A 35-year-old woman presents to the office complaining of right-sided low back and buttock pain intermittently radiating down the posterior aspect of her right leg. It started a few months prior, initially during certain rotational yoga poses. She stopped yoga, and it improved, but then she began noticing the pain and radiculopathy after walking prolonged distances. She recently changed jobs to one where she spends more time at a desk and is having increasingly difficulty sitting for long periods of time because of pain in the back and leg. She denies any symptoms on the contralateral side, fevers, chills, and weight loss. She has not tried any intervention or medications for the pain. On examination, she has tenderness over the greater sciatic notch, mildly positive straight-leg-raise, and significant reproduction of her symptoms with hip external rotation and abduction but no motor or sensory deficits. Radiographs taken in the office are normal.

For decades, piriformis syndrome has posed a clinical conundrum. The earliest mention of the piriformis muscle as a source of radicular low back pain was 1928, by Yeoman, who theorized fibrosis of the muscle could cause sciatica.¹ In 1947, Robinson formally established “piriformis syndrome,” setting forth six diagnostic features: a history of trauma to the sacroiliac or gluteal region, pain in the sciatic notch and piriformis muscle that extends down the leg and hinders walking, exacerbation with stooping or lifting that is relieved by traction, presence of a palpable and tender mass over the piriformis, a positive Lasègue sign, and gluteal atrophy.²

The piriformis muscle originates from the anterior aspect of S2 to S4, exits through the sciatic notch, and inserts on the posteromedial corner of the superior edge of the greater trochanter (Figure 20-1). With the hip extended, it acts as an external rotator; when in flexion, it acts as an abductor. The sciatic nerve (L4–S1) exits the
sciatic notch, most commonly, anterior to, or “under,” the piriformis. Anatomic studies identified variations on this normal anatomy, such as a high division of the nerve splitting above and below the muscle or a single nerve piercing a bifid muscle.³ A cadaveric study cadavers showed that 94% of the time, the piriformis and sciatic nerve follow the expected course, but that the most common variation, occurring in 4% of subjects, was a high-dividing nerve with the common peroneal division passing through and the tibial division below a bifid piriformis muscle⁴ (Figure 20-2). Other structures exiting the sciatic notch can potentially also be affected by piriformis pathology, including the superior gluteal artery (above the piriformis), the inferior gluteal artery and nerve, the pudendal nerve and internal pudendal vessels, and the posterior femoral cutaneous nerve.

Symptoms are thought to be due to compression of the sciatic nerve by the piriformis muscle, although some physicians attribute any sciatica without a clear discogenic cause to be piriformis syndrome. Any lesion or mass causing compression of the proximal sciatic nerve, such as endometriosis, hematoma, myositis ossificans, tumors, or aberrant vascular formations, can lead to symptoms of piriformis syndrome.⁵ Reports in the literature indicate that it may be simultaneously overdiagnosed and overlooked, causing between 0.33% and 6% of all low back pain and radicular symptoms.^6,7 Across all etiologies, there is a greater incidence among women, with the ratio in some studies reported as high as 6:1.⁸

Piriformis syndrome is most commonly attributed to trauma, either from acute hematoma or posttraumatic scar formation, as well as overuse syndromes or variant anatomy.^2,9 Benson and Schutzer presented a series of patients with piriformis syndrome, all of whom had symptoms only after a confirmed blow to or fall on the gluteal muscles. In this series, all patients had electromyographic (EMG) abnormalities in the muscles
supplied by the sciatic nerve that resolved with surgical release.⁹ Overuse theorists believe the gait cycle, during which the piriformis muscle is under strain throughout the entirety of a stride, leads to hypertrophy or spasm.¹⁰ Similarly, Fishman and Schaefer proposed it to be a functional entrapment of the sciatic nerve, occurring only when the leg or hip was in specific positions.¹¹ In more sedentary individuals, prolonged sitting on hard surfaces has been blamed, hence the nickname “wallet neuritis.”¹² Variant anatomy may make the nerve more susceptible to compression, but the incidence of piriformis syndrome in patients with documented anatomic aberrations is no higher than that in those with normal anatomy.¹³

