Doxycycline speeds up healing of chronic venous ulcers

Abstract

Venous ulcers are common, with an overall prevalence of up to 2% in the general population of western countries, and have significant socioeconomic impact. Matrix metalloproteinases (MMPs) are involved in the alteration of extracellular matrix that could lead to venous ulceration. Sixty‐four patients with venous ulcers were recruited in a 22‐month period. All patients were subjected to the most appropriate treatment considering also the patient's wishes (compression therapy followed or not by vein surgery). Patients were randomised into two groups of 32 persons in each (groups A and B). Patients of group A in addition to the basic treatment, described above, received the administration of oral low doses of doxycycline 20 mg b.i.d. for 3 months, whereas patients of group B received basic treatment only. Healing was assessed by means of direct ulcer tracing with computerised planimetry. Group A showed a higher healing rate compared with group B. In group B, the lower healing rate was related to higher levels of MMP‐9; neutrophil gelatinase‐associated lipocalin and vascular endothelial growth factor, documented in plasma; wound fluid and biopsies executed and compared between both groups. Pharmacological treatments, as doxycycline administration, which by means of its immunomodulatory and anti‐inflammatory actions, through the inhibition of MMP, could improve extracellular matrix functioning and represent a possible solution to support wound healing.

Introduction

Chronic venous disorders are widespread in western countries and their main complication, that is, venous leg ulcers, with an overall prevalence ranging up to 2% in the general population, is responsible for important socioeconomic problems 1, 2, 3, 4, 5. Chronic wounds, that is, venous ulcers, exhibit dysfunctions in extracellular matrix that cannot support healing, and some studies reported that both matrix metalloproteinases (MMPs) and neutrophil gelatinase‐associated lipocalin (NGAL) could play a role in the healing process in patients with chronic venous ulcers 6, 7.

NGAL seems to positively modulate the activity of the MMP‐9 enzyme 8, 9, which degrades basement membranes as well as the extracellular matrix, thus liberating vascular endothelial growth factor, which is involved in wound healing 10.

Several antibiotics have been reported to be able, at lower dosages, to have immunomodulatory and anti‐inflammatory actions, suppressing the ‘cytokine storm’ of inflammation and conferring an additional clinical benefit through their immunomodulatory properties 11, 12, 13.

In particular, tetracyclines are able to affect inflammation, immunomodulation, cell proliferation and angiogenesis and have been used to treat a wide variety of non‐infectious diseases including rheumatoid arthritis, acne and periodontitis 11. Moreover, clinical and experimental studies reported that doxycycline is able to inhibit, at lower doses, the activation of MMPs 14, 15, 16, 17, 18.

In this light, in this study, we evaluated the effects of subantimicrobial doses of doxycycline in patients with chronic venous leg ulcers.

Methods

We performed an open‐label, parallel‐group, single clinical centre study, conducted between January 2010 and October 2012 with prior approval from the Institutional Review Board, in accordance with the Declaration of Helsinki and the Guideline for Good Clinical Practice. Before the beginning of this study, all participants provided written informed consent.

Patients eligible for this study were of both sexes, older than 20 years with chronic venous leg ulcer [class 6 of Clinical‐Etiology‐Anatomy‐Pathophysiology (CEAP) clinical classification (for this classification, see reference 19)], evidence of venous reflux at duplex scanning and with no evidence of ischaemia (ankle‐brachial pressure index > 0·9).

Inclusion criteria were as follows: age: 20–70 years; the presence of a venous leg ulcer for a minimum of 6 weeks, as chronic leg ulcers often are defined as those wounds lasting longer than 6 weeks 2. Exclusion criteria were as follows: arterial diseases, the presence of bacterial infections within the previous 6 weeks, connective tissue disorders including rheumatoid arthritis, blood disorders, cancer, gastroenteritis, history of allergy to tetracyclines and a recent use of tetracyclines (within 3 months). In all patients at the time of admission, the medical history was recorded and clinical examination, laboratory analyses and duplex ultrasonography were performed.

