- Email: [email protected]
Risk factors for cerebrospinal fluid leak in pediatric patients undergoing endoscopic endonasal skull base surgery
Accepted Manuscript Risk Factors for Cerebrospinal Fluid Leak in Pediatric Patients Undergoing Endoscopic Endonasal Skull Base Surgery Amanda L. Stapleton, MD, Elizabeth C. Tyler-Kabara, MD, PhD, Paul A. Gardner, MD, Carl H. Snyderman, MD, MBA, Eric W. Wang, MD PII:
S0165-5876(16)30455-4
DOI:
10.1016/j.ijporl.2016.12.019
Reference:
PEDOT 8354
To appear in:
International Journal of Pediatric Otorhinolaryngology
Received Date: 7 October 2016 Revised Date:
14 December 2016
Accepted Date: 15 December 2016
Please cite this article as: A.L Stapleton, E.C Tyler-Kabara, P.A Gardner, C.H Snyderman, E.W Wang, Risk Factors for Cerebrospinal Fluid Leak in Pediatric Patients Undergoing Endoscopic Endonasal Skull Base Surgery, International Journal of Pediatric Otorhinolaryngology (2017), doi: 10.1016/ j.ijporl.2016.12.019. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Risk Factors for Cerebrospinal Fluid Leak in Pediatric Patients Undergoing Endoscopic Endonasal Skull Base Surgery
RI PT
Stapleton, Amanda L, MD1; Tyler-Kabara, Elizabeth C, MD, PhD2; Gardner, Paul A, MD2; Snyderman, Carl H, MD, MBA1, 2, Wang Eric W, MD1
Department of Otolaryngology Head and Neck Surgery 1; Department of Neurological Surgery2,
SC
University of Pittsburgh Medical Center; Children’s Hospital of Pittsburgh of UPMC.
Center
M AN U
Research Performed at Children’s Hospital of Pittsburgh of University of Pittsburgh Medical
Short Running Title: CSF leak risk factors in pediatric skull base surgery Financial Support/Funding: None
TE D
Conflict of Interest Statement: Carl Snyderman: consultant for SPIWay LLC Corresponding Author Address: Amanda Stapleton, MD
EP
Children’s Hospital of Pittsburgh of UPMC 7th Floor Faculty Pavilion
AC C
4401 Penn Ave
Pittsburgh, PA 15224
Phone: 412-692-5466
Email: [email protected] Research Presented as an Oral Podium Presentation at the 25th Annual North American Skull Base Society Meeting. February 20-22, 2015 Tampa, Florida, USA
ACCEPTED MANUSCRIPT
Key Words: pediatric, skull base, cerebrospinal fluid leak, endoscopic surgery
AC C
EP
TE D
M AN U
SC
RI PT
Level of Evidence: 4
ACCEPTED MANUSCRIPT
Abstract Objectives: To determine the risk factors associated with cerebrospinal fluid (CSF) leak
RI PT
following endoscopic endonasal surgery (EES) for pediatric skull base lesions. Methods: Retrospective chart review of pediatric patients (ages 1 month to 18 years) treated for skull base lesions with EES from 1999-2014. Five pathologies were reviewed:
SC
craniopharyngioma, clival chordoma, pituitary adenoma, pituitary carcinoma, and Rathke’s cleft cyst. Fisher’s exact tests were used to evaluate the different factors to determine which
M AN U
had a statistically higher risk of leading to a post-operative CSF leak.
Results: 55 pediatric patients were identified who underwent 70 EES’s for tumor resection. Of the 70 surgeries, 47 surgeries had intraoperative CSF leaks that were repaired at the time of surgery. 11 of 47 (23%) surgeries had post-operative CSF leaks that required secondary
TE D
operative repair. Clival chordomas had the highest CSF leak rate at 36%. There was no statistical difference in leak rate based on the type of reconstruction, although 28% of cases that used a vascularized flap had a post-operative leak, whereas only 9% of those cases not
EP
using a vascularized flap had a leak. Post-operative hydrocephalus and perioperative use of a
AC C
lumbar drain were not significant risk factors.
Conclusions:
Pediatric patients with an intra-operative CSF leak during EES of the skull base have a
high rate of post-operative CSF leaks. Clival chordomas appear to be a particularly high-risk group. The use of vascularized flaps and perioperative lumbar drains did not statistically decrease the rate of post-operative CSF leak.
ACCEPTED MANUSCRIPT
Introduction: Pediatric patients who undergo endoscopic endonasal surgery (EES) for skull base
RI PT
tumors have unique pre-operative, intra-operative, and post-operative challenges in comparison to adult patients with similar lesions. As the indications for endonasal resection of pediatric skull base tumors widens, so do the reconstructive challenges. A post-operative
SC
cerebrospinal fluid (CSF) leak can place the patient at risk for significant complications including pneumocephalus, mental status changes, and meningitis. Our 15-year history of treating
M AN U
pediatric skull base tumors using the endoscopic endonasal approach (EEA) has given us a unique perspective on the surgical and post-operative challenges that these patients face. Tatreau et al1 in 2010 evaluated the anatomic considerations in pediatric patients undergoing EES. They found three areas of limitation in the pediatric patient as compared to
TE D
adults. The nasal pyriform aperture, sphenoid sinus pneumatization and sphenoid bone thickness, and intercarotid distance within the sphenoid sinus were all areas where anatomic limitations could influence surgical outcomes. All of these factors can effect surgical outcomes
EP
and the reconstructive options available to the pediatric skull base surgeon. The adult literature on the use of reconstructive materials including bio-synthetic
AC C
products and local vascularized flaps to repair intra-operative CSF leaks has been a recent area of intense review. A systematic review of the adult literature evaluated 22 studies looking to create an evidence-based algorithm for skull base reconstruction techniques. However, due to the wide variety of reconstructive techniques used, inconsistent reporting of high or low flow intra-operative CSF leaks, and lack of uniform post-operative outcomes reporting, a true algorithm could not be created. They focused on two factors: location of skull base defect and
ACCEPTED MANUSCRIPT
degree of intraoperative CSF flow. They concluded that repair of high-flow intraoperative CSF leaks with pedicled vascularized flaps was superior to non-vascularized flaps and that location
RI PT
of the defect did not matter, with the exception of the clivus2. There are case series in the literature addressing the feasibility of nasoseptal flap repair of skull base defects after undergoing EES in the pediatric population 3-5 . However, the
M AN U
leaks in pediatric patients undergoing EES.
