The piriformis syndrome is overdiagnosed

Confusion reigns in the literature because the term piriformis syndrome (PS) has been used to denote four different entities. Which of these, if any, warrant the designation PS? Each will be discussed in turn. EMG, electromyography; MRI, magnetic resonance imaging; PS, piriformis syndrome Lesions of the proximal sciatic nerve in the area of the sciatic notch may occur from endometriosis, tumors, hematomas, fibrosis, aneurysms, false aneurysms, or arteriovenous malformations. Some authors have diagnosed such patients as having PS. Since the piriformis muscle plays no role in these situations, such causes of sciatic neuropathy are best included under the rubric “proximal sciatic neuropathies.” It has long been suggested that the proximal sciatic nerve can be compressed by the piriformis muscle where it crosses the nerve, and that the frequent anatomical variations occurring here predispose to this. Does such a condition exist? This author proposes that, ideally, the following five criteria need to be fulfilled to define such a syndrome: (1) Presence of symptoms and signs of sciatic nerve damage. (2) Presence of electrophysiological evidence of sciatic nerve damage. Paraspinal muscle electromyography (EMG) must be normal, to help in excluding a radiculopathy. (3) Imaging of the lumbosacral nerve roots and of the paravertebral and pelvic areas must be normal to exclude radiculopathy, or lower lumbar or sacral plexus infiltration or damage. Imaging of the pelvis and sciatic notch must show the absence of mass lesions there. The significance of suspected abnormalities of the piriformis muscle seen on imaging is uncertain, as discussed later. (4) Surgical exploration of the proximal sciatic nerve should confirm an absence of mass lesions. Ideally, compression of the sciatic nerve by the piriformis muscle or associated fibrous bands should be identified. However, it can sometimes be difficult to recognize a compressed nerve. (5) Relief of symptoms and improvement in neurological abnormalities should follow surgical decompression. However, as in other situations of chronic nerve damage, decompression may not always lead to symptom relief. Further, as discussed later, surgical division of the piriformis muscle has been described, surprisingly, as relieving pain in patients with lumbosacral radiculopathies. The older descriptions of alleged PS antedate modern imaging techniques, so these patients are excluded from further discussion. A few patients meet some of the criteria for PS, and a few others very nearly meet the criteria. In one of the latter, the surgical finding was a hypertrophied piriformis muscle compressing the sciatic nerve.7 Three patients had bifid piriformis muscles compressing the lateral trunk of the sciatic nerve.2, 4 Two patients had nerve compression by fibrous bands associated with the piriformis muscle.4, 8 Benson and Shuster1 reported a series of patients with sciatic nerve lesions that they appropriately termed “post-traumatic PS.” Symptoms began after blows to the buttocks. Several had electromyographic (EMG) studies showing abnormalities in muscles supplied by the sciatic and inferior gluteal nerves. One patient had myositis ossificans of the piriformis muscle, confirmed at operation. Others had adhesions between the piriformis muscle, the sciatic nerve, and the roof of the sciatic notch. One patient had an anatomical anomaly of the sciatic nerve and piriformis muscle. All patients had a release of the piriformis tendon and their symptoms improved. There are many reports of patients with the primary symptom of buttock pain (often with “sciatica”) but no neurological deficits. In some, the symptoms followed buttock trauma. These patients do not meet the criteria outlined above. They are labeled as having PS based on the belief that their symptoms are due to impingement on the proximal sciatic nerve by the piriformis muscle. The core issue here is the likelihood of chronic, or chronically recurrent compression of a peripheral nerve producing pain but no manifestations of nerve fiber damage. In clear-cut compressive neuropathies, pain from nerve trunk involvement is almost always accompanied by sensory or motor symptoms, clinical deficits, and electrophysiological abnormalities. When this combination of features is absent (e.g., in the “disputed neurological” thoracic outlet syndrome and resistant tennis elbow syndrome), there is little convincing evidence that nerve trunks are involved in the genesis of symptoms. A frequent accompaniment of the buttock pain in this group of patients is sciatica. Most would accept a definition for this term as being pain radiating down the leg from the lower back, buttock, or hip. Notwithstanding its name, such pain is not a specific indicator of sciatic nerve involvement. The most frequent neurological cause is L-5 or S-1 radiculopathy; others include lower lumbar and sacral plexopathy and proximal sciatic neuropathies. Non-neurological causes are more frequent and include musculoskeletal abnormalities of the lumbosacral spine and hip, and pelvic disease. Much of the so-called evidence for this type of “PS” rests on a variety of physical signs said to