Infected bone resection plus adjuvant antibiotic-impregnated calcium sulfate versus infected bone resection alone in the treatment of diabetic forefoot osteomyelitis

Diabetic foot osteomyelitis (DFO) is a common complication of patients with diabetic foot infections. It was reported that nearly 20–60% of patients with diabetic foot might suffer from DFO [1–3]. However, DFO is a difficult-to-treat infection disease, as the treatment of DFO might include the management of chronic ulcers, necrotic soft tissues, gangrenes and of course, the infected bones. Although various treatment methods have been adopted, unfortunately, suffered from DFO still means a high rate of amputation and mortality [4].

Currently, the mainstay treatments for DFO consist of antimicrobial therapy alone or in combination with surgical intervention depending on the severity of disease [5]. In the case of mild infection, antibiotic administration alone for several weeks received promising results [5–7]. However, damaged peripheral vessels condition may make it difficult for parental antimicrobial therapy to achieve satisfying local effects due to insufficient penetration [8]. Moreover, the optimal duration of antibiotic therapy is still controversial with formal studies reported the antibiotic therapy duration varied from 6 weeks to more than 40 weeks [6, 9, 10]. The prolonged duration of antibiotic therapy is limited by the advent of antibiotic-resistant bacteria and potential drug-induced gastrointestinal, liver and kidney injury [11]. For patients with pus, sequestrum, gangrene or antibiotic-resistant bacterial infection [5, 12], surgery is the cornerstone to remove dead tissues and eliminate the infections. For the latest decades, necrotic bones and tissues resection instead of amputation has been widely accepted in treating with DFO, as it removes the infected bones while preserves the healthy bones to minimize the biomechanical changes. However, when carrying out infected bone resection, the completely negative resection margin is relatively difficult to be identified, which may lead to the residue of pathogens. Furthermore, the removal of infected bone sometimes causes the formation of dead space, which will be filled with hematoma soon and provide an environment for the growth of bacteria. Muscle flap used to be a method to obliterate the defects caused by debridement, but it is limited when managing with deeper defects and may disturb the bone healing [13].

Local antibiotic delivery system has been widely used as an adjuvant after the surgical treatment of osteomyelitis and achieved good results [14, 15]. Compared with intravenous route, the local antibiotic delivery has the advantages of more accurate positioning, higher local concentration, less side effects and longer duration. At the same time, it works as a bone substitute which fills the dead space caused by bone resection and reduces the incidence of reinfection. Polymethyl-methacrylate cement (PMMA) has acted as an antibiotic carrier to fill the defects caused by debridement since Buchholz successfully applied it in joint prosthesis. However, its non-biodegradable characteristics, the high temperature it produces and a second surgery for removal all limit its application on osteomyelitis especially in DFO [16].

Nowadays, biodegradable antibiotic-impregnated materials such as calcium sulfate, calcium phosphate, bioactive glasses and collagen are gradually applied as a substitute for PMMA in the management of osteomyelitis. All materials mentioned above have advantages of biocompatibility and drug compatibility. Among those substitutes, calcium sulfate is most frequently used materials since it enjoyed some eminent advantages. To begin with, the elution characteristic of loaded antibiotics are now clearly illustrated, an initial burst of antibiotics releasing in the first 24 h or 48 h produces antibiotics levels about hundreds to thousands times higher than minimum inhibitory concentration (MIC), then the calcium sulfate releases all antibiotics it loaded gradually at relative slow pace and complete resorption in several weeks. This ideal elution duration makes it more available than collagen (too short) and calcium phosphate (too long) to be a bone graft. Furthermore, it hardly produces the foreign body reaction and helps the formation of new bone. The complications of calcium sulfate are also acceptable, including postoperative drainage and transient hypercalcemia [17].

Previous studies had reported that satisfying outcomes could be received when using antibiotic-impregnated calcium sulfate as an adjuvant after surgical treatment of DFO. However, few comparative studies had been carried out to confirm those results. This retrospective study was designed to observe the outcomes of surgical treatment combined with adjuvant antibiotic-impregnated calcium sulfate versus surgical treatment alone in the treatment of DFO and to compare the differences of healing rate, time to healing, osteomyelitis recurrence rate and amputation rate between two groups.