SC
objective of this study was to determine the risk factors associated with post-operative CSF
Methods:
This is an IRB approved retrospective chart review from January 1999 to September 2014 at Children’s Hospital of Pittsburgh of the University of Pittsburgh Medical Center. Inclusion
TE D
criteria were pediatric patients, ages 1 month to 18 years old, who had EES for surgical resection of five high-risk tumor pathologies Craniopharyngioma, Pituitary Tumors (Adenoma, Carcinoma, Rathke’s cleft cyst), and Clival Chordoma. Only patients who had intraoperative
EP
CSF leaks were included in the study. These 5 tumor types were included due to their relatively high prevalence and/or a high rate of intraoperative CSF leak.
AC C
Indications for surgery included endocrine abnormalities, cranial nerve palsies, vision loss, orbital complications, persistent headaches, hydrocephalus, or other mental status changes. Image-guided navigation was used for all cases. All tumors were removed by the skull base team, which consists of a neurosurgeon and otolaryngologist operating concurrently. Gross total resection was the primary goal in the surgical resection of craniopharyngiomas, pituitary adenomas, pituitary carcinoma, and clival chordomas. Rathke’s cleft cysts were opened and
ACCEPTED MANUSCRIPT
drained into the sphenoid sinus. All clival chordomas were referred for proton beam radiation therapy following surgical resection. The study parameters included: pathologic diagnosis, age, gender, weight, height,
RI PT
growth percentile, surgical corridor approach (transsellar, transclival, transpterygoid,
transplanum, transtuberculum), reconstruction technique, intra-operative CSF leak, postoperative complications, tumor recurrence, post-operative CSF leak, hydrocephalus, and
SC
lumbar drain use. Fisher’s exact test was used to determine which factors had a statistically
M AN U
significant risk (p<0.05) of leading to a post-operative CSF leak.
Results:
170 pediatric patients who underwent EES from January 1999 to September 2014 were
TE D
identified. Of those, 55 patients met our inclusion criteria.
The pathology included 20 craniopharyngiomas, 14 pituitary adenomas, 11 clival chordomas, 9 Rathke’s Cleft cysts, and 1 pituitary carcinoma. The overall average age for the
EP
cohort was 13.7 years with craniopharyngiomas being the youngest group averaging 8.43 years old (Table 1). Of the 14 pituitary adenomas, 5 presented with Cushing’s Disease, 3
AC C
Prolactinomas, 2 Pituitary Apoplexy, 2 with vision loss, 1 Acromegaly, and 1 incidental imaging finding.
Patient size was recorded based upon height, weight, BMI, and growth percentile. The
size of the patient was evaluated to determine if this affected the incidence of CSF leak. Pituitary adenomas were our largest patients averaging the 92nd percentile for weight and a BMI of 42. The smallest patients in our group were the craniopharyngiomas, averaging 55th
ACCEPTED MANUSCRIPT
percentile for weight and a BMI of 19. There was no statistical difference in CSF leak rates for the groups based on height, weight, or BMI (Table 2). Interestingly, age (analyzed by quartiles) did show a statistically significant difference in
RI PT
post-operative leaks, with higher leak rates in the 2nd quartile (ages 5.8-9.6) and 4th quartile (>16.2 years of age) (Table 2)
SC
The 55 patients that were included in the study underwent 70 EES for removal of tumor (Table 1). Five clival chordomas and one pituitary adenoma required planned staged surgical
M AN U
resections. Three clival chordomas, 7 craniopharyngiomas, and 2 pituitary adenomas required second surgeries for resection of residual/recurrent tumor. The most common surgical corridor used was the transsellar approach. Transclival, transpterygoid, transplanum, and transtuberculum approach modules were also used for resection, with several tumors requiring
TE D
multiple modules to remove the entire lesion.
Reconstruction of the tumor defect was most frequently accomplished using a vascularized flap (nasoseptal flap). 44 nasoseptal flaps were used as part of the reconstruction
EP
of these defects. Four patients had free mucosal middle turbinate grafts to reconstruct their defects. Fat grafts were used to augment the repair site in 20 patients.
AC C
Forty-seven patients had intra-operative CSF leaks. Of those 47 patients, 11 had post-
operative CSF leaks with a leak rate of 23.4%. The individual CSF leak rates were: clival chordomas 36% (4/11), pituitary pathology 25% (3/12), craniopharyngiomas 17% (4/24). The tumors were divided into sellar/suprasellar pathology versus posterior fossa pathology. Pituitary adenoma, pituitary carcinoma, Rathke’s cleft cyst, and craniopharyngioma were included in the sellar/suprasellar pathology group. Clival chordomas were included in the
ACCEPTED MANUSCRIPT
posterior fossa group. There was no statistical difference in leak rate comparing sellar/suprasellar pathology to posterior fossa pathology (p=0.256) (Table 2).
RI PT
All eleven patients with postoperative CSF leaks had operative repair of their CSF leaks. One patient required a pericranial flap. Three craniopharyngiomas, two clival chordomas, and 1 pituitary carcinoma developed meningitis due to their CSF leak (7/11). One craniopharyngioma
SC
developed aseptic meningitis in the absence of a CSF leak. All were treated with intravenous antibiotic therapy and surgical repair of the CSF leak with eventual resolution of the infection.
M AN U
Patients who were deemed high risk for post-operative hydrocephalus due to the size and location of their pathology had lumbar drains placed post-operatively. Twenty five patients were treated with a lumbar drain, 16 of which were craniopharyngiomas. Six patients had postoperative hydrocephalus, all of which were craniopharyngiomas. Six patients had unplanned
TE D
readmissions after hospital discharge. Post-operative hydrocephalus was not a risk factor for CSF leak on analysis and neither was post-operative use of a lumbar drain (p = 0.614 and 1.0 respectively).
EP
The type of reconstructive material/tissue used was also analyzed. The use of a vascularized rotational flap (nasoseptal flap) did not show a statistically different rate of post-
AC C
operative CSF leaks (p=.416). 28% of cases that used a vascularized flap had a post-operative leak, whereas 9% of those cases not using vascularized flap had a leak (Table 2). Other than CSF leak with an associated risk of meningitis, the post-operative
complication rate was low. One craniopharyngioma patient developed a hematoma that required evacuation in the operating room. The pituitary carcinoma patient developed an
ACCEPTED MANUSCRIPT
intracranial abscess that was drained. One clival chordoma patient developed occipital cervical
RI PT
instability that required cervical fusion, which is a known sequelae of resection.