Experimental protocol

The enrolled patients were randomised into two groups: group A: 32 subjects treated with doxycycline 20 mg b.i.d. for 3 months + basic treatment and group B: 32 subjects treated with basic treatment only.

All patients were subjected to the most appropriate treatment (defined as basic treatment), considering also the patient's wishes. The type of surgery, when it was accepted, was chosen according to the anatomical level of vein incompetence: superficial venous surgery [Cure Conservatrice et Haemodinamique de l'InsuffisanceVeineuse en Ambulatorie (CHIVA) procedure was used for the correction of superficial venous reflux] and/or subfascial endoscopic perforating surgery (SEPS) after computed haemodynamic mapping 20 (Table 1).

Table 1.

Baseline characteristics of the patients enrolled in this study

Enrolled	Group A	Group B
Number of subjects (n = 64)	32	32
Sex (male/female)	9/23	11/21
Age (range), years	50·5 ± 8 (41–60)	51·3 ± 7·5 (46–57)
Duplex ultrasound findings
Superficial vein incompetence	27 (84·37%)	29 (96·66%)
Perforating vein incompetence	3 (9·37%)	2 (6·25%)
Superficial + perforating incompetence	2 (6·25%)	1 (3·12%)
Surgery
Superficial vein surgery	27 (84·37%)	29 (96·66%)
Perforating vein surgery	3 (9·37%)	2 (6·25%)
Superficial + perforating vein surgery	2 (6·25%)	1 (3·12%)
Ulcer area
Range, cm2	2·9–19·5	3·1–17·5
Median, cm2	12·9	11·7

The patients, regardless of whether surgery was performed or not, were also treated with third‐class medical compression stocking (36–50 mmHg) for 3 months and then with second‐class medical compression stocking (25–40 mmHg) for 2 months. The median follow‐up was 13 months. Finally, an adapted local care was administered in agreement with the typology of the wound.

Healing evaluation

The healing was assessed, in agreement with the previous studies by means of computed planimetry 6, 21. Briefly, healing was calculated by subtracting the final ulcer area expressed in centimetre and assessed at the end of the study (T = 2) and from the initial ulcer area expressed in centimetre and evaluated at the time of the admission (T = 0). The result was divided by the number of weeks that the patient had been observed to obtain the total area healed per week.

Biopsies, wound fluid and plasma samples were collected at the time of admission (T = 0), after 4 weeks (T = 1) and 3 (T = 2) and 5 months later (T = 3) (see Figure 1). For wound‐healing evaluation, we considered as high‐healing ulcers those with a healing speed rate ≥ 1 cm²/week and slow‐healing ulcers those with healing speed rate < 1 cm²/week.

Figure 1.

Plasma levels of matrix metalloproteinase‐9 (MMP‐9) (up) and neutrophil gelatinase‐associated lipocalin (NGAL) (down) evaluated through enzyme‐linked immunosorbent assay (ELISA) test in both tetracycline‐treated (group A) and non treated patients (group B). The results represent mean ± SEM. **P < 0·01; *P < 0·05.

NGAL, MMP‐9 and VEGF evaluation in plasma and wound fluid

An enzyme‐linked immunosorbent assay (ELISA); Human ELISA System, Amersham Pharmacia Biotech, Buckinghamshire, UK kit was used to determine the concentrations of NGAL, MMP‐9 and vascular endothelial growth factor (VEGF) in both plasma and wound fluid, in agreement with the previous studies 6.

Protein extraction and immunoblot analysis

In both groups, the biopsies obtained at the time of the study were collected for Western blot and then evaluated, as previously described 22, 23, 24, 25, 26, 27, 28, 29. Immunoblotting was performed using anti‐MMP‐9 and anti‐NGAL monoclonal antibodies and expressed as arbitrary units, as recently described 6. All experiments were performed in triplicate.

