Alarming is the word Von Parat, Roos, and Roos choose to describe their findings of significant knee osteoarthritis in middle aged former soccer players who had sustained a rupture of the anterior cruciate ligament more than a decade previously.¹ Not only was there a high incidence of osteoarthritis in these former players, but reconstruction of the anterior cruciate ligament did not appear to provide protection from degenerative change in the knee.

In a 14 year follow up of subjects who had formed the basis of an earlier study,² the authors identified 238 male soccer players who were diagnosed with anterior cruciate ligament injuries in 1986. They were able to contact 205 of the players. Of these, 154 answered questionnaires and a further subgroup of 122 consented to have knee radiographs.

In just over half of the 95 subjects with radiographic changes, there was osteoarthritis equivalent to Kellgren-Lawrence grade 2 or higher. Of the subjects who answered questionnaires, 58% had undergone anterior cruciate ligament reconstruction. Interestingly, and perhaps surprisingly, there was no difference in radiographic outcome between those that had been treated with anterior cruciate ligament reconstruction and those who had not undergone reconstructive surgery. On the other hand, subjects who had sustained a meniscal tear had an increased prevalence of osteoarthritis, but the severity of radiographic changes was similar to those who did not have meniscal pathology associated with their anterior cruciate ligament rupture.

This study raises some interesting issues. The lack of benefit of anterior cruciate ligament reconstruction in terms of protection from osteoarthritis represents a challenge to knee surgeons. The authors do acknowledge that because there was no randomisation with respect to surgery, their findings are difficult to interpret. However, it is clearly difficult to design and implement a randomised comparison of reconstructive surgery with non-operative management of anterior cruciate ligament injuries, especially in active sportspeople. We therefore have to rely on studies such as the one under consideration to determine the best advice for a young, active sportsperson who has sustained a rupture of the anterior cruciate ligament. Although successful reconstruction will theoretically provide some protection from further damage to the menisci and articular cartilage, findings such as those being reported challenge this concept.

The authors state that there has been no study to date that shows that anterior cruciate ligament reconstruction protects the knee from osteoarthritis. However, Fink et al³ have reported a lesser severity of osteoarthritic change, albeit with a higher prevalence, in those who had undergone anterior cruciate ligament reconstruction and returned to sport compared with those who did not have reconstructive surgery but nonetheless returned to sporting activities. It is noteworthy that, in both the current study and that of Fink et al, a patellar tendon graft was used for all reconstructive procedures.

Recently Pinczewski et al⁴ reported an increased prevalence of osteoarthritis associated with patellar tendon grafts compared with hamstring grafts seven years after anterior cruciate ligament reconstruction. We have recently used three dimensional motion analysis to compare the biomechanical function of subjects who had undergone anterior cruciate ligament reconstruction using either patellar tendon or hamstring tendon grafts.⁵ We identified differing patterns of abnormal moments about the knee in the two groups. The patellar tendon group had a reduced external knee flexion moment at mid stance, whereas the hamstring group had a reduced external extension moment at terminal stance. A reduced external knee flexion moment associated with patellar tendon grafts has also been reported in the setting of single limb landing tasks.⁶ This may provide an explanation for the increased incidence of osteoarthritis in the patellar tendon group as reported by Pinczewski et al, as reduced external knee flexion moments may be associated with reduced attenuation of forces passing across the knee joint. The apparent lack of protection from osteoarthritis after anterior cruciate ligament reconstruction as reported by von Parat et al therefore needs to be evaluated cautiously.

Efforts have been made by various authors, including those of the current study, to determine the role that soccer participation in itself plays in the development of knee osteoarthritis. It continues to be difficult to establish the relative roles of recognised injury, unrecognised injury, and simple participation in soccer in the development of secondary osteoarthritis. Nonetheless, there is evidence to suggest that soccer participation, particularly at an elite level, does in itself contribute to development of knee osteoarthritis.⁷ The current study could have shed considerable further light on this issue by including radiographic assessment of the contralateral knee as part of the research protocol. This would have provided an excellent control cohort of knees with which to compare the anterior cruciate ligament deficient or reconstructed knees. The rate of return to soccer after the initial injury to the anterior cruciate ligament would also help to tease out the role of participation in soccer per se in the pathogenesis of osteoarthritis in this population. Unfortunately this information is not provided.

Whatever the cause of osteoarthritis in the current group of former soccer players, it did not appear to be associated with poorer function compared with no osteoarthritis. Similarly, anterior cruciate ligament reconstructive surgery and meniscal surgery did not appear to influence functional outcome. However, despite the lack of an association between these variables and functional outcome, the vast majority of subjects had reduced their level of activity after their knee injury, and most of these subjects had reduced their activity specifically because of their knee injury.

Somewhat paradoxically, subjects reported better knee function 14 years after injury than they had at seven years after injury. Given that at the more recent follow up there was a decreased rate of participation in soccer, this presumably reflects the assumption that the subjects had come to accept the limitations imposed by the knee injury. This highlights the difficulties associated with long term follow up of patients who have undergone anterior cruciate ligament reconstruction, or indeed those who have not had their anterior cruciate ligament injury treated surgically. Sports participation may not be a useful outcome variable in the longer term as reduced sports participation may simply reflect a change in priorities on the part of the individual, rather than be regarded by the individual as a significant functional loss.

An important methodological concern pertaining to the study under review is the potential for selection bias. As the authors correctly observe, it is possible that those subjects with knee symptoms would have been more likely to participate in the current follow up than those whose knee continued to function well. This could in turn have resulted in an overall poorer outcome than if all of the original 238 had been contactable, and had responded and undergone a radiographic assessment of their affected knee. Clearly the logistic problems of such long term follow up are considerable, and the authors should be congratulated rather than criticised for their efforts.

Whatever its shortcomings, the message from the current paper is clear. Anterior cruciate ligament rupture, whether treated surgically or not, is clearly associated with an increase in osteoarthritis in former soccer players. It remains to be seen whether improved surgical techniques of anterior cruciate ligament reconstruction and the use of grafts other than the patellar tendon can offer greater protection, while at the same time allowing resumption of sporting activities.