Myths and Facts of In-Office Regenerative Procedures for Tendinopathy: Literature Review

Abstract

Tendinopathy carries a large burden of musculoskeletal disorders seen in both athletes and aging population. Treatment is often challenging, and progression to chronic tendinopathy is common. Physical therapy, NSAIDs, and corticosteroid injections have been the mainstay of treatment but are not optimal given that most tendon disorders appear to involve degenerative changes in addition to inflammation. The field of regenerative medicine has taken the forefront, and various treatments have been developed and explored including prolotherapy, platelet rich plasma (PRP), stem cells, and percutaneous ultrasonic tenotomy. However, high-quality research with standardized protocols and consistent controls for proper evaluation of treatment efficacy is currently needed. This will make it possible to provide recommendations on appropriate treatment options for tendinopathy.

Introduction

Tendinopathies are common conditions seen in both athletic and aging population. Sustained pain and functional impairment due to tendon impairment significantly impede sports related activities and ability to work. Repetitive microtrauma leads to failure of tendon cells to regenerate normal tendon tissue and angiofibroblastic hyperplasia ensues¹. Histological findings demonstrate necrotic tenocytes, collagen disarray, and neovascularization in injured tendons. Without a rich blood supply, tendon tissue has an intrinsic low healing potential, which makes tendinopathies extremely difficult to treat, especially once they reach the chronic stage and fibrotic scarring has occurred. Current standard treatments in clinical practice include various rehabilitation programs, non-steroidal anti-inflammatory drugs (NSAIDs), extracorporeal shockwave therapy (ESWT), corticosteroids, and operative options. Diagnosis of tendinopathy is traditionally clinical. However, diagnostic ultrasound has gained popularity as it allows point of care tissue characterization of tendon, muscle, and nerve with ease of access². Ultrasound may reveal tendon thickening, loss of fibrillar continuity, and neovascularization via Doppler mode indicating tendon injury. Clinical regenerative therapies for tendinopathy can be divided into three large categories, 1) chemical procedures, 2) orthobiologic procedures, and 3) mechanical procedures^3,4. In contrast to conventional steroid injections which are used to modulate pain, these emerging options are being explored in order to eliminate/minimize degenerative tissues and to encourage regeneration⁵. The most common chemical procedure is called prolotherapy where an irritant such as dextrose or other chemicals are used to initiate an inflammatory cascade, which is considered to enable a healing process⁶. Orthobiologic procedures refer to the use of biological agents for improving orthopedically related disease processes. Most commonly used orthobiologic agents include autologous agents such as platelet-rich plasma (PRP), various types of cells (mainly stem cells from bone marrow and adipose), or allogenic agents derived from placental tissue⁷. Mechanical procedures use physical force to primarily remove degenerated tendon tissues and thereby help improve tendon structures. Such procedures include traditional percutaneous needle tenotomy (PNT) and more recently, percutaneous ultrasonic tenotomy (PUT). Research in this field has been progressing rapidly, though there is still much controversy surrounding the efficacy of these therapies, as few large randomized controlled trials (RCTs) exist. In this article, we will appraise available evidences and suggest the areas of necessary researches that relate to regenerative tendon procedures.

Methods

A literature review was performed on the use of regenerative medicine therapies as treatment for chronic tendinopathies focusing on the following procedures: prolotherapy, PRP, cell treatments (stem cells and skin-derived tenocyte like cells), and percutaneous ultrasonic tenotomy. The following terms were included: patellar tendon, Achilles tendon, common extensor tendon, rotator cuff tendon, gluteal tendon, and hamstring tendon. The search was conducted on the PubMed database in October 2017 using the following keywords:

1. For prolotherapy: (prolotherapy OR hyperosmolar dextrose OR hyperosmolar glucose)

2. For PRP: (PRP OR platelet rich plasma OR platelet derived growth factors)

3. For stem cells: (stem cells OR mesenchymal stem cells OR bone marrow stem cells OR adipose stem cells)

4. For skin-derived tenocyte-like cells: (skin-derived tenocyte-like cells OR SD-TLC OR skin derived cells)

5. For percutaneous ultrasonic tenotomy: (percutaneous ultrasonic tenotomy OR PUT or ultrasonic tenotomy)

6. For tendinopathy: (patellar tendon OR patellar tendinopathy OR jumper’s knee) OR (Achilles tendon OR Achilles tendinopathy) OR (common extensor tendon OR common extensor tendinopathy OR lateral elbow OR tennis elbow OR lateral epicondylitis) OR (rotator cuff tendon OR rotator cuff tendinopathy) OR (gluteal tendon OR gluteal tendinopathy OR gluteus medius tendon OR gluteus medius tendinopathy) OR (hamstring tendon OR hamstring tendinopathy).

Titles and abstracts were screened, and the following inclusion criteria were used: human studies, randomized controlled trials, case series, and use of above therapies in chronic tendinopathy. Exclusion criteria were articles analyzing other applications of the above regenerative therapies, other preparations of cell therapies (stromal vascular fraction, bone marrow aspirate concentrate), case reports, or animal studies. Reference lists from the selected articles were also screened. The full texts were read and relevant data was extracted for use in this review. This is not an all-inclusive systemic review of the previous literature.

Result/Discussion

Prolotherapy: Mechanism of Action

Prolotherapy has been in practice for over 80 years dating back to the 1937 when it was used on the thumb ulnar collateral ligament of a surgeon to treat both pain and ligamentous laxity⁸. The most common injected solution in prolotherapy is hyperosmolar dextrose. This solution is supposed to work by creating a hypertonic environment, causing cell rupture and upregulation of platelet-derived growth factors (PDGF). Sodium morrhuate is another agent used for its properties that are theorized to attract inflammatory mediators (e.g. CD43⁺ leukocytes, ED1⁺ and ED2⁺ macrophages)⁹ and act as a vascular sclerosant to reduce neovascularization. Cellular irritants such as phenol, glycerin, and glucose are no longer used in practice¹⁰.

