Journal of Surgical Oncology 48:232-238 (1991)
Diagnosis and Management of Sacrococcygeal Chordoma KEVIN P. BETHKE, MU, JAMES P. NEIFELD, MU, AND WALTER LAWRENCE JR., MD From the Division of Surgical Oncology, Department of Surgery, Medical College of Virginia, Richmond
Chordoma is a rare, slow-growing but locally aggressive malignant tumor derived from the primitive notochord and located along the axial skeleton. Between 1973 and 1991, of 15 patients with chordomas treated at the Medical College of Virginia, eight originated in the sacrococcygeal area. There was a median one year interval between the onset of symptoms and diagnosis (range of four months to six years) for this latter group. Two patients had undergone coccygectomies and one patient a lumbar discectomy prior to establishing the correct diagnosis of sacral chordoma. Seven patients underwent resection and one refused therapy. The four patients who had an initial wide radical resection had a longer disease-free survival than the three who underwent local excision. Three of four patients had metastatic disease at the time of death. Early diagnosis and aggressive initial surgical resection are necessary for long-term survival. KEYWORDS:sacral tumor, bone tumor, sacrum, notochord, radiotherapy
INTRODUCTION Chordomas are rare tumors, account for 1-2% of all bone malignancies, and are thought to originate from the primitive notochord. Approximately 50% originate in the sacral region, 35% in the clivus (base of skull), and 15% elsewhere along the axial skeleton. Sacrococcygeal chordomas have been described as slow-growing and difficult to diagnose; many patients are treated for an assortment of unrelated diagnoses before the correct diagnosis is made. Wide excision with postoperative radiation therapy for positive or close margins is the treatment of choice but, even with the best surgical technique, the local recurrence rate is very high. This leads to multiple resections, further radiation, and often terminal complications from local spread. The purpose of this retrospective review of sacrococcygeal chordomas was to determine the optimal preoperative work-up and management of patients with these tumors. PATIENTS AND METHODS A review of the records of the Division of Surgical Oncology, Division of Surgical Pathology, and the Hospital Tumor Registry revealed that 15 patients with the diagnosis of chordoma were treated at the Medical 0 1991 Wiley-Liss, Inc.
College of Virginia Hospitals (MCV) from 1973 to 1991. Eight patients had the chordoma located in the sacrococcygeal area, three in the clivus, and four elsewhere along the axial skeleton. This paper will focus on the eight patients with sacrococcygeal chordoma. Information was obtained regarding the patient’s sex, age, presenting symptoms, treatment, complications, and survival. Patients undergoing a sacral resection at MCV underwent a two-stage resection consisting of an anterior followed by a posterior approach [l-31. The initial anterior approach was via a lower midline incision. The sigmoid colon was mobilized and the peritoneal reflection incised to expose the sacral promontory. The rectum was then dissected from the presacral fascia after which a laparotomy pad was placed within the presacral space to protect the rectum during dissection and transection of the sacrum. The hypogastric arteries were ligated and divided followed by closure of the abdominal incision. The resections were then performed in the prone, jackknife position. An elliptical incision was made Accepted for publication August 20, 1991, Address reprint requests to James P. Neifeld, MD, Division of Surgical Oncology, Medical College of Virginia, Box 1 1 , MCV Station, Richmond, VA 23298-001 1 ,
Sacral Chordorna
233
DISCUSSION Approximately one half of patients with chordoma have tumors localized to the sacrococcygeal region, similar to our patient population. This distribution is somewhat surprising considering the theory that chordoma arises from the primitive notochord and may represent abnormal rests of notochord tissue (except in the nucleus pulposis) [4].Based on this theory, chordoRESULTS mas should arise in the areas with the majority of The median age of patients presenting with sacrococ- intervertebral discs, not in the sacrum or clivus where cygeal chordomas was 59 years (range 30-75 years); most chordomas are found. Sacrococcygeal chordomas are slow-growing, insidithere were four males and four females. Six of the eight patients presented initially with low-back or sacral pain, ous tumors which often reach a large size before the two with urinary retention, and one with radicular leg correct diagnosis is made. Our finding of a median pain. The median interval between onset of symptoms interval of one year between the onset of symptoms and and correct diagnosis was one year (range four months to correct diagnosis is typical of previously reported series six years). Two patients had undergone a prior coccygec- [5-111. Three of our patients had operations prior to the tomy for what was thought to be coccygodynia and one correct diagnosis being made. Higinbotham et al. [5] patient a previous lumbar discectomy for presumed disc reported that 17 of 30 sacrococcygeal chordomas had disease. On physical examination all patients had a been treated for other problems such as arthritis, hemorrhoids, pilonidal sinus, and vertebral disc disease prior to palpable mass extrinsic to the rectum. Seven of the eight patients