- Email: [email protected]
Piriformis syndrome: Correlation of muscle morphology with symptoms and signs
2036
Piriformis Syndrome: Correlation of Muscle Morphology With Symptoms and Signs Norman A. Broadhurst, MD, PhD, D. Neil Simmons, MD, Malcolm J. Bond, PhD ABSTRACT. Broadhurst NA, Simmons DN, Bond MJ. Piriformis syndrome: correlation of muscle morphology with symptoms and signs. Arch Phys Med Rehabil 2004;85:2036-9. Objective: To determine the relation of symptoms and pain provocation tests to abnormal piriformis morphology among people with chronic buttock pain. Design: Each of 2 clinical symptoms and 2 clinical signs were compared with the abnormal morphology found on ultrasound. The pain-free side was used as an internal control. Setting: A tertiary referral center. Participants: A series of 27 consecutive patients (26 women, 1 man; average age, 48y) with chronic low lumbosacral or buttock pain, who presented to a musculoskeletal clinic over a 12-month period, underwent ultrasound assessment of piriformis muscle morphology. Four patients were excluded because their body mass index was in excess of 30kg/m2. Intervention: The symptomatic piriformis muscle was injected with bupivacaine after pain was assessed on a visual analog scale (VAS), using the resisted abduction test. Main Outcome Measure: A 70% reduction of pain on the VAS was considered positive for pain in the piriformis muscle. Results: Odds ratios (ORs) and 95% confidence intervals were calculated comparing each of the signs and symptoms with normal morphology. The highest ORs were found for pain on walking up inclines (10.8), referred pain (5.3), and pain on needling the piriformis muscle (6.0). Conclusions: This study did not provide a criterion standard for the diagnosis of piriformis syndrome, but it did support the syndrome as a contributing factor in chronic buttock pain and very low back pain. Key Words: Buttocks; Low back pain; Rehabilitation; Ultrasonography. © 2004 by the American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation HE PIRIFORMIS MUSCLE has been disputed as a cause of very low back or buttock pain since the idea was first T postulated by Yeoman in 1928. Since then, several studies 1
have attempted to verify that the piriformis muscle is 1 source of such pain. Hallin2 suggested that patients with piriformis pathology had buttock pain that was exacerbated by sitting,
From the Department of Musculoskeletal Medicine, School of Medicine (Broadhurst), and Centre for Epidemiology and Biostatistics (Bond), Flinders University, Adelaide; and Jones and Partners Radiology, Adelaide (Simmons), Australia. No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit upon the author(s) or upon any organization with which the author(s) is/are associated. Reprint requests to Norman A. Broadhurst, MD, PhD, Dept of Musculoskeletal Medicine, School of Medicine, Flinders University, GPO Box 2100, Adelaide 5001, Australia, e-mail: [email protected]. 0003-9993/04/8512-8956$30.00/0 doi:10.1016/j.apmr.2004.02.017
Arch Phys Med Rehabil Vol 85, December 2004
occasionally with radiation of the pain, but no pain with the straight-leg raising test. Others3,4 have palpated the muscle to reproduce patients’ pain and then proceeded to provide pain relief by injecting local anesthetic into the body of the muscle. Beatty5 used forced adduction of the affected flexed hip, with the patient lying on the painful side, to produce buttock pain as a diagnostic criterion of piriformis syndrome. Fishman et al6 found that prolongation of the H-reflex with the hip flexed, adducted, and internally rotated (FAIR position) identified a piriformis problem, and proposed that such an effect resulted from the muscle pressing on the sciatic nerve. Identification of the symptomatic muscle, using various imaging modalities, has been anecdotal. Others7-9 did not find piriformis muscle abnormality in any form. When muscle pain is refractory to conservative measures, such as stretching, massage, and injections, the piriformis muscle has been separated from its insertion at the greater trochanter, resulting in significant pain relief even when there is no consistent abnormal morphology in the muscle as determined by computed tomography (CT) imaging.9,10 Despite some support for