Article Text

Download PDFPDF

Editorial in response to the systematic review by de Vos et al: ‘Strong evidence against platelet-rich plasma injections for chronic lateral epicondylar tendinopathy: a systematic review’
  1. Taco Gosens1,
  2. Allan K Mishra2
  1. 1 Department of Orthopaedics and Traumatology, St Elisabeth Hospital Tilburg Hilvarenbeekseweg 60, Tilburg, The Netherlands
  2. 2 Department of Orthopedic Surgery, Menlo Medical Clinic, Stanford University Medical Center, Menlo Park, California, USA
  1. Correspondence to Taco Gosens, Department of Orthopaedics and Traumatology, St Elisabeth Hospital Tilburg Hilvarenbeekseweg 60, Tilburg 5022 GC, The Netherlands; t.gosens{at}

Statistics from

We would like to compliment the authors on their recent efforts to construct a review article concerning the use of platelet-rich plasma (PRP) in the treatment of lateral epicondylar tendinopathy.1 As the authors of two of the studies evaluated in the paper, we would like to comment about the interpretation of our investigations and provide suggestions about the conclusions that may be drawn from the published data.

First, we would like to discuss the terminology used to describe the effect of a steroid injection in the published studies by Peerbooms et al 2 and Gosens et al 3 In these investigations, the group treated with steroids returned to baseline pain and function levels after 1 and 2 years (see figure 2 in ref 3). A steroid injection may have the negative effect implied by de Vos et al,1 but in this study we were able to show that the condition of patients did not worsen. It is therefore incorrect to use these papers2 ,3 as evidence to suggest that steroid injections are detrimental.

Second, the title of the article “Strong evidence against platelet-rich plasma injections for chronic lateral epicondylar tendinopathy: a systematic review” is inappropriate and misleading. We acknowledge the value of the PEDro scale to grade the manuscripts and agree that most often papers are graded correctly. It is important, however, to note that the authors recognise in their paper that there was disagreement about how to apply the PEDro score to the included investigations. We invite them to review the information below and appropriately re-grade one of the papers. The PEDro scale score for the Mishra et al 4 study is incorrect. According to supplementary table 4, item 3 (allocation was concealed) was inappropriately scored as negative. For that study, the person determining whether a patient was eligible for inclusion was unaware when that decision was made to which group the participant would be allocated. Random allocation was performed via a computerised protocol and therefore concealed from the person determining the eligibility. Criterion 3 should therefore have been scored as positive, not negative. PEDro criterion 9 (intention to treat) is also scored as negative in supplementary table 4 but should have been scored as positive. All participants in the Mishra et al 4 study for whom outcome measures were available received the PRP treatment or control treatment as allocated making criterion 9 positive. Therefore, a proper PEDro evaluation of this study would result in a seven on the scale, changing it from a low-quality study to a high-quality study.

The de Vos et al 1 review also inappropriately states the loss to follow-up percentage in the Mishra et al 4 study. The reasons for the extension from 12 to 24 weeks are clearly stated in the published manuscript. One such reason is the need for a longer term follow-up to better evaluate safety and efficacy. The study included 94 patients enrolled under the 12-week protocol and 136 patients enrolled in the 24-week protocol. It was not possible to fully evaluate the 94 patients in the 12-week cohort because they had already passed the 24-week follow-up period when the study was modified. Therefore, the incomplete data at 24 weeks are not the result of a high dropout rate. Again, this is specifically stated in the published article.4 Importantly, there were 119 patients with 24-week data that were studied in a prospective, randomised fashion. This represents the largest cohort studied until now using PRP. In this group, clinically meaningful improvements in outcomes at 6 months for pain scores, elbow tenderness and overall success rates were documented. These results confirm the value of PRP for chronic lateral epicondylar tendinopathy in a high-quality study. These conclusions should be mentioned in the de Vos et al 1 review along with the elite documented safety profile of PRP in the treatment of chronic lateral epicondylar tendinopathy across all the studies.

Apart from miscalculating the PEDro scale and the follow-up inaccuracies, only two studies (2, 3 and 8) used the exact treatment protocol. Both these studies found value in the treatment of lateral epicondylar tendinopathy using PRP. Krogh et al,5 for example, injected ‘10–15 mL of lidocaine’ into the peritendon of the common tendon origin prior to intervention. This is a very different protocol from the one used by Gosens et al 3 or Mishra et al. 4 Creaney et al 6 also used a different PRP preparation method and employed a two-injection protocol. It is therefore difficult to combine the results of all these studies and make comprehensive generalisations. It would be better to break out the results by specific study design and PRP type. We advise that future publications concerning the use of PRP should employ a classification of the PRP and the injection protocol used such as the Sports Medicine Platelet-Rich Plasma Classification System that has been published previously.7

We also want to advise the readers to compare this review to the paper by Ahmad et al 8 that actually shows an opposing conclusion from de Vos et al,1 that is, as there is limited but evolving evidence for the use of PRP in lateral epicondylitis (see figure 2: Forest plots for standardised mean differences for (A) function and (B) pain in their manuscript). We wonder why their conclusion is so different from the conclusion by de Vos et al 1?

Collectively, the information presented in this letter contradicts the conclusions of the de Vos et al 1 review. We suggest that the title, body and conclusions of the de Vos et al 1 article be modified to address the information presented in this letter.



  • Competing interests TG (Europe) and AM (USA) have a consultancy relationship with Biomet.

  • Provenance and peer review Not commissioned; internally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles