Abstract

CancerVolume 88, Issue 9 p. 2122-2134 Original ArticleFree Access Prognostic factors in chordoma of the sacrum and mobile spine A study of 39 patients Peter Bergh M.D., Corresponding Author Peter Bergh M.D. Department of Orthopedic Surgery, Musculoskeletal Tumor Center, Sahlgrenska University Hospital, Göteborg, SwedenDepartment of Orthopedic Surgery, Sahlgrenska University Hospital, Göteborg University, SE-413 45 Göteborg, Sweden===Search for more papers by this authorLars-Gunnar Kindblom M.D., Ph.D., Lars-Gunnar Kindblom M.D., Ph.D. Department of Pathology, Musculoskeletal Tumor Center, Sahlgrenska University Hospital, Göteborg, SwedenSearch for more papers by this authorBjörn Gunterberg M.D., Ph.D., Björn Gunterberg M.D., Ph.D. Department of Orthopedic Surgery, Musculoskeletal Tumor Center, Sahlgrenska University Hospital, Göteborg, SwedenSearch for more papers by this authorFabrizio Remotti M.D., Fabrizio Remotti M.D. Department of Pathology, Musculoskeletal Tumor Center, Sahlgrenska University Hospital, Göteborg, SwedenSearch for more papers by this authorWalter Ryd M.D., Ph.D., Walter Ryd M.D., Ph.D. Department of Pathology, Musculoskeletal Tumor Center, Sahlgrenska University Hospital, Göteborg, SwedenSearch for more papers by this authorJeanne M. Meis-Kindblom M.D., Jeanne M. Meis-Kindblom M.D. Department of Pathology, Musculoskeletal Tumor Center, Sahlgrenska University Hospital, Göteborg, SwedenSearch for more papers by this author Peter Bergh M.D., Corresponding Author Peter Bergh M.D. Department of Orthopedic Surgery, Musculoskeletal Tumor Center, Sahlgrenska University Hospital, Göteborg, SwedenDepartment of Orthopedic Surgery, Sahlgrenska University Hospital, Göteborg University, SE-413 45 Göteborg, Sweden===Search for more papers by this authorLars-Gunnar Kindblom M.D., Ph.D., Lars-Gunnar Kindblom M.D., Ph.D. Department of Pathology, Musculoskeletal Tumor Center, Sahlgrenska University Hospital, Göteborg, SwedenSearch for more papers by this authorBjörn Gunterberg M.D., Ph.D., Björn Gunterberg M.D., Ph.D. Department of Orthopedic Surgery, Musculoskeletal Tumor Center, Sahlgrenska University Hospital, Göteborg, SwedenSearch for more papers by this authorFabrizio Remotti M.D., Fabrizio Remotti M.D. Department of Pathology, Musculoskeletal Tumor Center, Sahlgrenska University Hospital, Göteborg, SwedenSearch for more papers by this authorWalter Ryd M.D., Ph.D., Walter Ryd M.D., Ph.D. Department of Pathology, Musculoskeletal Tumor Center, Sahlgrenska University Hospital, Göteborg, SwedenSearch for more papers by this authorJeanne M. Meis-Kindblom M.D., Jeanne M. Meis-Kindblom M.D. Department of Pathology, Musculoskeletal Tumor Center, Sahlgrenska University Hospital, Göteborg, SwedenSearch for more papers by this author First published: 20 November 2000 https://doi.org/10.1002/(SICI)1097-0142(20000501)88:93.0.CO;2-1Citations: 400AboutSectionsPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Share a linkShare onFacebookTwitterLinked InRedditWechat Abstract BACKGROUND The prognosis of patients with chordoma of the sacrum and mobile spine has been reported to be dismal and attributable in the majority of cases to intralesional surgery. The purpose of this study was to evaluate the clinical outcome of these patients using modern surgical principles aimed at complete resection and to identify prognostic factors. METHODS The clinical and morphologic features, type of surgery, and follow-up of 39 consecutive patients with chordoma were reviewed and analyzed statistically. RESULTS Thirty sacral and 9 mobile spine chordomas (size range, 3–20 cm; mean, 8 cm) occurring in 22 women and 17 men (median age, 55 years) were analyzed. The preoperative morphologic diagnosis was based on fine-needle aspiration (FNA) biopsy, core needle biopsy, or incisional