Management of recurrent cerebrospinal fluid leak, current practices and open challenges. A systematic literature review

Cerebrospinal fluid leak (CSFL) is a rare pathological condition which requires prompt and adequate management due to the high rate of associated major complications. At present, CSFL can be successfully treated through endoscopic endonasal approaches with success rates higher than 90%. Despite this, CSFL recurrence may occur, and its management is critical and still represents a matter of debate. A systematic review of the literature on the management of recurrent CSFL was conducted. It included a cohort of 1,083 cases of CSFL treated with surgical or conservative approaches; 112 cases of recurrence were reported for an overall recurrence rate of 10.3%. Most of the leaks were localised in the anterior skull base (68.3%) and identification of recurrent CSFL may be troublesome. Therefore, the entire skull base must be meticulously examined and long-term follow-up is mandatory. The risk for CSFL recurrence is higher in patients affected by idiopathic intracranial hypertension (IIH), but its role in CSFL recurrence is yet to be understood. Recurrent CSFL must be repaired as soon as possible in order to reduce the risk of intracranial complications. The use of early post-operative CSF diversion by lumbar drain (LD) is currently a matter of debate even in case of CSFL recurrences.

Venous sinus stenting after repair of skull base spontaneous cerebrospinal fluid leaks: A single-center retrospective cohort series with case-control analysis

BACKGROUND

Management of idiopathic intracranial hypertension (IIH) is recommended after surgical repair of spontaneous cerebrospinal fluid leaks (sCSF-leaks) of the skull base for prevention of recurrence.

PURPOSE

To assess the feasibility of venous sinus stenting, a treatment commonly used for the treatment of IIH associated with intracranial venous sinus stenosis (VSS), after sCSF-leaks closure.

MATERIALS AND METHODS

A single-center cohort series of consecutive patients who underwent sCSF-leak closure was retrospectively analyzed. Stenting was considered either for leak recurrence or in prophylactic manner after repair in patients with VSS as confirmed by cerebral venous imaging. Leak recurrence, need for new repair or adjunctive treatment of IIH, meningitis, and stenting complications were determined at the last follow-up. Cases who had prophylactic stenting were compared to historical controls before stenting option.

RESULTS

Twenty-two patients had intracranial venous stenting after sCSF-leak closure. Their median age was 58 years (Q1=45; Q3=68), BMI=31 kg.m^-2 (Q1=27; Q3=36), and female rate=85%. The overall rate of successful repair after stenting was 95% (95% CI = 87-100%) at a median follow-up of 2.4 years (Q1=1.2; Q3=3.3). Adjunctive treatment for IIH was needed in 4 patients (4/22, 18%) including 2 patients without leak recurrence. No meningitis, permanent morbidity or mortality was observed after stenting. Compared to 18 controls, cases had significantly less recurrence (P = 0.03), and a trend for less adjunctive treatment for IIH (P = 0.06).

CONCLUSIONS

Our study suggests that stenting might be a valid option for prevention of sCSF-leak recurrences after repair in patients with intracranial venous sinus stenosis.

The relationship between spontaneous cerebrospinal fluid leak and idiopathic intracranial hypertension

Surgical treatment of spontaneous cerebrospinal fluid (CSF) leak is now performed by ENT surgeons, endonasal endoscopy being preferred to craniotomy as less invasive. However, it is often the symptom of underlying idiopathic intracranial hypertension, which lies outside the traditional sphere of ENT competence. Surgery is a necessary step, but should not obscure the need to treat the underlying pathology. This treatment is complex, and requires multidisciplinary team-work between otorhinolaryngologist, ophthalmologist, neurologist, neurosurgeon, radiologist, dietician, endocrinologist and psychotherapist. The present update details this multidisciplinary management to which the ENT surgeons must be attentive before and after spontaneous CSF leak repair.

Multiple spontaneous skull base cerebrospinal fluid leaks: some insights from an international retrospective collaborative study

PURPOSE

Spontaneous skull base cerebrospinal fluid leaks occurring without any apparent cause are rare. But those patients that present such leaks simultaneously, or successively, in multiple locations are even rarer. Given the rarity of this condition, we collected data from other groups in a multicentre study to reach an adequate number of patients and draw some preliminary considerations.