Prolotherapy for Patellar Tendinopathy

Prolotherapy studies in patellar tendinopathy and Osgood-Schlatter disease (OSD) generally report good outcomes with reduction in pain. In the study by Ryan et al., 47 subjects with patellar tendinopathy refractory to conservative treatment received ultrasound-guided injections of 25% dextrose with lidocaine¹¹. Subjects received injections at six-week intervals based on symptom improvement with a mean of 4 ± 3.4 injections required. At 45-week follow up, there was a significant reduction in pain at rest and with activity, which were accompanied with sonographic improvement in both vascularity and echogenicity. Another RCT by Topol et al. demonstrated the efficacy of unguided prolotherapy in OSD, which is a traction apophysitis of patellar tendons (a patellar tendinopathy spectrum disease)¹². This study randomized 65 knees in 54 subjects aged 9–17 years with recalcitrant OSD to three treatment groups: 12.5% dextrose injection, lidocaine injection, or a therapeutic exercise group. The injection treatment groups received three injections at four-week intervals. Greater rates of returning asymptomatic patients to sports were observed in the prolotherapy group (84%) versus the lidocaine (46%) and exercise (14%) groups at one year.

Prolotherapy for Achilles Tendinopathy

An early case series showed reduced pain at six weeks after ultrasound-guided, intra-tendinous 25% dextrose injections in 33 cases of both insertional and mid-portion Achilles tendinosis¹³. The subjects required an average of 4 injections at 1.5 months interval, and although statistical significance was defined as p=0.10, the improvement was seen with pain at rest and pain with tendon-loading activities, with improvement on sonographic hyperemia in 55% of the treated tendons. Yelland et al. randomized 43 subjects with mid-portion AT into a 12-week therapy program of eccentric loading, weekly unguided 20% glucose prolotherapy injection with local anesthetics, or combined 20% prolotherapy/eccentric load exercise groups¹⁴. Prolotherapy alone or in combination with eccentric exercise showed greater reduction in pain at 6 months, although longer-term results at 12 months showed no significant difference. A potential attention bias should be noted as injection protocol included performance of injection until minimum clinically important change (MCIC), resulting in 3 times more visits in injection groups. Interestingly, 4 injections (range: 2 to 11) were required to achieve the MCIC, which was similar to Maxwell’s case series.

Prolotherapy for Common Extensor Tendinopathy

Common extensor tendinopathy (CET) is one of the most heavily studied anatomic structures for prolotherapy. Scarpone et al. performed a double blind RCT involving a total of 24 subjects with CET in whom three landmark guided injections at four-week intervals of 10.7% dextrose and 14.7% sodium morrhuate by volume versus saline control was used without tenotomy. Tenotomy refers to insertion of the needle into the tendon with or without injection of the solution, resulting in fiber disruption and induction of bleeding to promote healing. The intervention resulted in significant reduction in elbow pain and improvement in strength in the prolotherapy group compared to the control¹⁵. Carayannopoulos et al. randomized 24 subjects into two groups, administering various landmark guided prolotherapy agents including phenol, glycerin, 12.5% dextrose, and sodium morrhuate, and compared the response to corticosteroid injection⁶. Subjects were given two injections at four-week intervals. Improvements in pain and function, as evidenced by visual analogue scale (VAS) for pain and disability/symptom (DASH) scores, were noted in both groups with no significant inter-group differences or differences in subject satisfaction although an attenuation effect in VAS scores was seen in the corticosteroid group beyond six months as noted by a decreased change in VAS score compared to the three month follow up. This may suggest that studies involving benefits of prolotherapy would require a longer duration of follow up⁶. The study of 84 subjects by Shin et al. demonstrated decreased VAS pain scores and improved sonographic appearances of tendon after 3 land-mark guided injections of 15% dextrose¹⁶. Figure 1 demonstrates a sonographically guided CET prolotherapy injection.

Figure 1:

Sonographically guided CET prolotherapy injection with CET (arrow) and radial collateral ligament (asterisk) seen in anatomic long axis view with the needle (open arrow) guided in an in-plane, distal to proximal technique. The radius is distal, to the right and the humerus is proximal, to the left.

Prolotherapy for Rotator Cuff Tendinopathy

A few RCTs exist that have evaluated prolotherapy in chronic rotator cuff tendinopathy, and while they do not have consistent control groups, they all did find a significant reduction in pain and disability with prolotherapy. A recent RCT by Seven et al. treated 101 subjects with chronic rotator cuff lesions to either the ultrasound-guided 25% (sub-acromial bursa) and 15% (tendon insertion) dextrose prolotherapy injection group or to a physical therapy (PT) group¹⁷. The prolotherapy group required a mean of 5.23 injections (range 2–6). This study demonstrated significant improvement in pain score, disability index, and shoulder range of motion (except external rotation) at 6-week, 12-week, and 1-year follow up in the prolotherapy group compared to the PT group, though it should be noted that there was no difference at 3-weeks. There was 93% patient satisfaction in the prolotherapy group versus 57% in the control group at 1-year. In a case series of 126 subjects, three to eight prolotherapy injections with 16.5% dextrose by volume solution without imaging guidance were given to patients unresponsive to aggressive conservative treatment¹⁸. At one year follow up, strength, range of motion, and pain were significantly improved in the treatment group over the control group. See Figures 2A and 2B demonstrating sonographically guided prolotherapy injection for rotator cuff tendinopathy.

Figure 2:

Sonographically guided prolotherapy injection for rotator cuff tendinopathy with long axis view of subacromial bursa (asterisk) and supraspinatus tendon (closed arrow). Acromion is located medially, to the left. A. Needle (open arrow) is directed in plane, lateral to medial, into the sub-deltoid sub-acromial bursa. B. Needle is directed in plane, lateral to medial, into the tendinopathic supraspinatus tendon (closed arrow).

Current View of Prolotherapy Injection for Tendinopathies

Table 1 summarizes literature discussed in this section of the review. CET studies appear to show most robust data for success while prolotherapy appears to be promising for patellar, Achilles, and rotator cuff tendinopathies. Finally, studies suggest multiple injections are most likely required and appear to be safe. Little is known about the benefits of “prehab” and post-injection therapies. There have not been any clinical studies for the use of prolotherapy in gluteal or hamstring tendinopathies. Although high-level evidences are still few, with the low risk of adverse events given its long history in existence and efficacy, dextrose prolotherapy can be considered as an alternative to conventional treatments such as PT or corticosteroid injections in chronic refractory tendinopathies.

