Neuroimaging and treatment implications of patients with multiple epidural spinal metastases

BACKGROUND

Although multiple spinal epidural metastases (MEMs) commonly occur in cancer patients, their clinical significance remains uncertain. The authors attempted to ascertain the incidence of MEMs and their association with the completeness of spinal imaging by magnetic resonance (MR) scanning versus myelography to determine how often they are missed because of incomplete spinal imaging and to assess their prognostic and treatment implications.

METHODS

A review of 337 epidural spinal cord compression (ESCC) cases seen at the Mayo Clinic between 1985 and 1993 was conducted.

RESULTS

ESCC patients undergoing myelography only were significantly more likely to undergo complete spinal imaging (CSI)than patients undergoing either MR scan only or both imaging modalities (P < 0.0001). MEMs were detected in 32% of patients undergoing CSI and 18% of patients with incomplete spinal imaging (P=0.02). Failure to image the cervical spine in patients with symptomatic thoracic or lumbar epidural lesions would have missed secondary epidural lesions in only 1% of patients; however, this figure increased to 21% for failure to image either the thoracic or lumbosacral spine when symptomatic disease was located elsewhere. Radiation oncologists included secondary epidural deposits in treatment ports in 93% of MEM cases. In a multivariate model, the presence of MEMs was an independent prognostic factor for poorer survival.

CONCLUSION

The incidence of MEMs in patients with ESCC is approximately 30%, and their presence frequently alters treatment plans. It appears safe to forgo cervical spine MR scanning in patients with radiographically verified thoracic or lumbar ESCC; however, careful imaging of the thoracic and lumbar spine should be considered in all ESCC patients to detect MEMs.

Adrenoleukodystrophy: increased plasma content of saturated very long chain fatty acids. 1981

With a new method we measured the saturated very long chain fatty acids in the plasma of adrenoleukodystrophy (ALD) hemizygotes, ALD heterozygotes, and controls. ALD hemizygotes showed increased levels of hexacosanoate (C26 fatty acid) which represented 0.081 ± 0.0066% (SEM) of total fatty acids, compared to 0.015 ± 0.0032% in the controls. C25, C24, and C23 fatty acids were also increased, but the C22 and C20 fatty acids were normal. C26 levels were also increased in most ALD heterozygotes, with a mean level 0.057± 0.0063% of total fatty acids. The technique can be used for diagnosis and carrier identification, and in the evaluation of therapy.

Risk factors for primary central nervous system lymphoma: a case-control study

BACKGROUND

The incidence of primary central nervous system lymphoma (PCNSL) has increased over time in both immunocompetent and immunodeficient individuals. The reasons for the increase among immunocompetent patients are unclear.

METHODS

The authors conducted a case-control study of PCNSL at the Mayo Clinic based on cases from local community and referral practices. Cases were all PCNSL patients without immunodeficiency diagnosed between 1975 and 1994 (n = 109). Two groups of controls were selected from the same referral practice, namely, patients with other cancer (cancer controls; n = 101), and patients with a different neurologic disease (neurologic controls; n = 109) seen at our institution during the same time period.

RESULTS

PCNSL was significantly associated with lower education when cases were compared with cancer controls but only suggestively when cases were compared with neurologic controls. PCNSL cases were less likely to have had a history of tonsillectomy or oral contraceptive use compared with both control groups. The findings regarding autoimmune disorders and cardiovascular diseases were inconsistent for the two control groups. Neither farming nor prior personal or family history of cancer appeared to be risk factors for PCNSL.

CONCLUSIONS

The findings of this study warrant further investigation of tonsillectomy and oral contraceptives as possible factors for PCNSL.