underwent surgical resec- making the correct diagnosis. There was no history of tion; one patient refused both surgery and a complete prior sacral trauma in our group, unlike some series course of radiation therapy (Table I). Of the seven which report a 15-22% incidence [3,9,1 I]. The pain is gradual in onset and nonspecific as the resected patients, three had undergone a prior tumor excision elsewhere while four had their initial resections surrounding soft tissues are displaced and bone is eroded. performed at MCV. One of these four required a hemi- The tumor usually expands ventrally, displacing the pelvectomy for extensive disease which infiltrated the presacral fascia anteriorly without producing an obvious gluteus muscles. Five patients had sacral resections external mass. As the tumor continues to grow and between SllS2, one through S2, and one between S2lS3. encroaches on the sacral foramina, nerve root neuropaPostoperative complications included temporary uri- thies develop, leading to urinary retention, rectal sphincnary incontinence in five of the seven patients undergoing ter incompetence, and radicular pain. These neurologic resection; one patient has required chronic intermittent symptoms are a sign of advanced disease and may bladder catheterization. One patient experienced stool become less frequent as the newer diagnostic methods of incontinence postoperatively and was also the single computerized tomography (CT) and magnetic resonance postoperative death, caused by a massive pulmonary imaging (MRI) are employed to evaluate the early, embolus 30 days after resection. The patient requiring a nonspecific symptoms. Prior to the advent of CT and MRI, plain pelvic x-rays hemipelvectomy developed a dural leak which required were used to evaluate patients with an extrinsic rectal reoperation for closure. Examination of the data regarding recurrence and mass [8, lo]. These revealed lytic, expansile lesions with survival revealed that those who initially had undergone a soft tissue mass anterior to the sacrum on lateral films local excision had a much shorter disease-free survival (Fig. 1). With the advent of CT, better localization of the than those who underwent an initial aggressive wide tumor became possible (Fig. 2 ) . Currently, MRI is excision (Table I). The patient who required a hemipel- replacing CT as the imaging modality of choice [12] vectomy for extensive disease had a disease-free interval because it not only offers a higher resolution image but it of eight years. Recurrence, although a bad prognostic allows evaluation of the sacral canal, which acts as an sign, does not imply imminent death or significant unobstructed conduit for the superior spread of tumor disability as evidenced by the patient who had a second (Fig. 3 ) . Once a mass is noted on physical examination and recurrence resected three years after the first recurrence and is currently free of disease. Another patient had a x-ray studies confirm sacral involvement, a biopsy must recurrence 6.5 years after sacral resection, refused sur- be performed for definitive diagnosis. The biopsy should gery, and has remained asymptomatic with a stable pelvic be performed via an open or tru-cut needle technique recurrence for the past 4.5 years. Three patients devel- from a posterior approach. A transrectal biopsy should oped distant metastases, all of whom have died of their not be performed because it risks seeding the bowel, which normally is protected by the presacral fascia, a disease. encompassing the previous biopsy site. Lateral skin flaps were developed and the muscular and ligamentous attachments to the sacrum divided. The presacral space, which had been opened during the anterior approach, was entered and the appropriate level for sacral transection chosen. After transection, the subcutaneous tissue was closed over two suction drains and the skin closed.
None Coccygectomy, Partial sacral resection Local excision, Sacral resection through S2 bilaterally Local excision
15/M 49/M
38/F
64/F
54lM
66lM
63/F
30lF
I. 2.
3.
4.
5.
6.
7.
8.
MCV None Completion sacral resection between Sl IS2 bilaterally Resection of buttock recurrence Posterior pelvic exenteration Sacral resection between S2lS3 bilaterally Hemipelvectorny, Sacral resection between S 1IS2 bilaterally Sacral resection between SI/S2 bilaterally Sacral resection between SllS2 bilaterally Sacral resection between SI/S2 bilaterally
aRefused further radiation therapy. bMetastatic disease. 'Requires chronic intermittent bladder catheterization. dRemains free of disease.
Coccygectomy, Exploratory laparotomy
Lumbar discectomy
Elsewhere
Agelsex
Patient
Surgical therapy
TABLE I. Patient Data Summary-Sacrococcygeal Chordoma
68 Gy for microscopically positive margin 65 Gy for microscopically positive marEin
None
None
60 Gy after local excision
45 Gy for pelvic recurrence
10 Gy" None
Radiation therapy
Urinary retention
Urinary retention
Urinary retention'
Urinary retention Dural leak
None
None Urinary retention Stool incontinence Postoperative death Urinary retention
Complications
0.4d
2.0d
6.5
8.0
0.4
2.0
11.0
8.5b
5.0
5.3b
1.o
2.0
3.0b 4.0
0.1
0.0
Overall
Survival (years) Diseasefree
Alive and well
Alive and well
Alive with disease
Alive without disease (second recurrence resected) Dead of disease
Dead of disease
Dead of disease Postoperative death
Status
Sacral Chordoma
Fig. 1. Pelvic x-ray demonstrating a large pelvic soft tissue mass with extensive destruction of the sacrum.
