Comparison of Results Between Posterior Fossa Decompression with and without Duraplasty for the Surgical Treatment of Chiari Malformation Type I: A Systematic Review and Meta-Analysis

Literature Review

Comparison of Results Between Posterior Fossa Decompression with and without Duraplasty for the Surgical Treatment of Chiari Malformation Type I: A Systematic Review and Meta-Analysis Weiwei Lin1, Guman Duan2, Jinjin Xie3, Jiashen Shao2, Zhaoqi Wang1, Baohua Jiao1

Key words Chiari malformation (CM-1) - Duraplasty - Meta-analysis - Posterior fossa decompression - Systematic review

- BACKGROUND:

Abbreviations and Acronyms CI: Confidence interval CM-1: Chiari malformation type I CSF: Cerebrospinal fluid MRI: Magnetic resonance imaging PFD: Posterior fossa decompression PFDD: Posterior fossa decompression with duraplasty RR: Relative risk

- METHODS:

-

1

From the Department of Neurosurgery, Second Hospital of Hebei Medical University, Shijiazhuang City, Hebei Province, China; 2Department of Orthopedics, Third Hospital of Hebei Medical University, Shijiazhuang City, Hebei Province, China; and 3Departamento dental, Complejo Hospitalario de la Universidad de Santiago de Compostela, Rúa da Choupana, A Coruña, Spain To whom correspondence should be addressed: Baohua Jiao, M.D. [E-mail: [email protected]] Supplementary digital content available online. Citation: World Neurosurg. (2017). https://doi.org/10.1016/j.wneu.2017.10.161 Journal homepage: www.WORLDNEUROSURGERY.org Available online: www.sciencedirect.com 1878-8750/$ - see front matter ª 2017 Elsevier Inc. All rights reserved.

Posterior fossa decompression without (PFD) or with duraplasty (PFDD) for the treatment of type 1 Chiari malformation (CM-1) is controversial. We thus performed a systematic review and meta-analysis of studies to assess the effect on clinical and imaging improvement, operative time, complications, and recurrence rate between PFD and PFDD in patients with CM-1.

We systematically searched PubMed, Embase, Cochrane, Web of Knowledge, and ClinicalTrials.gov for retrospective or prospective studies comparing PFD with PFDD. Our main end points were clinical and imaging improvement, operative time, complications, and recurrence rate. We assessed pooled data by use of a fixed-effects or random-effects model according to the between-study heterogeneity.

- RESULTS:

Of 214 identified studies, 13 were eligible and were included in our analysis (N [ 3481 patients). Compared with PFD, PFDD led to a mean greater increase in operative time than did PFD [standardized mean difference, e2.35; 95% confidence interval [CI], (e2.70 to e1.99)], a higher likelihood of clinical improvement in patients with syringomyelia (relative risk [RR], 0.70; 95% CI, 0.49e0.98), no increased RR of clinical improvement in patients without syringomyelia, no increased RR of imaging improvement, but an increased RR of cerebrospinal fluiderelated complications (RR, 0.29; 95% CI, 0.15e0.58), cerebrospinal fluid leak, aseptic meningitis, pseudomeningocele, and a decreased likelihood of recurrence rate.

- CONCLUSIONS:

PFDD can be an optimal surgical strategy because of its higher clinical improvement and lower recurrence rate in the patients with syringomyelia. In patients without syringomyelia, PFD can be a preferred choice because of its similar clinical improvement and lower costs. Future randomized studies with large numbers and the power to provide illumination for surgical decision making in CM-1 are warranted.

BACKGROUND Chiari malformation type I (CM-1), in which the caudal poles of the cerebellar tonsils extend into the upper cervical spinal canal with no involvement of the brainstem, is the most common clinical type.1 The herniation of caudal descent of the cerebellar tonsils may lead to significant clinical signs and symptoms, including pain, altered sensation, weakness, dysphagia, sleep apnea, sensory deficit, weakness in extremities, and atrophy. In patients with CM-1, 70%e80% have been accompanying syringomyelia.2,3 There have been several treatment modalities to reduce the volume

of syringomyelic cavity and the pressure on the brainstem in CM-1.4 Posterior fossa decompression (PFD) with duraplasty (PFDD) has been proposed as a surgical treatment strategy for CM-1 that could provide better clinical improvement.5 Some studies have shown that PFDD has a higher likelihood of cerebrospinal fluid (CSF)-related complications than does PFD. However, some believe that the CSFrelated morbidity associated with autologous duraplasty for CM-1 in a uniformly treated population is negligible.6 However, a

WORLD NEUROSURGERY -: ---, MONTH 2017

study by Förander et al.7 shows that clinical improvement was not higher after primary decompression with duraplasty. Thus, we performed a meta-analysis to evaluate the advantages of the 2 strategies for the treatment of CM-1.

METHODS Search Strategy and Selection Criteria This systematic review and meta-analysis is reported in accordance with the PRISMA

www.WORLDNEUROSURGERY.org

1

LITERATURE REVIEW WEIWEI LIN ET AL.

PFD AND PFDD IN CM-I: META-ANALYSIS

“duraplasty,” and “duroplasty.” The complete search used for PubMed was: (chairi malformation) AND ((duraplasty) OR duroplasty). We considered all potentially eligible studies for review, irrespective of the primary outcome or language. We also performed a manual search, using the reference lists of key articles published in English.

214 studies identified 196 studies excluded on the basis of title and abstract No additional study identified through manual search

Study Selection and Data Extraction We regarded studies as eligible for inclusion if they were randomized or nonrandomized controlled trials undertaken with patients with CM-1, compared PFD with PFDD and reported changes in clinical or imaging improvements, or operative time, or complications, or recurrence rate. Exclusion criteria were as follows: case reports, comments, letters, animal studies, and studies that did not assess PFD and PFDD. The outcomes assessed were as follows: clinical and imaging improvements, operative time, complications (overall complications, CSF-related complications, and wound infection), and recurrence rate after surgery. Two independent investigators (W.L. and G.D.) reviewed study titles and abstracts, and studies that satisfied the

18 potentially relevant publications 5 studies excluded inappropriate study design did not only compare PFD and PFDD (n = 3) did not for the patients with CM-I (n = 1) no outcome required (n = 1)

13 studies included in meta-analysis Figure 1. Study selection process.

by searching Embase, PubMed, Cochrane, and Web of Science. We applied no language restrictions. We used the following combined text: “chairi malformation,”

(Preferred Reporting Items for Systematic Reviews and Meta-Analyses) Statement.8 We selected relevant studies published between January 1, 1980, and May 12, 2017, Table 1. Characteristics of Studies Included Operation Year

n

Munshi et al.13

2000

Lemonade and Seldon15

2004

15

Galarza et al.

Follow-Up Time

PFD

PFDD

PFD

34

11

23

9 months to 8 years

24

12

12

15.7 months (4e27 months)

Sex (Male/Female)

PFDD (Months)

14.8 (3e30)

PFD

PFDD

4/7 5/7

Mean Age (Standard Deviation), Years PFD

PFDD

7/16

38

29.9

6/6

7.6

10.8

2007

41

20

21

21 months (1e10 years)

16

McGirt et al.