Spinal epidural metastasis as the initial manifestation of malignancy: clinical features and diagnostic approach

The objective of our study was to delineate clinical features and specific diagnostic and therapeutic implications of spinal epidural metastasis (SEM) occurring as the initial manifestation of malignancy (IMM)-a less common event than SEM in the setting of previously established malignancy (PEM). We performed a retrospective review of the clinical histories of 337 patients seen at Mayo Clinic with a radiographically verified diagnosis of SEM between January 1, 1985, and December 31, 1993. Twenty percent of all cases of SEM occurred as SEM-IMM. Carcinoma of the lung, cancer of unknown primary site, multiple myeloma, and non-Hodgkin's lymphoma were disproportionately represented in SEM-IMM, together accounting for 78% of SEM-IMM patients and only 26% of SEM-PEM patients. Inversely, breast and prostate carcinoma contributed only 12% of SEM-IMM patients but 51% of SEM-PEM patients. Only 27% of patients with SEM-IMM had nonspecific symptoms suggesting malignancy, and in only 24% did the history or physical examination suggest the primary site of malignancy. Percutaneous needle biopsy of the vertebral lesion was diagnostic of malignancy in 18 of 19 patients (95%), and no complications ensued. Fifteen patients underwent diagnostic laminectomy with biopsy, and one had a fatal complication. Survival after the diagnosis of SEM did not differ between patients with SEM-IMM and those with SEM-PEM; the median survival was 6.6 months. SEM not uncommonly occurs as the presentation of malignancy, and often it produces the only symptoms or signs of malignancy. The great majority of neoplasms presenting with SEM are carcinomas of the lung, unknown primary lesions, and hematologic malignancies. Computed tomography-guided needle biopsy is a safe, efficacious means of diagnosing malignancy, allowing for rational diagnostic workup and staging. In most patients it obviates a diagnostic spinal surgical procedure.

Intramedullary spinal cord metastases: clinical features and treatment outcome

Our objective was to delineate clinical features and treatment outcome of patients with intramedullary spinal cord metastasis (ISCM). There are no reports of a large experience with this rare cancer complication. We reviewed records retrospectively from 1980 to 1993 to identify patients with histologically confirmed systemic cancer, clinical features of myelopathy, and either tissue-proven ISCM or abnormal neuroimaging findings consistent with ISCM. We identified 40 patients who fulfilled these criteria. In nine, ISCM was the initial presentation of cancer. Nineteen patients had lung primaries (small cell in 12). Twenty-one patients had pain, 35 had demonstrable sensory loss, 37 had weakness, and 25 had urinary incontinence at presentation. Nine patients had true Brown-Séquard syndrome and nine others had pseudo-Brown-Séquard syndrome. Median duration of symptoms at diagnosis was 28 days (range 3 days to 18 months). Thirteen patients had prior brain metastasis, nine had brain metastasis simultaneous with ISCM, and one had brain metastasis after ISCM; 11 had concomitant leptomeningeal metastases. Spinal magnetic resonance findings were abnormal in 30/30 patients, myelographic results were abnormal in 16/20, and eight had pathologic confirmation of ISCM. Thirty-five patients had radiotherapy and five had surgery; four were untreated or treated elsewhere. Median survival was 4 months for patients receiving radiotherapy and 2 months for patients not receiving radiotherapy. Eleven patients survived > 6 months. Twenty-three were ambulatory at ISCM diagnosis, and 21 were ambulatory at letest follow-up. We conclude that ISCM as the initial presentation of malignancy is not rare, and hemicord syndromes occur frequently. Although long-term survival is poor, treatment preserves ambulation in most patients still ambulatory at diagnosis. Focal radiotherapy is indicated in most patients.

Diagnosis and management of primary spinal epidural non-Hodgkin's lymphoma

OBJECTIVE

To describe the diagnosis, management, and outcome in 10 patients with histologically confirmed primary spinal epidural non-Hodgkin's lymphoma.

MATERIAL AND METHODS

We review the findings in a cohort of seven men and three women in whom this tumor was diagnosed between January 1979 and January 1993 and discuss the prognostic differences between primary and secondary spinal lymphomas.

RESULTS

All patients (median age at diagnosis, 70 years) underwent a decompressive laminectomy, subtotal tumor resection, and spinal irradiation (median dose, 3,800 cGy). Nine of 10 tumors were of B-cell origin. Six patients are alive and well. In four patients, recurrent disease developed from 15 to 62 months after the original diagnosis; of these, one has died. The median duration of survival of all patients was 42 months; of those living more than 24 months after diagnosis, the median duration of survival was 80 months.