Fig. 2.
Pelvic CT scan showing sacrum (S), tumor (T), and rectum (R).
235
236
Bethke et al.
Fig. 3. Sagittal MRI showing presacral tumor (T) and sacral canal (arrows).
strong barrier between the tumor and rectum. A transrectal biopsy would necessitate a proctectomy , which normally is not required for the treatment of chordoma. The early reports of sacrococcygeal resection advocated only a posterior approach performed in the prone position [7,13]. Currently the standard technique is a combined anterior-posterior approach. The anterior approach allows the surgeon to develop the presacral space, making subsequent sacral transection easier and safer because a laparotomy pad is placed in this space to protect the rectum during sacral transection. During the anterior approach the hypogastric arteries are divided and ligated in order to decrease the amount of bleeding which may be brisk during the sacral transection. Sung et al. [14] reported on 54 patients undergoing sacral tumor excision with prior hypogastric artery ligation. They reported no specific complications related to the ligation, although they did report an 11 % wound breakdown rate. We have not had any wound breakdowns and most patients have not required transfusions. In 1967 Localio et al. [6,15] suggested a combined anterior-posterior approach in the lateral decubitus position. He recommended that the laparotomy be performed through an oblique incision parallel to the inguinal ligament with mobilization of the colon and rectum as previously described. Simultaneously, a second team approaches the sacrum posteriorly through a transverse
incision, divides the muscular and ligamentous attachments to the sacrum, and transects the sacrum with no attempt made to preserve the sacral nerves below the level of transection. In our experience this approach is more difficult than turning the patient from the supine to the prone, jackknife position. It has been recommended [ 161 that the sacral canal be unroofed posteriorly in order to better define and protect the higher nerve roots; however, this might lead to spillage of tumor which has invaded the sacral canal. We believe better exposure can be obtained using the two-stage technique and have found total operating time to be three hours or less. The primary deterrent to complete resection of sacrococcygeal chordomas has been concern for potential iatrogenic nerve root injury. The sacral nerve roots supply sensation to the perineum and genitalia, innervation of the bladder as well as urinary and rectal sphincters, and motor function to the gluteus muscles and plantar extensors. Although the potential for iatrogenic neurologic injury exists, Kaiser et al. [8] reported that all patients who had an incomplete resection eventually lost bladder and bowel function secondary to recurrence. Thus, concern for injury should not compromise margins. Stener and Gunterberg [2] and Gunterberg et al. [ 17,181 have published detailed accounts of the consequences of various levels of sacral transection. They
Sacral Chordoma
found that bilateral sacrifice of S2-5 leads to bladder and urethral sphincter denervation. Lacking the detrusor function of the bladder, the patient will depend on abdominal muscles and posterior manual pressure for voiding. They may require chronic indwelling bladder catheterization or bladder neck resection which carries a risk of permanent incontinence. Our patients, like most reported cases, showed progressive improvement in urinary function over time. The impact of any form of treatment is difficult to gauge because of the slow-growing nature of chordomas. There are many reports of prolonged survival after recurrence. In the review of the literature by Gray et al. [ 111 they found that in 222 patients, the average survival from onset of symptoms was 5.7 years. Eleven patients lived for 10 years and only two of these were diseasefree. Five year survival appears to be an inadequate measure of outcome in patients with chordoma since there is no evidence of a decrease in the rate at which patients die, either prior to or following the five year mark [19]. The importance of complete removal of the tumor with a clear margin cannot be overstated. In reviewing our patients, we divided them into those who had an initial operation consisting of only local excision performed elsewhere and those who initially underwent radical wide resection at MCV. Those treated with aggressive resection had a much longer disease-free survival, which hopefully, will translate into a longer overall survival. This finding has been reported by others. Kaiser et al. [8] reported on 50 patients who underwent complete wide excision; 25 patients had tumor spillage with a 64% recurrence rate while 25 with no tumor spillage had a 28% recurrence rate. Their average survival after recurrence was 3.9 years and 50% who recurred developed distant metastases. Sundaresian et al. [lo] reported that in their series no patient with recurrent disease could be rendered free of disease. In our series three of the five patients with persistent or recurrent disease developed distant metastases. Despite the poor prognosis for patients who develop recurrence, radiation therapy can improve the quality of life for this group of patients who must deal with pain, depression, and who may be unable to function independently. Response of the tumor is quite slow, but regression will usually begin in a month or two and may continue for four to five months. There is controversy regarding the optimal dose of radiation therapy. Cummings et al. [20] concluded that symptomatic improvement for at least five years was no more likely at doses above 50 Gy than at lower doses. Saxton [21] reviewed the MD Anderson experience and reached the same conclusion. However, this view contrasts with that of others [5,9,22,23] who advocate higher
237
doses of approximately 70 Gy. Those who respond to the first course usually respond to a second [5]; however, the re-irradiated patient is at higher risk for radiation-related complications [22]. Most reports indicate that partial resection prior to radiation therapy does not improve long-term control [8,10,24]. Although no good studies have shown its efficacy, postoperative radiation for close or microscopically positive margins is generally accepted as standard therapy.
