 |
 |
 |
|
| home | CME | help | feedback | subscriptions | archive |
|
© 1999 The Journal of Bone and Joint Surgery, Inc. Posttraumatic Piriformis Syndrome: Diagnosis and Results of Operative Treatment*  , HARTFORD, CONNECTICUTInvestigation performed at Hartford Hospital, Hartford
Background: Posttraumatic piriformis syndrome is a rare disorder that is not clearly defined in the orthopaedic literature. We report on the specific diagnosis, operative treatment, and outcome of treatment of fifteen cases of piriformis syndrome (in fourteen patients), treated by one surgeon, in which the common etiology was blunt trauma to the buttock. We are unaware of any previously published report of this kind. Methods: Fourteen patients (fifteen cases of piriformis syndrome), with an average age of thirty-eight years (range, twenty-four to fifty-six years), were managed with an operative release of the piriformis tendon and sciatic neurolysis. All fourteen patients had a history of a blow to the buttock, and all had pain in the buttock, intolerance to sitting, tenderness to palpation of the greater sciatic notch, and pain with flexion, adduction, and internal rotation of the hip. Eleven patients (twelve cases) had severe radicular pain in the affected lower limb. All fourteen patients failed to improve after a prolonged period of conservative treatment with nonsteroidal medication or physical therapy, or both. On the average, the patients had been evaluated by three physicians who were not orthopaedic surgeons and by two orthopaedic surgeons before they were referred to the senior one of us. They had had an average of 4.5 diagnostic tests and an average delay of thirty-two months (range, four to seventy-one months) between the time of the injury and the operation. Preoperative electromyograms revealed extrapelvic compression of the sciatic nerve in six of the eight patients who had this study. Intraoperative findings revealed adhesions between the piriformis muscle, the sciatic nerve, and the roof of the greater sciatic notch. Results: Clinical examination at a minimum of twenty-four months (average, thirty-eight months) postoperatively revealed eleven excellent and four good results according our symptom-rating scale. All of the patients returned to work or to their usual daily activities at an average of 2.3 months postoperatively, and the time to maximum subjective improvement averaged 2.1 months. Complications included a seroma and an infected hematoma. Conclusions: Patients who have blunt trauma to the buttock and then have signs and symptoms that are suggestive of lumbar nerve-root compression may have posttraumatic piriformis syndrome. In our group of carefully selected patients, release of the piriformis tendon and sciatic neurolysis led to encouraging results with few complications.
The diagnosis and treatment of so-called posttraumatic piriformis syndrome have been ill defined in the orthopaedic literature. It is common for orthopaedic surgeons to manage patients who have pain in the low back or buttock associated with radicular pain in the lower limb (also known as sciatica). In 1983, a survey of 1221 American men who were between eighteen and fifty-five years old revealed that 368 (30 percent) had never had low-back pain, 565 (46 percent) had had or were having moderate low-back pain, and 288 (24 percent) had had or were having severe low-back pain7. Symptoms of low-back pain and sciatica frequently overlap. Although most patients with these symptoms have disorders of the lumbar spine, extraspinal conditions have also been implicated. In 1928, Yeoman19 reported that sciatica may be due to a so-called periarthritis involving the anterior sacroiliac ligament, the piriformis muscle, and adjacent branches of the sciatic nerve. In 1937, Freiberg5 described two findings on physical examination that were consistent with sciatic pain referable to the piriformis muscle: Lasègue's sign (pain in the vicinity of the greater sciatic notch with extension of the knee with the hip flexed to 90 degrees and tenderness to palpation of the greater sciatic notch) and Freiberg's sign (pain with passive internal rotation of the hip). The following year, Beaton and Anson2 identified certain anomalies of the piriformis muscle and postulated that sciatica could be secondary to an aberrant relationship between the piriformis muscle and the sciatic nerve. Pace and Nagle12 later described a diagnostic maneuver that is now referred to as Pace's sign: pain and weakness in association with resisted abduction and external rotation of the affected thigh. Robinson14 has been credited with introducing the term piriformis syndrome and describing its six classic findings: (1) a history of