Keywords
Chiari 1 malformation - duraplasty - foramen magnum decompression - posterior fossa
decompression - syringomyelia
Introduction
Chiari malformations, first described by Hans Chiari in 1891, are a group of defects
associated with the displacement of the cerebellar tonsils through the foramen magnum
into the upper cervical spinal canal.[1] The most commonly observed type is Chiari 1 malformation, which presents with symptoms
such as headaches, neck pain, limb paraesthesia, and, less commonly, visual changes,
unsteadiness, and syncope.[2] Surgical decompression is considered the gold standard of management; however, its
success rate remains poorly defined.
Two techniques are generally employed for patients offered surgery: foramen magnum
decompression with duraplasty (FMDD) and foramen magnum decompression without duraplasty
(FMD). FMDD involves removal of the posterior arch of C1, widening of the foramen
magnum, and release of the dura to open up space, followed by dural patching using
autologous tissue or other materials. FMD is identical except that only the outermost
layers of the dura are released.
FMDD is associated with longer operative times and higher complication rates but lower
reoperation rates compared to FMD.[3] Despite ongoing research, evidence regarding the impact of these techniques on overall
symptom improvement and syrinx reduction remains unclear. This retrospective analysis
aims to determine the differences in clinical improvement and syrinx reduction between
FMDD and FMD in patients with Chiari 1 malformation at a large tertiary referral centre
in Aotearoa New Zealand (AoNZ).
Materials and Methods
Institutional review board approval for the design and conduct of this study was sought
and granted by the Waikato Hospital Ethics Committee (number RD024049). Informed consent
from patients was not required due to the retrospective design of this study.
Study Design and Setting
This retrospective cohort study was conducted at Waikato Hospital (Te Whatu Ora Waikato,
Hamilton, AoNZ), which is a large tertiary-level and academic teaching institution
comprising 759 inpatient beds, and provides complex subspecialty neurosurgical care
to approximately 1 million patients in the Waikato and Midlands region of AoNZ.
Inclusion Criteria
All patients who underwent operative management for Chiari 1 malformation between
January 1, 2010 and December 31, 2022 at Waikato Hospital were eligible for inclusion.
Only symptomatic patients with Chiari 1 malformation, as evidenced by cerebellar tonsillar
descent of 5 mm below the McRae line, with or without concurrent syrinx, or asymptomatic
patients with concurrent syrinx with at least 4 mm cerebellar tonsillar descent, were
included. Exclusion criteria included patients with other types of Chiari malformation
and those who were transferred to other centers or migrated out-of-district and were
thus lost to follow-up.
Relevant patients were identified from both electronic and paper medical records of
all neurosurgical procedures pertaining to foramen magnum decompression and other
synonymous terms, such as posterior fossa decompression, completed at Waikato Hospital
during the specified study interval.
Operative Protocol
Patients were assessed clinically for symptoms and radiological evidence of Chiari
1 malformation with or without concurrent syrinx. Radiological assessment consisted
of the presence of cerebellar descent and concurrent syrinx. Among asymptomatic patients,
only those with radiological syrinx were offered surgery. Patients presenting closer
to the establishment of the service underwent FMDD, while those presenting more recently
underwent FMD. Other factors such as symptom severity and radiological findings did
not contribute to the choice of procedure offered to patients. Furthermore, some patients
initially planned for FMD underwent FMDD perioperatively due to inadequate decompression.
During this period, patients were followed up by a brain/spine magnetic resonance
imaging (MRI) between 3 and 6 months after surgery. In order to capture patients undergoing
revisional surgery, follow-up was undertaken until 2 years.
Data Extraction
Extracted data included information on basic patient demographics (age at time of
surgery, sex, and ethnicity), preoperative symptoms, radiological findings (e.g.,
presence or absence of syrinx preoperatively, syrinx reduction postoperatively, and
extent of cerebellar tonsillar herniation), as well as postoperative complications.
Ethnicity were self-identified by patients and recorded at the time of operation,
according to the AoNZ Ministry of Health ethnicity data collection protocols.[4] When more than one ethnicity was recorded, reporting was prioritized in the following
order: indigenous Māori, Pacific peoples, Asian, and European/Other. For comparisons,
patients were categorized into indigenous Māori and non-Māori strata.