Discussion:
Pediatric skull base reconstruction can be a challenge after extensive tumor resection or
SC
when treating a high flow leak. Pediatric patients have unique challenges that are not always encountered in the adult population. Patients as young as 2 years old have undergone
M AN U
extensive EES for excision of skull base lesions that often leave large skull base defects that require careful, watertight repair. The available tissue options in a 2 year old are significantly different than in a 17-year-old patient.
Pediatric nasoseptal flaps (NSF) were described in 2009 by Shah et al3. At that time, the
TE D
group concluded that nasoseptal flaps may not be a viable option in children < 10 years of age for anterior skull base reconstruction, because the average potential flap length was shorter than the average anterior skull base length. They found that transsellar/transplanum
EP
approaches could be adequately covered by a NSF in patients 6-7 years and older, however the length in patients <6 years was insufficient to cover the defect. They also stated that transclival
AC C
approaches were too deep to have adequate coverage by a NSF alone in all pediatric age groups. The UPMC group starting using nasoseptal flaps in pediatric patients in 2008. A second study in 2015 by Purcell et al4 evaluated CT scans of normal pediatric patients to assess reconstruction feasibility of transsellar defects only. They excluded patients who had incomplete pneumatization of the sphenoid sinus (defined as “ >2mm of marrow space anterior to the sella turcica”). However, pediatric skull base surgery often requires operating on
ACCEPTED MANUSCRIPT
patients who have incomplete pneumatization of the sphenoid, especially in patients under age 7 years. They report that out of 125 CT scans reviewed, only 5 patients had inadequate flap
RI PT
length to potential sellar defect length. They report that the ratio of flap length to sellar defect length is actually higher in younger patients then in older ones. This study is limited by the exclusion of incompletely pneumatized sinuses and they evaluated only 2 of 6 surgical patients
SC
in their analysis; 1 of the 2 developed a post-operative CSF leak for sellar pathology and reconstruction.
M AN U
Another small series of 10 pediatric skull base patients was reported by Giannoni and Whitehead in 2013.5 They report their experience with 10 pediatric patients that underwent EES for sellar pathology and had a nasoseptal flap repair of their skull base defect. They reported no post-operative CSF leaks in the 9 of 10 patients who underwent EES for
TE D
sellar/suprasellar lesions. The age range was reported as 5.1-17.4 years. In our series, vascularized versus non-vascularized tissue reconstruction did not show a statistical difference in post-operative CSF leak rate. A recent systematic review of the skull
EP
base literature by Soudry et al2 evaluated adult patients with skull base defects after EES. They focused on the location of the defect and whether a high or low flow defect was created during
AC C
surgery. They found that the vascularized versus non-vascularized flap reconstruction was not a risk factor for CSF leak in low-flow intraoperative CSF leaks. However there was a 94% vs 82% difference in successful skull base closure in high-flow leaks that used vascularized flaps for repair. In our study, there was not a statistical difference in post-operative CSF leak rates in clival resection cases compared to sellar/suprasellar resection. Similarly there was not a
ACCEPTED MANUSCRIPT
difference in vascularized versus non-vascularized reconstruction techniques. The higher leak rate observed with vascularized repair represents a selection bias in the use of vascularized
RI PT
repair. Larger defects are often treated with vascularized flaps. Whether this is due to the unique anatomic challenges in the pediatric patient or operative approach/philosophy was unable to be discerned in this study. The majority of our sellar/suprasellar tumor resections
SC
were for endocrinopathies (9 of 14 Pituitary Adenomas). This is significantly different from the adult literature where the primary indication for surgery is macroadenoma with optic nerve
M AN U
compression. The different pathophysiologic and anatomic differences between these two surgical indications for sellar/suprasellar resection may be a factor in why there was no difference in post-operative CSF leak rates.
Thorp et al6 reviewed a large series of adult patients with intra-operative CSF leaks after
TE D
EES. 151 patients with intra-operative CSF leaks were treated with vascularized reconstruction. Their success rate was 96% and they found that all of the post-operative CSF leaks were in highflow leak patients. They did not observe a higher rate of CSF leaks in patients who received
EP
radiation therapy post-operatively.
Clival chordomas had the highest leak rate within our study cohort. 36% of clival
AC C
chordoma patients had post-operative CSF leaks. Several studies have also noted the difficulty in obtaining water-tight closure after extensive clival resection. The systematic review by Soundry et al2 found that there was no difference in skull base reconstruction outcomes using a vascularized flap vs non-vascularized repair of skull base subsites except when trying to reconstruct the clivus. Similarly, in a study by Gruss et al7, there was a statistically significant higher risk of post-operative CSF leak after resection of a central skull base lesion (sella and
ACCEPTED MANUSCRIPT
clivus) as compared to an anterior skull base lesion. They also reported a statistically significant higher risk of post-operative CSF leak in patients with dural defects >2cm2. These studies were
RI PT
all performed in adults. Post-operative hydrocephalus and the use of lumbar drains were not significant risk factors for CSF leak in our study population. Soundry’s systematic review2 evaluated the use of
SC
lumbar drains in patients with intraoperative CSF leaks and they were unable to draw any
conclusions based on the data available. Interestingly, they state that lumbar drain placement
M AN U
is indicated as initial management of post-operative CSF leaks since 40% of patients had resolution of their CSF leak when a subsequent lumbar drain was placed. However, this means that 60% of patients required revision surgery for repair. Delayed repair of postoperative CSF leaks is also a risk factor for meningitis. A randomized trial of lumbar drains in an adult EES
TE D
population at UPMC confirmed benefit for patients with a transclival approach and large anterior cranial base defects but no benefit for sellar/suprasellar defects (unpublished data). Pediatric patients are unique in many ways and the size of the patient and their nasal
EP
corridor is a major consideration in considering the operative approach and method of reconstruction. In order to evaluate if patient size contributes to reconstructive options/repair
AC C
success, the patients’ weight, height, BMI, and growth percentiles based on age were evaluated. Our statistical analysis did not find a difference in CSF leak rates when comparing patients based on these factors. Interestingly, obese pediatric patients did not have higher CSF leak rates. This is in contrast to the adult population in which a higher BMI is a risk factor for higher CSF leak rate8.