trauma to the sacroiliac and gluteal regions; (2) pain in the region of the sacroiliac joint, greater sciatic notch, and piriformis muscle that usually extends down the limb and causes difficulty with walking; (3) acute exacerbation of pain caused by stooping or lifting (and moderate relief of pain by traction on the affected extremity with the patient in the supine position); (4) a palpable sausage-shaped mass, tender to palpation, over the piriformis muscle on the affected side; (5) a positive Lasègue sign; and (6) gluteal atrophy, depending on the duration of the condition. We believe that piriformis syndrome can be caused by blunt trauma to the buttock and is a result of hematoma formation and subsequent scarring between the sciatic nerve and the short external rotators. Local anatomical anomalies may contribute to the likelihood that symptoms will develop. In patients who have this condition, movement of the hip may cause radicular pain that is much like the nerve-root pain associated with lumbar disc disease. We have termed this clinical entity posttraumatic piriformis syndrome. Our hypothesis is that, because of both the lack of widespread understanding in the medical and orthopaedic communities about this condition and the lack of reliable objective tests to identify it, patients have had lengthy delays in diagnosis, additional expenses related to misleading diagnostic tests and to time lost from work, and unnecessary suffering. Patients who have a history of gluteal trauma, symptoms of pain in the buttock and intolerance to sitting, tenderness to palpation of the greater sciatic notch, and pain with flexion, adduction, and internal rotation of the hip may have posttraumatic piriformis syndrome. In addition, we believe that, in patients who have a typical history and typical findings on physical examination with intractable pain after failure of nonoperative treatment, release of the piriformis tendon and sciatic neurolysis can give predictably good results with few complications. Our goal is to make the medical and orthopaedic communities more aware of this condition.
Between July 1991 and November 1995, the senior one of us (S. F. S.) evaluated ninety-three patients who had a presumptive diagnosis of sciatica (International Classfication of Diseases, Ninth Revision [ICD-9], code 724.3) or a so-called contusion of the hip (ICD-9 code 924.01). These patients were referred by three other orthopaedic surgeons whose practices primarily involved the treatment of spinal disorders. Seventy-nine of these patients were managed conservatively with a combination of nonsteroidal anti-inflammatory medications, modification of activities, and physical therapy, with eventual improvement. Only three of these seventy-nine patients had a history of gluteal trauma, and two other patients had a history of intolerance to sitting on the involved side. None of the seventy-nine patients had a history of both of these conditions, in contrast to the fourteen patients who subsequently were managed with operative intervention. Many of these patients were believed to have degenerative lumbar disc disease, and all seventy-nine of them had enough of a decrease in the symptoms with conservative measures or enough acceptance of the symptoms so as not to pursue additional intervention by the senior one of us during the follow-up period. All of the patients were followed for at least twenty-four months. Fourteen patients did not respond to conservative measures and were thought to have recalcitrant posttraumatic piriformis syndrome (which was bilateral in one patient) on the basis of their history and findings on physical examination. The records of these fourteen patients, who were ultimately managed with release of the piriformis tendon and sciatic neurolysis, were retrospectively reviewed at a minimum of twenty-four months (average, thirty-eight months; range, twenty-four to seventy-three months). There were nine women and five men, with an average age of thirty-eight years (range, twenty-four to fifty-six years) at the time of the operation. Nine operations were on the left side, and six were on the right side. One patient was managed with a bilateral release, with the procedures performed eight months apart. This patient had fallen directly on both buttocks after slipping on ice, and she subsequently had symptoms bilaterally, with those on the right side worse than those on the left. She was managed successfully with operative intervention on the right side three years after the injury, and, after dramatic improvement, she had similar treatment of the still symptomatic left side. All of the procedures were performed between February 1992 and November 1995 (Table I).