Outcome Measures
The primary outcome was symptom improvement postoperatively. In the preoperative setting,
each symptom was scored as 0 (absent) or 1 (present), and at 6-month postoperative
follow-up as 0 (absent), 1 (present with no improvement), or 2 (present with improvement).
Secondary outcomes included syrinx reduction at 6-month postoperative follow-up, postoperative
length of stay (LOS), and overall as well as specific complications. Postoperative
complications comprised wound infection, presence of pseudomeningocele on MRI, and
the need for reoperation.
Statistical Analysis
All statistical analyses were conducted using RStudio (version 4.4.3; R Foundation
for Statistical Computing, Vienna, Austria). Categorical variables were reported as
frequencies (n) and percentage (%), while continuous variables were reported as the median and interquartile
range for nonparametric data and mean ± standard deviation for normally distributed
(parametric) data. Normality of data was investigated with the Shapiro–Wilk test.
Continuous data were compared with the independent samples' t-test (for normally distributed variables) and Mann–Whitney U test (for nonparametric variables) and reported as the mean difference (MD). Categorical
data were compared with the chi-squared test. Univariate and multivariate logistic
regression analyses were conducted to assess for independent predictive factors of
symptom improvement and syrinx reduction at 6-month postoperative follow-up. All explanatory
variables with probability (p) values of ≤ 0.15 on univariable analyses were included in the multivariable logistic
regression model in a forward stepwise approach. Overall effect estimates were presented
as an odds ratio (OR) for univariate analyses, and adjusted OR (aOR) for multivariate
analyses, with 95% confidence interval (CI). Effect estimates were considered statistically
significant if the 95% CI did not cross the no-effect line of 1 for OR and 0 for MD.[5]
Results
Baseline Demographics
A total of 102 patients were identified, from which 6 patients were excluded as they
underwent surgery for other types of Chiari malformation or were lost to follow-up
and had incomplete data due to migration outside the Waikato Hospital region. Therefore,
96 patients underwent foramen magnum decompression over the 12-year study period,
of which 58 (60.4%) patients underwent FMDD while 38 (39.6%) patients underwent FMD
([Table 1]).
Table 1
Baseline characteristics between patients who underwent foramen magnum decompression
with versus without duraplasty
|
Characteristic
|
Total cohort (n = 96)
|
FMDD (n = 58)
|
FMD (n = 38)
|
|
Age at surgery (y), median (IQR)
|
31 (23–47.25)
|
32.5 (23–47.75)
|
31 (23.25–45.25)
|
|
Sex, n (%)
|
|
|
|
|
Male
|
25 (26)
|
14 (24)
|
11 (29)
|
|
Female
|
71 (74)
|
44 (76)
|
27 (71)
|
|
Ethnicity, n (%)
|
|
|
|
|
Indigenous Māori
|
38 (40)
|
24 (41)
|
14 (37)
|
|
Non-Māori
|
58 (60)
|
34 (59)
|
24 (63)
|
|
Preoperative symptoms, n (%)
|
|
|
|
|
Headache
|
77 (80)
|
45 (78)
|
32 (84)
|
|
Neck pain
|
25 (26)
|
17 (29)
|
8 (21)
|
|
Limb weakness or paraesthesia
|
66 (69)
|
44 (76)
|
22 (58)
|
|
Visual symptoms
|
21 (22)
|
8 (14)
|
13 (34)
|
|
Dizziness or vertigo
|
14 (15)
|
9 (16)
|
5 (13)
|
|
Nausea and vomiting
|
8 (8)
|
1 (2)
|
7 (18)
|
|
Syncope
|
10 (10)
|
4 (7)
|
6 (16)
|
|
Tinnitus
|
5 (5)
|
1 (2)
|
4 (11)
|
|
Unsteadiness
|
14 (14)
|
9 (16)
|
5 (13)
|
|
Speech or swallowing difficulty
|
8 (8)
|
4 (7)
|
4 (11)
|
|
Presence of syrinx, n (%)
|
47 (49)
|
30 (52)
|
17 (45)
|
|
Extent of cerebellar tonsillar herniation (mm), median (IQR)
|
10 (7–12)
|
10 (7–12)
|
8 (6.1–10)
|
Abbreviations: FMD, foramen magnum decompression without duraplasty; FMDD, foramen
magnum decompression with duraplasty; IQR, interquartile range.
