127 Neurologic Complications of Ovarian Carcinoma Lauren E. Abrey, M.D. Josep O. Dalmau, M.D., BACKGROUND. Neurologic complications of ovarian carcinoma are uncommon and Ph.D. to the authors’ knowledge the full spectrum has not been delineated previously. METHODS. The authors reviewed the findings of 121 neurologic consultations on 83 Department of Neurology, Memorial Sloan-Kettering Cancer Center, New York, New York. ovarian carcinoma patients between 1993 and 1996; this represents 4% of all ovarian carcinoma patients seen at the study institution in this time period. RESULTS. The most common reasons for consultation were altered mental status, pain, weakness, numbness, headache, and seizure. Twenty-seven consultations diagnosed metastatic disease, 14 diagnosed cerebrovascular disease, and 4 diagnosed paraneoplastic syndromes; however, iatrogenic complications (n ⫽ 38) comprised the majority of diagnoses. Greater than 50% of patients improved neurologically after diagnosis and treatment. CONCLUSIONS. Neurologic disease accompanying ovarian carcinoma may be more common and more diverse than recognized previously. Definitive neurologic diagnosis and treatment benefits the majority of patients. Cancer 1999;85:127–33. © 1998 American Cancer Society. KEYWORDS: ovarian neoplasms, metastasis, brain, cerebrovascular disorders, syndrome, paraneoplastic, iatrogenic disease. N eurologic problems in patients with ovarian carcinoma may be the result of treatment, metastatic disease, paraneoplastic syndromes, or coincidental neurologic disease. Although brain metastases are uncommon and neurologic paraneoplastic syndromes are rare, the full spectrum of neurologic complications in ovarian carcinoma patients is not known. To describe the range and impact of neurologic disease in these patients more fully, we analyzed 121 consecutive neurology consultations performed on ovarian carcinoma patients at our institution over a 3-year period. METHODS Presented in part at the American Neurological Association Meeting, San Diego, California, September 28-October 1, 1997. Address for reprints: Lauren E. Abrey, M.D., Department of Neurology, Memorial Sloan-Kettering Cancer Center, 1275 York Avenue, New York, NY 10021. Received February 20, 1998; revision received April 22, 1998; accepted May 26, 1998. © 1998 American Cancer Society The charts of all ovarian carcinoma patients seen by the neurology service between January 1993 and December 1996 were identified from a departmental database and reviewed. A hospital database was used to identify the total number of patients with ovarian carcinoma seen at our institution in the same time period. Both inpatient and outpatient encounters were included. The following information was recorded: total number of patients, number of consultations, neurologic diagnoses, patient age, type and stage of ovarian carcinoma, date of diagnosis, presence or history of other malignancy, prior cancer treatment, date of last follow-up, and outcome of the neurologic problem. Analysis of discrete variables was performed using the chi-squared method; Yates correction was applied where expected values were ⬍ 5. 128 CANCER January 1, 1999 / Volume 85 / Number 1 FIGURE 1. Reasons for obtaining neurologic consultation. FIGURE 2. Neurologic diagnoses. Asterisk indicates that chemotherapyinduced peripheral neuropathy and medication-induced encephalopathy were excluded from this category. TIA: transient ischemic attack. RESULTS Between January 1993 and December 1996, 121 neurologic consultations were performed on 83 patients with ovarian carcinoma. During that same period 2050 patients with ovarian carcinoma were seen, indicating that the neurology service saw 4% of all ovarian carcinoma patients. The most common reasons for consultation and diagnosis are summarized in Figures 1 and 2. The median age of the patients was 51 years (range, 13– 86 years). Seventy-six patients had epithelial ovarian carcinoma and 7 patients had other pathology (4 germ cell tumors [2 teratomas, 1 endodermal sinus tumor, and 1 germinoma], 1 hemangioendothelioma, 1 ovarian lymphoma, and 1 müllerian cell carcinoma). Staging information was available in 66 of the 76 patients with epithelial ovarian carcinoma; 10 patients were Stage I (tumor limited to the ovaries), 3 patients were Stage II (tu- mor involving 1 or both ovaries with pelvic extension), 40 patients were Stage III (tumor involving 1 or both ovaries with microscopically confirmed peritoneal metastasis outside the pelvis and/or regional lymph node metastasis), and 13 patients were Stage IV (distant metastasis, excluding peritoneal metastasis).1 Seventy-five patients (78%) had active ovarian tumor at the time of neurology consultation. Fourteen patients had another malignancy and 3 patients had multiple malignancies. Breast carcinoma was the most common second malignancy (n ⫽ 11); 3 patients had endometrial