biopsy. The final surgical margins were wide in 23 patients and marginal or intralesional in 16. The mean follow-up was 8.1 years (range, 0.1–23 years). Seventeen patients (44%) developed local recurrences and 11 patients (28%) developed metastases. The estimated 5-, 10-, 15-, and 20-year survival rates were 84%, 64%, 52%, and 52%, respectively. Local recurrence was associated significantly with an increased risk of metastasis and tumor-related death (P 5%, and local recurrence were found to be adverse prognostic factors. FNA is the preferred method for establishing the preoperative morphologic diagnosis of chordoma. Cancer 2000;88:2122–34. © 2000 American Cancer Society. Chordoma is a rare malignant bone tumor that originates from notochordal remnants and occurs exclusively in the axial skeleton, with a predilection for the sacrum, base of the skull, and occasionally the mobile spine.1-4 Tumor location and large size at presentation often have precluded complete surgical removal in the past, and the long term prognosis typically has been poor.3, 5-8 The poor outcome primarily has been due to extensive local recurrences and secondary complications as well as metastases that occur in up to 60% of sacral-vertebral chordomas.1-3, 9, 10 The principles and techniques of major surgical sacral resections and complete removal of one or more vertebrae of the mobile spine were described over 20 years ago along with evaluation of the mechanical, functional, and physiologic results.11-17 The purpose of this study was to evaluate the clinical outcome as well as complications in chordoma patients whose tumors were removed using modern surgical principles and techniques, such as sacral amputation and en bloc vertebral resection, and to identify prognostic factors. MATERIALS AND METHODS A consecutive series of 39 patients with surgically treated sacral and vertebral chordomas was analyzed. All patients were treated at the Department of Orthopedic Surgery, Sahlgrenska University Hospital, Göteborg, Sweden; 38 patients were treated from 1968 to 1998 and one patient in 1963. Patient age, gender, symptoms, time from onset of symptoms to diagnosis, neurologic function, tumor size and location, type of preoperative diagnostic procedure, and treatment were recorded in all cases as well as follow-up information regarding adjuvant therapy, postoperative sequelae, local recurrences, metastases, and survival. The microscopic appearances of all preoperative fine-needle aspiration (FNA), core needle biopsies, open surgical biopsies, and resections of the primary and recurrent tumors were reviewed. The following histologic factors were systematically recorded: predominance of epithelioid cells versus physaliferous cells; degree of cellularity, pleomorphism and spindling (rated as absent, low, moderate, or high for each factor); presence or absence of microscopic tumor necrosis (defined as one high-power field [HPF] or more of necrosis, with one HPF = 1.13 mm2); mitotic activity (low, moderate, or high corresponding to 5 mitoses/10 HPF, respectively); and assessment of proliferative activity using Ki67 (MIB1) in formalin fixed, paraffin embedded tissues (rated as low, moderate, or high corresponding to 5%, respectively). Immunostains for cytokeratins (KL-1, AE1/AE3, CAM 5.2), epithelial membrane antigen (EMA), vimentin, S-100 protein, Ki67 (MIB1), and p53 were performed on formalin fixed, paraffin embedded tissues in 38 cases and FNA specimens in 12 cases. For the statistical analysis, three end points were used: local recurrence, metastasis, and death. Univariate analysis was performed with a nonparametric test.18 The Poisson model19, 20 was used to evaluate whether the time dependent variables, local recurrence and metastasis, influenced survival. The multivariate analysis was performed