Piriformis syndrome may present acutely, particularly if traumatic, or insidiously, as in the overuse or compression types. Patients describe buttock or posterior hip pain (or both), variable patterns of radiculopathy, and intolerable pain with prolonged sitting. In one study, buttock pain occurred in 95% of patients, more common than low back pain, and pain aggravated by sitting occurred in 97% of subjects.¹⁴ Distal radicular symptoms generally follow a pattern consistent with sciatic nerve distributions, but motor or reflex deficits are uncommon. Additionally, some have pain with bowel movements, and women have complained of dyspareunia.

No single physical examination finding is pathognomonic. Palpation of the muscle belly can cause focal tenderness or recurrence of radicular symptoms; a sausage-shaped mass may be present. Pain at this site is distinct from pathology at the insertion of the gluteus maximus on the greater trochanter or the sacroiliac joint.¹⁵ Palpation of the piriformis should be performed directly posterior to the hip joint, which is close to the sciatic notch. This is distinct from the location pain of greater trochanteric pain syndrome (GTPS), which is found overlying the greater trochanter on the lateral aspect of the thigh just distal and lateral to the hip joint (Figure 20-3).

Many studies advocate palpating the sciatic notch internally via a rectal or pelvic examination.⁸ Reproduction of symptoms is often quite obvious, and the absence of pain during these examinations should cause reconsideration of the diagnosis.

Some patients may report positive findings with straight-leg raise, but these are more likely seen because of a lumbar nerve root irritation. In a systemic review of piriformis syndrome, only 31% had a positive straight-leg raise.¹⁴ The cross straight-leg raise test is not reported in the piriformis syndrome population.¹⁶ The Freiberg sign, resisted external rotation of the hip with it in extension and internal rotation, is positive in 63% of patients.¹⁰ The Pace test,⁷ resisted abduction and external rotation of hip in flexion or in a seated position, is positive in 30% to 74% of patients.¹⁴ The piriformis stretch, passive adduction with internal rotation in 90° of flexion, may cause symptoms. The Beatty test, pain with flexion and elevation of the symptomatic side while the patient lies on the asymptomatic side, has also been described.¹⁰ These tests may be held for up to 1 minute before symptoms appear, similar to the Phalen test for carpal tunnel syndrome.⁵

There are no unique radiographic findings of piriformis syndrome. Plain films are recommended to rule out other pathology. MRI can be performed to rule out lesions causing mass effect. A full workup of the lumbar spine for possible source should be performed as well, particularly if there are motor deficits. Neurodiagnostic studies have revealed that a prolonged H-reflex latency can be indicative of piriformis syndrome.¹³ These are accentuated with the hip is tested in flexion, abduction, and internal rotation.¹⁴ If conductive abnormalities are seen, there should still be preservation of the superior gluteal nerve as it exits the notch above the piriformis muscle.⁹

Conservative treatment is the initial recommendation for suspected piriformis syndrome. Lifestyle and activity modifications and anti-inflammatory medications are principal. A general back and hip physical therapy program, with attention paid to specific piriformis stretches,
is beneficial. If there is limited or no response to these nonoperative measures, then an injection of local anesthetic agent, corticosteroids, or both is a next step. It can be done with ultrasound or computed tomography guidance or by gross palpation. There are several described techniques for injecting via anatomic landmarks,^13,15 but a cadaver study showed that only 30% of injections via any landmark technique for piriformis syndrome were accurate compared with 95% accuracy for ultrasound-guided.¹⁷ In studies, 79% of patients experience at least a 50% reduction in pain after a corticosteroid injection combined with therapy.¹² In some studies, this diagnostic injection was curative.¹⁵ Recently, data are available on Botox injections, particularly for persistent cases.^18,19,20 In a study by Michel et al, 77% of patient who did not respond to rehabilitation in 6 weeks experienced good to very good pain relief after three injections of botulinum toxin A.¹⁹ A small trial comparing Botox with corticosteroid injections in refractory patients showed superior improvement in pain and function with Botox A at 12 weeks.²⁰

