The effect of different irrigation solutions on fracture healing in a rat femur fracture model.

OBJECTIVES: This study aims to evaluate and compare radiological, biomechanical, histopathological, histomorphometric and immunohistochemical effects of povidone iodine (PVP-I), hydrogen peroxide (HPO) and chlorhexidine gluconate (CHG) on fracture healing in their minimum cytotoxic and most efficient concentrations.
MATERIALS AND METHODS: This experimental animal study, conducted between April 2018 and January 2019, included 48 male Sprague Dawley® rats (aging 9 weeks; weighing 356 g) which were randomly divided into four groups: control (saline), HPO, PVP-I and CHG. Rat model of femoral fracture was established and intramedullary fixation was applied. Solutions were applied to fracture region in determined concentration and time, and all subjects were sacrificed on Day 28. Extracted femurs were investigated radiologically by micro-computed tomography. Then, all groups were divided into two random groups to be evaluated biomechanically, histopathologically, histomorphometrically and immunohistochemically.
RESULTS: In histopathological evaluation, inflammation score of CHG group was significantly lower than other groups, and inflammation score of PVP-I group was significantly lower than control and HPO groups (p<0.05). Biomechanically, flexural strength (σbend) (megapascal) values of CHG and control groups showed similar results, but there was no significant difference between all groups (p>0.05). In immunohistochemical localization of bone morphogenic protein (BMP)-4, osteoblast and chondroblast histoscores (H-scores) of HPO group were significantly lower than other groups, and chondroblast H-score in CHG group was lower than control and PVP-I groups (p<0.05). In immunohistochemical localization of BMP-7, osteoblast H-score was significantly higher in CHG group than other groups (p<0.05).
CONCLUSION: We determined that CHG 0.05% solution had no negative effect on the fourth week of fracture healing histopathologically, immunohistochemically and biomechanically, and is an alternative irrigative to normal saline.

Introduction

Open fractures, which may be defined as incomplete amputations, are among orthopedic surgical emergencies. Open fracture management consists of evaluation of the patient, classification of the fracture, antibiotic treatment, debridement and irrigation, fracture stabilization, wound management and supportive procedures.[1] Infection, nonunion, sepsis, osteomyelitis and amputation are some of the complications that can emerge in case of inappropriate management of open fractures.[2]

Adequate debridement and irrigation are the most important interventions to diminish risks and avoid complications. Open fracture irrigation with normal saline or various irrigatives for optimizing wound and fracture healing has a vital role to diminish bacterial load. Despite this significance, there is no consensus in the literature on the amount, additives or form of irrigation.

Antiseptics and antibiotics have been used as adjuvants in irrigation solutions and there are many studies about the effect of these solutions on soft tissue toxicity and bacterial load, while there is limited number of studies about their effect on fracture healing. Many methods and devices were studied and used to avoid orthopedic infections in open fracture treatment, while changing the solution in irrigation phase is easily applicable, accessible and cost-effective.[3,4]

In this study, we hypothesized that chlorhexidine gluconate (CHG) 0.05% solution is an alternative irrigative in open fracture treatment, due to its antiseptic features without affecting fracture healing. Therefore, in this study, we aimed to evaluate and compare radiological, biomechanical, histopathological, histomorphometric and immunohistochemical effects of povidone iodine (PVP-I), hydrogen peroxide (HPO) and CHG on fracture healing in their minimum cytotoxic and most efficient concentrations.

Patients and Methods

This experimental animal study was conducted at University of Health Sciences Istanbul Bağcılar Training and Research Hospital between April 2018 and January 2019. The study included 48 male Sprague Dawley® rats (weighing 356 g; aging 9 weeks). The study was approved by the University of Health Sciences Istanbul Bağcılar Training and Research Hospital Ethics Committee (Approval Number: 2018-11) and achieved according to ETS-123 and European Convention for the Protection of Vertebrate Animals Used for Experimental and Other Scientific Purposes.[5] In this experimental study, all animals were subjected to all interventions according to the Guide for the Care and Use of Laboratory Animals and Animal Experiments Ethics Committee Approval Report.

