benefit of ventriculoperitoneal cerebrospinal fluid...

J Neurosurg 115:730–736, 2011

730 J Neurosurg / Volume 115 / October 2011

Neoplastic meningitis (NM) is an increasingly recognized complication of cancer that occurs in approximately 5%–10% of all cancer pa-

tients7,9,10,12,19,23,27 and almost 20% of those with neurologi-cal symptoms.24 The incidence of neoplastic meningitis is rising,12 most likely due to improved diagnostic tech-

niques, increasingly successful treatment of cancer out-side the nervous system, and the use of chemotherapeutic agents with little or no access to the CNS.2,9,12 The opti-mum treatment strategy for NM includes focal radiation to sites of bulky or symptomatic disease, systemic che-motherapy appropriate to the stage and histology of the underlying cancer, and intrathecal chemotherapy through a subcutaneously placed ventricular reservoir.9 Although the median survival of patients with NM is poor—2–4 months in randomized controlled trials4,14,15,18—about one-quarter of patients survive for 6 months, and ap-

Benefit of ventriculoperitoneal cerebrospinal fluid shunting and intrathecal chemotherapy in neoplastic meningitis: a retrospective, case-controlled study

Clinical articleNiNg LiN, M.D.,1,2 iaN F. DuNN, M.D.,1,2 MichaeL gLaNtz, M.D.,3 DaNa L. aLLisoN, B.a.,4 RaNDy JeNseN, M.D., Ph.D.,4 MaRk D. JohNsoN, M.D., Ph.D.,1,2 RoBeRt M. FRieDLaNDeR, M.D.,1,2 aND saNtosh kesaRi, M.D., Ph.D.5

1Department of Neurosurgery, Brigham and Women’s Hospital; 2Harvard Medical School, Boston, Massachusetts; 3Department of Neurosurgery, Milton S. Hershey Medical Center, Penn State College of Medicine, Hershey, Pennsylvania; 4Department of Neurosurgery, Huntsman Cancer Institute, University of Utah School of Medicine, Salt Lake City, Utah; and 5Department of Neurosciences, Moores Cancer Center, University of California, San Diego, La Jolla, California

Object. Neoplastic meningitis (NM) is a debilitating and increasingly frequent neurological complication of cancer characterized by infiltration of tumor cells into the leptomeninges and the subarachnoid space. Although NM is rarely curable, combined intrathecal chemotherapy and focal radiation can improve disease-related symptoms and survival. Hydrocephalus occurs in a significant proportion of patients, is associated with poor prognosis and reduced quality of life, and usually precludes the use of intrathecal therapy.

Methods. Since January of 2005, the authors have used a combined treatment approach for patients with both NM and hydrocephalus that employs a subcutaneously placed reservoir connected in series to an on/off valve and a ventriculoperitoneal shunt for both diversion of CSF and injection of intrathecal chemotherapy. They conducted a retrospective, case-controlled study from 2 independent institutions to review their experience.

Results. Twenty-four patients with NM and hydrocephalus underwent placement of a CSF reservoir-on/off valve-ventriculoperitoneal shunt (RO-VPS) construct. There was no perioperative mortality, and there were only 2 minor complications. One shunt failure and no shunt-associated infections were observed over a median of 28 weeks of follow-up. Symptomatic improvement and improved performance status were seen in 20 patients (83.3%) and were sustained over 6 months. Eighteen patients received intraventricular chemotherapy without unexpected toxicity, and cytological responses were found in 11 patients (61.1%). Median progression-free and overall survival was 14 and 31 weeks, respectively. Compared with a contemporaneous comparison group of 24 demographically matched patients with NM who underwent CSF reservoir placement only, those who received RO-VPS constructs (p = 0.02) and had primary diagnosis of breast cancer (p = 0.04) had significant advantage in overall survival.

Conclusions. A combined RO-VPS system is safe and practical to install, results in symptomatic improvement in most patients, and allows uncomplicated and effective administration of intrathecal chemotherapy in patients with NM. Cerebrospinal fluid diversion surgery should be considered in NM patients in conjunction with intrathecal and systemic treatments. (DOI: 10.3171/2011.5.JNS101768)

key WoRDs • ventriculoperitoneal shunt • intrathecal chemotherapy • neoplastic meningitis • leptomeningeal metastasis • on-off valve • oncology

Abbreviations used in this paper: KPS = Karnofsky Performance Scale; NM = neoplastic meningitis; RO-VPS = reservoir–on/off valve–ventriculoperitoneal shunt.

