Preoperative Lumbar Drain Use during AcousticNeuroma Surgery and Effect on CSF Leak Incidence

Matthew G. Crowson, MD1; Calhoun D. Cunningham III, MD1; Helen Moses, MD1; Ali Zomorodi, MD2; David Kaylie, MD, MS1

Duke University Medical Center, 1Division of Otolaryngology-HNS, 2Division of Neurosurgery

INTRODUCTION DISCUSSIONRESULTS

Table 1. Descriptive statistics of patient demographics and tumor factors.

ABSTRACT

METHODS AND MATERIALS

CONCLUSIONS

REFERENCESCONTACT

Objective: To determine if preoperative lumbar drain (LD) use reduces the incidence of postoperativecerebrospinal fluid (CSF) leak in patients undergoing acoustic neuroma resection.

Study design: Retrospective Review.

Setting: Tertiary Care HospitalPatients: 282 patients presenting for primary acoustic neuroma resection between 2005-2014.

Interventions: Clinical record review of tumor characteristics, imaging, procedures, patient outcomes, CSFleak incidence, and postoperative complications.

Main outcome measures: CSF leak frequency, LD complications, analysis of patient demographics and diagnosis, tumor size, surgical approach, and length of stay.

Results: 282 patients had a mean tumor size of 19.1 mm +/- 10.2 mm. 29 (10.3%) patients developed a postoperative CSF leak. 220 patients (78.0%) received a preoperative LD, and 20 (9.1%) developed a CSF leak. 62 (22.0%) patients did not receive a preoperative LD, and 9 (14.5%) developed a CSF leak. No significantdifference in CSF leak frequency observed with use versus no use of a LD (p > 0.05). 15 (5.3%) patients with a LD placed had a complication related to the LD. No significant difference in CSF leak frequency was observed with patient age, Neurofibromatosis Type-2 diagnosis, tumor size or sidedness.

Conclusions: Postoperative CSF leaks are amongst the most common complications of acoustic neuroma microsurgery. No formal guidelines exist for the elective placement of a preoperative LD to lower the incidence of CSF leaks. Our reported CSF leak incidence with preoperative LD placement is not significantly lower than without LD use, and there is a significant complication rate associated with LD use.

313 patients underwent resection for an acoustic neuroma. 282 patients were included in the analysis, and 31 excluded for prior resection attempt, or if they had a pre-operative external ventricular drain (EVD) placement.

220 (78.0%) patients received a pre-operative lumbar drain (LD) placed prior to acoustic neuroma resection (Table 3). Fifteen patients (5.32%) had a complication related to their LD. Complications included LD falling out early (5), LD leak (3), over-draining of CSF (2), LD non-functioning (2), retrained LD tip requiring laminectomy (1), LD clotted off (1), and positive CSF surveillance cultures (1). The length of stay was longer for patients who had a LD complication (6.31 days) vs no complication (5.83 days), however this was not statistically significant (p = 0.50).

29 patients had a CSF leak (10.3%). 11 patients (3.9% of all patients; 38.0% of patients with leak at any time) had a CSF leak during their operative admission. 19 patients (6.7% of all patients; 65.5% of patients with leak at any time) had a CSF leak discovered after discharge at a future encounter. Of the 62 patients who did not have a preoperative LD placed, 9 (14.5%) developed a CSF Leak. Of 220 patients who had a preoperative LD, 20 (9.1%) developed a CSF leak. There was no statistical difference between the two groups (p = 0.23).

CSF leak after acoustic neuroma resection is a common complication that carries the potential for significant morbidity. The objective of this study was to determine if the preoperative placement of a lumbar drain reduced the incidence of postoperative CSF leaks after primary acoustic neuroma resection. We report that CSF leak incidence with preoperative LD placement is not significantly lower than without LD use. Our patient population was unique as most patients had a LD placed prior to surgery compared to those who have not.

Reported CSF leak incidences with the use of routine preoperative lumbar drain placement are limited in current literature. At time of publication, no randomized control trial has directly measured the effect of preoperative lumbar drain placement on postoperative CSF leak incidence after acoustic neuroma microsurgery.

Placement of a LD is a relatively simple and commonly performed procedure. As with any procedure, the placement and maintenance of a LD is associated with a clinically significant complication rate. We found that 5.3% of our patients with a pre-operative LD had a complication related to their LD.

Patients – This retrospective study was completed with the approval of the IRB at Duke University Medical Center. The study population consisted of 282 patients who underwent acoustic neuroma resection at our institution between 1990 and 2014 at two Duke-affiliated hospitals.

Patient information collected included gender, age at presentation, race and ethnicity, acoustic neuroma sidedness (left or right), tumor size (dimensions reported in 3 axes), surgical resection procedure type and approach, neurofibromatosis diagnosis (yes or no), pre-operative stereotactic radiation (yes or no), pre-operative lumbar drain use (yes or no), and duration of lumbar drain placement (days).

Patient outcome variables collected included hospital stay length (days), post-operative complication type and frequency, lumbar drain complication type and frequency, CSF leak rate frequency and timing (same admission, or delayed), and recurrence on follow-up MRI imaging (yes or no).

Statistical Analysis – All analyses were completed using the JMP Pro 11 software suite (Cary, North Carolina, USA). Two-by-two contingency tables were created, and Fisher’s exact tests were performed on all variables . P-values were reported with statistical significance fixed at p = 0.05. Patients with insufficient demographic or outcomes data were excluded from statistical analysis. 