2008

256

116

140

25
15 months

60/56

61/79

8
5

12
4

Erdogan et al.4

2010

27

12

15

NA

9/3

15/0

31.58
11.4

25.86
13

Muchnik et al.17

2010

120

56

64

0.5 years

58/63

Romero et al.

2010

16

6

10

9 months to 2 years

3/3

Yilmaz et al.19

2011

82

24

58

NA

36/46

Lee et al.

2014

65

29

36

23.7
5.6 months

Gurbuz et al.21

2015

25

12

13

44 months (12e75 months)

13/12

36

2015

39

18

21

NA

NA

36

2649 1593

1056

NA

762/831

476/580

9.8
0.2

10.9
0.3

1 year

9/24

23/70

40.8
11.0

40.6
10.5

18

20

Gurbuz et al.22 23

Shweikeh et al.

2015

Chen et al.24

2017

103

33

70

NA 29
18

23.5
5.8

15/14

10

11.1 3/7

16/20

40.5

40.7

31

38.9

8.9
5.2

9.9
5.3

PFD, posterior fossa decompression; PFDD, posterior fossa decompression with duraplasty; NA, not available.

2

www.SCIENCEDIRECT.com

WORLD NEUROSURGERY, https://doi.org/10.1016/j.wneu.2017.10.161

LITERATURE REVIEW

8 1 1 1 1 1

Quality of Included Studies The 2 reviewers (W.L. and G.D.) independently evaluated the quality of the studies included using the NewcastleOttawa Scale9 based on the 3 main items: the selection of the study groups (0e4 points), the comparability of the groups (0e2 points), and the determination of either the exposure or the outcome of interest (0e3 points), with a perfect score of 9.

Total scores were calculated by adding up the points awarded in each item.

1 1 2017 Chen et al.

24

Shweikeh et al.23

Gurbuz et al.

22

Gurbuz et al.

21

Lee et al.20

Yilmaz et al.

19

Romero et al.

18

Muchnik et al.17

Erdogan et al.

4

McGirt et al.16

Galarza et al.

15

Limonadi and Seldon

WORLD NEUROSURGERY -: ---, MONTH 2017

inclusion criteria were retrieved for fulltext assessment. Trials selected for detailed analysis and data extraction were analyzed by 2 investigators (W.L. and G.D.) with an agreement value (k) of 96.5%; disagreements were resolved by a third investigator (J.X.). We extracted the following data from each selected study: total number of participants, age, gender, follow-up time, the choice of the 2 surgical techniques, surgical technique (the extent of bone decompression and materials used for the duraplasty), operative time for each group, number of participants who have clinical improvement, and number of participants with any complications.

1

7

6

0

1

0

1

1

0

1

1

1

1 1

1 2015

2015

1

1

1

1

8 1 1 2015

1

1

1

1

1

1

8

7

1

1

0

1

1

1

1

1

1

1 1

1 2011

2014

1

1

1

1

8 1 1 2010

1

1

1

1

1

1

8

7

1

1

0

1

1

1

1

1

1

1 1

1 2010

2010

1

1

1

1

8

7

1

1

1

1

1

0

1

1

1

1 1

1 2007

2008

1

1

1

1

8 1 1 1 2004

1

1

1

1

1

1 1 1 1 1 1 2000 Munshi et al.13

15

Year Study

Representativeness of the Exposed Cohort

1

Ascertainment of Exposure

1

Assessment of Outcome

Adequacy of Follow-Up of Cohorts Follow-Up Long Enough for Outcomes to Occur Basis of the Design or Analysis Outcome of Interest Was Not Present at Start of Study Selection of the Nonexposed Cohort

Outcome (Score) Comparability (Score) Selection (Score)

Table 2. Methodological Quality Assessments Based on the Newcastle-Ottawa Scale

8

PFD AND PFDD IN CM-I: META-ANALYSIS

Total Scores

WEIWEI LIN ET AL.

Statistical Analysis We assessed the surgical effects on 4 outcomes: clinical syndrome release, as assessed by both clinical and imaging improvements, operative time, incidence of any complications, and recurrence rate. We analyzed operative time as continuous variables and reported arithmetic means for each group. For analyses of the proportion of participants in clinical and imaging improvements, complications (overall complications, CSF-related complications, and wound infection), and recurrence rate, we calculated an overall relative risk (RR). We calculated pooled estimates of the mean differences in operative time between the 2 groups by using a fixed-effects model (Mantel-Haenszel) to adequately account for the additional uncertainty associated with interstudy variability in the effect of different surgical strategy. For categorical outcomes, we also calculated pooled estimates of the RR with a fixedeffects (Mantel-Haenszel) or randomeffects (DerSimonian-Laird) model judging by between-study heterogeneity. The outcomes of meta-analysis were summarized using a forest plot. In the

www.WORLDNEUROSURGERY.org

3

LITERATURE REVIEW WEIWEI LIN ET AL.

PFD AND PFDD IN CM-I: META-ANALYSIS

SMD (95% CI)

Limonadi et al. (2004)

–2.19 (–3.22 to –1.16) 11.92

Mutchnick et al. (2010)

–2.46 (–2.93 to –1.98) 55.66

Lee (2014)

–2.22 ( –2.84 to –1.59) 32.42

Overall (I2 = 0.0%, P = 0.798)

–2.35 ( –2.70 to –1.99) 100.00

-3

-2.35

-2

Figure 2. Meta-analysis of posterior fossa decompression without (PFD) or with duraplasty (PFDD), comparing operative time. The gray shaded

meta-analysis of each outcome, we performed preplanned sensitivity analysis (based on the following factors: assessment quality of included studies, the extent of bone decompression, materials used for the duraplasty, areas of participants, publication year of studies, and the choice of the 2 techniques by surgeons) restricted to trials that compared PFD with PFDD. This comparison is the most important clinical question pertaining to the role of PFDD and also reduced the heterogeneity of the treatment-induced changes in outcomes in the comparator arm seen in the overall analysis. We assessed the possibility of publication bias by constructing a funnel plot of the effect size of each trial against the standard error if more than 10 studies were selected in one forest plot. We assessed funnel plot asymmetry using the Egger test and defined significant publication bias as a P value < 0.1. The trim-and-fill computation was used to estimate the effect of publication bias on the interpretation of the results.10 We used the Cochran Q test to assess heterogeneity between studies with significance set at P < 0.10.11 We also performed I2 testing to assess the magnitude of the heterogeneity between studies, with

4

Weight %

www.SCIENCEDIRECT.com

-1

Favours PFD

0

Favours PFDD

area is the weight of the estimate in proportion to the overall effect. SMD, standardized mean difference.

