Prostate Cancer and Prostatic Diseases (2003) 6, 169173 & 2003 Nature Publishing Group All rights reserved

1365-7852/03 $25.00 www.nature.com/pcan

Treatment and prognosis of patients with paraplegia or quadriplegia because of metastatic spinal cord compression in prostate cancer
M Nagata1, T Ueda1*, A Komiya1, H Suzuki1, K Akakura1, M Ishihara1, T Tobe1, T Ichikawa1, T Igarashi1 & H Ito1
Department of Urology, Chiba University Graduate School of Medicine, Chiba, Japan

Metastatic spinal cord compression (MSCC) is a serious complication of metastatic prostate cancer (PCa). This study retrospectively evaluated patients who presented with paraplegia or quadriplegia because of MSCC of PCa. Of 847 patients with PCa who were treated between 1989 and 1998, 26 (3.1%) demonstrated paraplegia or quadriplegia because of MSCC. Characteristics, treatment efficacy, and prognosis of these patients were analyzed. In total, 15 cases became paraplegic despite androgen ablation therapy (Group I). Average time to paraplegia from initial hormonal treatment was 34 months. Out of nine cases who underwent radiation therapy (RT) to spinal lesions with/without chemotherapy, one patient became ambulatory. However, this patient subsequently had recurrent compression. Two cases had remission of paralysis. Two cases underwent laminectomy plus RT and in one case paralysis improved. MSCC was the first indication of PCa in 11 cases (Group II). Two cases underwent laminectomy plus hormone therapy and nine cases underwent hormone therapy alone. Four patients became ambulatory and two cases showed improved motor capacity. Average interval from paraplegia to death was 7.4 months in Group I and 27.1 months in Group II. However, there was no statistical difference in these two groups on diseasespecific survival from the start of initial treatment. It is difficult to recover the ability to walk if paraplegia or quadriplegia occurs in PCa patients although decompression surgery plus hormone therapy seemed to impair the prognosis. Stage M1 patients with paraplegia had survival rates as good as stage M1 patients without paralysis. This should encourage an aggressive treatment approach. However, for patients with hormone-independent disease there seems to be no effective treatment and prognosis is poor. Prostate Cancer and Prostatic Diseases (2003) 6, 169173. doi:10.1038/sj.pcan.4500641

Keywords: bone metastasis; spinal cord compression; paraplegia; quadriplegia

Introduction
If paraplegia or quadriplegia occurs because of metastatic lesions of the spine from malignancy other than prostate cancer (PCa), it usually means that the patient
Grant sponsor: Japan Society for the Promotion of Science; Grant numbers: Grant-in-Aid for Scientific Research (C2) 14571483.
*Correspondence: T Ueda, Department of Urology, Chiba University Graduate School of Medicine, 1-8-1 Inohana, Chuo-ku, Chiba-shi, Chiba 260-8670, Japan. E-mail: uedat@ho.chiba-u.ac.jp Received 7 June 2002; accepted 24 October 2002

will survive less than a year. The patients quality of life will be devastated in this situation. Skeletal metastases are found in 60% of cancer patients at autopsy and the vertebral column accounts for 50% of osseous metastases.1 Autopsy series on patients with PCa reveal that 8085% have bone metastasis, with the pelvis and the vertebrae being involved in nearly all cases.2,3 It is said that spinal cord compression with symptoms occurs in about 20% of patients with spinal metastases. The primary tumors most commonly causing spinal metastases are breast, lung, prostate, and kidney cancers.4 PCa is the second most common cause of metastatic spinal cord compression (MSCC) in men after lung cancer. This

Paraplegia in prostate cancer M Nagata et al

170

is recognized as a relatively common complication of metastatic PCa. Kuban et al5 reported that MSCC occurred in 6.7% of 611 patients with PCa. Nevertheless, there are no definite recommendations for its treatment. High-dose steroids, androgen ablation, radiation, decompression surgery (ie laminectomy), and combination of these therapies have all been used. Here we reviewed retrospectively 26 patients who presented paraplegia or quadriplegia because of MSCC with PCa. Patients characteristics, treatment efficacy, and prognosis were all analyzed.