ACCEPTED MANUSCRIPT
Surprisingly, the age of the patient was statistically significant when evaluating by quartiles. Patients ages 5.8-9.6 years old and those >16.2 years old had higher rates of leak. It
RI PT
does not appear that this is a surrogate marker for disease pathology as there was no statistical significance in pathology based on age quartiles. The difference in post op leaks by age group was also not caused by a difference in the use of a vascularized tissue flaps or not. Whether
SC
there is another underlying anatomic or developmental difference in these two age groups was unable to be ascertained from our data.
M AN U
The strengths of this study are the inclusion criteria. This is an exclusive pediatric study that evaluated the five most common pathologies. We evaluated only patients who sustained intraoperative CSF leaks so that we can have an accurate representation of the risk for postoperative CSF leak. An additional strength is that the same operative team with the same
TE D
reconstructive philosophy was used throughout the patient treatment cohort. Skill set and prior experience guide intra-operative decision making. We recognize that as surgical teams continue to grow and evolve as a unit that reconstructive approaches often change based on
EP
prior successes and failures.
The limitations of this study are the sample size. Even though it is the largest pediatric
AC C
only series, it is still a low number of patients: 55 total patients sustaining 47 intra-operative CSF leaks. There were only 11 post-operative CSF leaks which limited the statistical analysis of contributing factors. It is also a retrospective review and may suffer from incomplete documentation.
Conclusions:
ACCEPTED MANUSCRIPT
Pediatric patients have a higher rate of post-operative CSF leaks when undergoing EES of the skull base. The use of vascularized flaps and perioperative lumbar drains did not statistically
RI PT
decrease the rate of post-operative CSF leaks. Low body weight, short height, and postoperative hydrocephalus were not risk factors for CSF leaks in this population. Age of the
patient was a significant factor for post-operative leak rates. Clival chordomas appear to be a
AC C
EP
TE D
M AN U
reconstructive algorithm for high-risk pediatric patients.
SC
particularly high-risk group. Additional experience is needed to define the optimal
ACCEPTED MANUSCRIPT
References: 1. Tatreau JR, Patel MR, Shah RN, et al. Anatomic Considerations for Endoscopic Endonasal
RI PT
Skull Base Surgery in Pediatric Patients. Laryngoscope. 2010; 120: 1730-1737. 2. Soudry E, Turner JH, Nayak JV, Hwang PH. Endoscopic Reconstruction of Surgically
Created Skull Base Defects: A Systematic Review. Otolaryngol Head Neck Surg. 2014;
SC
150(5): 730-738.
3. Shah RN, Surowitz JB, Patel MR, Huang BY, Snyderman CH, Carrau RL, Kassam AB,
M AN U
Germanwala AV, Zanation AM. Endoscopic Pedicled Nasoseptal Flap Reconstruction for Pediatric Skull Base Defects. Laryngoscope. 2009; 119: 1067-1075. 4. Purcell PL, Shinn JR, Otto RK, Davis GE, Parikh SR. Nasoseptal Flap Reconstruction of Pediatric Sellar Defects: A Radiographic Feasibility Study and Case Series. Otolaryngol
TE D
Head Neck Surg. 2015;152(4): 746-751.
5. Giannoni CM, Whitehead WE. Use of endoscopic vascularized nasoseptal flap in children. Otolaryngol Head Neck Surg. 2013; 148: 344-346.
EP
6. Thorp BD, Sreenath SB, Ebert CS, Zanation AM. Endoscopic Skull Base Reconstruction: A Review and Clinical Case Series of 152 Vascularized Flaps Used for Surgical Skull Base
AC C
Defects in the Setting of Intraoperative Cerebrospinal Fluid Leak. Neurosurg Focus. 2014; 37: 1-7.
7. Gruss CL, Al Komser M, Aghi MK, Pletcher SD, Goldberg AN, McDermott M, El-Sayed IH. Risk Factors for Cerebrospinal Leak after Endoscopic Skull Base Reconstruction with Nasoseptal Flap. Otolaryngol Head Neck Surg. 2014; 151 (3): 516-521.
ACCEPTED MANUSCRIPT
8. Zanation AM, Carrau RL, Snyderman CH, Germanwala AV, Gardner PA, Prevedello DM, Kassam AB. Nasoseptal flap reconstruction of high flow intraoperative cerebral spinal
RI PT
fluid leaks during endoscopic skull base surgery. Am J Rhinol 2009; 23: 519-521. 9. Patel MR, Taylor RJ, Hackman TG, Germanwala AV, Sasaki-Adams D, Ewend MG,
Zanation AM. Beyond the Nasoseptal Flap: Outcomes and Pearls with Secondary Flaps in
SC
Endoscopic Endonasal Skull Base Reconstruction. Laryngoscope. 2014; 124: 846-852. 10. Emanuelli E, Bossolesi P, Borsetto D, D’Avella E. Endoscopic Repair of Cerebrospinal
AC C
EP
TE D
M AN U
Fluid Leak in Paediatric Patients. Int J Pediatr Otorhinolaryngol. 2014; 78(11) 1898-902.
ACCEPTED MANUSCRIPT
Table 1: Demographic information for surgical cohort
Craniopharyngioma Pituitary Carcinoma
11
10.23
15
4
14
16.72
17
1
9
15.43
10
0
20
8.46
27
0
1
17.80
1
0
55
13.7
70
5
AC C
EP
TE D
Totals:
# Staged Tumor Resections
RI PT
Rathke’s Cleft Cyst
# Resections
SC
Pituitary Adenoma
Age (yrs.)
