Proximal hamstring avulsions : surgery or not?

Abstract

In paper 1, a retrospective cohort study of patients treated for proximal hamstring avulsions at Danderyd Hospital from 2007–2013 was conducted. The study was performed in 2015. The primary outcome was the subjective patient reported outcome Lower Extremity Functional Scale (LEFS), and the exposure was surgical treatment. We hypothesized that there would be no difference in the LEFS, between the two treatments at this long-term follow-up. The results showed similar LEFS scores in the surgically treated 74 (SD±12) and nonsurgically treated 72 (SD±16) patients, which were also true after adjusting for confounders. There were some differences between the groups, with surgically treated patients having more severe injuries and reporting more hours of physical activity at the follow-up than nonsurgically treated patients. To obtain an evidence-based treatment decision for proximal hamstring avulsions, studies at higher scientific levels are needed.

Paper 2 consists of a cross-sectional cohort study. Patients treated for proximal hamstring avulsions at Danderyd Hospital between 2007 and 2016 were included. The study was performed from 2018-2019. The main outcomes were the correlation among subjective, patient-reported outcome measurements (PROMS), the Perth Hamstring Assessment Tool (PHAT) score and the Lower Extremity Functional Scale (LEFS) score and their correlation. The secondary outcomes were the correlation of PROMs with objective performance-based tests. Moreover, we explored which activity patients perceived to be most limited several years after injury. We hypothesized that there would be good correlations between the PHAT and LEFS scores and at least moderate correlations between these questionnaires and performance-based tests. We found strong correlations between the PHAT and LEFS scores (r=0.832, p<0.001). The LEFS was more appropriately aligned with the performance-based tests than the PHAT. Of all the physical performance tests performed at follow-up, only the isokinetic test could discriminate between injured and uninjured legs. Patients most frequently reported activity limitation was running. Since the PHAT, LEFS and physical performance-based tests seem to assess different aspects of recovery, both subjective and objective outcome measures are recommended to be used for follow-up after proximal hamstring avulsion.

In paper 3, we performed a cross-sectional study on the same cohort of patients in the same setting as in Study 2. The primary outcomes were hamstring muscle volume and fatty infiltration at least 2 years after injury. The secondary outcome was the correlation of these parameters with isokinetic muscle strength. The conditions of the injured and uninjured legs were compared. We hypothesized that the injured leg would have greater fatty infiltration and atrophy than the uninjured leg at follow-up and that these findings would correlate with muscle weakness. We found that, on average, the hamstring muscle volume was reduced by 9% (SD±11%) compared to that of the uninjured leg. Fatty infiltration was significantly more severe in the injured hamstrings than in the uninjured hamstrings (p<0.001). Reduced muscle volume and increased fatty infiltration were significantly weakly correlated with isokinetic strength test results (r=0.357-494, p< 0.001-0.013). At follow-up, we concluded that fatty infiltration and muscle atrophy are likely to occur as a result of proximal hamstring avulsions, and muscle quality impairment is weakly correlated with muscle weakness in the injured leg.

In Study 4, a randomized, noninferiority, multicentre, preference-tolerated clinical trial was performed. Patients from ten study sites in Sweden (8) and Norway (2) participated. Patients were eligible for inclusion if they had an acute (within 4 weeks) proximal hamstring injury and were aged 30 to 70 years. Patients were randomly selected to undergo surgical or nonsurgical treatment. Surgical treatment included reinsertion of the tendons followed by rehabilitation, and nonsurgical treatment included rehabilitation only. If the patients and doctors could not reach a consensus on treatment, the patients were invited to join an observational follow-up cohort. The primary outcome was the PHAT score at two years post treatment. A noninferiority margin of 10 points on the PHAT was set for the lower limit of the two-sided 95% CI. The secondary outcomes consisted of the LEFS score, physical performance-based test results and muscle quality analysis results on MRI. We enrolled 119 patients in the randomized trial and 97 in the observational cohort. According to the intention-to-treat analysis, the mean PHAT scores were similar, with mean (±SD) scores of 80.4(±19.3) and 77.7(±20.0) in the surgical and nonsurgical groups, respectively. The prespecified inferiority limit was not crossed (mean difference, -2.1; (95%CI -9.3 to 5.1) p =0.017 for noninferiority). According to the inverse probability weighted analysis of both cohorts combined, the mean difference in the PHAT score was -2.6 (95%CI, -7.9 to 2.8). Analyses of secondary outcomes including the mean LEFS score difference of -2.1 (95%CI, -5.7 to 1.5) supported noninferiority.

The conclusion was that patients with proximal hamstring avulsions who were treated nonsurgically do not have worse PHAT scores than patients who were treated surgically; therefore, the treatment of choice for middle-aged patients should be nonsurgical treatment.