a diagnosis of piriformis syndrome, its existence must be considered a clinical entity. There are no studies reporting an established pathology in enough patients to satisfactorily validate the existence of piriformis pathology. Benson and Schutzer9 performed CT scans on patients with suspected posttraumatic piriformis syndrome but concluded that the scans were not helpful. Thus, there is ongoing skepticism in the musculoskeletal community, as evidenced in an article by Silver and Leadbetter11 in which they assessed the opinions of 75 physiatrists in the United States, that the diagnosis is made clinically and the condition remains controversial. This debate is likely to continue until there are enough clinical trials to clarify the sources of buttock pain in patients with this symptom complex. We designed this observational study to determine whether patients with low lumbosacral buttock pain had any piriformis muscle abnormalities and to correlate the diagnostic findings with 2 symptoms and 2 clinical signs commonly associated with patients who have buttock pain. METHODS Patients for the study were selected from people who presented with lumbosacral and buttock pain at a tertiary musculoskeletal referral center. All patients had not improved with conservative management, and none had been suspected of having piriformis pain. These patients, selected over a 1-year period, had a mixture of symptoms that included deep-seated buttock pain, which was made worse by sitting for short periods and by walking up inclines or stairs, and that often referred down the back of the thigh. All patients had pain for more than 6 months, had good pain-free lumbar movement, had no history of radicular pain, and were pain free on stressing the sacrospinous and sacrotuberous ligaments per rectum. All study patients experienced pain with pressure at the intersection of the piriformis muscle and sciatic nerve when lying on their side
2037
PIRIFORMIS SYNDROME, Broadhurst Table 1: Statistical Results Comparing Signs and Symptoms With Morphology (Nⴝ27) Yes
No
Abnormal Morphology
Normal Morphology
Pain Reproduction
n
%
n
%
Pain
No Pain
Pain
No Pain
OR
Up inclines Referred to thigh Resisted abduction Needling muscle
23 20 25 24
86 76 95 90
4 7 2 3
14 24 5 10
18 16 18 18
1 4 1 1
5 3 7 6
3 4 1 2
10.8 5.3 2.6 6.0
with the ipsilateral hip flexed, adducted, and internally rotated (FAIR test). Assessment of both piriformis muscles and surrounding structures by ultrasound was selected as the imaging modality because of the availability of a high-resolution instrument to detect morphologic changes in soft tissue, and because of the expertise of a particular radiologist with a special interest in musculoskeletal ultrasound. Further comparison with magnetic resonance imaging was not considered because of the cost. Treatment regimens and outcomes were not a part of this study. Because there are no data available on the morphology of a painful muscle, we decided to compare the morphology of the asymptomatic surrounding muscles on both buttocks with that of the painful symptomatic piriformis muscle. The muscle on the symptomatic side was considered to be abnormal if it differed in size and/or texture from the muscle on the asymptomatic side. The subjective nature of the assessment is acknowledged, as is imaging reporting in general. Further details are reported below. Once the muscle on the symptomatic side was delineated, a total of 10mL of local anesthetic was injected into the origin, insertion, and the body of the muscle to ensure anesthesia. With the patient lying in a supine position, pain was assessed by asking the patient to indicate the level of pain felt on resisted abduction of the adducted flexed hip and with the femur parallel to the floor. A visual analog scale (VAS) score was obtained before and within 20 minutes of the injection. Patients with a body mass index greater than 30kg/m2 were excluded because of the technical difficulties involved in imaging the muscle. There were no other exclusion criteria. Several symptoms are listed as being features of the syndrome, and when we compared them with abnormal morphology, only 2 had an odds ratio (OR) greater than 2.12 These symptoms were pain going up inclines or stairs and pain referred to