CONCLUSIONS Sacrococcygeal chordoma is a rare, slow-growing , insidious, malignant tumor which presents a diagnostic challenge. The only hope for cure is early diagnosis followed by aggressive wide resection as the initial treatment. Even though radical surgery may lead to iatrogenic sacral nerve root injury, most urinary and bladder dysfunction improves with time and training. Recurrence caused by inadequate initial resection will progress to even greater problems. Although there are no good studies proving its efficacy, postoperative radiation therapy for close or microscopically positive margins seems indicated.
REFERENCES 1. Bowers RF: Giant cell tumor of the sacrum: a case report. Ann Surg 128:116&1172, 1948. 2. Stener B, Gunterberg B: High amputation of the sacrum for extirpation of tumors: principles and technique. Spine 3:351-366, 1978. 3. Cody HS, Marcove RC, Quan SH: Malignant retrorectal tumors: twenty-eight years’ experience at Memorial Sloan-Kettering Cancer Center. Dis Colon Rectum 24501-506, 1981. 4. Horwitz T: Chordal ectopia and its possible relation to chordoma. Arch Pathol Lab Med 31:354-362, 1941. 5. Higinbotham NL, Phillips RF, Farr HW, Hustu HO: Chordoma: thirty-five year study at Memorial Hospital. Cancer 20:18411850, 1967. 6. Localio SA, Eng K, Ranson JHC: Abdominosacral approach for retrorectal tumors. Ann Surg 191:555-560, 1980. 7. Dahlin DC, MacCarty CS: Chordoma: a study of fifty-nine cases. Cancer 5.1 170-1 178, 1952. 8. Kaiser TE, Pritchard DJ, Unni KK: Clinicopathologic study of saccrococcygeal chordoma. Cancer 53:2574-2578, 1984. 9. Rich TA, Schiller A, Suit HD, Mankin HJ: Clinical and pathologic review of 48 cases of chordoma. Cancer 56:182-187, 1985. 10. Sundaresian N, Huvos AG, Krol G, et al.: Surgical treatment of spinal chordomas. Arch Surg 122:1479-1482, 1987. 11. Gray SW, Singhabhandhu B, Smith RA, Skandalakis JE: Sacrococcygeal chordoma: report of a case and review of the literature. Surgery 78:573-582, 1975. 12. Wetzel LH, Levine E: MR Imaging of sacral and presacral lesions. AJR 154:771-775, 1990. 13. MacCarty CS, Waugh JM, Mayo CW, Coventry MB: The surgical treatment of presacral tumors: a combined problem. Mayo Clin Proc 27:73-84, 1952. 14. Sung HW, Shu WP, Wang HM, et al: Surgical treatment of primary tumors of the sacrum. Clin Orthop 215:91-98, 1987. 15. Localio SA, Francis KC, Rossano PG: Abdominosacral resection of sacrococcygeal chordoma. Ann Surg 166:394402, 1967. 16. Huth JF, Dawson EG, Eilber FR: Abdominosacral resection for malignant tumors of the sacrum. Am J Surg 148:157-161, 1984.
238
Bethke et al.
17. Gunterberg B, Norlen L, Stener B, Sundin T: Neurologic evaluation after resection of the sacrum. Invest Urol 13:183-187, 1975. 18. Gunterberg B, Petersen I: Sexual function after major resections of the sacrum with bilateral or unilateral sacrifice of sacral nerves. Fertil Steril 27:1146-1153, 1976. 19. Cummings BJ, Esses S , Harwood AR: The treatment of chordomas. Cancer Treat Rev 9:299-311, 1982. 20. Cummings BJ, Hodson DI, Bush RS: Chordoma: the results of megavoltage radiation therapy. Int J Radiat Oncol Biol Phys 9:633-642, 1983.
21. Saxton JP: Chordoma. Int J Radiat Oncol Biol Phys 7:913-915, 1981. 22. Pearlman AW, Friedman M: Radical radiation therapy of chordoma. Am J Roentgen01 Radium Ther Nucl Med 108:333-341, 1970. 23. Tewfik HH, McGinnis WL, Nordstrom DG, Latourette HB: Chordoma. Evaluation of clinical behavior and treatment modalities. Int J Radiat Oncol Biol Phys 2:959-962, 1977. 24. Fuller DB, Bloom JG: Radiotherapy for chordoma. Int J Radiat Oncol Biol Phys 15:331-339, 1988.