CONCLUSION

A rapidly progressive spinal cord or cauda equina syndrome with neuroimaging findings consistent with an extradural compressive lesion should alert caregivers to the possibility of spinal epidural lymphoma. Although the prognosis for patients with secondary spinal epidural non-Hodgkin's lymphoma is often poor, primary spinal epidural non-Hodgkin's lymphoma can be associated with a favorable outcome if diagnosed and treated early.

Primary central nervous system non-Hodgkin's lymphoma: survival advantages with combined initial therapy?

PURPOSE

Results of multiple radiation, chemotherapy, and combined treatment trials have shown that the fate of primary central nervous system lymphoma (PCNSL) patients is very different from that of patients with similarly treated systemic IE non-Hodgkin's lymphoma. This study was designed to improve the survival of PCNSL patients by the use of combined initial therapy.

METHODS AND MATERIALS

Forty-six eligible primary PCNSL patients were treated with whole brain irradiation and adjuvant chemotherapy consisting of preirradiation cyclophosphamide-adriamycin-vincristine-prednisone (CHOP) and postirradiation high-dose cytosine arabinoside (HDAC) as part of an ongoing Phase II Mayo/North Central Cancer Treatment Group/Eastern Cooperative Oncology Group (M/NCCTG/ECOG) intergroup effort, which opened in April 1986.

RESULTS

This cohort consisted of 23 men and 23 women with median age 63.5 years (range 24 to 75 years). Only 5% were under age 40; 36% were age 40 to 59, 37% were age 60 to 69, and 22% were age 70 and over. Forty-six percent had good performance scores of ECOG 0-1 at time of study entry. Forty-six patients were evaluable for treatment outcome as of October 6, 1993. Of these, 10 were still alive. Estimated median survival and 21-month survival were 45.3 weeks and 29%, respectively. There were four early deaths ranging from Day 9 to Day 15 (three drug-related, one from other complications), and two CHOP responders died at 32 and 35 days, soon after Cycle 2 of CHOP (one probably drug-related, one from other complications). There was no significant difference in survival according to baseline performance status. However, survival was consistently worse for patients > 60 years old than for the younger patients (< or = 60 years). With deaths recorded for 21 of 21 older patients, but only 9 of the 14 younger patients, 21-month survival for older vs. younger was 14 vs. 50% based on the 35 patients who entered the study at least 21 months ago (p = 0.0365). Of the 46 patients evaluable for response, 63% had objective remissions on CHOP and another 20% remained stable.

CONCLUSION

Combined modality therapy in this study did not produce an overall survival advantage in treating PCNSL. The 50% 21-month survival of younger patients may be a reflection of age only.

Primary intracerebral malignant lymphoma: a clinicopathological study of 89 patients

The authors report on a clinicopathological study of 89 surgical patients with histologically proven primary parenchymal brain lymphoma, all diagnosed between January 1975 and December 1990. The cohort included 60 men and 29 women whose median age at diagnosis was 60 years (range 14 to 84 years). The duration of symptoms was less than 8 weeks in 48% of the patients. Symptom groups included focal neurological deficit (73%), neuropsychiatric symptoms (28%), seizures (9%), and increased intracranial pressure (3%). A total of 132 tumors were seen in 89 patients: the most common sites were frontal (32 patients), temporoparietal (31 patients), and basal ganglia (17 patients); multiple lesions were reported in 23 patients. No patient had antecedent of human immunodeficiency virus positivity or acquired immunodeficiency syndrome. A family history of cancer was present in 33% of the patients, three-quarters of whom were first-degree relatives. Histological subtypes (National Cancer Institute Working Formulation) included 64 large cell (72%) and 13 immunoblastic (15%) tumors. Phenotype was determined in 66 patients: 63 were B-cell type and three were T-cell type. Surgical resection was performed in 47% of the cases, with the remainder undergoing biopsy only. All but six patients received radiation therapy. Thirty-one patients received chemotherapy, whereas 46 patients did not; data on the remaining 12 patients were unavailable. The end point of the study was death from any cause. At the time of last contact, 69 of the patients (78%) had died; the median survival time for this study group was 20.9 months. On univariate analysis, prognostic factors significantly associated with survival included age at diagnosis, family history of cancer, and focal neurological deficit. Multivariate analysis revealed four unfavorable prognostic factors: age greater than or equal to 60 years, history of cancer in first-degree relatives, focal deficit, and ependymal contact. After adjustment for these variables, clinical syndrome, size and number of lesions, extent of surgery, histological cell type, radiation dose, and use of chemotherapy were not significantly associated with survival.