Each patient had a history of trauma to the gluteal region that preceded the onset of symptoms; none had had similar symptoms before the injury. All but one patient was employed outside the home at the time of the initial injury. Two of the patients were waitresses, two were computer programmers, and one (who had bilateral symptoms) was a marketing manager. There was also a corrections officer, an owner of a maid service, a psychiatrist, a homemaker, an airport inspector, a secretary, a lawyer, an apartment supervisor, and an independent sales representative. Eleven patients (twelve cases) traced the onset of symptoms to a fall directly onto one or both buttocks. One patient sustained an injury to the gluteal region in a motor-vehicle accident when the door on the driver's side was crushed in by a second automobile. Another patient sustained a low-energy contusion to the right buttock when, as a pedestrian, she was trapped between two automobiles, and a third patient was struck in the right buttock by a sixty-pound (27.2-kilogram) cart of machinery. Nine injuries (ten cases) occurred at work and resulted in Workers' Compensation claims. Five patients (five cases) sustained injuries that were not work-related. After the initial injury, nine patients (ten cases) reported swelling and six patients (six cases) noticed ecchymoses in the involved gluteal region. Five patients (six cases) had radicular pain in the lower limb (sciatica) immediately after the accident, six patients (six cases) began to have such pain at an average of six weeks (range, one day to six months) after the injury, and three patients (three cases) did not have pain in the limb at any time. All eleven patients (twelve cases) with sciatica had this symptom until the time of the operation. Eight patients (eight cases) were out of work for an average of seventeen months (range, five to thirty-six months) between the time of the initial injury and that of the operative intervention. Before they were referred to an orthopaedic surgeon, these patients saw an average of three other physicians (range, one to six physicians) who were not orthopaedic surgeons. Thirteen patients were also evaluated by other orthopaedic surgeons (average for the series, two orthopaedic surgeons; range, zero to three orthopaedic surgeons) before they were referred to the senior one of us. Overall, the fourteen patients in the series had had an average of 4.5 diagnostic tests and had seen an average of 4.9 other physicians (range, one to nine physicians) before the diagnosis of posttraumatic piriformis syndrome was made. Thirteen patients (fourteen cases) had been managed with physical therapy for an average of five months (range, one to sixteen months) before operative intervention. At the time of the initial evaluation by the senior one of us, two patients were using walking aids. One patient had used crutches for two months, and the other had used a cane for four months. All but one patient had been taking various pain medications. Eight patients were taking nonsteroidal anti-inflammatory medications; four, narcotic analgesics; and one, muscle relaxants. The most common presenting symptoms were pain in the buttock and intolerance to sitting on the involved side. All fourteen patients had these symptoms, and none of them were able to sit comfortably for more than thirty minutes. Seven patients had numbness in various sensory distributions on the involved side, but none had muscle weakness in a corresponding distribution. The two most consistent findings on physical examination were tenderness to palpation of the greater sciatic notch and reproduction of the pain with maximum flexion, adduction, and internal rotation of the hip. Both of these findings were present in all fourteen patients (fifteen cases). The patient who had bilateral symptoms had sciatica bilaterally, but she had one positive and one negative straight-leg-raising test. Twelve of the fifteen cases of piriformis syndrome involved paresthesias and pain in the affected limb with straight-leg raising (average minimum angle of hip flexion, 67 degrees; range, 45 to 80 degrees) in the supine position. A positive test was associated with the presence or absence of sciatica on the involved side in only nine of the fifteen cases, as three of the eleven patients (twelve cases) who subjectively noted sciatica had a negative straight-leg-raising test. All of the patients in the present series had multiple ancillary studies as part of the evaluation. At least one series of radiographs of the hip and lumbar spine was made for each of the patients, and these studies were repeated for seven patients before the diagnosis was made. These radiographs showed