Chiari malformation type I (CM1) is a common pediatric neurosurgical disease characterized by ectopia of the cerebellar tonsils below the level of the foramen magnum and is often associated with syringomyelia (SM).1–4 Controversy continues regarding the treatment of CM1-SM by posterior fossa decompression with duraplasty (PFDD) versus extradural posterior fossa decompression (PFD).5–20 A number of single-institution studies and meta-analyses have evaluated PFD and PFDD in terms of complication rates, improvement in clinical symptoms, and reduction in syrinx size,5,18,21,22 but there have been no large-scale multicenter prospective or retrospective studies comparing PFD and PFDD in the pediatric population.
To address this critical void in the literature regarding pediatric CM1-SM, we conducted a retrospective analysis comparing PFD with PFDD using a large multicenter cohort of pediatric patients who underwent PFD or PFDD for CM1-SM at medical institutions across North America. Our primary research aim was to clarify differences in postoperative complications between PFD and PFDD, while our secondary research aim was to discern differences in postoperative syrinx size, symptom improvement, physical examination changes, and operative parameters between the two groups. To our knowledge, this is the largest study examining complication rates, postoperative outcomes, and other differences between PFD and PFDD for CM1-SM.
Methods
This study was a retrospective examination of retrospectively and prospectively acquired data in the Park-Reeves Syringomyelia Research Consortium (PRSRC) registry. The PRSRC is a multicenter research collaborative created to improve the understanding of CM1-SM and its treatment. Starting in 2011, the PRSRC has worked to collect data both retrospectively (from July 2011 to October 2014) and prospectively (since October 2014), with a total of 42 contributing centers. All study-related procedures were approved by institutional review boards at the host institution (Washington University in St. Louis Human Research Protection Office) and each participating center. For this study, patients were enrolled retrospectively and prospectively from surgery. Patients had their initial consultation for neurosurgical evaluation between September 2001 and February 2018. All patients had tonsillar ectopia ≥ 5 mm, syrinx diameter ≥ 3 mm, and ≥ 1-year postoperative follow-up from their index decompression procedure. The type of surgical intervention performed was at the discretion of the treating physician. Patients without documented SM; those older than 21 years of age; and those with Chiari malformation types II, III, and IV were excluded from the study. Presenting patient demographics, clinical signs and symptoms, surgical parameters, complications, and improvement in signs and symptoms were compiled and analyzed. Before data entry into the PRSRC registry, a signed consent form was required and obtained for all prospectively enrolled patients. A waiver of consent was obtained for retrospectively enrolled patients. For this current study, 349 patients (50.4%) were prospectively enrolled and 343 patients (49.6%) were retrospectively enrolled (Table 1). There were no differences between the retrospective and prospective cohorts for mean age at surgery (p = 0.75), mean follow-up time (p = 0.44), and sex (p = 0.2). The prospective cohort was more likely to receive PFDD (86.5%) compared with the retrospective cohort (79.6%) (p = 0.01).
Demographics of all patients stratified by treatment type
| Variable | PFD | PFDD | p Value |
|---|---|---|---|
| No. of pts | 117 | 575 | |
| Prospectively collected | 48 (41.0) | 301 (52.3) | 0.03 |
| Retrospectively collected | 69 (59.0) | 274 (47.7) | |
| Mean age at op, yrs | 9.46 ± 4.14 | 9.94 ± 4.64 | 0.26 |
| Mean follow-up, yrs | 2.73 ± 1.12 | 2.72 ± 1.16 | 0.93 |
| Sex | |||
| Male | 54 (46.2) | 224 (39.0) | 0.15 |
| Female | 63 (53.8) | 351 (61.0) | |
| Race | |||
| White | 88 (75.2) | 480 (83.5) | 0.05 |
| Black | 7 (6.0) | 60 (10.4) | 0.14 |
| Asian | 3 (2.6) | 7 (1.2) | 0.27 |
| Hispanic or Latino | 23 (19.7) | 62 (10.8) | 0.01 |
| Native American | 0 (0.0) | 2 (0.3) | >0.99 |
| Insurance | |||
| Private | 77 (65.8) | 403 (70.1) | 0.36 |
| Medicaid | 38 (32.5) | 180 (31.3) | 0.80 |
| Self-pay/no insurance | 3 (2.6) | 11 (1.9) | 0.71 |
Pt = patient. Values represent the number of patients (%) unless stated otherwise. Mean values are presented as mean ± SD. Boldface type indicates statistical significance.
Data Collection and Consistency
Demographic data, clinical symptoms, physical examination findings, and radiological measurements were collected and recorded in the PRSRC database at both pre- and postoperative encounters.
The data dictionary comprised 24 sections across multiple domains, including social and demographic parameters, preoperative and postoperative symptoms, treatment parameters, radiographic parameters, and outcome parameters. Most data elements offered specific designations (“yes/no,” “improved/stable/worse”). The same variables for each section were used for every subject at every site to ensure data consistency throughout the study. A data monitor at the lead PRSRC site (Washington University in St. Louis) remotely reviewed the completed data set for each participant as they were enrolled and with each follow-up visit. Sites were reimbursed for each participant enrolled but only after all data elements in the data dictionary were complete and all neuroimaging was uploaded to the Centralized Neuroimaging Data Archive. Radiological images for all PRSRC patients have been reviewed by two blinded, trained readers, one of whom is a board-certified pediatric neuroradiologist.
Radiological measurements included syrinx diameter in the anteroposterior (AP) dimension (mm), syrinx length (as measured in vertebral levels), tonsillar descent (in mm), clivoaxial angle (CXA), and a line perpendicular to a line from the basion to the posteroinferior aspect of the C2 body (pBC2; in mm). Clinical symptoms were subdivided into categories corresponding to headaches (including headache location), nausea/vomiting, visual disturbances, cranial nerve symptoms, subaxial pain (including upper- and lower-extremity pain, neck pain, back pain, and trunk pain), spinal cord symptoms, and cerebellar symptoms. Clinical signs were subdivided into categories corresponding to lethargy and papilledema, cranial nerve signs, spinal cord signs (weakness, reflexes, sensation, and proprioception), and cerebellar signs. Postoperative complications included pseudomeningocele, CSF leak or drainage, meningitis (chemical or infectious), hydrocephalus, infection, cervical instability (with any fusion procedures), and ventriculoperitoneal shunt requirement. Pseudomeningocele was diagnosed based on having both clinical signs or symptoms and positive imaging. They may have been treated conservatively or surgically. Asymptomatic pseudomeningoceles were not included. Chemical meningitis was diagnosed from clinical signs and symptoms with a negative CSF culture. Infectious meningitis was diagnosed from clinical signs and symptoms with a culture-positive lumbar puncture and antibiotic treatment.
