Charcot Foot Offloading in Stage 0 Is Associated With Shorter Total Contact Cast Treatment and Lower Risk of Recurrence and Reconstructive Surgery: A Pilot Study

Charcot neuropathic osteoarthropathy (i.e., Charcot foot [CF]) is an incompletely understood inflammatory condition affecting bones, joints, and surrounding structures in individuals with peripheral neuropathy, almost always in the foot and ankle, and is in the Western world predominantly caused by diabetes (1–3). In its early stage, CF is challenging to diagnose and treat and requires prompt and prolonged offloading of the affected foot. When not discovered and treated in a timely manner, CF can lead to fractures, joint dislocation, and severe foot deformity, increasing the risk of ulceration, reconstructive surgery, and amputation (2,4). Reports describing patient characteristics, treatment duration, and outcomes have been published (5–9). The International Working Group on the Diabetic Foot published guidelines for diagnosing and treating CF in 2023, emphasizing early recognition and consideration of CF when assessing an inflamed foot in individuals with diabetes or neuropathy (4). The current consensus is to offload the affected foot in a nonremovable knee-high device, preferably using a total contact cast (TCC), until signs of inflammation diminish and normal skin temperature is confirmed, and then gradually increase weight bearing using orthotic walkers, boots, and therapeutic footwear. Offloading should be initiated as soon as the diagnosis is considered, before it is radiologically confirmed; treating the condition early has been associated with a reduced risk of complications and deformity (2,4,10–12).

Several classifications of CF exist. Classifications of the anatomical distribution of affected joints have been proposed, of which the Sanders-Frykberg classification (13) is frequently used. Eichenholtz (14) proposed a classification dividing the disease course into three stages based on radiological appearance using conventional X-ray imaging. The original Eichenholtz classification did not encompass the early inflammatory phase, later referred to as stage 0 in a modified classification, that precedes fractures and dislocations without showing signs on conventional X-ray imaging (15). Accordingly, diagnosing stage 0 requires other imaging techniques, such as technetium-based bone scintigraphy, positron emission tomography, or MRI (16). A classification proposed by Chantelau and Grützner (17), combining clinical signs and radiological findings from X-ray and MRI examinations (Supplementary Table 1), dichotomizes the assessed CF into either stage 0, in which inflammatory signs are seen, or stage 1, in which macrofractures, joint dislocation, and deformity are evident.

In our long-term retrospective study, our primary aims were to investigate whether stage 0 compared with stage 1 at the time of CF diagnosis predicts duration of TCC treatment, future need for orthopedic surgery, or CF recurrence. We also investigated if factors beyond CF stage predict recurrence.

All adult patients diagnosed with CF between 2006 and 2019 (ICD-10 codes M90.8, M90.8H, M14.2, and M14.6) at the Department of Endocrinology, Skåne University Hospital, were considered for inclusion. Patients with inactive CF at the time of diagnosis and patients without diabetes were excluded. The date when active CF was first suspected and managed at our department was registered as the first visit.

Age, sex, diabetes type and duration, HbA_1c level, estimated glomerular filtration rate (eGFR), and presence of ipsilateral diabetic foot ulcer, osteomyelitis, or contralateral foot disease as well as a medical history of diabetic foot ulcer, amputation of the lower limb, ischemic heart disease, hypertension, peripheral artery disease (defined as previous revascularization or ankle-brachial pressure index >1.3 or <0.9 in either leg), dialysis, kidney transplantation, or osteoporosis at time of CF diagnosis were registered. eGFR was calculated according to the revised Lund-Malmö formula based on creatinine level, and renal impairment was defined as eGFR <60 mL/min/1.73 m² (18). Osteoporosis was defined as a T score of −2.5 or lower on dual-energy X-ray absorptiometry. Biochemistry analyzed as close as possible to inclusion, not exceeding 1 year, was used. Radiology reports for conventional X-ray and MRI examinations were used to classify CF as stage 0 or 1 in accordance with Chantelau and Grützner (17). Thus, cases of normal X-ray and MRI with bone marrow edema and without macrofractures or joint dislocation were classified as stage 0, and cases with presence of macrofractures, joint dislocation, or skeletal deformity were classified as stage 1. Anatomical engagement was classified according to Sanders-Frykberg (13).

