Diagnosis of piriformis syndrome based on the piriformis muscle cross-sectional area on hip MRI

Abstract

Piriformis syndrome (PS) is a condition in which the piriformis muscle characterized by buttock and hip pain. PS is frequently overlooked in clinical field because its symptoms are similar to that of primary sacral dysfunction, or lumbar radiculopathy. Thus exact diagnosis is very important. The piriformis muscle cross-sectional area (PMCSA) has not yet been proven to be an independent risk factor for the diagnosis of PS. The present study analyze the relationship between the PMCSA and PS. We hypothesized that PMCSA is a key diagnostic parameter in the PS. Both PMCSA and piriformis muscle thickness (PMT) samples were obtained from 30 patients with PS, and from 30 healthy individuals who underwent hip magnetic resonance imaging (H-MRI) with no evidence of PS. T1W H-MRI images were obtained. We investigated the PMCSA and PMT on H-MRI using a PACS system. The PMCSA was measured in coronal sections of the entire image by contour drawing. The PMT was measured primarily based on the hypertrophied piriformis muscle. Both PMCSA and PMT were significantly associated with PS, but PMCSA was measured as a much more sensitive parameter. Therefore, to evaluate patients with PS, physicians should examine PMCSA more carefully than PMT. The average PMCSA was 564.36 ± 121.61 mm² in the normal group and 736.88 ± 168.87 mm² in the PS group. The average PMT was 13.83 ± 2.61 mm in the control group and 15.99 ± 2.34 mm in the PS group. PS group had significantly higher PMCSA (P≤.001) and PMT (P≤.001). Regarding the validity of both PMCSA and PMT as predictors of PS, receiver operating characteristic curve analysis showed the best cutoff point for the PMCSA was 611.67 mm², with 75.0% specificity, 75.0% sensitivity, and the AUC of 0.81 (95% CI 0.68–0.94). The best cutoff value of the PMT was 14.24 mm, with 70.0% sensitivity, 70.0% specificity, and the AUC of 0.78 (95% CI 0.63–0.93).

1. Introduction

Piriformis syndrome (PS) is not uncommon peripheral neuritis caused by an abnormal inflammation of the piriformis muscle (PM).^[1–3] PS often is misdiagnosed or goes unrecognized in clinical fields. Delayed diagnosis of PS may lead to the chronic somatic dysfunction, pathologic disorder of the sciatic nerve, and compensatory changes resulting in paresthesia, pain, muscle weakness and hyperesthesia.^[4] The challenge for physicians is to recognize signs and symptoms that are unique to PS, enabling appropriate management in an appropriate time. Diagnosis of PS is made by the patient’s report of symptoms and by physical examination using movements to elicit discomfort to the PM.^[4–7] A tender or contracted PM can be also found on physical examination.^[4] However, all of these clinical and diagnostic tests are frequently insufficient to identify PS. Because, there are no objective, essential diagnostic tool for PS. Use of hip magnetic resonance imaging (H-MRI) to evaluate the PM in PS patients has been demonstrated. H-MRI can be a noninvasive and valuable diagnostic tool, revealing an enlarged PM.^[8–11]

H-MRI is very useful to diagnose PS correctly and to differentiate PS from another possible causes of sciatica and lower back pain, such as lumbar stenosis, lumbar disk herniation and mass effects in the region of the PM.^[1,12,13] Therefore, H-MRI can be an important image analysis modality for detection the changes of the PM in PS. Therefore, piriformis muscle thickness (PMT) may be a useful diagnostic variable in the evaluation of PS.^[6,14] However, few studies have analyzed the exact anatomical basis of PM hypertrophy. Therefore, we measured piriformis muscle cross-sectional area (PMCSA) to assess the relationship between PM hypertrophy and PS. PMCSA has not been investigated for association with PS. Additionally, no study has calculated the highest clinical cutoff values for PMT and PMCSA. In this paper, we compared the accuracy of PMT and PMCSA in diagnosing PS using H-MRI to determine which parameter is more sensitive.

2. Materials and methods

2.1. Patients

This was a single-center case-control study and CKU-IRB reviewed and approved the research protocol. We reviewed patients who visited our hospital from April 2017 to November 2022, and who were diagnosed with PS, retrospectively. The PS group included 30 patients (16 males and 14 females) with an average age of 46.95 ± 12.58 years (range, 16–63 years).

The inclusion criteria of the PS group were: numbness and tingling in the buttocks that may extend down the back of the leg; H-MR image taken for review; tenderness of the muscles in the buttocks; difficulty sitting comfortably; pain in the buttocks and legs that worsens with activity. Exclusion criteria were: hip fracture; L4, L5 lumbar radiculopathy; history of any hip surgery; history of any pelvic surgery; and meralgia paresthetica. To compare the PMT and PMCSA between individuals with and without PS, we also enrolled a control group of individuals who underwent H-MRI without PS. The normal group consisted of 20 individuals (7 men and 13 women) with an average age of 45.35 ± 14.03 years (range, 14–75 years; Table 1).