Table 1:

Summary of clinical trials discussed for Prolotherapy in tendinopathy

Reference	Level of evidence	Pathology	N patients	Avg Age	Therapeutic Protocol	US Guidance	Prolotherapy Solution	Outcome measure	Follow up	Main findings
Ryan et al11.	Level IV case series	Patellar tendinopathy	45 subjects (47 knees)	38.3 ± 14.3 years	Dextrose injections (mean of 4 ± 3.4)*	Yes	25% dextrose	VAS, US examination	Mean 45 weeks	Significant reduction in pain at rest, with ADLs, and during sport. Improved neovascularity and hypoechoic lesion appearance on US.
Topol et al12.	Level I RCT	Osgood-Schlatter Disease	54 subjects (65 knees) n= 17 dextrose n= 18 lidocaine n=19 control	n/a (range 9–17 years)	Dextrose injection monthly × 3 Lidocaine injection monthly × 3 Therapeutic exercise	No	12.5% dextrose	NPPS	1 year	Greater rates of asymptomatic return to sport in prolotherapy group.
Yelland et al14.	Level I RCT	Achilles tendinopathy	43 subjects n=14 glucose n=14 combo n=15 exercise	46 (exercises and combination) vs. 48 (glucose) years	Hypertonic glucose (average 9.5 injections) Hypertonic glucose + eccentric exercise (average 8.7 injections) 12 week eccentric exercise (mean 3.3 formal sessions)	No	20% hypertonic glucose	VISA-A, Likert scale, PGIC, VAS	12 months	More rapid pain improvement in prolotherapy and combo therapy groups over exercises alone. No difference in long-term functional scores.
Maxwell et al13.	Level IV case series	Achilles tendinopathy	36 subjects	52.6 years	Dextrose injections (repeat q6w until symptoms resolved)*	Yes	25% dextrose	VAS at rest, during ADLs, and during activity; tendon appearance on US	12 months	Reduced pain and neovascularity on US at 6 weeks after treatment. No improvement in hypoechoic lesions on US.
Scarpone et al15.	Level I RCT	Lateral epicondylosis	24 subjects n=12 prolotherapy n=12 control	48 years	Prolotherapy injections monthly × 3 (at supracondylar ridge, lateral epicondyle, and annular ligament) Saline injection monthly × 3 (same location)	No	10.7% dextrose, 14.7% sodium morrhuate, and local anesthetic	Likert scale, extension and grip strength	1 year	Significant improvement in pain and isometric strength in prolotherapy group that was maintained at 12 months.
Carayannopoulos et al6.	Level I RCT	Lateral epicondylosis	24 subjects n=11 prolotherapy n=13 corticosteroid	49 (prolotherapy) vs. 46 (steroid) years	2 prolotherapy injections at 4 week interval (annular ligament, lateral epicondyle, and radial collateral ligament) 2 Corticosteroid injections at 4 week interval (same locations)	No	12.5% dextrose, phenol glycerine, and sodium morrhuate	VAS, QVAS, DASH	6 months	Improved pain and function in both groups but no significant intergroup differences.
Shin et al16.		Lateral epicondylosis	84 subjects	n/a	3 prolotherapy injections	No	15% dextrose	VAS, US tendon appearance	6 months	Significantly improved pain, and decreased hypervascularity and improved tendon structure on US in prolotherapy
Seven et al17.	Level I RCT	Rotator cuff tendinopathy	101 subjects n= 57 prolotherapy n= 44 control	57 (prolotherapy) vs. 44 (control) years	Prolotherapy to sub-acromial bursa and tendon insertions (mean 5.23 injections)^ PT 3x/week × 12 weeks	Yes	25% dextrose (sub-acromial bursa) 15% dextrose (tendon insertions)	VAS, SPADI, WORC, patient satisfaction, shoulder ROM	Minimum 1 year	Significant improvements in pain, disability index, and shoulder ROM (except ER) with prolotherapy.
Lee et al18.	Level IV case series	Rotator cuff tendinopathy	126 subjects n=63 prolotherapy n=63 control	54.1 ± 7.8 (prolotherapy) vs. 55.8 ± 6.6 (control) years	3–8 prolotherapy injections (average 4.8 ± 1.3 injections) Continued conservative treatment	No	16.5% dextrose	VAS, SPADI, AROM, maximal tear size on US, analgesic use	1 year	Prolotherapy demonstrated significantly improved strength, ROM, and pain over control.

^*Intra-tendinous;

^#Peritendinous;

^{^}Intra-and Peritendinous

RCT: randomized controlled trial; US: Ultrasound; VAS: Visual Analogue Scale; NPPS: Nirschl Pain Phase Scale; VISA-A: Victorian Institute of Sport Assessment-Achilles; PGIC: Patient Global Impression of Change Scale; QVAS: Quadruple Visual Analogue Scale; DASH: Disabilities of the Arm, Shoulder, and Hand; SPADI: Shoulder Pain and Disability Index; WORC: Western Ontario Rotator Cuff Index; ROM: Range of Motion; AROM: Active Range of Motion; PT: Physical Therapy

Improving Prolotherapy Clinical Studies

The mechanism of action by prolotherapy has not been established and this will continue to be the area of active discussion, and will likely require further preclinical investigations. Optimal procedural protocols have not been determined including frequency, concentration of dextrose, and injection location (peri-tendon vs. intra-tendinous). This may depend on the targeted tendon location (lower vs. upper extremity tendon) and on the severity and chronicity of tendon injuries. Ultrasound-guidance will maximize injection accuracy and study reproducibility and allows assessment of varying degree of tendinopathy.

Platelet-Rich Plasma (PRP): Mechanism of Action

Since the first documented musculoskeletal application of PRP in the 1990s, PRP has been studied extensively for various conditions including tendinopathy. PRP is obtained by centrifugation of autologous blood and collecting concentrated platelets in plasma. PRP is considered to exert regenerative effect via growth factors (GFs). Platelets produce a variety of GFs such as PDGF, VEGF, bFGF, TGF, HGF, and IGF that are involved in stimulating chemotaxis, extracellular matrix synthesis, and cell migration and proliferation¹⁹. Along with GFs, chemokines (IL-1β, PF4), adhesive proteins (plasminogen, fibrinogen, vitamin D-binding protein), proteases (MMPs, ADAMTs), and smaller molecules (serotonin, histamine) are released from platelets providing a rich environment with tendon healing capability²⁰. The clotting cascade leads to a fibrin clot at the site of injury allowing for cessation of bleeding and a scaffold for cell migration and proliferation. The regenerative capacity of PRP has allowed it to be used as a conservative treatment option as well as in surgical augmentation in tendon injuries. While the safety of such injections is well established, recent debate has focused on optimization of PRP formula for improved results. Depending on the preparation protocols used, PRP varies in its contents and several variables such as numbers of white blood cells, their differentials, and presence or absence of platelet-activating agents are believed to affect overall treatment outcome. Results of PRP for tendinopathy treatment have been largely variable, as discussed below in regard to patellar, Achilles, rotator cuff, and common extensor tendons; however, these are the four clinical tendinopathies that have the most robust evidences.