See the corresponding editorial in this issue, pp 728–729.

http://thejns.org/doi/abs/10.3171/2011.4.JNS11492

J Neurosurg / Volume 115 / October 2011

Multimodal treatment for neoplastic meningitis

731

proximately 12% for a year following the diagnosis of NM3,4,14,15 using this treatment approach.

Because NM is characterized by infiltration of can-cer cells throughout the leptomeninges and subarachnoid space, obstruction of CSF flow is common8,13 and leads to elevated intracranial pressure and hydrocephalus in half of patients.9,22,25 Without CSF diversion, patients with hydrocephalus or increased intracranial pressure can de-velop debilitating headaches, and cognitive, gait, sphinc-ter, and other neurological deficits, sometimes including herniation and death. Cerebrospinal fluid diversion tech-niques such as ventriculoperitoneal and lumboperitoneal shunting are safe and effective in treating hydrocephalus from multiple etiologies, but these techniques have not been used extensively in patients with NM, in part due to therapeutic nihilism and in part due to concerns about peritoneal spread of cancer.1,21 Furthermore, administer-ing intrathecal chemotherapy to patients with ventricu-loperitoneal shunts is problematic because a functioning ventriculoperitoneal shunt will shunt chemotherapy out of the CSF space, making it difficult to maintain an ef-fective concentration of chemotherapy within the CSF. To address these concerns, we have used a combined treat-ment approach for NM patients with hydrocephalus that employs an in-line CSF reservoir for injection of intra-ventricular chemotherapy, an on-off valve for CSF flow control, and a programmable ventriculoperitoneal shunt for diversion of CSF (RO-VPS). In this paper, we review our experience with this approach.

MethodsPatient Selection

Demographic and clinical outcome data from pa-tients who underwent ventriculoperitoneal shunt place-ment were prospectively collected at the Brigham and Women’s Hospital/Dana-Farber Cancer Institute and the Huntsman Cancer Institute, University of Utah with in-stitutional-review-board –approved protocols. All patients treated with ventriculoperitoneal shunts were at least 18 years of age, carried the diagnosis of NM, and had symp-tomatic hydrocephalus. Most patients underwent lumbar puncture prior to shunt placement to demonstrate high opening pressure and/or volume-dependent neurologi-cal symptoms. Identical demographic and outcome data were collected from a contemporaneous group of patients with NM and no hydrocephalus treated at the Brigham and Women’s Hospital/Dana-Farber Cancer Institute who underwent Ommaya reservoir placement without shunt placement. Patients were monitored carefully during the study period for signs and symptoms of elevated intra-cranial pressure, such as worsening headache, nausea, vomiting, blurry vision, mental status changes, or gait changes. If elevated intracranial pressure was confirmed via lumbar puncture and an Ommaya reservoir had al-ready been placed, a ventriculoperitoneal shunt would be placed on the contralateral side. Informed consent was obtained from all patients for all surgical procedures.

Design of the RO-VPS ConstructThe RO-VPS construct consists of a Codman Holter

reservoir (Codman Inc.) connected in series with an on-off valve (Integra LifeSciences), a programmable shunt valve (Codman Hakim valve, in-line design), and distal peritoneal tubing (Fig. 1). The connection between each component is achieved via straight connectors with silk ties. The construct measures approximately 11 cm in length, compared with 3 cm for the valve only. We chose to use both an on-off valve and a programmable valve, which could be dialed from 30 mm Hg to 200 mm Hg, because of concerns that frequent adjustment of the pro-grammable valve setting could be associated with in-creased risk of shunt failure. It is also important to have the on-off valve in between the Ommaya reservoir and the shunt valve because switching it off prevents the shunt valve from being exposed to external pressure when the reservoir is used for injection. During the operative pro-cedure, a ventricular catheter is placed intracranially, and is then attached to the Ommaya reservoir. All patients re-ceived an intravenous injection of antibiotic agent (Van-comycin 1 gm at Brigham and Women’s Hospital, Ancef 1 gm at the Huntsman Cancer Institute) prior to incision. After the operation, a CT scan of the head and skull radio-graphs (shunt series) were obtained in all cases to confirm appropriate ventricular catheter placement and the appro-priate shunt pressure setting (Fig. 1). A CSF flow study (with the on-off valve closed or with the programmable valve turned up to the highest setting) was performed in all cases prior to the first administration of intraventricu-lar chemotherapy to demonstrate appropriate CSF distri-bution of tracer and patency of CSF pathways.5,9,13,16