The median age of patients presenting for surgical decompression was 31 years. There
was a female preponderance (74%), although self-identified ethnicity was similar among
the two groups. Headaches were the most common symptom reported by patients, with
almost 80% of patients from each group having these symptoms. This was followed by
limb symptoms, including weakness and paraesthesia (69%), followed by neck pain (26%).
Radiologically, almost half of patients had concurrent syrinx on MRI prior to surgery.
Primary Outcome
Overall Symptom Improvement
Overall, 79 (82%) patients undergoing foramen magnum decompression reported symptom
improvement at 6-month follow-up. Results of the univariate and multivariate regression
analyses are presented in [Table 2]. No difference in overall symptom improvement was found between those undergoing
FMDD and FMD on univariate (OR 0.72, 95% CI: 0.18–4.74) or multivariate analysis (aOR
0.96, 95% CI: 0.24–3.80). Patients who were symptomatic with headaches were more likely
to experience symptom improvement (aOR 4.83, 95% CI: 1.16–20.05), while those reporting
neck pain (aOR 0.23, 95% CI: 0.06–0.84) and unsteadiness (aOR 0.18, 95% CI: 0.04–0.81)
were less likely to report symptomatic improvement.
Table 2
Univariate and multivariate regression analysis for symptom improvement
|
Variable
|
Univariate analysis
|
Multivariate analysis[b]
|
|
OR (95% CI)
|
p-Value
|
OR (95% CI)
|
|
Demographics
|
|
|
|
|
Age at surgery
|
0.97 (0.93–1.02)
|
0.26
|
|
|
Male sex
|
1.02 (0.20–6.36)
|
0.98
|
|
|
Non-Māori ethnicity
|
0.91 (0.23–3.43)
|
0.89
|
|
|
Preoperative symptoms
|
|
|
|
|
Headache
|
6.98 (1.52–37.7)
|
0.02
|
4.83 (1.16–20.05)
|
|
Neck pain
|
0.17 (0.03–0.74)
|
0.02
|
0.23 (0.06–0.84)
|
|
Limb weakness or paresthesias
|
1.98 (0.47–8.22)
|
0.60
|
|
|
Visual symptoms
|
0.18 (0.03–1.00)
|
0.05
|
0.35 (0.07–1.67)
|
|
Dizziness or vertigo
|
3.21 (0.41–71.72)
|
0.34
|
|
|
Nausea and vomiting
|
3.53 (0.29–101.95)
|
0.37
|
|
|
Syncope
|
1.04 (0.08–30.81)
|
0.98
|
|
|
Tinnitus
|
0.12 (0.01–2.94)
|
0.15
|
0.15 (0.01–1.83)
|
|
Unsteadiness
|
0.19 (0.03–1.20)
|
0.08
|
0.18 (0.04–0.81)
|
|
Speech or swallowing difficulty
|
0.59 (0.05–14.56)
|
0.69
|
|
|
Presence of syrinx
|
0.83 (0.18–4.74)
|
0.82
|
|
|
Extent of cerebellar tonsillar herniation
|
0.98 (0.88–1.11)
|
0.76
|
|
|
FMD[a]
|
0.72 (0.18–4.74)
|
0.82
|
0.96 (0.24–3.80)
|
|
Postoperative factors
|
|
|
|
|
Postoperative wound infection
|
0.98 (0.08–25.17)
|
0.99
|
|
|
Postoperative pseudomeningocoele
|
0.26 (0.04–1.76)
|
0.15
|
0.46 (0.07–2.96)
|
|
Postoperative complications
|
1.53 (0.22–17.52)
|
0.69
|
|
|
Syrinx reduction
|
1.13 (0.18–6.18)
|
0.89
|
|
|
Length of stay
|
0.90 (0.72–1.15)
|
0.38
|
|
Abbreviations: CI, confidence interval; FMD, foramen magnum decompression without
duraplasty; OR, odds ratio.
a FMD were included a priori in the multivariate regression analysis model.
b Variables with p-values ≤ 0.20 on univariate analysis were included in the multivariate regression
analysis model in a forward stepwise manner.
