although several authors have already reported on the high local recurrence rate of sacral chordomas after surgical resection, there are no reports on the risk factors for recurrence after resection when combined with preoperative tumor-related blood vessel embolism by digital subtraction angiography (DSA) technique. PURPOSE: To investigate the factors related to the continuous disease-free survival time (CDFS) after the resection of sacral chordomas combined with embolization. STUDY DESIGN/SETTING: Retrospective review of the signs, images, and immunohistochemical data of patients with sacral chordomas treated with an initial operation combined with transcatheter arterial embolization. PATIENT SAMPLE: Twenty-two patients with sacral chordomas received initial resection combined with transcatheter arterial embolization. OUTCOME MEASURES: Recurrence, proliferating cell nuclear antigen (PCNA) expression, basic fibroblast growth factor (bFGF) expression, CDFS. METHODS: All cases were selected and followed for an average of 39.2 months. The roles of gender, age, tumor size, tumor location, surgical method, radiation therapy, PCNA expression, and bFGF expression in local recurrence were analyzed using the log-rank test. RESULTS: Sacral chordomas recurred in eight of 22 cases. The CDFS was significantly greater in tumors located below S3 as compared with those above S3. When evaluating PCNA and bFGF expression levels, the CDFS was greater in low expressions rather than high expressions. It was determined that the surgical method used was of prognostic significance to the CDFS. CONCLUSIONS: Higher tumor location and higher expressions of PCNA and bFGF will lead to a shorter CDFS. Resecting the tumor as completely as possible will decrease the chances of local recurrence of sacral chordomas. Published by Elsevier Inc.
Keywords:
Sacrum; Chordoma; Recurrence; Embolization; PCNA; bFGF; Continuous disease-free survival time
Introduction Sacral chordoma is a rare slow-growing malignant neoplasm arising from cellular remnants of the notochord. Because of its insidious onset and ambiguous symptoms, by the time the tumor has been detected, it is often already FDA device/drug status: not applicable. Author disclosures: none. * Corresponding author. Department of Orthopaedic Surgery, University of Toledo, 3065 Arlington Ave., Toledo, OH 43614, USA. Tel.: (419) 3836558; fax: (419) 383-3526. E-mail address: [email protected] (J. Liu) 1529-9430/09/$ – see front matter Published by Elsevier Inc. doi:10.1016/j.spinee.2009.08.447
large in size and invading the sacral nerves. Radiotherapy and chemotherapy are ineffective, making surgery the main treatment for sacral chordomas. Because of the anatomic characteristics of the region, removing a sacral tumor was once regarded as a difficult procedure with massive blood loss and severe surgical injury [1–5]. If the primary tumor is not completely resected, controlling the disease becomes more difficult, increasing the risk of recurrence. Although the local recurrence rate of sacral chordomas after surgical resections has been discussed in several studies [4,6–9], the present authors have not found any reports on the risk of recurrence after resection with preoperative
H. Yang et al. / The Spine Journal 9 (2009) 972–980
tumor-related blood vessel embolism by digital subtraction angiography (DSA) technique. The present study reviewed initial operations, combined with transcatheter arterial embolization (TAE), of 22 sacral chordomas from January 1994 to September 2006. Significant improvements in outcome were made when using this method. Symptoms, signs, images, and immunohistochemical data were collected to investigate the factors related to the continuous disease-free survival time (CDFS) after the resection of sacral chordoma. Materials and methods Patient selection and data collection There were a total of 22 patients (10 men and 12 women) in the present study. The average age was 47.1613.7 years (range: 22–76 years). Between 1994 and 2006, they were diagnosed and treated, selected for the present study, and their patient charts were reviewed. Original paraffin specimens were available from 16 of the 22 patients. These were immunohistochemically prepared for the detection of basic fibroblast growth factor (bFGF) and proliferating cell nuclear antigen (PCNA). Age at the time of diagnosis, gender, tumor location and size, surgical method, radiation therapy, PCNA and bFGF expression levels, and local recurrence were obtained. This information was encoded to establish a database on an Excel spreadsheet. The CDFS was calculated according to the Kaplan-Meier method and used as the index of recurrence. All data were accurately imported to SPSS 13.0 (SPSS Inc., Chicago, IL, USA). Embolization technique Local anesthesia was given over the level of the femoral artery, and a catheter was inserted in the bilateral internal iliac arteries. Using the DSA technique, arterial embolization was performed preoperatively across the abdominal aorta using a combination of Gelfoam, 35% to 40% omnipaque, and saline solution to create a cylindrical shaped embolus (222 mm3), which blocked all tumor-supplying vessels including the iliac arteries bilaterally, the middle sacral artery, and the numerous small tumor vessels. Surgical technique The optimal time to perform the surgery is within 24 hours of embolization. After general anesthesia was given, patients were placed in the prone position. An ‘‘I’’-shaped incision was made that was extensive enough to ensure the removal of the tumor. The posterior, caudal, and bilateral aspects of the tumor were excised either radically or marginally. The anterior aspect was separated by blunt dissection through retroperitoneal interstitial spaces. The sacral nerves were protected and preserved whenever possible. However, if nerve roots were contaminated, the membrane of these nerves was dissected.
