The effect of platelet-rich plasma injection on lateral epicondylitis following failed conservative management

Abstract

Objective

We assessed the effect PRP injection on pain and function in patients with lateral epicondylitis where conservative management had failed.

Methods

We prospectively reviewed 34 patients. The mean follow-up was 26 weeks (range 6–114 weeks). We used the Oxford Elbow Score (OES) and progression to surgery to assess outcomes.

Results

88.2% improved their OES. 8.8% reported symptom progression. One patient had no change. No patients suffered adverse reactions. Two patients underwent an open release procedure. One had the injection repeated.

Conclusion

An injection of PRP improves pain and function in patients suffering from LE where conservative management has failed.

1. Introduction

“Tennis Elbow”, or lateral epicondylitis (LE) is a condition which affects both sexes equally and is considered common in the UK as between 1% and 3%⁹ of the population suffer from it. The condition mainly affects those of middle age; predominately those aged between 35 and 55 years.

The condition has been said to affect tennis players more commonly and prevalence between 35 and 51% has been quoted in large studies.¹

The term epicondylitis suggests that this is an acute inflammatory condition however it is more aptly described as a “tendinosis”. The most commonly affected structure is the Extensor Carpe Radialis Brevis (ECRB) which attaches to the lateral epicondyle of the humerus. The other tendons which join the ECRB on the lateral epicondyle are extensor carpe ulnaris, extensor digiti mini and extensor digitorum.

The chronic degenerative changes which occur in LE as a result of the repetitive strain injury implicated bear the hallmarks of a tendinosis. These include neovascularisation and disorganisation of collagen fibres.⁹ The resultant shift in cell population to include fibroblasts and vascular hyperplasia has been described as angiofibroblastic hyperplasia.¹⁰

There are numerous treatment modalities for LE of which the mainstay is non-operative. Bisset et al.³ concluded that conservative management has very similar or slightly inferior outcomes when compared to steroid injections. Part of conservative management is physical therapy which has been shown to be effective² and generally focuses on eccentric strength exercises and the maintenance of a good range of motion. Conservative management includes using non-steroidal anti-inflammatorys (NSAIDs) and corticosteroid injections however their long-term effectiveness has been questioned.¹⁹ Corticosteroid injections have been known to cause skin atrophy, skin depigmentation and fatty atrophy. Three injections within a single year are considered a maximum after which surgical procedures are normally considered.

Platelet-rich plasma contains supra-physiological quantities of growth factors which are implicated in tissue healing. These include: platelet-derived growth factor, transforming growth factor beta, epidermal growth factor¹¹ and vascular endothelial growth factor. Macrophages, mesenchymal stem cells and osteoblasts are attracted to the local tissue and enhance wound healing.¹⁷

2. Methods

We prospectively gathered data on 34 consecutive patients suffering from LE who received platelet-rich plasma injection therapy. The study was approved by the local audit department which enabled us to complete a pre-injection and mid-term follow up Oxford Elbow Scores on each of the patients during clinic appointments and over the phone.

All patients were seen in clinic by the senior author following failed conservative management. The conservative management included analgesia, physiotherapy and importantly all patients had at least one steroid injection. Subsequently these patients were listed for operative treatment of their condition. Prior to their operation all patients were offered platelet-rich plasma Injections as an alternative to improve their symptoms.

For platelet-rich plasma preparation the Arthrex (Arthrex, Inc. Florida, USA) ACP system was used. This is a closed system which uses a double syringe method along with 15 ml of autologous blood drawn from the patient's contralateral arm. The blood is then centrifuged at 1500 rpm for 5 min and 4–7 ml of platelet-rich plasma is extracted. An anticoagulant is not used. Using a 22-ga needle the senior author gained access to the common extensor tendon on the lateral epicondyle using a single skin portal. Local anaesthetic was not employed as it can disrupt the pre-prepared pH of the platelet-rich plasma and result in suboptimal efficacy of the injection. The platelets were injected into the epicondyle using the pepper-pot method. The patients were asked not to move their arm for 30 min. The affected limb was then placed in a poly-sling for 48 h after which the patients were advised to carry out eccentric elbow exercises. Patients were prohibited from using NSAIDS for three months.