degenerative changes in the lumbar spine in three patients but were otherwise unrevealing. Ten patients had a bone scan, which revealed negative findings in all of them, and every patient had at least one magnetic resonance imaging scan of the lumbar spine. Two scans revealed nerve-root compression and degenerative changes. Two patients had two magnetic resonance imaging scans, and a third patient had three. No patient had a magnetic resonance imaging scan of the hip or pelvis. Eight patients had an electromyographic evaluation as part of the diagnostic workup. Six of the eight electromyograms revealed abnormal findings in the distribution of the inferior gluteal nerve and in the tibial and peroneal divisions of the sciatic nerve, with normal tracings in the distribution of the superior gluteal nerve. These findings were believed to be consistent with extrapelvic compression of the sciatic nerve. Six patients had had an invasive procedure before they were referred to the senior one of us: two had received epidural injections of steroids, one had received an injection of corticosteroids into the sacroiliac joint, and another had received local injections of corticosteroids in the vicinity of the piriformis tendon. None of these procedures alleviated the symptoms. The two patients who were seen to have nerve-root compression on magnetic resonance imaging scans and radicular pain that was believed to be secondary to that lesion were managed with posterior lumbar decompression and an arthrodesis with spinal instrumentation. Neither patient noted any relief of the pain in the lower limb or the buttock as a result of the operation. The average delay in the release of the piriformis tendon and the sciatic neurolysis from the time of the initial injury was thirty-two months (range, four to seventy-one months). Records and results of diagnostic studies were collected and analyzed. Jankiewicz et al.10 showed the utility of computed tomography in the evaluation of patients who have suspected piriformis syndrome, and they reported finding an asymmetrical enlargement of the piriformis muscle on the symptomatic side. In light of this, the senior one of us referred all of the patients who were suspected of having posttraumatic piriformis syndrome for computed tomography scans of the pelvis without the use of contrast medium. However, these scans were not helpful in confirming the diagnosis of posttraumatic piriformis syndrome. Two of the symptomatic piriformis muscles were larger than the muscle on the asymptomatic side, seven of the symptomatic piriformis muscles appeared to be smaller than the muscle on the unaffected side, and six (including those in the patient who had bilateral symptoms) were not obviously different from the muscle on the contralateral side. In addition, computed tomography scans of the pelvis were made for three of the seventy-nine patients who were managed nonoperatively. There were no obvious differences between the sides in any of these patients. All computed tomography scans were evaluated by a board-certified radiologist as well as by the senior one of us. Fourteen patients (fifteen hips) were managed with operative release of the piriformis tendon and sciatic neurolysis by the senior one of us. As mentioned, one patient, a thirty-two-year-old woman, slipped on ice and fell onto both buttocks. Three years later, she had an operation on the more symptomatic, right side. As a result of the dramatic improvement, she had similar treatment eight months later on the left side as well. An attempt was made to perform all of the operations on an outpatient basis, although three patients stayed in the hospital overnight for parenteral treatment with analgesics before they were discharged. The senior one of us performed or directly supervised all of the procedures. The operation was carried out with the patient under general endotracheal anesthesia and in the lateral position. A skin incision was made starting at the posterolateral tip of the greater trochanter and was carried posteriorly and proximally in the same direction as the gluteus maximus fibers. After the gluteus maximus fascia and muscle were split, the insertion of the piriformis tendon was palpated and then exposed. The tendon was divided from its insertion on the greater trochanter, grasped with an Allis clamp, and dissected proximally to its exit from the greater sciatic notch. The sciatic nerve was then identified, and neurolysis was performed by mobilizing the nerve from the overlying piriformis muscle and other underlying short rotators with use of blunt dissection proximally into the greater sciatic foramen and distally to the end of the wound. After freeing of the nerve, it was