Quantitative assessment of foot structure in rheumatoid arthritis by a foot digitizer: Detection of deformities even in the absence of erosions

Authors


Abstract

Objective

Foot involvement is a major feature in rheumatoid arthritis (RA), leading to structural deformities. Methods to allow a 3-dimensional (3-D) evaluation of foot structure in RA to be applicable in daily clinical practice have not been evaluated. This study assessed the use of a foot digitizer, a noninvasive 3-D scanner collecting objective quantitative data of the feet, to evaluate the presence of foot structure abnormalities in an RA outpatient cohort.

Methods

Foot digitizer data of RA patients were compared with healthy controls. Subanalyses were performed to find relationships with erosive disease and the presence of swollen and/or tender joints. Linear mixed models were applied with correction, including sex, age, body weight and height, foot length, Disease Activity Score in 28 joints, and disease duration.

Results

Forty-one percent of the patients showed >1 abnormal parameter, measured with the 3-D foot scanner. Most differences found were located in the forefoot, the most frequently affected area of the RA foot. Strikingly, even in the absence of joint erosions, marked alterations were found. Comparable differences were also observed between the patients with and without swollen and/or tender joints. Additionally, alterations were not strongly related to foot pain and disability, suggesting the capacity of the foot digitizer to detect early changes in foot structure.

Conclusion

The results highlight the impact of RA on foot structure, even in the absence of clinical signs of swelling or radiographic erosions. The foot digitizer offers a valuable tool to screen for such foot deformities before the presence of erosions.

INTRODUCTION

Foot involvement is a major feature of rheumatoid arthritis (RA). Twenty percent of patients initially present with foot problems. Increasing with disease duration, 90% of patients are affected by foot problems during the course of their disease (1). Foot deformities can affect the entire foot in RA. The most common forefoot deformities are hallux valgus (23–91%) and lesser toe deformities (40–94%) (2–9). There is often a fibular deviation of the toes, caused by the dorsolateral dislocation of the lesser metatarsophalangeal joints. Distal migration of the plantar fat pad in the forefoot and subluxation of the metatarsal joints leads to prominence of the metatarsal heads. Forty percent to 86% of the patients have widening of the forefoot (2, 3, 7, 9). In the midfoot, a flattening of the medial longitudinal arch (46–81%) and a marked abduction of the forefoot on the rearfoot can be seen (2, 3, 7, 9–11). The RA hindfoot is characterized by a heel eversion (25–35%), leading to pes planovalgus (2, 8, 9, 12). The deformities were documented using a variety of methods, such as based on previous literature, radiographs, and clinical examinations, often comparing RA patients with a healthy control group. The prevalence of foot-related impairment and disability ranges from 35–70% across the disease spectrum (13).

Yet in clinical research the rheumatoid foot is often neglected. There seems to be limited clinical interest, which could possibly lead to insufficiencies in treatment (1, 14, 15). Assessing the impact of RA on foot structure and function can be done by careful clinical examination, including goniometric range of motion measurements, and investigations such as imaging and gait analysis. However, in routine practice, assessment of foot problems in RA is usually restricted to the patient's symptoms evaluation and clinical examination.

The disease activity indices evolved from including the feet and ankles to omitting them in the joint count. The American College of Rheumatology 68 joint count developed over a 36 joint count to a 28 joint count examination, eliminating the hips, ankles, and feet (16). The Disease Activity Score (DAS) originally included a 44 swollen joint count and used the Ritchie Articular Index (RAI) to assess joint tenderness (16). The RAI includes the feet but groups the joints, e.g., the metatarsophalangeal joints (16, 17). Its 0–3 graded evaluation of the severity of joint tenderness may be somehow subjective and complicated (17). Currently, the DAS28, only using 28 joints, is frequently used in daily clinical routine, even though it omits the feet and ankles. However, it has been a matter of debate whether the exclusion of the ankles and feet may jeopardize the definition of remission or may not reflect the activity or severity of the patient (16, 18–21), although it is shown that the inclusion of the ankles and feet only rarely influences the definition of overall disease activity (18). Bakker et al showed that in RA patients with mainly disease progression in the feet, the DAS28 underestimates the disease activity and the expected joint damage (22). The clinical importance at the individual patient level of the ankle and feet information is generally recognized (18, 19, 23). Recently, simplified versions of the original DAS are validated. They include the feet and ankles but use a yes/no answer to score tenderness in the joints instead of the 0–3 score used previously (16–19, 23).

Patient-reported outcome measures can be used as well to subjectively assess and evaluate the multidimensional consequences, e.g., the effect on a patient's functional ability, of foot problems (24). The Foot Function Index (FFI) (25, 26) and the Foot Impact Scale for Rheumatoid Arthritis (FIS-RA) (13, 27) are examples of such indexes. The FFI was developed to measure the effect of foot problems on function in terms of pain and disability. The FIS-RA was developed to measure foot-related impairment and disability at the individual and group levels. It comprises 2 subscales for impairment/footwear and activities/participation. Even though the use of patient-reported outcome measures should be integrated in patient care, they are subjective, self-administered questionnaires that do not require inspection of the feet by the physician.