Surgical release of the piriformis muscle can also be performed for recalcitrant symptoms. The diagnosis must be convincingly proven without confounding sources. The response can be excellent, but it can also be incomplete if there is other involved pathology. Through the posterior approach, the piriformis attachment on the greater trochanter is released, the course of the muscle is explored, the nerve is decompressed if necessary, and the proximal stump of the tendon is amputated. Improvement in symptoms can be immediate or appear gradually over several weeks. The published results from this intervention are, overall, good, but they are generally from small case series or case reports^9,13,15 without uniform criteria or outcomes scores. Most report somewhere between 59% to 69% with good to excellent results.^16,21 Future research includes standardization of outcomes scores in the operative and nonoperative groups to be better able to interpret and compare data on diagnosis, etiology, and outcomes.

A 40-year-old woman presents to the office with complaints of increasing low back pain. She reports that it worsens throughout the day with activity, particularly when she tries to go running. She has always been a recreational runner but decided recently to run a marathon and has increased her training. She denies any radicular symptoms. She has tried massage therapy and notes that her back is sometimes so tender that she cannot tolerate the massage. She has not tried any further treatment. She denies trauma, fevers, and weight loss. Radiographs of her pelvis and lumbar spine are grossly normal.

Stress fractures account for a significant percentage of diagnosed overuse injuries and up to 15% of all injuries to runners.²² Stress fractures of the sacrum are being diagnosed in the athletic population with increasing frequency. They can occur in both high-level competitive athletes and “weekend warriors,” and a high clinical suspicion should be maintained. Most commonly seen in the tibia, fibula, and metatarsals, stress fractures of the pelvis are reported to account for only 1% to 7% of stress fractures.²³ The most prevalent sport is track and long-distance runners, but they have also been reported in tennis, gymnastics, and other sports with repetitive high-impact movements. Military trainees, particularly female recruits, constitute up to 22% of all stress fractures.²⁴

Stress fractures are skeletal defects caused by repeated application of lower intensity stress than is required to cause an outright fracture.²⁵ The first description of one in the sacrum was by Volpin et al in 1989.²⁶ The underlying cause of stress injuries is still debated. The overload theory is that repetitive contractions of muscles leads to stress at osseous insertions and impairs the bone’s ability to resist mechanical forces.²⁷ This theory is strongly advocated in stress fractures from non–weight-bearing activities. Second, the fatigue theory is that progressive exhaustion during activity reduces the ability of the muscles to function as shock absorbers, leading to abnormal load distributions and stress concentrations in bone.²⁸ The accumulation of microdamage to the bone cannot be repaired before new additional stress is placed. In the “female athlete triad”—disordered eating, amenorrhea, and osteoporosis—it is theorized that the decrease of estrogen and resulting loss of bone mineral density heighten risk for repetitive stress injuries.

The sacrum is the keystone arch of the pelvis, and it receives repetitive loading to the axial skeleton from ground reactive forces and muscle contraction.²⁹ Leg length discrepancies predisposing to uneven stride length and asymmetrical motion of the hips and lumbar spine have been proposed to potentiate these fractures, but they have not been shown in biomechanic studies.³⁰ Muscular imbalances may cause abnormal load transfer and increased risk of stress fracture.³¹ Many studies describe diagnosis after an increase of intensity or change in type
of training.^32,33 Women presenting with pelvic stress fractures have been shown to have an average body mass index of 21, an average age of 20 years old, and an average weekly running mileage of 25 to 33 miles per week.²⁹ In both populations, a history of amenorrhea^24,33 or prior stress fractures³⁴ are significant risk factors.

Sacral stress fractures have been classified to occur in one or more of three anatomic zones. Zone I is lateral to the foraminal line and involves the sacral ala. Zone II consists of one or several sacral foramina. Zone III is the central sacral canal. Zone I fractures can cause irritation of the L5 nerve root as it runs over the pelvic brim; this occurs in 6% of zone I fractures. The risks of neurologic symptoms with zone II and III fractures are 28% and 57%, respectively. Zone III fractures can present with symptoms of cauda equina syndrome.³⁵