possible to pass a right-angled clamp freely between the nerve and the short rotators. The wound was irrigated, and hemostasis was obtained. The gluteus maximus fascia was then closed with a running absorbable suture before the skin was closed. A bulky compressive dressing was used, and no patient needed a suction drain. Before the operation, all of the patients were given one gram of cefazolin intravenously for prophylaxis against infection. The patients were discharged and were told to bear weight while using crutches and to increase activities as tolerated on the involved side. All of the patients were seen again one week later to check the wound and to change the dressings. No patient needed physical therapy postoperatively. Functional results were determined on the basis of a symptom-rating scale that we devised and that was specific for the symptoms and limitations of the patients in this series (Table I). An excellent result was defined as no pain in the buttock or sciatica with prolonged periods (more than thirty minutes) of sitting or strenuous activity, including bending, twisting, stair-climbing, rapid walking, or jogging. Patients in this category had no limp, used no external supports, and did not take analgesics for pain in the buttock or sciatica. A good result was defined as no pain in the buttock or sciatica with short periods (thirty minutes or less) of sitting or normal daily activities or as mild pain or slight sciatica with prolonged periods of sitting or strenuous activity. Patients in this category had a slight limp, occasionally used an external support, or sometimes took non-narcotic analgesics for pain in the buttock or sciatica. A fair result was defined as occasional mild pain in the buttock or sciatica with short periods of sitting or normal daily activities or as moderate pain with prolonged sitting or strenuous activity. Patients in this category had a moderate limp, used one cane or crutch, or frequently used narcotic analgesics or muscle relaxants for pain in the buttock or sciatica. Finally, a poor result was defined as severe pain or sciatica with short periods of sitting or normal daily activities (that is, little subjective change from the preoperative level of pain). Patients were placed in this category if they had a severe limp, used a crutch or walker, or needed to take narcotic analgesics or muscle relaxants daily for pain in the buttock or sciatica.
All of the patients were followed by the senior one of us, and none were lost to follow-up. The results of operative release of the piriformis tendon and sciatic neurolysis were assessed at a minimum of twenty-four months (average, thirty-eight months; range, twenty-four to seventy-three months) postoperatively. Seven patients (eight cases of piriformis syndrome) needed to use external supports for an average of five weeks (range, one week to four months) after the operation. Three patients (four cases) used a cane, and four patients (four cases) used a single crutch. Eleven patients (twelve cases) took analgesics for an average of two weeks (range, one to four weeks) postoperatively. During this time, ten patients (eleven cases) took narcotic analgesics and one patient took only nonsteroidal anti-inflammatory medication. No patient continued to take analgesics or nonsteroidal medication on a regular basis. At the most recent follow-up evaluation, all fourteen patients reported a substantial decrease in pain in the buttock and in intolerance to sitting. One patient reported pain in the gluteal region that was much less severe than before the operation and that was only bothersome when he sat for more than one hour. Three patients reported minor residual pain in the buttock that was precipitated by strenuous activities such as repetitive bending, twisting, stair-climbing, rapid walking, or jogging. All eleven patients (twelve cases) who had had sciatica preoperatively had complete relief of pain in the limb. Of the seven patients who had had numbness on the affected side, two continued to have this symptom but in a smaller area than before the operation. Two patients had occasional numbness in the foot after prolonged periods of bending, twisting, or stair-climbing. No patient had muscle weakness in the involved extremity. Of the twelve patients (twelve cases) who had had a positive straight-leg-raising test with reproduction of the symptoms before the operation, only one continued to have a positive test after the operation. Two patients (two cases) continued to have tenderness to deep palpation of the greater sciatic notch, but none had persistent pain with flexion, adduction, or internal rotation of the hip. All of the patients, including the nine (ten