973
Context Risk factors for local recurrence of sacral chordomas following combined surgical resection and preoperative embolization are poorly understood. Contribution In this retrospective series, the authors found three factors associated with recurrence: tumors located cephalad to S3; tumors with high expression of PCNA and bFGF; and tumors treated operatively by intralesional resection. Implications This article provides useful prognostic information for patients with these rare tumors. —The Editors Paraffin specimens with hematoxylin stain and the examination of bFGF and PCNA Sixteen original paraffin specimens of sacral chordomas from 1994 to 2006 were examined in the present study. Basic fibroblast growth factor expression was detected using the immunohistochemistry streptavidin peroxidase (SP) method. Stain results were evaluated as follows: the pigmentation of the renal carcinoma specimen slice with antibody-1 was defined as positive. In contrast, the pigmentation of the renal carcinoma specimen slice without antibody-1 was defined as negative. When positive, the stain of the cell expressing bFGF was brownish yellow with prominent plasma and small vessel walls. Proliferating cell nuclear antigen expression was detected using the immunohistochemistry SP method. Stain results were evaluated as follows: the pigmentation of the colon carcinoma specimen slice with antibody-1 was defined as positive. In contrast, the pigmentation of the colon carcinoma specimen slice without antibody-1 was defined as negative. When positive, the stain of the cell expressing PCNA revealed brownish yellow granulation. Proliferating cell nuclear antigen and bFGF expression levels were scored based on the reagents’ instruction booklets. A four-tier system was used to classify the expression intensities. The positive index number was calculated as the positive cell number divided by the observed cell number multiplied by 100%. A positive index number from 0% to 25% was defined as Grade 1, 26% to 50% was Grade 2, 51% to 75% was Grade 3, and 76% to 100% was Grade 4. Grades 1 and 2 were classified as low expressions, and Grades 3 and 4 were classified as high expressions. Statistical analysis The dependent variable was the patients’ postoperative CDFS. The independent variables were age at the time of diagnosis, gender, tumor location and size, surgical method,
974
H. Yang et al. / The Spine Journal 9 (2009) 972–980
radiation therapy, PCNA expression, and bFGF expression. The log-rank test was used to perform statistical analysis. The Kaplan-Meier estimator was used to create the CDFS curve. The level of statistical significance was p!.05. Results General data Intraoperative blood loss ranged from 200 to 4,500 mL (mean: 1,200 mL). The average follow-up time was 38.2632.0 months (range: 6–144 months). Eight of the 22 cases experienced local recurrence (36.4%). Postoperative neurologic function In 17 patients, the sacral nerve root was reserved bilaterally at and above S3, and the sphincter muscle function of the bladder and bowel was normal. In two cases, the nerve root was reserved unilaterally at and above S3, resulting in a partial dysfunction of the bladder and bowel. The function of the sphincter muscle was impaired in two patients with nerve roots reserved at and above the S1 level. In one case, colostomy and ureterocutaneostomy were performed. Pathohistological data There was one case of chondroid chordoma among the 16 hematoxylin (HE)-stained specimens. The other 15 cases
were typical chordomas (Fig. 1). Scores for PCNA expression were divided into two groups: high expression (six cases) and low expression (10 cases) (Fig. 2). The average score (X6s) for PCNA expression was 2.460.8. Cells positive for PCNA included droplet cells, asteriated cells, and fusiform cells. Interstitial cells were not positive for PCNA. Scores for bFGF expression were divided into two groups: high expression (four cases) and low expression (12 cases) (Fig. 3). The average score (X6s) for bFGF expression was 2.360.7. Relevant factors of sacral chordoma recurrence The results of the log-rank analysis indicated that tumor locations, surgical methods (Fig. 4), PCNA expression (Fig. 5), and bFGF expression contributed to statistical significance in the CDFS. Gender, age, tumor size, and radiation therapy did not have a statistically significant impact on the CDFS. For details on the relevant factors of sacral chordoma recurrence and the CDFS median, see Tables 1 and 2. Discussion The bilateral internal iliac artery and the median sacral artery are the blood supply of the sacrum with numerous collateral branches between these arteries and the superior gluteal artery. Because of the short distance between the sacral tumor and the main vessels, severe blood loss often occurs
Fig. 1. (Top Left) Chondroid chordoma (hematoxylin, 100): tumor cells form sheets and clusters separated by fibrous bands. The composition of cartilage cells is apparent in the tumor tissue. Chondroid stroma appears to be grayish blue. (Top Right) Chondroid chordoma (hematoxylin, 400): the tumor cells appear to be circular and ovaloid. Uneven vacuoles are present in the plasma. Chondroid stroma appears to be grayish blue. (Bottom Left) Typical sacral chordoma (hematoxylin, 100): tumor cells are separated by fibrous bands. The tumor cells are scattered in myxoid stroma and arranged in strips. Plasma contains uneven vacuoles. (Bottom Right) Typical sacral chordoma (hematoxylin, 400): tumor cells are circular with a clear border. Uneven vacuoles are scattered in the plasma. Tumor cells are spread in myxoid stroma.
H. Yang et al. / The Spine Journal 9 (2009) 972–980
975
Fig. 2. PCNA-positive expression (streptavidin peroxidase, 200). (Top Left) Grade 1: 0% to 25%. (Top Right) Grade 2: 26% to 50%. (Bottom Left) Grade 3: 51% to 75%. (Bottom Right) Grade 4: 76% to 100%.
intraoperatively. This can be significantly reduced with TAE. In the present study, intraoperative blood loss ranged from 200 to 4,500 mL (mean: 1,200 mL). Previously published studies that did not use embolization reported significantly greater blood loss [10,11]. It has been noted that the optimal
time to perform surgery is within 24 hours after embolization. This is based on the observation, by the principle author, of the reestablishment of the collateral circulation after embolizing the bilateral internal iliac arteries with gelfoam particles in dogs. Twelve hours after embolization, a few
Fig. 3. bFGF-positive expression (streptavidin peroxidase, 200). (Top Left) Grade 1: 0% to 25%. (Top Right) Grade 2: 26% to 50%. (Bottom Left) Grade 3: 51% to 75%. (Bottom Right) Grade 4: 76% to 100%.
976
H. Yang et al. / The Spine Journal 9 (2009) 972–980
Fig. 4. CDFS curves. (Top Left) Based on age. (Top Right) Based on gender. (Bottom Left) Based on tumor location. (Bottom Right) Based on diameter of the tumor.
branches of collateral circulation could be observed. This increased at 24 hours after embolization. At 48 hours, the collateral circulation was abundant [12]. Sacral chordoma is a low-grade malignant neoplasm that originates from notochordal remnants and occurs exclusively in the axial skeleton. Approximately 50% of chordomas are located in the sacrococcygeal region. Sacral chordomas account for 4% of malignant bone tumors. After the removal of these tumors, local recurrence may occur, which is significantly influenced by several factors. Tumor location is an important factor in recurrence. Sacral tumors normally extend from the distal part of the sacrum to the proximal part, S1 or S2. If located in S3, the diameter of the tumor is usually smaller than the diameter of a tumor located in S2 or S1. Similarly, S2 is usually smaller than S1. In addition, the more proximal the tumor, the more it penetrates the cortex of the sacrum, leading to a higher recurrence rate. Statistically significant differences were found between sacral chordomas located at or below S3 and sacral chordomas located above S3.
Most tumors located above S3 require radical resections, involve a large amount of blood loss, have easily damaged tumor borders, and involve resections or preservations of sacral nerves. Thus, sacral chordomas at these levels are often difficult to resect completely and require numerous surgeries. In contrast, given that the infiltration is relatively small for sacral chordomas at or below S3, operations for these tumors generally do not involve sacral nerve resections. The tumors at these levels can usually be completely resected, dramatically decreasing the recurrence rate. In a review of 23 patients with sacral chordomas by Cheng et al. [13], the CDFS median for tumors above S3 was significantly less than the CDFS median for tumors at or below S3. The present study included 10 cases of tumors at or below S3 (CDFS median: 82 months) and 12 cases above S3 (CDFS median: 52 months). Two of the cases at or below S3 and six of the cases above S3 experienced recurrence. There was a statistically significant difference between the CDFS medians for the two groups (p5.026).
H. Yang et al. / The Spine Journal 9 (2009) 972–980
977
Fig. 5. CDFS curves. (Top Left) Based on surgical method. (Top Right) Based on radiation therapy. (Bottom Left) Based on low expression versus high expression of PCNA. (Bottom Right) Based on low expression versus high expression of bFGF.