Materials and methods

A total of 847 patients with PCa treated at Chiba University Hospital between 1989 and 1998 were evaluated in this study. The diagnosis of all the patients was confirmed histologically by needle biopsy or surgical specimens obtained from the compressing tissue if paraplegia or quadriplegia was the first sign of PCa. TNM classification was performed according to the TNM staging system.6 Bone metastasis of the patients was evaluated by conventional bone X-ray film and/or bone scan. Bone scan was performed using by 99mTcmethylene-diphosphate. The extent of bone metastasis (the extent of disease, EOD) in each patient was classified according to Soloway et al7 MSCC of PCa patients was diagnosed by neurological examinations, myelography, and/or magnetic resonance imaging (MRI). Based on the neurological examination at presentation, patients who were nonambulatory, paraplegic, or quadriplegic secondary to motor loss were reviewed in this study. Patients who were asymptomatic, ambulatory with minimal neurological findings or paraparetic, nonambulatory secondary to either extremity weakness from motor loss or exacerbation of back pain on walking were excluded from this study. When epidural disease was confirmed, patients were treated with steroids, androgen-deprivation therapy, radiation therapy (RT), and/or laminectomy. Patients undergoing external beam RT to the site of cord compression to 2039 gray were treated on a fractionation schedule over a 2 to 4-week period. Laminectomy was performed by the orthopedic surgeons in our hospital. Statistical analyses were carried out according to the w2 analysis. The disease-specific survival rates were calculated by the KaplanMeier method. Statistical differences in the survival rates between subgroups were evaluated using the log-rank test.

ment to paraplegia or quadriplegia was 34 months (range 3 days to 109 months). Incidence of paraplegia or quadriplegia because of MSCC was statistically higher in stage M1 than the other stages (Po0.001, w2 test). The other 11 cases were newly diagnosed stage M1 PCa patients with MSCC (Group II). Patients pathological findings in each group are shown in Table 1. No statistical difference in tumor grade was observed between Groups I and II. The sites of spinal cord compression and paralysis in each group are shown in Table 2. In 16 (62%) patients, paraplegia occurred at the thoracic level. Paraplegia because of MSCC tended to occur at the upper level of the spine in Group I and, on the contrary, at the lower level in Group II. We also examined the relations between the level of MSCC and EOD score. In 20 of 26 cases, the EOD score could be measured. Of these 20 patients, 15 (75%) revealed EOD score 0 or 1 at primary diagnosis. However, there is no statistically correlation between the site of compression and EOD score in each group. In Group I, the onset of paralysis was acute in six of 15 patients (40%). The time interval between primary diagnosis and MSCC in the other nine cases varied from 1 week to 33 months, with an average interval of 8.2 months. In these nine patients, seven had progressive back pain, one had numbness in sacral area 2 weeks before, and the other had weakness in the lower extremities 2 months before. In Group II, the onset of paralysis was acute in six of 11 patients (55%). The other five patients were aware of back pain before paraplegia (average 4.2 months; range 3 weeks to 8 months). High-dose steroid treatment to MSCC was given to all patients. In Group I, nine cases were treated with RT to spinal lesions combined with/without chemotherapy. Two cases underwent laminectomy plus RT, but they were unable to recover the ability to walk. Three cases underwent palliative treatments for the terminal stage (Table 3). From these treatments, one (7%) of 15 patients in Group I became ambulatory, three (20%) had remission of paralysis, and 11 (73%) cases did not have efficacious outcome to paraplegia. In Group II, nine cases underwent hormone therapy and two cases underwent laminectomy plus hormone therapy. Two cases treated with laminectomy regained the ability to walk. In these two patients, time intervals to laminectomy were 4 days after paralysis and 12 h after paraplegia, respectively. Two patients with quadriplegia had a poor prognosis

Table 1 Tumor grade of PCa patients with metastatic spinal cord compression

Results
Of 847 patients, 26 (3.1%) with PCa had paraplegia or quadriplegia because of MSCC. Clinical stages determined before treatment were one with T1N0M0, one with T3N0M0, one with TxN1M0, and 23 with TxNxM1. Average age and average follow-up period of these patients were 69 years (range 5586) and 33 months (range 1102), respectively. In all, 15 cases had been already treated with androgen ablation therapy without MSCC at primary diagnosis (Group I). Thereafter, these patients showed progression to hormone refractory status and relapsed. Average time from primary treatProstate Cancer and Prostatic Diseases

Group Ia Adenocarcinoma Well Moderate Poor Squamous cell carcinoma Unknown

a c b

Group IIb

Total

1c 8 4 0 2

0 6 4 1 0

1 14 8 1 2

Prostate biopsy was performed before primary treatment was started. Prostate biopsy was performed at the time of spinal cord compression. Obtained by prostatectomy. Classification of patients groups (I and II) were explained in the Results section.