the posterior thigh. Findings were similar for the 2 signs: pain on resisted abduction of the adducted flexed hip and pain produced by needling the muscle. Diminishing the patient’s pain by 70% after the intramuscular injection of a local anesthetic into the muscle was considered to be subjective evidence that implicated the piriformis as the source of pain. Because of the absence of a criterion standard for the diagnosis of piriformis syndrome, statistical calculations involved ORs (⫾95% confidence interval [CI]).13 Exact Mann-Whitney U tests were calculated for continuous variables. Ultrasound Scanning Technique We used an Acuson Sequoia ultrasounda for all scans, which were done by the musculoskeletal radiologist who always examined both buttocks—the asymptomatic side first, with the patient in the prone position. Each sciatic nerve was followed transversely and longitudinally, from the proximal thigh to the piriformis muscle. The thigh, with the knee flexed at 90°, was
internally and externally rotated to demonstrate the movement of the sciatic nerve and surrounding structures. This permitted assessment of the soft tissues in the region of pain, including the overlying gluteus maximus. The underlying external hip rotators (quadratus femoris, inferior gemellus, obturator internus, superior gemellus) were also assessed for structural abnormalities. The relation of the sciatic nerve to the piriformis was noted. The piriformis muscle was visually assessed for texture and size, in both longitudinal and transverse planes, to ensure that the surrounding muscle morphology was the same on both buttocks. Doppler ultrasound helped identify the sciatic nerve, inferior gluteal artery, and nerve complex. Under ultrasound guidance, a 22-gauge spinal needle was inserted until its tip was in the proximal piriformis, away from the neurovascular complex. Needle-tip position was confirmed with a test injection of bupivacaine, followed by 10mL of bupivacaine injected into the whole muscle. The material could be seen spreading along the plane of the piriformis. A similar technique was used for the musculotendinous junction of the piriformis, near the apex of the posterior part of the greater trochanter. Positioning for needle placement was aided by rotation of the femur under dynamic ultrasound to identify the insertion of the piriformis. RESULTS Twenty-seven patients were included in the study (26 women, 1 man; mean age, 48y). In 18 patients (67%), an abnormal piriformis muscle was identified by comparison with the asymptomatic side using ultrasound. The median age of those with abnormal morphology was 46 years and it was 52 years for patients without such abnormality. The average ages in the 2 groups did not differ significantly (P⫽.58). Data for all presenting symptoms and signs are shown in table 1 and indicate that 76% of patients presented with all 4 symptoms and signs. Abnormal morphology findings on ultrasound compared with selected symptoms and signs are also shown in table 1. The best discriminating symptom for an abnormal piriformis muscle was pain experienced while walking on inclines (OR-10.8). This means that a patient presenting with this symptom was nearly 11 times more likely to have abnormal morphology in the piriformis muscle than a patient who did not have pain while ascending inclines. Patients with referred pain had an OR of 5.3. Pain reproduction in needling, and pain on resisted abduction of the flexed hip, had ORs of 6.0 and 3.2, respectively, for clinical signs. Using a VAS, the pain on resisted abduction diminished by 70% after a local anesthetic was given. This occurred in all patients, including the 9 who had bilateral normal piriformis texture. Scarring and fibrosis were seen in 3 muscles, while 15 patients had enlarged and diffuse muscle bulk. Three aberrant sciatic nerves were identified, 1 passing through an asymptomatic muscle and 2 passing over the asymptomatic muscle. One patient had a markedly thickened sacrospinous-sacrotu-
Arch Phys Med Rehabil Vol 85, December 2004
2038
PIRIFORMIS SYNDROME, Broadhurst
Fig 1. (A) The normal asymptomatic muscle (left) and the swollen symptomatic muscle (right). (B) The normal asymptomatic muscle (right) and the sclerotic and atrophic symptomatic muscle (left).