Currently, surgical resections are the primary treatment of sacral chordomas [2,8]. Sundaresan et al. [14] observed 34 patients with sacral chordomas and reported that the postoperative local recurrence rate is directly related to the scope and degree of the resection. Fuchs et al. agreed that there is a significant relationship between the local recurrence rate and the margin achieved during surgery. They found that a wide surgical margin is the most important predictor of survival and local recurrence [4]. Cases treated with radical resections have a longer postoperative period without local recurrence than those treated with perilesional
resections (p!.05) [2,6,15]. The reason for this is that the tumor walls of patients with perilesional or intralesional resection are damaged, causing implantation of tumor cells in other areas and increasing the recurrence rate considerably, particularly with intralesional resections. Kaiser et al. [16] stated that the local recurrence rate after the radical resection of an entire tumor was 28%, whereas the local recurrence rate after an intralesional resection was 64%. Several authors have reported findings in agreement with this [5,17]. In the present study, 8 of the 22 cases had a radical resection, one of which experienced
Table 1 Relevant factors of sacral chordoma recurrence Related factors
Age
Result Recurrence No recurrence CDFS (mo) c2 value P value
!50 4 9 82 2.316 .128
Gender $50 4 5 52
M 3 7 58 0.071 .790
CDFS, continuous disease-free survival time; F, female; M, male.
Location F 5 7 82
I 2 8 82 4.982 .026
Size (mm) II 6 6 52
!100 5 10 82 1.093 .296
$100 3 4 44
978
H. Yang et al. / The Spine Journal 9 (2009) 972–980
Table 2 Relevant factors of sacral chordoma recurrence Related factors
Surgery methods
Result Recurrence No recurrence CDFS (mo) c2 value P value
I 6 4 24 9.126 .003
II 1 3 44
Radiation III 1 7 82
Yes 3 8 82 2.101 .147
PCNA No 5 6 52
I 3 7 58 4.066 .044
bFGF II 4 2 22
I 4 8 58 4.732 .030
II 3 1 22
CDFS, continuous disease-free survival time. Key—tumor location: I5S3 and below S3, II5above S3; surgical methods: I5intralesional, II5perilesional, III5radical; PCNA (proliferating cell nuclear antigen): I5low expression, II5high expression; bFGF (basic fibroblast growth factor): I5low expression, II5high expression.
recurrence (12.5%); 4 cases involved a perilesional resection, one of which experienced recurrence (25%); 10 cases had an intralesional resection, six of which experienced recurrence (60%). The lower recurrence rate for radical resections (12.5%) compared with that reported by other studies (Kaiser et al. reported 28%) may be attributed to the preoperative application of tumor-related blood vessel embolism by DSA technique. Reducing the amount of blood loss during surgery is the key measure to removing the tumor en bloc. Preoperative TAE for other types of tumors has been proven to reduce blood loss significantly and maintain a clean surgical field, both of which aid in the complete removal of the tumor [18–20]. Among the present cases, the CDFS medians were 82 months for radical resections, 44 months for perilesional resections, and 24 months for intralesional resections. Statistically significant differences were found in the CDFS medians between the various surgical methods (p5.003). Usually, the combination of the activation of various tumor genes and the inactivation of tumor-inhibiting genes causes malignant tumors to occur and proliferate. Proliferating cell nuclear antigen is a protein that acts as a processivity factor for DNA polymerase delta in eukaryotic cells. Because DNA polymerase delta is involved in synthesizing new DNA that replaces excised damaged DNA during DNA repair, PCNA is necessary for both DNA synthesis and DNA repair. Proliferating cell nuclear antigen is directly related to the generation of tumor vessels and has important effects on the tumor’s malignancy and biological behavior. Therefore, PCNA can be used as an index for tumor DNA proliferation. Proliferating cell nuclear antigen expression among the specimens of sacral chordoma is a valuable clinical guide for tumor recurrence [21]. Among the 16 cases on which the immunohistochemistry SP method was performed in the present study, 10 cases showed low expressions of PCNA (CDFS median: 58 months), of which three experienced recurrence; six cases showed high expressions of PCNA (CDFS median: 22 months), of which four experienced recurrence. The results of the statistical analysis indicated that the cases in the high expression group had significantly shorter CDFS medians than those in the low expression group (Fig. 5).
During tumor development, bFGF mediates the formation of new blood vessels, a process known as angiogenesis. Basic fibroblast growth factor also stimulates abnormal cell division and proliferation, as well as a variety of tumor development and malignant transformations through autocrine and paracrine processes [22,23]. Deniz et al. [24] found statistically significant differences in the recurrence rates between the high and low expression groups of bFGF (p!.05). Among the 16 cases on which the immunohistochemistry SP method was performed in the present study, 12 cases showed low expressions of bFGF (CDFS median: 58 months), of which four experienced recurrence; four cases showed high expressions of bFGF (CDFS median: 22 months), of which three experienced recurrence. The results of the statistical analysis indicated that the cases in the high expression group had significantly shorter CDFS medians than those in the low expression group. In the study by Thieblemont et al. [25], age was found to be an independent factor of postoperative recurrence and CDFS. The present study categorized patients into two groups based on age. Thirteen patients were younger than 50 years, four of which experienced recurrence (CDFS median: 82 months). Nine patients were 50 years and older, and four patients in this group experienced recurrence (CDFS median: 82 months). Statistical analysis of the CDFS for each group did not find them to be significantly different. The ratio of incidence rate of sacral chordomas in men versus women is 2:1 [26,27]. Although gender has been considered a possible factor of postoperative recurrence, a comparison of these studies reveals an inconsistency regarding the effect of gender on the CDFS [25,28]. Of the 22 cases involved in the present study, 10 were men (CDFS median: 58 months) and 12 were women (CDFS median: 82 months). Three of the male patients and five of the female patients experienced recurrence. There was no statistically significant difference between the CDFS medians for the men and the women. Tumor size is one of the factors that reflects the speed of tumor growth, but its effect on tumor recurrence has been contended [2,29]. The average diameter of the tumors among the 22 patients in the present study was 95 mm. Fifteen cases had tumors with diameters !100 mm, of which
H. Yang et al. / The Spine Journal 9 (2009) 972–980
five had recurred (CDFS median: 82 months). Seven cases had tumors with diameters $100 mm, of which four had recurred (CDFS median: 44 months). The present authors believe that these recurrence rates are the result of using the DSA technique to embolize the bilateral internal iliac arteries and the tumor supply vessels, thereby increasing the complete resection rate (Fig. 4). The use of the DSA embolism technique during surgery significantly decreased hemorrhaging and allowed for the resection of larger tumors. Although radiation therapy is a common supplemental treatment for patients after the resection of their tumors, it is debatable as to whether it can reduce local recurrence [2,4,5,9–11,13,14,24,30–32]. In the present study, postoperative radiation was found to have no significant influence on the patient’s overall survival time. Eleven of the 22 patients received postoperative radiation therapy, and three experienced recurrence (CDFS median: 82 months). The remaining 11 cases did not receive radiation therapy, and five experienced recurrence (CDFS median: 52 months). There was no statistically significant difference between the CDFS medians of the two groups. The relationship between the recurrence rate and the maintenance of the sacral root function has not been thoroughly explored. If the nerve root is saved, function is preserved but the recurrence rate may be higher and the CDFS lower. In contrast, if the nerve root is completely removed, nerve root function is lost but the recurrence rate may be lowered. Therefore, the present authors believe that a balance must be achieved between nerve root function and tumor recurrence. The physician must discuss the options thoroughly with the patient, helping to make the appropriate decisions. The improvement in technique as the surgeon’s experience increases may play some role in the recurrence rate. However, the present study used a small group of patients with the same principle surgeon performing all operations. The difference in technique between the first case and the last case carried out by the principle surgeon may not have been significant. The present authors could not draw conclusions about whether the surgeon’s experience influenced the recurrence rate. Conclusions Patients treated with radical resections had a significantly longer CDFS than patients treated with perilesional and intralesional resections. In the present study, a trend was discovered in which the higher the location of the tumor, the shorter the CDFS. In addition, greater expressions of PCNA and bFGF led to a shorter CDFS. Patient’s age at the time of diagnosis, gender, tumor size, and radiation therapy were found to have no statistical significance on the CDFS. References [1] Samson IR, Springfield DS, Suit HD, Mankin HJ. Operative treatment of sacrococcygeal chordoma. A review of twenty-one cases. J Bone Joint Surg Am 1993;75:1476–84.