Throughout the study period, management of active CF involved initial offloading with TCC until signs of inflammation diminished, confirmed by a local temperature difference <2°C, compared with the corresponding area on the contralateral foot. All dates of changes in offloading device (TCC, orthotic walker, boot, and shoe) and recurrences of active CF were recorded until 31 August 2022. Ipsilateral recurrences of active CF occurring within 1 year of the introduction of definitive therapeutic footwear (or, if unknown, discharge date) were recorded. Later ipsilateral recurrences and all contralateral active CF recurrences were registered as new events. Later ipsilateral reconstructive surgery and partial and above-ankle amputations were recorded.

Categorical data are given as percentages and continuous data as medians and interquartile ranges (IQRs). Categorical data were compared using the Fisher exact test and continuous data using the Mann-Whitney U test. Recurrences and amputations are presented as Kaplan-Meier curves, and the log-rank test was used to compare differences between patients treated with offloading in stage 0 versus 1. To reduce the risk of incorrect interpretation of stage and anatomical involvement, only the first CF event in either foot was included in the analyses. Univariable Cox regression was used to estimate hazard ratios (HRs). Variables with a P value <0.10 in univariable analyses were entered into a multivariable Cox regression model. P values <0.05 were considered statistically significant. SPSS software (version 28; IBM Corporation, Chicago, IL) was used for statistical analyses.

This study was approved by the Swedish Ethical Review Authority (Dnr 2022-05609-01; Uppsala, Sweden).

The data analyzed in in the current study are available from the corresponding author upon reasonable request.

A total of 185 patients presenting with active CF were identified (Supplementary Fig. 1). Two did not consent to study participation. The remaining 183 individuals were followed for a median of 7.0 (IQR 3.9–11) years. Diabetes duration was 32 (20–46) years in individuals with type 1 and 14 (7–20) years in those with type 2 diabetes (P < 0.001). Descriptive characteristics are listed in Table 1. Four individuals had developed CF within 3 months after orthopedic surgery. Fifteen had undergone a previous partial amputation in the CF-affected foot, and two had undergone a previous contralateral above-ankle amputation.

During follow-up of the 183 individuals, 27 developed new CF events. Of those, 13 had a second CF event in the ipsilateral foot and 14 in the contralateral foot, four of which occurred during offloading of the primarily affected foot. Five individuals developed a third new CF event. In total, 215 CF events were registered.

TCC was used for primary offloading in 203 (94%) CF events. Reasons for not using TCC included patient refusal in eight cases and physician decision to use an orthotic walker in four. TCC was aborted because of severe lymphedema in one case.

Recurrence of active CF occurred in 57 (27%) events, and in 40 (70%) of these, a new TCC was applied. In six events, TCC offloading was reintroduced more than once. Four patients were lost to follow-up after uncasting, because they did not attend scheduled follow-ups for unknown reasons. Five individuals died during CF treatment (as a result of arrhythmia in two cases, hypoglycemia, pulmonary embolism, or an unknown cause).

Including only primary and first contralateral events with active CF, 198 events were identified and used for comparison of outcomes between CF stages (Table 2). Those with CF treated with offloading in stage 0 had a significantly shorter duration in TCC compared with those treated with offloading in stage 1 (median 75 [IQR 51–136] vs. 111.5 [72–158] days; P = 0.001). This difference was sustained when only events confirmed by MRI were included (median 77.5 [IQR 53.5–150] vs. 115.5 [70–158]; P < 0.05). Furthermore, differences in offloading duration in TCC persisted, and seemed to increase, when evaluating the total duration in TCC, including acute recurrences requiring recasting. Anatomical distribution was similar in those treated with offloading in stage 0 compared with those treated with offloading in stage 1.

There was no difference in the acute CF recurrence rate between individuals treated with offloading in CF stage 0 versus 1 (Fig. 1). Baseline HbA_1c >8.6% (70 mmol/mol) and age <60 years at presentation were associated with a higher risk of acute recurrence (Supplementary Table 2). In a multivariable model including HbA_1c >8.6% (70 mmol/mol), age <60 years, ischemic heart disease, and diabetic foot ulcer history as factors, only high HbA_1c remained statistically significant (P = 0.01). Including diabetes type as a factor in the model did not alter the outcome (data not shown). A higher risk of developing a new active CF event was seen in those initially treated with offloading in stage 1 (P < 0.05) (Fig. 2). A history of kidney transplantation was also associated with an increased risk of developing a new active CF episode (P < 0.01) (Supplementary Table 3).