carcinoma and 1 patient each had Hodgkin’s lymphoma, non-Hodgkin’s lymphoma, renal cell carcinoma, and melanoma. Patients with a second malignancy were more likely to have Stage I or II ovarian carcinoma (P ⫽ 0.001). Three patients had evidence of an active second tumor at the time of neurology consultation. In 38 patients, neurologic symptoms were a direct result of treatment. Fifty percent were related to chemotherapy (n ⫽ 19), most often a peripheral neuropathy secondary to treatment with paclitaxel, cisplatin, or combination chemotherapy. Headache, encephalopathy, and ototoxicity were less frequent complications of chemotherapy. Opiates contributed to approximately 33% of encephalopathies. Five patients had postoperative complications (three strokes and two lumbosacral radiculopathies). Twenty-seven consultations resulted in a diagnosis of metastatic disease (12 brain metastases, 4 leptomeningeal metastases, 4 vertebral body metastases, 4 lumbosacral plexus metastases, and 3 bone metastases). Three of 11 patients with both breast and ovarian carcinoma developed brain metastases; in one patient surgical pathology confirmed ovarian carcinoma. The other two patients had active, advanced stage breast carcinoma in addition to their ovarian carcinoma; both died within 1 month of being diagnosed with brain metastases and the pathology of the brain metastases was unknown. The initial symptom in ⬎ 50% of all patients with brain metastases was altered mental status; 8 patients underwent a nonfocal neurologic examination. Brain metastases occurred a median of 38 months after the initial tumor diagnosis (range, 0 –91 months), but was the presenting site of ovarian carcinoma in 1 patient. All but one patient had advanced stage ovarian carcinoma at initial diagnosis. Six patients had a single supratentorial metastasis, two had multiple supratentorial metastases, and four had multiple supratentorial and infratentorial metastases. All patients with leptomeningeal metastases had a prior diagnosis of parenchymal brain metastases. Four patients with brain metastases were treated with surgical resection fol- Neurologic Complications of Ovarian Carcinoma/Abrey and Dalmau lowed by whole brain radiation, three patients received whole brain radiation alone, and one patient received whole brain radiation followed by surgical resection. One patient initially was treated with ifosfamide alone for her solitary brain metastasis; she had a partial radiographic response lasting 10 months (Fig. 3) an stereotactic radiation was given at time of tumor progression. Three patients received best supportive care. The median survival of patients receiving active treatment for brain metastases was 4.5 months; however, 3 of 4 patients surviving ⬎ 1 year after the diagnosis of brain metastases underwent resection and received whole brain radiation, including 2 patients with multiple lesions. The fourth patient was the patient treated with ifosfamide initially and stereotactic radiosurgery at recurrence. All 4 1-year survivors had active systemic tumor in addition to their brain metastases; 2 patients eventually died of progressive systemic tumor and 2 eventually developed leptomeningeal tumors and died of neurologic disease. The presence of multiple versus single metastases did not impact 1-year survival significantly (P ⫽ 0.54). Metastases outside the central nervous system accounted for ⬎ 33% of metastatic neurologic complications. Local pelvic disease resulted in lumbosacral plexopathy in four patients. Four patients with vertebral column metastases and three patients with other bone metastases presented with pain. No patient had spinal cord compression. Thirteen patients had 14 cerebrovascular events: 4 transient ischemic attacks and 10 infarctions. The median age of these patients was 67 years. Eight patients had cerebrovascular risk factors unrelated to their malignancy. Twelve patients had active ovarian tumor at the time of their vascular event; all had advanced stage disease. One patient also had metastatic breast carcinoma. Six patients were receiving chemotherapy at the time of their cerebrovascular event; however, no particular chemotherapy was identified in association with these events. Three patients had a perioperative stroke; two had concomitant deep venous thrombosis. A 26-year-old woman developed a hypercoaguable state and cerebrovascular symptoms as the presentation of her ovarian tumor. Nonbacterial thrombotic endocarditis (NBTE) was diagnosed by echocardiogram in another patient. Ten of the 13 patients improved neurologically; 4 received anticoagulants. Low molecular weight heparin was used to treat the patients with hypercoaguability and NBTE; both patients improved neurologically and had no further cerebrovascular events. Four patients had neurologic paraneoplastic disease; one developed dermatomyositis and three presented with subacute