berous ligament complex on the symptomatic side, symmetric muscle bulk bilaterally, and an aberrant sciatic nerve passing above the muscle on the asymptomatic side. Figure 1 shows the type of abnormal morphology seen in the symptomatic side compared with the asymptomatic side. DISCUSSION We recognize that this is a comparatively small sample and hence CIs for the ORs are wide. Nonetheless, we believe that the absolute size of the ORs warrants confidence in the findings. Our study found evidence that abnormal morphology may not be seen in one third of patients with buttock pain, although abnormal morphology was found in two thirds of symptomatic muscles when compared with the pain-free side. The signs and symptoms regarded as features of piriformis syndrome have variable ORs whose CIs all straddle 1.0 (table 1). Notably, none of the patients had all of the symptoms and signs. This highlights the central problem in the diagnosis of this condition, namely, the lack of a criterion standard. Until such a standard can be established, the diagnosis of piriformis syndrome remains controversial. Consequently, attention should be given to determining what, and if, other pelvic musculoskeletal and visceral structures might be contributing to buttock pain. This would entail rectal or vaginal examination to assess the sacrospinous-sacrotuberous ligament complex and possibly the sacrococcygeal junction, as we did in this study. Our study identified some abnormalities in the piriformis muscle on the affected side; with the blocking of pain by 70%, it would seem reasonable to conclude that this muscle is the cause of pain. However, there is no direct visualization because there is no arthroscopy to correlate clinical findings with the criterion standard (ie, visualization). The dilemma facing clinicians is how to be certain that buttock pain is not referred pain from the lumbar spine, but originates in buttock soft tissues. Most soft tissue symptoms resolve with time, but some do not. Although the piriformis muscle may be considered the primary source of pain, there are other external rotators that could be contributors. Is it because
Arch Phys Med Rehabil Vol 85, December 2004
the piriformis muscle is more readily delineated, larger and more superficial than the others, that it has been singled out as the cause of the pain? In treating patients with buttock pain, standard treatment of reassurance, massage, stretching, and even injection of the tender points can be efficacious. However, it is necessary to avoid a specific diagnosis when one is insufficiently supported. Specific causes of low back pain (LBP) are not easily identified; hence, a diagnosis of nonspecific LBP has become well established. This avoids a definitive comment about the particular tissue involved. It would seem reasonable to diagnose buttock pain as nonspecific and apply the same philosophy in its treatment as is applied with LBP. CONCLUSIONS Patients presenting with lumbosacral or buttock pain, and who have pain when ascending inclines, referred pain to the thigh in the symptomatic side, and reproduction of the pain on needling, provide tacit support for recognizing the piriformis muscle as a contributing factor to the pain. The quest for a criterion standard must continue. Our study’s results have demonstrated abnormal piriformis morphology in a significant number of patients with chronic LBP and buttock pain who failed to improve with conservative treatment. References 1. Yeoman W. The relation of arthritis of the sacroiliac joint to sciatica. Lancet 1928;2:1119-22. 2. Hallin RP. Sciatic pain and the piriformis muscle. Postcard Med 1983;74:69-72. 3. Wyant GM. Chronic pain syndromes and their treatment. III. The piriformis syndrome. Can Anaesth Soc J 1979;26:305-8. 4. Durrani Z, Winnie AP. Piriformis muscle syndrome: an underdiagnosed cause of sciatica. J Pain Symptom Manage 1991;6:374-9. 5. Beatty RA. The piriformis muscle syndrome: a simple diagnostic manoeuvre. Neurosurgery 1994;34:512-4. 6. Fishman LM, Dombi GW, Michaelsen C, et al. Piriformis syndrome: diagnosis, treatment and outcome—a ten-year study. Arch Phys Med Rehabil 2002;83:295-301.
PIRIFORMIS SYNDROME, Broadhurst
7. Karl RD Jr, Yedinak MA, Hartshorne MF, et al. Scintigraphic appearance of the piriformis muscle syndrome. Clin Nucl Med 1985;10:361-3. 8. Jankiewicz JJ, Hennrikus WL, Houkom JA. The appearance of the piriformis muscle syndrome in computed tomography and MRI. A case report and review of the literature. Clin Orthop 1991; Jan(262):205-9. 9. Benson ER, Schutzer SF. Posttraumatic piriformis: diagnosis and results of operative treatment. J Bone Joint Surg Am 1999;81: 941-9. 10. Solhein LF, Siewers P, Paus B. The piriformis muscle syndrome. Acta Orthop Scand 1981;52:73-5.
2039
11. Silver JK, Leadbetter WB. Piriformis syndrome: assessment of current practice and literature review. Orthopaedics 1998;21: 1133-5. 12. Benzon HT, Katz JA, Benzon HA, Iqbal MS. Piriformis syndrome: atomic considerations, a new injection technique, and a review of the literature. Anesthesiology 2003;98:1442-8. 13. Mehta CR, Patel NR, Gray R. On computing an exact confidence interval for the common odds ratio in several 2⫻2 contingency tables. J Am Stat Assoc 1985;80:969-73. Supplier a. Siemens, 1230 Shorebird Way, Mountain View, CA 94043-1344.