using the Cox regression model with stepwise forward procedure. The following factors were evaluated: type of diagnostic procedure (FNA, core needle biopsy, incisional biopsy, excision), performance of an invasive diagnostic procedure at our tumor center versus outside our center, surgical margins of the initial surgical resection, primary or secondary treatment at our tumor center, surgical margins at definitive surgery, tumor size, presence of microscopic tumor necrosis, individual morphologic features (see "Results"), and Ki67 immunolabeling index. RESULTS Patients and Presenting Symptoms Of 39 patients, 22 were women and 17 were men. The mean age at diagnosis was 55 years (range, 25–81 years). Thirty tumors were located in the sacrum and nine in the mobile spine (five cervical, one thoracic, and three lumbar) (Tables 1 and 2). Table 1. Clinical Data on Patients with Spinal Chordoma Case no. Age (yrs) Gender Location Surgical margins at center Time to first local recurrence (yrs) Time to metastases (yrs) Final disease status Follow-up time (yrs) 1 62 M CII Marginal, SP 9.4aa Operation at center for local recurrence. NED 12.4 2 25 M CII-CIII Intralesional, PP CDF 2.8 3 62 F CIII Intralesional, SP 3.6aa Operation at center for local recurrence. TRD 6.6 4 81 F CIII Intralesional, PP CDF, DOC 0.1 5 57 M CIII-CIV Intralesional, PP 2.7 2.3 AWD 4.6 6 40 M TI-TXII Wide, PP 2.0 TRD 2.1 7 51 F LIII Marginal, PP 10.0bb Decompressive surgery 1 year before surgery at center. DWD, DOC 10.5 8 59 F LIII Wide, SP NED, DOC 3.1 9 53 F LIV-SI Intralesional, SP 3.0aa Operation at center for local recurrence. NED, DOC 6.1 M: male; F: female; SP: secondary procedure; NED: no evidence of disease; PP: primary procedure; CDF: continuously disease free; TRD: tumor-related death; DOC: dead of other causes; DWD: dead with disease; AWD: alive with disease. a Operation at center for local recurrence. b Decompressive surgery 1 year before surgery at center. Table 2. Clinical Data on Patients with Sacral Chordoma Case no. Age (yrs) Gender Level of sacral amputation Nerves sacrificed right Nerves sacrificed left Surgical resection margins at center Time to first local recurrence (yrs) Time to metastases (yrs) Final disease status Follow-up time (yrs) 10 47 M SI S2-S5 S2-S5 Intralesional, SP 4.9aa Operation at center for local recurrence. TRD 9.4 11 58 M SI S2-S5 S2-S5 Wide, PP 13.2 TRD 15.2 12 52 F SI S1-S5 S1-S5 Intralesional, PP CDF, DOC 6.5 13 55 M S1 S1-S5 S1-S5 Wide, PP 10.1 DWD, DOC 12.2 14 45 F SI S1-S5 S1-S5 Wide, SP 9.6aa Operation at center for local recurrence. NED 21.5 15 42 M SI L5-S5 S1-S5 Intralesional, PP CDF 9.2 16 64 F SI-SII S2-S5 S2-S5 Wide, PP CDF, DOC 2.3 17 56 M SI-SII S2-S5 S2-S5 Wide, PP CDF 23.0 18 54 F SI-SII S2-S5 S2-S5 Wide, SP 5.5aa Operation at center for local recurrence. 5.4 TRD 7.2 19 53 M SI-SII S2-S5 S2-S5 Wide, PP CDF 13.1 20 66 F SI-SII S2-S5 S2-S5 Wide, PP CDF 11.1 21 70 F SI-SII S1-S5 S2-S5 Wide, PP 4.5 NED 6.3 22 58 M SI-SII S2-S5 S1-S5 Marginal, PP 1.4 0.2 TRD 1.9 23 67 F SI-SII S2-S5 S2-S5 Wide, SP CDF 5.1 24 34 F SI-SII S1-S5 S1-S5 Wide, PP CDF 4.1 25 69 F SI-SII S2-S5 S2-S5 Wide, PP CDF 2.6 26 72 F SI-SIIbb Indicates extent of tumor (treated by curettage). 0 0 Intralesional, PP CDF 2.7 27 61 M SI-SII S2-S5 S2-S5 Wide, PP CDF 1.0 28 52 M SII S2-S5 S2-S5 Wide, SP 4.9aa Operation at center for local recurrence. TRD 5.0 29 57 M SII S3-S5 S2-S5 Wide, PP 7.3cc Pulmonary metastasectomy. NED, DOC 17.9 30 32 M SII S3-S5 S3-S5 Wide, PP 12.7 13.3 AWD 18.0 31 63 F SII S2-S5 S2-S5 Wide, PP CDF 1.6 32 68 F SII-SIII S3-S5 S3-S5 Marginal, PP 7.1 DWD, DOC 10.4 33 74 M SII-SIII S3-S5 S3-S5 Intralesional, PP 1.6 2.1 TRD 2.7 34 42 F SIII S4-S5 S4-S5 Wide, PP CDF 21.8 35 47 M SIII S3-S5 S4-S5 Marginal, SP 7.1aa Operation at center for local recurrence. 7.9 TRD 11.0 36 60 F SIII-SIV S4-S5 S4-S5 Marginal, PP 1.8 5.3 DWD, DOC 12.8 37 62 F SIII-SIV S4-S5 S4-S5 Wide, PP CDF 7.0 38 50 F SIII-SIV S4-S5 S4-S5 Wide, PP 1.2 TRD 1.2 39 31 F SIV S5 S5 Wide, SP CDF 4.1 M: male; F: female; SP, secondary procedure; PP, primary procedure; TRD, tumor related death; CDF, continuously disease free; DOC, dead of other causes; DWD, dead with disease; mets, metastases; NED, no evidence of disease; AWD, alive with disease. a Operation at center for local recurrence. b Indicates extent of tumor (treated by curettage). c Pulmonary metastasectomy. Most patients (34/39) presented with pain. Five patients had a painless mass as the sole symptom. Twelve patients had neurologic symptoms with radiculopathy and/or bladder-bowel dysfunction. The average duration of symptoms before diagnosis was 2 years (range, 3 months to 9 years). Nine of 30 patients with sacral chordoma had a history of trauma to the sacrum 1–17 years before diagnosis. Preoperative Investigations Imaging studies All 39 patients were examined with plain roentgenograms, 13 with conventional tomography, 22 with computed tomography scan, and 19 with magnetic resonance imaging (MRI). Twelve patients were examined with angiography and eight patients with myelography. All tumors were osteolytic and sometimes difficult to detect on initial plain films. Computed tomography, and to a greater extent MRI, revealed considerable soft tissue extension in all but two cases (Figs. 1 and 2). None of the tumors was entirely confined to bone. Intratumoral calcifications occasionally were observed. Angiography showed that the tumors were poorly vascularized. Angiography and myelography were used to examine patients observed in the earlier part of this series and were at times helpful in determining tumor extent when vessels were displaced or the dura was compressed. Figure 1Open in figure viewerPowerPoint T1-weighted sagittal MR image of a sacral chordoma. White arrow indicates tumor extension in the sacral canal; black arrows indicate tumor extension into the gluteal subcutaneous fat. The tumor was removed en bloc, see Figure 3A. The patient is continuously disease free after 3 years. R: uterus. Figure 2Open in figure viewerPowerPoint T2-weighted sagittal magnetic resonance image of a cervical chordoma originating in the C4 vertebra, with tumor extension in the retropharyngeal space and spinal canal (arrows). The patient is alive with disease 4.6 years after intralesional surgery and stabilization followed by radiotherapy. He has no myelopathy. Preoperative diagnosis Twenty-nine patients were initially surgically treated at our tumor center. The preoperative diagnosis was based solely on FNA (16 cases), core needle biopsy (2 cases), combined core needle biopsy and FNA (2 cases), incisional biopsy alone (5 cases), and combined incisional biopsy and FNA (3 cases). One patient underwent surgery based solely on the clinical and radiographic findings. Of 10 patients whose first surgical resection was performed outside our tumor center, 3 had a preoperative diagnosis based on incisional biopsy, and the remaining 7 underwent surgery based on the clinical findings without morphologic confirmation. Treatment Surgery En bloc surgical removal (Figs. 1 and 3A ) was performed as the definitive surgical procedure at our center in 35 cases (6 vertebral, 29 sacral) with the objective to achieve at least a marginal resection. In the remaining four cases, the tumors were intentionally removed intralesionally (Fig. 2) (2 cervical, 1 thoracic, and 1 sacral tumor). In these four cases, a Cavitron Ultrasonic Surgical Aspirator (system 200; Valleylab. Inc., Boulder, CO) and a highspeed burr were used. Reconstruction with autologous bone grafts, screws, rods, and plates was performed in seven cases involving the mobile