Intrathecal Chemotherapy ProtocolAs early as 48 hours after placement of an RO-VPS

system, patients were treated with intraventricular che-motherapy selected by the treating neurooncology team. Prior to an intrathecal injection, the treating oncologist manually closed the on-off valve, effectively stopping the CSF flow toward the peritoneal space. The Ommaya reservoir was accessed via a 25-gauge needle, and the opening pressure was measured prior to chemotherapy administration to confirm a pressure less than 200 mm of water. The on-off valve stayed closed for 2–6 hours after the injection while the patient was closely monitored in the clinic. The valve was promptly switched open if the patient developed symptoms suggestive of increased intracranial pressure such as worsened headache, nau-sea, vomiting, or confusion. It was observed that over time most patients could tolerate longer duration of valve closure; as a result, in many patients the valve was kept closed for up to 24 hours. Intrathecal chemotherapy was mostly carried out in an outpatient setting and rarely was admission required after a treatment.

Data AnalysisFollow-up was performed twice weekly to once

monthly depending on the patient’s chemotherapy regi-men and clinical course. All electronic and paper medical records were reviewed. Demographic characteristics and perioperative medical conditions were recorded, includ-ing age, sex, extent of systemic cancer, CSF chemistry,

N. Lin et al.


cell count, cytology, radiographic findings, concurrent che motherapy and radiation therapy, pre- and postop-erative Karnosfky Performance Scale (KPS) scores, and progression-free and overall survival. Perioperative com-plications, length of hospital stay, and postoperative re-lief of symptoms were also extracted from the medical records. Disease progression was defined as either clini-cal deterioration attributed by the treating team to NM, positive results of CSF cytological analysis after initial clearing, or radiographic worsening of CNS disease. The KPS score, disease-related symptoms, and treatment-re-lated side-effects were reported by the patient’s treating neurooncology team in the medical record. Independent-sample t-tests and log-rank tests were performed. All sta-tistical analyses were performed with Excel (Microsoft Corp.) and SPSS 15.0 (SPSS, Inc.).

ResultsPatient Cohorts

Between 2005 and 2009, 24 patients underwent sur-gical placement of RO-VPS constructs at our 2 institu-tions. Two patients were lost to follow-up. A comparison (“control”) group consisted of 26 consecutive patients who were diagnosed with neoplastic meningitis, had Ommaya reservoirs placed, and received intrathecal che-motherapy at the Brigham and Women’s Hospital/Dana-Farber Cancer Institute during the same 4-year interval. Two patients in the control group who initially underwent

placement of an Ommaya reservoir alone subsequently developed symptomatic hydrocephalus and underwent placement of a ventriculoperitoneal shunt on the contra-lateral side. Both of these patients received intrathecal chemotherapy following the second operation and were included in the RO-VPS group. Thus, data from 24 pa-tients were available for analyses from each group. The 2 groups were nearly identical with respect to demographic characteristics, systemic cancer spread, functional status (measured by preoperative KPS score), and treatment, apart from the type of ventricular access or shunting de-vice they received (Table 1). A total of 29 patients were treated with whole brain radiation therapy prior to shunt or Ommaya reservoir placement, and 14 patients also re-ceived stereotactic radiosurgery for focal intracranial le-sions. The distribution of patients treated with radiation therapy is similar between the 2 groups.