M AN U
Clival Chordoma
Pathology
ACCEPTED MANUSCRIPT
Table 2: Risk of post-operative leak based on Age in quartiles, BMI, Site/Pathology,
Site/pathology
AC C
EP
TE D
Vascularized Flap
% with post op CSF leak 0.0 50.0 0.0 41.7 30.0 9.1 20.0 27.3 19.4 36.4 9.1 27.8
SC
BMI [quartile]
<=5.7 5.8-9.6 9.7-16.2 >16.2 <=15.8 15.9-18.9 19.0-28.4 >28.4 sellar/suprasellar posterior fossa no yes
# with post op CSF leak 0 6 0 5 3 1 2 3 7 4 1 10
M AN U
Age [quartile]
N 11 12 12 12 10 11 10 11 36 11 11 36
RI PT
and Vascularized Reconstruction. Fisher’s exact test used for statistical analysis.
p-value
0.001
0.692 0.256 0.416
S0165-5876(16)30455-4
DOI:
10.1016/j.ijporl.2016.12.019
Reference:
PEDOT 8354
To appear in:
International Journal of Pediatric Otorhinolaryngology
Received Date: 7 October 2016 Revised Date:
14 December 2016
Accepted Date: 15 December 2016
Please cite this article as: A.L Stapleton, E.C Tyler-Kabara, P.A Gardner, C.H Snyderman, E.W Wang, Risk Factors for Cerebrospinal Fluid Leak in Pediatric Patients Undergoing Endoscopic Endonasal Skull Base Surgery, International Journal of Pediatric Otorhinolaryngology (2017), doi: 10.1016/ j.ijporl.2016.12.019. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Risk Factors for Cerebrospinal Fluid Leak in Pediatric Patients Undergoing Endoscopic Endonasal Skull Base Surgery
RI PT
Stapleton, Amanda L, MD1; Tyler-Kabara, Elizabeth C, MD, PhD2; Gardner, Paul A, MD2; Snyderman, Carl H, MD, MBA1, 2, Wang Eric W, MD1
Department of Otolaryngology Head and Neck Surgery 1; Department of Neurological Surgery2,
SC
University of Pittsburgh Medical Center; Children’s Hospital of Pittsburgh of UPMC.
Center
M AN U
Research Performed at Children’s Hospital of Pittsburgh of University of Pittsburgh Medical
Short Running Title: CSF leak risk factors in pediatric skull base surgery Financial Support/Funding: None
TE D
Conflict of Interest Statement: Carl Snyderman: consultant for SPIWay LLC Corresponding Author Address: Amanda Stapleton, MD
EP
Children’s Hospital of Pittsburgh of UPMC 7th Floor Faculty Pavilion
AC C
4401 Penn Ave
Pittsburgh, PA 15224
Phone: 412-692-5466
Email: [email protected] Research Presented as an Oral Podium Presentation at the 25th Annual North American Skull Base Society Meeting. February 20-22, 2015 Tampa, Florida, USA
ACCEPTED MANUSCRIPT
Key Words: pediatric, skull base, cerebrospinal fluid leak, endoscopic surgery
AC C
EP
TE D
M AN U
SC
RI PT
Level of Evidence: 4
ACCEPTED MANUSCRIPT
Abstract Objectives: To determine the risk factors associated with cerebrospinal fluid (CSF) leak
RI PT
following endoscopic endonasal surgery (EES) for pediatric skull base lesions. Methods: Retrospective chart review of pediatric patients (ages 1 month to 18 years) treated for skull base lesions with EES from 1999-2014. Five pathologies were reviewed:
SC
craniopharyngioma, clival chordoma, pituitary adenoma, pituitary carcinoma, and Rathke’s cleft cyst. Fisher’s exact tests were used to evaluate the different factors to determine which
M AN U
had a statistically higher risk of leading to a post-operative CSF leak.
Results: 55 pediatric patients were identified who underwent 70 EES’s for tumor resection. Of the 70 surgeries, 47 surgeries had intraoperative CSF leaks that were repaired at the time of surgery. 11 of 47 (23%) surgeries had post-operative CSF leaks that required secondary
TE D
operative repair. Clival chordomas had the highest CSF leak rate at 36%. There was no statistical difference in leak rate based on the type of reconstruction, although 28% of cases that used a vascularized flap had a post-operative leak, whereas only 9% of those cases not
EP
using a vascularized flap had a leak. Post-operative hydrocephalus and perioperative use of a
AC C
lumbar drain were not significant risk factors.
Conclusions:
Pediatric patients with an intra-operative CSF leak during EES of the skull base have a
high rate of post-operative CSF leaks. Clival chordomas appear to be a particularly high-risk group. The use of vascularized flaps and perioperative lumbar drains did not statistically decrease the rate of post-operative CSF leak.
ACCEPTED MANUSCRIPT
Introduction: Pediatric patients who undergo endoscopic endonasal surgery (EES) for skull base
RI PT
tumors have unique pre-operative, intra-operative, and post-operative challenges in comparison to adult patients with similar lesions. As the indications for endonasal resection of pediatric skull base tumors widens, so do the reconstructive challenges. A post-operative
SC
cerebrospinal fluid (CSF) leak can place the patient at risk for significant complications including pneumocephalus, mental status changes, and meningitis. Our 15-year history of treating
M AN U
pediatric skull base tumors using the endoscopic endonasal approach (EEA) has given us a unique perspective on the surgical and post-operative challenges that these patients face. Tatreau et al1 in 2010 evaluated the anatomic considerations in pediatric patients undergoing EES. They found three areas of limitation in the pediatric patient as compared to
TE D
adults. The nasal pyriform aperture, sphenoid sinus pneumatization and sphenoid bone thickness, and intercarotid distance within the sphenoid sinus were all areas where anatomic limitations could influence surgical outcomes. All of these factors can effect surgical outcomes
EP
and the reconstructive options available to the pediatric skull base surgeon. The adult literature on the use of reconstructive materials including bio-synthetic
AC C
products and local vascularized flaps to repair intra-operative CSF leaks has been a recent area of intense review. A systematic review of the adult literature evaluated 22 studies looking to create an evidence-based algorithm for skull base reconstruction techniques. However, due to the wide variety of reconstructive techniques used, inconsistent reporting of high or low flow intra-operative CSF leaks, and lack of uniform post-operative outcomes reporting, a true algorithm could not be created. They focused on two factors: location of skull base defect and
ACCEPTED MANUSCRIPT
degree of intraoperative CSF flow. They concluded that repair of high-flow intraoperative CSF leaks with pedicled vascularized flaps was superior to non-vascularized flaps and that location
RI PT
of the defect did not matter, with the exception of the clivus2. There are case series in the literature addressing the feasibility of nasoseptal flap repair of skull base defects after undergoing EES in the pediatric population 3-5 . However, the
M AN U
leaks in pediatric patients undergoing EES.