Exclusion criteria were inconvenient implant position, fixation failure, infection, weight loss of more than 20%, another fracture line, insufficient nutrition of rats and failure to respond to stimuli.[6]

The rats were anesthetized by intraperitoneal ketamine 80 mg/kg (Ketalar®, Pfizer, Berlin, Germany) and xylazine 5 mg/kg (Rompun®, Bayer, Berlin, Germany). One single dose of gentamicin 8 mg/kg (Genta®, I.E. Ulagay, Istanbul, Türkiye) was injected as antibiotic prophylaxis. After surgical interventions, one dose of subcutaneous carprofen 3 mg/kg (Rimadyl®, Zoetis, Parsippany, New Jersey, USA) was applied to all animals daily for analgesic medication, for two days. Carprofen 3 mg/kg (Rimadyl®, Zoetis, Parsippany, New Jersey, USA) was injected to animals as analgesic medication after two days, if needed. So, animals were protected against pain or disturbance. Right mid-diaphyseal femur of all rats were explored through anterolateral approach, and osteotomized transversely by micromotor (ConMed Linvatec® Pro-6200 Motor, California, USA); thus, rat femur open fracture model was created (Figure 1).

Fracture region was exposed and irrigated with 100 mL 0.9% saline in the control group (n=12). HPO 3% solution was applied to fracture region for 2 min in HPO group (n=12). Fracture region was irrigated with PVP-I 1% solution for 2 min in PVP-I group (n=12). CHG 0.05% solution was used for irrigation for 1 min in CHG group (n=12). Then, fracture regions of all groups were irrigated with 100 mL 0.9% saline. After irrigation, all fractures were fixed with retrograde Kirschner wires and radiographically confirmed (Figure 2). At the end of the fourth week, all rats were sacrificed with high dose ketamine (Ketalar®, Pfizer, Berlin, Germany) and xylazine (Rompun®, Bayer, Berlin, Germany). After sacrification, right femurs of all rats were extracted, and Kirschner wires were removed. Then, femurs were fixed in formalin for histological, radiological and biomechanical investigations.

All samples were evaluated radiologically with micro-computed tomography (mCT) device (SkyScan®, Bruker, Billerica, Massachusetts, USA) for measurement of bone mineral density (BMD, g/cm3) and new callus formation (mm3) (Figure 3). After radiological evaluation, each group was divided into two groups randomly in equal numbers for biomechanical and histological evaluation. In biomechanical test, three point bending test was applied with class 1 calibrated universal mechanical testing device (Alsa Laboratory Devices, Ltd., Istanbul, Turkey). Ultimate bending force (Fmax) values were measured and flexural strength (σbend) was calculated by using the following equations (Eq. 1 to 4):

σbend=Mc/I (Eq. 1) M=FL/4 (Eq. 2) c=D/2 (Eq. 3) I=π(D4-d4)/64 (Eq. 4)

Where M is the maximum bending moment, L is the distance between supports, c is the distance to neutral axis, I is the moment of inertia, D is the outer diameter and d is the inner diameter of the bone. In the name of biomechanical results, Fmax values itself are not enough in understanding the strength of bone, because the strength is the force acting to the fracture surface cross-sectional area of the bone. Thus, we take the σbend values into consideration to understand the effect of healing by looking at strength enhancements.

Six samples of each group were fixed in formalin 10% for one week and decalcified for three days. After alcohol, acetone, xylene and paraffin processes, samples were dehydrated and embedded in paraffin. They were sliced longitudinally with a 3-4 μm microtome and stained with hematoxylin and eosin and Masson’s trichrome and evaluated with the numerical scoring scheme used for histologic evaluation of fracture healing (Table I).[7] Inflammation scoring was based on lymphocyte infiltration numerated as 0 for no inflammation, 1 for mild inflammation, 2 for moderate inflammation and 3 for severe inflammation.

Slices of 3-4 μm were used for quantitative histomorphometric evaluation of fracture region and two regions were measured. These were cortex area (CtAr, mm2) and total callus area (CAr, mm2) that contain ossifying tissue, cartilage tissue and fibrous callus. CAr/CtAr was measured as %.