The median age of the patients at surgery was 57 years (range 23–75 years) in the RO-VPS group and 53.5 years (range 27–87 years) in the control group. Breast cancer was the most common primary cancer diagnosis in both groups. In the RO-VPS group, 7 patients had lep-tomeningeal spread of primary brain cancer (4 anaplastic astrocytomas, 2 glioblastomas, 1 medulloblastoma), as did 5 patients in the control group (2 anaplastic astrocy-tomas, 2 glioblastomas, 1 oligodendroglioma). At diagno-sis of NM, the majority of patients (76.9%) complained of headache, although ataxia, nausea/vomiting, visual changes, and cranial nerve palsies were also frequently

Fig. 1. Postoperative CT scan (A) and skull radiograph (B), and photograph (C) showing the RO-VPS construct. The intra-ventricular catheter is placed through a bur hole at the Kocher point and connected to the Ommaya reservoir (A). The RO-VPS construct is located in an extended subgaleal pocket above the pericranium (B). The 3-part apparatus (C) includes an Ommaya reservoir (1), an on-off valve (2), and a programmable shunt valve (3). All connections are tied with 2-0 silk sutures.



733

present. Approximately two-thirds of patients (66.7% in the RO-VPS group and 70.8% in control group) had sys-temic cancer spread or other brain parenchymal metas-tases prior to the diagnosis of NM. Most patients in both groups had had prior systemic chemotherapy or radia-tion therapy, and about one-third (45.8% in the RO-VPS group and 29.3% in the control group) had undergone a craniotomy for either resection or biopsy of intracranial metastases.

Cerebrospinal fluid was examined in all patients in both groups prior to surgery. Fourteen patients (53.8%) in the RO-VPS group had elevated intracranial pressure prior to shunt placement, documented either by lumbar puncture or during the surgical procedure itself, although not all opening pressures were quantitatively record-ed. All patients had evidence of NM by CSF analysis or imaging: CSF was positive for malignant cells in 21 patients (87.5%), and MR imaging of the brain or spine demonstrated leptomeningeal enhancement in 15 patients (62.5%). The CSF protein level was elevated (≥ 40 mg/

dl) in 17 patients (70.8%; mean 83 mg/dl, range 19–404 mg/dl), and the white blood cell count was elevated (≥ 4 cells/mm3) in 9 (37.5%; median 3 cells/mm3, range 0–490 cells/mm3). No patients in the control group were found to have intracranial hypertension prior to Ommaya reservoir placement.

Safety of RO-VPS PlacementThere was no perioperative mortality. Two postop-

erative complications were observed (8.3%). One patient had an obvious kink in the shunt tubing on the postopera-tive skull film and required immediate revision. The oth-er patient suffered a fall postoperatively and developed a subdural hematoma, which required surgical evacuation. The shunt had to be revised in another patient 3 months after the original operation due to valve failure. There was no shunt infection during the follow-up period. The median length of hospital stay was 2.5 days after RO-VPS placement. Four patients were hospitalized for more than 10 days perioperatively because of medical complications unrelated to their surgery or treatments required for their cancer.

During this study, no routine abdominal imaging was performed unless indicated by new onset of symptoms or for staging purposes. No clinical symptoms were ob-served that suggested the development of peritoneal car-cinomatosis or toxicity from intrathecal chemotherapy agents. In one patient with isolated leptomeningeal breast cancer, whole-body PET CTs were performed every 6 months after RO-VPS placement, and no new abdominal pathology was discovered during 17 months of follow-up.

Benefits of RO-VPS SurgeryPatients who were referred to the neurosurgery clinic

with a diagnosis of NM and with symptoms of hydro-cephalus usually underwent RO-VPS placement within 3 days. Patients exhibited steep functional decline (as mea-sured by KPS score) from the time of the initial diagnosis of cancer to the time when they developed symptoms of NM. This downward trend was reversed after RO-VPS placement (Fig. 2), and the beneficial effect was sus-tained over time. Most patients who received RO-VPS constructs had rapid relief of preoperative symptoms of headache, nausea, and gait disturbance (20 [83.3%] of 24). The mean KPS score at the first routine postoperative fol-low-up appointment (usually 1–2 weeks after the opera-tion) was significantly higher than that at the preoperative evaluation (67.6 vs 53.6, p < 0.05). This improvement was sustained for the entire 6-month follow-up period of this series in 10 (41.7%) of 24 patients.