Statistical Analysis
Comparisons between the PFD and PFDD groups were performed for demographic data, symptoms, physical examination findings, radiological measurements, postoperative complications, and intraoperative data. The Student t-test and the Mann-Whitney U-test were used to compare continuous variables (e.g., syrinx size, changes in radiological measurements, and mean hospital stay). A two-sided p value < 0.05 was considered statistically significant for all statistical tests. Odds ratios were calculated with 95% confidence intervals. The reference group was PFDD. All statistical analyses were performed using SAS version 9.4 (SAS Institute, Inc.) and Stata version 16.1 (StataCorp). Chi-square and Fisher’s exact tests were used to compare categorical variables (e.g., differences in complication rates or differences in the rates of symptoms or physical examination findings).
Results
A total of 692 patients met inclusion criteria: 117 (16.9%) underwent PFD and 575 (83.1%) underwent PFDD. Among all patients, 414 (59.8%) were female. The mean age at surgery for the cohort was 9.86 ± 4.56 years, and the mean follow-up time for all patients was 2.73 ± 1.16 years. The mean age at time of surgery and mean follow-up time did not significantly differ between the PFD and PFDD groups (p = 0.26 and p = 0.93, respectively). Demographic information for the individual groups is listed in Table 1.
Presenting Signs and Symptoms
The differences between groups for presenting signs and symptoms are listed in Table 2. For presenting symptoms, the only statistically significant difference between groups was for the proportion of patients who presented with holocranial headaches (26.5% PFD vs 14.6% PFDD, p = 0.003). For presenting physical examination findings, neck rotation weakness (4.3% PFD vs 1.0% PFDD, p = 0.02) was significantly different between groups, although the overall number of patients with this symptom was too small to make meaningful comparisons. No other presenting symptoms or examination findings significantly differed between groups.
Presenting symptoms and examination findings
| Variable | No. of Pts | p Value | OR* | 95% CI | |
|---|---|---|---|---|---|
| PFD (n = 117) | PFDD (n = 575) | ||||
| Symptoms | |||||
| Headache | |||||
| Occipital/suboccipital | 26 | 138 | 0.68 | 0.90 | 0.54–1.48 |
| Frontal | 12 | 69 | 0.75 | 0.84 | 0.40–1.63 |
| Holocranial | 31 | 84 | 0.003 | 2.11 | 1.27–3.44 |
| Parietal | 2 | 13 | 0.71 | 0.75 | 0.08–3.40 |
| Temporal | 2 | 18 | 0.55 | 0.54 | 0.06–2.30 |
| Worse w/ Valsalva | 8 | 57 | 0.38 | 0.67 | 0.27–1.46 |
| Nausea/vomiting | 4 | 46 | 0.11 | 0.41 | 0.10–1.15 |
| Cranial nerve | |||||
| Double vision | 4 | 21 | 0.90 | 0.93 | 0.23–2.84 |
| Facial numbness | 1 | 9 | 0.56 | 0.54 | 0.01–3.98 |
| Facial weakness | 1 | 4 | 0.85 | 1.23 | 0.02–12.68 |
| Difficulty swallowing | 11 | 62 | 0.74 | 0.86 | 0.39–1.71 |
| Hoarseness | 1 | 7 | 0.74 | 0.70 | 0.01–5.53 |
| Choking | 7 | 30 | 0.66 | 1.16 | 0.42–2.77 |
| Vocal cord dysfunction | 1 | 9 | 0.56 | 0.54 | 0.01–3.98 |
| Sleep apnea | 7 | 24 | 0.46 | 1.46 | 0.52–3.60 |
| Subaxial pain | |||||
| Upper-extremity pain | 5 | 28 | 0.78 | 0.87 | 0.26–2.36 |
| Lower-extremity pain | 4 | 35 | 0.25 | 0.55 | 0.14–1.57 |
| Neck pain | 24 | 99 | 0.43 | 1.24 | 0.72–2.08 |
| Back pain | 22 | 89 | 0.41 | 1.26 | 0.72–2.16 |
| Trunk pain | 0 | 10 | 0.23 | ||
| Spinal cord | |||||
| Incontinence | 4 | 36 | 0.28 | 0.53 | 0.13–1.52 |
| Cerebellar | |||||
| Tremors | 3 | 9 | 0.44 | 1.65 | 0.28–6.77 |
| Balance/gait ataxia | 17 | 84 | 0.98 | 0.99 | 0.53–1.78 |
| Exam findings | |||||
| Lethargic/stuporous | 0 | 1 | >0.99 | ||
| Papilledema | 0 | 5 | 0.59 | ||
| Cranial nerve | |||||
| Nystagmus | 5 | 39 | 0.41 | 0.61 | 0.18–1.61 |
| Dysconjugate gaze | 2 | 10 | 0.98 | 0.98 | 0.10–4.70 |
| Extraocular palsies | 1 | 3 | 0.52 | 1.64 | 0.03–20.66 |
| Facial weakness | 0 | 5 | 0.59 | ||
| Decreased facial sensation | 0 | 9 | 0.17 | ||
| Hearing loss | 3 | 13 | 0.74 | 1.13 | 0.20–4.23 |
| Hoarseness | 0 | 3 | >0.99 | ||
| Tongue deviation | 1 | 1 | 0.31 | 4.95 | 0.06–389.06 |
| Weak gag | 8 | 33 | 0.67 | 1.20 | 0.47–2.75 |
| Weak shrug | 1 | 0 | 0.17 | ||
| Weak neck rotation | 5 | 6 | 0.02 | 4.23 | 1.01–16.92 |
| Spinal cord | |||||
| Upper-extremity weakness | 7 | 24 | 0.46 | 1.46 | 0.52–3.60 |
| Lower-extremity weakness | 4 | 18 | 0.78 | 1.09 | 0.26–3.41 |
| Light touch deficit | 5 | 39 | 0.41 | 0.61 | 0.18–1.61 |
| Pinprick deficit | 3 | 30 | 0.34 | 0.48 | 0.09–1.58 |
| Joint position deficit | 0 | 3 | >0.99 | ||
| Deep tendon reflexes | |||||
| Absent | 6 | 27 | 0.81 | 1.10 | 0.36–2.80 |
| Hypoactive | 6 | 34 | >0.99 | 0.86 | 0.29–2.14 |
| Hyperactive | 13 | 51 | 0.48 | 1.28 | 0.62–2.50 |
| Ankle clonus | 4 | 18 | 0.78 | 1.09 | 0.26–3.41 |
| Babinski reflex | 2 | 13 | 0.71 | 0.75 | 0.08–3.39 |
| Hoffman’s reflex | 3 | 12 | 0.73 | 1.23 | 0.22–4.67 |
| Cerebellar | |||||
| Gait instability | 10 | 57 | 0.73 | 0.85 | 0.37–1.75 |
| Romberg response | 2 | 4 | 0.27 | 2.48 | 0.22–17.52 |
| Dysmetria | 1 | 12 | 0.71 | 0.40 | 0.01–2.79 |
Boldface type indicates statistical significance.