PRP for Patellar Tendinopathy

All patellar tendon RCTs treated subjects who are in their 20s. Two RCTs have investigated the effect of one versus two injections of PRP at one- to two-week intervals and found varying results^21,22. Kaux et al. randomized 20 surgical candidates to one versus two US guided PRP injections. While both groups showed improvements including approximately 40–50% pain improvement at 12 months, the study did not document any superior benefit of two PRP injections over one. Zayni et al. randomized 40 subjects to one vs two injections, and demonstrated that two ultrasound guided PRP injections resulted in improved pain and function as evidenced by better VAS, Tegner, and Victorian Institute of Sport Assessment-Patellar (VISA-P) scores. This difference may be partly explained by the variability in PRP preparations in the above studies in which Kaux et al. formulated a solution with a higher concentration of platelets than Zayni et al. (4–5 times vs. 2 times of whole blood) and started the rehabilitation stage sooner (5–7 days post first injection vs 2 weeks post last injection). The other two RCTs investigated PRP versus ESWT and dry needling^23,24. Vetrano et al. compared two weekly US guided PRP injections (n=23) to three weekly ESWT sessions (n=23) and found that 2xPRP was superior to 3xESWT in terms of functional recovery at 6- and 12-months²³. In Zayni and Vetrano’s investigations, approximately 20% of subjects ended up receiving surgical interventions due to non-responsiveness to PRP injections. Dragoo et al. randomized 23 subjects to a single leukocyte-rich US guided PRP injection with dry needling or dry needling alone groups, and concluded that PRP plus dry needling provided greater reduction in pain and improvement in function scores at 12-week follow up although such difference was not seen at 26 weeks²⁴. Dragoo concluded that PRP may be a viable option in managing patellar tendinopathies and may provide faster recovery.

There have been a total of 317 indexed patellar tendinopathy cases treated using various PRP formulations in case series. Most common formula reported consisted of a platelet concentration of 2–3 times of whole blood without reference to white cell counts, treating chronic patellar tendinosis of average duration of 22.5 months, with average follow up of 6–24 months, resulting in an average of 66% reduction in pain. Average age of indexed case series patients was 34.6 years old, and most common activity level represented was recreational athletics. There was a predominance of male subjects. Also, 63% of studies performed cell counts or other forms of PRP characterization, and ultrasound guidance was implemented in 45% of these studies^25–34.

A closer look at some of these case series reveal that pain relief can last up to 2 years following injections even in individuals who are at professional / semi-professional levels of athletic competitions²⁵. An investigation by Ferrero demonstrated that US guided PRP injection can result in improved tendon echo-texture and decreased hypervascularity²⁸. Filardo’s investigation showed both increased chronicity and bilateral nature of the presentation carried negative prognostication³⁰.

PRP for Achilles Tendinopathy

There have been four published RCTs on AT. Three RCTs (n=98 patients at follow up duration of 6–12 months) reported no benefit of a single PRP injection when compared to saline injections^35,36 or physical therapy³⁷. All three studies treated chronic mid-portion Achilles tendinosis without a tear. It should be noted that when comparing 3 negative RCTs to over 200 indexed cases in case series, subjects in case series tended to be slightly younger (average age= 49 years for RCTs vs. 45 years for case series subjects) and 37% of the case series required 2–3 PRP injections for improvement. Age is known to affect activity of tendon specific stem/progenitor cells and platelet function, which may explain the discrepancy^38,39. Further, Salini et al. divided 44 subjects with recalcitrant non-insertional AT based on age (29 subjects with mean age 39.5 years and 15 subjects with mean age 61.5 years) and showed that US guided PRP treatment was less effective in the elderly population⁴⁰. The most recent RCT by Boesen et al. (n=60, average age=42 years) has been the only positive RCT investigation where 4 US guided PRP injections at two-week intervals were found to be superior to control saline injection for mid-portion Achilles tendinosis at 6, 12, and 24 weeks in terms of pain control and functional outcome⁴¹. In addition, the PRP groups demonstrated a significant reduction in tendon thickness and vascularity on ultrasound examination, most evident at 6 weeks. The inconsistency of these results compared to earlier RCTs may be related to the higher PRP doses or injection numbers and due to younger age of included subjects (42 in this RTC vs. 49 in previous 3 RTCs for chronic mid-portion AT studies). The inclusion of an eccentric training regimen likely potentiated PRP benefits also⁴². See Figure 3 demonstrating short axis view of a sonographically guided PRP injection for Achilles tendinopathy.

Figure 3:

Sonographically guided PRP injection for Achilles tendinopathy with transducer in anatomic short axis view utilizing an in plane, lateral to medial, injection technique. Achilles tendon (asterisk), needle (open arrow).

PRP for Common Extensor Tendinopathy

Common extensor tendinopathy (CET) is the most extensively studied tendinopathy with PRP. Most studies have used a single injection protocol and compared it to autologous whole blood, local anesthetic, saline injection, laser therapy, or corticosteroid injection. Compared to corticosteroids, PRP has shown longer lasting benefit in pain and function up to two years after injection^43,44. There are 2 RCTs that did not show improved clinical outcome over corticosteroid. This may however be due to a shorter follow up period (3–6 months for negative studies versus 6 to 24 months for positive studies)^45,46. In four other RCTs, there was no statistically significant advantage of PRP compared to whole blood beyond the 8-week follow up⁴⁷. Based on additional data, PNT appears to be complementary and should be considered as an adjunct when performing PRP injections⁴⁸.