values greater than 40% regarded as being indicative of moderate-to-high heterogeneity12 A random-effect model was used to calculate pooled RRs in the case of significant heterogeneity (P < 0.10 or I2 > 50%); otherwise, a fixed-effects model was used. We used Stata (version 14.2 [StataCorp LLC, College Station, Texas, USA]) for all statistical analyses. Role of the Funding Source The study was supported by intramural funds, with no commercial entity involved. The funding source had no role in study design, data collection, data analysis, data interpretation, or writing of the report. The corresponding author had full access to all the data in the study and had final responsibility for the decision to submit for publication. RESULTS We identified 214 studies, of which 13 (with data for 3481 participants, 1593 of whom adopted PFD and the remaining 1056 of whom received PFDD) were included in our analysis (Figure 1). The 13 trials were published between 2000 and 2017 (Table 1).4,13-24 Mean follow-up duration of patients in 9 studies was

more than 6 months and in the other 4 studies it was not mentioned, as shown in Table 1. Eleven of 13 studies performed a small suboccipital craniectomy and a C1 laminectomy and 2 studies did not mention it (Table 3). Of the 13 studies, materials used for the duraplasty varied. In 7 studies, the specific surgical procedure (nonduraplasty or duraplasty) was chosen by each surgeon based on training and personal preference. In 4 studies, the choice of 2 techniques was based on clinical presentation and in 2 studies, it was not mentioned (Table 3). We hence performed a prespecified sensitivity analysis in our study. The outcome of quality assessment for studies included was as follows: 8 studies scored 8, 4 studies scored 7, and 1 study scored 6 (Table 2). In a pooled analysis of 3 trials,14,17,20 the PFDD led to a mean greater increase in operative time than did PFD [standardized mean difference, e2.35; 95% confidence interval [CI], (e2.70 to e1.99); P < 0.05], with no statistically significant betweenstudy heterogeneity (Figure 2). In this analysis, sensitivity analysis showed a stable outcome. Pooled analysis of 4 studies13,15,18,22 that assessed the proportion of patients with

WORLD NEUROSURGERY, https://doi.org/10.1016/j.wneu.2017.10.161

LITERATURE REVIEW WEIWEI LIN ET AL.

PFD AND PFDD IN CM-I: META-ANALYSIS

A

RR (95% CI)

Munshi et al. (2000)

0.78 (0.47–1.28) 29.30

Galarza et al. (2007)

1.25 (0.56–2.77) 14.46

Romero et al. (2010)

0.76 (0.41–1.42) 18.07

Gurbuz et al. (Karaaslan;2015)

0.39 (0.17–0.91) 38.17

Overall (I2 = 27.7%, P = 0.246)

0.70 (0.49–0.98) 100.00

.4

.7

1 Favours PFDD

Weight %

1.7 Favours PFD

B

RR (95% CI)

Weight %

Munshi et al. (2000)

0.92 (0.49–1.72)

19.80

Galarza et al. (2007)

0.45 (0.19–1.07)

40.33

Romero et al. (2010)

1.11 (0.56–2.20)

12.37

Gurbuz et al. (Karaaslan;2015)

1.07 (0.64–1.77)

27.50

Overall (I2 = 25.7%, P = 0.257)

0.80 (0.56–1.13)

100.00

.4

.8 1 Favours PFDD

Figure 3. Meta-analyses of posterior fossa decompression without (PFD) or with duraplasty (PFDD), comparing clinical

syringomyelia achieving clinical improvement after surgery showed a higher likelihood of clinical improvement when patients were operated on with PFDD (absolute risk difference of 25%) compared with PFD (RR, 0.70; 95% CI, 0.49e0.98; P

1.7 Favours PFD

improvement in (A) patients with syringomyelia, (B) patients without syringomyelia (continues)

< 0.05), with no statistically significant between-study heterogeneity (I2 ¼ 27.7%; P ¼ 0.246) (Figure 3). In the sensitivity analysis comparison, a stable outcome was observed. Four studies13,15,18,22 assessed the proportion of patients

WORLD NEUROSURGERY -: ---, MONTH 2017

without syringomyelia achieving clinical improvement after surgery. Pooling the data of these studies showed no significant difference in the RR of clinical improvement, comparing PFDD with PFD (RR, 0.80; 95% CI, 0.56e1.13; P > 0.05),

www.WORLDNEUROSURGERY.org

5

LITERATURE REVIEW WEIWEI LIN ET AL.

PFD AND PFDD IN CM-I: META-ANALYSIS

Weight

C

RR (95% CI)

(M-H)

adult Munshi et al. (2000)

0.84 (0.56–1.24) 10.29

Romero et al. (2010)

0.93 (0.61–1.40) 5.37

Yilmaz et al. (2011)

0.88 (0.71–1.10) 24.21

Gurbuz et al. (2015)

0.39 (0.17–0.91) 8.40

Chen et al. (2017)

0.99 (0.92–1.06) 32.89

M-H Subtotal (I2 = 75.0%, P = 0.003)

0.87 (0.78–0.97) 81.16

D+L Subtotal

0.86 (0.67–1.11)

. child 0.69 (0.40–1.21) 11.06

Galarza et al. (2007)

Erdogan et al. (2010)

1.14 (0.76–1.69) 7.78

M-H Subtotal (I2 = 56.9%, P = 0.128)

0.88 (0.62–1.24) 18.84

D+L Subtotal

0.92 (0.55–1.54)

. M-H Overall (I2 = 66.9%, P = 0.006)

0.87 (0.78–0.97) 100.00

D+L Overall

0.88 (0.72–1.08)

.4

.88

Favours PFDD

1

1.7

Favours PFD

Weight

D

RR (95% CI)

(M-H)

adult Munshi et al. (2000)

0.84 (0.56–1.24) 17.53

Romero et al. (2010)

0.93 (0.61–1.40) 9.15

Yilmaz et al. (2011)

0.88 (0.71–1.10) 41.24

M-H Subtotal (I2 = 0.0%, P = 0.939)

0.88 (0.73–1.05) 67.92

D+L Subtotal

0.88 (0.74–1.05)

. child Galarza et al. (2007)

0.69 (0.40–1.21) 18.84

Erdogan et al. (2010)

1.14 (0.76–1.69) 13.25

M-H Subtotal (I2 = 56.9%, P = 0.128)

0.88 (0.62–1.24) 32.08

D+L Subtotal

0.92 (0.55–1.54)

. M-H Overall (I2 = 0.0%, P = 0.654)

0.88 (0.74–1.03) 100.00

D+L Overall

0.90 (0.77–1.05)

.4

.88

Favours PFDD

Figure 3. (continued). (C) overall patients, and (D) overall patients in further sensitive analysis. For each estimate, the gray shaded area is the weight of the

6

www.SCIENCEDIRECT.com

1

1.7

Favours PFDD

estimate in proportion to the overall effect. CI, confidence interval; RR, relative risk. H-M, Mantel-Haenszel; D-L, DerSimonian-Laird.