Paraplegia in prostate cancer M Nagata et al Table 2 EOD score and the level of metastatic spinal cord compression in the prostate cancer patients Group I EOD score Cervical Upper thoracic Lower thoracic Lumber Total
ND: not done.

171

Group II ND Sub total 0 7 5 3 15 0 0 1 1 2 3 6 1 1 0 1 1 3 2 3 4 ND 1 1 1 Sub total 2 1 3 5 11 Total 2 8 8 8 26

2 1 3

1 4 1 6 0

1 2

Table 3 Outcome of each combination of therapies for prostate cancer patients with paraplegia or quadriplegia Recover of motor function Treatments Group I RTa RT+chemotherapy Chemotherapy Laminectomy+RT Others HTb Laminectomy+HT Total
a b

Ambulatory

Improved 2

Not efficacious 5 1 1 1 4 12

Total 7 2 1 2 3 9 2 26

1 1 2 2 5 3 6

Group II

RT: radiation therapy. HT: hormone therapy.

and died within 3 weeks. Four of 11 (36%) patients in Group II became ambulant, three (27%) had remission of paralysis, and four (36%) did not obtain any improvement in motor function. Once paraplegia occurred, average follow-up until death in Group I was 7.4 months (range 117), and 12 of 15 patients died within a year. Average follow-up in Group II was 27.1 months with a range of 12 days to 98 months (Figure 1). A further analysis was carried out in these two groups, exploring whether the difference of the onset of MSCC in each group affects clinical outcomes and disease-specific survival. The 5-year survival rate of Group I was 38.9%, whereas that of Group II was 45%. There was no statistical difference in these two groups on diseasespecific survival from the start of initial treatment. Of five patients who became ambulatory after treatment, four patients (one in Group I and three in Group II) had recurrent paraplegia. These four cases remained ambulatory for an average of 21 months with a range of 465 months. Recurrent site of spinal cord compression was the same in two cases (50%), and different in two cases (50%). The other one patient had been alive for 96 months without recurrent paraplegia (see Figure 2).

1 a. Group I (n=15) Survival Rate (%) .8 .6 .4 .2 0 0 20 40 60 Time (month) 80 100 b. Group II (n=11)

Figure 1 Survival curves after paraplegia or quadriplegia because of MSCC in prostate cancer. Classification of groups I and II are described in the Results section.

Discussion
In this study, paraplegia or quadriplegia occurred in 26 of 847 PCa patients (3.1%) either following treatment or at initial presentation. This incidence seems to be lower

than previous reports, which included both paraplegic and paraparetic cases. Paraparetic patients tend to have a good prognosis and to achieve their nonambulatory status under combined therapy.8 On the other hand, paraplegic patients at either presentation or recurrence remained nonambulatory after combined therapy even if decompression surgery was performed. From this point of view, we analyzed the patients who were both nonambulatory and paraplegic (or quadriplegic), and
Prostate Cancer and Prostatic Diseases

Paraplegia in prostate cancer M Nagata et al

172
1
a. Group I (n=15)

.8 Survival Rate (%) .6 .4 .2 0 0 20 40

b. Group II (n=11)

60 80 100 Time (month)

120

140

Figure 2 Disease-specific survival curves in PCa with paraplegia or quadriplegia because of MSCC.