PRP for Rotator Cuff Tendinopathy

In contrast to the other types of tendinopathies, rotator cuff tendinopathy research has been more geared toward surgical augmentation rather than injection mono-therapy. In surgical context, there are a total of 16 RCTs combining for a total of 929 cases with no clear benefit of PRP in surgical augmentation during or after the procedure^49–51.

For PRP monotherapy, the injection protocols have been variable among studies. Rha et al. used 40–50 PNT using 25-gauge needle with two US guided PRP injections compared to PNT alone for supraspinatus partial tear or tendinosis and discovered a greater decrease in pain and disability in the PNT+PRP group⁵². Alternatively, Kesikburun et al. injected PRP versus saline with US guidance into the subacromial space but not into the tendon, and did not observe any inter-group differences⁵³.

PRP for Gluteal Tendinopathy

There have not been any RCTs for the use of PRP in gluteal tendinopathy but clinical case series advocate for safety and possible benefits. A case series of 21 subjects with refractory gluteus medius tendinosis or partial tear assessed US guided intra-tendinous leukocyte-rich PRP injection with needle tenotomy and found that at a mean follow-up of 19.7 months (range 12.1–32.3 months) there was a statistically significant improvement in all functional scores⁵⁴. Conversely, Jacobson et al. included 30 subjects (24 female) whom either received needle tenotomy alone or with US guided leukocyte-rich PRP injection and found improvement in pain for both groups⁵⁵. The early advantage of tenotomy alone may be explained by the early inflammatory effect of leukocytes in the PRP group as there was no differences between groups at 3 months follow up.

PRP for Proximal Hamstring Tendinopathy

Davenport et al. performed a double-blind RCT comparing US guided leukocyte-rich PRP to autologous whole blood injections with tenotomy for the treatment 15 cases of proximal hamstring tendinopathy⁵⁶. There were greater improvements in pain and function in the whole blood group at 12 weeks but the PRP group showed greater improvement at 6 months. Case series have indexed a total of 28 subjects with hamstring tendinopathy treated with PRP with improvement in pain at 2–6 months time frame^57,58.

Current View of PRP Injection for Tendinopathy

Based on literature (See Table 2), the strongest evidence for the beneficial effect of PRP remains as treatment for both common extensor tendinopathy and patellar tendinopathy, while there is limited evidence to show the benefit for Achilles, rotator cuff, gluteal, and proximal hamstring tendinopathies. Younger and active individuals seem to benefit from this procedure though this needs to be further studied. In case of chronic rotator cuff tendinopathy, it appears prolotherapy might be an appropriate consideration given current minimum evidence to recommend costly PRP over prolotherapy. PRP as a surgical augmentation has shown marginal benefits. Likewise, PRP in gluteal and hamstring tendinopathies are common occurrences in clinics, without robust RCTs.

Table 2:

Summary of clinical trials discussed for PRP in tendinopathy treatment

Reference	Level of evidence	Pathology	N patients	Avg Age	Therapeutic Protocol	US Guidance	Platelet count and leukocytes	Outcome measure	Follow up	Main findings
Kaux et al21.	Level I RCT	Patellar tendinopathy	20 subjects: n=10 one PRP n=10 two PRP	n/a	1 vs. 2 PRP injections at 1 week interval^	Yes	Platelet count: 8.5–9 × 103 per mm3 Leukocytes: no	VAS, IKDC, VISA-P, isokinetic strength	12 months	No benefit of 2 injections over 1 in pain relief, function, or strength.
Zayni et al22.	Level I RCT	Patellar tendinopathy	40 subjects: n=20 one PRP n=20 two PRP	24.6 (one injection) vs. 24.1 (two injections) years	1 vs. 2 PRP injections at 2 week intervals*	Yes	Platelet count: 2x basal value Leukocytes: no	VISA-P, VAS, Tegner	2 year (minimum)	Significantly better outcome in terms of pain relief and function with 2 injections vs. 1 injection.
Vetrano et al23.	Level I RCT	Patellar tendinopathy	46 subjects: n=23 PRP n=23 ESWT	26.85 years	2 injections of PRP at one week interval^ 3 focused extracorporeal shock wave therapies	Yes	Platelet count: 0.89 −1.1 × 106 per mm3 Leukocytes: n/a	VAS, VISA-P, modified Blazina scale	12 months	PRP was superior to ESWT in functional recovery and pain reduction at 6 and 12 months.
Dragoo et al24.	Level I RCT	Patellar tendinopathy	23 subjects: n=10 PRP n=12 guided dry needling	35 ± 13 years	1 PRP injection + PNT* 1 PNT*	Yes	Platelet count: n/a Leukocytes: yes	VISA-P, Lysholm	6 months	PRP resulted in faster recovery at 12 weeks but no difference at 26 weeks.
Charousset et al25.	Level IV Case Series	Patellar tendinopathy	28 subjects	27 years	3 PRP injections^	Yes	Platelet count: 2x basal value Leukocytes: no	VISA-P, VAS, Lysholm	24 months	Improved pain, function, and tendon healing on MRI with 75% return to sport at pre-level injury within 3 months.
Ferrero et al28.	Level IV Case Series	Patellar tendinopathy	28 subjects	37.4 years	3 PRP injections at a mean of 3 week intervals*	Yes	Platelet count: n/a Leukocytes: n/a	VISA-P, US exam	6 months	Statistically improved pain and US tendon appearance at 6 months but not at 20 days.
Filardo et al30.	Level IV Case Series	Patellar tendinopathy	43 subjects	30.6 years	3 PRP injections at 2 week intervals	No	Platelet count: n/a Leukocytes: n/a	Blanzina, VISA-P, EQ-VAS, Tegner	Mean 48.6 ± 8.1 months	Improved function up to 4-year follow up. Bilateral pathology or longer chronicity resulted in poorer outcome.
de Vos et al35.	Level I RCT	Achilles tendinopathy (mid-portion)	54 subjects: n= 27 PRP n=27 saline	49.5 years	1 PRP injection* 1 saline injection*	Yes	Platelet count: n/a Leukocytes: n/a	VISA-A, patient satisfaction, return to sport	12 months	No significant difference between groups in pain or activity level.
Krogh et al36.	Level I RCT	Achilles tendinopathy (mid-portion)	24 subjects: n=12 PRP n=12 saline	49.2 ± 9.4 years	1 PRP injection* 1 saline injection*	Yes	Platelet count: 8x basal value Leukocytes: n/a	VISA-A, pain at rest, pain at activity, tendon thickness on US, color Doppler activity	12 months	No significant differences in pain or function at 3 months but PRP did show an increased tendon thickness on US. Drop out rate high after 3 months.
Kearney et al37.	Level I RCT	Achilles tendionpathy (mid-portion)	20 subjects: n=10 PRP n=10 eccentric exercises	49 years	1 injection of PRP Eccentric training program 2x/day, 7d/week × 12 weeks	No	Platelet count: n/a Leukocytes: n/a	VISA-A	6 months	No significant intergroup differences.
Salini et al40.	Level IV case series	Achilles tendinopathy	44 subjects: n=29 young n=15 elderly	Young group: 39.5 ± 6.9 years Elderly group: 61.5± 5.3 years	3 PRP injections at one week intervals*	Yes	Platelet count: 1.6x basal value Leukocytes: n/a	VISA-A	12 months	PRP less effective in elderly population.
Boesen et al41.	Level I RCT	Achilles tendinopathy	60 subjects: n=20 PRP n= 20 HVI n=20 saline injection	43.1 (PRP) vs. 41.9 (HVI) vs. 40.9 (control) years	4 PRP injections at two week intervals# HVI (steroid, saline, or local anesthetic)# Control: few drops of saline under skin	Yes	Platelet count: n/a Leukocytes: n/a	VISA-A, VAS, Intratendinous vascularity and tendon thickness on US, heel-rise test	6 months	PRP superior to control in terms of pain control and patient outcome. PRP with reduction in tendon thickness and vascularity on US, most evident at 6 weeks.
Peerbooms et al.43, Gosens et al44.	Level I RCT	Lateral epicondylosis	100 subjects: n=51 PRP n=49 corticosteroid	47.3 (PRP) vs. 46.9 (steroid) years	1 injection of PRP Control: 1 injection of saline	No	Platelet count: n/a Leukocytes: n/a	VAS, DASH	2 years	Significantly better pain relief and functional improvement in PRP group.
Krogh et al45.	Level I RCT	Lateral epicondylosis	60 subjects: n=20 PRP n=20 glucocorti-coid n=20 saline control	45.4 ± 8.0 years	1 PRP injection* 1 corticosteroid injection* 1 saline injection*	Yes	Platelet count: n/a Leukocytes: n/a	PRTEE, Doppler signal and tendon thickness on ultrasound	3 months	No difference in pain reduction or function at 3 months but glucocorticoid showed a significant reduction in pain at 1 month and decreased color Doppler and tendon thickness on US.
Palacio et al46.	Level I RCT	Lateral epicondylosis	60 subjects: n= 20 neocaine n=20 dexameth-asone n=20 PRP	47.9 (neocaine) vs. 46.2 (dexamethasone) vs. 46.6 (PRP) years	1 injection of each treatment	No	Platelet count: n/a Leukocytes: n/a	DASH, PRTEE	6 months	PRP did not provide improved results over saline or steroid.
Raeissadat et al47.	Level I RCT	Lateral epicondylosis	64 subjects: n=33 PRP n=31 AWB	45.3 ± 5.9 years	1 injection of each treatment	No	Platelet count: 1,227,000± 250,000 per mm3 in PRP group (4.8x basal value) Leukocytes: 6740 ± 1396 per mm3 in PRP group	VAS, PPT, modified Mayo clinic performance index for the elbow	12 months	PRP and AWB were effective at reducing pain and improving function but there were no significant differences.
Mishra et al48.	Level II RCT	Lateral epicondylosis	225 subjects n= 112 PRP n=113 bupivicaine	48.4 (PRP) vs. 47.4 (control) years	1 PRP injection +NT 1 bupivacaine injection +NT	No	Platelet count: 5x basal value Leukocytes: yes	VAS, PRTEE	24 weeks	PRP with statistically significant improvement in pain over control group at 24 weeks but not 12 weeks. NT appears to be complementary to PRP.
Rha et al52.	Level I RCT	Supraspinatus tendinosis or partial tear	39 subjects: n= 20 PRP n= 19 NT	52.2 (PRP) vs. 53.9 (dry needling) years	2 PRP injection with PNT at 4 week intervals^ 2 PNT at 4 week intervals*	Yes	Platelet count: n/a Leukocytes: n/a	SPADI, shoulder ROM, physician global rating scale	6 months	PRP resulted in greater decrease in pain and disability than dry needling alone.
Kesikburun et al53.	Level I RCT	Rotator cuff tendinopathy	40 subjects: n= 20 PRP n= 20 saline control	45.5 (PRP) vs. 51.4 (control) years	1 PRP injection into sub-acromial space# 1 saline injection into sub-acromial space#	Yes	Platelet count: n/a Leukocytes: n/a	WORC, SPADI, VAS	12 months	No inter-group differences regarding QOL, pain, disability, or shoulder ROM.
Lee et al54.	Level IV Case Series	Gluteal tendinopathy or partial tear	19 subjects	48 years	1 PRP injection with PNT*	Yes	Platelet count: n/a Leukocytes: yes	mHHS, HOS-ADL, HOS-Sport, iHOT-33	Mean 19.7 months	Statistically significant improvement in function
Jacobson et al55.	Level II Prospective Study	Gluteal tendionpathy or partial tear	30 subjects: n= 15 PRP n=15 PNT	53 (PRP) vs. 60 (NT) years	1 PRP injection with PNT (10 passes)* 1 PNT (20–30 passes)*	Yes	Platelet count: 4–6x basal value Leukocytes: yes	VAS at rest and with activity	Mean 3 months	Faster improvement in pain scores (2 weeks) in PNT group but no difference at 3 months.
Davenport et al56.	Level I RCT	Proximal hamstring tendinopathy	15 subjects (17 hamstrings): n= 11 hamstrings PRP n= 6 hamstrings AWB	46.6 (PRP) vs. 45.4 (whole blood) years	1 PRP injection with PNT* 1 AWB injection with PNT*	Yes	Platelet count: n/a Leukocytes: yes	mHHS, HOS-ADL, HOS-Sport, iHOT-33, US tendon appearance	6 months	Greater improvement in pain and function in AWB group at 12 weeks but better in PRP at 6 months. No significant difference in US appearance
Fader et al57.	Level IV Case Series	Proximal hamstring tendinopathy	18 subjects	42.6 years	1 PRP injection*	Yes	Platelet count: n/a Leukocytes: no	VAS	6 months	Overall 63% improvement in pain
Wetzel et al58.	Level III Retrospective Study	Proximal hamstring tendinopathy	15 subjects (17 hamstrings): n= 12 hamstrings PRP n=5 control	37.1 (PRP) vs. 42.8 (control) years	1 PRP injection Traditional conservative therapy	No	Platelet count: n/a Leukocytes: n/a	VAS, NPRS	Mean 4.5 months (PRP) Mean 2 months (control)	Significant reduction in VAS in PRP group and all patients returned to prior level of activity.

