Symptomatic presentation of hamstring injuries varies based on severity of pathology. Acute injuries generally follow rapid acceleration or deceleration movements and frequently occur in soccer, football, skiing, and hockey.1–7 Chronic injuries present with an insidious onset of unrelenting pain made worse with sports activities and sitting.8–13 Chronic injuries are often found in runners.7 Weakness and difficulty running are common.9 The hamstring tendon may be injured at a variety of locations along its anatomical course. A variety of forms of damage exist, including inflammation, degeneration, partial tearing, complete tearing, or a combination of these pathologies.
Chronic proximal hamstring tendinosis and partial tearing of the proximal hamstring origin are known causes of chronic posterior thigh and posterior hip pain. Tendinosis is defined as a degenerative condition of tendons; partial tearing is defined as an incomplete rupture. Chronic proximal hamstring tendinosis and partial tearing of the proximal hamstring origin may present in the acute, chronic, or acute on chronic setting.11,14–18 Complete or near-complete high-grade partial tearing of the hamstring origin with multiple tendon involvement has been reported more extensively, with surgical treatment generally being recommended.1–4,19,20 In contrast, relatively few publications exist related to diagnosis of and management strategies for chronic proximal hamstring tendinosis and partial tearing of the proximal hamstring origin. Injuries are frequently diagnosed late as a result of poor recognition and understanding of them.9 Partial tears often occur in the setting of degenerative tendinosis of various degrees, and occur more frequently with older age. This systematic review discusses the incidence, treatment, and prognosis of chronic proximal hamstring tendinosis and partial tearing of the proximal hamstring origin.21
Activity Level/Athletic Status. All 6 articles that reviewed surgical outcomes reported data on preinjury activity and sports level of participation. Twenty-eight professional athletes were included across 4 articles. Seventy-two patients were competitive athletes and 126 patients were recreational athletes. Of the professional athletes, 10 soccer players, 1 basketball player, 1 body builder, 1 ice hockey player, 1 aerobic athlete, 1 triathlete, 1 judo athlete, 1 karate athlete, 1 pole vaulter, 3 hurdlers, 3 long-distance runners, 2 middle-distance runners, and 2 professional triple jumpers were identified.6,9,10,12 Of the competitive and recreational athletes, 3 hurdlers, 2 triple jumpers, 8 middle-distance runners, 15 long-distance runners, 1 field hockey player, 1 track runner, 3 sprinters, 14 soccer players, 1 basketball player, 4 ice hockey players, 2 ballerinas, 1 cross-country skier, 1 floorball player, 1 handball player, 1 orienteering (a timed racing and navigation sport) athlete, 1 tennis player, 1 figure skater, 3 Finnish baseball players, 3 heptathlon athletes, 1 decathlon athlete, 2 long jumpers, 1 aerobics athlete, and 3 power lifters were identified.9,10,12
Two PRP articles specifically indicated the activity level of patients.7,28 Wetzel et al7 described 11 competitive athletes. Hamid et al28 discussed 15 national-level, 3 state-level, 2 club-level, and 8 school athletes. Further, there were 12 track runners, 9 soccer players, and 7 hockey, netball, basketball, rugby, tennis, or shot put athletes. All 40 patients were professional athletes in the SWT article.8 No specific mention of type of professional sport was made in the Wetzel et al7 or SWT article.8
In the conservative, noninvasive articles, athletes were identified as follows: 541 soccer players, 63 sprinters and jumpers, 5 baseball players, 6 football players, 3 basketball players, 1 ultimate Frisbee player, 1 racquetball player, 5 softball players, 1 triathlete, 1 tennis player, 1 hockey player, 1 cross-country runner, and 1 runner.29–32
Age. The average age of patients for all of the articles was 34 years. Average age across PRP, SWT, and surgical treatment groups was 36 years. Within the surgical treatment group, average age was 36 years. Average age across the conservative, noninvasive treatment group was 27 years.