Reconstructive surgery was performed in 11 (8.9%) patients with CF treated with offloading in stage 1 compared with none treated with offloading in stage 0 (P < 0.01) (Table 2). In seven feet, surgery was performed because of recurrent ulceration, and in the remainder, surgery was performed because of joint misalignment or disabling deformity restricting ambulation. The median time to reconstructive surgery was 28 (IQR 19–58) months from referral. The rate of future partial and above-ankle amputations did not differ between the groups (Supplementary Figs. 2 and 3). Above-ankle amputation was performed because of CF fracture–related collapse of the foot (after reconstructive surgery) in one case.

CF is a potentially disabling condition, impairing quality of life and increasing the risk of long-term complications, such as deformity, ulceration, and amputation (19,20). Current guidelines emphasize that CF should always be considered when a patient with diabetes and neuropathy presents with a warm, red, and swollen foot and that offloading treatment should start as soon as CF is suspected (4). Previous studies have shown that individuals with CF diagnosed in stage 1 have an increased risk of complications during treatment and development of foot deformities (11,12). The presented retrospective cohort study investigates the impact of initiating treatment in CF stage 0 compared with 1, elucidating the effect on TCC offloading duration, risk of recurrence, and future surgery. The study results show that CF treated in stage 0 is associated with improved outcomes and further strengthen the case for the importance of early recognition of CF where it is typically first presented, including primary care and emergency departments and across specialties handling diabetes and its complications.

In our study, the duration of offloading in TCC was shorter in individuals diagnosed and treated with offloading in stage 0 (i.e., in an early phase without established cortical fractures, joint dislocation, or foot deformity). Because MRI is currently considered to be the golden standard for reliable confirmation of CF in either stage (17,21), we performed an additional comparison including only CF cases confirmed by MRI (n = 118 [60%]), which showed similar differences in TCC offloading duration. Anatomical distribution of CF involvement was similar in both groups and cannot explain the difference in TCC time. Furthermore, offloading in CF stage 0 was associated with a lower risk of developing a new ipsilateral CF event and diminished need for reconstructive surgery. Our findings are consistent with previously published data indicating a lower risk of complications and foot deformities in those diagnosed in CF stage 0 (9,11,22). The rates of partial and above-ankle amputations did not seem to depend on CF stage when offloading treatment was initiated. Similar to results in a previously reported Danish cohort (23), amputations usually occurred several years after the resolution of active CF and were usually performed due to severe infection. One possible explanation for this result could be that individuals treated with offloading in either stage have an inherent risk of developing neuropathic foot ulcers at sites of the foot not affected by previous CF, and they risk worsened limb perfusion and ischemia in the future because of underlying diabetes and cardiovascular risk factors. Therefore, increased risk of amputation is present in all those affected by CF, and acquired foot deformity from CF is only one of several potential risk factors possibly affecting this outcome. Therefore, the study may lack statistical power to detect a potential difference between the stages. The overall above-ankle amputation rate seen in this study is comparable to those in several other reported CF cohorts (19,20,23,24).

Although our findings should be interpreted with caution, they support the clinical use of staging in the treatment of CF, as suggested by Chantelau and Grützner (17), and support the notion that early recognition and treatment of CF in stage 0 is important in order to improve outcomes. The shorter time offloading in TCC, lower rate of future recurrence, and apparent prevention of reconstructive surgery associated with offloading in stage 0, as seen in this study, are all clinically important. Arguably, longer offloading treatment further reduces the lower quality of life associated with CF (19), increases the risk of muscle atrophy and worsened glycemic control, and leads to financial consequences resulting from the use of health care resources, sick leave, and unemployment. Likewise, new events of active CF as well as ulceration and joint instability necessitating reconstructive surgery can lead to additional morbidity. In order to increase the ratio of CF diagnosed in stage 0, awareness of CF as a differential diagnosis likely has to increase in the various fields managing and educating individuals with diabetes and its complications, because knowledge of the disorder has been shown to be lacking among health care practitioners not handling diabetic foot complications (25), which likely contributes to CF being diagnosed late in many cases (26). Apart from educational efforts, reliably diagnosing CF in stage 0 requires access to MRI; a full health economic evaluation of diagnosing and treating more CF cases in stage 0 has yet to be performed.

Impairment of renal function was more prevalent in CF stage 1 compared with stage 0. One might speculate that renal impairment could lead to faster CF progression because of possible osteopenia, renal osteodystrophy, or other factors. Also, because measurements of vibration perception thresholds (e.g., using biothesiometry) were not consistently registered, we cannot exclude the possibility that more pronounced neuropathy delayed requests for health care services in individuals with impaired renal function.