spine. The mean operation time was 9.7 hours, and median blood loss was 5.3 L. Surgically removed segments, sacrificed nerve roots, surgical margins, and outcome are given in Tables 1 and 2. Figure 3Open in figure viewerPowerPoint (a) Sacral chordoma, sagittal section of the tumor seen in Figure 1. The arrows correspond in both figures. (b and c) FNA appearance of chordoma demonstrating abundant intercellular myxoid matrix and large, polygonal cells with abundant cytoplasm (May-Grünwald Giemsa stain). Some of the cells have cytoplasmic vacuoles and represent physaliferous cells. (d) Immunohistochemical stain for cytokeratin in an FNA of a sacral chordoma. (e) Permeative invasion of bone characteristic of a sacral chordoma. (f) Vacuolated physaliferous cells predominated in this sacral chordoma. (g) Epithelioid cells with abundant eosinophilic cytoplasm impart a "hepatoid" appearance to this particular chordoma. (h) Some conventional chordomas had considerable nuclear pleomorphism and atypia as seen here. (i) A variable degree of focal spindling in conventional chordomas was observed in few cases; such foci should not be misconstrued as dedifferentiation. (j) A chordoma with microscopic tumor necrosis. Of 29 patients primarily treated at our center, wide surgical margins were achieved in 17 cases (16 sacral, 1 thoracic), marginal resection in 5 (3 sacral, 1 cervical, 1 lumbar), and intralesional excision in 7 (4 sacral, 3 cervical). The surgical resections were intralesional in all 10 chordomas primarily treated outside our center. After referral to our center, the final surgical margins were wide in six cases, marginal in one and intralesional in three cases. Radiotherapy and chemotherapy Two patients were treated primarily with radiotherapy after diagnostic incisional biopsies; one patient had a cervical chordoma and the other a sacral chordoma. Both of these patients underwent surgery later due to continued tumor growth. Five patients (four with sacral chordomas and one with a cervical chordoma) received adjuvant postoperative radiation therapy. Twelve patients received radiotherapy for local recurrences or skeletal metastases. None of the patients received chemotherapy. Morphology and Immunohistochemistry FNA Cytologically the FNAs were usually moderately to highly cellular, containing small and large cell groups as well as single cells. In May-Grünwald Giemsa stains, there was usually an abundant, purplish, sometimes fibrillar mucoid matrix in the background surrounding the cells or forming fragments. The primary cell type was large, round or polygonal, and had one or two centrally located nuclei and abundant pale cytoplasm that was well defined (Fig. 3B,C). The cytoplasm was frequently bubbly (as in the physaliferous cell), containing clear or occasionally mucin-filled vacuoles of various sizes and occasional granules. The chromatin was evenly distributed, and there were usually one or two nucleoli. Varying numbers of smaller, rounded epithelioid and short spindled and stellate cells also were observed. A diagnosis of chordoma was rendered in all cases. In the single case of "dedifferentiated" chordoma, large, bizarre, pleomorphic rounded to spindled cells arranged singly or in small groups with unevenly distributed chromatin and large nucleoli were observed in addition to the cytologic features of typical chordoma. Macroscopic and microscopic features The tumors ranged from 3 to 20 cm in size (median, 8 cm) and were usually well demarcated. All cases were located in bone and involved adjacent soft tissues, frequently with massive extension into adjacent skeletal muscle and subcutaneous adipose tissue (Fig. 3A). The cut surfaces of the tumors were characteristically soft, mucoid, and hemorrhagic. Yellow-gray friable foci (presumably necrotic) were occasionally