979
[2] York JE, Kaczaraj A, Abi-Said D, et al. Sacral chordoma: 40-year experience at a major cancer center. Neurosurgery 1999;44:74–9; discussion 79–80. [3] Yonemoto T, Tatezaki S, Takenouchi T, et al. The surgical management of sacrococcygeal chordoma. Cancer 1999;85:878–83. [4] Fuchs B, Dickey ID, Yaszemski MJ, et al. Operative management of sacral chordoma. J Bone Joint Surg Am 2005;87:2211–6. [5] Hulen CA, Temple HT, Fox WP, et al. Oncologic and functional outcome following sacrectomy for sacral chordoma. J Bone Joint Surg Am 2006;88:1532–9. [6] Bergh P, Kindblom LG, Gunterberg B, et al. Prognostic factors in chordoma of the sacrum and mobile spine: a study of 39 patients. Cancer 2000;88:2122–34. [7] Berven S, Zurakowski D, Mankin HJ, et al. Clinical outcome in chordoma: utility of flow cytometry in DNA determination. Spine 2002;27:374–9. [8] Boriani S, Chevalley F, Weinstein JN, et al. Chordoma of the spine above the sacrum. Treatment and outcome in 21 cases. Spine 1996;21:1569–77. [9] Conlon KC, Boland PJ. Laparoscopically assisted radical sacrococcygectomy. A new operative approach to large sacrococcygeal chordomas. Surg Endosc 1997;11:1118–22. [10] Simpson AH, Porter A, Davis A, et al. Cephalad sacral resection with a combined extended ilioinguinal and posterior approach. J Bone Joint Surg Am 1995;77:405–11. [11] Tomita K, Tsuchiya H. Total sacrectomy and reconstruction for huge sacral tumors. Spine 1990;15:1223–7. [12] Yang H, Ni C, Tang T. Surgical treatment of sacral tumors after transcatheter arterial embolization. Chinese J Orthop 1998;18:646–8. [13] Cheng EY, Ozerdemoglu RA, Transfeldt EE, Thompson RC Jr. Lumbosacral chordoma. Prognostic factors and treatment. Spine 1999;24: 1639–45. [14] Sundaresan N, Huvos AG, Krol G, et al. Surgical treatment of spinal chordomas. Arch Surg 1987;122:1479–82. [15] Fourney DR, Gokaslan ZL. Current management of sacral chordoma. Neurosurg Focus 2003;15:E9. [16] Kaiser TE, Pritchard DJ, Unni KK. Clinicopathologic study of sacrococcygeal chordoma. Cancer 1984;53:2574–8. [17] Puri A, Agarwal MG, Shah M, et al. Decision making in primary sacral tumors. Spine J 2009;9:396–403. [18] Gellad FE, Sadato N, Numaguchi Y, Levine AM. Vascular metastatic lesions of the spine: preoperative embolization. Radiology 1990;176: 683–6. [19] Hilal SK, Michelsen JW. Therapeutic percutaneous embolization for extra-axial vascular lesions of the head, neck, and spine. J Neurosurg 1975;43:275–87. [20] Ellman BA, Parkhill BJ, Curry TS III, et al. Ablation of renal tumors with absolute ethanol: a new technique. Radiology 1981;141:619–26. [21] Essers J, Theil AF, Baldeyron C, et al. Nuclear dynamics of PCNA in DNA replication and repair. Mol Cell Biol. 2005;25:9350–9. [22] Chandler LA, Sosnowski BA, Greenlees L, et al. Prevalent expression of fibroblast growth factor (FGF) receptors and FGF2 in human tumor cell lines. Int J Cancer 1999;81:451–8. [23] Griffioen AW, Damen CA, Mayo KH, et al. Angiogenesis inhibitors overcome tumor induced endothelial cell anergy. Int J Cancer 1999;80:315–9. [24] Deniz ML, Kilic¸ T, Almaata I, et al. Expression of growth factors and structural proteins in chordomas: basic fibroblast growth factor, transforming growth factor alpha, and fibronectin are correlated with recurrence. Neurosurgery 2002;51:753–60. [25] Thieblemont C, Biron P, Rocher F, et al. Prognostic factors in chordoma: role of postoperative radiotherapy. Eur J Cancer 1995;31A: 2255–9. [26] Ashwood N, Hoskin PJ, Saunders MI. Metastatic chordoma: pattern of spread and response to chemotherapy. Clin Oncol (R Coll Radiol) 1994;6:341–2.
980
H. Yang et al. / The Spine Journal 9 (2009) 972–980
[27] Forsyth PA, Cascino TL, Shaw EG, et al. Intracranial chordomas: a clinicopathological and prognostic study of 51 cases. J Neurosurg 1993;78:741–7. [28] O’Connell JX, Renard LG, Liebsch NJ, et al. Base of skull chordoma. A correlative study of histologic and clinical features of 62 cases. Cancer 1994;74:2261–7. [29] Baratti D, Gronchi A, Pennacchioli E, et al. Chordoma: natural history and results in 28 patients treated at a single institution. Ann Surg Oncol 2003;10:291–6.
[30] Chandawarkar RY. Sacrococcygeal chordoma: review of 50 consecutive patients. World J Surg 1996;20:717–9. [31] Catton C, O’Sullivan B, Bell R, et al. Chordoma: long-term follow-up after radical photon irradiation. Radiother Oncol 1996;41:67–72. [32] Fourney DR, Rhines LD, Hentschel SJ, et al. En bloc resection of primary sacral tumors: classification of surgical approaches and outcome. J Neurosurg Spine 2005;3:111–22.