Diagnosis and Management of Sacrococcygeal Chordoma KEVIN P. BETHKE, MU, JAMES P. NEIFELD, MU, AND WALTER LAWRENCE JR., MD From the Division of Surgical Oncology, Department of Surgery, Medical College of Virginia, Richmond
Chordoma is a rare, slow-growing but locally aggressive malignant tumor derived from the primitive notochord and located along the axial skeleton. Between 1973 and 1991, of 15 patients with chordomas treated at the Medical College of Virginia, eight originated in the sacrococcygeal area. There was a median one year interval between the onset of symptoms and diagnosis (range of four months to six years) for this latter group. Two patients had undergone coccygectomies and one patient a lumbar discectomy prior to establishing the correct diagnosis of sacral chordoma. Seven patients underwent resection and one refused therapy. The four patients who had an initial wide radical resection had a longer disease-free survival than the three who underwent local excision. Three of four patients had metastatic disease at the time of death. Early diagnosis and aggressive initial surgical resection are necessary for long-term survival. KEYWORDS:sacral tumor, bone tumor, sacrum, notochord, radiotherapy
INTRODUCTION Chordomas are rare tumors, account for 1-2% of all bone malignancies, and are thought to originate from the primitive notochord. Approximately 50% originate in the sacral region, 35% in the clivus (base of skull), and 15% elsewhere along the axial skeleton. Sacrococcygeal chordomas have been described as slow-growing and difficult to diagnose; many patients are treated for an assortment of unrelated diagnoses before the correct diagnosis is made. Wide excision with postoperative radiation therapy for positive or close margins is the treatment of choice but, even with the best surgical technique, the local recurrence rate is very high. This leads to multiple resections, further radiation, and often terminal complications from local spread. The purpose of this retrospective review of sacrococcygeal chordomas was to determine the optimal preoperative work-up and management of patients with these tumors. PATIENTS AND METHODS A review of the records of the Division of Surgical Oncology, Division of Surgical Pathology, and the Hospital Tumor Registry revealed that 15 patients with the diagnosis of chordoma were treated at the Medical 0 1991 Wiley-Liss, Inc.
College of Virginia Hospitals (MCV) from 1973 to 1991. Eight patients had the chordoma located in the sacrococcygeal area, three in the clivus, and four elsewhere along the axial skeleton. This paper will focus on the eight patients with sacrococcygeal chordoma. Information was obtained regarding the patient’s sex, age, presenting symptoms, treatment, complications, and survival. Patients undergoing a sacral resection at MCV underwent a two-stage resection consisting of an anterior followed by a posterior approach [l-31. The initial anterior approach was via a lower midline incision. The sigmoid colon was mobilized and the peritoneal reflection incised to expose the sacral promontory. The rectum was then dissected from the presacral fascia after which a laparotomy pad was placed within the presacral space to protect the rectum during dissection and transection of the sacrum. The hypogastric arteries were ligated and divided followed by closure of the abdominal incision. The resections were then performed in the prone, jackknife position. An elliptical incision was made Accepted for publication August 20, 1991, Address reprint requests to James P. Neifeld, MD, Division of Surgical Oncology, Medical College of Virginia, Box 1 1 , MCV Station, Richmond, VA 23298-001 1 ,
Sacral Chordorna
233
DISCUSSION Approximately one half of patients with chordoma have tumors localized to the sacrococcygeal region, similar to our patient population. This distribution is somewhat surprising considering the theory that chordoma arises from the primitive notochord and may represent abnormal rests of notochord tissue (except in the nucleus pulposis) [4].Based on this theory, chordoRESULTS mas should arise in the areas with the majority of The median age of patients presenting with sacrococ- intervertebral discs, not in the sacrum or clivus where cygeal chordomas was 59 years (range 30-75 years); most chordomas are found. Sacrococcygeal chordomas are slow-growing, insidithere were four males and four females. Six of the eight patients presented initially with low-back or sacral pain, ous tumors which often reach a large size before the two with urinary retention, and one with radicular leg correct diagnosis is made. Our finding of a median pain. The median interval between onset of symptoms interval of one year between the onset of symptoms and and correct diagnosis was one year (range four months to correct diagnosis is typical of previously reported series six years). Two patients had undergone a prior coccygec- [5-111. Three of our patients had operations prior to the tomy for what was thought to be coccygodynia and one correct diagnosis being made. Higinbotham et al. [5] patient a previous lumbar discectomy for presumed disc reported that 17 of 30 sacrococcygeal chordomas had disease. On physical examination all patients had a been treated for other problems such as arthritis, hemorrhoids, pilonidal sinus, and vertebral disc disease prior to palpable mass extrinsic to the rectum. Seven of the eight patients underwent surgical resec- making