cerebellar degeneration. Para- 129 neoplastic markers were identified in the serum of two patients with cerebellar degeneration; one had anti-Yo antibodies and the other had anti-Ri antibodies. The third patient had no detectable antibodies, but immunohistochemistry of her tumor demonstrated expression of Hu and Ri neuronal antigens. At last follow-up, two patients with subacute cerebellar degeneration had tumors in complete remission; both were Stage I at diagnosis, but have permanent neurologic impairment. The patient with anti-Ri antibodies and the patient with dermatomyositis both had advanced stage tumors and died of progressive disease. It is interesting to note that the patient with dermatomyositis had a disease course closely paralleling that of her tumor, improving as her tumor responded to chemotherapy and worsening at the time of tumor recurrence. After initial neurologic evaluation, follow-up information was available for 78 consultations; in 61.5% there was clear neurologic improvement documented by neurologic examination and in 32% the neurologic findings were unchanged. In particular, 87% of patients with toxic metabolic encephalopathy, 77% with cerebrovascular events, and 48% with metastatic disease improved. DISCUSSION Neurologic complications affect up to 50% of patients with cancer over the course of their disease and its treatment. Although prior studies of ovarian carcinoma detail specific neurologic problems such as brain metastases or chemotherapy-induced peripheral neuropathy, to our knowledge none has sought to elaborate the full spectrum of neurologic complications in patients with ovarian carcinoma. In this series we found that nearly 33% of neurologic problems were related to complications of therapy, and 25% were related to metastatic involvement of the nervous system. Cerebrovascular disease and paraneoplastic syndromes were more common than anticipated.2,3 Definitive neurologic evaluation and diagnosis often led to appropriate intervention and the majority of patients improved neurologically. Our observations and patient sample may be limited by the fact that our hospital is a major referral center for cancer and neurologic complications of cancer; wherever possible we compared our results with prior series or reports that either support or contradict our conclusions. Altered mental status was the most common reason for consultation and our findings were representative of the larger population of cancer patients. The majority of patients had a multifactorial, toxic metabolic encephalopathy with the use of opiates contributing to ⬎ 33% of cases. Encephalopathy most often is Neurologic Complications of Ovarian Carcinoma/Abrey and Dalmau a dose-limiting event during opiate titration, and typically is observed in older patients receiving an intravenous infusion of any opiate. The only opiate with a particular risk of encephalopathy is meperidine; the toxic accumulation of its metabolite, normeperidine, occurs with chronic use and causes a florid delirium, myoclonus, and seizures. Greater than 50% of the patients with brain metastases presented with altered mental status and a nonfocal neurologic examination. Similar findings have been reported in other malignancies2; 30 –75% of patients with brain metastases develop behavioral or cognitive abnormalities at presentation, most likely because these patients are more prone to encephalopathy induced by opiates or other metabolic derangements. Complications of chemotherapy were observed in 16% of patients. Cisplatin and paclitaxel alone or in combination accounted for the majority of peripheral neuropathies. Cisplatin typically produces a large fiber neuropathy with loss of vibratory and position sense, whereas paclitaxel affects all sensory fibers and also may cause a proximal motor neuropathy.2 Either concomitant or sequential administration of cisplatin and paclitaxel enhances neurotoxicity and can result in debilitating peripheral neuropathy.4,5 Twenty percent of patients in this series who received ifosfamide developed a reversible encephalopathy. Ifosfamide-related encephalopathy is reported to occur in as many as 30% of patients treated; it typically begins within 24 hours of infusion and usually is reversible.6 The only available intervention for chemotherapyinduced peripheral neuropathy is to discontinue the offending chemotherapy. This is frustrating for the oncologist, the neurologist, and the patient because it often means discontinuing the optimal antitumor drug. Our practice has been to recommend either discontinuing or reducing the dose of the offending agent when the patient develops objective weakness or sensory loss that results in functional impairment. An important contribution of the neurologist is to eliminate other potential causes of peripheral neuropathy so that effective chemotherapy is not withdrawn unnecessarily. Several potential