Arch Phys Med Rehabil Vol 85, December 2004
Piriformis Syndrome: Correlation of Muscle Morphology With Symptoms and Signs Norman A. Broadhurst, MD, PhD, D. Neil Simmons, MD, Malcolm J. Bond, PhD ABSTRACT. Broadhurst NA, Simmons DN, Bond MJ. Piriformis syndrome: correlation of muscle morphology with symptoms and signs. Arch Phys Med Rehabil 2004;85:2036-9. Objective: To determine the relation of symptoms and pain provocation tests to abnormal piriformis morphology among people with chronic buttock pain. Design: Each of 2 clinical symptoms and 2 clinical signs were compared with the abnormal morphology found on ultrasound. The pain-free side was used as an internal control. Setting: A tertiary referral center. Participants: A series of 27 consecutive patients (26 women, 1 man; average age, 48y) with chronic low lumbosacral or buttock pain, who presented to a musculoskeletal clinic over a 12-month period, underwent ultrasound assessment of piriformis muscle morphology. Four patients were excluded because their body mass index was in excess of 30kg/m2. Intervention: The symptomatic piriformis muscle was injected with bupivacaine after pain was assessed on a visual analog scale (VAS), using the resisted abduction test. Main Outcome Measure: A 70% reduction of pain on the VAS was considered positive for pain in the piriformis muscle. Results: Odds ratios (ORs) and 95% confidence intervals were calculated comparing each of the signs and symptoms with normal morphology. The highest ORs were found for pain on walking up inclines (10.8), referred pain (5.3), and pain on needling the piriformis muscle (6.0). Conclusions: This study did not provide a criterion standard for the diagnosis of piriformis syndrome, but it did support the syndrome as a contributing factor in chronic buttock pain and very low back pain. Key Words: Buttocks; Low back pain; Rehabilitation; Ultrasonography. © 2004 by the American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation HE PIRIFORMIS MUSCLE has been disputed as a cause of very low back or buttock pain since the idea was first T postulated by Yeoman in 1928. Since then, several studies 1
have attempted to verify that the piriformis muscle is 1 source of such pain. Hallin2 suggested that patients with piriformis pathology had buttock pain that was exacerbated by sitting,
From the Department of Musculoskeletal Medicine, School of Medicine (Broadhurst), and Centre for Epidemiology and Biostatistics (Bond), Flinders University, Adelaide; and Jones and Partners Radiology, Adelaide (Simmons), Australia. No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit upon the author(s) or upon any organization with which the author(s) is/are associated. Reprint requests to Norman A. Broadhurst, MD, PhD, Dept of Musculoskeletal Medicine, School of Medicine, Flinders University, GPO Box 2100, Adelaide 5001, Australia, e-mail: [email protected]. 0003-9993/04/8512-8956$30.00/0 doi:10.1016/j.apmr.2004.02.017
Arch Phys Med Rehabil Vol 85, December 2004
occasionally with radiation of the pain, but no pain with the straight-leg raising test. Others3,4 have palpated the muscle to reproduce patients’ pain and then proceeded to provide pain relief by injecting local anesthetic into the body of the muscle. Beatty5 used forced adduction of the affected flexed hip, with the patient lying on the painful side, to produce buttock pain as a diagnostic criterion of piriformis syndrome. Fishman et al6 found that prolongation of the H-reflex with the hip flexed, adducted, and internally rotated (FAIR position) identified a piriformis problem, and proposed that such an effect resulted from the muscle pressing on the sciatic nerve. Identification of the symptomatic muscle, using various imaging modalities, has been anecdotal. Others7-9 did not find piriformis muscle abnormality in any form. When muscle pain is refractory to conservative measures, such as stretching, massage, and injections, the piriformis muscle has been separated from its insertion at the greater trochanter, resulting in significant pain relief even when there is no consistent abnormal morphology in the muscle as determined by computed tomography (CT) imaging.9,10 Despite some support for a diagnosis of piriformis syndrome, its