Complications associated with the approach to metatarsal head resection in diabetic foot osteomyelitis

Abstract

The aim of this study was to evaluate the recovery time and the development of complications in the dorsal and plantar approach to metatarsal head resections (MHR) in patients with diabetic foot ulcers complicated by osteomyelitis. A retrospective study was carried out involving 108 patients who underwent MHRs for the treatment of diabetic foot osteomyelitis. Two cohorts were defined: dorsal approach with incision closed with sutures and plantar approach with ulcer healed using conservative treatment. The main outcomes were the weeks until healing and complications related to the approaches. Fifty‐three patients (49.1%) underwent a plantar approach and 55 (50.9%) a dorsal approach. Both approaches rendered similar healing times. However, the patients undergoing a dorsal approach developed more post‐surgical complications than patients treated through a plantar approach. The dorsal approach intervention was performed on smaller and shallower ulcers; however, more complications developed at follow up using this approach than through a plantar approach for MHR complicated with osteomyelitis.

1. INTRODUCTION

Diabetic foot ulcers (DFUs) are one of the major complications associated with diabetes mellitus (DM), and diabetic foot osteomyelitis (DFO)1 is a frequent and severe complication of DFUs. DFO is one of the main causes of limb amputation,2 and it is well recognised to complicate a significant proportion of diabetic infections.3

The forefoot plantar surface is a common location of DFUs, where there is a protrusion of the metatarsal heads. Two techniques have been described to address metatarsal head resection (MHR). In 1990, Griffiths and Wieman4 described MHR in diabetic patients through a dorsal approach. The plantar approach is used as part of the surgical routine for metatarsal head osteomyelitis.5 In addition, some studies6, 7, 8 have suggested that MHR should also be used to decrease the peak plantar pressure of the forefoot to avoid possible recurrence and to allow faster healing of the plantar foot ulcer. The complications associated with the procedure may be promoted by technique‐related factors and patient‐related factors.9 The most common complication associated with this procedure was a pressure transfer lesion. This condition will move the peak pressure to the adjacent metatarsal joints, and consequently, a new ulcer may develop.10, 11, 12

The possibility of complications related with the approach to the MHR has not been adequately investigated. Some procedural options are available for the treatment of DFUs, but the selection of the appropriate procedure for these patients (either osteotomy or joint resection) should be determined based on the consideration of several underlying factors, for instance: age, vascular status, the presence of metabolic disturbances, patient compliance, the presence of soft tissue infection, bone quality, and the patient's social support after surgery.8 Nevertheless, until now, guidelines for the decision on a dorsal or plantar approach to MHR have not been described, and surgeons commonly decide on the approach according to their own experience. There is no clear indication regarding the most suitable approach to MHR in the literature or knowledge of complications related to the approach.

The aim of this study was to evaluate the healing time and the development of complications for the dorsal and plantar approach to MHR in patients with DFUs complicated by DFO.

2. MATERIALS AND METHODS

Between January 2013 and September 2017, a retrospective study was carried out involving 108 patients who underwent MHR (dorsal or plantar approach) for the treatment of DFO at a specialised diabetic foot unit.

The inclusion criteria were as follows: patients with DM aged >18 years, the presence of DFUs with osteomyelitis classified as a curative or emergent surgical indication13, and patients who had agreed to be included in the study by means of written consent.

The exclusion criteria were as follows: patients suffering critical limb ischaemia, patients who had concomitantly undergone related toe amputation or transmetatarsal amputation before or during metatarsal head resection, pregnant or lactating women, and patients who did not understand the purpose of the study or refused to be included.

The diagnosis of DFO was established through a combination of the probe‐to‐bone (PTB) test and a plain X‐ray examination as previously published.14 The PTB test was performed using metal forceps (Halsted mosquito forceps), and the result was considered positive when the researcher could feel a hard or gritty surface. We considered the plain X‐ray examinations (two standard views) positive for osteomyelitis if they showed cortical disruption, periosteal elevation, a sequestrum or involucrum, or gross bone destruction.

