Hamstring Injuries in Athletes: Evidence-Based Treatment

Justin W. Arner, MD; Michael P. McClincy, MD; James P. Bradley, MD


J Am Acad Orthop Surg. 2019;27(23):868-877. 

In This Article

Treatment and Outcomes

Proximal Hamstring Avulsions

Patients with a suspected proximal avulsion injury are initially imaged with an AP pelvis radiograph to evaluate for possible bony avulsion, which is most commonly seen in skeletally immature patients. MRI is recommended when a hamstring avulsion is suspected to evaluate the insertion site and quantify the number of involved tendons as well as their degree of retraction. A careful examination of the sciatic nerve is essential because scaring in chronic proximal hamstring injuries may cause neuropathy. MRI is also useful in these cases because the sciatic nerve can be traced to identify points of potential tethering to the retracted tendons.[16] A recent meta-analysis reported a 23.17% complication rate with surgical fixation, 7.99% being neurologic.[17]

Multiple studies with small patient numbers have evaluated surgical repair of hamstring avulsions with many reporting good outcomes in acute ruptures, whereas chronic cases showed more variable improvement.[16,18–20] In a seminal article in 2008, Wood et al[21] evaluated 72 patients with proximal hamstring avulsions at an average 2-year follow-up using a three-anchor modified Mason-Allen suture technique. They reported 84% return of strength and 89% return in endurance, with poorer prognosis in severe retracted tears or with scaring to the sciatic nerve, both causing a more technically challenging repair. However, repair afforded an improved prognosis even in chronic cases where 80% returned to their preinjury level at 6 months. In the first study to use validated outcome measures, our group reported on 40 acute and 12 chronic avulsions at an average 2.8-year follow-up. We found both acute and chronic repairs to be successful regarding outcome scoring, with 67% of patients participating in strenuous activities and 98% of patients reporting satisfaction with their outcome.[22] At midterm follow-up, Sandmann et al[23] reported that functional outcomes returned to baseline, knee strength was equivalent to the uninjured leg, and high levels of return to sport were achieved in 15 patients at an average 4.6-year follow-up.

In the skeletally immature patient, ischial tuberosity avulsion fractures account for 10% to 11% of all pelvic avulsion fractures.[24] These are the most common pelvic apophyseal avulsion fractures to undergo surgical fixation, typically by acute open reduction and internal fixation (Figure 7). Case reports have described symptomatic nonunions in cases with >2 cm of fracture displacement, so this degree of displacement is now considered a relative indication for acute repair.[25] Delayed symptoms caused by ischial malunions have also been reported after conservative treatment of ischial tuberosity avulsion, with delayed fragment excision with or without hamstring repair showing notable clinical improvement.[24,25]

Figure 7.

A, Radiograph showing a bony avulsion fracture of the right ischial tuberosity. B, Radiograph status after acute open reduction and internal fixation of the avulsion fracture.

Overall the authors recommend surgical repair if more than one tendon is involved in a soft-tissue injury or if a bony avulsion is displaced >2 cm.[21,26] There is good evidence that if tendon retraction is >2 cm, surgical fixation leads to improved outcomes; however, no studies exist comparing different distances of retraction. Yet, higher amounts of retraction have been linked to poorer outcomes.[21] One study found that 40% of nonsurgically treated proximal hamstring avulsions of <2 cm eventually required surgical intervention.[26] We think that if proximal hamstring retraction is <2 cm, an individualized treatment decision should be made in concert with the patient. Although no studies have yet evaluated the number of avulsed tendons to failure rates, we think that if more than one tendon is torn, greater tension is likely placed on the remaining poor quality tendon thus making surgery necessary. Discussed below are our current techniques for soft-tissue repair of proximal hamstring ruptures.

Surgical Technique. The patient is placed prone on the operating table with the bed slightly flexed at the hips to improve exposure. The leg is prepped free so the knee can be flexed to relieve hamstring tension during reduction and repair. A transverse incision is made in the gluteal crease which can be extended distally in a "T" configuration to access large retracted tears; however, the authors find this rarely necessary. After dissection through the subcutaneous tissues, the gluteal fascia is transversely incised to reveal the gluteus maximus. The authors prefer to split the muscle rather than to dissect inferior to its border because superior retraction of the entire gluteus maximus makes exposure and anatomic repair more difficult.

A longitudinal incision is then made in the hamstring septum over the ischium with care to avoid injury to the sciatic nerve. The hamstring fascia is typically intact; the surgeon must not allow this to mislead them into believing that the hamstrings are intact. A fluid collection or hematoma is often encountered after a vertical fascial incision. Evacuation of fluid and blood allows identification of the torn hamstring tendons which are then tagged, then edges débrided to healthy tissue. Next, the hamstring insertional anatomy at the ischium is identified with the facet being more inferior and lateral than most appreciate. The leg may be rotated during dissection to confirm the ischium is indeed encountered and not the lesser trochanter. Subperiosteal dissection starts proximal and medial and then goes distal and lateral where the tear is typically appreciated. Once the ischium is exposed, the insertion site is curetted and an osteotome is used to fish scale the cortical bone to improve the healing environment. No motorized instruments are used to decrease risk of injury to the sciatic nerve.