Side of Injury. The surgical articles used for analysis included right or left side of injury. A total of 120 injuries occurred on the right leg and 121 on the left leg; 7 cases were bilateral. The SWT article included 22 right leg injuries and 18 left leg injuries. Handedness and side of injury did not correlate across these studies.
Male vs Female. Across all surgical and injection treatment studies, 229 men and 204 women were included in followup. Surgery was performed for 160 men and 94 women. Platelet-rich plasma was injected in 24 men and 29 women. Whole blood was injected in 5 women and 1 man. Shockwave therapy was administered to 27 men and 13 women. The conservative, noninvasive study groups included 668 men and 98 women.29–35
Radiographs are of little benefit in the diagnosis of hamstring tendinopathy.18 Prior to surgical treatment, and in the setting of chronic hamstring pain, MRI is typically performed. Magnetic resonance imaging reveals the extent of tendon involvement, but positive findings do not consistently correlate with symptoms. De Smet et al26 found a higher correlation of tendinopathy symptoms in patients with edema noted at the ischial tuberosity, and in those with increased signal in a feathery pattern at the peritendinous junction on T2 images.
Two studies detailed specific tendon involvement seen on MRI.9,10 Twenty-one (32.8%) patients had semitendinosus and biceps tendon involvement. Thirty-one (48.4%) patients had semimembranosus and 3-tendon involvement. Of those with single-tendon involvement, 5 (7.8%) patients had only semimembranosus involvement, 7 (11%) patients had only biceps femoris involvement, and 1 (1.5%) patient had only semitendinosus involvement.9,10
Ultrasound has lower reliability than MRI for diagnosis of hamstring tendinopathy.24,26,27 Zissen et al27 reported evidence of peritendinous fluid or edema in 62.9% of patients on MRI and in only 20% of patients on ultrasound. In patients who underwent both ultrasound and MRI, the findings noted on ultrasound were compatible with those on MRI in only 13 (37%) of 35 patients. Twenty (57%) of 35 studies had abnormal MRI findings with normal ultrasound findings. Despite the lower reliability, ultrasound may be used as a therapeutic tool for injection of corticosteroid.24,26,27
Zissen et al27 performed 65 corticosteroid injections with ultrasound guidance as both a diagnostic and therapeutic tool. All but 9 patients reported immediate symptom relief. There were no side effects identified following injection. Half of the patients experienced symptomatic improvement for 1 month, and 24% reported prolonged symptom resolution for more than 6 months.
Conservative Measures. No consensus exists in the literature regarding an appropriate algorithm for nonoperative management. The standard “rest, ice, compression, elevation” method is used acutely.18,24,36,37 Nonsteroidal anti-inflammatory drugs may also be used, along with physical therapy.18,24,37 The literature does not contain quantifiable benefits of conservative methods; conservative measures are mostly historically represented.18,24
Across the literature, most patients have a trial of conservative treatment prior to more invasive surgical management. Specific conservative treatment protocols varied across all publications.7,8,12,15,27
In a double-blind, placebo-controlled trial investigating 45 patients allocated into a meclofenamate, diclofenac, or placebo treatment group for 7 days after sustaining a hamstring injury, Reynolds et al35 analyzed patients on days 1, 3, and 7. At the end of day 7, the visual analog scale (VAS) was employed, swelling was assessed, and muscle testing was performed. When assessing the most severe injuries, the placebo group experienced a significantly lower pain score than the treatment groups (P<.05). Reynolds et al35 concluded that nonsteroidal anti-inflammatory drugs offered no benefit in the treatment of an acute hamstring injury.