METHODS

We carried out a multicentre retrospective study on a cohort of patients treated at third level hospitals in Italy, Spain, United Kingdom and Greece for multiple spontaneous-CSF leaks and we compared them with a control group of patients treated for recurrent spontaneous-CSF leaks. Data regarding clinical aspects, radiological findings, surgical techniques and outcome were collected and preliminary considerations on the results were discussed.

RESULTS

A total of 25 patients presented multiple simultaneous spontaneous CSF leaks while 18 patients fit with the criteria of recurrent spontaneous CSF leaks. Data analysis was conducted separately.

CONCLUSIONS

Our understanding of the pathogenesis of this condition is currently very limited. A causative role of IIH may be present but the differences that emerged from the comparison with patients with recurrent fistulas seem to promote the possible role of other cofactors. A longer follow-up period is needed, and, in our opinion, prospective and multicentre studies are the only solution to seriously deal with such a complex topic.

Resumption of Positive-Pressure Ventilation Devices for Obstructive Sleep Apnea following Transsphenoidal Surgery: An Institutional Experience of a Surgical Cohort

Objectives  Transsphenoidal surgery creates a skull base defect that may cause postoperative cerebrospinal fluid (CSF) leakage or pneumocephalus. This study reviewed the institutional experience of a pituitary center in managing patients who use positive-pressure ventilation (PPV) devices for obstructive sleep apnea (OSA) after transsphenoidal surgery, which risks disturbing the skull base repair. Design  Retrospective review. Setting  Pituitary referral center in a major metropolitan medical center. Methods  PPV was resumed at the discretion of the treatment team based on intraoperative findings and OSA severity. Perioperative complications related to resuming and withholding PPV were recorded. Participants  Transsphenoidal surgery patients with OSA using PPV devices. Main Outcome Measures  Intracranial complications before and after resuming PPV. Results  A total of 42 patients met the study criteria. Intraoperative CSF leakage was encountered and repaired in 20 (48%) patients. Overall, 38 patients resumed PPV (median: 3.5 weeks postsurgery; range: 0.14-52 weeks) and 4 patients did not resume PPV. Postoperatively, no patient experienced CSF leakage or pneumocephalus before or after resuming PPV. Four (10%) patients required temporary nocturnal supplemental oxygen at home, one patient was reintubated after a myocardial infarction, and one patient had a prolonged hospital stay due to chronic obstructive pulmonary disease exacerbation. Conclusions  Resuming PPV use after transsphenoidal surgery did not result in intracranial complications. However, delay in resuming PPV resulted in four patients requiring oxygen at home. We propose a preliminary PPV device management algorithm based on the size of the intraoperative CSF leak to facilitate future studies.

Association of Intracranial Hypertension With Calvarial and Skull Base Thinning

OBJECTIVE

Determine if patients with increased opening pressure (OP) on lumbar puncture (LP) have thinner calvaria and skull bases.

STUDY DESIGN

Retrospective cohort study.

SETTING

Tertiary referral center.

PATIENTS

Patients (≥18 yr of age) who had a recorded OP on LP and high-resolution computed tomography imaging of the head. Patient age, sex, body mass index were calculated. Intracranial hypertension (IH) was defined with an OP≥25 cm-H2O and low intracranial pressure with an OP<15 cm-H2O.

INTERVENTION

Measurement of calvarial, zygoma, and skull base thickness when blinded to OP with three-dimensional slicer and radiologic calipers.

MAIN OUTCOME MEASURES

Association of calvarial, skull base, and zygoma thickness with OP and age.