The reference group is PFDD.
Complications and Operative Parameters
The intraoperative complication rate was low (0.0% PFD vs 0.7% PFDD, p > 0.99), and the rate of complications beyond 6 months after surgery was comparable (0.9% PFD vs 2.8% PFDD, p = 0.33) between groups; however, PFD yielded significantly fewer complications within 6 months after surgery (13.7% PFD vs 24.3% PFDD, p = 0.01). When analyzing individual types of complications, patients who underwent PFDD were found to have a significantly higher rate of clinically significant postoperative pseudomeningocele than those who underwent PFD (7.7% PFDD vs 2.6% PFD, p = 0.04). No other specific postoperative complications were different between groups. Data regarding intraoperative and postoperative complications are listed in Table 3 (refer to Data Collection and Consistency above).
Intraoperative complications and complications within and beyond 6 months
| Variable | PFD (n = 117) | PFDD (n = 575) | p Value | OR* | 95% CI |
|---|---|---|---|---|---|
| No. of pts w/ intraop complications (%) | 0 (0) | 4 (0.7) | >0.99 | ||
| Vascular injury | 0 | 2 | >0.99 | ||
| Hemorrhage | 0 | 2 | >0.99 | ||
| Neurological injury | 0 | 2 | >0.99 | ||
| Death | 0 | 0 | |||
| Other complications | 0 | 3 | >0.99 | ||
| No. of pts w/ complications w/in 6 mos (%) | 16 (13.7) | 140 (24.3) | 0.01 | 0.49 | 0.26–0.87 |
| Pseudomeningocele | 3 | 44 | 0.04 | 0.32 | 0.06–0.94 |
| CSF leak | 5 | 35 | 0.52 | 0.69 | 0.21–1.82 |
| Received surgical treatment (wound oversewing or surgical wound revision) | 4 | 19 | >0.99 | 1.03 | 0.24–3.20 |
| Received no surgical treatment | 1 | 16 | 0.33 | 0.30 | 0.01–1.98 |
| External CSF drainage required | 2 | 17 | 0.75 | 0.57 | 0.05–2.46 |
| Meningitis | 2 | 25 | 0.18 | 0.38 | 0.04–1.57 |
| Chemical meningitis | 2 | 20 | 0.32 | 0.48 | 0.05–2.03 |
| Infectious meningitis | 0 | 5 | 0.59 | ||
| Postop hydrocephalus | 1 | 12 | 0.71 | 0.40 | 0.01–2.79 |
| Cervical instability | 1 | 3 | 0.52 | 1.64 | 0.03–20.66 |
| Managed w/ cervical collar | 1 | 2 | 0.43 | 2.47 | 0.04–47.73 |
| Necessitating fusion | 0 | 1 | >0.99 | ||
| Infection | 2 | 2 | 0.13 | 4.98 | 0.36–69.14 |
| No. of pts w/ complications beyond 6 mos (%) | 1 (0.9) | 16 (2.8) | 0.33 | 0.30 | 0.01–1.98 |
| Cervical instability | 0 | 1 | >0.99 | ||
| Fusion requirement | 0 | 1 | >0.99 | ||
| Infection | 0 | 1 | >0.99 | ||
| Pseudomeningocele | 0 | 3 | >0.99 | ||
| Hydrocephalus | 0 | 3 | >0.99 | ||
| Shunt placement | 0 | 2 | >0.99 | ||
| Syrinx shunt placement | 1 | 5 | >0.99 | 0.98 | 0.02–8.90 |
Boldface type indicates statistical significance.
The reference group is PFDD.
The mean operating room time was shorter for the PFD group (1.98 ± 1.06 hours) than for the PFDD group (2.92 ± 1.32 hours) (p < 0.0001). The number of patients with > 50 mL of blood loss was greater in the PFDD group (18.3%) compared with the PFD group (10.2%) (p = 0.04). The mean length of hospital stay was shorter in the PFD group (3.45 ± 2.58 days) than in the PFDD group (4.57 ± 3.70 days) (p = 0.0001). There were no differences between the groups in terms of need for postdecompression shunting (p = 0.41) or fusion (p = 0.54). The PFD group was more likely to require revision surgery compared with the PFDD group (17.9% vs 8.3%, p = 0.002), with most of these patients undergoing PFDD (85.7%) for their revision decompression. The mean time to revision surgery did not differ between the groups (p = 0.76). The operative characteristics and information on revision surgeries are listed in Table 4.