Clinical Study
Analysis of risk factors for recurrence after the resection of sacral chordoma combined with embolization Huilin Yang, MD, PhDa, Lifan Zhu, MDa, Nabil A. Ebraheim, MDb, Xiaochen Liu, BSb, Sharmaine Castillo, BSb, Tiansi Tang, MDa, Jiayong Liu, MDb,*, Hongjuan Cui, PhDc a
Department of Orthopaedic Surgery, First Hospital Affiliated to Suzhou University, 188 Shizi Street, Suzhou 215006, China b Department of Orthopaedic Surgery, University of Toledo, 3065 Arlington Avenue, Toledo, OH 43614, USA c Department of Stem Cell Biology and Regenerative Medicine, Cleveland Clinic, 9500 Euclid Avenue, Cleveland, OH 44195, USA Received 27 May 2009; revised 28 July 2009; accepted 21 August 2009
Abstract
BACKGROUND CONTEXT: Although several authors have already reported on the high local recurrence rate of sacral chordomas after surgical resection, there are no reports on the risk factors for recurrence after resection when combined with preoperative tumor-related blood vessel embolism by digital subtraction angiography (DSA) technique. PURPOSE: To investigate the factors related to the continuous disease-free survival time (CDFS) after the resection of sacral chordomas combined with embolization. STUDY DESIGN/SETTING: Retrospective review of the signs, images, and immunohistochemical data of patients with sacral chordomas treated with an initial operation combined with transcatheter arterial embolization. PATIENT SAMPLE: Twenty-two patients with sacral chordomas received initial resection combined with transcatheter arterial embolization. OUTCOME MEASURES: Recurrence, proliferating cell nuclear antigen (PCNA) expression, basic fibroblast growth factor (bFGF) expression, CDFS. METHODS: All cases were selected and followed for an average of 39.2 months. The roles of gender, age, tumor size, tumor location, surgical method, radiation therapy, PCNA expression, and bFGF expression in local recurrence were analyzed using the log-rank test. RESULTS: Sacral chordomas recurred in eight of 22 cases. The CDFS was significantly greater in tumors located below S3 as compared with those above S3. When evaluating PCNA and bFGF expression levels, the CDFS was greater in low expressions rather than high expressions. It was determined that the surgical method used was of prognostic significance to the CDFS. CONCLUSIONS: Higher tumor location and higher expressions of PCNA and bFGF will lead to a shorter CDFS. Resecting the tumor as completely as possible will decrease the chances of local recurrence of sacral chordomas. Published by Elsevier Inc.
Keywords:
Sacrum; Chordoma; Recurrence; Embolization; PCNA; bFGF; Continuous disease-free survival time
Introduction Sacral chordoma is a rare slow-growing malignant neoplasm arising from cellular remnants of the notochord. Because of its insidious onset and ambiguous symptoms, by the time the tumor has been detected, it is often already FDA device/drug status: not applicable. Author disclosures: none. * Corresponding author. Department of Orthopaedic Surgery, University of Toledo, 3065 Arlington Ave., Toledo, OH 43614, USA. Tel.: (419) 3836558; fax: (419) 383-3526. E-mail address: [email protected] (J. Liu) 1529-9430/09/$ – see front matter Published by Elsevier Inc. doi:10.1016/j.spinee.2009.08.447
large in size and invading the sacral nerves. Radiotherapy and chemotherapy are ineffective, making surgery the main treatment for sacral chordomas. Because of the anatomic characteristics of the region, removing a sacral tumor was once regarded as a difficult procedure with massive blood loss and severe surgical injury [1–5]. If the primary tumor is not completely resected, controlling the disease becomes more difficult, increasing the risk of recurrence. Although the local recurrence rate of sacral chordomas after surgical resections has been discussed in several studies [4,6–9], the present authors have not found any reports on the risk of recurrence after resection with preoperative
H. Yang et al. / The Spine Journal 9 (2009) 972–980
tumor-related blood vessel embolism by digital subtraction angiography (DSA) technique. The present study reviewed initial operations, combined with transcatheter arterial embolization (TAE), of 22 sacral chordomas from January 1994 to September 2006. Significant improvements in outcome were made when using this method. Symptoms, signs, images, and immunohistochemical data were collected to investigate the factors related to the continuous disease-free survival time (CDFS) after the resection of sacral chordoma. Materials and methods Patient selection and data collection There were a total of 22 patients (10 men and 12 women) in the present study. The average age was 47.1613.7 years (range: 22–76 years). Between 1994 and 2006, they were diagnosed and treated, selected for the present study, and their patient charts were reviewed. Original paraffin specimens were available from 16 of the 22 patients. These were immunohistochemically prepared for the detection of basic fibroblast growth factor (bFGF) and proliferating cell nuclear antigen (PCNA). Age at the time of diagnosis, gender, tumor location and size, surgical method, radiation therapy, PCNA and bFGF expression levels, and local recurrence were obtained. This information was encoded to establish a database on an Excel spreadsheet. The CDFS was calculated according to the Kaplan-Meier method and used as the index of recurrence. All data were accurately imported to SPSS 13.0 (SPSS Inc., Chicago, IL, USA). Embolization technique Local anesthesia was given over the level of the femoral artery, and a catheter was inserted in the bilateral internal iliac arteries. Using the DSA technique, arterial embolization was performed preoperatively across the abdominal aorta using a combination of Gelfoam, 35% to 40% omnipaque, and saline solution to create a cylindrical shaped embolus (222 mm3), which blocked all tumor-supplying vessels including the iliac arteries bilaterally, the middle sacral artery, and the numerous small tumor vessels. Surgical technique The optimal time to perform the surgery is within 24 hours of embolization. After general anesthesia was given, patients were placed in the prone position. An ‘‘I’’-shaped incision was made that was extensive enough to ensure the removal of the tumor. The posterior, caudal, and bilateral aspects of the tumor were excised either radically or marginally. The anterior aspect was separated by blunt dissection through retroperitoneal interstitial spaces. The sacral nerves were protected and preserved whenever possible. However, if nerve roots were contaminated, the membrane of these nerves was dissected.