We used the Oxford Elbow Score (OES) to evaluate outcomes following platelet-rich plasma injection.⁶ The OES is a 12 item score which comprises three unidimensional domains: elbow pain, elbow function and psychological–social. Each domain contains four items. The responses to each item range from 0 to 4 where a 4 represents the most significant severity. The scores for each domain can be calculated as the sum of the individual items which related to that particular domain. The maximum score obtainable using the OES is 48 which is a raw scale score. The individual domain scores are then converted to a metric sore out of 100 (100 representing the greatest severity).

3. Results

Our study group included 34 (18 women and 16 men) patients who all suffered from LE. The mean age at operation was 45 years (range 33–60 years). All patients were successfully followed up and the mean follow up time was 26 weeks (range 6–114 weeks).

All patients had reported compliance with the post-procedure protocol outlined. There were no reported complications following the injection of platelet-rich plasma. No patients suffered adverse reactions including either infections or neurovascular problems.

Thirty patients (88.2%) showed an improvement in their self reported OES. Three patients (8.8%) reported that their symptoms had progressed in severity following the injection. From this group only one had chosen and undergone surgery successfully at the time of last follow-up and one had been listed for the operation. Three patients who did show a very small improvement in symptoms following injection were listed for an open release procedure; one of them had at the time of writing successfully undergone the procedure. One patient had chosen to undergo a second injection. A single patient (2.9%) was found to have no change in symptoms.

All patients successfully completed pre and post injection OES. In order to determine the success of this procedure we used the figures set by Dawson et al.⁵ on minimal clinically important difference (MCID). These were roughly 10 (metric scores) for elbow function and 18 for pain and psychological–social. 70.6% of our patients reported a clinically important difference in Elbow function while 70.6% and 61.8% reported a clinically important difference in pain and psychological–social, respectively.

To compare pre and post-injection scores a Wilcoxon matched-pairs signed-ranks test was used. The P values generated are displayed in Table 1 along with the mean difference. We used a non-parametric test to accommodate for the distribution of results. The results indicate that a single injection of platelet-rich plasma significantly improves each of the domains of the OES.

Table 1.

Mean comparison and Wilcoxon results.

	Pre-op	Post-op	Difference	P value
	Mean (sd)	Mean (sd)	Mean
Elbow function	51.1 (2.8)	29.3 (3.9)	−22.8	P < 0.0000
Pain	70.2 (3.0)	36.6 (4.5)	−33.3	P < 0.0000
Psychological–social	74.6 (2.8)	43.9 (5.6)	−30.7	P < 0.0000

As there was a varied distribution of follow-up time in our data set we used linear models to investigate if there was a relationship between post-op scores and follow-up time. The results are displayed in Table 2. Week number was scaled such that a unit represents 4 weeks for easier interpretation. Each model also controlled for the pre-op outcome score. Confidence intervals were produced using a non-parametric bootstrap with 2000 repetitions. These results suggest that both pain and psychological–social scores improve over time. The pain score reduces by 1.96 for each month increase in follow up time and the psychological–social score by 2.15. These relationships have been demonstrated in Fig. 1, Fig. 2.

Table 2.

The effect of follow up time on outcome.

Outcome		Coefficient	Bootstrapped 95% CI	P value
Elbow function	Follow up time	−1.03	(−2.40, 0.10)	0.109
Pain	Follow up time	−1.96	(−3.15, −1.00)	<0.000
Psychological–social	Follow up time	−2.15	(−3.35, −0.84)	0.001

Fig. 1.

Effect of follow up time on pain score.

Fig. 2.

Effect of follow up time on psychological–social score.

To find out if failure to meet MCID predicted the need for further intervention we created indicator variables for each domain in the OES. We used exact regression models to see if failure to meet the MCID resulted in the need for further intervention. Due to the sample size exact logistic regression was used rather than logistic regression. The results shown in Table 3 suggest that failure to meet the MCID in elbow function and psychological–social domains was associated with the need for further intervention.

Table 3.

Does failure to meet MCID result in further intervention?

Outcome	Odds ratio	Std. error	95% CI	P value
Elbow function	20.2	23.7	(1.8, 1132.7)	0.0093
Pain	2.9	2.6	(0.3, 27.1)	0.4567
Psychological–social	11.5	13.3	(1.1, 619.8)	0.0427

Though these results seem striking, they are due to the fact that this is a small sample size. A cross tabulation of elbow function and surgery shows the very small numbers in the off-diagonal cells. Therefore although these are accurate results they should be interpreted as results for this sample only.