RESULTS

Fifty-eight patients were included with a mean (SD) age of 53.1 (16.2) years and average (SD) body mass index of 30.1 (9.1) kg/m. Patients with IH had thinner mean (SD) calvaria (3.01 [0.81] versus 2.70 [0.58] mm; p = 0.036) and skull bases (5.17 [1.22] versus 4.60 [1.42] mm; p = 0.043) when compared with patients without IH. The mean (SD) extracranial zygoma thickness was similar between the two groups (5.09 [0.76] versus 5.00 [0.73] mm; p = 0.56). General linear model regression demonstrated advancing age was associated with increasing calvarial thickness in patients without IH and calvarial thinning in patients with IH (p = 0.038).

CONCLUSION

IH is independently associated with intracranial bone (calvaria and skull base) thinning and not extracranial (zygoma) thinning. Skull thinning occurs with IH and advancing age. These findings support a possible role of increased ICP in the pathophysiologic development of spontaneous cerebrospinal fluid leaks.

Layered sellar reconstruction with avascular free grafts: Acceptable alternative to the nasoseptal flap for repair of low-volume intraoperative cerebrospinal fluid leak

BACKGROUND

Although the nasoseptal flap has become the method of choice for reconstruction of intraoperative cerebrospinal fluid (CSF) leak in endoscopic minimally invasive surgery of the skull base, layered avascular graft techniques, including allografts and middle turbinate mucosal autografts, may provide comparable reconstructive success with decreased nasal morbidity.

OBJECTIVE

To describe a method of reconstruction of intraoperative CSF leak in endoscopic surgery of the sella turcica and analyze its postoperative success rate and associated comorbidities.

METHODS

A retrospective review of expanded endonasal sellar tumor resections from 2008-2014 was performed, and cases of layered intraoperative skull base reconstruction with avascular free grafts were identified. Demographic factors and comorbidities that predisposed to reconstruction failure (obstructive sleep apnea, obesity) were determined. Reconstruction-related nasal complications were also identified. Postoperative CSF leak rate was determined, and statistical analysis was performed to identify predictive factors for reconstructive failure.

RESULTS

Seventy-three cases were identified. Layered closure with avascular free grafts was performed. There were five cases of postoperative CSF leak (6.85%). The mean follow-up was 19 months (range, 1-76 months). Intraoperative high-flow CSF leak was a significant predictor of reconstruction failure on univariate (odds ratio 22 [95% confidence interval, 2.26-214]; p = 0.008) and multivariate analysis (odds ratio 33.6 [95% confidence interval, 2.30-492]; p = 0.010). There were no significant differences in postoperative leak rates among bony overlay graft types. There were five patients (7.9%) who experienced persistent crusting after surgery. There were no significant differences in crusting rates between allografts and mucosal grafts. There were no postoperative mucoceles.

CONCLUSION

In cases of low-volume intraoperative CSF leak, layered skull base repair with avascular free grafts was an acceptable alternative to the nasoseptal flap, which may reduce prolonged sinonasal healing and donor-site morbidities.

Intervention for elevated intracranial pressure improves success rate after repair of spontaneous cerebrospinal fluid leaks

OBJECTIVES/HYPOTHESIS

Spontaneous cerebrospinal fluid (CSF) leaks are associated with increased intracranial pressure (ICP) and considered a manifestation of idiopathic intracranial hypertension. Although postoperative acetazolamide and placement of CSF shunt systems are considered valuable interventions for elevated ICP, the impact on recurrence rate remains unclear. The objective of this study was to systematically review evidence from reported literature to evaluate whether postoperative ICP management reduces recurrence rates after primary endoscopic repair.

STUDY DESIGN

Prospective case series and systematic review.

METHODS

Demographics, defect location, success rates, and ICP management in spontaneous CSF leak patients were prospectively collected over 8 years. A search was also conducted in PubMed to identify studies reporting cases of spontaneous CSF rhinorrhea.

RESULTS

Fifty-six articles with nonduplicated data were identified and combined with a prospective series of 108 patients for a total of 679 patients treated for spontaneous CSF rhinorrhea. Average age was 50.4 years with 77% female. Average body mass index was 35.8 kg/m² . Defects were most commonly located in the sphenoid sinus (n = 334) followed by the ethmoid (n = 318) and the frontal sinus (n = 46). Successful primary repair was 92.82% in patient cohorts where ICP evaluation and intervention with acetazolamide or CSF shunt systems was performed, but was significantly decreased to 81.87% in series with no active management of elevated ICP (P < .001).