Syrinx Reduction
Overall, 32 (68%) patients with syrinx demonstrated reduction in size on postoperative
MRI at 6 months. Univariate and multivariate regression analyses are detailed in [Table 3]. No difference in the degree of syrinx reduction was found between those undergoing
FMDD and FMD on univariate (OR 1.49, 95% CI: 0.33–7.62) or multivariate analysis (aOR
1.69, 95% CI: 0.24–16.41). Patients experiencing limb weakness or paraesthesia preoperatively
had an increased likelihood for syrinx reduction postoperatively (aOR 16.45, 95% CI:
1.83–287.13). Older age (aOR 0.95, 95% CI: 0.81–0.98) and male sex (aOR 0.04, 95%
CI: 0.00–0.43) were associated with a reduced likelihood for syrinx reduction.
Table 3
Univariate and multivariate regression analysis for syrinx reduction
|
Variable
|
Univariate analysis
|
Multivariate analysis[b]
|
|
OR (95% CI)
|
p-Value
|
OR (95% CI)
|
|
Demographics
|
|
|
|
|
Age at surgery
|
0.97 (0.92–1.01)
|
0.15
|
0.95 (0.81–0.98)
|
|
Male sex
|
0.21 (0.04–0.90)
|
0.04
|
0.04 (0.00–0.43)
|
|
Non-Māori Ethnicity
|
1.04 (0.27–4.18)
|
0.95
|
|
|
Preoperative symptoms
|
|
|
|
|
Headache
|
0.69 (0.17–3.03)
|
0.62
|
|
|
Neck pain
|
0.68 (0.17–2.56)
|
0.58
|
|
|
Limb weakness or paresthesias
|
9.51 (1.86–79.6)
|
0.02
|
16.45 (1.83–287.13)
|
|
Visual symptoms
|
0.14 (0.02–0.71)
|
0.03
|
0.08 (0.00–2.13)
|
|
Dizziness or vertigo
|
3.68 (0.18–75.85)
|
0.54
|
|
|
Nausea and vomiting
|
4.35 (0.22–100)
|
0.34
|
|
|
Syncope
|
0.12 (0.00–1.68)
|
0.17
|
NA
|
|
Tinnitus
|
NA
|
|
|
|
Unsteadiness
|
0.35 (0.02–2.87)
|
0.38
|
|
|
Speech or swallowing difficulty
|
6.28 (0.86–77)
|
0.10
|
NA
|
|
Extent of cerebellar tonsillar herniation[a]
|
0.98 (0.85–1.13)
|
0.78
|
|
|
FMD
|
1.49 (0.33–7.62)
|
0.61
|
1.69 (0.24–16.41)
|
Abbreviations: CI, confidence interval; FMD, foramen magnum decompression without
duraplasty; OR, odds ratio.
a Extent of cerebellar tonsillar herniation was included a priori in the multivariate regression analysis model.
b Variables with p-values ≤ 0.20 on univariate analysis were included in the multivariate regression
analysis model in a forward stepwise manner.
Secondary Outcomes
Overall and Specific Postoperative Complications
Postoperative complications occurred in 9 (9.4%) patients undergoing foramen magnum
decompression ([Table 4]). There was no difference between FMD and FMDD regarding complication risk (5.3%
vs. 12.1%; OR 0.41, 95% CI: 0.08–2.06). This was also the case for postoperative wound
infections (0% vs. 8.6%; OR 0.13, 95% 0.01–2.35).