973
Context Risk factors for local recurrence of sacral chordomas following combined surgical resection and preoperative embolization are poorly understood. Contribution In this retrospective series, the authors found three factors associated with recurrence: tumors located cephalad to S3; tumors with high expression of PCNA and bFGF; and tumors treated operatively by intralesional resection. Implications This article provides useful prognostic information for patients with these rare tumors. —The Editors Paraffin specimens with hematoxylin stain and the examination of bFGF and PCNA Sixteen original paraffin specimens of sacral chordomas from 1994 to 2006 were examined in the present study. Basic fibroblast growth factor expression was detected using the immunohistochemistry streptavidin peroxidase (SP) method. Stain results were evaluated as follows: the pigmentation of the renal carcinoma specimen slice with antibody-1 was defined as positive. In contrast, the pigmentation of the renal carcinoma specimen slice without antibody-1 was defined as negative. When positive, the stain of the cell expressing bFGF was brownish yellow with prominent plasma and small vessel walls. Proliferating cell nuclear antigen expression was detected using the immunohistochemistry SP method. Stain results were evaluated as follows: the pigmentation of the colon carcinoma specimen slice with antibody-1 was defined as positive. In contrast, the pigmentation of the colon carcinoma specimen slice without antibody-1 was defined as negative. When positive, the stain of the cell expressing PCNA revealed brownish yellow granulation. Proliferating cell nuclear antigen and bFGF expression levels were scored based on the reagents’ instruction booklets. A four-tier system was used to classify the expression intensities. The positive index number was calculated as the positive cell number divided by the observed cell number multiplied by 100%. A positive index number from 0% to 25% was defined as Grade 1, 26% to 50% was Grade 2, 51% to 75% was Grade 3, and 76% to 100% was Grade 4. Grades 1 and 2 were classified as low expressions, and Grades 3 and 4 were classified as high expressions. Statistical analysis The dependent variable was the patients’ postoperative CDFS. The independent variables were age at the time of diagnosis, gender, tumor location and size, surgical method,
974
H. Yang et al. / The Spine Journal 9 (2009) 972–980
radiation therapy, PCNA expression, and bFGF expression. The log-rank test was used to perform statistical analysis. The Kaplan-Meier estimator was used to create the CDFS curve. The level of statistical significance was p!.05. Results General data Intraoperative blood loss ranged from 200 to 4,500 mL (mean: 1,200 mL). The average follow-up time was 38.2632.0 months (range: 6–144 months). Eight of the 22 cases experienced local recurrence (36.4%). Postoperative neurologic function In 17 patients, the sacral nerve root was reserved bilaterally at and above S3, and the sphincter muscle function of the bladder and bowel was normal. In two cases, the nerve root was reserved unilaterally at and above S3, resulting in a partial dysfunction of the bladder and bowel. The function of the sphincter muscle was impaired in two patients with nerve roots reserved at and above the S1 level. In one case, colostomy and ureterocutaneostomy were performed. Pathohistological data There was one case of chondroid chordoma among the 16 hematoxylin (HE)-stained specimens. The other 15 cases
were typical chordomas (Fig. 1). Scores for PCNA expression were divided into two groups: high expression (six cases) and low expression (10 cases) (Fig. 2). The average score (X6s) for PCNA expression was 2.460.8. Cells positive for PCNA included droplet cells, asteriated cells, and fusiform cells. Interstitial cells were not positive for PCNA. Scores for bFGF expression were divided into two groups: high expression (four cases) and low expression (12 cases) (Fig. 3). The average score (X6s) for bFGF expression was 2.360.7. Relevant factors of sacral chordoma recurrence The results of the log-rank analysis indicated that tumor locations, surgical methods (Fig. 4), PCNA expression (Fig. 5), and bFGF expression contributed to statistical significance in the CDFS. Gender, age, tumor size, and radiation therapy did not have a statistically significant impact on the CDFS. For details on the relevant factors of sacral chordoma recurrence and the CDFS median, see Tables 1 and 2. Discussion The bilateral internal iliac artery and the median sacral artery are the blood supply of the sacrum with numerous collateral branches between these arteries and the superior gluteal artery. Because of the short distance between the sacral tumor and the main vessels, severe blood loss often occurs
Fig. 1. (Top Left) Chondroid chordoma (hematoxylin, 100): tumor cells form sheets and clusters separated by fibrous bands. The composition of cartilage cells is apparent in the tumor tissue. Chondroid stroma appears to be grayish blue. (Top Right) Chondroid chordoma (hematoxylin, 400): the tumor cells appear to be circular and ovaloid. Uneven vacuoles are present in the plasma. Chondroid stroma appears to be grayish blue. (Bottom Left) Typical sacral chordoma (hematoxylin, 100): tumor cells are separated by fibrous bands. The tumor cells are scattered in myxoid stroma and arranged in strips. Plasma contains uneven vacuoles. (Bottom Right) Typical sacral chordoma (hematoxylin, 400): tumor cells are circular with a clear border. Uneven vacuoles are scattered in the plasma. Tumor cells are spread in myxoid stroma.
H. Yang et al. / The Spine Journal 9 (2009) 972–980
975
Fig. 2. PCNA-positive expression (streptavidin peroxidase, 200). (Top Left) Grade 1: 0% to 25%. (Top Right) Grade 2: 26% to 50%. (Bottom Left) Grade 3: 51% to 75%. (Bottom Right) Grade 4: 76% to 100%.
intraoperatively. This can be significantly reduced with TAE. In the present study, intraoperative blood loss ranged from 200 to 4,500 mL (mean: 1,200 mL). Previously published studies that did not use embolization reported significantly greater blood loss [10,11]. It has been noted that the optimal
time to perform surgery is within 24 hours after embolization. This is based on the observation, by the principle author, of the reestablishment of the collateral circulation after embolizing the bilateral internal iliac arteries with gelfoam particles in dogs. Twelve hours after embolization, a few
Fig. 3. bFGF-positive expression (streptavidin peroxidase, 200). (Top Left) Grade 1: 0% to 25%. (Top Right) Grade 2: 26% to 50%. (Bottom Left) Grade 3: 51% to 75%. (Bottom Right) Grade 4: 76% to 100%.
976
H. Yang et al. / The Spine Journal 9 (2009) 972–980
Fig. 4. CDFS curves. (Top Left) Based on age. (Top Right) Based on gender. (Bottom Left) Based on tumor location. (Bottom Right) Based on diameter of the tumor.
branches of collateral circulation could be observed. This increased at 24 hours after embolization. At 48 hours, the collateral circulation was abundant [12]. Sacral chordoma is a low-grade malignant neoplasm that originates from notochordal remnants and occurs exclusively in the axial skeleton. Approximately 50% of chordomas are located in the sacrococcygeal region. Sacral chordomas account for 4% of malignant bone tumors. After the removal of these tumors, local recurrence may occur, which is significantly influenced by several factors. Tumor location is an important factor in recurrence. Sacral tumors normally extend from the distal part of the sacrum to the proximal part, S1 or S2. If located in S3, the diameter of the tumor is usually smaller than the diameter of a tumor located in S2 or S1. Similarly, S2 is usually smaller than S1. In addition, the more proximal the tumor, the more it penetrates the cortex of the sacrum, leading to a higher recurrence rate. Statistically significant differences were found between sacral chordomas located at or below S3 and sacral chordomas located above S3.
Most tumors located above S3 require radical resections, involve a large amount of blood loss, have easily damaged tumor borders, and involve resections or preservations of sacral nerves. Thus, sacral chordomas at these levels are often difficult to resect completely and require numerous surgeries. In contrast, given that the infiltration is relatively small for sacral chordomas at or below S3, operations for these tumors generally do not involve sacral nerve resections. The tumors at these levels can usually be completely resected, dramatically decreasing the recurrence rate. In a review of 23 patients with sacral chordomas by Cheng et al. [13], the CDFS median for tumors above S3 was significantly less than the CDFS median for tumors at or below S3. The present study included 10 cases of tumors at or below S3 (CDFS median: 82 months) and 12 cases above S3 (CDFS median: 52 months). Two of the cases at or below S3 and six of the cases above S3 experienced recurrence. There was a statistically significant difference between the CDFS medians for the two groups (p5.026).
H. Yang et al. / The Spine Journal 9 (2009) 972–980
977
Fig. 5. CDFS curves. (Top Left) Based on surgical method. (Top Right) Based on radiation therapy. (Bottom Left) Based on low expression versus high expression of PCNA. (Bottom Right) Based on low expression versus high expression of bFGF.