observed. Histologic features are illustrated in Figure 3E–I. Most cases had an admixture of epithelioid and physaliferous cells. Physaliferous cells predominated in 21 cases (53%) and epithelioid cells in 18 cases (47%). Most tumors had either no (6 cases) or a low degree (25 cases) of nuclear pleomorphism. Seven cases were moderately pleomorphic, and only one, a dedifferentiated chordoma, was classified as highly pleomorphic due to marked overgrowth by a malignant fibrous histiocytoma like component. Only eight cases had noticeable spindling of the chordoid cells (these were not classified as dedifferentiated), thus simulating a high grade sarcoma at least focally. Mitotic activity was negligible in the overwhelming majority of cases, with 34 cases having low mitotic activity, 4 having a moderate amount, and only 1 having a high degree of mitotic activity (the latter being the dedifferentiated chordoma). Microscopic tumor necrosis was observed in 11 cases (28%). Immunohistochemical findings The 39 formalin fixed specimens and all 12 FNAs were stained strongly and diffusely for cytokeratins (Fig. 3D), EMA, and vimentin and somewhat more variably and weakly for S100 protein. The dedifferentiated component of one chordoma was stained only for vimentin. Immunostains for p53 were uniformly negative in the classic chordomas but positive in the sarcomatous component of the dedifferentiated chordoma. There was a wide range of Ki67 positively stained cells ( 5%) of proliferative activity. The single case of dedifferentiated chordoma had 40% positivity in the high grade sarcomatous component of the tumor. Ultrastructural features Ultrastructural examination was performed in 27 surgical specimens and in 6 FNAs. Evidence of epithelial differentiation, such as well developed desmosomes, tonofilament-like bundles of intermediate filaments, and intracytoplasmic lumina with microvilli, was characteristically observed in all cases. Additional findings included abundant glycogen, a prominent network of intermediate filaments, and cytoplasmic vacuoles. The abundant intercellular matrix contained scattered collagen fibers and numerous proteoglycan particles. Follow-Up Follow-up information was obtained in all cases. The mean follow-up interval was 8.1 years (range, 0.1–23 years; see Table 3). Table 3. Follow-Up of Patients with Sacral and Spinal Chordoma Final disease status No. of patients Mean follow-up time (yrs) Range of follow-up (yrs) CDF 14 7.8 1.0–23.0 CDF, DOC 3 3.0 0.1–6.5 NED 3 13.4 6.3–21.5 NED, DOC 3 9.0 3.1–17.9 AWD 2 4.6 and 18 4.6 and 18 DWD, DOC 4 11.5 10–13 TRD 10 6.2 1–15 Total 39 8.1 0.1–23.0 CDF: continuously disease free; DOC: dead of other causes; NED: no evidence of disease; AWD: alive with disease; DWD: dead with disease; TRD: tumor-related death. Complications and functional outcome Vertebral tumors (nine cases). Postoperative complications occurred in three of five patients with cervical chordoma. One of these patients had pharyngeal paresis and died 1 month postoperatively due to aspiration. Another patient required percutaneous gastric feeding for 6 months due to pharyngeal paresis but eventually recovered completely. The third patient developed a vocal cord palsy and Horner syndrome postoperatively. None of the five patients with cervical tumors developed weakness of the extremities. One patient with a lumbar chordoma developed weakness of the lower extremities, and another patient with a lumbar tumor became paraplegic 6 weeks postoperatively due to anticoagulant-related hemorrhage in the surgical area. The remaining two patients with thoracic and lumbar chordomas had no complications or loss of function. Sacral tumors (30 cases): fractures. Six of 18 patients with high sacral amputations (through the S1 vertebra or through the disc between S1 