Operative characteristics and postdecompression surgeries
| Variable | PFD (n = 117) | PFDD (n = 575) | p Value |
|---|---|---|---|
| Mean operating room time, hrs | 1.98 ± 1.06 | 2.92 ± 1.32 | <0.0001 |
| EBL >50 mL, n (%) | 12 (10.2) | 105 (18.3) | 0.04 |
| Necessitating transfusion, n (%) | 0 (0) | 4 (0.7) | >0.99 |
| Mean length of postop hospital stay, days | 3.45 ± 2.58 | 4.57 ± 3.70 | 0.0001 |
| Postop VP shunt placement, n (%) | 5 (4.3) | 39 (6.8) | 0.41 |
| Postop spinal fusion, n (%) | 16 (13.7) | 68 (11.8) | 0.54 |
| Revision Chiari decompression op | |||
| 1 revision decompression, n (%) | 21 (17.9) | 48 (8.3) | 0.002 |
| PFD → PFDD | 18 | ||
| PFD → PFD | 3 | ||
| PFDD → PFDD | 47 | ||
| PFDD → PFD | 1 | ||
| Multiple decompression ops, n (%) | 0 (0.0) | 4 (0.7)* | >0.99 |
| Mean time to revision op, mos | 22.4 ± 17.8 | 21.0 ± 17.7 | 0.76 |
Mean values are presented as the mean ± SD. Boldface type indicates statistical significance.
All repeat decompressions in the multiple decompression group were PFDD.
Postoperative Signs and Symptoms
Table 5 shows the clinical outcomes for the patients’ postoperative symptoms and physical examination findings. Patients who underwent PFDD reported headache resolution or improvement significantly more often (89.6%) than those who received PFD (80.8%) (p = 0.04). PFDD was more often associated with resolved or improved back pain (86.5% vs 59.1%, p = 0.01), and patients who had experienced preoperative tremor only reported improvement postoperatively if they had undergone PFDD (PFDD 8/9 patients, PFD 0/3 patients; p = 0.02). There were no differences between the groups for the development of new postoperative signs or symptoms.
Postoperative changes in exam findings and symptoms
| Variable | No. of Pts | p Value | OR* | 95% CI | |
|---|---|---|---|---|---|
| PFD | PFDD | ||||
| Exam findings | |||||
| Cranial nerve | |||||
| Nystagmus | 95 | 399 | |||
| None | 90 | 382 | 0.59 | 0.80 | 0.27–2.85 |
| Resolved/improved | 2 | 2 | 0.17 | 4.27 | 0.30–59.35 |
| Stable/worse | 3 | 15 | 0.78 | 0.83 | 0.15–3.04 |
| Dysconjugate gaze | 105 | 412 | |||
| None | 103 | 410 | 0.18 | 0.25 | 0.02–3.52 |
| Resolved/improved | 1 | 0 | 0.20 | ||
| Stable/worse | 1 | 2 | 0.49 | 1.97 | 0.03–38.14 |
| Facial weakness | 77 | 417 | |||
| None | 77 | 416 | >0.99 | ||
| Resolved/improved | 0 | 0 | |||
| Stable/worse | 0 | 1 | >0.99 | ||
| New facial weakness | 0 | 0 | |||
| Hearing loss | 61 | 356 | |||
| None | 61 | 349 | 0.60 | ||
| Resolved/improved | 0 | 6 | 0.60 | ||
| Stable/worse | 0 | 1 | >0.99 | ||
| New hearing loss | 0 | 1 | >0.99 | ||
| Hoarseness | 51 | 353 | |||
| None | 51 | 352 | >0.99 | ||
| Resolved/improved | 0 | 1 | >0.99 | ||
| Stable/worse | 0 | 0 | |||
| New hoarseness | 0 | 0 | |||
| Tongue deviation | 95 | 411 | |||
| None | 95 | 410 | >0.99 | ||
| Resolved/improved | 0 | 0 | |||
| Stable/worse | 0 | 1 | >0.99 | ||
| New tongue deviation | 0 | 0 | |||
| Weak gag | 52 | 323 | |||
| None | 51 | 319 | >0.99 | 0.64 | 0.06–32.10 |
| Resolved/improved | 0 | 2 | >0.99 | ||
| Stable/worse | 1 | 2 | 0.36 | 3.15 | 0.05–61.21 |
| New gag weakness | 1 | 3 | 0.52 | 2.09 | 0.39–25.56 |
| Weak shrug | 53 | 340 | |||
| None | 53 | 340 | |||
| Resolved/improved | 0 | 0 | |||
| Stable/worse | 0 | 0 | |||
| New shrug weakness | 1 | 0 | 0.17 | ||
| Weak neck rotation | 46 | 359 | |||
| None | 45 | 355 | 0.45 | 0.51 | 0.05–25.51 |
| Resolved/improved | 1 | 4 | 0.45 | 1.97 | 0.04–20.46 |
| Stable/worse | 0 | 0 | |||
| New weak neck rotation | 1 | 1 | 0.31 | 8.04 | 0.10–633.20 |
| Spinal cord | |||||
| Upper-extremity weakness | 80 | 366 | |||
| None | 78 | 359 | 0.67 | 0.76 | 0.14–7.64 |
| Resolved/improved | 1 | 6 | >0.99 | 0.76 | 0.02–6.40 |
| Stable/worse | 1 | 1 | 0.33 | 4.62 | 0.06–363.62 |
| New upper-extremity weakness | 2 | 2 | 0.13 | 4.67 | 0.33–64.96 |
| Lower-extremity weakness | 80 | 366 | |||
| None | 79 | 360 | 0.80 | 1.32 | 0.16–61.31 |
| Resolved/improved | 0 | 3 | >0.99 | ||
| Stable/worse | 1 | 3 | 0.55 | 1.53 | 0.03–19.34 |
| New lower-extremity weakness | 0 | 2 | >0.99 | ||
| Light touch deficit | 25 | 158 | |||
| None | 25 | 158 | |||
| Resolved/improved | 0 | 0 | |||
| Stable/worse | 0 | 0 | |||
| New light touch deficit | 1 | 1 | 0.31 | 6.54 | 0.08–516.60 |
| Deep tendon reflexes | 84 | 359 | |||
| Absent | 6 | 16 | 0.28 | 1.65 | 0.51–4.61 |
| Normal | 63 | 285 | 0.38 | 0.78 | 0.44–1.43 |
| Hypoactive | 7 | 25 | 0.64 | 1.21 | 0.43–3.02 |
| Hyperactive | 8 | 33 | 0.92 | 1.04 | 0.40–2.41 |
| Cerebellar | |||||
| Gait instability | 100 | 454 | |||
| Normal | 93 | 425 | 0.82 | 0.91 | 0.37–2.53 |