Eccentric training may be performed as both a preventive and treatment modality in hamstring tendinopathy. In a study by Peterson et al32 of male soccer players, the Nordic Training Program reduced the rate of new hamstring injury by more than 60%. Askling et al30,31 performed 2 different studies investigating rehabilitation protocols among Swedish football players and sprinters/jumpers. Athletes were placed into 2 protocols. The L-protocol loaded the hamstrings with eccentric motions. The C-protocol used standard exercises without a focus on eccentric motions. In both studies, there was a significantly shorter time to sport return using the L-protocol (mean of 28 days for football players and 49 days for sprinters/jumpers). Askling et al30,31 concluded that a rehabilitation focus on lengthening exercises will lead to a faster return to sport. Sherry and Best29 discussed a prospective randomized study of 24 athletes assigned to static stretching vs agility and trunk stabilization rehabilitation. Quicker return to function was noted with the agility and trunk stabilization exercises.29 Malliaropoulos et al33 discussed 80 Greek athletes with hamstring injuries; they treated all athletes in the first 48 hours with protection and rest, ice, compression, and elevation. Then, these athletes were assigned to identical rehabilitation stretching groups that varied by the number of stretching sessions. The group with a greater frequency of stretching had a statistically significant shorter time to return to normal range of motion.33 Cibulka et al38 assessed hamstring injuries and their relationship to associated sacroiliac joint dysfunction. The experimental group underwent manipulation of the sacroiliac joint and had statistically significant increased peak torque in the hamstrings following manipulation.38
In addition to noninvasive rehabilitation protocols, more invasive but nonsurgical methods exist. Zissen et al27 showed fair results with ultrasound-guided injection of corticosteroid for 65 patients. Thirty-eight (58%) of the 65 were available for phone survey following injection. All but 9 patients reported immediate relief of symptoms. Fifty percent of the patients experienced improvement of symptoms for at least 1 month, and 29% reported prolonged resolution of symptoms. No complications were identified. Despite the risks, corticosteroids have been shown to alleviate symptoms.27
Wetzel et al7 conducted a study of 15 patients with 17 proximal hamstring tendinopathies managed by 3 surgeons. Their protocol involved 6 to 12 weeks of physical therapy with 1 week of nonsteroidal anti-inflammatory drugs, followed by as-needed nonsteroidal anti-inflammatory drug use. For 10 patients who failed to improve, PRP injections were offered. Five of these patients continued to be treated with conservative measures without PRP injections. This study showed statistically significant improvement in the VAS and Nirschl phase rating scale for patients treated with PRP injections, but no statistically significant improvement for those patients treated with conservative measures only.7
Fader et al15 treated 18 patients with PRP injections, all of whom had failed conservative treatment options consisting of a 6-month course of activity modifications, physical therapy, and oral anti-inflammatory medications. Following PRP injection, patients were restricted from taking anti-inflammatory medications and in their activities to minimal chores for 1 week. Jogging with progression to incline running was initiated from 2 to 3 weeks after injection. Following this protocol, 5% of patients had 80% or greater improvement in subjective pain at 6 months after injection, and the total average improvement was 63%.15
Davenport et al24 also required all patients to have trialed conservative measures prior to treatment with PRP or whole blood. Patients must not have had a corticosteroid injection within 6 months of treatment with PRP to be included in the study.24
Cacchio et al8 performed the first level I study investigating SWT for chronic proximal hamstring tendinopathy in professional athletes. This was a randomized controlled study that enrolled 20 patients in the SWT group and 20 in a traditional conservative treatment group. At 3 months, the SWT group had statistically significant improvement in the VAS; the traditional conservative treatment group did not. At an average of 9 weeks, 16 of the patients in the SWT group returned to their professional sport. None of the patients in the traditional conservative treatment group returned to sport.8
In the treatment of chronic proximal hamstring tendinosis and partial tearing of the proximal hamstring origin, a trial of conservative measures is typically completed prior to surgical intervention. Bowman et al12 required a minimum of 6 months of nonoperative management with activity modification, physical therapy, nonsteroidal anti-inflammatory drugs, and PRP prior to surgical intervention. Benazzo et al9 required all 17 patients to undergo a trial of conservative therapy before surgery. Specific therapy included local steroid injections in 7 patients, physical therapy in 8 patients, soft tissue mobilization and isometric exercises in 12 patients, local and systemic nonsteroidal anti-inflammatory drugs in 5 patients, and eccentric contraction exercises in all 17 patients.9 In a retrospective study of 28 athletes, Lempainen et al10 noted that 18 of 28 were in too much pain to trial nonsurgical management and proceeded with surgical treatment. All patients in the Lempainen et al10 and Puranen and Orava36 studies required a trial of conservative measures prior to surgery.