The neurological examination was conducted using a Semmes‐Weinstein 5.07/10 g monofilament (Novalab Ibérica, Alcal. de Henares, Madrid, Spain) and Horwell's biotensiometer (Novalab Ibérica). Neuropathy was diagnosed in patients who felt nothing during one of the two tests.15 Peripheral arterial disease (PAD) was diagnosed if the patient met the following criteria: absence of both distal pulses (dorsalis pedis and posterior tibial pulse) and ankle brachial index (ABI) <0.9; in patients with ABI >1.4 (non‐compressible measurement resulting in medial arterial calcification), we considered PAD to be defined by a toe brachial index <0.7 and a transcutaneous oxygen pressure (TcPO2) (using a TCM4 transcutaneous monitor; Radiometer Medical, Brønshøj, Denmark) <30 mm Hg.16, 17 Critical limb ischaemia was diagnosed if the patient met the following criteria:

Absence of both distal pulses and ankle pressure lower than 70 mm Hg or ABI < 0.5 or a toe systolic pressure lower than 50 mm Hg.17, 18

Once the patients had agreed to enrol in the study, the patient cohorts were decided according to the surgical approach:

• The dorsal approach4 was defined as follows: the skin incision was made on the dorsal surface, extending from the base of the toe to the midpoint of the metatarsal shaft; the extensor tendon was protected by retraction, while the incision was deepened down to the bone. The dorsal incision was closed by primary closure.

• Plantar approach5 was defined as follows: the incision was made on the plantar surface, removing the ulcer by means of conic ulcerectomy, which removes both the walls and the bottom of the lesion; the MHR was performed, and the ulcer healed using conservative treatment.

The patients received surgical treatment according to the description above. Bone samples were sent for pathology and microbiology studies. All surgical procedures were performed by the same surgeon (JLM) who is specialist in diabetic foot surgery for more than 20 years. All patients received post‐surgical antibiotic treatment for at least 10 days after surgery. Patients took empirical antibiotics during the pre‐procedure period, choosing according to IDSA guidelines19 recommendations, until we had the results from bone culture.20 After we received bone culture results, we adjusted the antibiotic therapy for targeting isolated bacteria. For plantar ulcers, the local treatment was antimicrobial dressing used for 2 weeks (Actisorb Plus 25; Systagenix, Cardiff, UK). Local treatment for sutures included antiseptic and non‐adherent dressing. After that, the patients were treated according to the wound care protocol of our department. Removable cast walkers were used for offloading. Podiatrist specialists in diabetic foot evaluated patients twice per week.

An experienced podiatrist prescribed therapeutic footwear that consisted of either an extra‐depth shoe or custom insoles, according to foot biomechanical characteristics and plantar pressure measurements (Rsscan International, Olen, Belgium), after the ulcer was completely healed. Patients were subjected to periodic footwear checks and forefoot examination for new ulcers according to the recommendations of the International Working Group on the Diabetic Foot (IWGDF).21

The main outcomes were related to the weeks until healing and the complications associated with the approaches to MHR.

Ethical approval was obtained (18/227‐E), and the study was completed in accordance with the ethical standards of the Ethics Committee. The authors declare that they have conformed to the Declaration of Helsinki code of ethics.22

The data were entered and processed using SPSS version 22.0 for Macintosh (SPSS, Inc. Chicago, Illinois). Descriptive analyses were performed. For quantitative variables, we calculated means and standard deviations. For qualitative variables, we calculated frequency distributions and percentages. A Student's t test was performed on the independent samples to compare the means of the quantitative variables. The χ ² test was used to identify differences in qualitative variables. Odds ratios (OR) and the 95% confidence intervals were determined through univariate and multivariate logistic regression models. Differences <5% were assumed significant for type I errors (P < 0.05).

3. RESULTS

Of the 189 individuals identified as eligible, 108 diabetic patients with DFO who required MHR were included in the study. Ninety‐three patients were excluded because of: 28 prophylactic metatarsal head resections, 32 minor amputations, 7 transmetatarsal amputations, and 14 critical limb ischaemia cases. In 53 patients (49.1%), we performed plantar approach, and in 55 (50.9%), a dorsal approach was used. Available pathology results were collected from 65 patients (60.2%) and microbiology results from 57 (52.8%). The demographic and clinical characteristics of the ulcers in the study population are shown in Table 1. Mean time to healing was 14.1 ± 10.3 weeks.

Table 1.