cases) who had filed Workers' Compensation claims, returned to work or to their usual daily activities at an average of 2.3 months (range, zero to eight months) after the operation. Of the eight patients (eight cases) who had been out of work for an average of seventeen months (range, five to thirty-six months) before the operation, all were back at work at an average of three months (range, one week to eight months) postoperatively, and seven of the eight returned to the position that they had had preoperatively. The remaining patient did not return to her job as a waitress but subsequently gained employment as a secretary. There were unusual operative findings in two patients. In one of them, the sciatic nerve split such that one branch passed through the piriformis muscle and the other branch passed proximal (rather than distal) to the muscle. In this patient, the piriformis tendon was released from its insertion and was further released adjacent to the branch of the sciatic nerve that passed through it. In the second patient, myositis ossificans of the piriformis tendon was found. This patient had fallen down a flight of stairs, landing on the left buttock, and subsequently had sciatic pain on the left side. Three months later, computed tomography and magnetic resonance imaging scans revealed myositis ossificans of the piriformis muscle. At the time of the operation, the piriformis muscle contained a large osseous mass that compressed and flattened the sciatic nerve against the roof of the greater sciatic notch. The area of myositis ossificans of the piriformis muscle was excised as far proximally into the pelvis as possible to decompress the sciatic nerve3. In nine of the remaining thirteen operations the piriformis muscle was found to be more firm to palpation than expected, and in all fifteen operations the sciatic nerve was noted to be adherent to the posterior pelvic column. There were no obvious gross alterations (narrowing or inflammation) of any sciatic nerve, although the adhesions between the nerve and the underlying short external rotators in all cases markedly limited the mobility of all of the sciatic nerves. No persistent sciatic artery or venae comitantes were discovered, and no other space-occupying lesions were noted. The patients were asked when they believed that they had obtained maximum improvement. Two patients (three cases) reported a dramatic decrease in preoperative symptoms on awakening from the operation in the recovery room. Both of these patients (three cases) believed that they had obtained maximum improvement immediately, as they thought that their preoperative symptoms had resolved after the operation despite discomfort from the operative incision. The remaining twelve patients (twelve cases) obtained maximum improvement at an average of 2.6 months (range, one week to six months) after the operation. Based on the criteria in our symptom-rating scale, there were eleven excellent, four good, and no fair or poor results (Table I). Three of the patients with a good result had residual pain in the buttock with strenuous activity, and the other patient with a good result had pain in the buttock with prolonged periods of sitting. All fourteen patients (fifteen cases) stated that they would elect to have the operation again under similar circumstances. There was one minor and one major complication in our series. The patient who had myositis of the piriformis muscle was noted to have a large amount of swelling in the vicinity of the wound two weeks postoperatively. There was no wound erythema or drainage and no systemic evidence of infection. Under sterile conditions, forty milliliters of deep hematoma was aspirated without recurrence. All of the cultures were negative. This patient had an excellent result at the time of the most recent follow-up. A second patient had persistent swelling of the buttock and clear drainage from the incision three weeks postoperatively. She was immediately taken back to the operating room for irrigation and débridement of a deep hematoma. An intraoperative gram stain revealed the presence of bacteria, and Staphylococcus aureus ultimately grew on culture. The wound was treated with wet-to-dry dressings twice a day, and it was allowed to heal by secondary intention to facilitate maximum continued drainage and to decrease the risk of recurrent infection. The patient received antibiotics (one gram of cefazolin intravenously every eight hours) at home for four weeks and then a cephalosporin (Keflex [cephalexin], 500 milligrams orally four times a day) for an additional two weeks. There was no recurrent drainage or erythema, and the patient had a good result at the time of the most recent follow-up.