Occult systemic non-Hodgkin's lymphoma (NHL) in patients initially diagnosed as primary central nervous system lymphoma (PCNSL): how much staging is enough?

OBJECTIVE

To demonstrate the extent of staging necessary to exclude occult systemic stage IV NHL before making a diagnosis of stage I AE PCNSL.

BACKGROUND

The diagnosis of PCNSL requires the demonstration of malignant lymphocytes within the CNS (usually by biopsy) and finding no evidence of systemic NHL. Different staging approaches have been recommended, ranging from extensive systemic evaluation (including bone marrow examination) to a more focused approach (abdominal and pelvic CT) to no systemic evaluation. We have employed a staging regimen that included: ophthalmologic evaluation (including slit lamp examination); CT of chest, abdomen, and pelvis; bilateral iliac crest aspirate and biopsy; flow cytometry of circulating lymphocytes; and, in men, testicular ultrasound.

DESIGN/METHODS

We carried out a retrospective review of 128 patients entered into the Mayo Lymphoma Project data bank (1975-1994).

RESULTS

Between the years 1975 and 1994, five patients (3.9%) were identified who fulfilled criteria for the diagnosis of PCNSL (typical clinical history, pathognomonic neuro-imaging, and histologic proof of NHL in brain tissue) but who had occult systemic NHL on staging (bone marrow 1, abdominal lymph nodes 3), or at autopsy (colon 1). Case histories are presented.

CONCLUSIONS

Patients with apparent PCNSL may have systemic NHL. Complete staging is essential to the initial management of patients presenting as PCNSL to exclude systemic stage IV disease.

Brain metastatic lesions

OBJECTIVE

To describe current concepts in the diagnosis and treatment of brain metastases.

RESULTS

More than 25% of all autopsy-proven brain metastases have a pulmonary source. Most brain metastases manifest with a combination of focal and generalized symptoms and signs. Typically, patients have subacute, progressive symptoms. In most situations, a computed tomographic scan of the head provides sufficient neuroimaging and allows one to monitor the effects of therapy. Magnetic resonance imaging has become increasingly useful in the diagnosis and management of brain metastases. It can detect computed tomographic occult metastases, identify associated leptomeningeal disease, and reveal early therapeutic complications.

CONCLUSION

Treatment options for patients with brain metastases include corticosteroids, whole-brain radiation therapy (WBRT), surgical intervention, stereotactic radiosurgical techniques, and chemotherapy. Corticosteroids produce prompt improvement in most patients; however, prolonged use is associated with considerable risks. For most patients, WBRT is the preferred treatment. Nonetheless, it has associated nonneurologic and neurologic complications, some of which are serious. In patients with a single metastasis, surgical removal should be considered. Recent studies have suggested that resection of a single metastatic lesion followed by radiation therapy offers better survival than does radiation therapy alone. The subsequent administration of WBRT after radiosurgical treatment has become standard practice. The role of chemotherapy is uncertain.

Neurological complications of peripheral and cutaneous T-cell lymphomas

To determine the frequency and pattern of neurological complications of T-cell lymphoma (TCL), we retrospectively reviewed the medical records of 316 patients with TCL diagnosed between January 1984 and May 1991. Disease entities not included in this study were lymphoblastic lymphoma, primary central nervous system lymphoma, CD30-positive anaplastic large cell lymphoma, and lymphomas secondary to human immunodeficiency virus or human T-cell lymphotropic virus type I. Cases were classified as having direct complications (parenchymal, leptomeningeal, epidural, or peripheral) or indirect complications (paraneoplastic, disease related, or treatment related). Preexisting neurological conditions were excluded. The overall rate of neurological complications was 7.9%. The frequency of neurological complications in peripheral TCL and cutaneous TCL was 17% and 3%, respectively, with at least half of the neurological complications in both conditions due to direct involvement of the nervous system. Direct neurological complications of TCL were primarily due to leptomeningeal and parenchymal involvement. There were no cases of epidural spinal cord disease.