existence must be considered a clinical entity. There are no studies reporting an established pathology in enough patients to satisfactorily validate the existence of piriformis pathology. Benson and Schutzer9 performed CT scans on patients with suspected posttraumatic piriformis syndrome but concluded that the scans were not helpful. Thus, there is ongoing skepticism in the musculoskeletal community, as evidenced in an article by Silver and Leadbetter11 in which they assessed the opinions of 75 physiatrists in the United States, that the diagnosis is made clinically and the condition remains controversial. This debate is likely to continue until there are enough clinical trials to clarify the sources of buttock pain in patients with this symptom complex. We designed this observational study to determine whether patients with low lumbosacral buttock pain had any piriformis muscle abnormalities and to correlate the diagnostic findings with 2 symptoms and 2 clinical signs commonly associated with patients who have buttock pain. METHODS Patients for the study were selected from people who presented with lumbosacral and buttock pain at a tertiary musculoskeletal referral center. All patients had not improved with conservative management, and none had been suspected of having piriformis pain. These patients, selected over a 1-year period, had a mixture of symptoms that included deep-seated buttock pain, which was made worse by sitting for short periods and by walking up inclines or stairs, and that often referred down the back of the thigh. All patients had pain for more than 6 months, had good pain-free lumbar movement, had no history of radicular pain, and were pain free on stressing the sacrospinous and sacrotuberous ligaments per rectum. All study patients experienced pain with pressure at the intersection of the piriformis muscle and sciatic nerve when lying on their side
2037
PIRIFORMIS SYNDROME, Broadhurst Table 1: Statistical Results Comparing Signs and Symptoms With Morphology (Nⴝ27) Yes
No
Abnormal Morphology
Normal Morphology
Pain Reproduction
n
%
n
%
Pain
No Pain
Pain
No Pain
OR
Up inclines Referred to thigh Resisted abduction Needling muscle
23 20 25 24
86 76 95 90
4 7 2 3
14 24 5 10
18 16 18 18
1 4 1 1
5 3 7 6
3 4 1 2
10.8 5.3 2.6 6.0
with the ipsilateral hip flexed, adducted, and internally rotated (FAIR test). Assessment of both piriformis muscles and surrounding structures by ultrasound was selected as the imaging modality because of the availability of a high-resolution instrument to detect morphologic changes in soft tissue, and because of the expertise of a particular radiologist with a special interest in musculoskeletal ultrasound. Further comparison with magnetic resonance imaging was not considered because of the cost. Treatment regimens and outcomes were not a part of this study. Because there are no data available on the morphology of a painful muscle, we decided to compare the morphology of the asymptomatic surrounding muscles on both buttocks with that of the painful symptomatic piriformis muscle. The muscle on the symptomatic side was considered to be abnormal if it differed in size and/or texture from the muscle on the asymptomatic side. The subjective nature of the assessment is acknowledged, as is imaging reporting in general. Further details are reported below. Once the muscle on the symptomatic side was delineated, a total of 10mL of local anesthetic was injected into the origin, insertion, and the body of the muscle to ensure anesthesia. With the patient lying in a supine position, pain was assessed by asking the patient to indicate the level of pain felt on resisted abduction of the adducted flexed hip and with the femur parallel to the floor. A visual analog scale (VAS) score was obtained before and within 20 minutes of the injection. Patients with a body mass index greater than 30kg/m2 were excluded because of the technical difficulties involved in imaging the muscle. There were no other exclusion criteria. Several symptoms are listed as being features of the syndrome, and when we compared them with abnormal morphology, only 2 had an odds ratio (OR) greater than 2.12 These symptoms were pain going up inclines or stairs and pain referred to the posterior thigh. Findings