Table 4
Secondary outcomes between patients who underwent foramen magnum decompression without
versus with duraplasty
|
Outcome
|
Total cohort (n = 96)
|
FMD (n = 38)
|
FMDD (n = 58)
|
Effect estimate (95% CI)
|
|
Postoperative complication(s), n (%)[a]
|
9 (9.4)
|
2 (5.3)
|
7 (12.1)
|
0.41 (0.08–2.06)[b]
[c]
|
|
Postoperative infection, n (%)
|
5 (5.2)
|
0 (0.0)
|
5 (8.6)
|
0.13 (0.01–2.35)[b]
[c]
|
|
Pseudomeningocoele on postoperative MRI, n (%)
|
8 (8.3)
|
0 (0.0)
|
8 (13.8)
|
0.08 (0.00–1.38)[b]
[c]
|
|
Reoperation, n (%)
|
8 (8.3)
|
4 (10.5)
|
4 (6.9)
|
1.58 (0.27–9.09)[b]
[c]
|
|
Mean postoperative length of stay (d)
|
5.18
|
4.00
|
5.95
|
–1.94 (–2.93 to –0.96)[b]
[d]
|
Abbreviations: CI, confidence interval; FMD, foramen magnum decompression without
duraplasty; FMDD, foramen magnum decompression with duraplasty; MRI, magnetic resonance
imaging.
a Comprising hydrocephalus (n = 1), benign intracranial hypertension (n = 1), cerebrospinal fluid (CSF) leak (n = 3), respiratory arrest (n = 1), pneumonia (n = 1), hygromas (n = 2).
b Statistically significant differences are denoted by 95% CI which did not cross the
no-effect line (0 for mean difference and 1 for odds ratio).
c Odds ratio.
d Mean difference.
Pseudomeningocele
Eight patients (8.3%) developed pseudomeningocele following foramen magnum decompression
([Table 4]). Risk of pseudomeningocele did not differ between FMD and FMDD (0% vs. 13.8%; OR
0.08, 95% CI: 0.00–1.38).
Postoperative LOS
Mean postoperative LOS after foramen magnum decompression was 5.2 days. Compared with
FMDD, FMD was associated with significantly shorter postoperative LOS (4.00 vs. 5.95
days; MD –1.94 days, 95% CI: –2.93 to –0.96).
Reoperation
The overall rate of redo operations following foramen magnum decompression was 8.3%
(n = 8). Risk of reoperation did not differ between FMD and FMDD (10.5% vs. 6.9%; OR
1.58, 95% CI: 0.27–9.09).
Discussion
Many studies have proposed guidelines and protocols to determine when surgery should
be pursued, although further validation of these protocols is still required.[6] Among patients deemed suitable for surgery, this study aimed to clarify which surgical
technique, FMDD or FMD, offer greater clinical improvement and syrinx reduction in
patients with Chiari 1 malformation.
Although both clinical and radiological improvements are goals of surgical decompression,
symptom improvement remains the primary objective. This study found that FMDD was
associated with a higher likelihood of overall symptom improvement, although the difference
between FMDD and FMD did not reach statistical significance. Several meta-analyses
have demonstrated that FMDD has higher likelihood for symptom improvement.[6] However, other meta-analyses, such as those completed in the adult population show
no equivalence for symptom improvement.[7] Thus, the existing literature presents conflicting findings regarding the optimal
surgical technique for optimal clinical improvement, highlighting the need for high-quality
evidence such as randomized controlled trials. Furthermore, long-term clinical improvement
is also indirectly measured through the need for redo operation. No statistically
significant difference in redo operation rates was seen between the FMDD and FMD.
While most studies contradict these findings, the apparent difference in redo operations
may relate to the intrinsic nature of technique employed. FMD is a less extensive
procedure, and there is opportunity for further duraplasty if revisional surgery is
required, while FMDD is generally more extensive which limits the extent of further
intervention.[8]
To the best of the authors' knowledge, this is the first study to highlight preoperative
symptoms as predictive of treatment efficacy in terms of symptomatic improvement or
radiological evidence of syrinx reduction. Notably, the presence of limb weakness
or paraesthesia was associated with a significantly increased likelihood of syrinx
reduction. One possible explanation is that these patients may have had a pressure-transmitting
syrinx involving the spinothalamic or anterior horn tracts, which is more amenable
to collapse following decompression.[9] Additionally, preoperative headaches were associated with greater likelihood of
symptom improvement, whereas patients reporting neck pain were less likely to benefit
symptomatically on multivariate analysis. Although the underlying mechanisms remain
unclear, headaches may identify a subgroup of patients who are more responsive to
decompression, while isolated neck pain in the absence of neurological symptoms may
reflect a more complex Chiari malformation phenotype that is less responsive to surgical
decompression.[10]
As seen with prior studies, older age was negatively associated with syrinx reduction.[11] We have also demonstrated sex as independently predicting a reduced likelihood of
syrinx reduction, which has not been identified previously. With relation to the surgical
technique employed, no significant difference was seen for syrinx reduction although
FMD had a trend for higher likelihood for syrinx reduction (aOR 1.49 [0.33–7.62]).