demonstrate pinching of the sciatic nerve by the piriformis muscle during certain leg and hip maneuvers. These signs are all of doubtful validity in terms of specifically demonstrating compression of the sciatic nerve by the piriformis muscle. Several classic signs in medicine have been shown to have poor sensitivity and specificity; for instance, Tinel's and Phalen's signs for carpal tunnel syndrome, and Adson's maneuver for thoracic outlet syndrome. No sign said to indicate PS has been critically evaluated. Tenderness on deep palpation in the buttock is often found in patients with conditions such as lumbosacral radiculopathy, tumors or other masses at the sciatic notch, and posttraumatic scarring in this area. Tenderness therefore does not reliably indicate an abnormality of the piriformis muscle. Further so-called evidence that the piriformis muscle plays a role in these patients is the occurrence of pain relief following local anesthetic or corticosteroid injections into the piriformis muscle and sciatic notch area. Unfortunately, this does not elucidate the underlying pathology. Such injections will tend to relieve local symptoms regardless of the cause. Moreover, it is well established that nerve blocks distal to a nerve lesion can produce pain relief.5 Deep buttock injections have been shown to relieve pain in patients with lumbosacral radiculopathies and carcinomatous sacral root infiltration. One extraordinary study found that division of the piriformis muscle in patients with lumbosacral radiculopathies produced pain relief.6 Thus, improvement of pain from injections and even from surgical division of the piriformis muscle cannot be used as proof of sciatic nerve compression at that site. The reports of a swollen piriformis muscle or abnormal signals in that muscle as seen on computerized tomography scanning or magnetic resonance imaging (MRI) are difficult to interpret. Do such so-called abnormalities occur occasionally as incidental findings in otherwise normal persons? In 2002, Fishman and colleagues3 reported a series of 918 patients (1014 legs) with alleged PS. These investigators made a valiant attempt to create diagnostic criteria, to validate an electrophysiological test demonstrating sciatic nerve dysfunction, as well as to evaluate a nonsurgical treatment protocol and results of surgery. Unfortunately, there are serious flaws in their methodology. The entry criteria consisted of nonspecific symptoms and signs. Exclusionary criteria (imaging abnormalities of the lower spine and pelvis) were not described. Standard electrophysiological studies of sciatic nerve function were not performed. The H-reflex testing protocol used was based on normal values derived from volunteers who may not have been age-matched to the patients. Treatments were broad-based and could benefit patients with a variety of painful musculoskeletal disorders of the lower spine, pelvis, and hips. Some treatment measures and their alleged effects were implausible; for instance, “myofascial release at the lumbosacral paraspinal muscles” and “conservative treatment that lengthens the piriformis muscle.” Most patients, regardless of the number of clinical criteria for PS, and whether the H-reflex test was abnormal or not, responded well to nonsurgical therapy. The statistical results are difficult to interpret. In summary, their study, regrettably, neither defined a distinct syndrome nor clarified any of the issues surrounding PS. There remains no consensus as to how to use the term PS, let alone whether there is an entity deserving of this appellation. Sciatic nerve lesions from masses in the region of the sciatic notch and piriformis muscle should be designated as proximal sciatic neuropathies; to use the term PS misleadingly implies a pathogenic role for the piriformis muscle. There is certainly a rare syndrome of proximal sciatic nerve damage by compression from the piriformis muscle or associated fibrous bands. Sometimes this is associated with an anatomical anomaly of the piriformis muscle and sciatic nerve. The term PS is apt here. When such a condition occurs following buttock trauma, it is reasonable to designate this as posttraumatic PS. It is unlikely that the piriformis muscle plays a pathogenic role in most patients with buttock pain (with or without sciatica) yet no clinical or electrophysiological evidence of sciatic nerve damage, and in whom imaging studies of the lumbosacral spine and pelvis are normal. The use of the term PS here is therefore strongly discouraged. In such patients a diligent search for alternative causes of their pain should be undertaken. If no abnormalities are found, the patient should be treated symptomatically. However, it is impossible to know which of the many conservative measures described, including local anesthetic and corticosteroid injections into the piriformis area, are likely to be beneficial. The indications for, and the outcome from, surgical exploration and division of the piriformis muscle in these patients are entirely uncertain. Dr. Gillian Bartlett-Esquilant of the Department of Medicine, McGill University, provided valuable help with this study.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.