were similar for the 2 signs: pain on resisted abduction of the adducted flexed hip and pain produced by needling the muscle. Diminishing the patient’s pain by 70% after the intramuscular injection of a local anesthetic into the muscle was considered to be subjective evidence that implicated the piriformis as the source of pain. Because of the absence of a criterion standard for the diagnosis of piriformis syndrome, statistical calculations involved ORs (⫾95% confidence interval [CI]).13 Exact Mann-Whitney U tests were calculated for continuous variables. Ultrasound Scanning Technique We used an Acuson Sequoia ultrasounda for all scans, which were done by the musculoskeletal radiologist who always examined both buttocks—the asymptomatic side first, with the patient in the prone position. Each sciatic nerve was followed transversely and longitudinally, from the proximal thigh to the piriformis muscle. The thigh, with the knee flexed at 90°, was
internally and externally rotated to demonstrate the movement of the sciatic nerve and surrounding structures. This permitted assessment of the soft tissues in the region of pain, including the overlying gluteus maximus. The underlying external hip rotators (quadratus femoris, inferior gemellus, obturator internus, superior gemellus) were also assessed for structural abnormalities. The relation of the sciatic nerve to the piriformis was noted. The piriformis muscle was visually assessed for texture and size, in both longitudinal and transverse planes, to ensure that the surrounding muscle morphology was the same on both buttocks. Doppler ultrasound helped identify the sciatic nerve, inferior gluteal artery, and nerve complex. Under ultrasound guidance, a 22-gauge spinal needle was inserted until its tip was in the proximal piriformis, away from the neurovascular complex. Needle-tip position was confirmed with a test injection of bupivacaine, followed by 10mL of bupivacaine injected into the whole muscle. The material could be seen spreading along the plane of the piriformis. A similar technique was used for the musculotendinous junction of the piriformis, near the apex of the posterior part of the greater trochanter. Positioning for needle placement was aided by rotation of the femur under dynamic ultrasound to identify the insertion of the piriformis. RESULTS Twenty-seven patients were included in the study (26 women, 1 man; mean age, 48y). In 18 patients (67%), an abnormal piriformis muscle was identified by comparison with the asymptomatic side using ultrasound. The median age of those with abnormal morphology was 46 years and it was 52 years for patients without such abnormality. The average ages in the 2 groups did not differ significantly (P⫽.58). Data for all presenting symptoms and signs are shown in table 1 and indicate that 76% of patients presented with all 4 symptoms and signs. Abnormal morphology findings on ultrasound compared with selected symptoms and signs are also shown in table 1. The best discriminating symptom for an abnormal piriformis muscle was pain experienced while walking on inclines (OR-10.8). This means that a patient presenting with this symptom was nearly 11 times more likely to have abnormal morphology in the piriformis muscle than a patient who did not have pain while ascending inclines. Patients with referred pain had an OR of 5.3. Pain reproduction in needling, and pain on resisted abduction of the flexed hip, had ORs of 6.0 and 3.2, respectively, for clinical signs. Using a VAS, the pain on resisted abduction diminished by 70% after a local anesthetic was given. This occurred in all patients, including the 9 who had bilateral normal piriformis texture. Scarring and fibrosis were seen in 3 muscles, while 15 patients had enlarged and diffuse muscle bulk. Three aberrant sciatic nerves were identified, 1 passing through an asymptomatic muscle and 2 passing over the asymptomatic muscle. One patient had a markedly thickened sacrospinous-sacrotu-
Arch Phys Med Rehabil Vol 85, December 2004
2038
PIRIFORMIS SYNDROME, Broadhurst
Fig 1. (A) The normal asymptomatic muscle (left) and the swollen symptomatic muscle (right). (B) The normal asymptomatic muscle (right) and the sclerotic and atrophic symptomatic muscle (left).