Table 1.

Comparison of the characteristics of normal and PS groups.

Variable	Normal group n = 30	PS group n = 30	Statistical significance
Gender (male/female)	14/16	16/14	NS
Age (yr)	46.35 ± 14.03	46.95 ± 12.58	NS
PMT (mm)	13.83 ± 2.61	15.99 ± 2.34	P < .001
PMCSA (mm2)	564.36 ± 121.61	736.88 ± 168.87	P < .001

Data represent the mean ± standard deviation (SD) or the numbers of patients.

NS = not statistically significant, PMCSA = piriformis muscle cross-sectional area, PMT = piriformis muscle thickness, PS = piriformis syndrome (P > .05), SD = standard deviation.

2.2. MRI scanning protocol

All individuals had 3 T MRI examinations (MRI 3 T Area Siemens, Erlangen, Germany) with 3 T scanners (Achieva; Philips Healthcare, The Netherlands). The protocol was as follows: H-MRI with turbo spin echo T1W acquisition of transverse cross-sections (thickness: 3.0 mm, TE: 15 ms and TR: 1349 ms, 399 × 399 field of view, and 534 × 493 matrix).

2.3. Image analysis

The PMCSA was calculated on the transverse angled sections through the whole images by drawing outlines. The PMT was measured by most hypertrophied piriformis muscle (Fig. 1A, B). We performed test-retest reliability assessments by having the same observer measure PMCSA and PMT on separate occasions, and we calculated the intra-class correlation coefficient (ICC). The ICC values were high, indicating excellent test-retest reliability. Moreover, we carried out inter-rater reliability analyses with 2 independent observers, and again found high ICC values. These results suggest that our measurement methods are consistent and reliable, thereby minimizing the impact of observer bias.

Figure 1.

Transverse turbo spin echo T1W H-MR images of the (A) piriformis muscle cross-sectional area (PMCSA) and (B) piriformis muscle thickness (PMT). H-MR = hip magnetic resonance, PMCSA = piriformis muscle cross-sectional area, PMT = piriformis muscle thickness.

2.4. Statistical analyses

Data are expressed as mean ± standard deviation or number of cases. Differences in demographic characteristics between control and PS groups were assessed using student t test. Receiver operating characteristic (ROC) curves were used to compare and illustrate the diagnostic validity of PMT and PMCSA. The quality of each diagnostic parameter was assessed based on sensitivity, specificity, and area under the curve (AUC). P values <.05 were considered significant. SPSS version 22.0 (IBM SPSS) was used for statistical analysis.

3. Results

The average PMCSA was 564.36 ± 121.61 mm² in the healthy group and 736.88 ± 168.87 mm² in the PS group (Table 1). The mean PMT was 13.83 ± 2.61 mm in the control group and 15.99 ± 2.34 mm in the PS group. PS patients had significantly higher PMCSA (P≤.001) and PMT (P≤.001; Table 1). Regarding the validity of both PMCSA and PMT as predictors of PS, ROC curve analysis showed that the best cutoff point for the PMCSA was 611.67 mm², with sensitivity: 75.0%, specificity: 75.0%, and the AUC of 0.81 (95% CI 0.68–0.94; Table 2, Fig. 2). The best cutoff point of the PMT was 14.24 mm, with 70.0% sensitivity, 70.0% specificity, and the AUC of 0.78 (95% CI 0.63–0.93; Table 3, Fig. 2).

Table 2.

Sensitivity and specificity of each cutoff value of the PMCSA.

PMCSA (mm2)	Sensitivity (%)	Specificity (%)
387.21	100	0
442.92	100	15.0
492.02	95.0	40.0
611.67*	75.0	75.0
706.69	60.0	85.0
813.80	30.0	95.0

PMCSA = piriformis muscle cross-sectional area, ROC = receiver operating characteristic.

^* The optimal cutoff value on the receiver operating characteristic (ROC) curve.

Figure 2.

Receiver operating characteristic (ROC) curve of PMCSA and PMT for prediction of PS. The best cut off point of PMT was 14.24 mm versus 611.67 mm² of PMCSA, with sensitivity 70.0% versus 75.0%, specificity 70.0% versus 75.0% and AUC 0.78 versus 0.81, respectively. PMCSA AUC (95% CI) = 0.81 (0.68–0.94); PMT AUC (95% CI) = 0.78 (0.63–0.93). AUC = area under the curve, PMT = piriformis muscle thickness, PMCSA = piriformis muscle cross-sectional area, PS = piriformis syndrome, ROC = receiver operating characteristic.

Table 3.

Sensitivity and specificity of each cutoff value of the PMT.

PMT (mm)	Sensitivity (%)	Specificity (%)
7.92	100	0
12.39	100	25.0
13.33	95.0	50.0
14.24*	70.0	70.0
15.97	45.0	85.0
20.28	10.0	100

PMT = piriformis muscle thickness, ROC = receiver operating characteristic.