SC
objective of this study was to determine the risk factors associated with post-operative CSF
Methods:
This is an IRB approved retrospective chart review from January 1999 to September 2014 at Children’s Hospital of Pittsburgh of the University of Pittsburgh Medical Center. Inclusion
TE D
criteria were pediatric patients, ages 1 month to 18 years old, who had EES for surgical resection of five high-risk tumor pathologies Craniopharyngioma, Pituitary Tumors (Adenoma, Carcinoma, Rathke’s cleft cyst), and Clival Chordoma. Only patients who had intraoperative
EP
CSF leaks were included in the study. These 5 tumor types were included due to their relatively high prevalence and/or a high rate of intraoperative CSF leak.
AC C
Indications for surgery included endocrine abnormalities, cranial nerve palsies, vision loss, orbital complications, persistent headaches, hydrocephalus, or other mental status changes. Image-guided navigation was used for all cases. All tumors were removed by the skull base team, which consists of a neurosurgeon and otolaryngologist operating concurrently. Gross total resection was the primary goal in the surgical resection of craniopharyngiomas, pituitary adenomas, pituitary carcinoma, and clival chordomas. Rathke’s cleft cysts were opened and
ACCEPTED MANUSCRIPT
drained into the sphenoid sinus. All clival chordomas were referred for proton beam radiation therapy following surgical resection. The study parameters included: pathologic diagnosis, age, gender, weight, height,
RI PT
growth percentile, surgical corridor approach (transsellar, transclival, transpterygoid,
transplanum, transtuberculum), reconstruction technique, intra-operative CSF leak, postoperative complications, tumor recurrence, post-operative CSF leak, hydrocephalus, and
SC
lumbar drain use. Fisher’s exact test was used to determine which factors had a statistically
M AN U
significant risk (p<0.05) of leading to a post-operative CSF leak.
Results:
170 pediatric patients who underwent EES from January 1999 to September 2014 were
TE D
identified. Of those, 55 patients met our inclusion criteria.
The pathology included 20 craniopharyngiomas, 14 pituitary adenomas, 11 clival chordomas, 9 Rathke’s Cleft cysts, and 1 pituitary carcinoma. The overall average age for the
EP
cohort was 13.7 years with craniopharyngiomas being the youngest group averaging 8.43 years old (Table 1). Of the 14 pituitary adenomas, 5 presented with Cushing’s Disease, 3
AC C
Prolactinomas, 2 Pituitary Apoplexy, 2 with vision loss, 1 Acromegaly, and 1 incidental imaging finding.
Patient size was recorded based upon height, weight, BMI, and growth percentile. The
size of the patient was evaluated to determine if this affected the incidence of CSF leak. Pituitary adenomas were our largest patients averaging the 92nd percentile for weight and a BMI of 42. The smallest patients in our group were the craniopharyngiomas, averaging 55th
ACCEPTED MANUSCRIPT
percentile for weight and a BMI of 19. There was no statistical difference in CSF leak rates for the groups based on height, weight, or BMI (Table 2). Interestingly, age (analyzed by quartiles) did show a statistically significant difference in
RI PT
post-operative leaks, with higher leak rates in the 2nd quartile (ages 5.8-9.6) and 4th quartile (>16.2 years of age) (Table 2)
SC
The 55 patients that were included in the study underwent 70 EES for removal of tumor (Table 1). Five clival chordomas and one pituitary adenoma required planned staged surgical
M AN U
resections. Three clival chordomas, 7 craniopharyngiomas, and 2 pituitary adenomas required second surgeries for resection of residual/recurrent tumor. The most common surgical corridor used was the transsellar approach. Transclival, transpterygoid, transplanum, and transtuberculum approach modules were also used for resection, with several tumors requiring
TE D
multiple modules to remove the entire lesion.
Reconstruction of the tumor defect was most frequently accomplished using a vascularized flap (nasoseptal flap). 44 nasoseptal flaps were used as part of the reconstruction
EP
of these defects. Four patients had free mucosal middle turbinate grafts to reconstruct their defects. Fat grafts were used to augment the repair site in 20 patients.
AC C
Forty-seven patients had intra-operative CSF leaks. Of those 47 patients, 11 had post-
operative CSF leaks with a leak rate of 23.4%. The individual CSF leak rates were: clival chordomas 36% (4/11), pituitary pathology 25% (3/12), craniopharyngiomas 17% (4/24). The tumors were divided into sellar/suprasellar pathology versus posterior fossa pathology. Pituitary adenoma, pituitary carcinoma, Rathke’s cleft cyst, and craniopharyngioma were included in the sellar/suprasellar pathology group. Clival chordomas were included in the
ACCEPTED MANUSCRIPT
posterior fossa group. There was no statistical difference in leak rate comparing sellar/suprasellar pathology to posterior fossa pathology (p=0.256) (Table 2).
RI PT
All eleven patients with postoperative CSF leaks had operative repair of their CSF leaks. One patient required a pericranial flap. Three craniopharyngiomas, two clival chordomas, and 1 pituitary carcinoma developed meningitis due to their CSF leak (7/11). One craniopharyngioma
SC
developed aseptic meningitis in the absence of a CSF leak. All were treated with intravenous antibiotic therapy and surgical repair of the CSF leak with eventual resolution of the infection.
M AN U
Patients who were deemed high risk for post-operative hydrocephalus due to the size and location of their pathology had lumbar drains placed post-operatively. Twenty five patients were treated with a lumbar drain, 16 of which were craniopharyngiomas. Six patients had postoperative hydrocephalus, all of which were craniopharyngiomas. Six patients had unplanned
TE D
readmissions after hospital discharge. Post-operative hydrocephalus was not a risk factor for CSF leak on analysis and neither was post-operative use of a lumbar drain (p = 0.614 and 1.0 respectively).
EP
The type of reconstructive material/tissue used was also analyzed. The use of a vascularized rotational flap (nasoseptal flap) did not show a statistically different rate of post-
AC C
operative CSF leaks (p=.416). 28% of cases that used a vascularized flap had a post-operative leak, whereas 9% of those cases not using vascularized flap had a leak (Table 2). Other than CSF leak with an associated risk of meningitis, the post-operative
complication rate was low. One craniopharyngioma patient developed a hematoma that required evacuation in the operating room. The pituitary carcinoma patient developed an
ACCEPTED MANUSCRIPT
intracranial abscess that was drained. One clival chordoma patient developed occipital cervical
RI PT
instability that required cervical fusion, which is a known sequelae of resection.