therefore paraparetic or asymptomatic patients were excluded in this study. It is still controversial as to how to manage the patients who present with paraplegia or quadriplegia because of MSCC of PCa. The clinician who cares for these patients is faced with the problems of accurate and urgent diagnosis, and needs to give appropriate treatment for the patients. Treatment of MSCC requires a multidisciplinary approach combining the expertise of orthopedic surgeons, neurosurgeons, and medical and radiation oncologists in addition to urologists. The treatment options available for MSCC include steroids, RT, hormonal manipulation, chemotherapy, surgical intervention, and a combination of two or more of these treatments.9 In general, the management of spinal metastasis is decided by three factors: neurological status, stability, and pain. Harringtons classification is the most widely used as follows: no significant neurological involvement (Class I), bone involvement without collapse (Class II), neurological impairment in the absence of body involvement (Class III), vertebral collapse or instability without significant neurological involvement (Class IV), and vertebral collapse with major neurological impairment (Class V).10 In Class I or II patients, who are neurologically intact and without evidence of instability, chemotherapy, or local radiation is usually indicated for local control. In Class III patients, who have neurological impairment without structural compromise, RT alone is usually as effective as laminectomy in relieving epidural compression. In cases with an acute onset, additional steroid administration has been recommended.11 In Class IV or V patients, who have mechanical instability or progressive deformity with or without neurological involvement, surgical intervention is indicated. Laminectomy alone resulted in only 23% neurological improvement.12 Generally, laminectomy does not remove the neoplastic mass when there is vertebral body involvement. Recently, because of the poor outcome of laminectomy and progress of spinal surgery techniques, decompression and reconstruction have been recommended by the anterior approach. Anterior compression resulted in neurological improvement rates of 78%. In case of severe collapse with
Prostate Cancer and Prostatic Diseases

disruption of the posterior element of the vertebral body, a posterior approach may be required. In PCa patients, neurological findings are very important factors determining the management of spinal metastases. In all, 90% of bone metastases in PCa are osteoblastic and this may make instability less common.13 Urgent androgen ablation therapy with high-dose steroids and RT to metastatic lesions is standard therapy in hormone na ve patients, although radiation plus steroids is the only therapy for hormone refractory PCa (Group I).8,14 However, the efficacy of decompression surgery combined with other therapies is still unclear. Some reports claim a benefit for decompression surgery, others deny it. Zelelfsky et al14 reported that 67% patients with MSCC because of PCa treated with external RT plus steroids had a significant neurological improvements. However, only one out of seven (14%) nonambulatory patients regained the ability to walk after treatment. Shoskes and Perrin16 reported that 22 of 28 (79%) patients had improvement of the presenting complaints following laminectomy, and that eight of 13 (62%) bedridden patients at presentation were ambulatory postoperatively, although the number of paraplegic patients was not clearly defined in their study. Smith et al 8 showed that 83% (10/12) of paraparetic patients at presentation were ambulatory after treatment (RT+steroids+hormone), and that 80% (4/5) of paraplegic or quadriplegia patients at presentation or recurrence remained nonambulatory after RT+steroids with or without laminectomy. In these series of PCa patients, RT+steroids+hormone therapy was not sufficiently efficacious for nonambulatory, especially paraplegic patients at either initial presentation or recurrence.8 It is said that in case of acute neurological deterioration, only the use of surgical decompression and steroids has been shown to be efficacious in reversing and stabilizing neurological dysfunction, the former by physically removing the offending structure and the latter by decreasing the edema associated with MSCC.9 The decision to proceed with surgical treatment should be made according to the prognosis and the patients general status. Whereas untreated paraplegic patients showed a good survival comparable to other stage metastatic patients, prognosis was very poor in the patients who were nonambulatory after hormonal therapy in this study. There is no consensus as to the life expectancy required to justify surgical intervention. Some investigators recommend a 6-month prognosis when considering surgery.16,17 Other reports recommend a more aggressive approach and require only a 3-month life expectancy.1821 Since it is frequently impossible to determine within months a patients life expectancy, the decision for surgery should not be based entirely on presumed survival. It is important to consider the patients quality of life, which can be profoundly improved with timely surgical intervention. The occurrence of paraplegia from metastatic disease is thought to indicate that the life expectancy of the patients is limited. However, prognosis for prolonged survival in patients with untreated (hormone dependent) metastatic PCa is better than other cancers as has been shown in this study, encouraging an aggressive approach of treatment. In Group II patients who were untreated at the time of paraplegia in the present study, only two out of nine (22%) regained mobility after androgen ablation plus

Paraplegia in prostate cancer M Nagata et al

steroids therapy; however, two of two (100%) were ambulatory after laminectomy in addition to androgen ablation plus steroids therapy. Untreated quadriplegic patients had very poor prognoses. Although the number of patients in this study was limited, these findings also suggest, at least in hormone-dependent paraplegic patients (Group II in this study without quadriplegia), that aggressive treatment combined with surgical spinal cord decompression could be directed. In the hormone-independent state, treatment outcome was miserable. Any sort of therapy or any combinations of therapies are not efficacious enough to walking ability. In this context, it is important to diagnose MSCC as early as possible before it becomes symptomatic, if possible,22 and prophylactic radiation of vertebral metastases discovered concurrently with compressive metastases may be valuable in preventing paraplegia.8