4. Discussion

Our study shows that a single injection of platelet-rich plasma can improve the pain, function and psychological–social outcomes of patients who suffer from LE where conservative management has failed. Our group of patients suffered from symptoms which were refractory to conservative management; which included corticosteroid injections in all cases. The entire cohort would have been offered an open release procedure to treat their severe symptoms of LE prior to being given the option of platelet-rich plasma injection as per previous practice. We believe that avoiding an operative procedure in 85% of our patients is a successful outcome as none of our patients suffered adverse reactions to the injection. Only five patients from our group were listed for an open procedure following a single PRP injection with only two of those having had it at time of last follow up. We saw a significant improvement in elbow pain and psychological–social scores in our group as the length of follow up increased however the association between achieving a MCID and avoiding an operation was not established.

Mishra et al.¹¹ conducted a pilot study looking at the effect of platelet-rich plasma injection in patients suffering from chronic LE who had failed conservative management. They concluded that in a group of 15 patients treated with PRP injection there were significant improvements in pain (visual-analogue score) and function (MAYO elbow score). More recently Mishra et al. have produced a double-blind, prospective, multicentre randomised control trail of 230 patients¹² which demonstrated clinically meaningful improvements in pain for patients treated with PRP injection for chronic LE. Although their evidence is strengthened by a relatively large cohort the more recent study does not assess improvement in elbow function.

Other studies with longer follow-up times by Gosens et al.¹³ and Peerbooms et al.⁷ have shown similarly positive results when PRP injection has been compared to corticosteroid injections. These larger studies had the advantage of assessing pain (using VAS scores) as well as function (DASH score) however lacked a control group. A study published in 2011 containing 31 elbows⁸ also showed positive results at 1 year follow-up using various outcome measures including a modified American Shoulder and Elbow Surgeons (ASES) score for both self-reported and physician reported outcomes. We consider their outcome measures may be overcomplicated however show a similarly positive result to our study in avoiding operative intervention. Only two patients elected for surgery at 1 month post-injection; perhaps before the benefits of the injection were felt.

Studies conducted in 2011 by Thanasas¹⁸ and Creaney⁴ compared autologous whole blood (AWB) injection with PRP injections in a cohort of difficult to treat LE sufferers. Their studies differed in size greatly (28 patients compared to 150 patients, respectively) and were unable to conclude significant improvements in longer-term outcome for PRP injection compared to AWB injection. The study be Creaney et al. may have been unsuccessful in producing a difference in long-term outcome due to its relatively low concentration of PRP (2.8 times whole blood) however recent studies by Raeissadat et al.15, 16 using PRP concentrations above 4 times that of normal blood failed to demonstrate significant differences in outcomes also.

We acknowledge the distribution of follow-up times in our study to be varied. However we believe that the generation of statistically significant values in two out of three domains within the OES when comparing length of follow-up time demonstrates a delayed benefit from PRP injection supported in other studies.7, 13 Other limitations of our study include the lack of comparison group and thus lack of randomisation.

Our study although relatively small has produced statistically significant results which are comparable to existing data. We believe the selection of the OES to have been appropriate due to its relevance to LE when compared to other scores such as the Mayo or disabilities of the arm shoulder and hand (DASH) scores.

It is clear that outcome measures for studies on the effect of PRP in LE are varied which makes it difficult to compare studies and a systematic review¹⁴ comparing various injections for LE relied mainly on VAS pain scores to compare studies. This is ineffective as VAS pain scores do not accommodate for age or health status. We believe that using progression to surgery as an outcome measure is useful as it provides a simple bench mark for comparison in patients who have received the normally effective conservative management techniques and have had no response to corticosteroid injections as these patients are likely to be considered for surgery. The relative economic implications of this are thought to be significant although no cost-analysis has been conducted. Relevant and validated radiological, biomechanical and healing response biomarkers are needed to make comparison easier and establish a definitive superiority among available treatments.

5. Conclusions

Our results show that a single injection of PRP improves pain and function in patients suffering from LE where conservative management has failed and would have otherwise required surgery. Improvements in pain and psychological–social domains increased as the length of follow-up increased suggesting PRP injection has a delayed beneficial effect.

Conflicts of interest

The authors have none to declare.