From 2015 to 2017, 46 patients (48 limbs) met the criteria were included in the study. The locations of DFO were 16 in phalanges (7 in the CS group), 13 in metatarsal bones (5 in the CS group) and 19 in both phalanges and metatarsal bones (8 in the CS group).

Twenty limbs (41.7%) as the CS group were applied with antibiotic calcium sulfate after infected bone resection compared to 28 limbs (58.3%) as the control group underwent infected bone resection alone. The follow-up outcomes of two groups were presented in Table 3. During follow up, 2 patients in the CS group died of cardiovascular disease after surgery, but the wounds had healed before death and did not recur within 1 year after surgery. 2 patients applied antibiotic-impregnated CS are presented in Figs. 1 and 2.

	The CS group	The control group	P value
Mean hospital stay (days)	24.8 (7–59)	28.3 (9–56)	0.341
Mean follow-up duration(months)	17.6 (12–38)	20.1 (12–30)	0.120
Preoperative antibiotics duration (days)	13.5 (4–30)	15.6 (6–28)	0.142
Postoperative IV antibiotics duration (days)	8.8 (3–14)	9.8 (2–14)	0.569
Postoperative healing rate	90.0% (18/20)	78.6% (22/28)	0.513
Mean healing duration (weeks)	13.3 (5–30)	11.2 (2–26)	0.132
Recurrence rate	0.0% (0/18)	36.4% (8/22)	0.014
Postoperative leakage rate	30% (6/20)	–	–
Postoperative leakage duration (weeks)	8.5 (4–13)	–	–
Amputation rate	0.0% (0/20)	7.1% (2/28)	0.153

According to the severity of infections and local blood supply, diabetic foot osteomyelitis can be managed with conservative treatment or surgery. When accompanied with pus, substantial bone necrosis, gangrene, recurrent ulcer or antibiotic-resistant bacteria infection, surgery is recommended to remove necrosis tissues, reduce the antibiotic therapy duration and correct bone deformity to promote healing [23]. However, even surgical treatment also has its own limitations. Due to the bone removal, biomechanics in foot is inevitably changed, which may lead to the ulcer formation in a new position. Thus, postoperative offloading with customized insoles and shoes was essential to promote healing and prevent infection recurrence. Furthermore, during infected bone resection, the clear margin of bone and soft tissue is fairly difficult to identify. The exact extent for infected bone resection is largely depended on the intra-operative judgement of the surgeons and thus sometimes lead to the non-healing or recurrence of osteomyelitis. To eradicate the residual infection, local antibiotic-impregnated calcium sulfate was applied in our study as the high local antibiotic level it produced.

In total, 90.0% (18/20) limbs in CS group and 78.6% (22/28) limbs in control group healed after the first operation. The higher healing rate in CS group is in accordance with the retrospective study of Rajesh M. Jogia et al., who reported all DFO limbs (20 patients) achieved healing after surgical debridement combining with antibiotic-impregnated calcium sulfate beads application [24]. Similarly, Noman Shakeel Niazia et al. retrospectively studied 70 patients with DFO who received debridement and adjunctive antibiotic-loaded calcium sulfate treatment. During an average follow-up of 10 months, infection eradication and wound healing were achieved in 90 and 81% limbs respectively [25]. Proper explanation of the high healing rate is that much higher antibiotic concentration reached topically can eradicate more residual organisms with the resorption of the calcium sulfate. Former studies have shown antibiotic levels surpass 200 times the MIC [14] for organisms over days and still retain antimicrobial effect after 6 weeks to 3 months [26, 27], which is sufficient to penetrate the biofilm and eradicate the residual organisms. Unfortunately, a significant difference in the healing rate between the two groups was not found through statistical analysis (90% in CS group versus 78.6% in control group, P>0.05). Combining with the previously-reported efficacy of antibiotic-impregnated calcium sulfate in eradicating infections and well-controlled variables (similar blood supply condition, appropriate wound care and postoperative offloading between two groups) in our study, we deem that the small group of patients included may cause the absence of significant differences in the healing rate. This explains why the healing rates of two groups are not statistically significant. With regard to the recurrence rate, such high topical antibiotic concentration and long therapeutic duration explained that the much lower recurrence rate of antibiotic-impregnated calcium sulfate group than the control group. Rajesh M. Jogia et al. reported no recurrence in 20 patients who received surgical debridement combined with gentamicin or vancomycin-impregnated calcium sulfate beads within the follow-up of 12 months after surgical intervention, which is similar with the result in our study [26, 27].