WORLD NEUROSURGERY, https://doi.org/10.1016/j.wneu.2017.10.161

LITERATURE REVIEW WEIWEI LIN ET AL.

with no statistically significant betweenstudy heterogeneity (I2 ¼ 25.7%; P ¼ 0.257) (Figure 3). The pooled analysis of overall clinical improvement including 7 studies4,13,15,18,19,21,24 (Figure 3) showed no significant difference in the RR of clinical improvement between PFD and PFDD (RR, 0.88; 95% CI, 0.72e1.08; P > 0.05), with statistically significant between-study heterogeneity (I2 ¼ 66.9%; P ¼ 0.006). The sensitivity analysis comparing PFDD showed no clinical improvement with PFD (RR, 0.88; 95% CI, 0.74e1.03; P > 0.05); we detected no significant between-study heterogeneity (I2 ¼ 0.0%; P ¼ 0.654) (Figure 3). The subgroup analysis of adults showed no significant difference in the RR of overall clinical improvement between PFD and PFDD with no statistically significant between-study heterogeneity. The subgroup analysis of children also showed no significant difference in the RR of clinical improvement between PFD and PFDD, yet with moderate between-study heterogeneity (I2 ¼ 56.9%; P ¼ 0.128). Pooled analysis of 7 studies4,13,17,20,22-24 that assessed the proportion of patients achieving improvement on magnetic resonance imaging (MRI) showed no significant difference in the RR of MRI improvement between PFD and PFDD (RR, 0.86; 95% CI, 0.61e1.22; P > 0.05), with statistically significant between-study heterogeneity (I2 ¼ 64.0%; P ¼ 0.011). The sensitivity analysis and Egger test showed no statistical significance and no publication bias (Appendix 1). Thirteen studies assessed the RR of any complications, including CSF-related complications (CSF leak, aseptic meningitis, pseudomeningocele, and others), wound infection, and overall complications. Pooling the data of 8 studies4,13,18-21,23,24 showed significant difference in the RR of overall complications between the 2 groups in favor of PFD (RR, 0.78; 95% CI, 0.66e0.93; P < 0.05), with no statistically significant betweenstudy heterogeneity (I2 ¼ 7.7%; P ¼ 0.371) (Figure 5). There was a significant difference in the rate of CSFrelated complications (RR, 0.29; 95% CI, 0.15e0.58; P < 0.05), CSF leak (RR, 0.39; 95% CI, 0.15e0.99; P < 0.05), aseptic meningitis (RR, 0.25; 95% CI, 0.08e0.78; P < 0.05), and pseudomeningocele (RR, 0.54; 95% CI, 0.37e0.79; P < 0.05) between the 2 groups in favor of PFD with no

PFD AND PFDD IN CM-I: META-ANALYSIS

statistically significant between-study heterogeneity (Figure 4). On the other hand, there was no significant difference between the 2 groups in the rate of wound infection (RR, 0.85; 95% CI, 0.34e2.10; P > 0.05) and other CSFrelated complications (including subgaleal CSF, seroma collections, subcutaneous hydrops, and so on) (RR, 0.65; 95% CI, 0.20e2.06; P > 0.05), with no statistically significant between-study heterogeneity (Figure 4). Pooled analysis of 7 studies showed that there was a significant difference in the recurrence rate between the 2 groups in favor of PFDD (RR, 3.13; 95% CI, 1.37e7.16; P < 0.05), with statistically significant between-study heterogeneity (I2 ¼ 67.7%; P ¼ 0.009) (Figure 5). In the sensitivity analysis, when the study by Shweikeh et al.23 was removed, PFDD was still associated with a lower risk of recurrence than was PFD, with no significant between-study heterogeneity (Figure 5). We also assessed funnel plot asymmetry using the Egger test (Appendices 2 and 3). DISCUSSION Our results show that there is no significant difference in overall clinical improvement between PFD and PFDD, whereas for patients with syringomyelia, PFDD is better than PFD. It seems that there is no significant difference in imaging improvement between the 2 groups. Furthermore, PFDD led to a greater increase in operative time than did PFD. PFDD is associated with a higher rate of overall complications compared with PFD, especially CSF-related complications. However, compared with PFD, PFDD has a lower recurrence rate, which may decrease the reoperation rate. Generally, neurosurgeons usually consider clinical improvement as the most important priority for surgical decision making. A study by Hao Xu et al.5 showed a significant difference in the (overall) clinical improvement rate between the 2 groups in favor of the PFDD group. Thus, these investigators conclude that although PFDD is related to longer operation time and higher CSF leak rate, it can still be considered as a preferable treatment option for most patients with CM-1 because of its higher improvement rate. However, according to our results,

WORLD NEUROSURGERY -: ---, MONTH 2017

PFDD achieves the same clinical outcomes as PFD for overall patients with CM-1. In the subgroup analysis of age, we found no significant differences between the 2 groups in adult and child patients. Further analysis shows that in patients with syringomyelia, PFDD has more satisfactory outcomes than does PFD in clinical improvement, but for patients without syringomyelia, there is no significant statistical difference between the 2 groups. The imaging improvement in patients with syringomyelia was also analyzed, and we found no significant difference between the 2 surgical groups. Some studies also present no significant relationship between the reduction in syrinx size on MRI and the degree of clinical improvement. According to Förander et al.,7 at the end of follow-up, there were almost similar proportions of (overall) clinical improvement and syringomyelia improvement (on MRI) in both treatment groups. Our study shows more satisfactory results in PFDD than in PFD in clinical improvement; we suspect that the clinical symptoms of CM-1 do not depend only on the syrinx size but also on the nervous structures and the CSF circulation at the craniocervical junction. So, the aim of surgery is to relieve the impact on the nervous structures and to improve CSF circulation. On the other hand, the decrease in syrinx on MRI can be an indicator of sufficient decompression, which can also be seen as a prognostic indicator. Both clinical and imaging improvement should be considered in the choice of surgical option. In terms of complications, our results confirm significant differences in the RR of general complications between the 2 groups in favor of PFD, especially in CSFrelated complications, which mainly include postoperative CSF leak, aseptic meningitis, pseudomeningocele, subgaleal CSF, seroma collections, subcutaneous hydrops, and so on. Those complications may be caused by CSF exposure to blood and the technique of duraplasty, which is why PFD comes with fewer complications. According to Lee et al.,25 dural substitutes generated from porcine collagen, compared with those from bovine collagen, were associated with a higher likelihood of pseudomeningocele development in adult patients undergoing CM-1 decompression and duraplasty. The materials used for

www.WORLDNEUROSURGERY.org

7

LITERATURE REVIEW WEIWEI LIN ET AL.