the correct diagnosis. There was no history of tion; one patient refused both surgery and a complete prior sacral trauma in our group, unlike some series course of radiation therapy (Table I). Of the seven which report a 15-22% incidence [3,9,1 I]. The pain is gradual in onset and nonspecific as the resected patients, three had undergone a prior tumor excision elsewhere while four had their initial resections surrounding soft tissues are displaced and bone is eroded. performed at MCV. One of these four required a hemi- The tumor usually expands ventrally, displacing the pelvectomy for extensive disease which infiltrated the presacral fascia anteriorly without producing an obvious gluteus muscles. Five patients had sacral resections external mass. As the tumor continues to grow and between SllS2, one through S2, and one between S2lS3. encroaches on the sacral foramina, nerve root neuropaPostoperative complications included temporary uri- thies develop, leading to urinary retention, rectal sphincnary incontinence in five of the seven patients undergoing ter incompetence, and radicular pain. These neurologic resection; one patient has required chronic intermittent symptoms are a sign of advanced disease and may bladder catheterization. One patient experienced stool become less frequent as the newer diagnostic methods of incontinence postoperatively and was also the single computerized tomography (CT) and magnetic resonance postoperative death, caused by a massive pulmonary imaging (MRI) are employed to evaluate the early, embolus 30 days after resection. The patient requiring a nonspecific symptoms. Prior to the advent of CT and MRI, plain pelvic x-rays hemipelvectomy developed a dural leak which required were used to evaluate patients with an extrinsic rectal reoperation for closure. Examination of the data regarding recurrence and mass [8, lo]. These revealed lytic, expansile lesions with survival revealed that those who initially had undergone a soft tissue mass anterior to the sacrum on lateral films local excision had a much shorter disease-free survival (Fig. 1). With the advent of CT, better localization of the than those who underwent an initial aggressive wide tumor became possible (Fig. 2 ) . Currently, MRI is excision (Table I). The patient who required a hemipel- replacing CT as the imaging modality of choice [12] vectomy for extensive disease had a disease-free interval because it not only offers a higher resolution image but it of eight years. Recurrence, although a bad prognostic allows evaluation of the sacral canal, which acts as an sign, does not imply imminent death or significant unobstructed conduit for the superior spread of tumor disability as evidenced by the patient who had a second (Fig. 3 ) . Once a mass is noted on physical examination and recurrence resected three years after the first recurrence and is currently free of disease. Another patient had a x-ray studies confirm sacral involvement, a biopsy must recurrence 6.5 years after sacral resection, refused sur- be performed for definitive diagnosis. The biopsy should gery, and has remained asymptomatic with a stable pelvic be performed via an open or tru-cut needle technique recurrence for the past 4.5 years. Three patients devel- from a posterior approach. A transrectal biopsy should oped distant metastases, all of whom have died of their not be performed because it risks seeding the bowel, which normally is protected by the presacral fascia, a disease. encompassing the previous biopsy site. Lateral skin flaps were developed and the muscular and ligamentous attachments to the sacrum divided. The presacral space, which had been opened during the anterior approach, was entered and the appropriate level for sacral transection chosen. After transection, the subcutaneous tissue was closed over two suction drains and the skin closed.
None Coccygectomy, Partial sacral resection Local excision, Sacral resection through S2 bilaterally Local excision
15/M 49/M
38/F
64/F
54lM
66lM
63/F
30lF
I. 2.
3.
4.
5.
6.
7.
8.
MCV None Completion sacral resection between Sl IS2 bilaterally Resection of buttock recurrence Posterior pelvic exenteration Sacral resection between S2lS3 bilaterally Hemipelvectorny, Sacral resection between S 1IS2 bilaterally Sacral resection between SI/S2 bilaterally Sacral resection between SllS2 bilaterally Sacral resection between SI/S2 bilaterally
aRefused further radiation therapy. bMetastatic disease. 'Requires chronic intermittent bladder catheterization. dRemains free of disease.
Coccygectomy, Exploratory laparotomy
Lumbar discectomy
Elsewhere
Agelsex
Patient
Surgical therapy
TABLE I. Patient Data Summary-Sacrococcygeal Chordoma
68 Gy for microscopically positive margin 65 Gy for microscopically positive marEin
None
None
60 Gy after local excision
45 Gy for pelvic recurrence
10 Gy" None
Radiation therapy
Urinary retention
Urinary retention
Urinary retention'
Urinary retention Dural leak
None
None Urinary retention Stool incontinence Postoperative death Urinary retention
Complications
0.4d
2.0d
6.5
8.0
0.4
2.0
11.0
8.5b
5.0
5.3b
1.o
2.0
3.0b 4.0
0.1
0.0
Overall
Survival (years) Diseasefree
Alive and well
Alive and well
Alive with disease
Alive without disease (second recurrence resected) Dead of disease
Dead of disease
Dead of disease Postoperative death
Status
Sacral Chordoma
Fig. 1. Pelvic x-ray demonstrating a large pelvic soft tissue mass with extensive destruction of the sacrum.
