Original Article
Leptomeningeal Carcinomatosis Prognostic Implications of Clinical and Cerebrospinal Fluid Features Jordi Bruna, MD1, Laura Gonza ´lez, MD1, Ju ´lia Miro ´, MD1, Roser Velasco, MD1, Miguel Gil, MD2, and 3 Avelina Tortosa, MD, PhD , for the Neuro-Oncology Unit of the Institute of Biomedical Investigation of Bellvitge
BACKGROUND: Leptomeningeal carcinomatosis (LC) represents a devastating complication of systemic cancer, and patients with LC have a dismal prognosis and increased mortality. The few studies that have focused on the evaluation of prognostic factors in patients with LC have resulted in contradictory results. Thus, the treatment of LC remains controversial, and no straightforward guidelines exist in the literature. The objective of the current study was to identify prognostic markers related to LC survival to better select patients who are eligible for intensive treatment. METHODS: Seventy patients who had a diagnosis of LC were reviewed, and clinical data, cerebrospinal fluid (CSF) parameters, tumor-related characteristics, and treatment information were registered. The impact of single parameters on overall survival was determined by both univariate and multivariate analyses. RESULTS: The multivariate analysis revealed that Radiation Therapy Oncology Group score 2 (P ¼ .028), a glucose level in CSF 2.7 mmol/L (P ¼ .001), the presence of infratentorial symptoms at onset (P ¼ .026), and intrathecal treatment (P < .001) were associated independently with longer overall survival in patients with LC. In addition, the same clinical factors also predicted response to treatment in such patients. CONCLUSIONS: The predictive factors for patients with LC that were identified in this study could help to better select patients who are more likely to benefit from C 2009 American Cancer Society. chemotherapy. Cancer 2009;115:381--9. V KEY WORDS: cerebrospinal fluid, leptomeningeal carcinomatosis, meningeal carcinomatosis, prognostic factors.
Leptomeningeal carcinomatosis (LC) is a devastating complication of systemic cancer that occurs in 5% to 10% of patients with solid tumors, more frequently adenocarcinoma, and is observed most commonly in patients with breast cancer, lung cancer, or melanoma.1-11 Recently, an increased incidence of LC was observed because of the development of new, effective antineoplastic treatments against primary tumors and improvements in neuroimaging techniques.7,12 Overall, current treatments offer a poor outcome and a high frequency of treatment-related toxicity. Without therapy, the median survival is 4 to 6 weeks, which can be extended to 4 months to 6 months with chemotherapy in selected patients.13 In most studies, at least 55% of treated patients die within a few weeks after starting treatment. There is only a 20% response to treatment, and the median survival is 4 months, although Corresponding author: Avelina Tortosa, MD, PhD, Department of Basic Nursing, School of Nursing, Campus of Health of Bellvitge, University of Barcelona, IDIBELL, Pavello´ Central 3a pl., C/feixa Llarga s/n, 08907 L’Hospitalet de Llobregat, Barcelona, Spain; Fax: (011);
[email protected] 1 Department of Neurology, University Hospital of Bellvitge, L’Hospitalet de Llobregat, Barcelona, Spain; 2Department of Oncology, Hospital Duran and Reynals Catalan Institute of Oncology, L’Hospitalet de Llobregat, Barcelona, Spain; 3Department of Basic Nursing, School of Nursing, University of Barcelona, IDIBELL, L’Hospitalet de Llobregat, Barcelona, Spain
We thank Mr. Tom Yohannan for editorial assistance and Dr. Francesc Graus for his review of the article and advice. Received: July 23, 2008; Accepted: August 27, 2008 C 2008 American Cancer Society Published online: December 24, 2008, V
DOI: 10.1002/cncr.24041, www.interscience.wiley.com
Cancer
January 15, 2009
381
Original Article
some patients can survive for >1 year. To date, these significant differences in survival rates in LC have not been explained and remain unpredictable.14,15 Prognostic factors in LC may help the clinician to select the patients who would benefit most from chemotherapy. However, studies focused on the evaluation of prognostic markers in patients with LC have produced contradictory results, most likely because of methodologic differences among them. There are weaknesses in the studies: Some have included in their series hematologic or primary brain tumors, they show important differences in sample size and analyzed variables, they have evaluated selected subsets of patients, and only univariate analyses were considered.5,9,11,13-20,22 Accordingly, currently, there are no useful prognostic markers, except for cerebrospinal fluid (CSF) blocks,23,24 that could help clinicians to decide the best therapeutic approach for patients with LC. The objective of this study was to identify prognostic factors related to LC survival and enable clinicians to select patients who are more likely to benefit from chemotherapy treatment.
MATERIALS AND METHODS Seventy patients who were diagnosed with LC at the University Hospital of Bellvitge between 1990 and 2007 were evaluated. Patients with leptomeningeal infiltration either from hematologic malignancies or primary brain tumors were not included in the current analysis. The diagnosis of LC was based on the detection of malignant cells in CSF (n ¼ 62 patients) or on the demonstration of findings consistent with LC on neuraxis magnetic resonance imaging (MRI), biochemical abnormalities in the CSF, and suitable clinical signs and symptoms.
Evaluated Parameters Demographic data, clinical features, primary tumor characteristics, CSF findings, and treatment were evaluated. The clinical features at LC presentation were classified into 5 groups according to their origin: 1) supratentorial symptoms (seizures, cognitive impairment, difficulty in walking, sensory or unexplained weakness from brain metastases, decreased visual acuity without endocranial hypertension, and dysphasia), 2) infratentorial symptoms
382
Table 1. Radiation Therapy Oncology Group Neurologic Function Classification
RTOG Neurologic Description Function Classification 1 2
3
4
Able to work or perform normal activities; neurologic findings minor or absent Able to carry out normal activity with minimal difficulties; neurologic impairment does not require nursing care or hospitalization Seriously limited in performing normal activities; requiring nursing care or hospitalization; patients confined to bed or wheelchair or have significant intellectual impairment Unable to perform even minimal normal activities; requiring hospitalization and constant nursing care feeding; patients unable to communicate or in coma
RTOG indicates Radiation Therapy Oncology Group.
(cranial nerves, radicular or spinal involvement), 3) endocranial hypertension, 4) endocranial hypertension plus other symptoms, and 5) encephalopathy. Patients were included in only 1 group according to their main clinical manifestations at onset. To further evaluate patients’ physical capability, the Radiation Therapy Oncology Group (RTOG) neurologic functional status score (Table 1) at the time of LC diagnosis was calculated.25 The RTOG score was used because it was estimated more reliably from medical records when data regarding Karnofsky performance status (KPS) were not available. Regarding tumor-related characteristics, the histology and location of the primary tumor, the primary disease status (stable or response vs progression), and the presence or absence of brain metastases at the time of LC diagnosis were analyzed. CSF findings included fluid cytology as well as measurement of glucose and protein concentrations at LC presentation. With respect to treatment, previously administered chemotherapy regimens for the primary neoplasm, time from the primary tumor diagnosis to the onset of LC, and therapy for LC were evaluated. Indeed, regarding treatment of LC, 36 patients (51.4%) received intrathecal methotrexate administered by lumbar puncture at a dose of 10 mg twice weekly. Ten of these patients also received systemic concomitant chemotherapy, and 6 patients also received systemic adjuvant chemotherapy and wholebrain irradiation. Finally, the remaining 34 patients Cancer
January 15, 2009
Leptomeningeal Carcinomatosis Prognosis/Bruna et al
Table 2. Clinical and Cerebrospinal Fluid Characteristics of Patients with Leptomeningeal Carcinomatosis
Characteristic
No. of Patients (%)
Mean age6SD, y
55.5611.38
Sex 29 (41.4) 41 (58.6)
Men Women
Median RTOG neurologic function classification [range]
2 [1-4]
Clinical onset Infratentorial symptoms Endocranial hypertension plus other symptoms Supratentorial symptoms Encephalopathy Endocranial hypertension
Mean6SD duration of symptoms prior to diagnosis, d
32 (45.7) 17 (24.3) 12 (17.1) 6 (8.6) 3 (4.3)
23.16619.14
Tumor Characteristics
No. of Patients (%)
Histology Adenocarcinoma Oat cell Carcinoma Melanoma Mesothelioma
50 10 6 3 1
(71.4) (14.3) (8.6) (4.3) (1.4)
30 27 4 3 3 3
(42.9) (38.6) (5.7) (4.3) (4.3) (4.3)
Primary tumor Breast Lung Gastrointestinal Skin Genitourinary Others
Brain metastasis Yes No
24 (34.3) 46 (65.7)
Status of primary tumor
CSF characteristics, median [range] Cells Proteins, g/L Glucose, mmol/L
Table 3. Tumor and Treatment Characteristics
Response or stable Progression 4 [0-130] 2.4 [0.15-25.8] 2.7 [0.07-8]
SD indicates standard deviation; RTOG, Radiation Therapy Oncology Group; CSF, cerebrospinal fluid.
received only best supportive care. For purposes of the current analysis, patients were classified into a treatment group and a palliative group. Therapeutic decisions were made according to physicians’ criteria.
6 (8.6) 64 (91.4)
Median no. of chemotherapy treatments previous to LC [range]
1 [0-4]
Median time since primary diagnosis to LC, wk [range]
49.7 [0-685.1]
Breast Lung Others
179.8 [21.7-685.1] 2 [0-73.3] 80.1 [0-498]
Treatment of LC Yes No
36 (51.4) 34 (48.6)
LC indicates leptomeningeal carcinomatosis.
Statistical Analysis The primary endpoint of the study was overall survival, which we measured from the date of LC diagnosis to the date of either last follow-up visit or death. The date of LC diagnosis was defined as the date of the first positive cytologic study or the date of MRI diagnostic scans, depending on the diagnostic criteria. The analysis was performed for both in the whole series, to identify prognostic variables, and for the treatment group, to identify predictive factors of response to treatment. The univariate analysis was made by constructing probability curves according to the Kaplan-Meier method and comparing them by using the log-rank test. Variables that achieved a P value