PFD AND PFDD IN CM-I: META-ANALYSIS

A

% RR (95% CI)

Weight

Munshi et al. (2000)

0.40 (0.02–7.69)

10.43

McGirt et al. (2008)

0.24 (0.01–4.97)

14.19

Erdogan et al. (2010)

0.41 (0.02–9.25)

8.42

Romero et al. (2010)

0.31 (0.02–5.63)

12.17

Yilmaz et al. (2011)

0.34 (0.02–6.29)

13.04

Lee et al. (2014)

0.14 (0.01–2.45)

25.22

Gurbuz et al. (Karaaslan;2015)

0.23 (0.01–4.53)

14.50

Chen et al. (2017)

6.26 (0.26–149.80)

Overall (I2 = 0.0%, P = 0.814)

0.39 (0.15–0.99)

.1 Favours PFD

.39

1

100.00

Favours PFDD

B

% RR (95% CI)

Weight

Munshi (2000)

0.67 (0.03–15.17)

9.09

McGirt (2008)

0.24 (0.01–4.97)

20.61

Romero (2010)

0.52 (0.02–11.14)

10.60

Yilmaz (2011)

0.79 (0.03–18.66)

8.11

Lee (2014)

0.03 (0.00–0.57)

51.58

Overall (I2 = 0.0%, P = 0.552)

0.25 (0.08–0.78)

100.00

.01

.25 Favours PFD

Figure 4. Meta-analyses of posterior fossa decompression without (PFD) or with duraplasty (PFDD), comparing

8

2.03

www.SCIENCEDIRECT.com

1

10

Favours PFDD

cerebrospinal fluid (CSF)-related complications. Outcomes assessed are (A) CSF leak, (B) aseptic meningitis (continues)

WORLD NEUROSURGERY, https://doi.org/10.1016/j.wneu.2017.10.161

LITERATURE REVIEW WEIWEI LIN ET AL.

PFD AND PFDD IN CM-I: META-ANALYSIS

C

%

RR (95% CI)

Weight

Mutchnick et al. (2010)

0.23 (0.01–4.65)

3.29

Gurbuz et al. (Karaaslan;2015)

0.39 (0.02–8.93)

1.96

Shweikeh et al. (2015)

0.56 (0.38–0.81)

94.76

Overall (I2 = 0.0%, P = 0.827)

0.54 (0.37–0.79)

100.00

.012

Favours PFD

.54

1

Favours PFDD

80

D

% RR (95% CI)

Weight

Munshi et al. (2000)

0.15 (0.01–2.40) 12.48

Erdogan et al. (2010)

0.18 (0.01–3.11) 8.79

Romero et al. (2010)

0.17 (0.01–2.77) 9.80

Yilmaz et al. (2011)

0.26 (0.01–4.69) 7.50

Lee et al. (2014)

0.08 (0.00–1.38) 18.81

Gurbuz et al. (2015)

0.22 (0.01–4.08) 6.74

Chen et al. (2017)

0.53 (0.22–1.29) 35.88

Overall (I2 = 0.0%, P = 0.803)

0.29 (0.15–0.58) 100.00

.005

.29

Favours PFD

Figure 4. (continued). (C) pseudomeningocele, and (D) CSF-related complications. For each estimate, the gray shaded

WORLD NEUROSURGERY -: ---, MONTH 2017

1

Favours PFDD

80

area is the weight of the estimate in proportion to the overall effect. CI, confidence interval; RR, relative risk.

www.WORLDNEUROSURGERY.org

9

LITERATURE REVIEW WEIWEI LIN ET AL.

PFD AND PFDD IN CM-I: META-ANALYSIS

A

% RR (95% CI)

Weight

Munshi et al. (2000)

0.70 (0.08–5.96)

17.00

McGirt et al. (2008)

0.40 (0.04–3.82)

23.81

Romero et al. (2010)

1.67 (0.13–22.00) 6.57

Yilmaz et al. (2011)

2.42 (0.16–37.08) 5.13

Lee et al. (2014)

3.70 (0.16–87.58) 3.92

Gurbuz et al. (Karaasian;2015)

0.29 (0.04–2.38)

Chen et al. (2017)

1.06 (0.10–11.28) 11.23

Overall (I2 = 0.0%, P = 0.789)

0.81 (0.35–1.87)

.012

Favours PFD

.811

100.00

80

Favours PFDD

B

% RR (95% CI)

Weight

Munshi et al. (2000)

0.21 (0.03–1.43)

2.64

Erdogan et al. (2010)

0.18 (0.01–3.11)

1.28

Romero et al. (2010)

0.33 (0.05–2.21)

1.53

Yilmaz et al. (2011)

0.69 (0.15–3.09)

1.67

Lee et al. (2014)

0.18 (0.02–1.36)

2.55

Gurbuz et al. (2015)

0.27 (0.04–2.10)

1.57

Shweikeh et al. (2015)

0.85 (0.71–1.03)

82.76

Chen et al. (2017)

0.74 (0.37–1.47)

6.01

Overall (I2

0.78 (0.66–0.93)

100.00

= 7.7%, P = 0.371)

.012

Favours PFD

Figure 5. Meta-analyses of posterior fossa decompression without (PFD) or with duraplasty (PFDD), comparing

10

32.34

www.SCIENCEDIRECT.com

1

Favours PFDD

80

complications and recurrence. Outcomes assessed are (A) wound infection, (B) overall complications. (continues)

WORLD NEUROSURGERY, https://doi.org/10.1016/j.wneu.2017.10.161

LITERATURE REVIEW WEIWEI LIN ET AL.

PFD AND PFDD IN CM-I: META-ANALYSIS

%

C

Weight RR (95% CI)

(M-H)

Mutchnick et al. (2010)

4.00 (0.87–18.47) 5.72

Munshi et al. (2000)

6.00 (0.26–136.52)1.02

Lee et al. (2014)

2.48 (0.24–26.04) 2.73

Gurbuz et al. (Karaaslan;2015)

1.56 (0.40–6.05)

8.48

Shweikeh et al. (2015)

0.36 (0.18–0.73)

81.05

Chen et al. (2017)

10.44 (0.52–211.55) 0.99

M-H Overall (I = 67.7%, P = 0.009)

0.89 (0.56–1.41)

D+L Overall

1.90 (0.57–6.36)

2

.012

Favours PFD1

Favours PFDD

100.00

80

%

D

Weight RR (95% CI)

(M-H)

Mutchnick et al. (2010)

4.00 (0.87–18.47)

30.18

Munshi et al. (2000)

6.00 (0.26–136.52) 5.39

Lee et al. (2014)

2.48 (0.24–26.04)

14.43

Gurbuz et al. (Karaaslan;2015)

1.56 (0.40–6.05)

44.77

Chen et al. (2017)

10.44 (0.52–211.55)5.24

M-H Overall (I2 = 0.0%, P = 0.747)

3.13 (1.37–7.16)

D+L Overall

2.87 (1.22–6.77)

.012

Favours PFD 1

Figure 5. (continued). (C) recurrence rate, and (D) recurrence rate in the sensitive analysis. For each estimate, the gray shaded area is the weight of the estimate in proportion to the

WORLD NEUROSURGERY -: ---, MONTH 2017

Favours PFDD

100.00

80

overall effect. CI, confidence interval; RR, relative risk; H-M, Mantel-Haenszel; D-L, DerSimonian-Laird.

www.WORLDNEUROSURGERY.org

11

www.SCIENCEDIRECT.com

Surgical Technique Removal of All Dural Scarring or Bands on the Outside of the Dura or Others (PFD)

Materials Used for the Duraplasty (PFDD)

Positioning of Patient

Extent of Bone Decompression (PFD and PFDD)