Currently, surgical resections are the primary treatment of sacral chordomas [2,8]. Sundaresan et al. [14] observed 34 patients with sacral chordomas and reported that the postoperative local recurrence rate is directly related to the scope and degree of the resection. Fuchs et al. agreed that there is a significant relationship between the local recurrence rate and the margin achieved during surgery. They found that a wide surgical margin is the most important predictor of survival and local recurrence [4]. Cases treated with radical resections have a longer postoperative period without local recurrence than those treated with perilesional
resections (p!.05) [2,6,15]. The reason for this is that the tumor walls of patients with perilesional or intralesional resection are damaged, causing implantation of tumor cells in other areas and increasing the recurrence rate considerably, particularly with intralesional resections. Kaiser et al. [16] stated that the local recurrence rate after the radical resection of an entire tumor was 28%, whereas the local recurrence rate after an intralesional resection was 64%. Several authors have reported findings in agreement with this [5,17]. In the present study, 8 of the 22 cases had a radical resection, one of which experienced
Table 1 Relevant factors of sacral chordoma recurrence Related factors
Age
Result Recurrence No recurrence CDFS (mo) c2 value P value
!50 4 9 82 2.316 .128
Gender $50 4 5 52
M 3 7 58 0.071 .790
CDFS, continuous disease-free survival time; F, female; M, male.
Location F 5 7 82
I 2 8 82 4.982 .026
Size (mm) II 6 6 52
!100 5 10 82 1.093 .296
$100 3 4 44
978
H. Yang et al. / The Spine Journal 9 (2009) 972–980
Table 2 Relevant factors of sacral chordoma recurrence Related factors
Surgery methods
Result Recurrence No recurrence CDFS (mo) c2 value P value
I 6 4 24 9.126 .003
II 1 3 44
Radiation III 1 7 82
Yes 3 8 82 2.101 .147
PCNA No 5 6 52
I 3 7 58 4.066 .044
bFGF II 4 2 22
I 4 8 58 4.732 .030
II 3 1 22
CDFS, continuous disease-free survival time. Key—tumor location: I5S3 and below S3, II5above S3; surgical methods: I5intralesional, II5perilesional, III5radical; PCNA (proliferating cell nuclear antigen): I5low expression, II5high expression; bFGF (basic fibroblast growth factor): I5low expression, II5high expression.
recurrence (12.5%); 4 cases involved a perilesional resection, one of which experienced recurrence (25%); 10 cases had an intralesional resection, six of which experienced recurrence (60%). The lower recurrence rate for radical resections (12.5%) compared with that reported by other studies (Kaiser et al. reported 28%) may be attributed to the preoperative application of tumor-related blood vessel embolism by DSA technique. Reducing the amount of blood loss during surgery is the key measure to removing the tumor en bloc. Preoperative TAE for other types of tumors has been proven to reduce blood loss significantly and maintain a clean surgical field, both of which aid in the complete removal of the tumor [18–20]. Among the present cases, the CDFS medians were 82 months for radical resections, 44 months for perilesional resections, and 24 months for intralesional resections. Statistically significant differences were found in the CDFS medians between the various surgical methods (p5.003). Usually, the combination of the activation of various tumor genes and the inactivation of tumor-inhibiting genes causes malignant tumors to occur and proliferate. Proliferating cell nuclear antigen is a protein that acts as a processivity factor for DNA polymerase delta in eukaryotic cells. Because DNA polymerase delta is involved in synthesizing new DNA that replaces excised damaged DNA during DNA repair, PCNA is necessary for both DNA synthesis and DNA repair. Proliferating cell nuclear antigen is directly related to the generation of tumor vessels and has important effects on the tumor’s malignancy and biological behavior. Therefore, PCNA can be used as an index for tumor DNA proliferation. Proliferating cell nuclear antigen expression among the specimens of sacral chordoma is a valuable clinical guide for tumor recurrence [21]. Among the 16 cases on which the immunohistochemistry SP method was performed in the present study, 10 cases showed low expressions of PCNA (CDFS median: 58 months), of which three experienced recurrence; six cases showed high expressions of PCNA (CDFS median: 22 months), of which four experienced recurrence. The results of the statistical analysis indicated that the cases in the high expression group had significantly shorter CDFS medians than those in the low expression group (Fig. 5).
During tumor development, bFGF mediates the formation of new blood vessels, a process known as angiogenesis. Basic fibroblast growth factor also stimulates abnormal cell division and proliferation, as well as a variety of tumor development and malignant transformations through autocrine and paracrine processes [22,23]. Deniz et al. [24] found statistically significant differences in the recurrence rates between the high and low expression groups of bFGF (p!.05). Among the 16 cases on which the immunohistochemistry SP method was performed in the present study, 12 cases showed low expressions of bFGF (CDFS median: 58 months), of which four experienced recurrence; four cases showed high expressions of bFGF (CDFS median: 22 months), of which three experienced recurrence. The results of the statistical analysis indicated that the cases in the high expression group had significantly shorter CDFS medians than those in the low expression group. In the study by Thieblemont et al. [25], age was found to be an independent factor of postoperative recurrence and CDFS. The present study categorized patients into two groups based on age. Thirteen patients were younger than 50 years, four of which experienced recurrence (CDFS median: 82 months). Nine patients were 50 years and older, and four patients in this group experienced recurrence (CDFS median: 82 months). Statistical analysis of the CDFS for each group did not find them to be significantly different. The ratio of incidence rate of sacral chordomas in men versus women is 2:1 [26,27]. Although gender has been considered a possible factor of postoperative recurrence, a comparison of these studies reveals an inconsistency regarding the effect of gender on the CDFS [25,28]. Of the 22 cases involved in the present study, 10 were men (CDFS median: 58 months) and 12 were women (CDFS median: 82 months). Three of the male patients and five of the female patients experienced recurrence. There was no statistically significant difference between the CDFS medians for the men and the women. Tumor size is one of the factors that reflects the speed of tumor growth, but its effect on tumor recurrence has been contended [2,29]. The average diameter of the tumors among the 22 patients in the present study was 95 mm. Fifteen cases had tumors with diameters !100 mm, of which
H. Yang et al. / The Spine Journal 9 (2009) 972–980
five had recurred (CDFS median: 82 months). Seven cases had tumors with diameters $100 mm, of which four had recurred (CDFS median: 44 months). The present authors believe that these recurrence rates are the result of using the DSA technique to embolize the bilateral internal iliac arteries and the tumor supply vessels, thereby increasing the complete resection rate (Fig. 4). The use of the DSA embolism technique during surgery significantly decreased hemorrhaging and allowed for the resection of larger tumors. Although radiation therapy is a common supplemental treatment for patients after the resection of their tumors, it is debatable as to whether it can reduce local recurrence [2,4,5,9–11,13,14,24,30–32]. In the present study, postoperative radiation was found to have no significant influence on the patient’s overall survival time. Eleven of the 22 patients received postoperative radiation therapy, and three experienced recurrence (CDFS median: 82 months). The remaining 11 cases did not receive radiation therapy, and five experienced recurrence (CDFS median: 52 months). There was no statistically significant difference between the CDFS medians of the two groups. The relationship between the recurrence rate and the maintenance of the sacral root function has not been thoroughly explored. If the nerve root is saved, function is preserved but the recurrence rate may be higher and the CDFS lower. In contrast, if the nerve root is completely removed, nerve root function is lost but the recurrence rate may be lowered. Therefore, the present authors believe that a balance must be achieved between nerve root function and tumor recurrence. The physician must discuss the options thoroughly with the patient, helping to make the appropriate decisions. The improvement in technique as the surgeon’s experience increases may play some role in the recurrence rate. However, the present study used a small group of patients with the same principle surgeon performing all operations. The difference in technique between the first case and the last case carried out by the principle surgeon may not have been significant. The present authors could not draw conclusions about whether the surgeon’s experience influenced the recurrence rate. Conclusions Patients treated with radical resections had a significantly longer CDFS than patients treated with perilesional and intralesional resections. In the present study, a trend was discovered in which the higher the location of the tumor, the shorter the CDFS. In addition, greater expressions of PCNA and bFGF led to a shorter CDFS. Patient’s age at the time of diagnosis, gender, tumor size, and radiation therapy were found to have no statistical significance on the CDFS. References [1] Samson IR, Springfield DS, Suit HD, Mankin HJ. Operative treatment of sacrococcygeal chordoma. A review of twenty-one cases. J Bone Joint Surg Am 1993;75:1476–84.