CONCLUSIONS

Evaluation and intervention for elevated ICP in spontaneous CSF leaks is associated with significantly improved success rates following primary endoscopic repair.

LEVEL OF EVIDENCE

4. Laryngoscope, 127:2011-2016, 2017.

Outcomes of Endoscopic Repair of Cerebrospinal Fluid Rhinorrhea without Lumbar Drains

Objective

Lumbar drains (LD) are commonly used during endoscopic repair of cerebrospinal fluid (CSF) rhinorrhea, either to facilitate graft healing or to monitor CSF fluid dynamics. However, the indications and necessity of LD placement remains controversial. The current study sought to evaluate endoscopic CSF leak repair outcomes in the setting of limited LD use.

Methods

Patients who underwent endoscopic repair of CSF rhinorrhea between 2004 and 2014 were identified by a review of medical records. Demographic and clinical data were extracted and compared between patients who had surgery with and patients who had surgery without a perioperative LD. A univariate analysis was performed to identify factors predictive of recurrence.

Results

A total of 107 patients (116 surgical procedures) were identified, with a mean follow-up of 15.8 months. Eighty-eight of 107 patients (82.2%) had surgery without an LD. The mean hospital stay was 4.48 days in the LD group versus 1.03 days in the non-LD group (p ≤ 0.00001). There was no difference in recurrence rate between the LD and non-LD groups. Predictors of recurrence included repair technique (p = 0.04) and size of defect (p = 0.005). Body mass index, leak site (ethmoid, sphenoid, frontal), and etiology (spontaneous, iatrogenic, traumatic) were not predictive of leak recurrence.

Conclusion

Use of LDs in endoscopic CSF leak repair was not associated with reduced recurrence rates, regardless of leak etiology, and resulted in a significant increase in hospital length of stay. Although the use of perioperative LDs to monitor CSF dynamics may have some therapeutic and diagnostic advantages, it may not be associated with clinically significant improvements in patient outcomes or recurrence rates.

Randomized clinical trial of acetazolamide administration and/or prone positioning in mitigating wound complications following untethering surgeries

OBJECTIVE No evidence-based guideline has been approved for the postoperative management of pediatric patients with tethered cord syndrome (TCS). The purpose of this randomized clinical trial was to evaluate the effectiveness of prone positioning and acetazolamide administration on complication rates following spinal cord untethering surgeries. METHODS From October 2012 to February 2015, patients with a primary diagnosis of TCS who were admitted to the Children's Medical Center Hospital of Iran were randomly allocated to 1 of 4 intervention modality groups postoperatively: 1) Group A, acetazolamide administration for 10 days; 2) Group B, prone positioning for 10 days; 3) Group C, acetazolamide administration and prone positioning for 10 days; and 4) Group D, no intervention. CSF leakage, CSF collection, wound dehiscence, operative site infection, and secondary surgical wound repair were considered failure. RESULTS A total of 161 patients were enrolled in this study (Group A, n = 39 [24.2%]; Group B, n = 41 [25.5%]; Group C, n = 39 [24.2%]; and Group D, n = 42 [26.1%]). The overall failure rate was 12.42% (20 patients). Complication rates through pooled analyses were as follows: CSF leakage (n = 9, 5.6%), CSF collection (n = 12, 7.5%), wound dehiscence (n = 2, 1.2%), and infection of operation site (n = 3, 1.9%). Two patients (1.2%) required surgical secondary wound repair due to complications. CSF leakage and collection rates were significantly lower in patients who underwent prone positioning (p = 0.042 and 0.036, respectively). The administration of acetazolamide, either isolated or in combination with prone positioning, not only could not significantly lower the complication rates, but also added the burden of side effects. CONCLUSIONS The current study demonstrates the possible role of prone positioning in mitigating the complication rates subsequent to untethering surgeries. Clinical trial registration no.: NCT01867268 ( clinicaltrials.gov ).