Munshi et al.13

2000

Prone

At least 2 cm above the foramen magnum, with bilateral removal of the C1 laminae

Removal of all dural scarring or bands on the outside of the dura

Cadaveric dura, bovine pericardium, fascia lata, or autologous pericranium

1. Chosen by each surgeon on the basis of training and personal preference 2. No surgeon performed both procedures in this series

Limonadi and Seldon15

2004

Prone

A small suboccipital and foramen magnum craniectomy, removal of the C1 arch

Dura-splitting

Allograft

1. Patients with symptomatic Chiari I malformation and syringomyelia: PFDD 2. Patients with symptomatic Chiari I malformation and no syringomyelia: PFD

Galarza et al.15

2007

Prone

A limited occipital craniectomy, removal of the C1 arch (mostly)

The outer layer of the dura was incised

Bovine pericardium

McGirt et al.16

2008

Prone

A small suboccipital craniectomy and a C1 laminectomy

NA

Surgeon-specific (B.C.: pericranial autograft; G.J.: Gortex, W. L. Gore and associates; and J.W.: Dura-guard, Bio-Vascular, Inc.).

Each surgeon’s subjective interpretation of evidence of adequate bone hindbrain decompression on intraoperative ultrasonography

Erdogan et al.4

2010

Prone

At least 3 cm above the foramen magnum with a width of 3 cm, with C1 laminectomy

Atlanto-occipital ligament and dural scarring or bands on the outside of the dura was removed

Cadaveric dura

The specific surgical procedure was chosen randomly

Muchnik et al.17

2010

NA

NA

NA

NA

1. Patients with CM and syrinx/the family‘s strong preference for PFDD/intraoperative observations requiring duraplasty: PFDD 2. Patients without a syrinx but with symptoms clearly attributable to a radiographically proven CM: PFD

Romero et al.18

2010

Prone

At least 2 cm above the foramen magnum, with bilateral removal of the C1 laminae

Removal of all dural scarring or bands on the outside of the dura

Fascia lata

Chosen by each surgeon on the basis of training and personal preference

Choice of the Two Techniques

The surgeon’s particular preference

LITERATURE REVIEW

PFD AND PFDD IN CM-I: META-ANALYSIS

WORLD NEUROSURGERY, https://doi.org/10.1016/j.wneu.2017.10.161

Year

WEIWEI LIN ET AL.

12

Table 3. The Information of Included Studies About the Choice of the Two Surgical Techniques and Surgical Technique (the Extent of Bone Decompression and Materials Used for the Duraplasty)

LITERATURE REVIEW

Chosen by each surgeon on the basis of training and personal preference. No surgeon performed both procedures in this series 2017 Chen et al.24

WORLD NEUROSURGERY -: ---, MONTH 2017

PFD, posterior fossa decompression; PFDD, posterior fossa decompression with duraplasty; NA, not available; FMD, foramen magnum decompression.

NA NA

Occipital fascia or artificial dura The thick fasciculationlike tissue was removed

NA NA

The inferior part of the occipital bone, the posterior lamina of C1, and the tip of the spinous process of C2 were removed.(approximately 4 cm  4 cm)

2015 Shweikeh et al.23

NA

2015 Gurbuz et al.22

Prone

NA NA NA At least 3 cm above the foramen magnum with a width of at least 4 cm, and total C1 laminectomy

2015 Gurbuz et al.21

NA

Criterion for selecting duraplasty in addition to FMD was the presence of a relatively aggressive, large-sized syringomyelia Galeal autograft NA At least 3 cm above the foramen magnum with a width of at least 4 cm, and a C1 laminectomy

2014 Lee et al.20

Prone

2011

Prone

From the foramen magnum to the inferior nuchal line, and C1 laminectomy

Intraoperative ultrasonography was used before and after lysis of the atlanto-occipital membrane to ensure adequate decompression

Durepair dura, regeneration matrix, or a pericranial graft

Although treatment decisions varied by patient, the choice to perform PFD most often resulted from of a comprehensive shift in the senior author’s practice

PFD AND PFDD IN CM-I: META-ANALYSIS

Yilmaz et al.19

Prone

Modest superior extension (approximately 1.5e2.0 cm) and lateral extension to the lateralmost aspect of the foramen magnum and cervical spinal canal, with C1 laminectomy (and C2 if necessary)

Only bone removal was performed

Cadaveric dura, bovine pericardium, fascia lata, or autologous pericranium

Chosen based on the surgeon’s experience and preference

WEIWEI LIN ET AL.

duraplasty may influence surgical outcomes and complication rate after PFDD. The materials used for the duraplasty (Table 3) in our involved studies are so varied that we cannot perform a metaanalysis to establish the influence of different dura materials. If we found a material to decrease the rate of complications, we might have more reasons and confidence to choose PFDD. As mentioned earlier, compared with the PFDD group, the PFD group has a shorter operative time, which may also reduce the rate of complications, especially wound infection. Furthermore, surgery with fewer complications may even reduce the reoperation rate. In addition, some researchers have reported that the PFD group has shorter hospital stay lengths and lower hospital costs.14,23,26 In our opinion, in patients with rapidly progressive symptoms, syringomyelia, or severe neurologic deficits, the optimal surgery strategy may be PFDD. On the other hand, for patients with mild symptoms and neurologic deficits, or no syringomyelia, PFD could be a better choice. Patients with syringomyelia may have a higher likelihood of improvement after undergoing duraplasty.13,18 Our study shows that for patients with syringomyeli, PFDD has better outcomes in clinical improvement but it has higher rate of complications. So, which is the optimal surgery strategy? According to Iskandar et al.,27 the impaction of the cerebellar tonsils in the cisterna magna, without herniation into the cervical spinal canal also causes disturbances of the CSF flow to the foramen magnum and can provoke neurologic signs and symptoms by compression of the brainstem and the cerebellar pathways. A study by Yilmaz et al.19 shows that PFD and duraplasty for the treatment of CTD grade 3 Chiari malformation may lead to a more reliable reduction in the volume of concomitant syringomyelia and Japanese Orthopaedic Association scores. In patients with CTD grade 1 and 2, PFD without duraplasty may be performed. According to the theory of Gardner and Angel,18 the obstruction of CSF flow at the craniocervical junction causes CSF to enter the cervical central canal. The aim of decompression surgery is to restore normal CSF dynamics at the foramen magnum.14,28-30 Some studies also show

www.WORLDNEUROSURGERY.org

13

LITERATURE REVIEW WEIWEI LIN ET AL.

