4.1 Summary of evidence
This systematic review assessed the effects of platelet concentrates on graft maturation and graft-bone interface healing in ACL reconstruction in human patients. The, somewhat limited, evidence from the current literature suggests that platelet concentrates may improve the rate at which grafts achieve a low signal on T2 imaging and histologic evidence of graft remodeling but have only little to no effect on tunnel healing. ()
Conclusion presented in included studies
Currently, it is estimated that 125,000 ACL reconstructions are performed per year in the US alone(36
). The outcome of such a procedure depends on two biological events that occur after the implantation–maturation of the graft and integration and secure fixation of the graft into the osseous tunnel. After implantation, the tendon graft matures by changing into a more ACL-like structure by changes in ultrastructure(38
), and innervation(43
). During the first two weeks after implantation there is histological evidence for central necrosis and subsequent hypocellularity in the graft, followed by a phase of vascularization and repopulation by host cells from week 4 to 10. Finally, the graft becomes more similar to the native ACL over 4 to 6 months. During this time the biomechanical strength of the graft is significantly compromised, but recovers over 6 to 12 months. At the same time, a fibrous interface develops between the implant and the surrounding bone. Weiler et al defined this interface as a disorganized, highly cellular and vascular granulation tissue(15
). Over time this tissue matures into a hypocellular and hypovascular dense connective tissue with Sharpey-like collagen fibres. Platelet concentrates are being used to stimulate both processes.
The first outcome for which we collected evidence in this study was graft maturation. Seven studies report on this outcome, but use different methods, different scores, and different follow-up time points. For those that use MRI, it is noteworthy that a healthy ligament shows a homogenous, low intensity signal in T2- and PD-weighted images. Generally, four out of seven studies observed positive effects of platelets on the intra-articular grafts. Such effects included reaching a higher level of tissue homogeneity on MRI or in histology, or reaching such a state earlier. Maybe more interesting is taking a look at three studies that did not find a difference in graft maturation. Nin et al measured graft intensity on proton density-weighted (PD) and T2-weighted images. While they found no significant differences, the mean scores were 230 vs 190 on PD and 75 vs 61 on T2 images for platelet treated grafts and controls, respectively, consistent with a 20–25% improvement based on platelet use. This numbers suggest the possibility of a type-II error, or “not seeing what truly is there”, based on insufficient sample size(48
). Figueroa et al found generally low numbers of hypotense grafts on MRI for both groups (63.2% in the platelet group, 42.11% in the control group) despite the relatively long follow-up of 6 months, but again found a 20% difference and the same argument of low power made for Nin et al above can be made here.(28
). Vogrin et al, in turn, focused specifically on difference in vascularization, not overall graft maturation, and found no significant difference in this parameter between ACL reconstructions with or without platelets(34
). Interestingly, the studies that did not find a difference had various lengths of follow-ups (3, 6, and 24 months) and high concentrations of platelets (12× and 5×), ruling these parameters out as reasons for different outcomes. One explanation that is in agreement with all the available evidence and biologically plausible is that the addition of platelet concentrates, i.e. a source of growth factors, speeds up the maturation process, including cellular repopulation and remodeling, as shown by Orrego, Radice, and Ventura, as well as animal studies(30
). However, the beneficial effect of platelets wears off rapidly and no difference can be seen at later time points, as reported by Figueroa and Nin, other than on the microscopic level as demonstrated by Sanchez et al(32
). Also, some (late stage) ligamentization processes, such as vascularization, seem not to be affected by platelets, as Vorgin and coworkers observed at 3 and 6 months(34
The second determinant of success of ACL reconstruction studied herein was graft-bone interface healing. Data are available from 5 studies, but only two found a beneficial effect of platelets. Vogrin et al saw a higher level of interface vascularization, which is essential for bone remodeling, at 3 months, but not 6 months(34
). This “early stage” effect is consistent with the mode of action posited above. Orrego et al found a borderline significant positive effect of platelets at six months postoperatively, while all other studies reporting on tunnel healing found no differences (30
). All studies but Vogrin et al assess tunnel healing at 6 months of follow-up, which is an interesting fact because earlier studies have suggested that graft integration is largely complete by 12 weeks(50
Finally, we were interested in assessing whether differences in graft and tunnel healing, or the lack thereof, would translate into differences in clinical outcomes. The three studies that evaluated clinical outcomes used different assessment tools but uniformly reported no differences between groups. Two issues should be considered when interpreting these data. First, the rationale for the use of platelets is to improve (the speed of) graft maturation, which is not necessarily visible in clinical scores focusing on pain, range of motion, and activities-of-daily-living. Second, considering that the most likely “clinical” effect of improved graft and tunnel healing is reduced rates of clinical graft failure, the current studies would be critically underpowered to assess differences in this rare outcome (even if they had included assessment of such outcomes). However, the current best evidence indicates no difference in clinical outcomes between ACL reconstruction with or without platelet concentrates.
After summarizing the findings from the included studies, we also want to focus on the second, no less important task of a good systematic review, identifying blank spots in the literature and describing ways to address them and to improve and expand the available evidence. One of the most obvious differences and the most commonly raised question in the included primary studies was the concentration of the used platelet concentrate. On the one hand, if a beneficial effect is to be assumed, one could also assume a dose-response relationship and opt for higher dosages to maximize the treatment effect. On the other hand, overdosing of platelet-released growth factors can be detrimental for two reasons, first, because of an over stimulation of cells leading to a poorly differentiated, chaotic scar, and second, because some of the released growth factors might precipitate adverse events, such as suppression of osteoclast generation (51
). While, to the best of our knowledge, there are no human studies to assess the dose-response relationship of platelet concentrates and ACL healing, data is available from large animal models, suggesting that 3× and 5× concentrates lead to equivalent biomechanical outcomes(52
). Thus reducing platelet concentrations would be a safe way to minimize the potential for detrimental effects and the volume of blood needed from the patient without jeopardizing mechanical outcomes.
This study has potential shortcomings. Like any systematic review the quality of this study depends on the quality of the primary data. We assessed the validity of these 8 studies based on levels-of-evidence and the Jadad score. Both methods are composite scores that focus on study design, such as randomization, concealed allocation, and attrition. The included studies scored in the middle to high range, which is representative for orthopedic research studies. However, neither score gives a comprehensive assessment of study validity, as they do not include parameters such as sample size estimation.
Beyond study design, another potential shortcoming is the clinical heterogeneity of the included studies. While age and gender distribution were uniform among all groups, as were postoperative rehabilitation algorithms, there were considerable differences in the used surgical techniques as well as the choice and placement of platelet concentrates. While some of these differences are only little, some might very well have major implications on healing, such as adding the platelet concentrate by injecting it on the graft under arthroscopy versus wrapping the graft in a biomaterial soaked in a platelet gel. As a matter of fact, these differences spurred a number of letters criticizing the lack of clarity when it comes to the description of the platelet concentrates uses (53
)–after some authors have used platelet concentrates but referred to them as, for example, PDGF(29
). Interestingly, the authors of the included studies wrote letters commenting on the methods used in other studies, also included here, showing the extent of disagreement among investigators as what platelet concentration to use, and how to assess its effects(54
). These differences stopped us from conducting a formal quantitative data synthesis, but it is still possible and valid to deduce evidence for a general, overall effect of platelet concentrates.