A prospective randomized controlled trial of bone metabolism in patients with charcot foot

Diabetic foot disease (DFD) is one of the serious late complications of diabetes mellitus (DM). It induces a rapid decrease in the patient’s quality of life and requires significant financial costs of treatment for both the patient and the entire healthcare system [1].

Diabetic neuroosteoarthropathy (DNOAP) is a form of DFD that occurs when a neurological deficiency is caused by diabetic neuropathy and is manifested as the destruction of bones, joints and cartilage tissue lesion of feet as a result of non-infectious inflammation. This may result in osteolysis, decreased bone density, foot deformity and fractures. According to literature, DNOAP incidence varied from 0.08% to 18% among all patients with diabetes [2].

DNOAP is known, like other microvascular and macrovascular complications, to reduce significantly the quality of life of patients and increase mortality [3].

With improper treatment, DNOAP leads to the formation of gross deformities of feet, which results in the loss of the extremity supporting ability. Also, prolapse bones can result in the formation of ulcerative defects, which significantly increases the risk of extremity amputations. Diabetic polyneuropathy (autonomic and sensorimotor), trauma, foot surgery and calcium-phosphorus metabolism disorders are currently recognised as the main reasons for DNOAP formation.

Most often, DNOAP develops unilaterally. The relation with a previous injury and severe prolonged hyperglycaemia is not always clearly visible.

There are works demonstrating the role of various pro-­inflammatory factors in the formation of DNOAP, but there have not yet been studies published that evaluate comprehensively the role of molecular, hormonal and metabolic disorders in this process.

According to the literature, with various types of diabetes, some features in bone remodelling are noted. In patients with type 1 diabetes (DM1), low bone mineral density is registered as well as six to seven times higher risk of fractures than in individuals with normal carbohydrate metabolism. In patients with type 2 diabetes (DM2), normal or high bone mineral density is registered, and the risk of fractures is three times higher than in the control group [4]. Despite the similarity of chronic hyperglycaemia, DM1 and DM2 have different pathophysiological aspects that can affect bone metabolism in various ways. In both cases, the basic mechanisms of reduced bone strength are not entirely understood [5].

According to histological studies of biopsy specimens of the foot bones of DNOAP patients, a significant increase in the number of osteoclast activity was revealed, as well as their high sensitivity to pro-inflammatory cytokines, such as interleukin-1 and interleukin-6, as well as tumour necrosis factor (TNF)-alpha [6].

Osteoprotegerin (OPG) and receptor activator of ­nuclear factor kappa-B ligand (RANKL) are currently considered as DNOAP markers, which are members of the ligand and TNF receptor families. The RANKL/OPG pathway is associated with an increase in the number of diseases affecting bones, and several review articles have hypothesised its possible role in Charcot foot [7, 8].

Thus, it is of interest to evaluate markers of bone metabolism (both bone formation and bone resorption) in patients with DNOAP and without DM complication, to understand better the pathophysiological mechanisms of the development of this form of DFD.

Osteocalcin is a late marker of osteoblast activity. The ­osteocalcin gene is regulated at the transcription level using 1.25-dihydroxyvitamin D, as well as vitamin K, which leads to an increase in the sensitivity of bone cells to calcium and hydroxyapatite [9].

Type 1 collagen produced by osteoblasts is secreted as a collagen precursor molecule into the extracellular matrix, where it is cleaved into procollagen type I carboxyterminal propeptide (PICP) and procollagen type I N (amino) terminal propeptide with their subsequent release into the bloodstream.

These markers are related to bone formation ­indices. Quantitative analysis of PICP enables estimation of the amount of newly synthesised type 1 collagen [10].

Beta-CrossLaps is an indicator of the severity of bone resorption and a marker for the destruction of type 1 ­collagen [11].

Sclerostin represents a molecule that slows down bone formation and promotes osteoblast apoptosis. According to the literature, in DM2 patients, there is an increase in sclerostin, which is proposed for use as a predictor of femoral neck fractures in post-menopausal women [12, 13].

Another marker of collagen degradation is the matrix metalloproteinase (MMP)-1 enzyme, which destroys this molecule.

Some prospective studies have shown that in patients with non-consolidating fractures, significantly higher concentrations of pro-MMP-1 and MMP-8 in the blood serum were determined compared with the group of patients with normal consolidation of fractures [14]. Given the increase in the incidence of DM and, as a consequence, DNOAP, the search for new markers of bone metabolism in patients of this category is extremely relevant and will enable to perform a deeper analysis of pathophysiological mechanisms of DNOAP occurrence and subsequently reduce the risk of developing this complication of DM and subsequent ­amputations.