Total callus area was measured as mean mm2 value of fracture region that contains all tissues inside and outside of cortical bone. Cortical area was measured as mean mm2 value of total area that is fulfilled by cortical tissue.[8,9]

For immunohistochemical evaluation, positive charged slides were used and sections were waited in 60°C oven for 60 min. Samples were stained with DAB (3,3’-Diaminobenzidine) V3 protocol (Ventana BenchMark® XT, V3, Roche Diagnostics, Istanbul, Turkey). After alcohol and xylene, all slides were closed in automatic tissue embedder device (Sakura Tissue-Tek Film® Automated Coverslipper, Sakura Finetek, Torrance, California, USA). Five regions were evaluated by two researchers for positive immune marking of bone morphogenic protein (BMP)-4, BMP-7 and cluster of differentiation (CD) 34 and scored between 0-300 with semiquantitative histoscore (H-score) (Figure 4).[10] H-score is a combined semiquantitative scoring system determined by adding the results of multiplication of the percentage of cells with staining intensity ordinal value (scored from 0 for no signal to 3 for strong signal) with 300 possible values.[11] Staining intensity ordinal values of BMP-4, BMP-7 and CD34 proteins on osteoblasts, chondroblasts and fibroblasts of all groups were evaluated and scored by H-score scoring system. Thus, numerical data were recorded for statistical analysis of immunohistochemical evaluation.

Statistical analysis

Statistical analysis was performed using the IBM SPSS version 22.0 software (IBM Corp., Armonk, NY, USA). Mean, standard deviation, median, ratio and frequency values were used in descriptive analysis of data. Distribution of variables was analyzed with Kolmogorov-Smirnov test. T test, Kruskal-Wallis and Mann-Whitney U tests were used for analysis of independent quantitative variables. The results were evaluated according to the significance value of p<0.05.

Results

In radiological evaluation, CAr/CtAr (%) values were lower in CHG group, while there was no statistically significant difference between all groups (p>0.05). BMD (g/cm3) values of PVP-I group were significantly lower than other groups (p<0.05) with no statistically significant difference between other groups (p>0.05) (Table II).

In biomechanical evaluation, Fmax Newton (N) values of CHG group were higher than other groups, with no statistically significant difference between all groups (p>0.05). σbend (megapascal) values of CHG and control groups were similar and higher than other groups, with no statistically significant difference between all groups (p>0.05) (Table III).

In histopathological evaluation, higher histological fracture healing scores were detected in control and CHG groups, with no statistically significant difference between all four groups (p>0.05). Inflammation scores of CHG group were significantly lower than other groups (p<0.05) (Table IV).

In histomorphometric evaluation, CAr/CtAr % values of HPO group were significantly lower than other groups (p<0.05). In PVP-I group, CAr/CtAr % values were significantly lower than control and CHG groups (p<0.05). CAr/CtAr % values of CHG group were significantly lower than control group (p<0.05).

In H-score values of BMP-4 evaluation, chondroblast H-scores of HPO and CHG groups were significantly lower than control and PVP-I groups (p<0.05), with no statistically significant difference between CHG and HPO groups (p>0.05). H-scores for osteoblasts in HPO group were significantly lower than control and CHG groups (p<0.05), and there was no statistically significant difference between control, PVP-I and CHG groups (p>0.05). When we evaluated BMP-4 expression on fibroblast cells, there was no statistically significant difference between H-scores of CHG, PVP-I, HPO and control groups (p>0.05).

When we evaluated H-scores of BMP-7 localization and immunoreactivity, fibroblast H-scores in CHG group were significantly lower than other groups (p<0.05). Fibroblast H-scores in PVP-I group were significantly lower than control group (p<0.05). There was no statistically significant difference between groups in respect to chondroblast H-scores (p>0.05). Osteoblast H-scores in CHG group were significantly higher than other groups (p<0.05), with no statistically significant difference between PVP-I and control groups (p>0.05). There was no statistically significant difference between all groups in respect to CD34 H-scores (p>0.05).

Discussion

Prevention of orthopedic infections due to open fractures or surgical procedures is of paramount importance. Because of increased resistance to antibiotics and complication rates after orthopedic surgeries, appropriate selection of intraoperative irrigation solutions may be an effective effort to reduce infection rates. Also, an antiseptic solution used in open fracture management as an irrigative could diminish complication rates related to infection.[12,13]

Chlorhexidine gluconate is an antiseptic that uses mechanical features of a liquid removing bacteria and combines antimicrobial features without affecting host cells. It has a broad spectrum for gram-positive and gram-negative bacteria, fungi and prokaryotic cell membrane with low toxicity to mammalian tissue.[14] Interestingly, CHG, which has a wide range of use in modern practice with low toxicity and with effective antiseptic activity, has not been investigated as an irrigation solution in clinical or animal experimental open fracture model. Despite this lack of evidence, some surgeons advise the use of this solution in surgical field instead of normal saline.[5]

