The gracilis tendon autograft is a safe choice for orthopedic reconstructive procedures: a consecutive case series studying the effects of tendon harvesting

The gracilis tendon is often used as an autograft for ligament or tendon reconstruction. Most commonly it is used in combination with the semitendinosus tendon as a hamstring graft in anterior cruciate ligament (ACL) reconstructions, but it is also frequently used in other knee reconstructive procedures as well as in shoulder, elbow, hip and ankle surgery [1‐6]. While allografts are an alternative to autografts there are indications that they stretch more postoperatively and they are not readily available in all countries [7, 8]. In 2011, at our institution, we started using gracilis autografts in coracoclavicular (CC) ligament reconstructions for treatment of chronic acromioclavicular joint dislocations [9].

While graft site morbidity and knee flexor strength after ACL reconstruction using hamstring grafts has been thoroughly studied [10, 11], there is little evidence regarding the knee related outcome after isolated gracilis tendon harvesting from a limb in which no reconstructive procedure is planned. Two studies touching on this subject concern patients undergoing primary, unilateral ACL reconstructions that were randomized with regard to which leg the hamstring graft would be harvested from [12, 13]. Both of these trials report that graft harvesting is associated with a decrease in knee flexor strength that is resolved within 12 months postoperatively and they do not advice against using the unaffected leg as a graft donor site.

While providing important information, the trials above studied the effect of harvesting a two-tendon hamstring autograft for use in ACL-reconstruction and these results cannot be directly translated to a situation where an isolated gracilis graft is used for reconstructive procedures in other parts of the body than the knee. After reviewing the literature we could find only one study investigating this issue in which 22 patients were retrospectively included after having undergone gracilis tendon harvesting to provide grafts for various foot and ankle procedures [1]. The patients were significantly weaker in knee flexion in the donor leg compared to the contralateral one but displayed minimal donor site morbidity.

In our study we aim to present further evidence regarding the outcome and possible morbidity related to gracilis tendon harvesting. This will allow surgeons to perform a more detailed risk-benefit analysis and patients to receive a more accurate preoperative information. Our hypothesis is that gracilis tendon harvesting leads to a weakening in knee flexion but not a reduction in subjective knee function.

This study comprised 24 consecutive patients operated between 2011 and 2016 with CC ligament reconstruction using an autogenous gracilis tendon graft. The hypothesis for this study was formulated in the end of 2011 after 5 patients had already undergone surgery, only postoperative outcome measures were available for these 5 patients and the study is therefore of combined prospective and retrospective design.

A total of 24 patients were eligible for the study and two were excluded. One due to mental illness and one because of chronic pain syndrome, both of these patients were from the group included prospectively. Out of the 22 patients available for follow up there were 4 females, the mean age was 44 years (22–62) and the graft was harvested from the right leg in 14 cases.

The five KOOS subscales, pain, symptoms, activities of daily living (ADL), sports and recreation (SR) and quality of life (QoL), were analyzed separately and the mean scores are presented in Fig. 1. There was no statistically significant change between baseline and 12 month values for any of the KOOS subscales.

Testing of isometric hamstring strength was performed at a median of 26,5 months (14–56) postoperatively and was completed by all of the 22 patients. The operated leg was significantly weaker in both 60° and 90° of flexion performing a mean force of 93 and 83% of the non-operated leg in the two angles respectively. Complete results are shown in Table 1.

Analysis of the hospitals’ medical journal database revealed that one patient had problems with a slow healing incision with no signs of infection and the same patient suffered a minor hamstring sprain after stumbling. Other than these two episodes that both resolved spontaneously no donor site complications were found.

This study shows that there was no statistically significant decrease in KOOS scores in any of the subscales 12 months after gracilis tendon harvesting compared to baseline values. This indicates that the use of autogenous gracilis grafts is well tolerated by patients and does not impair subjective knee function. However, the donor limb was significantly weaker in isometric knee flexion compared to the unoperated side.