Survival After RO-VPS SurgeryEighteen of the 24 patients in the RO-VPS group re-

ceived intrathecal chemotherapy. Other concurrent post-operative treatment consisted of systemic chemotherapy in 21 patients (87.5%) and radiation therapy in 16 (66.7%). In 11 of the 18 patients who received intrathecal chemo-therapy, the best cytological response to treatment was complete clearing of malignant cells on at least 2 sepa-rate, consecutive CSF samplings at least 1 week apart.

TABLE 1: Clinical and demographic characteristics of patients who underwent surgical placement of an RO-VPS construct or an Ommaya reservoir alone*

Characteristic RO-VPS Group Ommaya Group p Value

no. of pts 24 24 NAage at op (yrs) 0.95 median 57.0 53.5 range 23–75 27–87 sex 0.67 male 7 8 female 17 16primary cancer 0.74 breast 9 10 brain/glioma 7 5 lung 4 3 lymphoma 3 4 other 1 2systemic cancer spread† 16 (66.7) 17 (70.8) 0.67positive cytology 21 (87.5) 19 (79.2) 0.31preop KPS score 53.6 53.1 0.88preop Tx 0.46 chemo 20 (83.3) 20 (83.3) WBRT 15 (62.5) 14 (58.3) SRS 9 (37.5) 5 (20.8)prior craniotomy 11 (45.8) 7 (29.2) 0.07postop IT chemo 18 (75.0) 20 (83.3) 0.27postop systemic Tx 0.34 chemo 21 (87.5) 19 (79.2) RT 16 (66.7) 13 (54.2)

* Values represent numbers of patients (%) unless otherwise indicated. Abbreviations: chemo = chemotherapy; IT = intrathecal; NA = not appli-cable; RT = radiation therapy; SRS = stereotactic radiosurgery; WBRT = whole-brain radiation therapy.† Includes patients with other brain parenchymal metastases.

N. Lin et al.


The median duration of follow-up was 28 weeks. The median overall survival was 31 weeks (95% CI 21.7–40.2 weeks, range 10–88 weeks), and 11 patients (45.8%) were alive at end of follow-up. Twenty-two patients (91.7%) eventually showed disease progression during the follow-up period. The median progression-free survival was 14 weeks (95% CI 12.2–15.8 weeks, range 4–54 weeks).

In the control group, 20 patients (83.3%) received in-trathecal chemotherapy, 13 (54.2%) underwent postopera-tive radiation, and 19 (79.2%) had systemic chemothera-py. The median follow-up period was 14 weeks. Twenty-three patients (95.8%) showed disease progression during follow-up. The median overall survival was 19.5 weeks (95% CI 5.9–29.5 weeks, range 2.7–67.7 weeks) and pro-gression-free survival 8.4 weeks (95% CI 1.0–13.9 weeks, range 1.8–54.6 weeks). When analyzing the length of overall survival according to primary cancer diagnosis, breast cancer patients in the RO-VPS group had signifi-cantly better survival than those in the control group (35 vs 21 weeks, p = 0.04), whereas there was no statistically significant difference in patients with other primary di-agnoses (Table 2). The overall survival in the RO-VPS group altogether was also better than that in the control group, using the Kaplan-Meier method (p = 0.008, Fig. 3)

Intrathecal chemotherapy agents included Depocyt (cytarabine liposome injection), cytosine arabinoside, trastuzumab, and thiotepa. The choice of intrathecal che-motherapy agents was determined by the treating neu-rooncologist based on the patient’s primary cancer his-tory. No severe intrathecal chemotherapy–related side ef-fects were observed and all patients received oral steroid therapy during intrathecal chemotherapy. Two patients developed new-onset seizures after intrathecal chemo-therapy was started; they were found to have progression of disease on brain MR imaging.