Phylogenetic development of the piriformis muscle in bipedal species reportedly derives from the large bipennate caudifemoralis muscle, which is a major muscle of locomotion in lizards15. The piriformis muscle arises from the anterior surface of the sacrum, the gluteal surface of the ilium (near the posterior superior iliac spine), and the capsule of the sacroiliac joint. This muscle functions as an external rotator of the hip joint when the thigh is extended and as an abductor of the hip joint when the thigh is flexed4. It passes through the greater sciatic foramen and is attached by a rounded tendon to the superior border of the greater trochanter, and its tendon often blends with that of the obturator internus and the gemelli18. A study involving 240 consecutive dissections revealed four different relationships of the piriformis muscle to the sciatic nerve2: (1) 90 percent (216) of the specimens had a sciatic nerve distal to the piriformis muscle that emerged undivided from below the piriformis muscle; (2) 7 percent (seventeen) had a divided piriformis muscle, with one branch of the sciatic nerve passing through the split and the other branch passing distal to the muscle; (3) 2 percent (five) had a sciatic nerve that divided proximal to the piriformis muscle, with one branch of the nerve passing distal to the muscle and the other branch passing proximal to it; and (4) 1 percent (two) had a divided piriformis muscle, with the sciatic nerve passing through its midportion. In our series, one patient had a sciatic nerve that divided proximal to the piriformis muscle, with one branch passing through the muscle and the other branch passing superficial to it. We are unaware of any previous reports of this variation. Several mechanisms have been proposed to explain the symptoms of patients who have posttraumatic piriformis syndrome. Robinson14 believed that the piriformis muscle and fascia became inflamed secondary to trauma and the sciatic nerve became compressed between the swollen muscle fibers and the osseous pelvis, leading to so-called entrapment neuropathy. Pecina13 postulated that patients who have an anomalous sciatic nerve (with the nerve or one of its branches passing through the tendinous portion of the piriformis muscle) are especially susceptible to compression of the nerve by the edge of the muscle or tendon with internal rotation of the hip. Pace and Nagle12 proposed that focal irritability of the piriformis muscle, which is usually caused by trauma, results in a so-called trigger-point syndrome. We believe that Robinson's explanation is the most accurate on the basis of our intraoperative observations of scarring around the piriformis muscle and around the sciatic nerve as well as the adherence of the nerve to the posterior pelvic column. Only one of the fifteen cases of piriformis syndrome in the present study was associated with an anatomical variant, and the patient in whom this variant was seen also had sustained blunt trauma. Therefore, we do not know whether such findings are associated with the development of piriformis syndrome. We have no experience with the operative treatment of piriformis syndrome that is not caused by blunt trauma. The nonoperative treatment of piriformis syndrome has been described in the literature1,8,9,11,17. Nonsteroidal anti-inflammatory agents may be prescribed to reduce local prostaglandin-mediated inflammation, pain, and spasms1. Hallin8 reported a reduction of pain with ultrasound treatments in eight of eleven patients who had piriformis syndrome. Physical therapy involving stretching of the piriformis muscle with flexion, adduction, and internal rotation of the hip followed by strengthening of the hip abductors has been advocated as well9. Transrectal massage17 and the application of a transcutaneous electrical stimulator unit11 have also been recommended. To our knowledge, the success rate of these conservative measures has not been reported in the literature, although in our series 85 percent (seventy-nine) of the ninety-three patients who had been referred to the senior one of us with a presumptive diagnosis of sciatica or a so-called contusion of the hip were successfully managed nonoperatively. Whether these patients actually had piriformis syndrome is not known. It is also possible that there were other cases of piriformis syndrome that were coded differently (and therefore were not included in our study) and that improved without an operation. The use of local injections has been recommended for patients for whom other noninvasive measures have failed. Pace and Nagle12 described a double-injection technique in which an anesthestic is injected locally with corticosteroids through the buttock, toward the painful piriformis muscle, which is localized by transrectal or vaginal digital palpation. Forty-five patients received these injections, and only four of them had a recurrence of pain. The authors did not provide objective follow-up data or describe the functional outcomes. Half of the patients in their series had a history of trauma. To the best of our knowledge, this is the first report of the results of operative intervention specifically for recalcitrant posttraumatic piriformis syndrome. In the largest series of patients managed with release of the piriformis tendon in the