Tendon injury is challenging owing to two main causes: reinjury and adhesion (27,44). Tendons have poor vascularity and cellularity compared to other soft tissues, contributing to delayed healing and the gradual formation of imperfect connective tissue scars, which increases the risk of re-injury (24,33). Fibrous adhesion between the healing tendon and surrounding tissue occurs in the early healing phase and may reduce the tendon function (16). Thus, outcomes vary and are often unsatisfactory in many cases (43). Early rehabilitation can prevent adhesion formation, but excessive mobilization can increase the risk of re-injury (17,20). Thus, fast regaining of strength is essential, and to achieve this, various studies have been performed to augment the tendon healing process (4,5,15,27,44,45).

Several regenerative injection therapies (RIT) have been attempted to promote tendon healing (4,7,28). Representative RIT includes platelet-rich plasma (PRP), prolotherapy, and mesenchymal stem cell injection. PRP is a concentrate of platelets commonly containing more than 1×10⁶ cells/µL, which can be obtained by centrifugation of blood samples (41). Although the clinical effects of PRP have been proven through many clinical trials on patient satisfaction, the exact mechanisms are not fully understood (7). Previous in vivo or in vitro studies revealed some mechanisms by which PRP stimulates cell differentiation and growth and promotes the tendon healing process by releasing growth factors such as tumor growth factor β (TGF-β), platelet-derived growth factor (PDGF), insulin-like growth factor-1 (IGF-1), and vascular endothelial growth factor (VEGF) (43). The released growth factors mimic the initial stage of the inflammatory response and augment the natural healing process (41).

Prolotherapy, also called sclerotherapy, is an RIT in which irritants are injected and has been used for various purposes for a long time (7,10,28,38). Prolotherapy was first introduced by American surgeon George Hackett in the 1950s (14). After Hackett (14), prolotherapy elicited satisfactory responses in many clinical trials. Prolotherapy has been commonly used in human medicine for osteoarthritis (3), lateral epicondylalgia (2), and supraspinatus tendinopathy (23), and in veterinary medicine for cranial cruciate ligament injury, spinal pain, and hip laxity (10-12). Hypertonic dextrose, phenol, sodium morrhuate, and polidocanol are used as irritants; the most common clinically reported irritant is hypertonic dextrose. Although prolotherapy has been used for a long time, there is a minimal volume of robust evidence to support its use. The definite mechanisms of prolotherapy remain unclear, with only a few in vitro or in vivo studies (7,9,13,25,26,28). The basic principle of prolotherapy is to create controlled acute inflammation (13). The proposed mechanism particularly in dextrose prolotherapy was that injected dextrose induces “osmotic shock” to tissue, which stimulates inflammation cascade (7).

Although prolotherapy using hypertonic dextrose has been consistently applied in human and veterinary medicine for a long time, only a little robust evidence supports its use. Ekwueme et al. (9) and Güran et al. (13) revealed the mechanisms of dextrose prolotherapy in vitro. However, there are limitations to its in vitro nature. Since the injected dextrose may exert its effect through various interactions with the surrounding soft tissues, experimental in vivo studies are necessary. Martins et al. (25) performed an in vivo study that revealed that dextrose injection had no harmful effects on tendons in the early healing phase, but they could not prove healing-promoting effects owing to the short experiment duration. They also performed a study using an intact tendon. Thus, this study was designed to evaluate the tendon healing-promoting effect of dextrose prolotherapy through an in vivo experiment with a defect-created tendon. Concurrently, no previous in vivo studies compared the effects of dextrose prolotherapy and PRP therapy. By applying the same process as PRP, this study compared the healing effect of PRP or dextrose on tendons.

This study aimed to evaluate the effect of dextrose prolotherapy and PRP on tendon healing using in vivo experiments.

Experimental design

Twenty-one 13-week-old skeletally normal healthy male New Zealand white rabbits were used in this study. No remarkable findings were observed on physical examination or complete blood count (CELL-DYN 3700, Abbott Laboratories, USA). The mean body weight was 2.7 kg (range, 2.4-2.86 kg). The animals were housed in individual standard rabbit cages and kept on a standard rabbit diet without limited access to food or water. This study was approved by the Institutional Animal Care and Use Committee of the Kyungpook National University (KNU2020-0093).

Rabbits were randomly allocated into three groups as follows.