DiscussionNeoplastic meningitis is a devastating complication

of cancer and is increasing in frequency.12 Elevated in-tracranial pressure and hydrocephalus are important con-tributors to clinical symptoms and poor prognosis and can be present in as many as half of all patients.9,22,25,26 These complications produce neurological deficits that significantly impair quality of life. Although ventricu-loperitoneal shunting is a well-established, easily per-formed, and effective treatment for hydrocephalus, few patients with NM undergo this procedure.22,25 Reasons for withholding this intervention include: 1) therapeutic ni-hilism, 2) concerns about the morbidity of the procedure and the frequent failure of the shunt apparatus in this pa-tient population, 3) concerns about producing peritoneal carcinomatosis, and 4) concerns that shunt placement will either render intraventricular chemotherapy ineffective or increase its toxicity by shunting the chemotherapy drugs directly into the peritoneal cavity.

The use of an on-off valve in conjunction with a ven-triculoperitoneal shunt system to administer intrathecal chemotherapeutic agents or antibiotic therapy has been reported in the literature.11,20 With the advancement of programmable valves, the on-off valve has been less frequently used, as adjusting a programmable valve to the maximum setting prior to intrathecal injection has become a more common practice.28 We chose to use an on-off valve because it was easily accessible to neuroon-cologists and because it reduced the need for frequent manipulation of the programmable shunt valve. In the current study, the placement of an RO-VPS construct was associated with a very low rate of morbidity (8.3%), no perioperative infection, and no perioperative death. This is especially notable because patients with NM are poten-tially high-risk candidates for ventriculoperitoneal shunt placement. Prior chemotherapy and radiation therapy can result in impaired wound healing; treatment- and disease-related immunosuppression can cause increased suscep-tibility to infection; and prior craniotomies and the pres-ence of malignant cells and high protein in the CSF can result in a higher failure rate for the ventriculoperitoneal shunt in these patients than in patients who do not have cancer. Our experience compares favorably with similar published reports, in which significant complications oc-curred in about 10% of patients.22,25

In addition to being associated with a low complica-

TABLE 2: Median overall survival according to primary cancer diagnosis

Primary CancerMedian Overall Survival (wks)

p ValueRO-VPS Group Ommaya Group

breast 35 21 0.04brain/glioma 31 10.5 0.11lung 25.5 15 0.23lymphoma 32 24 0.30other 12 25 0.27total 31 19.5 0.02

Fig. 2. Karnofsky Performance Scale scores before and after shunt placement. The KPS score drops precipitously from the initial diagnosis of primary cancer to just prior to the RO-VPS placement procedure. This trend is reversed after the operation and the effect is sustained for up to 6 months. The time stamps indicate diagnosis of primary cancer (Initial Dx), diagnosis of NM (NM Dx), prior to surgical placement of RO-VPS construct (Pre-op), 2 weeks after the operation (Post-op), 3 months after the operation, and 6 months after the operation.



735

tion rate, the RO-VPS constructs were effective at reliev-ing the symptoms of hydrocephalus (in 20 [83.3%] of 24 patients) and permitted intraventricular chemotherapy to be administered in all patients in whom this intervention was deemed appropriate. In those patients in whom intra-ventricular chemotherapy was delivered, we saw no un-usual or clinically concerning toxicity, and we observed evidence of treatment efficacy, with cytological clearing of the CSF in 11 (61.1%) of 18 patients. The observed median survival of 31 weeks in this study also compares favorably with published studies reporting a median over-all survival of 8–30 weeks.4,14,15,18 The survival advantage was especially notable in breast cancer patients who were treated with RO-VPS systems compared with those who had an Ommaya reservoir only (p = 0.04). This difference could most likely be explained by the effective CSF di-version provided by the ventriculoperitoneal shunt, given that patients in both groups received Depocyt intrathecal chemotherapy. Moreover, in a separate analysis within the RO-VPS group, we compared the length of overall survival in those who received intrathecal chemotherapy and those who did not. The patients who underwent both RO-VPS placement and intrathecal chemotherapy had longer median overall survival (data not shown), although the difference was not statistically significant. These re-sults suggest that both CSF diversion and intrathecal che-motherapy are important treatment modalities for NM patients with hydrocephalus, a subgroup of patients usu-ally associated with poor outcomes.6,17,19

There are several limitations to be considered in interpreting these results. The study was a retrospec-tive case-control study, treatment assignment was not randomized, and sample sizes were small. Meanwhile, important differences between groups were present de-spite demographic similarities; for example, the RO-VPS and Ommaya reservoir (no–ventriculoperitoneal shunt) groups also differed with respect to the presence of hy-drocephalus and increased intracranial pressure. Never-

theless, hydrocephalus and intracranial pressure elevation are usually considered to herald advanced leptomenin-geal disease and thus portend a poor prognosis. Another observation that we could not analyze was the duration of closure of the on-off valve (which depended on patient tolerability) after injection of intrathecal medications, but we felt clinically that longer duration of closure was observed after repeated intrathecal chemotherapy trials, probably due to improved CSF circulation and reabsorp-tion.