^* The most suitable cutoff value on the receiver operating characteristic (ROC) curve.

4. Discussion

This study demonstrates the association of PMCSA and PS. PS patients had significantly higher PMCSA than healthy group. In my research, the optimal cutoff point for PMCSA was 611.67 mm², with 75.0% sensitivity, 75.0% specificity, and AUC of 0.81. And the most suitable cutoff value of the PMT was 14.24 mm, with 70.0% sensitivity, 70.0% specificity, and AUC of 0.78. I hope this value will be the standard about both PMCSA and PMT, because of there is no study about both PMT and PMCSA’s optimal cutoff value. I have demonstrated that PMCSA and PMT were both significantly associated with PS, with PMCSA being a more sensitive measurement parameter.

The PS is an uncommon entrapment nerve disorder in which the sciatic neuropathy is compromised by an abnormal PM. In clinical practice, evaluation of PS patients is based on performing electrodiagnostic studies and physical examination and obtaining the past history.^[4,7,15] However, these information are sometimes insufficient to identify PM abnormality.

H-MRI has a major role in the assessment of abnormalities of the skeletal muscles, especially changes in muscle morphology due to traumatic, inflammatory, degenerative, and neurologic disorders.^[10] The usefulness of MRI for evaluation of the PM in PS has been reported. Al-Shaikh et al^[8] have reported that H-MRI demonstrates hypertrophy of the PM, an anatomical abnormality of the PM or the course of the sciatic nerve signal change, and this change affects their anatomical relationships. Lee et al^[11] have also reported that H-MRI can be an important noninvasive diagnostic image modality, typically revealing an enlarged PM.

Therefore, H-MRI can differentiate PS effectively from other possible causes of sciatica and lower lumbar pain, such as mass lesions in the region of the PM, lumbar spinal stenosis, and lumbar disc herniation. The PMT is a valuable morphological method for the diagnosis of PS. Previous researches analyzed the PMT using 1 single technique at the “halfway” of the PM. However, an asymmetrical inflammatory thickening or partial atrophy of the PM can occur anywhere. Therefore, measurement mistakes can frequently occur. Russell et al^[16] have reported that a population of 100 asymptomatic individuals (200 buttocks) to demonstrate that over 90% of cases showed asymmetric PMT ranging from 3.0 to 8.0 mm. However, there is no research analyzing the anatomical exact basis of asymmetric inflammatory PM hypertrophy.

Unlike PMT, PMCSA measures the entire cross-sectional area of the PM, so this measurement error does not occur in the cross-sectional area of the PM. Therefore, we designed PMCSA as a new imaging parameter to evaluate asymmetric hypertrophy of PM. I hypothesized that PMCSA is an important morphological parameter for PM hypertrophy. Therefore, I compared PMCSA and PMT between PS subjects and healthy individuals using H-MRI.

This study demonstrates the association of PMCSA and PS. PS patients had significantly higher PMCSA than healthy group. We think this value will be the standard about both PMT and PMCSA, because of there is no study about both PMT and PMCSA’s optimal cutoff value. We have demonstrated that PMCSA and PMT were both significantly associated with PS, with PMCSA being a more sensitive morphological diagnostic method. We insist that the PMCSA could be an objective measurement parameter to assess PS. In our study, the PMCSA was measured from turbo spin echo transverse H-MR T1W image.

The present research had several limitations. First, it included only a small number of patients. Second, there might be some errors related with measuring the PMCSA and PMT on H-MRI. Even though, I maintained a good quality of morphologic measurement in the turbo spin echo transverse T1W H-MR images that best showed the piriformis muscle, the single H-MRI slice we measured the PMCSA and PMT can be inhomogeneous because of differences in the cutting phase in H-MR images resulting from individual anatomic variations and technical limit. Third, PS have multiple causes, including muscle spasm in the PM, either because of irritation in the PM itself, or irritation of a nearby anatomic structure such as the hip or sacroiliac joint or tightening of the leg muscle, in response to spasm or injury.^[17–26] However, we only focused on PM.^[1,4] In spite of these limitations, this is the very first research to demonstrate the relationship of PMCSA with PS. The PMCSA is a simple and reliable measurement method with high sensitive value to evaluate PS.

5. Conclusion

PMCSA is a new, objective and useful morphological parameter for assessing PS. I think that this adjuvant measurement method will be helpful to assess patients with PS.

Acknowledgments

Author thanks the International St. Mary’s Hospital.

Author contributions

Conceptualization: Young Uk Kim.

Data curation: Young Uk Kim.

Formal analysis: Young Uk Kim.

Investigation: Young Uk Kim.

Methodology: Young Uk Kim.

Project administration: Young Uk Kim.

Resources: Young Uk Kim.

Software: Young Uk Kim.

Supervision: Young Uk Kim.

Validation: Young Uk Kim.

Visualization: Young Uk Kim.

Writing – original draft: Young Uk Kim.

Writing – review & editing: Chaewan Lim, Hyung-Bok Park.