that intraoperative ultrasonography can effectively guide the decision making of the surgical strategy.14,16,31 More objective CSF flow or volumetric measures may be needed intraoperatively to guide duraplasty in patients with more pronounced tonsillar herniation.16 Thus, the Tonsillar Descending Grading Scale should be considered in the decision making, and ultrasonography plays an important role in this process. A limitation of this analysis is the insufficient number of studies to make a metaanalysis and establish the relationship between CTD grade and outcomes, and the choice of duraplasty material. The studies that were included in the meta-analysis applied different techniques by different choices, which might have led to performance bias. Similarly, follow-up periods ranged broadly from several months to 10 years. However, after sensitive analysis, I2 values decreased to less than 50%, indicating a robust analysis and a reliable result. Second, most of the studies included were published in English, and we do not have enough data about studies reported in other languages. According to our subgroup analysis, race may be a factor affecting between-study heterogeneity. Thus, a further analysis is necessary to discuss the relationship between race and outcomes of surgery for CM-1. All our studies included are nonrandomized controlled trials; future randomized studies with large numbers should be the subjects of meta-analysis to provide illumination in surgical decision making in CM-1. CONCLUSIONS PFDD can be an optimal surgical strategy because of its higher clinical improvement and lower recurrence rate in patients with syringomyelia, despite its longer operative time and increased CSF-related complications. In patients without syringomyelia, PFD can be a preferred choice for its similar clinical improvement and lower costs. The Tonsillar Descending Grading Scale, the choice of duraplasty material, and ultrasonography should be considered in the process. There is a shortage of highquality studies that offer guidance in the choice of decompressive technique in adult patients with CM-1. Future randomized studies with large numbers and the power to provide illumination in

14

www.SCIENCEDIRECT.com

PFD AND PFDD IN CM-I: META-ANALYSIS

surgical decision making in CM-1 are warranted.

12. Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med. 2002;21: 1539-1558.

ACKNOWLEDGMENTS

13. Munshi I, Frim D, Stine-Reyes R, Weir BK, Hekmatpanah J, Brown F. Effects of posterior fossa decompression with and without duraplasty on Chiari malformation-associated hydromyelia. Neurosurgery. 2000;46:1384-1389 [discussion 13891390].

W.L. had the idea for the study. W.L., G.D., and J.X. selected studies for inclusion and abstracted data. W.L. performed the statistical analyses. W.L., G.D., J.X., and J.S. interpreted the data. W.L., G.D., and J.X. wrote the first draft. W.L., G.D., Z.W., and J.X. critically revised the article for important intellectual content. All authors approved the final draft. REFERENCES 1. Zhao JL, Li MH, Wang CL, Meng W. A systematic review of Chiari I malformation: techniques and outcomes. World Neurosurg. 2016;88:7-14. 2. Isu T, Iwasaki Y, Akino M, Abe H. Hydrosyringomyelia associated with a Chiari I malformation in children and adolescents. Neurosurgery. 1990;26:591-596 [discussion: 596-597]. 3. Tubbs RS, Smyth MD, Wellons JC 3rd, Oakes WJ. Arachnoid veils and the Chiari I malformation. J Neurosurg. 2004;100(5 suppl pediatrics):465-467. 4. Erdogan E, Cansever T, Secer HI, Temiz C, Sirin S, Kabatas S, et al. The evaluation of surgical treatment options in the Chiari Malformation Type I. Turk Neurosurg. 2010;20:303-313. 5. Xu H, Chu L, He R, Ge C, Lei T. Posterior fossa decompression with and without duraplasty for the treatment of Chiari malformation type Iea systematic review and meta-analysis. Neurosurg Rev. 2017;40:213-221. 6. Hoffman CE, Souweidane MM. Cerebrospinal fluid-related complications with autologous duraplasty and arachnoid sparing in type I Chiari malformation. Neurosurgery. 2008;62(3 suppl 1): 156-160 [discussion: 160-161]. 7. Forander P, Sjavik K, Solheim O, Riphagen I, Gulati S, Salvesen O, et al. The case for duraplasty in adults undergoing posterior fossa decompression for Chiari I malformation: a systematic review and meta-analysis of observational studies. Clin Neurol Neurosurg. 2014;125:58-64. 8. Moher D, Liberati A, Tetzlaff J, Altman DG, Group P. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med. 2009;6:e1000097. 9. Stang A. Critical evaluation of the NewcastleOttawa scale for the assessment of the quality of nonrandomized studies in meta-analyses. Eur J Epidemiol. 2010;25:603-605. 10. Duval S, Tweedie R. Trim and fill: a simple funnel-plot-based method of testing and adjusting for publication bias in meta-analysis. Biometrics. 2000;56:455-463. 11. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ. 2003;327:557-560.

14. Limonadi FM, Selden NR. Dura-splitting decompression of the craniocervical junction: reduced operative time, hospital stay, and cost with equivalent early outcome. J Neurosurg. 2004;101(2 suppl):184-188. 15. Galarza M, Sood S, Ham S. Relevance of surgical strategies for the management of pediatric Chiari type I malformation. Childs Nerv Syst. 2007;23: 691-696. 16. McGirt MJ, Attenello FJ, Datoo G, Gathinji M, Atiba A, Weingart JD, et al. Intraoperative ultrasonography as a guide to patient selection for duraplasty after suboccipital decompression in children with Chiari malformation type I. J Neurosurg Pediatr. 2008;2:52-57. 17. Mutchnick IS, Janjua RM, Moeller K, Moriarty TM. Decompression of Chiari malformation with and without duraplasty: morbidity versus recurrence. J Neurosurg Pediatr. 2010;5:474-478. 18. Romero FR, Pereira CA. Suboccipital craniectomy with or without duraplasty: what is the best choice in patients with Chiari type 1 malformation? Arq Neuropsiquiatr. 2010;68:623-626. 19. Yilmaz A, Kanat A, Musluman AM, Colak I, Terzi Y, Kayaci S, et al. When is duraplasty required in the surgical treatment of Chiari malformation type I based on tonsillar descending grading scale? World Neurosurg. 2011;75:307-313. 20. Lee A, Yarbrough CK, Greenberg JK, Barber J, Limbrick DD, Smyth MD. Comparison of posterior fossa decompression with or without duraplasty in children with Type I Chiari malformation. Childs Nerv Syst. 2014;30:1419-1424. 21. Gurbuz MS, Berkman MZ, Unal E, Akpinar E, Gok S, Orakdogen M, et al. Foramen magnum decompression and duraplasty is superior to only foramen magnum decompression in Chiari malformation type 1 associated with syringomyelia in adults. Asian Spine J. 2015;9:721-727. 22. Gurbuz MS, Karaaslan N, Caliskan T, Unal E, Berkman MZ. Comparison of the surgical results for foramen magnum decompression with and without duraplasty in Chiari malformation type 1. Turk Neurosurg. 2015;25:419-424. 23. Shweikeh F, Sunjaya D, Nuno M, Drazin D, Adamo MA. National trends, complications, and hospital charges in pediatric patients with Chiari malformation type I treated with posterior fossa decompression with and without duraplasty. Pediatr Neurosurg. 2015;50:31-37. 24. Chen J, Li Y, Wang T, Gao J, Xu J, Lai R, et al. Comparison of posterior fossa decompression with and without duraplasty for the surgical treatment of Chiari malformation type I in adult

WORLD NEUROSURGERY, https://doi.org/10.1016/j.wneu.2017.10.161

LITERATURE REVIEW WEIWEI LIN ET AL.

patients: a retrospective analysis of 103 patients. Medicine. 2017;96:e5945. 25. Lee CK, Mokhtari T, Connolly ID, Li G, Shuer LM, Chang SD, et al. Comparison of porcine and bovine collagen dural substitutes in posterior fossa decompression for Chiari I malformation in adults [e-pub ahead of print]. World Neurosurg. 2017;108:33-40. 26. Chotai S, Medhkour A. Surgical outcomes after posterior fossa decompression with and without duraplasty in Chiari malformation-I. Clin Neurol Neurosurg. 2014;125:182-188. 27. Iskandar BJ, Hedlund GL, Grabb PA, Oakes WJ. The resolution of syringohydromyelia without hindbrain herniation after posterior fossa decompression. J Neurosurg. 1998;89:212-216.