Statistical analysis

All data are expressed as mean ± standard error of the mean (SEM). Statistical evaluation of the results was performed by the Student's t‐test. The threshold of statistical significance was set at *P < 0·05.

Results

During the study period, 64 patients consisting of 44 females and 20 males were enrolled in this study and the informed consent process was completed. In particular, 32 patients (23 females and 9 males, age range: 41–60 years) were enrolled in group A and 32 patients (21 females and 11 males, age range: 46–57 years) were enrolled in group B (see Table 1), without differences in demographic characteristics (data not shown).

ELISA findings showed lower levels (P < 0·01) of MMP‐9, NGAL and VEGF in both plasma and wound fluid of patients treated with tetracycline (group A) with respect to those not treated with tetracycline (group B) (Figures 2, 3, 4). These effects were significantly higher in patients with high‐healing ulcers with respect to patients with low‐healing ulcers (P < 0·01) (Figures 1 and 2).

Figure 2.

Wound fluid levels of matrix metalloproteinase‐9 (MMP‐9) (up) and neutrophil gelatinase‐associated lipocalin (NGAL) (down) evaluated through enzyme‐linked immunosorbent assay (ELISA) test in both tetracycline‐treated (group A) and non‐treated patients (group B). The results represent mean ± SEM. **P < 0·01; *P < 0·05.

Figure 3.

Tissue expression of matrix metalloproteinase‐9 (MMP‐9) (up) and neutrophil gelatinase‐associated lipocalin (NGAL) (down) in both doxycycline‐treated (group A) and non‐treated patients (group B). The results represent mean ± SEM. **P < 0·01; *P < 0·05.

Figure 4.

Plasma (up) and wound fluid (down) levels of vascular endothelial growth factor (VEGF) evaluated through enzyme‐linked immunosorbent assay (ELISA) test in both tetracycline‐treated (group A) and non‐treated patients (group B). The results represent mean ± SEM. **P < 0·01; *P < 0·05.

Western blot analysis showed an increased expression of MMP‐9 and NGAL in ulcer tissues of group B with respect to group A (P < 0·01; Figure 3), which was significantly higher in patients with low‐healing ulcers with respect to patients with high‐healing ulcers (Figure 4).

Discussion

In this study, we documented that doxycycline is able to improve the healing of ulcer through the inhibition of MMP‐9, NGAL and VEGF activation. Recently, several authors postulated that increased levels of proinflammatory cytokines and the dysregulation of MMP‐9 and NGAL are involved in impaired healing in patients with chronic venous leg ulcers 6, 7, 8, 9.

In particular, a previous study showed that patients with low‐healing ulcers and non‐healing ulcers had significantly higher levels of MMP‐9 and NGAL with respect to patients with high‐healing ulcers 6. In agreement with these findings, in the present study, we documented higher levels of MMP‐9, NGAL and VEGF in low‐healing ulcers with respect to high‐healing ulcers.

Pharmaceutical strategies speeding up wound healing in chronic venous ulcers, by supporting surgical and compression therapies, thereby reducing the need for wound care, could have major advantages for patients socio‐economically 21, 30, 31. Moreover, it has been reported that doxycycline through its anti‐inflammatory and immunomodulatory effects represents a potential treatment for some conditions mediated by MMPs (e.g. arterial aneurysms, periodontitis and arthritis) 32.

In this study, we documented that the long‐lasting treatment with lower doses of doxycycline is able to significantly reduce both MMP‐9 and NGAL expression in ulcer tissues. Moreover, we also documented decreased levels of MMP‐9, NGAL and VEGF in both plasma and wound fluid in doxycycline‐treated patients with respect to untreated patients.

These data support the hypothesis that doxycycline could be able to improve other clinical conditions other than infectious diseases, such as chronic ulcers. However, we were not able to evaluate the impact of lower doses of antimicrobial on the development of bacterial resistance and this could represent a limitation of this study.