berous ligament complex on the symptomatic side, symmetric muscle bulk bilaterally, and an aberrant sciatic nerve passing above the muscle on the asymptomatic side. Figure 1 shows the type of abnormal morphology seen in the symptomatic side compared with the asymptomatic side. DISCUSSION We recognize that this is a comparatively small sample and hence CIs for the ORs are wide. Nonetheless, we believe that the absolute size of the ORs warrants confidence in the findings. Our study found evidence that abnormal morphology may not be seen in one third of patients with buttock pain, although abnormal morphology was found in two thirds of symptomatic muscles when compared with the pain-free side. The signs and symptoms regarded as features of piriformis syndrome have variable ORs whose CIs all straddle 1.0 (table 1). Notably, none of the patients had all of the symptoms and signs. This highlights the central problem in the diagnosis of this condition, namely, the lack of a criterion standard. Until such a standard can be established, the diagnosis of piriformis syndrome remains controversial. Consequently, attention should be given to determining what, and if, other pelvic musculoskeletal and visceral structures might be contributing to buttock pain. This would entail rectal or vaginal examination to assess the sacrospinous-sacrotuberous ligament complex and possibly the sacrococcygeal junction, as we did in this study. Our study identified some abnormalities in the piriformis muscle on the affected side; with the blocking of pain by 70%, it would seem reasonable to conclude that this muscle is the cause of pain. However, there is no direct visualization because there is no arthroscopy to correlate clinical findings with the criterion standard (ie, visualization). The dilemma facing clinicians is how to be certain that buttock pain is not referred pain from the lumbar spine, but originates in buttock soft tissues. Most soft tissue symptoms resolve with time, but some do not. Although the piriformis muscle may be considered the primary source of pain, there are other external rotators that could be contributors. Is it because
Arch Phys Med Rehabil Vol 85, December 2004
the piriformis muscle is more readily delineated, larger and more superficial than the others, that it has been singled out as the cause of the pain? In treating patients with buttock pain, standard treatment of reassurance, massage, stretching, and even injection of the tender points can be efficacious. However, it is necessary to avoid a specific diagnosis when one is insufficiently supported. Specific causes of low back pain (LBP) are not easily identified; hence, a diagnosis of nonspecific LBP has become well established. This avoids a definitive comment about the particular tissue involved. It would seem reasonable to diagnose buttock pain as nonspecific and apply the same philosophy in its treatment as is applied with LBP. CONCLUSIONS Patients presenting with lumbosacral or buttock pain, and who have pain when ascending inclines, referred pain to the thigh in the symptomatic side, and reproduction of the pain on needling, provide tacit support for recognizing the piriformis muscle as a contributing factor to the pain. The quest for a criterion standard must continue. Our study’s results have demonstrated abnormal piriformis morphology in a significant number of patients with chronic LBP and buttock pain who failed to improve with conservative treatment. References 1. Yeoman W. The relation of arthritis of the sacroiliac joint to sciatica. Lancet 1928;2:1119-22. 2. Hallin RP. Sciatic pain and the piriformis muscle. Postcard Med 1983;74:69-72. 3. Wyant GM. Chronic pain syndromes and their treatment. III. The piriformis syndrome. Can Anaesth Soc J 1979;26:305-8. 4. Durrani Z, Winnie AP. Piriformis muscle syndrome: an underdiagnosed cause of sciatica. J Pain Symptom Manage 1991;6:374-9. 5. Beatty RA. The piriformis muscle syndrome: a simple diagnostic manoeuvre. Neurosurgery 1994;34:512-4. 6. Fishman LM, Dombi GW, Michaelsen C, et al. Piriformis syndrome: diagnosis, treatment and outcome—a ten-year study. Arch Phys Med Rehabil 2002;83:295-301.
PIRIFORMIS SYNDROME, Broadhurst
7. Karl RD Jr, Yedinak MA, Hartshorne MF, et al. Scintigraphic appearance of the piriformis muscle syndrome. Clin Nucl Med 1985;10:361-3. 8. Jankiewicz JJ, Hennrikus WL, Houkom JA. The appearance of the piriformis muscle syndrome in computed tomography and MRI. A case report and review of the literature. Clin Orthop 1991; Jan(262):205-9. 9. Benson ER, Schutzer SF. Posttraumatic piriformis: diagnosis and results of operative treatment. J Bone Joint Surg Am 1999;81: 941-9. 10. Solhein LF, Siewers P, Paus B. The piriformis muscle syndrome. Acta Orthop Scand 1981;52:73-5.
2039
11. Silver JK, Leadbetter WB. Piriformis syndrome: assessment of current practice and literature review. Orthopaedics 1998;21: 1133-5. 12. Benzon HT, Katz JA, Benzon HA, Iqbal MS. Piriformis syndrome: atomic considerations, a new injection technique, and a review of the literature. Anesthesiology 2003;98:1442-8. 13. Mehta CR, Patel NR, Gray R. On computing an exact confidence interval for the common odds ratio in several 2⫻2 contingency tables. J Am Stat Assoc 1985;80:969-73. Supplier a. Siemens, 1230 Shorebird Way, Mountain View, CA 94043-1344.
Arch Phys Med Rehabil Vol 85, December 2004