PFD AND PFDD IN CM-I: META-ANALYSIS

28. Fukushima T, Matsuda T, Tsuchimochi H, Yamamoto M, Tsugu H, Tomonaga M, et al. Symptomatic Chiari malformation and associated pathophysiology in pediatric and adult patients without myelodysplasia. Neurol Med Chir (Tokyo). 1994;34:738-743. 29. Panigrahi M, Reddy BP, Reddy AK, Reddy JJ. CSF flow study in Chiari I malformation. Child Nerv Syst. 2004;20:336-340.

invasive surgical technique in patients with Chiari type I malformation. World Neurosurg. 2017;101: 466-475.

Conflict of interest statement: The authors declare that the article content was composed in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Received 14 July 2017; accepted 28 October 2017

30. Bond AE, Jane JA Sr, Liu KC, Oldfield EH. Changes in cerebrospinal fluid flow assessed using intraoperative MRI during posterior fossa decompression for Chiari malformation. J Neurosurg. 2015;122:1068-1075.

Citation: World Neurosurg. (2017). https://doi.org/10.1016/j.wneu.2017.10.161

31. Brock RS, Taricco MA, de Oliveira MF, de Lima Oliveira M, Teixeira MJ, Bor-Seng-Shu E. Intraoperative ultrasonography for definition of less

1878-8750/$ - see front matter ª 2017 Elsevier Inc. All rights reserved.

WORLD NEUROSURGERY -: ---, MONTH 2017

Journal homepage: www.WORLDNEUROSURGERY.org Available online: www.sciencedirect.com

www.WORLDNEUROSURGERY.org

15

LITERATURE REVIEW WEIWEI LIN ET AL.

PFD AND PFDD IN CM-I: META-ANALYSIS

% Weight RR (95% CI)

(M -H)

Munshi et al. (2000)

0.53 (0.25–1.12)12.75

Galarza et al. (2007)

2.50 (0.17–37.26)1.09

Erdogan et al. (2010)

1.03 (0.72–1.48)10.06

Romero et al. (2010)

0.55 (0.22–1.35)8.15

Yilmaz et al. (2011)

0.92 (0.75–1.15)39.67

Lee et al. (2014)

1.26 (0.90–1.77)13.39

Gurbuz et al. (Karaaslan;2015)

0.14 (0.02–0.85)14.90

M -H Overall (I2 = 64.0%, P = 0.011)

0.80 (0.66–0.97)100.00

D+L Overall

0.86 (0.61–1.22)

0.025

Favours PFDD

0.81

Favours PFD

40

Appendix 1.1. Meta-analyses of posterior fossa decompression without (PFD) or with duraplasty (PFDD), comparing imaging improvement. CI, confidence interval; RR, relative risk.

Meta-analysis estimates, given named study is omitted Lower CI Limit Estimate Upper CI Limit Munshi et al. (2000)

Galarza et al. (2007)

Erdogan et al. (2010)

Romero et al. (2010)

Yilmaz et al. (2011)

Lee et al. (2014)

Gurbuz et al. (Karaaslan;2015) 0.43

0.61

0.86

1.22

1.36

Appendix 1.2. Sensitive analysis of posterior fossa decompression without (PFD) or with duraplasty (PFDD), comparing imaging improvement. CI, confidence interval.

15.E1

www.SCIENCEDIRECT.com

WORLD NEUROSURGERY, https://doi.org/10.1016/j.wneu.2017.10.161

LITERATURE REVIEW WEIWEI LIN ET AL.

PFD AND PFDD IN CM-I: META-ANALYSIS

Number of studies = Std_Eff slope bias

7

Root MSE

Coefficient Standard Error 0.1264857 –1.010428

0.1790745 0.8518411

Test of H0: no small-study effects

t 0.71 –1.19

P>|t| 0.512 0.289

=

1.336

[95% Confidence Interval] –0.3338401 –3.200156

0.5868115 1.179299

P = 0.289

Appendix 1.3. Egger test of posterior fossa decompression without (PFD) or with duraplasty (PFDD), comparing imaging improvement in the sensitive analysis. It showed no publication bias (P > 0.1). MSE, mean squared error; Std_Eff, standardized effect; H0, null hypothesis.

Appendix 2.1. Sensitive analysis and Egger test of posterior fossa decompression without (PFD) or with duraplasty (PFDD), comparing cerebrospinal fluid leak rate, which showed publication bias (P < 0.1). CI, confidence interval; MSE, mean squared error; Std_Eff, standardized effect; H0, null hypothesis.

WORLD NEUROSURGERY -: ---, MONTH 2017

www.WORLDNEUROSURGERY.org

15.E2

LITERATURE REVIEW WEIWEI LIN ET AL.

PFD AND PFDD IN CM-I: META-ANALYSIS

Appendix 2.2. Sensitive analysis and Egger test of posterior fossa decompression without (PFD) or with duraplasty (PFDD), comparing cerebrospinal fluiderelated complications, which showed publication bias (P < 0.1). CI, confidence interval; MSE, mean squared error; Std_Eff, standardized effect; H0, null hypothesis.

15.E3

www.SCIENCEDIRECT.com

WORLD NEUROSURGERY, https://doi.org/10.1016/j.wneu.2017.10.161

LITERATURE REVIEW WEIWEI LIN ET AL.

PFD AND PFDD IN CM-I: META-ANALYSIS

Appendix 3.1. Sensitive analysis and Egger test of posterior fossa decompression without (PFD) or with duraplasty (PFDD), comparing wound infections rate, which showed a publication bias (P < 0.1). CI, confidence interval; MSE, mean squared error; Std_Eff, standardized effect; H0, null hypothesis.

WORLD NEUROSURGERY -: ---, MONTH 2017

www.WORLDNEUROSURGERY.org

15.E4

LITERATURE REVIEW WEIWEI LIN ET AL.

PFD AND PFDD IN CM-I: META-ANALYSIS

Appendix 3.2. Sensitive analysis and Egger test of posterior fossa decompression without (PFD) or with duraplasty (PFDD), comparing overall complications rate, which showed publication bias (P < 0.1). CI, confidence interval; MSE, mean squared error; Std_Eff, standardized effect; H0, null hypothesis.

15.E5

www.SCIENCEDIRECT.com

WORLD NEUROSURGERY, https://doi.org/10.1016/j.wneu.2017.10.161