While the outcome after ACL reconstruction using ipsilateral hamstring grafts has been thoroughly studied there are very few articles available on the effects of isolated gracilis tendon grafting from a knee not planned for reconstructive surgery. To our knowledge the study by Cody et al. is the only one investigating this issue [1]. In their retrospective study 70 patients had undergone hamstring tendon autografting for use in foot and ankle reconstructive procedures and in 22 of these an isolated gracilis tendon graft had been harvested. The patients were interviewed postoperatively and 95% of the patients responded that they were either satisfied or very satisfied with their operative result and all patients would recommend the surgery to someone else. There were no serious donor site sequelae and the authors conclude that hamstring autografts can be used with high patient satisfaction and minimal morbidity. The study by Cody et al. is limited by a follow up rate of only just above 50% and by the fact that they did not use a patient related outcome measure for follow up. In our study we used the validated outcome measure KOOS that provides a more comprehensive evaluation of the knee and leg function and we can confirm the findings by Cody et al. [1].

Cody et al. also measured isokinetic knee flexion strength between 13 and 56 months postoperatively [1]. They found that gracilis tendon harvest lead to a significant decrease in strength compared to the unoperated side. In our study we measured isometric and not isokinetic force but as the two are very closely related [16] our results are comparable. In accordance with Cody et al. [1] we found that the donor leg strength was 93% of the non-operated side when tested at 60° of knee flexion and 83% at 90°. Considering that both Cody et al. [1] and this study showed that the subjective outcome of the donor leg was not affected by the surgery the measurable decrease in strength seems to be clinically unimportant in our study populations. It is, however, possible that the hamstring strength deficit might be of clinical importance in patients with very high demands such as those participating in certain sports. It has been suggested that a weakness in knee flexion might increase the risk of ACL tears [17] and it has also been shown that low hamstring strength is a risk factor for hamstring strain injuries in professional football players [18]. Structured physiotherapy has been shown effective in rebuilding hamstring strength [12] and should, therefore, be considered after gracilis tendon harvest.

There are two further studies concerning the outcome after hamstring tendon harvesting without further knee related procedures in the donor limb. In the study by Yasuda et al. 70 patients were randomized to have a two-tendon hamstring graft harvested from either the ipsilateral or contralateral leg when undergoing planned ACL-reconstructions, 34 patients were included in the contralateral group [12]. All subjective problems, such as activity related soreness, restricted range of motion, and reduction in the points in the Cincinnati sports medicine center rating scale, were resolved by 3 months postoperatively. In the second, similar, study by McRae et al., 50 patients were randomized to the contralateral harvest group [13]. The main outcome variable was the ACL quality of life outcome measure and there was no difference between groups at any point in time during follow up. The score did increase with time at each of the three follow up visits until 12 months postoperatively but no further improvement was seen at 24 months. While our study cannot be directly compared to the two above because of the differences in tendons harvested and follow up protocols we are able to provide further evidence that the negative effects of hamstring tendon grafting are resolved by one year postoperatively.

A strength of our study is that we can confirm the previous findings of Cody et al. [1] that isolated gracilis tendon harvesting leads to a reduction of knee flexion strength. We also deepen the understanding of the subjective outcome after this procedure by being the only study presenting the results of a validated patient related outcome measure. As gracilis autografts are widely used in orthopedic procedures outside of knee surgery these results are important to establish the safety and outcome of the harvesting procedure in such settings and allow surgeons to provide patients a detailed preoperative information [1‐6].

The main weakness of this study is that 5 of the 22 patients underwent surgery before the study was started and are therefore lacking preoperative outcome measures. In these 5 cases we compared postoperative KOOS scores to age and gender matched normative values acquired from a population based study of 568 persons [15]. Further, while there was no statistically significant decrease in KOOS scores one year postoperatively we did see a slight reduction in each of the five subscales and with a larger sample size this might have turned out to be significant. The differences were, however, not larger than what is considered the minimal clinically important change in any of the subscales [19]. Another weakness is that we did not systematically examine the donor site in the post operative period to look for local complications such as wound infection, hematoma or muscle rupture. Instead we used medical journal review in the Hospitals’ database to register complications that the patients sought care for. It is possible that patients with complications have been treated at other caregivers but this is unlikely as we are the only orthopedic clinic in the area and postoperative complications are usually referred to us. Further, we only measured isometric knee flexion strength and did not assess isokinetic performance, but as the two are closely related and our results are in accordance with other studies we do not believe this choice of method to have caused erroneous conclusions [1, 16].

Gracilis tendon harvesting results in a weakness of knee flexion but does not have significant impact on subjective knee function and is a procedure that can be recommended when an autogenous tendon graft is needed.