| Improved | 3 | 20 | 0.78 | 0.67 | 0.12–2.33 |
| Stable/worse | 4 | 9 | 0.27 | 2.06 | 0.45–7.54 |
| Symptoms | |||||
| Prior headache | 73 | 346 | |||
| Resolved/improved | 59 | 310 | 0.04 | 0.49 | 0.24–0.76 |
| Stable/worse | 14 | 36 | |||
| New headache | 20 | 97 | |||
| Occipital/suboccipital | 5 | 14 | 0.31 | 1.98 | 0.48–6.95 |
| Frontal | 4 | 23 | 0.70 | 0.79 | 0.18–2.81 |
| Holocranial | 11 | 48 | 0.81 | 1.25 | 0.42–3.74 |
| Parietal | 0 | 4 | >0.99 | ||
| Temporal | 0 | 8 | 0.36 | ||
| Motor/sensory deficits | |||||
| Prior weakness | 12 | 75 | |||
| Resolved/improved | 9 | 58 | 0.86 | 0.88 | 0.19–5.61 |
| Stable/worse | 3 | 17 | |||
| New weakness | 0 | 2 | >0.99 | ||
| Prior sensory deficits | 18 | 130 | |||
| Resolved/improved | 16 | 104 | 0.37 | 2.0 | 0.42–18.95 |
| Stable/worse | 2 | 26 | |||
| New sensory deficit | 4 | 28 | 0.63 | 0.69 | 0.17–2.03 |
| Cranial nerve | |||||
| Prior double vision | 4 | 21 | |||
| Resolved/improved | 2 | 20 | 0.06 | 0.05 | 0.001–1.65 |
| Stable/worse | 2 | 1 | |||
| New double vision | 0 | 2 | >0.99 | ||
| Prior dysphagia | 11 | 62 | |||
| Resolved/improved | 11 | 57 | >0.99 | ||
| Stable/worse | 0 | 5 | |||
| New dysphagia | 1 | 2 | 0.43 | 2.47 | 0.04–47.73 |
| Prior choking | 7 | 30 | |||
| Resolved/improved | 7 | 26 | 0.57 | ||
| Stable/worse | 0 | 4 | |||
| New choking | 1 | 0 | 0.17 | ||
| Prior sleep apnea | 7 | 24 | |||
| Resolved/improved | 4 | 17 | 0.65 | 0.55 | 0.07–4.84 |
| Stable/worse | 3 | 7 | |||
| New sleep apnea | 0 | 0 | |||
| Neck & back pain | |||||
| Prior neck pain | 51 | 205 | |||
| Resolved/improved | 43 | 181 | 0.48 | 0.71 | 0.28–1.97 |
| Stable/worse | 8 | 24 | |||
| New neck pain | 3 | 18 | >0.99 | 0.81 | 0.15–2.86 |
| Prior back pain | 22 | 89 | |||
| Resolved/improved | 13 | 77 | 0.01 | 0.22 | 0.07–0.74 |
| Stable/worse | 9 | 12 | |||
| New back pain | 2 | 7 | 0.65 | 1.41 | 0.14–7.53 |
| Cerebellar | |||||
| Prior tremors | 3 | 9 | |||
| Resolved/improved | 0 | 8 | 0.02 | ||
| Stable/worse | 3 | 1 | |||
| New tremors | 0 | 1 | >0.99 | ||
| Prior balance/gait ataxia | 17 | 84 | |||
| Resolved/improved | 13 | 69 | 0.73 | 0.71 | 0.18–3.40 |
| Stable/worse | 4 | 15 | |||
| New balance/gait ataxia | 0 | 3 | >0.99 | ||
Boldface type indicates statistical significance.
The reference group is PFDD.
Radiological Measurements
Preoperative and postoperative radiological measurements are displayed in Table 6. Both PFD and PFDD yielded significantly smaller mean AP syrinx diameters after surgery (p < 0.0001 for both). However, when comparing percent changes in mean syrinx size between groups, patients who received PFDD had a significantly greater percent change in syrinx diameter (43.7% vs 26.9%, p = 0.0001) and syrinx length (18.9% vs 5.6%, p = 0.04) compared with those who received PFD. A significantly greater proportion of PFDD patients (37.6%) had postoperative syringes that were ≤ 2 mm in AP diameter compared with PFD patients (27.3%) (p = 0.04).
Postoperative changes in radiological measurements
| Variable | PFD | PFDD | p Value |
|---|---|---|---|
| AP syrinx diameter, mm | |||
| Mean preop | 6.36 ± 2.79 | 7.75 ± 3.07 | <0.0001 |
| Mean postop | 4.65 ± 2.90 | 4.36 ± 3.28 | 0.34 |
| % change in AP diameter | 26.9% | 43.7% | 0.0001 |
| p value (preop vs postop) | <0.0001 | <0.0001 | |
| No. of pts w/ syrinx ≤2 mm after decompression op | 32 (27.3%) | 216 (37.6%) | 0.04 |
| Syrinx length (vertebral segments) | |||
| Mean preop | 7.44 ± 4.43 | 9.15 ± 4.61 | 0.0002 |
| Mean postop | 7.02 ± 4.68 | 7.42 ± 4.29 | 0.39 |
| % change in length | 5.6% | 18.9% | 0.04 |
| p value (preop vs postop) | 0.48 | <0.0001 | |
| No. of pts w/ preoperative syringes >10 mm | 94 | 360 | |
| AP syrinx diameter | |||
| Mean preop | 5.37 ± 1.94 | 6.14 ± 1.88 | 0.001 |
| Mean postop | 3.36 ± 1.91 | 2.86 ± 2.00 | 0.03 |
| % change | 37.4% | 53.4% | <0.0001 |
| p value (preop vs postop) | <0.0001 | <0.0001 | |
| Syrinx length (vertebral segments) | |||
| Mean preop | 6.68 ± 4.16 | 8.16 ± 4.56 | 0.003 |
| Mean postop | 6.27 ± 4.40 | 6.54 ± 4.22 | 0.59 |
| % change | 6.1% | 19.8% | 0.02 |
| p value (preop vs postop) | 0.51 | <0.0001 | |
| CXA, ° | |||
| Mean preop | 142.93 ± 14.10 | 145.77 ± 12.81 | 0.05 |
| Mean postop | 142.47 ± 21.66 | 142.02 ± 13.48 | 0.83 |
| % change | 0.32% | 2.6% | 0.19 |
| p value (preop vs postop) | 0.85 | <0.0001 | |
| pBC2, mm | |||
| Mean preop | 6.71 ± 1.23 | 6.89 ± 1.26 | 0.15 |
| Mean postop | 6.56 ± 1.17 | 6.67 ± 1.21 | 0.36 |
| % change | 2.2% | 3.2% | 0.68 |
| p value (preop vs postop) | 0.34 | 0.003 |
Boldface type indicates statistical significance.