Discussion:
Pediatric skull base reconstruction can be a challenge after extensive tumor resection or
SC
when treating a high flow leak. Pediatric patients have unique challenges that are not always encountered in the adult population. Patients as young as 2 years old have undergone
M AN U
extensive EES for excision of skull base lesions that often leave large skull base defects that require careful, watertight repair. The available tissue options in a 2 year old are significantly different than in a 17-year-old patient.
Pediatric nasoseptal flaps (NSF) were described in 2009 by Shah et al3. At that time, the
TE D
group concluded that nasoseptal flaps may not be a viable option in children < 10 years of age for anterior skull base reconstruction, because the average potential flap length was shorter than the average anterior skull base length. They found that transsellar/transplanum
EP
approaches could be adequately covered by a NSF in patients 6-7 years and older, however the length in patients <6 years was insufficient to cover the defect. They also stated that transclival
AC C
approaches were too deep to have adequate coverage by a NSF alone in all pediatric age groups. The UPMC group starting using nasoseptal flaps in pediatric patients in 2008. A second study in 2015 by Purcell et al4 evaluated CT scans of normal pediatric patients to assess reconstruction feasibility of transsellar defects only. They excluded patients who had incomplete pneumatization of the sphenoid sinus (defined as “ >2mm of marrow space anterior to the sella turcica”). However, pediatric skull base surgery often requires operating on
ACCEPTED MANUSCRIPT
patients who have incomplete pneumatization of the sphenoid, especially in patients under age 7 years. They report that out of 125 CT scans reviewed, only 5 patients had inadequate flap
RI PT
length to potential sellar defect length. They report that the ratio of flap length to sellar defect length is actually higher in younger patients then in older ones. This study is limited by the exclusion of incompletely pneumatized sinuses and they evaluated only 2 of 6 surgical patients
SC
in their analysis; 1 of the 2 developed a post-operative CSF leak for sellar pathology and reconstruction.
M AN U
Another small series of 10 pediatric skull base patients was reported by Giannoni and Whitehead in 2013.5 They report their experience with 10 pediatric patients that underwent EES for sellar pathology and had a nasoseptal flap repair of their skull base defect. They reported no post-operative CSF leaks in the 9 of 10 patients who underwent EES for
TE D
sellar/suprasellar lesions. The age range was reported as 5.1-17.4 years. In our series, vascularized versus non-vascularized tissue reconstruction did not show a statistical difference in post-operative CSF leak rate. A recent systematic review of the skull
EP
base literature by Soudry et al2 evaluated adult patients with skull base defects after EES. They focused on the location of the defect and whether a high or low flow defect was created during
AC C
surgery. They found that the vascularized versus non-vascularized flap reconstruction was not a risk factor for CSF leak in low-flow intraoperative CSF leaks. However there was a 94% vs 82% difference in successful skull base closure in high-flow leaks that used vascularized flaps for repair. In our study, there was not a statistical difference in post-operative CSF leak rates in clival resection cases compared to sellar/suprasellar resection. Similarly there was not a
ACCEPTED MANUSCRIPT
difference in vascularized versus non-vascularized reconstruction techniques. The higher leak rate observed with vascularized repair represents a selection bias in the use of vascularized
RI PT
repair. Larger defects are often treated with vascularized flaps. Whether this is due to the unique anatomic challenges in the pediatric patient or operative approach/philosophy was unable to be discerned in this study. The majority of our sellar/suprasellar tumor resections
SC
were for endocrinopathies (9 of 14 Pituitary Adenomas). This is significantly different from the adult literature where the primary indication for surgery is macroadenoma with optic nerve
M AN U
compression. The different pathophysiologic and anatomic differences between these two surgical indications for sellar/suprasellar resection may be a factor in why there was no difference in post-operative CSF leak rates.
Thorp et al6 reviewed a large series of adult patients with intra-operative CSF leaks after
TE D
EES. 151 patients with intra-operative CSF leaks were treated with vascularized reconstruction. Their success rate was 96% and they found that all of the post-operative CSF leaks were in highflow leak patients. They did not observe a higher rate of CSF leaks in patients who received
EP
radiation therapy post-operatively.
Clival chordomas had the highest leak rate within our study cohort. 36% of clival
AC C
chordoma patients had post-operative CSF leaks. Several studies have also noted the difficulty in obtaining water-tight closure after extensive clival resection. The systematic review by Soundry et al2 found that there was no difference in skull base reconstruction outcomes using a vascularized flap vs non-vascularized repair of skull base subsites except when trying to reconstruct the clivus. Similarly, in a study by Gruss et al7, there was a statistically significant higher risk of post-operative CSF leak after resection of a central skull base lesion (sella and
ACCEPTED MANUSCRIPT
clivus) as compared to an anterior skull base lesion. They also reported a statistically significant higher risk of post-operative CSF leak in patients with dural defects >2cm2. These studies were
RI PT
all performed in adults. Post-operative hydrocephalus and the use of lumbar drains were not significant risk factors for CSF leak in our study population. Soundry’s systematic review2 evaluated the use of
SC
lumbar drains in patients with intraoperative CSF leaks and they were unable to draw any
conclusions based on the data available. Interestingly, they state that lumbar drain placement
M AN U
is indicated as initial management of post-operative CSF leaks since 40% of patients had resolution of their CSF leak when a subsequent lumbar drain was placed. However, this means that 60% of patients required revision surgery for repair. Delayed repair of postoperative CSF leaks is also a risk factor for meningitis. A randomized trial of lumbar drains in an adult EES
TE D
population at UPMC confirmed benefit for patients with a transclival approach and large anterior cranial base defects but no benefit for sellar/suprasellar defects (unpublished data). Pediatric patients are unique in many ways and the size of the patient and their nasal
EP
corridor is a major consideration in considering the operative approach and method of reconstruction. In order to evaluate if patient size contributes to reconstructive options/repair
AC C
success, the patients’ weight, height, BMI, and growth percentiles based on age were evaluated. Our statistical analysis did not find a difference in CSF leak rates when comparing patients based on these factors. Interestingly, obese pediatric patients did not have higher CSF leak rates. This is in contrast to the adult population in which a higher BMI is a risk factor for higher CSF leak rate8.