literature, Freiberg6 reported favorable results in ten of twelve patients who had piriformis syndrome and for whom conservative treatment consisting of traction, mechanical support, and physical therapy had failed. There was no mention of trauma, and the duration of follow-up was not specified. Robinson14 also described favorable results (alleviation of preoperative pain in the buttock and sciatica) after release of the piriformis tendon and lysis of adhesions between the piriformis muscle and the sciatic nerve in two patients who had had symptoms that were consistent with piriformis syndrome after a blow to the buttock. Solheim et al.16 reported successful results after release of the piriformis tendon in two women who had piriformis syndrome. One of the women had a history of trauma. Barton1 reported on four patients who had atraumatic piriformis syndrome, only one of whom had reduction of symptoms after release of the piriformis tendon. Vandertop and Bosma17 reported successful results after release of the piriformis tendon in a man who had atraumatic piriformis syndrome. Hughes et al.9 reported on five patients who had release of the piriformis tendon for symptoms of extrapelvic compression of the sciatic nerve. Two of their five patients had a history of a fall, and conservative treatment had failed in all five. Preoperative electromyograms revealed normal tracings for muscles innervated proximal to the piriformis tendon by the superior gluteal nerve (the gluteus medius and minimus and the tensor fasciae latae) and abnormal tracings for muscles innervated by the inferior gluteal nerve (the gluteus maximus), the peroneal nerve, and the tibial nerve. At the time of the operation, each of the five patients had evidence of an anatomical variant, including two bifid piriformis muscles, two persistent sciatic arteries, a traumatic venous aneurysm, and a thick, tendinous band on the undersurface of the muscle. All five patients had prompt resolution of the symptoms postoperatively, although no attempt was made to use a functional outcome scale and no follow-up interval was longer than a year. Finally, Beauchesne and one of us (S. F. S.) reported on a patient who had had severe trauma to the buttock and subsequently had sciatic-nerve compression related to myositis ossificans of the piriformis muscle. The patient was successfully managed with excision of the mass, release of the piriformis tendon, and sciatic neurolysis and was included in the present study. The present study has several limitations. First, it is a retrospective review. Second, it lacks standardization of nonoperative treatment, including physical therapy. Our preoperative evaluation did not include rectal or vaginal examinations, which may be helpful for making the diagnosis and ruling out other causes of pain. Female patients were not questioned about the presence of dyspareunia. Electromyograms were not made for all of the patients. In retrospect, this may be the most valuable objective preoperative test, as six of eight patients in our series had electromyographic findings that were consistent with extrapelvic compression of the sciatic nerve at the level of the piriformis muscle. In our study, preoperative computed tomography scans were of limited value and had no relative predictive value concerning the final outcomes. Diagnostic or therapeutic injections of local anesthestics or steroids, or both, were not used in these patients. Although we know of only one study12 that has demonstrated a large number of successful results with this treatment, it may be worth considering in the future. Our conclusion is that, although posttraumatic piriformis syndrome is a diagnosis of exclusion, it is a distinct clinical entity. The constellation of symptoms in this series of patients who were managed operatively necessitated the exclusion of degenerative disc disease and lumbar nerve-root entrapment associated with disc herniation or another spinal abnormality. Patients who have a history of blunt trauma to the gluteal region, prolonged symptoms of pain in the buttock and intolerance to sitting, tenderness to palpation of the greater sciatic notch, and reproduction of symptoms with flexion, adduction, and internal rotation of the hip may have posttraumatic piriformis syndrome. Patients with this background for whom prolonged conservative treatment fails should be considered candidates for confirmatory nerve-conduction studies and referral to an experienced hip surgeon who is familiar with this syndrome. An increased awareness of this disorder within the medical and orthopaedic communities should lead to earlier diagnosis and a more cost-efficient evaluation. In the group of carefully selected patients in this report, release of the piriformis tendon and sciatic neurolysis led to encouraging results with few complications.
*No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. No funds were received in support of this study.
  CiteULike  Connotea  Del.icio.us  Facebook  Technorati  Twitter What's this?
This article has been cited by other articles:
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Top
Abstract
Introduction
| home | CME | help | feedback | subscriptions | archive | site index |
|
Stanford University Libraries' HighWire Press (TM) assists in the publication of JBJS Online. |