In an attempt to limit the effects of procedural bias related to particularly large syringes, subanalyses were performed for patients with a preoperative AP syrinx diameter < 10 mm (94 PFD patients and 360 PFDD patients; Table 6). These subanalyses demonstrated that both PFD and PFDD yielded significant decreases in mean syrinx diameter (p < 0.0001). However, the PFDD group experienced a higher percent decrease in syrinx diameter compared with the PFD group (53.4% vs 37.4%, p < 0.0001). Postoperatively, only PFDD produced a significant mean decrease in syrinx length (p < 0.0001). PFDD also yielded a significantly higher mean percent decrease in syrinx length compared with PFD (19.8% vs 6.1%, p = 0.02).
Preoperative and postoperative CXA and pBC2 were not different between the PFD and PFDD groups. However, the calculated change from preoperative to postoperative values was larger for the PFDD group (p < 0.0001 for reduction in CXA and p = 0.003 for reduction in pBC2).
Univariate and Multivariate Analyses
We conducted univariate and multivariate analyses to assess for the effect of confounding variables on complications within 6 months and found no significant affects from potential confounding variables (Table 7). Specifically, age, pBC2 distance, and syrinx diameter did not impact complication rates. After accounting for these factors, type of surgery was still significantly different, with PFDD patients having 1.5 greater odds of complications compared with PFD patients (p = 0.03). PFDD patients had 1.66 greater odds of complications compared with PFD in univariate analysis (p = 0.01).
Logistic regression results for potential factors influencing postoperative complications within 6 months
| Variable | p Value | OR | 95% CI |
|---|---|---|---|
| Univariate analysis | |||
| Age | 0.10 | 1.09 | 1.05–1.18 |
| PFDD (vs PFD) | 0.01 | 1.66 | 1.15–2.38 |
| CXA | 0.54 | 0.99 | 0.98–1.01 |
| pBC2 distance | 0.09 | 1.10 | 0.98–1.22 |
| Frontal-occipital ratio | 0.65 | 2.98 | 0.03–33.42 |
| Syrinx diameter | 0.03 | 1.05 | 1.005–1.10 |
| Syrinx length | 0.28 | 1.02 | 0.98–1.05 |
| Preop hydrocephalus | 0.95 | 0.98 | 0.43–2.22 |
| Multivariate analysis | |||
| Age | 0.30 | 1.02 | 0.98–1.05 |
| PFDD (vs PFD) | 0.03 | 1.54 | 1.05–2.25 |
| pBC2 distance | 0.27 | 1.07 | 0.95–1.20 |
| Syrinx diameter | 0.07 | 1.04 | 0.98–1.09 |
Boldface type indicates statistical significance.
Discussion
In this study, we conducted retrospective analyses of retrospectively and prospectively collected data from the PRSRC to compare complications and outcomes between PFD and PFDD in patients with CM1-SM. This represents the first large-scale, multi-institutional study of pediatric patients undergoing PFD compared with those undergoing PFDD for CM1-SM. Our results indicated that the overall complication rate between the groups was equivalent intraoperatively and beyond 6 months after surgery. There was a higher complication rate in the PFDD group within 6 months postoperatively, especially regarding pseudomeningocele rates. Of note, pseudomeningoceles were only documented in the registry if they were both clinically significant and radiographically evident. Asymptomatic pseudomeningoceles were not included. Interestingly, there were several pseudomeningoceles and CSF leaks in the PFD group, which may represent an unintended durotomy. The PFD group did have shorter operative times, less blood loss, and shorter hospital stays, but these patients more frequently required revision surgery compared with PFDD patients. PFDD was associated with better headache and back pain improvement and more significant improvements in syrinx size and other radiographic measures. Other than headache and back pain, both groups seemed to have similar outcomes for the various postoperative signs and symptoms reported.
Our data indicating a significantly higher overall complication rate within 6 months after PFDD are in line with previously published studies. In a prior meta-analysis, Chai et al. found a higher rate of aseptic meningitis (relative risk 4.02) and CSF leak (relative risk 5.23) in the PFDD group.5 Munshi et al. found that in a group of 11 patients undergoing PFD, there was just one complication of a superficial incision infection compared with 10 of 23 patients who underwent PFDD and had some form of minor complication.6 Yeh et al. found that patients undergoing PFDD were much more likely to have complications than those who underwent PFD.9 Chotai and Medhkour found a complication rate at 2 times odds for the PFDD group compared with the PFD group,15 and Lee et al. found no complications in the PFD group compared with 19.4% in the PFDD group.16 In a study using a large national registry, Shweikeh et al. found a higher rate of complications in the PFDD group.18 We also found a higher rate of reoperation in the PFD group compared with the PFDD group. Other studies have found similar results. Mutchnick et al. found that among 64 patients who underwent PFDD, only 2 required repeat decompression compared with 7 of 56 patients in the PFD group.12 The higher reoperation rate, however, needs to be weighed against the lower complication rate.