Our findings suggest different possible causes of CF recurrence during the weaning of offloading and new ipsilateral events after the introduction of definitive therapeutic footwear. In our study, we attempted to separate cases of acute recurrence of the inflammatory active stage of an ongoing CF process (caused by excessive reloading of the foot) from new active CF events after full resolution of a previous CF event (when the previous CF process has become completely inactive, and the individual is fully ambulatory). A threshold of 1 year after therapeutic footwear introduction was used because some patients were introduced to footwear immediately after TCC treatment. CF recurrence was independent of CF stage at primary diagnosis but was associated with younger age and higher baseline HbA_1c. A later new ipsilateral event was more common in CF stage 1 as well as in those who had undergone kidney transplantation. Our clinical impression is that this might imply that acute recurrences were associated with lower compliance and possibly a more active lifestyle. After uncasting, a more frequent follow-up schedule could therefore be recommended to improve adherence and metabolic control in these groups of patients. In contrast, later new events seemed to depend on risk factors related to CF per se. An increased risk of developing CF after kidney transplantation has previously been reported and is likely related to the introduction of glucocorticoid immunosuppressive therapy (27). Possibly, acquired foot deformity in stage 1 leads to altered biomechanics, increasing the risk of new skeletal injury and CF.

The strength of this study is its relatively large cohort size and long follow-up time, with few patients lost to follow-up. However, our retrospective design might have introduced bias, and all data depended on available documentation. For example, exact time points for change into new footwear were not always available, because this sometimes occurred between visits at a hospital-linked unit manufacturing footwear not registered in our patient records. Data for biochemistry were missing for a few individuals, and laboratory results obtained differed in time from the first visit. Furthermore, all patients were treated at a single center, possibly reducing generalizability to different populations and health care systems. The use of MRI has increased since 2010, possibly leading to delayed diagnosis or differing management in the earlier years. However, the proportion of individuals with stage 0 CF did not differ between time periods in our study. In some cases, TCC treatment was started without conventional X-ray and before MRI. In these cases, we cannot exclude progression from stage 0 to stage 1 in the period between casting and MRI. In cases examined with both X-ray and MRI, however, staging differed in only one case, implying that offloading in TCC is usually effective in preventing progression. The time between initiation of TCC treatment and MRI examination was shorter in the stage 0 group, possibly reflecting a higher priority on the MRI waiting list, because differential diagnostics sometimes included osteomyelitis. Staging was based on radiology reports. Therefore, we cannot exclude the possibility that selected cases could have been reassessed differently by a radiologist. Because the staging scheme is dichotomous, we deemed radiology reports sufficiently detailed and reliable for staging. Finally, adequate documentation of subsequent ulceration, and its relation to CF deformity, was not consistently available.

In conclusion, in this retrospective cohort study, offloading treatment in active CF before the onset of cortical fractures and joint dislocation (stage 0) was associated with shorter TCC treatment, lower risk of developing a new ipsilateral CF event, and diminished need for reconstructive surgery. However, the risk of future amputation was not affected. Younger age and higher HbA_1c at the time of diagnosis were associated with an increased risk of acute recurrence. A history of kidney transplantation increased the risk of a future new event of active CF.

Funding. This study was funded by unrestricted grants from Gyllenstiernska Krapperup Foundation and Sweden’s Southern Health Care Region.

Duality of Interest. K.F. is on the speakers’ list for and has received consultant fees from Sanofi, AstraZeneca, Novo Nordisk, Boehringer Ingelheim, Amgen, Amarin, and Santen Pharmaceutical. M.L. is on the speakers’ list for and has received consultant fees from Abbott, Amgen, AstraZeneca, Bayer, Boehringer Ingelheim, Eli Lilly, Merck, Novartis, Novo Nordisk, and Sanofi and has received unrestricted research grants from Boehringer Ingelheim and Sanofi. No other potential conflicts of interest relevant to this article were reported.

Author Contributions. J.S. conceived the study, collected data, and wrote the first draft of the manuscript. J.S. and M.L. were involved in the design of the study. All authors analyzed and interpreted the results and reviewed, edited, and approved the final version of the manuscript. J.S. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Prior Presentation. This study was presented in part in abstract form at the Ninth International Symposium on the Diabetic Foot, The Hague, the Netherlands, 10–13 May 2023.