One of the advantages of CHG is that there is no resistance to its antimicrobial effect. Also, it does not lose its effect in body fluids like blood, unlike other antiseptic solutions. Moreover, it distributes slowly from solid or soft tissues and shows its antimicrobial effect in an appropriate time period.[15]

Different types of animals including rabbits, dogs, and sheep, etc. are preferred for use in studies.[16] We preferred Sprague Dawley® rats because of standardization of fracture model, repeatability, accessibility to antibodies in study and cost. Experimental fracture models can be created by open (osteotomy) and closed (blunt trauma) techniques. We chose open fracture model to serve the purpose of our study, while blunt trauma may cause high death ratios, failure to standardize fracture type and fragment number and soft tissue damage.[17]

Combining biomechanical and radiological results with quantitative histological results is extremely important for the objectivity of the study. Gerstenfeld et al.[18] determined some parameters of bone healing on histomorphometric investigation of fracture healing. In our study, CAr/CtAr % value of these parameters was chosen to correlate with mCT data and compared with radiological results. CHG and control groups showed similar and parallel radiological and histomorphometric results, and there was no statistically significant difference between other groups. Siwicka et al.[19] showed that different immunohistochemical receptor markers have roles in different weeks of bone healing. Also, Haque et al.[20] indicated bone morphogenic factor, signal molecules in fibroblast and osteoblast cells in fracture healing sites and used semi-quantitative scoring system. In the same study, they suggested semi-quantitative or quantitative evaluation for statistical purposes. Gerstenfeld et al.[21] stated in their study that expression of fracture healing markers changes by stage of fracture healing. Immunohistochemical markers evaluated in our study had important roles in fourth week of fracture healing. In our study, we used BMP-4, BMP-7 and CD34 as immunohistochemical markers of fracture healing in the fourth week of bone healing and evaluated semi-quantitatively between 0 and 300 H-scores for objective results.[22]

Ramchandani and Weber[23] showed that low concentration PVP-I increases mesenchymal stem cell differentiation and new bone formation by increasing BMP-2 and osteopontin expression. In our study, PVP-I 1% had no negative effect on fracture healing in histopathological evaluation. Also, we found no negative effect on osteoblast BMP-4 and BMP-7 expression in immunohistochemical evaluation. However, we showed the negative effect of PVP-I 1% on BMD in the fourth week of bone healing.

Taylor et al.[24] stated that CHG 0.05% eliminates 99.8% of contaminating bacteria within 1 min in a tissue model when used in lavage, which is the rationale behind different exposure times between these three solutions. Moreover, it has no negative effect on cartilage tissue without osteoarthritis. In a study, it was thought that a high concentration of CHG is toxic to human cartilage in open or arthroscopic surgery, in undefined contact times. However, their effect on human osteoblast is still not clear.[25] In a recent study, Vörös et al.[26] showed that CHG has a toxic effect on osteoblasts, in vitro. In our study, we showed that CHG 0.05% has no negative effect on osteoblasts, immunohistochemically. In a contaminated open fracture model, saline and CHG solutions were compared, and usage of CHG alone was not suggested while lavage with saline after CHG irrigation was recommended.[27] In our study, we irrigated the osteotomy field with saline after 1 min of CHG 0.05% usage. Liu et al.[13] showed the toxic effect of CHG on osteoblast, fibroblast and myoblast, in vitro, and they stated that new in vivo studies are needed on this topic. In our study, we showed that CHG 0.05% had no negative effect on BMP-4 and BMP-7 expression of osteoblast, fibroblast and chondroblast while we detected lower H-scores of BMP-7 expression of fibroblast.

In a recent study, Shiels et al.[28] used CHG releasing implant coating on intramedullary nail as fixation device and showed that it reduces infection in rat fracture model. However, they did not study its effect on fracture healing. In our study, we found that CHG 0.05% solution did not have any negative effect on fracture healing, histopathologically. Also, inflammation score of CHG group was lower than other groups. In several studies that compared the effect of different antiseptic solutions on bacterial load and soft tissues, CHG showed similar results with normal saline and also superiority to other antiseptics.[29] In our study, similar results were detected in CHG and control groups in the fourth week of fracture healing. CHG group showed superior results to other antiseptic agents radiologically, histomorphometrically and biomechanically; besides, it showed no negative effect on fracture healing. In immunohistochemical evaluation during which we investigated BMP-4, BMP-7 and CD34 immunoreactivity on osteoblast, fibroblast and chondroblast in the fourth week of fracture healing, we found superior results in CHG group than PVP-I and HPO groups and similar results with control group.