We found that hospital stay, operative time, and blood loss were greater in the PFDD group than in the PFD group, which is perhaps not surprising given that PFDD is a more complex and invasive procedure. However, transfusion rates were similar between the groups. Yeh et al. found similar benefit in the PFD group compared with the PFDD group in hospital stay.9 Mutchnick et al. found similar reductions in operating room time and hospital stay in the PFD group,12 while Chotai and Medhkour found a shorter length of hospital stay in the PFD group.15 Lee et al. found that operating room times were roughly an hour shorter and hospital stays a day shorter in the PFD group compared with the PFDD group.16 Looking at a large national registry, Shweikeh et al. found longer hospital stays and higher charges in the PFDD group.18 Overall, it seems that PFD may be associated with a reduction in operative and hospital needs, which may portend a faster recovery and a reduction in resource utilization.
Our results demonstrated significant improvements in several radiographic parameters in the PFDD group compared with the PFD group, which is in line with previous research. However, it should be noted that the improvement in radiographic parameters seen in the PFDD group did not necessarily yield a more substantial improvement in symptoms over PFD, except for patients presenting with headaches. Additionally, our study design of a 1-year follow-up did not enable us to assess the effects of syrinx reduction on long-term spinal deformity. Using the PRSRC, Hale et al. found that younger age and PFDD were associated with greater rates of syrinx resolution.21 In a meta-analysis, Chai et al. saw a significant decrease in syrinx in the PFDD group (relative risk 1.57).5 Munshi et al. also saw similar improvement in radiographic syrinx in the PFDD group compared with the PFD group.6 Yeh et al. found similar results in a study that used intraoperative ultrasound to guide decision-making between the PFDD and PFD.9 Intraoperative ultrasound may be a useful adjunct to make a determination between these two types of surgery, but better understanding of this tool is needed in CM1-SM patients. Additionally, while radiographic parameters such as syrinx size or length may change, this was not necessarily associated with a greater clinical benefit in spine-related parameters other than back pain. Studies examining the association between syrinx reduction and improvements in symptoms or physical examination findings would help clarify the benefits that patients may experience from significantly smaller syrinxes following PFDD. Studies comparing PFDD for patients with and without SM may also help elucidate whether PFDD is particularly useful in the population of patients with a syrinx. Moreover, there were no postoperative differences between PFD and PFDD for either CXA or pBC2, and the mean pBC2 remained < 7 mm regardless of surgery type. These findings, we believe, merit further investigation in future studies.
We did find that PFDD was more likely to benefit patients with headache or back pain, which may support this more invasive approach for the elimination of these particular symptoms. However, if these symptoms are not the main focus of the presentation, then PFD may be sufficient since the outcomes for other signs and symptoms are equivalent with a lower risk of postoperative complications and decrease resource utilization.
Limitations
There are limitations to the current study. While PRSRC data have been prospectively collected for a number of patients, a large proportion of the data was collected retrospectively. Unfortunately, there are no data from the prospective cohort regarding the number of patients or legal guardians who refused to give consent for their child to be included in the PRSRC database, which could hide a bias in patient selection. The retrospective cohort was obtained using a waiver of consent, and thus enrollment could not be declined. These patients were selected and enrolled by participating study teams based on qualification criteria. While efforts were made to consecutively capture all patients meeting criteria both retrospectively and prospectively, there is a possibility of convenience sampling in the retrospective cohort. However, across the three largest centers, the patient enrollment rate was 95%. Fortunately, the prospectively and retrospectively acquired cohorts did not differ based on mean age at surgery, mean follow-up time, and sex. Additionally, this is a descriptive study that applies to pediatric patients with CM1-SM from the PRSRC database only, limiting its generalizability to CM1 patients without SM. Therefore, the results are only applicable to CM1 patients with SM. This study also only pertains to surgically treated patients and pediatric patients with CM1-SM and is limited to 1 year of follow-up. The PRSRC includes institutions with a specific interest in CM1-SM. Institutional and surgeon biases between and within centers may also contribute to the study findings. Data reporting in the database is dependent on the efforts of busy investigators or research administrators at the various institutions. While every attempt has been made to standardize data entry, there is still some variability in data collection and how data are reported. The data monitor for the study remotely reviewed data sets for the participants to minimize this variability. All data were processed with the reports available, and this possibly leads to a higher frequency of categorical outcomes, as negative findings are less likely reported than positive ones. There are additionally a high number of analyses performed, which does increase the chance of alpha error in this study. Finally, some signs and symptoms had small sample sizes, making both clinical and statistical significance testing challenging for those parameters.
Conclusions
Analyses of a large multicenter cohort of pediatric patients with CM1-SM demonstrated that patients undergoing PFD had a lower complication rate within 6 months of surgery and had shorter operative times, less blood loss, and a shorter length of hospital stay compared with those undergoing PFDD. However, PFDD was associated with better headache and syrinx outcomes as well as a reduced revision rate. The improved headache and radiographic outcomes seen with PFDD must be weighed against the reduced complication rate and resource utilization seen with PFD.
Acknowledgments
This publication was made possible through the support of Sam and Betsy Reeves, the Spears and O’Keefe families, and the many other contributors to the Park-Reeves Syringomyelia Research Consortium. Support was also provided by the Eunice Kennedy Shriver National Institute of Child Health & Human Development of the National Institutes of Health under award number U54 HD087011 to the Intellectual and Developmental Disabilities Research Center at Washington University. We furthermore appreciate the help of Timothy M. George, MD, and Sanjiv Bhatia, MD, who both contributed to this study but passed away before this manuscript was written.
Research reported in this publication was supported by the Park-Reeves Syringomyelia Research Consortium and the National Center for Advancing Translational Sciences of the National Institutes of Health under award number UL1 TR002345. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. Additional funding was provided by an American Syringomyelia Alliance Project, Inc. grant.
Disclosures
Dr. Limbrick: research and educational support for non–study-related clinical or research effort from Medtronic, Inc. and Microbot Medical, Inc. Dr. Alden: honoraria from Takeda.