AIM

The work aimed to evaluate the hormonal and metabolic markers of bone formation and resorption in patients with DNOAP and DM without this complication.

METHODS

Study Design

The work had the design of an observational single-­centre cross-sectional controlled study.

Inclusion Criteria

1.DM2 with the use of insulin therapy

2.Age under 65 years

3. DM2 duration of less than 30 years

4.DNOAP (subacute or chronic stage)

5.Lack of osteoporosis (T-test below –2.5)

Exclusion Criteria

1.DM1

2.DM2 without the use of insulin therapy

3.Hyperparathyroidism

4.Osteoporosis (T-test above –2.5)

5.Diabetic nephropathy and chronic kidney disease stage 4–5

6.An acute stage of DNOAP (based on the history, clinical presentation and results of radiography and magnetic resonance imaging (MRI))

7.The presence of DFD, injuries and surgeries on the foot in the anamnesis

8.Hypercalcaemia

9.Hyperparathyroidism (primary and secondary)

10.Therapy with drugs affecting bone mass (­anabolic ­steroids, bisphosphonates, denosumab and ­teriparatide).

Terms and Conditions

The study was conducted in a multi-field specialised endocrinological hospital (Diabetic Foot ­Department of the ­Scientific Research Centre for Endocrinology of the Russian Ministry of Health).

The study of blood samples was performed by a medical laboratory scientist at the clinical diagnostic laboratory of the same institution.

X-ray, MRI research and bone mineral density study were performed at the Department of Radiation Diagnostics, Research Centre for Endocrinology of the Russian Ministry of Health.

Duration of the Study

The study was conducted from January 2018 to June 2019 and included enrolment of patients, the formation of study groups and laboratory testing of blood samples.

Description of Medical Intervention

Patients underwent fasting blood sampling in Vacutainer SST II Advance and K2E (ethylenediaminetetraacetic acid) tubes, followed by centrifugation to obtain serum and ­plasma.

All patients underwent peripheral sensitivity testing. ­Vibration sensitivity was measured using a graduated tuning fork (Kircher & Wilhelm, Germany) at standard points (medial ankle and the pollex base). Tactile sensitivity was determined using a 10-g monofilament (North Coast Medical, Inc., USA) at standard points. Temperature sensitivity was evaluated using a standard type-therm cylinder (Neue Medizintechnic GmbH, Germany).

The patients underwent standard X-ray osteodensitometry (Lunar GE iDEXA, Prodigy, USA) with measurement of bone mineral density, T-critera and Z-criteria in the lumbar vertebrae and proximal femur.

If it was difficult to determine the DNOAP stage based on the clinical and radiological findings, MR imaging was performed. MRI was performed on a GE Optima MR450w 1.5T tomograph (USA) using a specialised 8-channel GEM coil for studies of the shin and foot joints. The DNOAP stage was determined according to the existing classification of Chantelau and Grutzner, as well as the diagnostic criteria presented in the current algorithms of specialised medical care for diabetic patients [15].

The blood flow through the dorsalis pedis, anterior tibial and posterior tibial arteries, as well as the state of the venous outflow, was assessed by duplex scanning on a ­Voluson 730 apparatus (General Electric, USA) in the Diabetic Foot Department of the Scientific Research Centre for Endocrinology.

Primary Study Outcome

The aspects of bone metabolism in DNOAP patients were studied based on laboratory determination of indicators such as 1.25-OH vitamin D, sclerostin, pro-MMP-1, PICP and type 1 collagen.

Additional Study Outcomes

Bone resorption markers studied in routine clinical practice were additionally evaluated, namely, osteocalcin, parathyroid hormone, total calcium and ionised calcium, C-terminal telopeptide, 25-OH vitamin D, bone ­mineral density determination by X-ray densitometry and MRI of the foot. To assess the condition of arterial blood flow, all patients underwent duplex ultrasound (DU) of lower ­extremity arteries.

Subgroup Analysis

According to the study design, two groups were formed. Group 1 consisted of patients with DM2 and DNOAP (subacute or chronic stage).

Group 2 (comparison group) included patients with DM2 without DNOAP or DFD in the history.