and S2) developed fatigue fractures in the remaining part of the sacrum (Fig. 4). Three of these patients had received radiotherapy. Four of six fractures healed spontaneously. Two of the six patients developed persistent pseudarthrosis, causing mild pain for one patient and disabling pain for the other despite surgery aiming at fusion. None of the 11 patients with resections through or below the S2 level developed fatigue fractures. Figure 4Open in figure viewerPowerPoint Pelvic radiogram with a fatigue fracture (arrow) after sacral amputation through the S1 vertebra for chordoma is shown. Sacral tumors (30 cases): bowel function. Rectal resection with a colostomy was performed for nine patients as part of the primary surgery. Two additional patients underwent a colostomy as a secondary procedure due to intolerable bowel dysfunction. The rectum was preserved in 19 patients. Ten of these had bilateral loss of the S2-S5 nerves and 8 of these 10 developed major bowel problems, i.e., incontinence or emptying difficulties. Four of the 19 patients lost the S3-S5 nerves bilaterally, and 2 of them developed major bowel problems. The remaining five patients with preserved S3 nerves had no significant bowel problems. Most patients eventually learned to cope with their bowel problems through dietary precautions. Sacral tumors (30 cases): bladder function. Nineteen of 30 patients with sacral chordoma had bilateral loss of the S2-S5 nerves. Seven required an indwelling catheter, 11 had major voiding problems but managed without a catheter, and 1 patient had minor problems. Five patients lost their S3-S5 nerves bilaterally; one of them needed an indwelling catheter, three developed major problems but managed without a catheter, and one had minor problems. Another patient with loss of S3-S5 nerves on one side and S4-S5 nerves on the contralateral side developed major voiding problems but did not require a catheter. There were five patients with intact S3 nerves bilaterally who had minimal or no problems. The patients who developed major bladder problems but managed without an indwelling catheter either emptied their bladders by self-catheterization or used manual pressure and abdominal contractions. They also used incontinence pads. Sacral tumors (30 cases): motor function. One patient with unilateral loss of the L5 and S1 nerves developed a paralytic foot and needed to use a brace. None of the six patients who lost their S1 nerves had complete loss of active plantar flexion. Surprisingly, five of these patients could even stand tiptoe, apparently due to preserved innervation of the medial gastrocnemius muscle. The remaining 23 patients with preserved S1 nerves had no motor dysfunction. Local recurrence and metastasis There were local recurrences in 12 of 30 patients (40%) with sacral chordoma and 5 of 9 patients (55%) with spinal chordoma; these occurred after surgery had been performed outside our center in 5 and 3 cases, respectively. Thus, a total of 17 patients (44%) developed local recurrences 1.5–12.5 years after the primary surgery (average 6 years). Five of these had a second local recurrence 1–3 years after the first recurrence, and 3 patients had a third local recurrence after another year. One patient had additional local recurrences. Treatment of local recurrences is detailed in Table 4. Of the 17 patients with local recurrences, 4 had no evidence of disease at last follow-up, 2 were alive with disease, and 11 were dead with disease or died of tumor. Table 4. Treatment of 17 Sacral and Spinal Chordoma Patients with Local Recurrences Surgery RTX Surgery and RTX No treatment First recurrence 8 1 7 1 Second recurrence 3 1 1 Third recurrence 2 0 1 RTX: radiation therapy. Metastases developed in 9 of 30 patients (30%) with sacral chordoma and 2 of 9 (22%) with spin