Husodo et al.[30] compared the histomorphometric effects of PVP-I and HPO on fracture healing in a rat fracture model and suggested the use of PVP-I 1% as an irrigative in fracture surgery. To our knowledge, ours is the first and only original study evaluating CHG, PVP-I and HPO solutions radiologically, biomechanically, histopathologically and immunohistochemically[31] with such different parameters.

Our study has some limitations. We did not investigate different fracture types and different weeks of fracture healing due to ethical concerns caused by the increased number of rats and loss of standardization. Furthermore, because of the limited number of rats, we did not evaluate different ratios and contact times of solutions. In conclusion, we recommend the usage of CHG 0.05% solution in fracture surgery and open fracture treatment as an irrigative because of its antiseptic features that do not affect fracture healing, while further comprehensive studies are needed on this topic. We believe that usage of CHG 0.05% as an irrigation solution is a simple, accessible and applicable way to diminish the complications of orthopedic infections.

Table 1

Histological scoring system of fracture healing defined by Huo et al.[12]

Score	Assessment of histology
1	Fibrous tissue
2	Predominant fibrous tissue with minimal cartilage tissue
3	Cartilage tissue and fibrous tissue in a uniform manner
4	Predominant cartilage tissue with minimal fibrous tissue
5	Cartilage tissue
6	Predominant cartilage tissue with minimal immature bone
7	Immature bone and cartilage tissue in a uniform manner
8	Predominant immature bone with minimal cartilage tissue
9	Bone healing with immature bone
10	Bone healing with matured bone

Table 2

Statistical comparison of radiological parameters between groups

	Control		PVP-I		CHG		HPO
	Mean±SD	Median	Mean±SD	Median	Mean±SD	Median	Mean±SD	Median	p*
Bone mineral density	3.7±0.8f	3.7	2.5±0.6	2.5	3.5±0.6f	3.5	3.4±0.6f	3.5	0.003
Callus volume	79.9±26.0	75.5	92.8±46.9	81.3	86.0±18.6	88.1	75.8±27.8	69.8	0.466
Cortex volume	82.4±22.4	87.1	76.3±22.6	81.6	79.0±14.9	76.9	81.2±11.1	82.8	0.832
Cortex/callus volume	1.0±0.4	0.9	1.1±0.5	1.1	0.9±0.2	1.0	1.2±0.4	1.2	0.353
PVP-I: Povidone iodine; CHG: Chlorhexidine gluconate; HPO: Hydrogen peroxide; SD: Standard deviation; * Kruskal-Wallis (Mann-Whitney U test); † Difference with povidone iodine group (p<0.05).

Table 3

Statistical comparison of biomechanical parameters between groups

	Control		PVP-I		CHG		HPO		p*
		Mean±SD	Median	Mean±SD	Median	Mean±SD	Median	Mean±SD		Median
Fmax (N)	64.5±27.2	58.9	56.0±29.6	63.2	78.8±31.7	71.8	60.0±56.8	44.9	0.435
σbend(MPa)	81.8±31.1	69.9	60.1 ±17.8	60.1	71.4±6.4	72.3	65.2±16.5	61.8	0.324
PVP-I: Povidone iodine; CHG: Chlorhexidine gluconate; HPO: Hydrogen peroxide; SD: Standard deviation; Fmax: Ultimate bending force; N: Newton; σbend: Flexural strength; MPa: Megapascal; * Kruskal-Wallis test.

Table 4

Statistical comparison of histopathological findings between groups

	Control		PVP-I		CHG		HPO		p*
		Mean±SD	Median	Mean±SD	Median	Mean±SD	Median	Mean±SD		Median
Fracture Healing Score	4.7±1.0	5.0	4.3±1.2	4.5	5.2±0.8	5.0	3.3±1.5	3.0	0.124
Inflammation Score	1.7±0.5†‡	2.0	1.0±0.6‡	1.0	0.2±0.4	0.0	2.0±0.8†‡	2.0	0.003
PVP-I: Povidone iodine; CHG: Chlorhexidine gluconate; HPO: Hydrogen peroxide; SD: Standard deviation; Fmax: Ultimate bending force; N: Newton; σbend: Flexural strength; MPa: Megapascal; * Kruskal-Wallis test (Mann-Whitney U test); † Difference with povidone iodine group (p<0.05); ‡ Difference with chlorhexidine gluconate group (p<0.05).