^*Intra-tendinous;

^#Peritendinous;

^{^}Intra-and Peritendinous

RCT: randomized controlled trial; US: Ultrasound; PNT: Percutaneous Needle Tenotomy; VAS: Visual Analogue Scale; IKDC: International Knee Documentation Committee; VISA-A: Victorian Institute of Sport Assessment-Patellar; EQ-VAS: self rated health score; VISA-A: Victorian Institute of Sport Assessment-Achilles; PRTEE: Patient-Rated Tennis Elbow Evaluation; DASH: Disabilities of the Arm, Shoulder, and Hand; PPT: Pain Pressure Threshold; SPADI: Shoulder Pain and Disability Index; WORC: Western Ontario Rotator Cuff Index; OSS: Oxford Shoulder Score; ASES: American Shoulder and Elbow Surgeons Score; mHHS: modified Harris Hip Score; HOS-ADL: Hip Outcome Score-Activities of Daily Living subscale; HOS-Sport: Hip Outcome Score-Sport-Specific subscale; iHOT-33: International Hip Outcome Tool-33; AWB: Autologous Whole Blood; NPRS: Nirschl Phase Rating Scale

Improving PRP Clinical Studies

A strict subject- and disease demographic selection combined with larger sample size, improved PRP-characterization, and long term follow-ups will be some of the key factors to better define PRP efficacy on tendinopathies. While cell counts have become popular method to objectify dose-response nature of this intervention, the quantification of growth factors, chemokines, cytokines, and other micro-particles might be necessary in order to elucidate dose-response relationship. Several animal studies have focused on leukocyte rich versus leukocyte poor PRP in tendinopathy and this should be further studied in human models as well⁵⁹. While percutaneous needle tenotomy has been shown to be beneficial in treatment of tendinopathy, the combination with PRP should be further studied⁶⁰. Finally, in recent years, tissue conditioning or “prehab” is seen as a way to optimize ultimate outcome in PRP treatments for tendons, and this can be another area of active research⁴².

Mesenchymal Stem Cells: Mechanism of Action

The use of stem cells in tendon injury treatment was inspired by discoveries that mesenchymal stem cells (MSCs) have the ability to differentiate into tenocytes in vitro. In vivo, however, such differentiation ability is inconsistent. Recent studies showed that stem cells are believed to benefit damaged tendons by exerting a paracrine effect from various secretomal molecules⁶¹. In prior animal experiments, MSCs were found to promote early tendon healing and lower re-injury risk⁶². Clinical investigations using stem cells have focused on joint diseases, and this treatment is relatively novel for tendinopathy. Systematic review of use of MSCs in arthritis research found that local irritation and transient fever were the most common adverse effects and the use of MSCs for musculoskeletal injuries appear to carry high safety^63,64. Three sources of MSCs studied for tendon injuries are bone marrow via aspirate concentrate or cultured cells, adipose tissue via stromal vascular fracture or cultured cells, and skin.

Bone Marrow MSCs

One French study that explored the efficacy of BM-MSCs harvested from bone marrow aspirate as orthobiologic augmentation during rotator cuff repair found that 100% of the subjects who underwent MSCs treatment with repair (n = 45) had healed based on ultrasound and MRI imaging versus 67% in the control repair only group at 6 month follow up⁶⁵. In addition, a lower re-tear rate (13% in MSCs group vs 66% in control group) and improved tendon integrity via MRI imaging in the MSCs group was seen at the 10-year follow up⁶⁵. The subjects that were found to have a re-tear in the MSCs group had received fewer cells as compared to those that maintained successful repair (mean cells =14000±9000 in re-tear vs 54000±23000 in successful repair group) indicating that cell concentration may play a significant role. Since 14 ml of concentrate was injected in this study, they determined that there was a risk of absence of healing when the MSCs concentration was lower than 1500 per ml. The size of the tear at the time of surgery, age, and time between diagnosis and repair were not found to affect healing in the MSCs group but were predictive in the control group. Both groups underwent a rehabilitation program emphasizing early range of motion, which might have had a synergistic effect to the intervention⁶⁵. Another study by Kim et al. suggests that bone marrow aspirate concentrates (BMACs) combined with PRP enhances proliferation and migration of tendon derived stem cells (TDSCs), aids in rotator cuff tendon tear healing seen on ultrasound, and improves pain as evidenced by decreased VAS scores at three month follow up. While the effects of PRP and BMACs cannot be delineated, the study suggests that PRP plays a role in providing a scaffold for the BMACs to allow for regeneration and prevention of abnormal differentiation of TDSCs. Limitations of the study included small sample size, lack of control group, and inability to perform tendon biopsy from a patient with rotator cuff tendinopathy⁶⁶.

Adipose Stem Cells

A small pilot case series of 12 subjects with chronic refractory lateral epicondylosis investigated the effects of ultrasound-guided injection of various concentrations of ASC, discovering 79% improved pain scores, improved functional performance (mayo elbow performance index), and reduced structural tendon defects on ultrasound examination up to 1-year post treatment⁶⁷. There were no significant differences in degree of pain reduction or performance but the higher concentration (10⁷ vs 10⁶ cells in 1 ml) group experienced more rapid pain improvement with earlier performance plateau. Mild local swelling in the first 48 hours in 50% of patients was reported with spontaneous resolution in 2 weeks, and no long term adverse effects for one year.