Conclusion
It is difficult to regain mobility if paraplegia or quadriplegia occurs in PCa patients, although decompression surgery plus hormone therapy seemed to give a better prognosis in previously untreated patient. Previously untreated paraplegic patients had survival rates as good as stage M1 patients without paralysis. This should encourage an aggressive approach to treatment. However, for patients with hormone-independent disease, there are no effective treatments and the prognosis is poor. The best outlook for these patients is the early diagnosis of MSCC and prophylactic radiation to vertebral metastases.

References
1 Drew M, Dickson RB. Osseous complications of malignancy. In: Lokich JJ (ed). Clinical Cancer Medicine: Treatment Tactics. MTP Press: Boston, 1980, pp 97124. 2 Arnheim FK. Carcinoma of the prostate: a study of the postmortem findings in one hundred and seventy-six cases. J Urol 1948; 60: 599603. 3 Jacobs SC. Spread of prostatic cancer to bone. Urology 1983; 21: 337344. 4 Schaberg J, Gainor BJ. A profile of metastatic carcinoma of the spine. Spine 1985; 10: 1920.

5 Kuban DA et al. Characteristics of spinal cord compression in adenocarcinoma of prostate. Urology 1986; 28: 364369. 6 Sobin JH, Wittekind C. International Union Against Cancer: TNM Classification of Management of Tumours, 5th edn. Wiley, New York, 1997. 7 Soloway MS et al. Stratification of patients with metastatic prostate cancer based on extent of disease on initial bone scan. Cancer 1988; 61: 195202. 8 Smith EM et al. Spinal cord compression secondary to prostate carcinoma: treatment and prognosis. J Urol 1993; 149: 330333. 9 Asdourian PL. Metastatic disease of the spine. In: Bridwell KH, DeWald RL (eds) The Textbook of Spinal Surgery, 2nd edn, Vol. 2. Lippincott-Raven, Philadelphia, 1997, pp 20072050. 10 Harrington KD. Anterior cord decompression and spinal stabilization for patients with metastatic lesions of the spine. J Neurosurg 1984; 61: 107117. 11 Boland PJ, Lane JM, Sundaresan N. Metastatic disease of the spine. Clin Orthop 1982; 169: 95102. 12 Siegal T. Current considerations in the management of neoplastic spinal cord compression. Spine 1989; 14: 223228. 13 Cook GB, Watson FR. Events in the natural history of prostate cancer: using salvage curves, mean age distributions and contingency coefficients. J Urol 1968; 99: 8796. 14 Zelefsky MJ et al. Spinal epidural tumor in patients with prostate cancer. Clinical and radiographic predictors of response to radiation therapy. Cancer 1992; 70: 23192325. 15 Shoskes DA, Perrin RG. The role of surgical management for symptomatic spinal cord compression in patients with metastatic prostate cancer. J Urol 1989; 142: 337339. 16 Cybulski GR, Von Roenn KA, DAngelo CM, DeWald RL. Luque rod stabilization for metastatic disease of the spine. Surg Neurol 1987; 28: 277283. 17 Harrington KD. Anterior decompression and stabilization of the spine as a treatment for vertebral collapse and spinal cord compression from metastatic malignancy. Clin Orthop 1988; 233: 177197. 18 Hammerberg KW. Surgical treatment of metastatic spine disease. Spine 1992; 17: 11481153. 19 Phillips E, Levine AM. Metastatic lesions of the upper cervical spine. Spine 1989; 14: 10711077. 20 ONeil J, Gardner V, Armstrong G. Treatment of tumors of the thoracic and lumbar spinal column. Clin Orthop 1988; 227: 103112. 21 Lee CK, Rosa R, Fernand R. Surgical treatment of tumors of the spine. Spine 1986; 11: 201208. 22 Maranzano E, Latini P. Effectiveness of radiation therapy without surgery in metastatic spinal cord compression: final results from a prospective trial. Int J Radiat Oncol Biol Phys 1995; 32: 959967.

173

Prostate Cancer and Prostatic Diseases