However, our study failed to provide evidence that antibiotic-impregnated calcium sulfate will shorten the wound healing duration, which is different from the similar former study. Martin Varga et al. reported that application of gentamicin-impregnated collagen sponge shortened nearly 2 weeks of wound healing duration after minor amputation [28]. Fabian G. Krause et al. found that time to a dry wound was 5.2 weeks in the antibiotic-impregnated beads group and 7.0 weeks in the control group, even if no significance in two groups [29]. In this study, however, the mean duration of healing in CS group was 2.1 weeks longer than the mean healing duration in control group. Actually, we hypothesize that the prolonged leakage in CS group may interfere with the wound healing, but no previous study was found to support this hypothesis.

Prolonged postoperative leakage was found to be the most common complication in calcium sulfate treated patients. 30.0% limbs (6/20) suffered from prolonged postoperative leakage in CS group with a mean duration of 8.5 weeks. With regular dressing, all wounds achieved healing eventually. The drainage rate is similar with the former studies about chronic osteomyelitis and varies from 4.2 to 32% [13, 17, 30]. Other studies have reported the prolonged postoperative leakage in treating chronic osteomyelitis after using antibiotic-impregnated calcium sulfate, but achieving healing with appropriate wound care [13, 31]. In fact, prolonged postoperative leakage itself is neither an indication for a second surgery nor did it relate to reinfection of the wound [31]. Regularly dressing in outpatient is enough in dealing with the postoperative leakage. Vacuum-assisted Closure (VAC) was not used because we deemed that it might lower the topical antibiotic level when pumping the drainage. During surgical treatment, good soft tissue coverage and primary closure are essential methods in the prevention of postoperative leakage.

Severe side effects were not found excluding postoperative leakage in CS group. The explanation is that the dose of vancomycin or gentamicin administered locally was less than 0.5 g (vancomycin) or 100 mg (gentamicin), which means the systemic concentration reached was much lower than the same dose applied intravenously because of the slow releasing of antibiotics with the degradation of calcium sulfate. Unfortunately, systemic drug concentration was not obtained to confirm our hypothesis. In their study, Zhang et al. measured the blood vancomycin levels in 24 osteomyelitis patients locally applied with vancomycin-impregnated calcium sulfate beads (range from 1.5 ml to 5 ml with a ratio of 1 g vancomycin:5 ml calcium sulfate). The results showed that the mean blood vancomycin level was still within a safe range for application [32]. P. Wahl et al. found that even 6 g vancomycin was applied locally, the systemic concentration remained within a safe range and local concentration was still below the reported cell toxicity thresholds [27].

To our knowledge, our study is the first retrospective comparative study comparing the outcomes of infected bone resection combined with adjunctive antibiotic-impregnated calcium sulfate versus infected bone resection in the treatment of DFO. The limitations of our study are mainly in two aspects. Firstly, the follow-up duration in two groups may not be enough to show the outcomes of all patients, which may influence the healing rate, recurrence rate and amputation rate in our study. Furthermore, it is a retrospective study for forefoot DFO with a small group of patients, the additional studies are necessary to confirm our findings.

Application of the antibiotic-impregnated calcium sulfate as an adjuvant can be regarded as efficacious for preventing the recurrence of forefoot DFO. However, evidence is not found that the use of antibiotic-impregnated calcium sulfate improves the healing rate, shorten the healing duration or reduce the amputation rate. Prolonged postoperative leakage as a common complication can be dealt with regular dressing.