Chronic meningitis: the role of meningeal or cortical biopsy

Meningeal and cortical biopsies were evaluated in 37 patients (25 men and 12 women; mean age, 54 yr) who had chronic meningitis of an unknown cause between 1985 and 1993 (the era of magnetic resonance imaging). Magnetic resonance imaging with gadolinium contrast was the most useful diagnostic imaging technique, demonstrating meningeal enhancement in 15 of 32 patients (47%). Only 2 of 32 (6%) computed tomographic scans revealed enhancement. A definitive diagnosis was made in 16 of 41 biopsies (39%), but in cases where enhancement was present on either magnetic resonance imaging or computed tomography, a diagnosis was obtained in 80% (12 of 15 cases). Only 2 of 22 biopsies (9%) from nonenhancing regions were diagnostic. Although the locations of enhancement were distributed evenly, biopsies through suboccipital and pterional craniotomies gave the highest diagnostic yields (50%). Furthermore, if the biopsies were obtained from enhancing regions, the yield of these two approaches increased to 84 and 100%, respectively. Of 18 cases in which biopsy samples were taken from both the meninges and cortex, only 1 had cortical involvement alone. The meninges were therefore diagnostic in 15 of the 16 definitive diagnostic cases (94%). Second biopsies were necessary in four cases, of which the three biopsies from enhancing regions were diagnostic. The most frequent causes of chronic meningitis were sarcoid (31%) and metastatic adenocarcinoma (25%).(ABSTRACT TRUNCATED AT 250 WORDS)

[Stereotaxic biopsy for primary lymphomas of the central nervous system]

Primary lymphomas of the CNS are rare. However the incidence of these lesions has increased recently. These tumors are often situated in the thalamus and basal ganglia, but can be multifocal. The treatment for primary CNS lymphoma is radiotherapy alone or in association with chemotherapy. Because the associated morbidity is very low stereotactic biopsy is an ideal method for determining the histologic diagnosis in patients with suspected CNS lymphoma. The authors present a study of 49 patients with primary CNS lymphomas in which a stereotactic biopsy was performed.

Primary spinal epidural non-Hodgkin's lymphoma: report of eight patients and review of the literature

Eight patients treated for histologically confirmed primary spinal epidural non-Hodgkin's lymphoma diagnosed between January 1979 and August 1989 (6.6% of all cases of intraspinal lymphoma) were studied. There were six men and two women. The median age was 70 years (range, 43-80 yr). Patients sought treatment for a prodrome of back pain (median duration, 3 mo) followed by an acute neurological deterioration (median duration, 6 d). The most common findings were a discrete sensory level in 5 patients, hyperreflexia in 5 patients, and paraparesis or paraplegia in 5 patients. Radiographically, there was an absence of bony destruction by these tumors. All patients underwent a decompressive laminectomy, subtotal tumor resection, and spinal irradiation (median dose, 3800 cGy). Two patients had low-grade lymphomas (one B cell and one T cell), and 6 patients had intermediate-grade lesions (six B cell). Two patients with B-cell lymphomas (one low-grade and one intermediate-grade) developed metastatic disease 15 and 17 months after the initial diagnosis; no evidence of lymphoma developed in the other 6 patients. The median survival was 22 months (range, 2-71 mo). Lymphoma was the cause of death in only 1 of the 4 patients who died, and the 4 younger patients are alive and well. Primary spinal epidural non-Hodgkin's lymphoma should be a diagnostic consideration in the older patient who seeks treatment for spinal cord compression manifested by a prodrome of back pain, followed by a rapid neurological deterioration, normal plain spine radiographs, and neuroimaging consistent with an extradural compressive lesion. Surgery for this diagnosis followed by spinal irradiation should result in significant neurological improvement.