ACCEPTED MANUSCRIPT
Surprisingly, the age of the patient was statistically significant when evaluating by quartiles. Patients ages 5.8-9.6 years old and those >16.2 years old had higher rates of leak. It
RI PT
does not appear that this is a surrogate marker for disease pathology as there was no statistical significance in pathology based on age quartiles. The difference in post op leaks by age group was also not caused by a difference in the use of a vascularized tissue flaps or not. Whether
SC
there is another underlying anatomic or developmental difference in these two age groups was unable to be ascertained from our data.
M AN U
The strengths of this study are the inclusion criteria. This is an exclusive pediatric study that evaluated the five most common pathologies. We evaluated only patients who sustained intraoperative CSF leaks so that we can have an accurate representation of the risk for postoperative CSF leak. An additional strength is that the same operative team with the same
TE D
reconstructive philosophy was used throughout the patient treatment cohort. Skill set and prior experience guide intra-operative decision making. We recognize that as surgical teams continue to grow and evolve as a unit that reconstructive approaches often change based on
EP
prior successes and failures.
The limitations of this study are the sample size. Even though it is the largest pediatric
AC C
only series, it is still a low number of patients: 55 total patients sustaining 47 intra-operative CSF leaks. There were only 11 post-operative CSF leaks which limited the statistical analysis of contributing factors. It is also a retrospective review and may suffer from incomplete documentation.
Conclusions:
ACCEPTED MANUSCRIPT
Pediatric patients have a higher rate of post-operative CSF leaks when undergoing EES of the skull base. The use of vascularized flaps and perioperative lumbar drains did not statistically
RI PT
decrease the rate of post-operative CSF leaks. Low body weight, short height, and postoperative hydrocephalus were not risk factors for CSF leaks in this population. Age of the
patient was a significant factor for post-operative leak rates. Clival chordomas appear to be a
AC C
EP
TE D
M AN U
reconstructive algorithm for high-risk pediatric patients.
SC
particularly high-risk group. Additional experience is needed to define the optimal
ACCEPTED MANUSCRIPT
References: 1. Tatreau JR, Patel MR, Shah RN, et al. Anatomic Considerations for Endoscopic Endonasal
RI PT
Skull Base Surgery in Pediatric Patients. Laryngoscope. 2010; 120: 1730-1737. 2. Soudry E, Turner JH, Nayak JV, Hwang PH. Endoscopic Reconstruction of Surgically
Created Skull Base Defects: A Systematic Review. Otolaryngol Head Neck Surg. 2014;
SC
150(5): 730-738.
3. Shah RN, Surowitz JB, Patel MR, Huang BY, Snyderman CH, Carrau RL, Kassam AB,
M AN U
Germanwala AV, Zanation AM. Endoscopic Pedicled Nasoseptal Flap Reconstruction for Pediatric Skull Base Defects. Laryngoscope. 2009; 119: 1067-1075. 4. Purcell PL, Shinn JR, Otto RK, Davis GE, Parikh SR. Nasoseptal Flap Reconstruction of Pediatric Sellar Defects: A Radiographic Feasibility Study and Case Series. Otolaryngol
TE D
Head Neck Surg. 2015;152(4): 746-751.
5. Giannoni CM, Whitehead WE. Use of endoscopic vascularized nasoseptal flap in children. Otolaryngol Head Neck Surg. 2013; 148: 344-346.
EP
6. Thorp BD, Sreenath SB, Ebert CS, Zanation AM. Endoscopic Skull Base Reconstruction: A Review and Clinical Case Series of 152 Vascularized Flaps Used for Surgical Skull Base
AC C
Defects in the Setting of Intraoperative Cerebrospinal Fluid Leak. Neurosurg Focus. 2014; 37: 1-7.
7. Gruss CL, Al Komser M, Aghi MK, Pletcher SD, Goldberg AN, McDermott M, El-Sayed IH. Risk Factors for Cerebrospinal Leak after Endoscopic Skull Base Reconstruction with Nasoseptal Flap. Otolaryngol Head Neck Surg. 2014; 151 (3): 516-521.
ACCEPTED MANUSCRIPT
8. Zanation AM, Carrau RL, Snyderman CH, Germanwala AV, Gardner PA, Prevedello DM, Kassam AB. Nasoseptal flap reconstruction of high flow intraoperative cerebral spinal
RI PT
fluid leaks during endoscopic skull base surgery. Am J Rhinol 2009; 23: 519-521. 9. Patel MR, Taylor RJ, Hackman TG, Germanwala AV, Sasaki-Adams D, Ewend MG,
Zanation AM. Beyond the Nasoseptal Flap: Outcomes and Pearls with Secondary Flaps in
SC
Endoscopic Endonasal Skull Base Reconstruction. Laryngoscope. 2014; 124: 846-852. 10. Emanuelli E, Bossolesi P, Borsetto D, D’Avella E. Endoscopic Repair of Cerebrospinal
AC C
EP
TE D
M AN U
Fluid Leak in Paediatric Patients. Int J Pediatr Otorhinolaryngol. 2014; 78(11) 1898-902.
ACCEPTED MANUSCRIPT
Table 1: Demographic information for surgical cohort
Craniopharyngioma Pituitary Carcinoma
11
10.23
15
4
14
16.72
17
1
9
15.43
10
0
20
8.46
27
0
1
17.80
1
0
55
13.7
70
5
AC C
EP
TE D
Totals:
# Staged Tumor Resections
RI PT
Rathke’s Cleft Cyst
# Resections
SC
Pituitary Adenoma
Age (yrs.)
M AN U
Clival Chordoma
Pathology
ACCEPTED MANUSCRIPT
Table 2: Risk of post-operative leak based on Age in quartiles, BMI, Site/Pathology,
Site/pathology
AC C
EP
TE D
Vascularized Flap
% with post op CSF leak 0.0 50.0 0.0 41.7 30.0 9.1 20.0 27.3 19.4 36.4 9.1 27.8
SC
BMI [quartile]
<=5.7 5.8-9.6 9.7-16.2 >16.2 <=15.8 15.9-18.9 19.0-28.4 >28.4 sellar/suprasellar posterior fossa no yes
# with post op CSF leak 0 6 0 5 3 1 2 3 7 4 1 10
M AN U
Age [quartile]
N 11 12 12 12 10 11 10 11 36 11 11 36
RI PT
and Vascularized Reconstruction. Fisher’s exact test used for statistical analysis.
p-value
0.001
0.692 0.256 0.416