1. Defect control group: No treatment was applied after empty defect creation.

2. Dextrose prolotherapy group (DP group): 20% dextrose was injected into the tendon after empty defect creation.

3. PRP group: PRP was injected into the tendon after empty defect creation.

After a 2 mm round empty defect was surgically created on the right superficial digital flexor tendon (SDFT) of the Achilles tendon complex in all groups, the tendons of each group were treated. All experimental groups were evaluated preoperatively (pre-op) and at 2, 4, and 8 weeks postoperatively. At week 8, SDFTs of both limbs were transected for histological examination. The tendons of the left limb were analyzed as intact controls.

PRP preparation

Autologous PRP was obtained as previously described with some modifications (21). For PRP preparation, 10 mL of autologous blood was collected from the saphenous vein of each animal. Blood was collected in a sterile tube containing 2 mL of the anticoagulant acid citrate dextrose (C.T.G. solution, Daehan New Pharm Co., Korea). A double centrifugation protocol was used to concentrate the platelets to obtain PRP. First, the blood was centrifuged at 1,800 RPM (636 × g) for 10 min at 4°C to separate erythrocytes, buffy coat, and plasma (1580 MGR, GYROZEN Co., Korea). After gentle shaking to float the buffy coat, the plasma was pipetted and transferred into another sterile tube. The second centrifugation was performed at 3,000 RPM (1766 × g) for 15 min at 4°C to concentrate platelets to separate platelet-poor plasma and platelet buttons. Platelet-poor plasma was discarded, except for 1 mL. The tubes were gently agitated to promote resuspension and platelet-rich plasma was obtained. The total number of platelets was measured (CELL-DYN 3700, Abbott Lab., USA), and plasma with more than 1×10⁶ platelets/L was used in this study. PRP was applied to each rabbit in subsequent experiments.

Treatment procedure

Rabbits were anesthetized by intramuscular injection of a combination of 35 mg/kg ketamine (Ketamine 50 inj., Yuhan Co., Korea) and 5 mg/kg xylazine (Rompun inj., Bayer, Germany). The skin over the right hind paw was shaved and aseptically prepared for the surgery. The skin and subcutaneous tissue in the Achilles tendon area were incised 3 cm longitudinally. The crural fascia and paratenon surrounding the Achilles tendon complex were split to expose the Achilles tendon. The SDFT was isolated from the surrounding tendons, and an empty round defect was created at the mid-point of the SDFT on 10 mm upper side from the calcaneus bone, using a 2 mm biopsy punch. In the defect control group, the defect remained untreated whereas 0.2 mL of the obtained PRP and 0.2 mL of 20% dextrose were injected into the proximal and distal ends of SDFT mass, in the PRP and DP groups, respectively (Fig. 1). Approximately, 0.2 mL of volume was decided because flow-out was observed when a larger volume was injected into the tendon in the pilot study. The SDFT and surrounding tendons were repositioned in the normal anatomic site, and then the paratenon and the crural fascia were closed with a simple continuous suture using a 4-0 multifilament absorbable suture (Vicryl, Ethicon, Germany). The subcutaneous tissue was closed with intradermal sutures using 3-0 multifilament absorbable sutures. Ten mg/kg enrofloxacin (Baytril 50 inj., Bayer, Germany) was injected subcutaneously for antibiotics and 4 mg/kg tramadol (Maritrol inj., Jeil Pharmaceutical Co., Korea) was injected subcutaneously for perioperative analgesia and were prescribed twice daily for 7 days postoperatively. Rabbits were allowed to move freely without immobilization or restriction.

Figure 1. (A) Treatment procedure. Defect creation with 2 mm biopsy punch. (B) Created round defect. (C) Injection of materials to the distal end of the superficial distal flexor tendon (SDFT). (D) Injection of materials to the proximal end of the SDFT.

Diameter measurement

The diameter of SDFT (exposed tendon diameter: ETD) was directly measured when the tendon was surgically exposed during defect creation and before sampling on week 8 using an electronic digital caliper (DC200-1, CAS, Korea). For a straightforward comparison, the percentage ratio was calculated by comparing preoperative values.

Radiography

Radiographic examinations were performed preoperatively and at weeks 2, 4, and 8. Laterally and craniocaudally positioned radiographs were obtained using a digital radiographic system (AccuRay-603R; DK Medical Systems, Korea). The musculoskeletal changes were also evaluated. All the radiographs were evaluated by a veterinary radiologist.