979
[2] York JE, Kaczaraj A, Abi-Said D, et al. Sacral chordoma: 40-year experience at a major cancer center. Neurosurgery 1999;44:74–9; discussion 79–80. [3] Yonemoto T, Tatezaki S, Takenouchi T, et al. The surgical management of sacrococcygeal chordoma. Cancer 1999;85:878–83. [4] Fuchs B, Dickey ID, Yaszemski MJ, et al. Operative management of sacral chordoma. J Bone Joint Surg Am 2005;87:2211–6. [5] Hulen CA, Temple HT, Fox WP, et al. Oncologic and functional outcome following sacrectomy for sacral chordoma. J Bone Joint Surg Am 2006;88:1532–9. [6] Bergh P, Kindblom LG, Gunterberg B, et al. Prognostic factors in chordoma of the sacrum and mobile spine: a study of 39 patients. Cancer 2000;88:2122–34. [7] Berven S, Zurakowski D, Mankin HJ, et al. Clinical outcome in chordoma: utility of flow cytometry in DNA determination. Spine 2002;27:374–9. [8] Boriani S, Chevalley F, Weinstein JN, et al. Chordoma of the spine above the sacrum. Treatment and outcome in 21 cases. Spine 1996;21:1569–77. [9] Conlon KC, Boland PJ. Laparoscopically assisted radical sacrococcygectomy. A new operative approach to large sacrococcygeal chordomas. Surg Endosc 1997;11:1118–22. [10] Simpson AH, Porter A, Davis A, et al. Cephalad sacral resection with a combined extended ilioinguinal and posterior approach. J Bone Joint Surg Am 1995;77:405–11. [11] Tomita K, Tsuchiya H. Total sacrectomy and reconstruction for huge sacral tumors. Spine 1990;15:1223–7. [12] Yang H, Ni C, Tang T. Surgical treatment of sacral tumors after transcatheter arterial embolization. Chinese J Orthop 1998;18:646–8. [13] Cheng EY, Ozerdemoglu RA, Transfeldt EE, Thompson RC Jr. Lumbosacral chordoma. Prognostic factors and treatment. Spine 1999;24: 1639–45. [14] Sundaresan N, Huvos AG, Krol G, et al. Surgical treatment of spinal chordomas. Arch Surg 1987;122:1479–82. [15] Fourney DR, Gokaslan ZL. Current management of sacral chordoma. Neurosurg Focus 2003;15:E9. [16] Kaiser TE, Pritchard DJ, Unni KK. Clinicopathologic study of sacrococcygeal chordoma. Cancer 1984;53:2574–8. [17] Puri A, Agarwal MG, Shah M, et al. Decision making in primary sacral tumors. Spine J 2009;9:396–403. [18] Gellad FE, Sadato N, Numaguchi Y, Levine AM. Vascular metastatic lesions of the spine: preoperative embolization. Radiology 1990;176: 683–6. [19] Hilal SK, Michelsen JW. Therapeutic percutaneous embolization for extra-axial vascular lesions of the head, neck, and spine. J Neurosurg 1975;43:275–87. [20] Ellman BA, Parkhill BJ, Curry TS III, et al. Ablation of renal tumors with absolute ethanol: a new technique. Radiology 1981;141:619–26. [21] Essers J, Theil AF, Baldeyron C, et al. Nuclear dynamics of PCNA in DNA replication and repair. Mol Cell Biol. 2005;25:9350–9. [22] Chandler LA, Sosnowski BA, Greenlees L, et al. Prevalent expression of fibroblast growth factor (FGF) receptors and FGF2 in human tumor cell lines. Int J Cancer 1999;81:451–8. [23] Griffioen AW, Damen CA, Mayo KH, et al. Angiogenesis inhibitors overcome tumor induced endothelial cell anergy. Int J Cancer 1999;80:315–9. [24] Deniz ML, Kilic¸ T, Almaata I, et al. Expression of growth factors and structural proteins in chordomas: basic fibroblast growth factor, transforming growth factor alpha, and fibronectin are correlated with recurrence. Neurosurgery 2002;51:753–60. [25] Thieblemont C, Biron P, Rocher F, et al. Prognostic factors in chordoma: role of postoperative radiotherapy. Eur J Cancer 1995;31A: 2255–9. [26] Ashwood N, Hoskin PJ, Saunders MI. Metastatic chordoma: pattern of spread and response to chemotherapy. Clin Oncol (R Coll Radiol) 1994;6:341–2.
980
H. Yang et al. / The Spine Journal 9 (2009) 972–980
[27] Forsyth PA, Cascino TL, Shaw EG, et al. Intracranial chordomas: a clinicopathological and prognostic study of 51 cases. J Neurosurg 1993;78:741–7. [28] O’Connell JX, Renard LG, Liebsch NJ, et al. Base of skull chordoma. A correlative study of histologic and clinical features of 62 cases. Cancer 1994;74:2261–7. [29] Baratti D, Gronchi A, Pennacchioli E, et al. Chordoma: natural history and results in 28 patients treated at a single institution. Ann Surg Oncol 2003;10:291–6.
[30] Chandawarkar RY. Sacrococcygeal chordoma: review of 50 consecutive patients. World J Surg 1996;20:717–9. [31] Catton C, O’Sullivan B, Bell R, et al. Chordoma: long-term follow-up after radical photon irradiation. Radiother Oncol 1996;41:67–72. [32] Fourney DR, Rhines LD, Hentschel SJ, et al. En bloc resection of primary sacral tumors: classification of surgical approaches and outcome. J Neurosurg Spine 2005;3:111–22.