ConclusionsIn many patients who suffer from NM, hydrocepha-

lus and elevated intracranial pressure are responsible for rapidly declining neurological status and eventual poor outcome. Our experience suggests that CSF diversion surgery in conjunction with intrathecal chemotherapy may provide important benefits to an appreciable propor-tion of patients with NM. Prospective clinical studies are needed to validate this approach in patients with NM.

Disclosure

The authors report no conflict of interest concerning the mate-rials or methods used in this study or the findings specified in this paper.

Author contributions to the study and manuscript preparation include the following. Conception and design: Kesari, Lin, Dunn, Glantz, Jensen, Friedlander. Acquisition of data: Lin, Allison, John-son, Glantz, Jen sen. Analysis and interpretation of data: Kesari, Lin. Drafting the article: Kesari, Lin. Critically revising the article: Kesari, Lin, Dunn, Glantz, John son, Friedlander. Statistical analysis: Kesari, Lin. Administrative/technical/material support: Kesari, Lin. Study supervision: Kesari, Friedlander.

References

1. Berger MS, Baumeister B, Geyer JR, Milstein J, Kanev PM, LeRoux PD: The risks of metastases from shunting in chil-dren with primary central nervous system tumors. J Neuro-surg 74:872–877, 1991

2. Bernstein WB, Kemp JD, Kim GS, Johnson VV: Diagnosing leptomeningeal carcinomatosis with negative CSF cytology in advanced prostate cancer. J Clin Oncol 26:3281–3284, 2008

3. Bokstein F, Lossos A, Siegal T: Leptomeningeal metastases from solid tumors: a comparison of two prospective series treated with and without intra-cerebrospinal fluid chemother-apy. Cancer 82:1756–1763, 1998

4. Boogerd W, van den Bent MJ, Koehler PJ, Heimans JJ, van der Sande JJ, Aaronson NK, et al: The relevance of intraventricu-lar chemotherapy for leptomeningeal metastasis in breast can-cer: a randomised study. Eur J Cancer 40:2726–2733, 2004

5. Brem SS, Bierman PJ, Brem H, Butowski N, Chamberlain MC, Chiocca EA, et al: Central nervous system cancers. J Natl Compr Canc Netw 9:352–400, 2011

6. Bruna J, González L, Miró J, Velasco R, Gil M, Tortosa A: Leptomeningeal carcinomatosis: prognostic implications of clinical and cerebrospinal fluid features. Cancer 115:381–389, 2009

7. Bruno MK, Raizer J: Leptomeningeal metastases from solid tumors (meningeal carcinomatosis). Cancer Treat Res 125: 31–52, 2005

8. Chamberlain MC, Glantz M: Ventriculoperitoneal shunt in patients with leptomeningeal metastasis. Neurology 66:783, 2006

9. Chowdhary S, Chamberlain M: Leptomeningeal metastases:

Fig. 3. Kaplan-Meier curves for patients in the Ommaya reservoir group and the RO-VPS group. Patients in the RO-VPS group have sig-nificantly better overall survival (p = 0.008).