Ultrasonography

SDFTs of the right hind limb of all animals were evaluated by ultrasonography pre-operatively and at weeks 2, 4, and 8 using a 12-MHz linear array transducer (Prosound F75, Tokyo, Japan). Changes in tendon characteristics such as echogenicity, margination, and degree of swelling were evaluated. Additionally, the width and thickness of the SDFT were measured using a picture archiving and communication system (INFINITT PACS, INFINITT Healthcare, Korea). Cross-sectional area (CSA) was calculated using the measured width and thickness. For a straightforward comparison, the percentage ratio was calculated by comparing pre-operative values.

Sampling

All rabbits were anesthetized on week 8, and SDFTs of both hindlimbs were transected below the musculotendinous junction and over the calcaneal attachment. Tendons of three rabbits from each group were fixed with 10% neutral buffered formalin for histomorphometric examination, and tendons of four rabbits from each group were stored at 4°C untreated for biomechanical analysis. All tendons of the right hindlimb were included in the examination, and seven tendons of the left hindlimbs were randomly selected for intact control.

Biomechanical analysis

The tensile strength was measured to analyze the biomechanical characteristics of the SDFT. After fixing each SDFT on a computerized testing machine (SV-H1000, Japan Instrumentation System Co., Japan), the peak value of the force generated during constant and bilateral stretching (300 mm/min) of the tendon until it reached 5 mm was measured.

Histopathologic evaluation

Each tendon fixed with formalin was crossly trimmed as one part around defected region and re-fixed in 10% neutral buffered formalin for 24 h. Paraffin blocks were prepared for all samples using an automated tissue processor (Shandon Citadel 2000, Thermo Scientific, USA) and embedding center (Shandon Histostar, Thermo Scientific, USA). Four serial paraffin sections (3-4 µm) were prepared from each paraffin block using an automated microtome (RM2255, Leica Biosystems, Germany). Representative sections were stained using different methods for histopathological analysis.

The H&E staining was performed to evaluate the histopathological changes. Histomorphometric properties, such as cell density, nuclear roundness, and fiber arrangement and structure, were evaluated using a previously described semi-quantitative grading score system with some modifications (6). A perfectly normal tendon was graded 0, mild and moderate abnormal tendons were scored 1 and 2, respectively, and a severely abnormal tendon was scored 3.

The number of infiltrated inflammatory cells was analyzed by H&E staining using a computer-assisted image analysis program and a histological camera system. To measure the collagen fiber-occupied region, Masson’s trichrome staining (MT staining) was performed. Additionally, avidin-biotin-peroxidase complex (ABC)-based immunohistochemical staining was performed for collagen types I and III (type I, 1:100, ab34710, Abcam, UK; type III, 1:100, ab7778, Abcam, UK). Immunostained samples were detected using a Vectastain Elite ABC Kit (1:50, PK-6200, Vector Lab. Inc., USA; peroxidase substrate kit, 1:50; SK-4100, Vector Lab. Inc., USA).

Tissues stained by MT stain or immunohistochemical staining were regarded as positive if the degree of staining and immunoreactivity of the stained region exceeded 20% compared to the background. The histological profiles were observed under a light microscope (Model Eclipse 80i, Nikon, Japan) equipped with a histological camera system (ProgResTM C5, Jenoptik Optical Systems GmbH, Germany) and computer-assisted automated image analyzer (iSolution FL ver 9.1, IMT i-solution Inc., Vancouver, BC, Canada).

Statistical analysis

All numerical values are expressed as mean ± standard deviation. Multiple comparison tests were conducted for different groups. Variance homogeneity was examined using Levene’s test. If the Levene test indicated no significant deviations from variance homogeneity, the obtained data were analyzed using a one-way ANOVA followed by the least-significant differences multi-comparison test to determine which pairs of group comparisons were significantly different. In case significant deviations from variance homogeneity were observed using the Levene test, a non-parametric comparison test, the Kruskal-Wallis H test, was conducted followed by the Mann-Whitney U test to determine the specific pairs of group comparisons that were significantly different. Statistical analyses were conducted using SPSS for Windows (Release 14.0, IBM SPSS Inc., USA). Differences were considered significant at p < 0.05, p < 0.01, and p < 0.001.

Body weight

The mean body weight of the defect control, DP, and PRP groups was 2.69 ± 0.14, 2.66 ± 0.16, and 2.70 ± 0.13 kg at arrival, respectively. On week 8, mean body weight significantly increased (p < 0.01) in all groups: 2.99 ± 0.20, 2.99 ± 0.19, and 3.09 ± 0.12 kg for the defect control, DP, and PRP groups, respectively. No significant differences were observed between the groups.