Patients with acute stage of Charcot foot were not included in the group since the study aimed to identify possible predictors of DNOAP development. According to previous studies, the most pronounced increase in bone resorption markers is found in the acute stage of complications, which can lead to false or ambiguous data.

The groups were formed by DM2 patients (due to the aspects of bone metabolism and increased osteoclast activity in DM1 patients [7]).

Methods of Outcome Registration

All laboratory studies of blood serum samples were conducted in the clinical and diagnostic laboratory of the Scientific Research Centre for Endocrinology of the Russian Ministry of Health. Routine determination of bone resorption markers was performed according to standard methods.

Radioimmunoassay was used to study 1.25-dihydroxyvitamin D and pro-MMP-1, and sclerostin levels were determined using the Sandwich enzyme-linked immunosorbent assay (ELISA) method. PICP and type I collagen were investigated using ELISA.

Ethical Examination

The local ethics committee under the Scientific ­Research Centre for Endocrinology of the Ministry of Health of the Russian Federation decided to approve the possibility of conducting this research work, extract from protocol no. 2 dated 31 January 2018. All the patients signed informed consent to participate in the study.

STATISTICAL ANALYSIS

Principles for Calculating Sample Size

The sample size was not preliminary calculated.

Methods of Statistical Data Analysis

For the study result analysis, the Statistica 7.0 program (TIBCO Software Inc., USA) was used. The description of quantitative data is presented in the form of a median and quartiles (percentiles 25 and 75, Q25–75). To describe the qualitative data, absolute (n) and relative values (%) were calculated. The type of distributions was analysed using the Shapiro–Wilk test. The independent groups were compared by quantitative characteristics using the non-parametric method using the Mann–Whitney U-test. The differences were considered statistically significant at p < 0.05.

RESULTS

Objects (Participants) of the Study

The study included 70 patients. Group 1 consisted of patients with DM and DNOAP (27 patients), and Group 2 (control) included patients with DM2 without DNOAP (43 ­patients).

The groups under study were comparable in terms of age, gender characteristics and diabetes duration and its control.

Table 1 presents the clinical and laboratory characteristics of the patients examined.

Table 1. Clinical and laboratory characteristics of the patients examined (n = 70).

Notes: The groups were comparable in the characteristics listed (p > 0.05).

DM, diabetes mellitus; CKD, chronic kidney disease; HbA1c, glycated haemoglobin; LDLC, low-density lipoprotein cholesterol; GFR, glomerular filtration rate; ESR, erythrocyte sedimentation rate; MDRD, Modification of Diet in Renal Disease.

Primary Study Outcomes

The absence of changes in the values of parathyroid hormone and calcium confirmed the absence of hyperparathyroidism and other disorders of calcium-phosphorus metabolism in the patients studied.

During the study, no significant differences in the values of sclerostin, pro-MMP-1, 25-OH vitamin D, type I collagen and osteocalcin were obtained.

However, a significant difference in the indices of 1.25-OH vitamin D is noteworthy, as in DNOAP patients. The level of this vitamin D metabolite was lower than in patients of the comparison group (p < 0.05). Beta-CrossLaps and PICP markers were statistically significantly higher than in the comparison group.

The main results are presented in Table 2 and Figures 1–3.

Fig. 1. Results of the study of 1.25-dihydroxyvitamin D

Fig. 2. Results of the study of beta-CrossLaps in the study groups

Fig. 3. Results of the PICP study in Group 1 and the comparison group

Table 2. Results of laboratory studies of bone metabolism markers in the studied groups.

Me, median; pro-MMP-1, promatrix metalloproteinase-1; PICP, procollagen type I carboxyterminal propeptide.

Additional Study Results

During the study, most patients of both groups had a pronounced deficiency of 25-OH vitamin D, which required the administration of vitamin D preparations at the recommended doses.

According to X-ray densitometry, there was no decrease in bone density in 49% of the examined patients of both groups; osteopenia was recorded in 35%, and pronounced osteoporosis of different localisation was noted in 9.6% of cases. In 5.4% of patients, this study was not conducted for technical reasons.

No data were obtained for the presence of haemodynamically significant stenosis of the arteries of the lower extremities, according to DU data of the arteries of the lower extremities.

Adverse Events

During the study, no adverse events were recorded.