• We have found that the pre-operative use of a LD for acoustic neuroma resection does not significantly decrease the post-operative CSF leak rate.

• The use of a LD carries a considerable risk of complication related to its use that could cause significant morbidity and increase hospital length of stay.

• While a CSF leak is a common and potentially serious complication of acoustic neuroma resection, we believe that the routine use of a pre-operative LD with the intent to prevent a CSF leak should be avoided unless specific case-by-case indications are present.

1. Mahboubi, H., et al., Complications of surgery for sporadic vestibular schwannoma. Otolaryngol Head Neck Surg, 2014. 150(2): p. 275-81.

2. Mangus, B.D., et al., Management of cerebrospinal fluid leaks after vestibular schwannoma surgery. Otol Neurotol, 2011. 32(9): p. 1525-9.

3. Laing, R.J., et al., A study of perioperative lumbar cerebrospinal fluid pressure in patients undergoing acoustic neuroma surgery. Skull Base Surg, 2000. 10(4): p. 179-85.

4. Nonaka, Y., et al., Contemporary surgical management of vestibular schwannomas: analysis of complications and lessons learned over the past decade. Neurosurgery, 2013. 72(2 Suppl Operative): p. 103-15

5. Bien, A.G., et al., Utilization of preoperative cerebrospinal fluid drain in skull base surgery. Skull Base, 2007. 17(2): p. 133-9

Matthew G. Crowson, MDResident PhysicianDuke University Medical CenterDivision of Otolaryngology-HNSDurham, NCTel.: +1 603 306 1182E-mail address: matthew.crowson@dm.duke.edu

Post-operative cerebrospinal fluid (CSF) leaks are amongst the most common major complications of acoustic neuroma (AN) microsurgery. CSF leaks after AN microsurgery typically present with otorrhea, headache, dizziness, and can lead to grave sequellae such as meningitis, and brain abscess formation.[1] Meticulous surgical technique and tight wound closure are cornerstones for CSF leak prevention, but no formal guidelines exist for the routine use of preoperative lumbar drain placement to lower the incidence of CSF leaks following AN microsurgery.

Consistent with basic fluid dynamics, CSF will flow from the relatively high intracranial pressure region to the relatively low-pressure regions of the eustachian tube, middle ear space, or dehiscence in the periosteal or skin wounds.[2] To characterize the intracranial CSF pressures following AN microsurgery, Laing et al. measured changes in CSF pressure and cerebrovascular hemodynamics following AN microsurgery in humans.[3] All patients showed a statistically significant rise in CSF pressure from normal levels with completely reversal within 48 hours of surgery.

CSF leak incidences with the use of routine preoperative lumbar drain placement are limited in current literature. Of the few studies published, CSF leak incidence has been reported to be within 7.6-12%. [4, 5] The purpose of this study is to retrospectively compare clinical outcomes with the preoperative use of lumbar drain placement during acoustic neuroma microsurgery. Patient Variables No. of patients (% of total)

Mean Age 52.3, range 14-87

Gender 161 female (57.1), 121 male (42.9)Neurofibromatosis Type 2 Diagnosis 12 (4.3)Mean Tumor Size (greatest dimension) 19.1 mm +/- 10.2 mmPre-operative stereotactic radiation 13 (4.6)Post-operative stereotactic radiation 12 (4.3)

Duke University Hospital 118 (41.8)

Duke Raleigh Hospital 164 (58.2)

Table 2. Tumor, Surgical factors and CSF Leak rate.

  No. of patients (% of row)Patient and Tumor

VariablesCSF Leak No CSF Leak p-value

No Lumbar Drain 9 (14.5) 53 (85.5) } 0.24Lumbar Drain 20 (9.1) 200 (90.9)

Translab. Approach 15 (12.4) 106 (87.6)  

Retrosigmoid Approach 12 (9.2) 118 (90.8)Middle Fossa Approach 2 (6.5) 29 (93.5)  

Patient Age, Mean 50.4 years +/- 2.4 52.6 years +/- 0.8 0.4

Length of Stay 9.8 days +/- 0.78 5.7 days +/- 0.16 0.0001

Tumor Size, Mean 18.3 mm +/- 1.92 19.2 mm +/- 0.64 0.67

Tumor Side, Left 15 (11.8) 112 (88.2) } 0.56Tumor Side, Right 14 (9.0) 141 (91.0)

NF-2, Yes 1 (8.33) 11 (91.7) } 1.00NF-2, No 28 (10.4) 242 (89.6)Pre-Operative Radiation, Yes

1 (7.69) 12 (92.3)

} 1.00Pre-Operative Radiation, No

28 (10.4) 241 (89.6)

Table 3. Duration of lumbar drain placement by approach

  No. of patients (% of row)

Mean, Mode(Std. Dev)

Approach LD Placed No LD LD Placement Duration (Days)

Retrosigmoid 82 (63.1) 48 (37.0) 2.22, 1 (1.44)

Translab. 110 (91.1) 11 (9.1) 2.86, 2 (1.36)

Middle Fossa 28 (90.3) 3 (9.7) 2.00, 2 (0.75)

NF-2: Neurofibromatosis Type 2

Translab: Translabyrinthine