Time Until Surgery. An average time of 15 months from symptom onset to surgical procedure was discussed across all surgical articles.9,10,12,13,36 Among the 4 PRP publications, 2 discussed time to treatment. Patients differed significantly regarding time from onset to PRP treatment, averaging 10 months7 and 32 months.15
Surgical Technique. Surgical options include debridement alone vs debridement and tendon repair with or without neurolysis. All studies employed prone positioning, and all surgeons used some degree of leg flexion during the procedure, ranging from 20° to 90°. Three of the 6 surgeons performed neurolysis regularly; Bowman et al12 rarely performed this procedure. Lempainen et al6 always investigated the sciatic nerve, but commented that full neurolysis was rarely indicated. Tendon repair was performed with suture or suture anchors.
Benazzo et al9 performed a partial transverse tenotomy of the involved tendon, followed by a sciatic nerve release from the ischial tuberosity to 10 cm distal. Lempainen et al6,10 preferred a vertical incision in one article and a transverse incision in another. In the first, they performed sciatic neurolysis, followed by tendon reattachment.10 In the second, they performed a transverse incision and tenotomy of the semimembranosus, followed by tendon reattachment and sciatic nerve exploration.6 Bowman et al12 reported using a transverse incision with debridement of all diseased tissue, followed by anatomic primary repair of the native footprint with suture anchors placed in an “X” pattern. Aldridge et al13 performed a longitudinal incision over the ischium and frequently exposed the sciatic nerve; the ischial tuberosity was prepared to reveal the defect, then 2 or 3 anchors were inserted.
Rehabilitation Protocol. Rehabilitation protocols differed across publications. All postoperative protocols recommended partial weight bearing for 2 to 6 weeks.9,10,12,13 Two of the 6 studies reported use of aspirin or enoxaparin for deep venous thrombosis prophylaxis.9,12
Lempainen et al6,10 discouraged sitting for 2 weeks. Three studies specifically recommended return to swimming at 3 weeks.6,9,10 All recommended limiting hip flexion for several weeks. A brace was used for immobilization postoperatively in all studies, but the type of brace differed. Return to biking and concentric stretching occurred at an average of 4 weeks, and running at an average of 11 weeks.6,9,10 Full return to sport averaged 5 months.9,10,12
Rehabilitation protocol varied between PRP studies. Fader et al15 recommended no anti-inflammatory drugs for 6 weeks, limited activity for 1 week, return to jogging at 3 weeks, and activity as tolerated after 3 weeks. Wetzel et al7 advocated for limited hip flexion not past 30° for 3 weeks, then physical therapy for 6 weeks. An abduction brace was used for 5 patients.7 Hamid et al28 recommended reduced activity for 48 hours following injection, and then required physical therapy weekly.
In the SWT article, no activity limitations were recommended. All patients applied ice for 4 hours following the procedure.8
Follow-up. Surgical patients were followed for an average of 37 months. Patients who received PRP were followed for an average of 5.5 months. Patients who received SWT were followed for 11 months.
Anecdotal evidence exists related to high-level athletes having complications, including muscle atrophy, skin hypopigmentation, delayed healing, and repeat injury, following corticosteroid injection.7
Complications were reported in all studies involving surgical treatment. Complications occurred with an overall incidence of 11% (n=30). The most common complication was paresthesias, which were mostly intermittent and related to activity.6,9,10 Three hematomas developed postoperatively.9,36 One patient developed a deep venous thrombosis. There were 8 skin complications with no incidence of deep infection.6,10,12,36 Five patients (1.05%) had discomfort at the incision site with prolonged sitting.12,36 There was 1 instance of muscle atrophy.36 No complications were reported in the corticosteroid, whole blood, or SWT groups. In the PRP group, 1 patient developed sciatic nerve irritation that resolved.