Like many repair techniques, bone tunnels with suture have been abandoned in favor of multiple bioabsorbable suture anchors. Miller et al first described using four anchors in a box configuration with simple mattress sutures. The authors have transitioned from five anchors in an "X" pattern to six anchors (3.0 mm PushLock; Arthrex) positioned like the face on a die. Sutures are tied in a simple mattress configuration distal to proximal with the knee flexed. More recently, the authors have also placed two luggage tag sutures at the medial leading edge of the tendon, which are dunked into smaller suture anchors (2.4 mm PushLock; Arthrex) to help improve bony contact (Figures 8 and 9). After a layered closure, patients are made toe-touch weight bearing and placed in a knee brace flexed to 40° for 4 weeks. The brace is then discontinued allowing progressive weight bearing over the next 4 weeks with pain-free arch of motion against gravity. We prefer a knee brace rather than a hip brace because these are too cumbersome for the patient and we do not think they provide a notable advantage. The authors think a knee brace locked at 40° sufficiently decreases tension on the repair and adequately slows patient's activity. A progressive return to sport protocol is then carried out over the ensuing 6 months as previously described.[2]

Figure 8.

Illustration showing the authors' preferred proximal hamstring avulsion fixation technique with six anchors (3.0-mm PushLock; Arthrex) on the ischial tuberosity, in an anatomic simple mattress suture configuration, tying them distal to proximal with the knee flexed. Then, two luggage tag sutures are placed at the medial leading edge of the tendon and dunked into their own smaller suture anchors (2.4-mm PushLock).

Figure 9.

Illustration showing postfixation of proximal hamstring avulsion repair showing the authors' preferred technique with six anchors tied in an anatomic simple mattress suture configuration with two luggage tag sutures placed at the medial leading edge of the tendon.

Timing of surgical intervention appears to be critical, with studies reporting more difficult exposure and fixation as well as more sciatic nerve neurolysis being required in chronic complete tears. Subbu et al[27] defined early repair as within 6 weeks of injury and found those patients to have quicker return to sport and better outcomes, whereas the chronic group displayed increased morbidity and complications. Similarly, Blakeney et al[28] defined chronic as over 3 months from injury, and found both acute and chronic repairs had improved outcomes. However, early repairs did have better outcomes in comparison with chronic repairs in their cohort of 94 patients.

Overall, good results have been reported in acute repairs when matched with those treated nonsurgically with increased strength, less pain, improved return to sport, and higher patient satisfaction. Chronic complete retracted repairs are less predictable and more difficult technically, with added challenges of mobilizing the tendons to the ischium and freeing tendons that may have scarred to the nearby sciatic nerve. Furthermore, chronic tears have higher rates of re-rupture and chronic pain. Despite these difficulties, studies have reported improved outcomes with surgical fixation when compared with those treated nonsurgically.[22,29]

Partial Proximal Hamstring Tears

Partial proximal hamstring ruptures are less well studied, but typically are attributed to chronic overload, commonly encountered in running and Pilates activities. One source of confusion is the differentiation between tendinopathy and partial tears. The literature is unclear regarding the progression of hamstring tendinopathy to partial tendon rupture, but the authors think these injuries begin as tendinosis/tendinopathy patterns, which progress to partial tears with continued stress. MRI is useful to distinguish these pathologies. Tendinopathy is seen as increased signal on T1 but not on proton density sequences, whereas partial tears have an increased signal focus on T2 as the classic sickle sign on coronal cuts, which indicates a poor scar response[2] (Figure 10).

Figure 10.

Sagittal pelvis T2 MRI showing a right partial avulsion of the proximal hamstring origin with the sickle sign.

Our group published the first cohort of partial repairs in 2013 that included 17 patients with an average 32-month follow-up. All 17 patients who had a tear intraoperatively had a sickle sign on MRI. The authors found improvements in all outcome scores with surgical repair. All returned to sport at the same level; however, one had persistent symptoms during competitive distance running.[15] More recently, Piposar et al[26] compared 15 patients undergoing nonsurgical and 10 undergoing surgical treatment for high-grade partial or complete ruptures and found both groups had acceptable subjective and objective outcomes. However, 40% (4/10) of those treated nonsurgically required future surgery for persistent pain or loss of function, indicating that more aggressive early operative treatment may be beneficial. A recent meta-analysis compared partial and complete avulsion repairs and found better strength and endurance in partial tears, but higher patient satisfaction, less pain, and higher complication rates in complete tears.[17]

Adjuncts to nonsurgical treatment, including platelet-rich plasma (PRP), have shown promise in partial hamstring avulsions. Fader et al[30] found that 10 of 18 patients who previously failed physical therapy had greater than 80% improvement in their visual analogue score after leukocyte-reduced PRP injections. They concluded that leukocyte-reduced PRP treatment was safe and benefited most patients, with the major adverse reaction being 72 hours of postinjection soreness. Park et al[31] compared 32 patients who received PRP and 24 who received steroid injections for grade 2 proximal hamstring injuries and found a more favorable response in the PRP group at 1 week. At 4 weeks, PRP had better outcome scores but did not reach statistical significance. The authors' protocol for PRP treatment is one double dose with injection under CT or ultrasonography. Anecdotally, the authors have observed that approximately 50% are able to return to their preinjury level after injection, whereas those with a sickle sign on MRI receive little benefit. Partial tears that fail injections, physical therapy, activity modifications, or patients who want to continue a very active lifestyle are typically recommended for surgical intervention.