Fig. 2.
Pelvic CT scan showing sacrum (S), tumor (T), and rectum (R).
235
236
Bethke et al.
Fig. 3. Sagittal MRI showing presacral tumor (T) and sacral canal (arrows).
strong barrier between the tumor and rectum. A transrectal biopsy would necessitate a proctectomy , which normally is not required for the treatment of chordoma. The early reports of sacrococcygeal resection advocated only a posterior approach performed in the prone position [7,13]. Currently the standard technique is a combined anterior-posterior approach. The anterior approach allows the surgeon to develop the presacral space, making subsequent sacral transection easier and safer because a laparotomy pad is placed in this space to protect the rectum during sacral transection. During the anterior approach the hypogastric arteries are divided and ligated in order to decrease the amount of bleeding which may be brisk during the sacral transection. Sung et al. [14] reported on 54 patients undergoing sacral tumor excision with prior hypogastric artery ligation. They reported no specific complications related to the ligation, although they did report an 11 % wound breakdown rate. We have not had any wound breakdowns and most patients have not required transfusions. In 1967 Localio et al. [6,15] suggested a combined anterior-posterior approach in the lateral decubitus position. He recommended that the laparotomy be performed through an oblique incision parallel to the inguinal ligament with mobilization of the colon and rectum as previously described. Simultaneously, a second team approaches the sacrum posteriorly through a transverse
incision, divides the muscular and ligamentous attachments to the sacrum, and transects the sacrum with no attempt made to preserve the sacral nerves below the level of transection. In our experience this approach is more difficult than turning the patient from the supine to the prone, jackknife position. It has been recommended [ 161 that the sacral canal be unroofed posteriorly in order to better define and protect the higher nerve roots; however, this might lead to spillage of tumor which has invaded the sacral canal. We believe better exposure can be obtained using the two-stage technique and have found total operating time to be three hours or less. The primary deterrent to complete resection of sacrococcygeal chordomas has been concern for potential iatrogenic nerve root injury. The sacral nerve roots supply sensation to the perineum and genitalia, innervation of the bladder as well as urinary and rectal sphincters, and motor function to the gluteus muscles and plantar extensors. Although the potential for iatrogenic neurologic injury exists, Kaiser et al. [8] reported that all patients who had an incomplete resection eventually lost bladder and bowel function secondary to recurrence. Thus, concern for injury should not compromise margins. Stener and Gunterberg [2] and Gunterberg et al. [ 17,181 have published detailed accounts of the consequences of various levels of sacral transection. They
Sacral Chordoma
found that bilateral sacrifice of S2-5 leads to bladder and urethral sphincter denervation. Lacking the detrusor function of the bladder, the patient will depend on abdominal muscles and posterior manual pressure for voiding. They may require chronic indwelling bladder catheterization or bladder neck resection which carries a risk of permanent incontinence. Our patients, like most reported cases, showed progressive improvement in urinary function over time. The impact of any form of treatment is difficult to gauge because of the slow-growing nature of chordomas. There are many reports of prolonged survival after recurrence. In the review of the literature by Gray et al. [ 111 they found that in 222 patients, the average survival from onset of symptoms was 5.7 years. Eleven patients lived for 10 years and only two of these were diseasefree. Five year survival appears to be an inadequate measure of outcome in patients with chordoma since there is no evidence of a decrease in the rate at which patients die, either prior to or following the five year mark [19]. The importance of complete removal of the tumor with a clear margin cannot be overstated. In reviewing our patients, we divided them into those who had an initial operation consisting of only local excision performed elsewhere and those who initially underwent radical wide resection at MCV. Those treated with aggressive resection had a much longer disease-free survival, which hopefully, will translate into a longer overall survival. This finding has been reported by others. Kaiser et al. [8] reported on 50 patients who underwent complete wide excision; 25 patients had tumor spillage with a 64% recurrence rate while 25 with no tumor spillage had a 28% recurrence rate. Their average survival after recurrence was 3.9 years and 50% who recurred developed distant metastases. Sundaresian et al. [lo] reported that in their series no patient with recurrent disease could be rendered free of disease. In our series three of the five patients with persistent or recurrent disease developed distant metastases. Despite the poor prognosis for patients who develop recurrence, radiation therapy can improve the quality of life for this group of patients who must deal with pain, depression, and who may be unable to function independently. Response of the tumor is quite slow, but regression will usually begin in a month or two and may continue for four to five months. There is controversy regarding the optimal dose of radiation therapy. Cummings et al. [20] concluded that symptomatic improvement for at least five years was no more likely at doses above 50 Gy than at lower doses. Saxton [21] reviewed the MD Anderson experience and reached the same conclusion. However, this view contrasts with that of others [5,9,22,23] who advocate higher
237
doses of approximately 70 Gy. Those who respond to the first course usually respond to a second [5]; however, the re-irradiated patient is at higher risk for radiation-related complications [22]. Most reports indicate that partial resection prior to radiation therapy does not improve long-term control [8,10,24]. Although no good studies have shown its efficacy, postoperative radiation for close or microscopically positive margins is generally accepted as standard therapy.