Platelet count

The number of platelets in PRP used in this study ranged from 1.018 × 10⁶ to 1.871 × 10⁶ cells/µL. The mean number of platelets was 1.420 × 10⁶, and the standard deviation was 3.14 x 10⁵ cells/µL.

Diameter evaluation

The exposed tendon diameter (ETD) was measured (Fig. 2) during surgery and immediately before sampling. The pre-operative mean diameter of tendons was 2.81 ± 0.25, 2.66 ± 0.26, and 2.74 ± 0.26 mm for the defect control, DP, and PRP groups, respectively. Mean ETD on week 8 was increased compared to pre-op: 3.35 ± 0.48, 3.88 ± 0.48, and 4.00 ± 0.89 mm for the defect control, DP, and PRP groups on week 8, respectively. The mean rates of change in diameter between pre-operation and week 8 were 120.54 ± 23.62, 146.50 ± 16.88, and 147.30 ± 37.93% for the defect control, DP, and PRP groups, respectively. No significant differences were observed between the groups.

Figure 2. Changes in exposed tendon diameter (ETD). DP, dextrose prolotherapy; PRP, platelet-rich plasma.

Radiography

No remarkable musculoskeletal changes were observed during the experimental period in any of the groups. Some degree of soft tissue swelling was observed in all groups after defect creation and decreased gradually; however, the quantitative assessment was limited.

Ultrasonography

Owing to the limited observation of the defect area on the tendon after treatment, the midpoint between the musculotendinous junction and calcaneal attachment of the SDFT was evaluated (Fig. 3). On pre-operative examination, the SDFT was identified as relatively homogeneously hypoechoic and relatively distinguishable from other tendons of the Achilles tendon complex. At week 2, the heterogeneity of SDFT increased, similar to other tendons of the Achilles tendon complex, and SDFT was relatively indistinct from other tendons due to poor margination. At week 4, the echogenicity of the tendons increased or decreased, independent of the groups. On week 8, the echogenicity of the tendons was observed as relatively hypoechoic compared to week 4 and relatively well-marginated. No remarkable consistency was observed between the groups.

Figure 3. Representative images of ultrasonographic evaluation of the superficial distal flexor tendon (SDFT) (Dashed circle). (A) Pre-op, (B) week 2, (C) week 4, and (D) week 8.

Additionally, the width and thickness of the SDFT were measured using ultrasonography, followed by the calculation of CSA, assuming that SDFT has a complete oval shape due to an ill-defined margin and irregular shape of the tendon (Fig. 4). Different trends were observed among the groups. The CSA of the tendons gradually increased throughout the duration of the experiment in the defect control group. The CSA of the tendons increased until week 4 and gradually decreased thereafter in the DP group. The CSA of the tendons increased until week 2 and then decreased in the PRP group. The mean CSAs of the SDFT at week 8 were not significantly different between the groups.

Figure 4. Mean rate of changes in the cross-sectional area (CSA) of tendon. *p < 0.05, **p < 0.01, ***p < 0.001 as compared with the pre-op in each group. DP, dextrose prolotherapy; PRP, platelet-rich plasma.

Tensile strength

The mean tensile strength of the intact control, defect control, DP, and PRP groups were 72.46 ± 10.33, 48.53 ± 5.22, 68.91 ± 5.77, and 83.84 ± 6.48 N, respectively (Fig. 5). The mean tensile strength of the defect control group was significantly lower than that of the other groups (p < 0.01). The mean tensile strength of the DP group was significantly lower than that of the PRP group (p < 0.05). No significant differences were observed between the intact control and DP groups (Fig. 6). In summary, RITs enhanced the tensile strength of the treatment group, and PRP showed a superior effect.

Figure 5. Tensile strength. *p < 0.05 and **p < 0.01 as compared with the intact control group, ^##p < 0.01 as compared with the defect control group, ^†p < 0.05 as compared with the dextrose prolotherapy (DP) group. DP, dextrose prolotherapy; PRP, platelet-rich plasma.

Figure 6. The representative general histopathological profiles under hematoxylin and eosin (H&E) and Masson’s trichrome (MT) stains of tendon. Scale bars = 100 μm. DP, dextrose prolotherapy; PRP, platelet-rich plasma.

Histopathologic evaluation

Histomorphometric properties, such as cell density, nuclear roundness, fiber arrangement, and fiber structure, were semi-quantitatively analyzed using H&E staining (Table 1, Fig. 6). In the treatment groups, semi-quantitative scores of cell density, nuclear roundness, fiber arrangement, and fiber structure were significantly lower than those in the defect control group and significantly higher than those in the intact control group (p < 0.05). The semi-quantitative scores of nuclear roundness in the PRP group were significantly lower than those in the DP group (p < 0.05).