N. Lin et al.


current concepts and management guidelines. J Natl Compr Canc Netw 3:693–703, 2005

10. DeAngelis LM, Boutros D: Leptomeningeal metastasis. Can-cer Invest 23:145–154, 2005

11. DeAngelis LM, Posner JB (eds): Leptomeningeal metastases, in Neurologic Complications of Cancer, ed 2. New York: Oxford University Press, 2008, pp 240–274

12. Drappatz J, Batchelor TT: Leptomeningeal neoplasms. Curr Treat Options Neurol 9:283–293, 2007

13. Glantz MJ, Hall WA, Cole BF, Chozick BS, Shannon CM, Wahlberg L, et al: Diagnosis, management, and survival of patients with leptomeningeal cancer based on cerebrospinal fluid-flow status. Cancer 75:2919–2931, 1995

14. Glantz MJ, Jaeckle KA, Chamberlain MC, Phuphanich S, Recht L, Swinnen LJ, et al: A randomized controlled trial comparing intrathecal sustained-release cytarabine (Depo-Cyt) to intrathecal methotrexate in patients with neoplastic meningitis from solid tumors. Clin Cancer Res 5:3394–3402, 1999

15. Grossman SA, Finkelstein DM, Ruckdeschel JC, Trump DL, Moynihan T, Ettinger DS: Randomized prospective compari-son of intraventricular methotrexate and thiotepa in patients with previously untreated neoplastic meningitis. Eastern Co-operative Oncology Group. J Clin Oncol 11:561–569, 1993

16. Grossman SA, Trump DL, Chen DC, Thompson G, Camargo EE: Cerebrospinal fluid flow abnormalities in patients with neoplastic meningitis. An evaluation using 111indium-DTPA ventriculography. Am J Med 73:641–647, 1982

17. Harstad L, Hess KR, Groves MD: Prognostic factors and out-comes in patients with leptomeningeal melanomatosis. Neuro Oncol 10:1010–1018, 2008

18. Hitchins RN, Bell DR, Woods RL, Levi JA: A prospective randomized trial of single-agent versus combination chemo-therapy in meningeal carcinomatosis. J Clin Oncol 5:1655–1662, 1987

19. Kesari S, Batchelor TT: Leptomeningeal metastases. Neurol Clin 21:25–66, 2003

20. Mason WP: Leptomeningeal metastases, in Schiff D, Wen PY

(eds): Cancer Neurology in Clinical Practice. Totowa: Hu-mana Press, 2003, p 115

21. Melisko ME, Glantz M, Rugo HS: New challenges and op-portunities in the management of brain metastases in patients with ErbB2-positive metastatic breast cancer. Nat Clin Pract Oncol 6:25–33, 2009

22. Omuro AM, Lallana EC, Bilsky MH, DeAngelis LM: Ven-triculoperitoneal shunt in patients with leptomeningeal metas-tasis. Neurology 64:1625–1627, 2005

23. Pace A, Fabi A: Chemotherapy in neoplastic meningitis. Crit Rev Oncol Hematol 60:194–200, 2006

24. Pavlidis N: The diagnostic and therapeutic management of leptomeningeal carcinomatosis. Ann Oncol 15 (Suppl 4): iv285–iv291, 2004

25. Sandberg DI, Bilsky MH, Souweidane MM, Bzdil J, Gutin PH: Ommaya reservoirs for the treatment of leptomeningeal metastases. Neurosurgery 47:49–55, 2000

26. Wasserstrom WR, Glass JP, Posner JB: Diagnosis and treat-ment of leptomeningeal metastases from solid tumors: experi-ence with 90 patients. Cancer 49:759–772, 1982

27. Yoshida S, Morii K: Intrathecal chemotherapy for patients with meningeal carcinomatosis. Surg Neurol 63:52–55, 2005

28. Zada G, Chen TC: A novel method for administering intrathe-cal chemotherapy in patients with leptomeningeal metastases and shunted hydrocephalus: case report. Neurosurgery 67 (3 Suppl Operative):onsE306–onsE307, 2010

Manuscript submitted October 15, 2010.Accepted May 23, 2011.Current affiliation for Dr. Friedlander: Department of Neurosur-

gery, University of Pittsburgh Medical Center, Pittsburgh, Penn-syl vania.

Please include this information when citing this paper: pub-lished online July 1, 2011; DOI: 10.3171/2011.5.JNS101768.

Address correspondence to: Santosh Kesari, M.D., Ph.D., Univer-sity of California, San Diego, Moores Cancer Center, 3855 Health Sciences Drive, MC 0819, La Jolla, California 92093-0819. email: [email protected].

benefit of ventriculoperitoneal cerebrospinal fluid...

Documents