neuroprotective and therapeutic agents (Org 2766, nerve growth factors, and ethiofus) are in different stages of preclinical and clinical evaluation.7–9 In longitudinal studies of ovarian carcinoma, brain metastases are estimated to occur in 0.29 – 4% of patients.10-15 There has been concern that the im- 131 proved treatment of systemic disease may lead to an increase in the incidence of brain metastases, either as a result of prolonged survival and/or as a result of the brain acting as a sanctuary site for tumor cells. Our findings are in keeping with other reports in which the median reported interval from the time of the diagnosis of ovarian carcinoma to brain metastases ranges from 14.5–34 months and the median survival after diagnosis of brain metastases ranges from 1–9 months.10 –14,16 –18 A recent report by Rodriguez et al. suggests that aggressive treatment with a combination of surgery, radiation, and chemotherapy leads to improved survival. In their meta-analysis patients treated with a combination of surgery, radiation, and chemotherapy had a median survival of 16.5 months compared with 10 months for patients treated with surgery and radiation, and 3 months for patients treated with radiation alone.19 Similarly, we found that our long term survivors received more aggressive, multimodality treatment with surgery, radiation, and chemotherapy. Prior studies have reported a predilection of pelvic tumors to metastasize to the posterior fossa.20 Our small series does not support this predilection. All solitary metastases were located in the cerebral hemispheres and all patients with multiple metastases had supratentorial lesions. The presence of solitary versus multiple metastases had no impact on patient survival; even patients with multiple lesions benefited from an aggressive combination of surgical resection or chemotherapy followed by radiation. Cerebrovascular disease was common in this series, accounting for 12% of consultations. Large retrospective series indicate that 15% of cancer patients have cerebrovascular lesions at autopsy; only approximately 50% are ever symptomatic and the etiology of cerebrovascular symptoms usually is cancer-related.3 Our patients with cerebrovascular events were older than the overall patient group and the majority had risk factors for stroke that were not related to their cancer. However, all these patients had active, advanced stage tumor and in at least 43% the etiology of the cerebrovascular event was related to the tumor or its treatment. Dehydration, electrolyte abnormalities, and alteration in baseline blood pressure are all potential side effects of abdominal surgery and chemotherapy, and may increase the risk of stroke in patients with preexisting cerebrovascular risk factors. Hyper- Š FIGURE 3. Coronal T1-weighted magnetic resonance imaging of the brain with gadolinium demonstrating response of a solitary brain metastasis to treatment with ifosfamide. (a and b) Scans taken prior to treatment. (c and d) Scans taken after five cycles of ifosfamide. 132 CANCER January 1, 1999 / Volume 85 / Number 1 coaguable states and NBTE are complications of disseminated malignancy and are independent cerebrovascular risk factors. It is interesting to note that no patient had a hemorrhagic stroke or cerebral sinus thrombosis. Paraneoplastic syndromes are rare, and considered to occur in 1–3% of patients with cancer. The incidence of subacute cerebellar degeneration associated with ovarian carcinoma is estimated to be 0.1%.2 Paraneoplastic dermatomyositis is even more uncommon and some authors question whether its association with cancer is coincidental rather than a true paraneoplastic event. However, in patients age ⬎ 50 years, dermatomyositis appears to have a higher than expected association with malignancy, among which ovarian carcinoma is a common offender.21 Anti-Yo antibodies are the most common paraneoplastic marker in patients with subacute cerebellar degeneration and ovarian carcinoma and these were identified in one of our three patients with cerebellar degeneration. Anti-Ri antibodies, found in another patient, develop in association with breast and ovarian carcinoma; these patients have a syndrome characterized by opsoclonus and truncal ataxia. The fact that one patient had no detectable paraneoplastic antibody in her serum but her tumor expressed both Hu and Ri antigens is unusual. A prior study comparing the tumors of patients with and without paraneoplastic cerebellar degeneration demonstrated that only the tumors of patients with paraneoplastic cerebellar degeneration expressed neural antigens.22 This study has several limitations. First, selection bias may have affected our patient population and may account for the high rate of multiple malignancies. The frequency of multiple tumors illustrates the importance of considering a second malignancy in patients with one cancer