Muscle Belly Injuries

Hamstring muscle belly injuries are common and can cause notable pain and prolonged disability. The biceps femoris is most commonly injured (84%), with the semimembranosus making up 12% and semitendinosus 4% of injuries.[32] MRI grade and size of tear have been shown to be useful in predicting return to sport; however, Reurink et al reported that many of these studies have a risk of bias.[2,32,33] T2 hyperintensity is the basis of the most common classification, where grade I is signal of a tendon or muscle without disruption, grade II is disruption of less than half of the width, and grade III being over half disrupted (Figure 11). A more detailed MRI scoring system was developed by our group that has been found to be highly predictive in determining time missed by assigning points based on (1) age, (2) number of muscles involved, (2) location, (3) if the insertion was involved, (4) percent of cross-sectional injury, (5) amount of retraction, and (6) proximal to distal length of injury (Table 2, Supplemental Digital Content 1, http://links.lww.com/JAAOS/A370). These data were based on data from two National Football League (NFL) teams with 43 injuries. Our group was able to separate players by rapid (<1 week), intermediate (2 to 3 weeks), or prolonged return to play (>3 weeks) and correlated these with MRI scores of <10 for rapid return and >15 in those with prolonged return to play. Table 3 (Supplemental Digital Content 1, http://links.lww.com/JAAOS/A370) shows more detailed findings of this study regarding MRI parameters and return to play in these three categories.[8]

Figure 11.

T2 axial MRIs showing (A) grade 1 hamstring strain without disruption, (B) grade 2 hamstring strain with tearing <50%, and (C) grade 3 hamstring strain with >50% tearing.

Conservative treatment of muscle belly injuries is well established including rest, ice, compression, modalities, and NSAIDs with gradual return to activities. The use of steroid injections is controversial, with some reporting quicker return to play and better recovery of contractile tension without ill effects.[8] A recent meta-analysis found rehab to produce the best results with limited evidence to support PRP, agility, or trunk stabilization. Specifically, eccentric strengthening was shown to decrease reinjury and improve muscle fiber length in 14 days; however, many of these studies were markedly underpowered and of fair or poor quality.[34] With data from our group and the authors' experience, Table 4 (Supplemental Digital Content 1, http://links.lww.com/JAAOS/A370) shows the expected time off for in-season hamstring injuries.[8] The authors' protocol for treatment of elite athletes typically involves MRI to localize and grade the injury, followed by a PRP injection within 24 to 48 hours. Second and third injections are repeated 5 to 7 days apart based on response to preceding injections. Physical therapy is instituted and functional testing (ie, plank strength, knee flexion strength at 15 degrees, etc.) is used to determine when the athlete can return to sport. The authors also now incorporate and compare baseline Catapult GPS (Melbourne, Australia) data as RTP criteria where NFL players must be within 1 mile per hour of running speed, have the same work load, and same angular velocity before being released to full play.

A study currently in progress by our group compared 31 NFL football players who received PRP with 30 who did not for grade 2 hamstring injuries. We found PRP injections allowed NFL players to return 1 game sooner (P < 0.05) compared with those who received rehabilitation alone. No re-tears occurred in the PRP group, which is noteworthy because RTP is prolonged if a re-tearing occurs at the same location. Because of the potential financial impact of expedited return to play and decreased re-tear rates, PRP injections for treatment of grade 2 hamstring injuries may be advantageous in professional athletes. The authors believe the use of leukocyte-poor PRP is essential and is the main contributor of the varied results regarding the efficacy of PRP in the literature.[35,36]

Distal Hamstring Injuries

Literature regarding distal hamstring injuries is sparse. Isolated distal biceps femoris tears typically occur from noncontact mechanisms and are commonly part of a multiligament knee injury. Case reports of isolated injuries suggest that acute repair results in improved outcomes.[37] Repair techniques include suture, suture anchors, or transosseous, with the latter having the highest load to failure.[38]

Isolated distal semimembranosus ruptures are rare and have even more limited literature. Case reports indicate that acute surgical repair is indicated and yields good results, while myotendinous junction repairs results in poor outcomes.[39] Distal semitendinosus injuries typically occur during sprinting and present with popliteal tenderness and skin puckering following retraction. Treatments include rest followed by physical therapy, débridement, or surgical repair. The literature is unclear regarding what factors may predict success of nonsurgical treatment. Cooper and Conway[40] evaluated 25 elite athletes and found nonsurgical treatment resulted in poor outcomes with 42% unable to recover, whereas acute débridement quickened recovery in those with a painful mass and issues with knee extension. Although rarely isolated, surgical intervention of distal hamstring injuries should be considered in athletes.