DISCUSSION

Summary of the Main Outcome of the Study

The examination revealed that the groups of patients did not differ significantly in age, duration of diabetes and the degree of its control, and there were no significant differences in lipid values. In addition, the presence of anaemia, renal failure signs (glomerular filtration rate corresponded to chronic kidney disease stage 1–3B), impaired calcium and phosphorus metabolism, and an acute inflammatory reaction and hypoproteinemia were ruled out in all patients included in the study. The increase in ESR was non-specific.

Discussion of the Main Outcome of the Study

The examination revealed the absence of significant differences in the values of sclerostin, pro-MMP-1 and type I collagen, as well as the values of parathyroid hormone and osteocalcin.

Interestingly, the study recorded a significant difference in the indices of 1.25-OH vitamin D in the main and control groups (the level of this vitamin D metabolite was lower in patients of the comparison group than that of the DNOAP group (p < 0.05)).

According to the literature, this vitamin D isoform can reduce the expression of fibroblast growth factor-23 (FGF23), which in turn plays a key role in the regulation of elements such as phosphates. In addition, FGF23 is the main regulator of bone homeostasis and mineralisation, as it is an inhibitor of mineralisation; however, its mechanism of action is unknown yet [16].

Thus, 1.25-dihydroxyvitamin D and FGF23 are antagonists, which suggest that in patients with Charcot foot, bone mineralisation processes proceed more slowly compared with the control group.

In addition, we studied the indicators of collagen, its metabolites and enzymes that destroy this protein.

Collagen type 1 and PICP are markers that relate to bone formation indices.

During the study, an increase in PICP was registered in the DNOAP group in the absence of differences with the control group in terms of the level of collagen type 1.

These results enable the increased collagen synthesis in patients with Charcot foot; however, the absence of differences in the values of collagen requires further study.

The results showed an increase in this marker compared with the control group, which suggests an increase in collagen destruction in DNOAP patients.

Based on the data obtained, it can be assumed that in patients with Charcot foot, the synthesis process and destruction of collagen are disturbed.

According to the results obtained in this study, there were no differences between groups in sclerostin values; also, this indicator in both groups was within the reference values. This is not consistent with the literature that describes an increase in the level of sclerostin in DM2 ­patients [12].

Another marker of collagen degradation is the MMP-1 enzyme, which destroys this molecule [14].

In the current study, pro-MMP-1 values were within reference values and did not differ between the groups.

Study Limitations

Because of the relatively low incidence of DNOAP, the comparison group was slightly larger than the main group. Patients receiving only insulin therapy in various modes were selected because the difference in hypoglycemic therapy could affect the results obtained. It is planned to continue this study in the future with a larger sample and subject to more stringent inclusion criteria.

CONCLUSION

During the study, a pronounced deficiency of vitamin D was recorded in all the patients examined. Given the increased risk of osteoporosis and fractures, it is extremely important to control this indicator in patients with DM.

A disorder of the synthesis and degradation of collagen in patients with Charcot foot (increased beta-CrossLaps and PICP) was revealed, which, perhaps, enables to clarify the mechanisms of DNOAP development; however, studies with a larger sample are required to understand more fully the pathogenesis of DNOAP, as well as population studies with the identification of genetic markers for the development of this complication in DM2 patients.

ADDITIONAL INFORMATION.

Source of Financing. The study was financially supported by research work within the state task No. АААА-А18-118051590055-8 ‘Metabolic and genetic markers of repair of soft tissues and bone structures of the lower extremities in patients with DM, their role in prediction, personalized therapy and prevention diabetic foot syndrome’.

Conflict of Interest. The authors declare no obvious or potential conflicts of interest related to the publication of this article.

Contribution of Authors. E.L. Zaitseva collected the clinical material, analysed and wrote the article. A.Yu. Tokmakova developed the study design and wrote the article. V.M. Zhilyaev collected the clinical material and performed statistical data processing. N.M. Malysheva conducted laboratory research. N.I. Sazonova performed X-ray densitometry. G.R. Galstyan developed the study design and wrote the article. A.V. Vorontsov performed the MRI scanning.

Acknowledgments. The authors express gratitude to the staff of the Department of Diabetic Foot, Department of Diabetic Kidney Disease and Post-transplant Rehabilitation, Department of Diabetes Therapy, Clinical and Diagnostic Laboratory of the Scientific Research Centre for Endocrinology and to the resident doctors E.S. Gracheva and D.A. Gadzhalieva for their help in preparing this publication.