Scores. There is no internationally recognized scoring system for hamstring tendinopathies.9 All studies employed a questionnaire to measure outcomes. The VAS was used in the PRP study by Wetzel et al7 and showed improvement after treatment from an average of 8.7 to 0.7 at 5-month follow-up (P<.5). No statistically significant improvement was achieved in the conservative cohort treatment group, which improved from 7.4 to 1.2 with treatment.7 Fader et al15 showed an average improvement of 63%, with 55.6% of patients having more than 80% improvement following treatment with PRP. Patients treated with SWT had statistically significant improvement in VAS from a mean of 7.1 to 2.1 at 3-month follow-up. The conservative therapy group in this SWT study changed from a VAS of 7.0 to 6.8. Significant differences were seen with SWT vs conservative treatment.8
Four surgical outcome studies described patient-reported outcomes as excellent, good, fair, or poor for a total of 213 patients.6,9,10,36 A total of 110 patients (52%) had excellent outcomes, 93 patients (44%) had good outcomes, 9 patients (4%) had fair outcomes, and 15 patients (7%) had poor outcomes. Eighty-three patients (82%) were satisfied or somewhat satisfied following surgery.9,10,12,13
Wetzel et al7 also employed the Nirschl phase rating scale to assess outcomes. This scale is a 7-phase assessment of pain and activity limitations. The PRP cohort group had a statistically significant reduction in the Nirschl phase rating scale score from 5.5 to 1.5 following treatment. The conservative therapy group had a nonsignificant reduction in the Nirschl phase rating scale score from 4.4 to 2. Despite significant reduction of the Nirschl phase rating scale score in the PRP cohort group, subsequent comparative analysis between the PRP and conservative groups using the Nirschl phase rating scale both before and after treatment revealed no statistically significant differences.7 Hamid et al28 used a secondary measure of change in severity of pain score, the Brief Pain Inventory–Short Form. Patients in the PRP subgroup had lower pain scores at all measures.
The SWT study also used the Nirschl phase rating scale at baseline and 3 months following treatment. A statistically significant improvement in the Nirschl phase rating scale score was seen in patients treated with SWT, and a nonstatistically significant worsening in the Nirschl phase rating scale score was noted in the conservative group.8
Conservative, noninvasive articles assessed outcome based on return to sport. No standardized scoring systems were used to assess outcome.
Return to Sport/Activity. Across all surgical outcome studies consisting of 266 patients, 99% of patients reported improvement following surgery. Ninety-nine percent of patients returned to strenuous activities and sports following surgery.6,9,10,12,13,36 In the SWT group, 16 of 20 patients returned to sport at an average of 9 weeks.8 No patients in the SWT study by Cacchio et al8 in the conservative therapy group returned to sport. In the PRP studies, Fader et al15 noted that 55% of patients had more than 80% improvement, but return to sport was not reported. In the study by Wetzel et al,7 2 of 5 patients in the conservative group returned to sport, and 9 of 13 in the PRP group returned to sport. Hamid et al28 used return to sport as a primary outcome measure in their PRP randomized controlled trial. Half of the patients in the PRP group made a full recovery by 26 weeks; the control group required a much longer time of 39 weeks for half to return to sport. The PRP group elucidated a statistically significant time to return to play.28
Davenport et al24 showed that patients injected with whole blood, compared with PRP, had greater improvement in all measures at 12 weeks. Patients treated with PRP injections, compared with whole blood, had greater improvements in all measures at 6 months. Statistically significant improvements occurred in activities of daily living scores at 6 months, sport-specific function scores at 2 weeks, and international Hip Outcome Tool-33 scores at 2 weeks, 12 weeks, and 6 months after injection with PRP. Patients injected with whole blood had significantly decreased pain with 15 minutes of sitting at 6 months and a greater proportion of improved tendon appearance on ultrasound at 6 months. All patients injected with either PRP or whole blood had increased evidence of calcifications on ultrasound (45% for PRP and 50% for whole blood).24
Conservative, noninvasive treatment programs published regarding outcomes of rehabilitation concluded that increased frequency of sessions with an emphasis on lengthening and core exercises yielded quicker return to play.29–33, 35