Table 1 . Histomorphometric analysis.

Histomorphometrical index	Control groups			Treatment groups
	Intact control group	Defect control group		DP group	PRP group
Semi-quantitative score
Cell density	0.33 ± 0.52	2.83 ± 0.41**		1.83 ± 0.41**,##	1.33 ± 0.52**,##
Nuclear roundness	0.50 ± 0.55	2.83 ± 0.41**		2.17 ± 0.41**,#	1.50 ± 0.55**,##,†
Fiber arrangement	0.17 ± 0.41	2.83 ± 0.41**		1.83 ± 0.41**,##	1.33 ± 0.52**,##
Fiber structure	0.17 ± 0.41	2.83 ± 0.41**		2.00 ± 0.63**,##	1.67 ± 0.52**,##
Quantitative histomorphometric analysis
Mean inflammatory cell numbers (cells/mm2)	7.00 ± 3.03	238.00 ± 109.92**		72.00 ± 16.97*	43.67 ± 14.17#
Mean MT+ collagen occupied regions (%)	86.35 ± 8.13	45.51 ± 11.04**		71.883 ± 10.03*,##	78.90 ± 11.03##
Mean COL I immunolabeled regions (%)	87.63 ± 5.46	18.22 ± 4.88**		33.98 ± 5.38**,##	59.80 ± 10.77**,##,††
Mean COL III immunostained regions (%)	5.90 ± 2.62	46.38 ± 6.78**		31.19 ± 5.76**,##	21.21 ± 5.81**,##,††

DP, dextrose prolotherapy; PRP, platelet-rich plasma; MT, Masson’s trichrome..

*<0.05 and **<0.01 as compared with the intact control group, ^#<0.05 and ^##<0.01 as compared with the defect control group, †<0.05 and ††<0.01 as compared with the DP group..

The mean number of infiltrated inflammatory cells under H&E staining was significantly higher in the defect control and DP groups than in the intact control group (p < 0.01). The mean number of infiltrated inflammatory cells was significantly lower in the PRP group than in the DP group (p < 0.01). No significant differences were observed between the defect control and DP groups (Table 1, Fig. 6).

The mean number of collagen fiber-occupied regions under MT staining in the treatment groups was significantly higher than that in the defect control group (p < 0.01). No significant differences were observed between the treatment groups or between the intact control and PRP groups (Table 1, Fig. 6).

The mean number of COL I-immunoreactive regions in the treatment groups was significantly higher than that in the defect control group (p < 0.01). The mean COL I immunoreactive region was significantly higher in the PRP group than that in the DP group (p < 0.01) (Table 1, Fig. 7).

Figure 7. The representative avidin-biotin- peroxidase complex (ABC)-based immunohistochemically stained profiles of COL I and COL III of tendon. Scale bars = 100 μm. DP, dextrose prolotherapy; PRP, platelet-rich plasma.

The mean number of COL III-immunoreactive regions in the treatment groups was significantly lower than that in the defect control group (p < 0.01). The mean COL III immunoreactive region was significantly lower in the PRP group than that in the DP group (p < 0.01) (Table 1, Fig. 7).

In summary, tendons that received RITs had histomorphometric characteristics that were closer to those of intact tendons.

In the present study, the SDFT of the rabbit Achilles tendon complex was selected. Although the Achilles tendon complex is the strongest and largest tendon, it is the most commonly injured tendon owing to its superficial location and strong load generated by tendons (39). Rabbits are favorably selected animals for Achilles tendon complex studies because of their low cost, availability, and relatively large size, which allows for a surgical process (5). The SDFT can be easily exposed and is more accessible than other tendons of the Achilles tendon complex (31).

Therefore, a tendon injury model is required to study tendon healing. Several tendon injury models, such as the full transverse transection, partial transverse transection, multiple longitudinal incision, and partial defect models, have been described in previous studies (1,22,32,40,42). Stoll et al. (40) first described a partial tendon defect model with a biopsy punch and suggested that the model is useful for evaluating the effect of biomaterials on tendon healing. This model not only induced an injury-like situation but also had sufficient stability to obviate additional surgical intervention or immobilization after defect creation. However, biomechanical properties, which are essential characteristics of injury models, have not been analyzed in previous studies.