Neurological features of cranial vault lymphomas: report of two cases

Two patients who had focal neurological deficits as the initial manifestation of a malignant non-Hodgkin's lymphoma involving the skull are described. Soft-tissue masses and variable bone destruction were the predominant computed tomographic findings, and magnetic resonance imaging studies revealed meningeal involvement in one case and sinus thrombosis in the other. Malignant non-Hodgkin's lymphomas initially appearing in the skull are rare, but these lesions should be considered in patients with a palpable scalp mass who have focal neurological signs.

Granulomatous angiitis of the spinal cord associated with Hodgkin's disease

A 28-year-old man had a 5-month history of focal and generalized neurologic symptoms culminating in a thoracic myelopathy. Evaluation revealed granulomatous angiitis of the spinal cord in association with occult nodular sclerosing Hodgkin's disease. In previous reports, manifestations indicative of intracranial involvement have dominated the clinical presentation of granulomatous angiitis associated with Hodgkin's disease. Successful therapy for Hodgkin's disease may result in marked improvement of associated granulomatous angiitis, whereas the lack or failure of therapy results in a uniformly fatal outcome. Definitive antemortem diagnosis of granulomatous angiitis requires a biopsy of involved tissue. The cause of granulomatous angiitis, as well as the nature of its association with Hodgkin's disease, remains unexplained.

Paraneoplastic brachial plexopathy in a patient with Hodgkin's disease

We describe a case of inflammatory brachial plexopathy that occurred in the context of a mild, diffuse sensorimotor peripheral neuropathy associated with Hodgkin's disease. Clinical, electrophysiologic, and pathologic studies helped distinguish this disorder from other causes of brachial plexopathy in patients with cancer. Treatment with corticosteroids seemed beneficial in this patient. We suggest that this may be another type of paraneoplastic condition associated with Hodgkin's disease.

Primary leptomeningeal lymphoma: report of 9 cases, diagnosis with immunocytochemical analysis, and review of the literature

We describe 9 patients who presented with a neoplastic meningitis of lymphomatous origin. No evidence of parenchymal central nervous system or systemic tumor was identified either at the time of presentation or throughout the course of their disease. We have chosen to call this entity "primary leptomeningeal lymphoma" (PLML). This unusual form of neurologic lymphoma must be differentiated from the more common clinical situations of primary parenchymal lymphoma with meningeal involvement and systemic lymphoma complicated by lymphomatous meningitis.

Gynecologic cancer in patients with subacute cerebellar degeneration predicted by anti-Purkinje cell antibodies and limited in metastatic volume

Between 1982 and 1989, 19 patients with gynecologic carcinoma, paraneoplastic cerebellar degeneration, and seropositivity for anti-Purkinje cell cytoplasmic antibodies were identified at our institution. Seven of the patients had no clinical, computed tomographic, or magnetic resonance imaging evidence of cancer but had undergone laparotomy solely because anti-Purkinje cell antibodies were found in their serum; all had high-grade adenocarcinoma. Cerebellar symptoms preceded or coincided with the initial cancer diagnosis in 15 patients and preceded the diagnosis of recurrent cancer in 4 patients. The cancers were 14 ovarian, 2 fallopian tube, 2 surface papillary, and 1 poorly differentiated metastatic adenocarcinoma in a periaortic lymph node. Two remarkable surgical observations in patients with high-grade ovarian and tubal cancers were the conspicuous lack of peritoneal implants and the small metastatic volume. A comparison of the 8 patients who had primary stage III cancer with 24 matched control patients without paraneoplastic cerebellar degeneration revealed no difference in primary tumor volume but a significantly smaller volume of metastatic tumor in the seropositive group (P = 0.05). Anti-Purkinje cell antibodies were not detected in 111 neurologically normal patients with advanced ovarian cancer. The small metastatic volume in the face of high-grade and advanced stage malignancy in seropositive patients with paraneoplastic cerebellar degeneration suggests that an immune response to the tumor (presumably cross-reactive with cerebellar cells) may impair the metastatic process. Earlier diagnosis and treatment of cancer, based on prompt serologic testing, may offer an improved neurologic and oncologic prognosis.