Baseline clinical data of subjects

Variables (N = 108)	Plantar approach (n = 53)	Dorsal approach (n = 55)	P‐value; OR [CI]
Male/female, n (%)	42 (79.2)/11 (20.8)	51 (92.7)/4 (7.3)	0.043; 0.2 [0.1–1.0] *
Age (years), mean ± SD	63.3 ± 11.6	65.6 ± 7.8	0.231
DM (years), mean ± SD	16.4 ± 8.8	16.5 ± 9.0	0.924
DM type 1/type 2, n (%)	9 (17.0)/44 (83.0)	6 (10.9)/49 (89.1)	0.362
Retinopathy, n (%)	15 (29.4)	16 (30.2)	0.931
Nephropathy, n (%)	10 (19.6)	12 (22.6)	0.705
Neuropathy, n (%)	53 (100.0)	54 (98.2)	0.324
PAD, n (%)	25 (47.2)	20 (36.4)	0.524
Duration from ulcer (weeks), mean ± SD	14.4 ± 13.9	17.0 ± 24.0	0.492
Positive bone biopsy, n (%)	28 (73.7)	16 (59.3)	0.220
Bone culture positive, n (%)	23 (88.5)	25 (80.6)	0.650
Body mass index (kg/m2), mean ± SD	28.2 ± 5.1	28.9 ± 4.9	0.417
HbA1c (%), mean ± SD	8.1 ± 1.4	8.1 ± 1.6	0.824
Signs of infection, n (%)	41 (77.4)	36 (65.5%)	0.172
Radiological signs of DFO, n (%)	38 (76.0)	40 (76.9)	0.913
Probe to bone positive n (%)	52 (98.1)	48 (87.3)	0.032; 0.1 [0.2–1.1] *
Size of the ulcer >1 cm2/<1 cm2, n (%)	33 (62.3)/20 (37.7)	23 (41.8) / 32 (58.1)	0.034; 0.4 [0.2–0.9] *

Abbreviations: CI, confidence interval; DM, diabetes mellitus; HbA1c, glycated haemoglobin; PAD, peripheral arterial disease.

^*Differences were assumed significant at p<0.05 for a confidence interval of 95%.

We found no statistical differences in the time of healing between the plantar or dorsal approach (P = 0.818; 14.3 ± 8.3 weeks for the plantar approach vs 13.9 ± 12.1 weeks for the dorsal approach).

During follow up (mean time 77.5 ± 66.5 weeks), different complications were observed; 21 (39.6%) developed in 55 plantar approach patients (50.9%) and 34 (61.8%) in dorsal approach patients, with P = 0.021; OR 2.5 [1.1‐5.3]. The complications for both approaches are shown in Table 2.

Table 2.

Complications during follow up

Complications N = 108	Plantar approach (n = 53)	Dorsal approach (n = 55)	P‐value; OR [CI]
Post‐surgical infection, n (%)	12 (22.6)	12 (21.8)	0.918
Dehiscence, n (%)	—	18 (32.7)	0.000 *
Haematoma, n (%)	0 (0)	5 (9.1)	0.025 *
Re‐intervention, n (%)	11 (20.8)	10 (18.5)	0.771
Minor amputation, n (%)	6 (11.3)	16 (29.1)	0.022; 3.2 [1.1–9.0] *
Major amputation, n (%)	1 (1.9)	2 (3.6)	0.580
Re‐ulceration, n (%)	28 (52.8)	23 (42.6)	0.289
Recurrence, n (%)	9 (17.0)	5 (9.3)	0.236

Abbreviation: CI, confidence interval.

^*Differences were assumed significant at p<0.05 for a confidence interval of 95%.

4. DISCUSSION

Our study showed that both approaches rendered similar times of healing. However, the dorsal approach patients developed more post‐surgical complications than plantar approach patients.

García‐Morales et al23 disclose the only study that analysed the result of the primary closure in the surgical treatment of DFO compared with those healed by secondary intention. For 23 surgical procedures of MHR, patients undergoing primary closure healed faster than those healing by secondary intention (9.99 ± 8.4 vs 19.19 ± 16.9 weeks, P = 0.008). Moreover, they found no differences regarding complications between the primary and secondary closure. Nevertheless, this study described different type of surgeries and locations, mainly toes, and did not reference the approach used in the MHR.

The majority of the studies in the literature describing MHR through a plantar approach performed the closure of the plantar ulcer by sutures. Only Piaggesi et al,24 who performed closure in plantar incision, showed a mean healing time of 6.67 ± 5.56 weeks in 21 procedures. Our results showed a longer time (14.3 ± 8.3 weeks) with the plantar approach because closure by sutures following MHR was not conducted, and ulcer management included local treatment.

Several authors4, 8, 9, 25 used primary closure of a dorsal incision, leaving the ulcer open and achieving good healing in most cases, with a range between 5.3 and 12 weeks. Our results in patients who underwent dorsal approach also showed a longer time of healing (13.9 ± 12.1 weeks); this could be because our study population comprised patients with DFO, and some of them had moderate limb ischaemia, and the rest of the studies included patients with neuropathic non‐healing ulcers.

In our sample, 41% of patients had moderate limb ischaemia without statistical differences between both groups. Revascularisation is an important consideration in the management of DFUs, and it is commonly performed in patients with CLI and DFU and also in patients with less severe arteriopathy.26 In a systematic review,27 the ulcer‐healing wide rate obtained after revascularisation ranges widely from 46% to 91% at 1 year. However, in those studies, healing was often a secondary criterion, there was no report of the severity of the disease and no description of the management of the wound. For this reason, the revascularisation in our sample could be useful but probably not determinant for obtaining the outcomes.