There is one level 1 randomized controlled trial by Usuelli et al. that studies the effects of PRP versus stromal vascular fraction (SVF) for the treatment of mid-portion Achilles tendinopathy⁶⁸. SVF can be derived from native adipose tissue or lipoaspirate and contains mature, progenitor, and stem cells. Usuelli et al. randomized 23 subjects to the PRP group and 21 subjects to the SVF group to undergo a unilateral or bilateral sonographically guided Achilles tendon injection for a total of 28 tendons in each group. The SVF group demonstrated faster improvement with a significantly better outcome in regards to VAS, VISA-A, and American Orthopaedic Foot and Ankle Society (AOFAS) Ankle-Hindfoot scores when compared to the PRP group at 15 and 30 day follow up although both groups showed improvement. There was no statistically significant difference between the groups at follow up beyond 30 days, though the SVF group continued to score slightly better than the PRP group on all outcome measures. MRI evaluation of the tendon lesion area did not demonstrate statistically significant improvement at 180-day follow up, thus not correlating with pain and functional outcome measures⁶⁸. The faster improvement in the SVF group suggests that SVF has a higher anti-inflammatory and immunomodulatory effect than PRP. The use of adipose SVF provides an advantage over purified ASCs given the native microcellular environment that may act as a scaffold for regeneration and the simplified and less costly preparatory steps.

Skin-Derived Tenocyte-Like Cells

Skin-derived tenocyte-like cells (SD-TLCs) are dermal fibroblasts that may be culturally expanded for use in regenerative therapies. There have been two RCTs for the use of SD-TLCs in tendinopathy. Clarke et al. (n=60) compared ultrasound-guided injection of SD-TLCs (17 million cells) plus PRP to PRP only for treatment of patellar tendinopathy⁶⁹. There was a statistically significant improvement in VISA-P scores at 6 months for the SD-TLCs plus PRP group compared to PRP alone. Connell et al. injected SD-TLCs (10 million cells) under ultrasound-guidance in 12 subjects with lateral epicondylosis and demonstrated improved tendon structure on ultrasound as evidenced by decreased thickness, hypoechogenicity, vascularity, and number of tears⁷⁰. Functional scores (Patient Rated Tennis Elbow Evaluation) also improved at 6-week, 3-month, and 6-month follow-ups. These results suggest that SD-TLCs are a relatively safe treatment option for tendinopathy although further high-level research with longer-term follow-up is needed.

Current View of Stem Cell Therapies for Tendon Injuries

The efficacy of cell therapies has been difficult to establish due to the requirement of elaborate injection preparation steps. Many preclinical questions still remain such as mechanism of action, optimum cell processing protocols (centrifugation, enzymatic digestion, and cultural expansion), or best injection/transplantation protocols (dose, procedure interval, addition of other agents such as PRP, and timing of procedure in relation to surgical intervention when applicable). To date, BMAC as a surgical augmentation for rotator cuff tear repair and adipose SVF for Achilles tendinopathy seem to be the only indications where cell therapy should be considered although other cell therapies also appear to be safe.

Improving Clinical Stem Cell Studies

Future clinical studies can be considered to elucidate relative efficacy of cell therapies in a RCT fashion. However, clear mechanism of action and optimization of each cell therapy are still being investigated in the preclinical realm. Consorted effort between scientists and clinicians from various disciplines will be helpful in designing future trials although the high cost for cell processing may be a barrier.

Percutaneous Ultrasonic Tenotomy (PUT)-Brief Discussion

Although typically not considered as a regenerative procedure, PUT is worthy of mentioning. PUT is performed under local anesthesia in clinics with a small incision. The probe emits ultrasonic energy that debrides tendinopathic tissue, which is then emulsified and collected for removal using saline irrigation system. Removal of tendinopathic tissue then results in subsequent regeneration of healthy tendon tissue as so believed in traditional surgical tendon debridement procedures. Clinical evidences are currently limited to case series although they are showing early successes in treating recalcitrant common extensor/flexor tendons, patellar tendons, and Achilles tendons. In a study by Seng et al. for treatment of chronic lateral elbow tendinopathy, for example, 20 subjects who underwent PUT reported reduced pain (mean VAS score decreased from 5.4 to 0.4) and improved function (DASH-Compulsory score from 27.8 to 0.4) that continued to 3-year follow up and 100% of patients had decreased tendon thickness and reduction in hypoechoic lesion size on ultrasound⁷¹. As ultrasound allows direct visualization of pathologic tendon regions, PUT in theory improves accuracy of debridement while keeping the procedure minimally invasive compared to conventional tendon debridement.

Conclusion

Tendinopathy is a prevalent condition leading to impaired sports performance in athletes and disability in the working population. Traditional measures such as physical therapy, corticosteroid injections, or even surgical debridement are sometimes unsuccessful in relieving pain and improving function. The emerging evidences for prolotherapy, PRP, cellular injections, and more recently, PUT have added options that appear to be safe and potentially effective, which patients can consider prior to contemplating a surgical option. Large variability in procedural protocol should be standardized. However, for such standardization to happen, preclinical studies need to be continued to better characterize various types of tendinopathies. Prehab and post-procedural rehabilitation protocol should also be an emphasis for tendinopathy related researches. A shift from the “point-of-care,” procedure-focused treatment paradigm, to the “spectrum-of-care,” prehab/post-procedure rehabilitation approach may aid in optimizing the outcome of tendinopathy treatment.

Abbreviations:

ASCs

adipose stem cells

AT

Achilles tendinopathy

bFGF

basic fibroblast GF

BMAC

bone marrow aspirate concentrate

BM-MSCs

bone marrow mesenchymal stem cells

CET

common extensor tendinopathy

DASH

disabilities of the arm, shoulder, and hand

ESWT

extracorporeal shockwave therapy

GF

growth factors

HGF

hepatocyte GF

IGF

insulin-like GF

NSAIDs

non-steroidal anti-inflammatory drugs

OSD

Osgood-Schlatter disease

PDGF

platelet-derived growth factors

PNT

percutaneous needle tenotomy

PRP

platelet-rich plasma

PUT

percutaneous ultrasonic tenotomy

RCT

randomized controlled trial

SD-TLCs

skin-derived tenocyte like cells

SVF

stromal vascular fraction

TDSC

tendon derived stem cells

TGF

transforming GF

VAS

visual analogue scale

VEGF

vascular endothelial GF

VISA –P/A

Victorian Institute of Sport Assessment-Patellar/Achilles