In the present study, the tensile strength was evaluated as a biomechanical property. Significant differences in tensile strength between the defect-created and intact tendons were observed; furthermore, these results corresponded well with the histological changes. Since early rehabilitation is important for tendon healing (17,20), it is essential to restore sufficient force earlier than tensile force to allow for rehabilitation. Biomechanical analysis in this study revealed that the tensile strength of the tendon that received RIT was significantly higher than that of the defect control group at week 8. These results indicate that RIT helps restore tensile strength earlier and can contribute to early rehabilitation and that PRP showed superior effects compared to dextrose prolotherapy. These results seemed to be consistent with previous studies. In addition to promote tendon healing, PRP was reported to improve tensile strength (18), while effect of DP on tensile strength was not remarkable, in Achilles tendon (25).

In the semi-quantitative scoring system, higher scores were graded when the tendon appeared abnormal, while lower scores were graded for tendons that had closer to normal characteristics. Injured tenocytes appear round in shape, whereas normal tenocyte appears spindle-shaped (30). Tenocyte density and cellularity increase after injury and then gradually decrease (16). Cell arrangement and fiber structure exhibit waviness as abnormalities in the early healing phase, and subsequently become normal with a parallel appearance under physical stimulation (16,30). In the treatment groups, scores were significantly lower than those of the control group and significantly higher than those of the intact control group (Table 1). These results suggest that the tendons of the treatment groups did not fully recover at week 8; however, the tendons of the treatment groups had characteristics closer to those of the normal tendon. There were no remarkable differences between the treatment groups in the semi-quantitative histologic analysis.

Inflammatory cells are prominent in the early healing phase and decrease during the healing process (32). The number of inflammatory cells was significantly lower than that in the control group (Table 1). Collagen fibers decreased in the early healing phase with an increase in cellularity. As the healing process progresses, cellularity decreases, and collagen fibers increase. COL I is the main component of normal tendon ECM. COL III is initially synthesized after injury, is replaced by COL I, and becomes more fibrous during the late healing phase (29,44). In this study, the mean collagen fiber-occupied and mean COL I-immunoreactive regions were significantly higher, while the mean COL III-immunoreactive regions were lower in the treatment groups than in the defect control group. The mean COL I-immunoreactive region was higher, and the mean COL III-immunoreactive region was lower in the PRP group than that in the DP group. Similar to the semi-quantitative analysis, these results could be interpreted as the tendon of the treatment groups having characteristics closer to the normal tendon, and the tendon of the PRP group appearing closer to the normal tendon than the DP group. The results of the histological examination were also supported by those of the ultrasonographic examination in this study.

Tendon CSA was evaluated pre-operatively and at weeks 2, 4, and 8. CSA is related to the degree of swelling. In the early healing phase of the tendon, the water content associated with swelling is higher than that of the normal tendon owing to increased proteoglycan and disruption of the collagen network (34). The CSA in the PRP- or dextrose-treated groups increased earlier than that in the defect control group. This could be interpreted as RITs accelerating the early phase of tendon healing. These accelerative changes appeared earlier in the PRP group than that in the DP group (Fig. 4).

Dextrose is the most common substance used clinically for prolotherapy. Usage of various dextrose concentrations from 12.5 to 25%, is documented in previous studies. In general, a combination with other drugs is used. In this study, dextrose, without other drugs, was used to evaluate its effectiveness. There are no distinct guidelines for dextrose concentration. In this study, one of the common concentrations of dextrose of 20% was selected (3,28,36,45). Further studies are required to determine the optimal concentration and combination for dextrose prolotherapy in tendon repair.

This is the first in vivo study to evaluate the effect of dextrose prolotherapy in the remodeling phase and to compare the effects of dextrose prolotherapy and PRP therapy on tendon healing. Both RITs, dextrose prolotherapy, and PRP therapy promoted the tendon healing process and enhanced the tensile strength in this study; the effect of PRP therapy was greater than that of dextrose prolotherapy. These results are consistent with those of previous systemic reviews in humans that compared the effects of dextrose prolotherapy and PRP therapy (19,35,37). The most important finding of this study was that RIT using dextrose or PRP promoted the healing process of the injured tendon with a significant increase in tensile strength.

PRP could be influenced by various factors, including state and composition of blood origin, and production and activation protocols (8). Therefore, optimization and standardization for preparing PRP seemed to be followed for clinical applications, not only counting platelet population.

In conclusion, this study suggests that PRP therapy and dextrose prolotherapy can promote the healing of Achilles tendon injury, and the healing promotive effects of PRP were superior to those of dextrose prolotherapy.

The authors have no conflicting interests.