The short‐term complication more frequently described in the populations undergoing MHR, mainly through a dorsal approach, was post‐surgical infection (soft tissue or bone). Wieman et al25 reported post‐surgical infection in 10% of cases and Armstrong et al8 18.2%. Post‐surgical infection was the short‐term complication more frequently found in our study in both cohorts (22%), without differences based on a dorsal or plantar approach. Nevertheless, we found a higher percentage of dehiscence (P = 0.00) and haematoma (P = 0.025) in the dorsal approach cohort. These complications, directly related to primary closure, could be associated with the post‐surgical infection management and cannot be developed in patients with an open plantar wound.

After MHR surgery, re‐ulceration appears to be the most prevalent medium‐ to long‐term complication, widely ranging from 10.3% to 41%10, 28, 29 according to studies based on a plantar approach and from 4.5% to 70%4, 8, 9, 25 for dorsal approach. In our study, the percent of re‐ulceration in both groups is similar (43% vs 53%), and these rates should not be underestimated. The value is closer than that obtained by Molines‐Barroso et al10 at 41% and Sanz‐Corbalan et al12 at 46.7% because the surgery procedure, follow‐up time of the sample, and the demographic and biomechanical characteristics of the patients are similar. Although patients were subjected to an intensive follow‐up programme, the results show a relatively high re‐ulceration frequency (47%). This concern must be taken into account when indicating an MHR, especially in non‐specialised centres or when there is no experience in designing individualised offloading treatments, because these numbers may increase considerably over time. At recruitment, our study population showed a high proportion of previous amputations and structured joint deformities; this leads to differences in the risk of re‐ulceration compared with a healthy population with no deformities or previous amputations.

Of the patients, 4.5% underwent minor amputation after MHR according to the study by Armstrong et al,8 while Wieman et al25 reported that 7.8% of patients ended up undergoing major amputation. We found a higher percent of minor amputations in the dorsal approach cohort (29%) during the follow up (P = 0.022), which could be a late complication when dehiscence or haematoma are not resolved. The dorsal approach could compromise toe viability because vessel damage causes haematoma, which facilities digital necrosis or infection. Minor amputation is a complication to consider when opting for MHR through the dorsal approach.

Some studies30, 31 included patients with superficial recurrent neuropathic ulcers that were often treated with metatarsal osteotomy. This procedure maintains a metatarsal parabola and allows for a relatively normal transfer of weight to the adjacent metatarsals, which prevents the development of transfer ulcers.

When the metatarsal head is involved in an infection and excision is required, MHR is necessary, and a greater transfer of weight‐bearing forces occurs to the adjacent metatarsal heads. This is why MHR is not routinely performed for superficial ulcers.32

Thus, it is important to correctly select the technique and previously evaluate the ulcer to establish the best approach in metatarsal head surgery.

We found that the previous size of the ulcer influenced the type of the approach, whereby for ulcers smaller than 1 cm², the dorsal approach was advised, while for ulcers larger 1 cm², the surgical approach was preferred (P = 0.034). In addition, the intervention of ulcers based on a plantar approach presented more evident bone exposure than that appearing in dorsal approach interventions with a previous positive PTB test (P = 0.032).

Ulcers were deeper and larger in cases of plantar approach than in cases of dorsal approach. MHR is more feasible, and ulcerectomy is advisable when the plantar approach is used. Furthermore, keeping the ulcer open benefits wound drainage and reduces post‐surgical complications. Ulcers smaller than 1 cm² allow easier intervention through dorsal approach in cases of minor plantar trauma and surgical injury.

The retrospective analysis of the study shows as limitations the selection of the sample and the fact that the procedure could not be evaluated before surgery. As we have not included the antibiotic used pre‐procedure and post‐procedure, we could not analyse the influence, or lack thereof, on the post‐surgical complications.

The main strength is that it is the first study that compares two types of approaches to MHR in patients with DFO and describes the associated complications. It would help surgeons decide on the approach in each case.

In conclusion, the healing time of the post‐surgical ulcer complicated with osteomyelitis following MHR is similar in cases of the dorsal or plantar approach. Nevertheless, the dorsal approach was adopted for smaller and shallower ulcers, but more complications developed in follow up than plantar approach cases in MHR procedures complicated with osteomyelitis.

CONFLICTS OF INTEREST

The authors have no conflicts of interest to report.

Tardáguila‐García A, Sanz‐Corbalán I, Molines‐Barroso RJ, Álvaro‐Afonso FJ, García‐Álvarez Y, Lázaro‐Martínez JL. Complications associated with the approach to metatarsal head resection in diabetic foot osteomyelitis. Int Wound J. 2019;16:467–472. 10.1111/iwj.13055