Author Contributions
Conception and design: Akbari, Park, Limbrick. Acquisition of data: Ackerman, Adelson, Ahmed, Albert, Aldana, Alden, Anderson, Bauer, Bethel-Anderson, Bierbrauer, Brockmeyer, Chern, Couture, Daniels, Dlouhy, Durham, Ellenbogen, Eskandari, Fuchs, Grant, Graupman, Greene, Greenfield, Gross, Guillaume, Hankinson, Heuer, Iantosca, Iskandar, Jackson, Jallo, Johnston, Kaufman, Keating, Khan, Krieger, Leonard, Maher, Mangano, McComb, McEvoy, Meehan, Menezes, Muhlbauer, O’Neill, Olavarria, Ragheb, Selden, Shah, Shimony, Smyth, Stone, Strahle, Tamber, Tuite, Tyler-Kabara, Wait, Wellons, Whitehead, Park, Limbrick. Analysis and interpretation of data: Akbari, Yahanda, Limbrick. Drafting the article: Akbari, Yahanda. Critically revising the article: Akbari, Yahanda, Anderson, Bauer, Limbrick. Reviewed submitted version of manuscript: Akbari, Ackerman, Adelson, Ahmed, Albert, Aldana, Alden, Bethel-Anderson, Bierbrauer, Brockmeyer, Chern, Couture, Daniels, Dlouhy, Durham, Ellenbogen, Eskandari, Fuchs, Grant, Graupman, Greene, Greenfield, Gross, Guillaume, Hankinson, Heuer, Iantosca, Iskandar, Jackson, Jallo, Johnston, Kaufman, Keating, Khan, Krieger, Leonard, Maher, Mangano, McComb, McEvoy, Meehan, Menezes, Muhlbauer, O’Neill, Olavarria, Ragheb, Selden, Shah, Shannon, Shimony, Smyth, Stone, Strahle, Tamber, Torner, Tuite, Tyler-Kabara, Wait, Wellons, Whitehead, Park, Limbrick. Approved the final version of the manuscript on behalf of all authors: Akbari. Statistical analysis: Akbari, Yahanda, Torner. Administrative/technical/material support: Park, Limbrick. Study supervision: Limbrick.
Supplemental Information
Previous Presentations
This work was presented as a podium talk at the 2021 AANS/CNS Section on Pediatric Neurological Surgery Annual Meeting, Salt Lake City, Utah, December 7–10, 2021.
References
- 1
Aitken LA, Lindan CE, Sidney S, et al. Chiari type I malformation in a pediatric population. Pediatr Neurol. 2009;40(6):449–454.
- 2
Arnautovic A, Splavski B, Boop FA, Arnautovic KI. Pediatric and adult Chiari malformation Type I surgical series 1965-2013: a review of demographics, operative treatment, and outcomes. J Neurosurg Pediatr. 2015;15(2):161–177.
- 3
Strahle J, Muraszko KM, Kapurch J, Bapuraj JR, Garton HJL, Maher CO. Chiari malformation Type I and syrinx in children undergoing magnetic resonance imaging. J Neurosurg Pediatr. 2011;8(2):205–213.
- 4
Wu YW, Chin CT, Chan KM, Barkovich AJ, Ferriero DM. Pediatric Chiari I malformations: do clinical and radiologic features correlate?. Neurology. 1999;53(6):1271–1276.
- 5↑
Chai Z, Xue X, Fan H, et al. Efficacy of posterior fossa decompression with duraplasty for patients with Chiari malformation type I: a systematic review and meta-analysis. World Neurosurg. 2018;113:357–365.e1.
- 6↑
Munshi I, Frim D, Stine-Reyes R, Weir BKA, Hekmatpanah J, Brown F. Effects of posterior fossa decompression with and without duraplasty on Chiari malformation-associated hydromyelia. Neurosurgery. 2000;46(6):1384–1390.
- 7
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.
- 8
Navarro R, Olavarria G, Seshadri R, Gonzales-Portillo G, McLone DG, Tomita T. Surgical results of posterior fossa decompression for patients with Chiari I malformation. Childs Nerv Syst. 2004;20(5):349–356.
- 9↑
Yeh DD, Koch B, Crone KR. Intraoperative ultrasonography used to determine the extent of surgery necessary during posterior fossa decompression in children with Chiari malformation type I. J Neurosurg. 2006;105(1 suppl):26–32.
- 10
McGirt MJ, Attenello FJ, Datoo G, 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(1):52–57.
- 11
Erdogan E, Cansever T, Secer HI, et al. The evaluation of surgical treatment options in the Chiari Malformation Type I. Turk Neurosurg. 2010;20(3):303–313.
- 12↑
Mutchnick IS, Janjua RM, Moeller K, Moriarty TM. Decompression of Chiari malformation with and without duraplasty: morbidity versus recurrence. J Neurosurg Pediatr. 2010;5(5):474–478.
- 13
Romero FR, Pereira CAB. Suboccipital craniectomy with or without duraplasty: what is the best choice in patients with Chiari type 1 malformation?. Arq Neuropsiquiatr. 2010;68(4):623–626.
- 14
Yilmaz A, Kanat A, Musluman AM, 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(2):307–313.
- 15↑
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.
- 16↑
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(8):1419–1424.
- 17
Gürbüz MS, Berkman MZ, Ünal E, 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(5):721–727.
- 18↑
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(1):31–37.
- 19
Chen J, Li Y, Wang T, et al. Comparison of posterior fossa decompression with and without duraplasty for the surgical treatment of Chiari malformation type I in adult patients: a retrospective analysis of 103 patients. Medicine (Baltimore). 2017;96(4):e5945.
- 20
Litvack ZN, Lindsay RA, Selden NR. Dura splitting decompression for Chiari I malformation in pediatric patients: clinical outcomes, healthcare costs, and resource utilization. Neurosurgery. 2013;72(6):922–929.
- 21↑
Hale AT, Adelson PD, Albert GW, et al. Factors associated with syrinx size in pediatric patients treated for Chiari malformation type I and syringomyelia: a study from the Park-Reeves Syringomyelia Research Consortium. J Neurosurg Pediatr. 2020;25(6):629–639.
- 22↑
Durham SR, Fjeld-Olenec K. Comparison of posterior fossa decompression with and without duraplasty for the surgical treatment of Chiari malformation Type I in pediatric patients: a meta-analysis. J Neurosurg Pediatr. 2008;2(1):42–49.