CONCLUSIONS Sacrococcygeal chordoma is a rare, slow-growing , insidious, malignant tumor which presents a diagnostic challenge. The only hope for cure is early diagnosis followed by aggressive wide resection as the initial treatment. Even though radical surgery may lead to iatrogenic sacral nerve root injury, most urinary and bladder dysfunction improves with time and training. Recurrence caused by inadequate initial resection will progress to even greater problems. Although there are no good studies proving its efficacy, postoperative radiation therapy for close or microscopically positive margins seems indicated.
REFERENCES 1. Bowers RF: Giant cell tumor of the sacrum: a case report. Ann Surg 128:116&1172, 1948. 2. Stener B, Gunterberg B: High amputation of the sacrum for extirpation of tumors: principles and technique. Spine 3:351-366, 1978. 3. Cody HS, Marcove RC, Quan SH: Malignant retrorectal tumors: twenty-eight years’ experience at Memorial Sloan-Kettering Cancer Center. Dis Colon Rectum 24501-506, 1981. 4. Horwitz T: Chordal ectopia and its possible relation to chordoma. Arch Pathol Lab Med 31:354-362, 1941. 5. Higinbotham NL, Phillips RF, Farr HW, Hustu HO: Chordoma: thirty-five year study at Memorial Hospital. Cancer 20:18411850, 1967. 6. Localio SA, Eng K, Ranson JHC: Abdominosacral approach for retrorectal tumors. Ann Surg 191:555-560, 1980. 7. Dahlin DC, MacCarty CS: Chordoma: a study of fifty-nine cases. Cancer 5.1 170-1 178, 1952. 8. Kaiser TE, Pritchard DJ, Unni KK: Clinicopathologic study of saccrococcygeal chordoma. Cancer 53:2574-2578, 1984. 9. Rich TA, Schiller A, Suit HD, Mankin HJ: Clinical and pathologic review of 48 cases of chordoma. Cancer 56:182-187, 1985. 10. Sundaresian N, Huvos AG, Krol G, et al.: Surgical treatment of spinal chordomas. Arch Surg 122:1479-1482, 1987. 11. Gray SW, Singhabhandhu B, Smith RA, Skandalakis JE: Sacrococcygeal chordoma: report of a case and review of the literature. Surgery 78:573-582, 1975. 12. Wetzel LH, Levine E: MR Imaging of sacral and presacral lesions. AJR 154:771-775, 1990. 13. MacCarty CS, Waugh JM, Mayo CW, Coventry MB: The surgical treatment of presacral tumors: a combined problem. Mayo Clin Proc 27:73-84, 1952. 14. Sung HW, Shu WP, Wang HM, et al: Surgical treatment of primary tumors of the sacrum. Clin Orthop 215:91-98, 1987. 15. Localio SA, Francis KC, Rossano PG: Abdominosacral resection of sacrococcygeal chordoma. Ann Surg 166:394402, 1967. 16. Huth JF, Dawson EG, Eilber FR: Abdominosacral resection for malignant tumors of the sacrum. Am J Surg 148:157-161, 1984.
238
Bethke et al.
17. Gunterberg B, Norlen L, Stener B, Sundin T: Neurologic evaluation after resection of the sacrum. Invest Urol 13:183-187, 1975. 18. Gunterberg B, Petersen I: Sexual function after major resections of the sacrum with bilateral or unilateral sacrifice of sacral nerves. Fertil Steril 27:1146-1153, 1976. 19. Cummings BJ, Esses S , Harwood AR: The treatment of chordomas. Cancer Treat Rev 9:299-311, 1982. 20. Cummings BJ, Hodson DI, Bush RS: Chordoma: the results of megavoltage radiation therapy. Int J Radiat Oncol Biol Phys 9:633-642, 1983.
21. Saxton JP: Chordoma. Int J Radiat Oncol Biol Phys 7:913-915, 1981. 22. Pearlman AW, Friedman M: Radical radiation therapy of chordoma. Am J Roentgen01 Radium Ther Nucl Med 108:333-341, 1970. 23. Tewfik HH, McGinnis WL, Nordstrom DG, Latourette HB: Chordoma. Evaluation of clinical behavior and treatment modalities. Int J Radiat Oncol Biol Phys 2:959-962, 1977. 24. Fuller DB, Bloom JG: Radiotherapy for chordoma. Int J Radiat Oncol Biol Phys 15:331-339, 1988.