diagnosis; those patients with diagnoses of breast and ovarian carcinoma may harbor a BRCA gene mutation. The presence of multiple malignancies also complicates the neurologic evaluation. The toxicity of cumulative treatments may be additive or synergistic, and metastases have more than one possible tumor of origin. Because only a few patients had more than one active tumor at time of neurologic consultation, we did not exclude them from our analysis, but sought to clarify their etiology or pathology when possible. Second, this group of patients was selected retrospectively from a departmental database and we cannot exclude the possibility that some patients with ovarian carcinoma and neurologic complications were not included. Particularly in the instance of chemotherapy-induced peripheral neuropathy, the neurology service often is consulted only for severe or unremitting cases. Also, many patients may come to our institution for a single consultation, but receive their care elsewhere, thus artificially inflating the total number of ovarian carcinoma patients seen over the 3-year period. Therefore, although we can describe the range of neurologic complications observed in this ovarian carcinoma population, we cannot predict the incidence of specific complications accurately. REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. Beahrs OH, Henson DE, Hutter RVP, Kennedy BJ, editors. American Joint Commission on Cancer manual for staging of cancer. 4th edition. Philadelphia: J.B. Lippincott Co., 1992:155–79. Posner JB. Neurologic complications of cancer. Philadelphia: F.A. Davis, 1995. Graus F, Rogers LR, Posner JB. Cerebrovascular complications in patients with cancer. Medicine (Baltimore) 1985;64: 16 –35. Gordon AN, Stringer CA, Matthews CM, Willis DL, Nemunaitis J. Phase I dose escalation of paclitaxel in patients with advanced ovarian cancer receiving cisplatin: rapid development of neurotoxicity is dose-limiting. J Clin Oncol 1997;15: 1965–73. Cavaletti G, Bogliun G, Marzorati L, Zincone A, Marzola M, Colombo N, et al. Peripheral neuropathy of taxol in patients previously treated with cisplatin. Cancer 1995;75:1141–50. Watkin SW, Husband DJ, Green JA, Warenius HM. Ifosfamide encephalopathy: a reappraisal. Eur J Cancer Clin Oncol 1989;25:1303–10. Apfel SC, Arezzo JC, Lipson L, Kessler JA. Nerve growth factor prevents experimental cisplatin neuropathy. Ann Neurol 1992;31:76 – 80. Mollman JE, Glover DJ, Hogan WM, Furman RE. Cisplatin neuropathy risk factors, prognosis, and protection by WR 27721. Cancer 1988;61:2192–5. Van Der Hoop RG, Vecht CJ, Van Der Burg MEL, Elderson A, Boogerd W, Heimans JJ, et al. Prevention of cisplatin neurotoxicity with an ACTH(4 –9) analogue in patients with ovarian cancer. N Engl J Med 1990;322:89 –94. Larson D, Copeland L, Moser R, Malone JM Jr., Gershenson DM, Wharton JT, et al. Central nervous system metastases in epithelial ovarian cancer. Obstet Gynecol 1986;68:746 –50. Dauplat J, Nieberg RK, Hacker NF. Central nervous system metastases in epithelial ovarian cancer. Cancer 1987;60: 2559 – 62. Stein M, Steiner M, Klein B, Beck D, Atad J, Kuten A, et al. Involvement of the central nervous system by ovarian carcinoma. Cancer 1986;58:2066 – 8. Deutsch M, Beck D, Manor D, Brandes J. Metastatic brain tumor following second look operation in ovarian carcinoma. Gynecol Oncol 1987;27:116 –20. Ross WM, Carmichael VA, Shelley WE. Advanced carcinoma of the ovary with central nervous system relapse. Gynecol Oncol 1988;30:398 – 406. Ziegler J, Gliedman P, Fass D, Beckman M, Neophytides A, Steinfeld A. Brain metastases from ovarian cancer. J Neurooncol 1987;5:211–5. Neurologic Complications of Ovarian Carcinoma/Abrey and Dalmau 16. Mayer RJ, Berkowitz R, Griffiths T. Central nervous system involvement by ovarian carcinoma: a complication of prolonged survival with metastatic disease. Cancer 1978;41: 776 – 83. 17. Le Roux PD, Berger MS, Elliott P, Tamimi HK. Cerebral metastases from ovarian carcinoma. Cancer 1991;67:2194 –9. 18. Barker GH, Orledge J, Wiltshaw E. Involvement of the central nervous system in patients with ovarian carcinoma. Br J Obstet Gynaecol 1981;88:690 – 4. 19. Rodriguez GC, Soper JT, Berchuck A, Oleson J, Dodge R, Montana G, et al. Improved palliation of cerebral metastases in epithelial ovarian cancer using a combined modality ap- 133 proach including radiation therapy, chemotherapy and surgery. J Clin Oncol 1992;10:1553– 60. 20. Delattre JY, Krol G, Thaler HT. Distribution of brain metastases. Arch Neurol 1988;45:741– 4. 21. Ashour AA, Verschragen CF, Kudlelka AP, Kavanagh JJ. Paraneoplastic syndromes of gynecologic neoplasms. J Clin Oncol 1997;15:1272– 82. 22. Furneaux HM, Rosenblum MK, Dalmau J, Wong E, Woodruff P, Graus F, et al. Selective expression of Purkinje-cell antigens in tumor tissue from patients with paraneoplastic cerebellar degeneration. N Engl J Med 1990;322: 1844 –51.