Chai et al reported FMDD as being associated with an increased likelihood of syrinx
reduction, while Förander et al and Lu et al (pediatric) found no difference between
the FMD and FMDD groups.[8]
[12]
[13] With further stratification by age, Chai et al did note that this effect was only
seen within the adult population. However, given these meta-analyses draw from very
small and heterogeneous studies, conclusions are difficult to make due to the scarcity
of available literature.
Postoperative wound infections and pseudomeningocoele development were only seen in
patients undergoing FMDD. Although not statistically significant, this finding was
consistent with prior studies.[6] What may be important to extract from this is that both groups report minimal differences
in overall complications, but differences arising from the risks inherent to the technique
employed are likely to be seen with larger groups.[6]
[8] Importantly, patients undergoing FMD had shorter LOSs compared to those undergoing
FMDD. Given no significant differences were seen between clinical symptoms, syrinx
reduction, and postoperative complications, this finding supports FMD as potentially
having resource-saving implications.
There are several limitations to this study. First, the retrospective and single-center
design introduces selection bias and limits the generalizability of our findings.
To mitigate this, the study were conducted at Waikato Hospital, which is one of the
largest tertiary referral centers in AoNZ, and is the only institution to provide
neurosurgical service to the Midlands and Waikato region of AoNZ. Inherently, following
foramen magnum decompression surgery, a small proportion of patients included within
our study may be returned to their home domicile for repatriation and convalescence.
As a result, this introduces risk of attrition bias as patients may be lost to follow-up,
potentially leading to under- or overestimation of our findings. However, this is
unlikely to substantially influence our overall results as only six patients (< 5%
of the overall cohort) had incomplete follow-up data. Second, the small sample size
of our cohort has resulted in imprecise effect estimates, with some analyses having
extremely wide 95% CI. Consequently, outcomes considered to be statistically significant
should be interpreted with caution. Third, symptom improvement in the postoperative
period was assessed via chart review of electronic and handwritten records, which
introduces subjectivity, interobserver variability, and limits comparability between
other centers and institutions. Future studies should utilize standardized and validated
instruments such as the Chicago Chiari Outcome Scale, to improve reliability and facilitate
reproducibility of results.[14] Fourth, the study was conducted over a 12-year period (2010–2022), which potentially
also introduces temporal confounding resulting from a time period bias, as evolving
surgical expertise, perioperative care, and imaging capacity and advent of novel technology
could influence outcomes independently of surgical technique. However, due to the
infrequent occurrence of Chiari 1 malformation, this study interval was selected to
ensure comprehensive inclusion of all patients with Chiari 1 malformation at our institution.
Lastly, only operatively managed patients were included, and while surgical decision-making
is typically guided by symptom duration and severity, individual surgeon judgment
may also have an influence.[15] Notwithstanding, we aimed to examine the clinical and radiological efficacy of two
different surgical techniques for foraminal decompression, and thus the natural history
and outcomes of patients with Chiari 1 malformation managed conservatively was outside
the scope of our present study.
Conclusion
This analysis did not identify any difference in symptom improvement and radiological
evidence of syrinx reduction among patients with Chiari 1 malformation undergoing
FMD compared with FMDD. Interestingly, preoperative unsteadiness and neck pain was
predictive of a poor outcome, whereas patients with headache were more likely to experience
symptomatic improvement postoperatively, irrespective of surgical technique. Additionally,
presence of limb weakness or paraesthesia was associated with syrinx reduction. These
findings suggest that there may be a subset of patients with Chiari 1 malformation
who are more or less responsive to surgical decompression irrespective of the technique
employed, supporting the need for accurate patient selection when determining potential
operative candidates. Furthermore, FMD was associated with significantly shorter postoperative
LOS without any differential risk of postoperative complications or reoperation, supporting
this technique as potentially having resource-saving and economic implications. Larger
prospective studies will be required to corroborate and validate our findings.