Nevertheless, it is clear that the feet need proper attention and an objective evaluation. However, there is a lack of validated instruments for the physician's assessment of foot pain, deformity, and function (28). Therefore, it is recommended that research should focus on development and timing of evidence-based assessments of the foot with RA (29). A foot digitizer, based on an optical laser scanning procedure, is a quick, easy-to-use, and noninvasive scanner device to collect objective 3-dimenstional (3-D) data of the feet. Measurements made by means of a 3-D scanner proved to be valid and reliable (30).

Here, we assessed the use of a foot digitizer to collect quantifiable structural data of the feet that can easily be used in daily clinical practice to evaluate the presence of foot structure abnormalities in patients with RA.

Significance & Innovations

  • Foot involvement is still underestimated in rheumatoid arthritis.

  • The Infoot foot digitizer offers a valuable noninvasive tool for the screening of foot deformities.

  • Foot deformities can be detected with a foot digitizer even in the absence of radiographic damage.

SUBJECTS AND METHODS

Subjects.

One hundred six patients (74 women and 32 men) volunteered to participate. The patients all had a confirmed diagnosis of RA based on the American College of Rheumatology criteria (31). Standard clinical and disease-specific information was collected for the RA group. Patient characteristics are shown in Table 1. The patients were included in a consecutive manner when presenting for their normal followup visit at the rheumatology department of Ghent University Hospital, irrespective of foot symptoms/deformities being present or not.

Table 1. Patient characteristics*
 All RA patients (n = 106)Infoot normal (n = 63)Infoot deviations (n = 43)
  • *

    Values are the mean ± SD unless otherwise indicated. RA = rheumatoid arthritis; DAS28 = Disease Activity Score in 28 joints; RF = rheumatoid factor; anti-CCP = anti–cyclic citrullinated peptide; ESR = erythrocyte sedimentation rate; CRP = C-reactive protein; DMARDs = disease-modifying antirheumatic drugs; anti-TNF = anti–tumor necrosis factor; NSAIDs = nonsteroidal antiinflammatory drugs.

Age, years60 ± 1257 ± 1364 ± 9
Body weight, kg73 ± 1573 ± 1474 ± 16
Body height, meters1.66 ± 0.81.65 ± 0.71.67 ± 0.9
Disease duration, years14 ± 1113 ± 1016 ± 13
DAS28 score2.5 ± 1.22.2 ± 0.72.7 ± 1.5
RF positive, %636166
Anti-CCP positive, %737373
ESR, mm/hour19.7 ± 17.417.7 ± 14.322.5 ± 20.9
CRP level, mg/liter10.2 ± 19.78.4 ± 19.112.8 ± 20.4
Swollen/tender foot joints, %494849
 No. of tender joints6 ± 143 ± 710 ± 20
 No. of swollen joints2 ± 52 ± 33 ± 7
Medication, %   
 None452
 DMARDs only555654
 DMARDs + anti-TNF383642
 Anti-TNF only332
 NSAIDs475437

One hundred thirty-five healthy adults (80 women and 55 men), community dwelling and some living in a retirement home, participated as the control group. Their mean ± SD age was 50 ± 18 years and their mean ± SD body mass and height were 73 ± 14.6 kg and 1.69 ± 0.13 meters, respectively. Inclusion criteria were no systemic diseases; 6-month symptom-free history for lower extremity bones, joints, and soft tissues; no obvious foot deformities; no history of foot surgery; and age between 30 and 80 years. Ethical committee approval was obtained and all of the participants signed an informed consent.

Data acquisition.

All measurements were performed in the late morning to avoid influence of changes in foot volume, which might occur in the patient group. Scanning with the Infoot digitizer (I-Ware Laboratory) was performed according to the description in previous work (30). The Infoot is a 3-D foot scanner that is based on an optical laser scanning system that scans a foot form and the anatomic landmark points. Placement of the green velvet markers (I-Ware Laboratory) was done bilaterally with the subjects standing so that the feet were in a loaded condition to minimize skin movement between the marker placement and the measurement. The markers were placed on the landmarks following the instructions of the manufacturer (see Supplement 1, available in the online version of this article at http://onlinelibrary.wiley.com/journal/10.1002/(ISSN)2151-4658). The subjects were then asked to step with 1 foot in the digitizer while the other foot was placed next to the scanner on a step, with the exact same height as the glass plate in the scanner. They were asked to stand still and relaxed and to distribute their body weight equally over both feet, which led to an estimated half–body weight bearing. Scanning was performed with a scan pitch of 1.0 mm using an optical laser scanning procedure. The subjects were not allowed to move their feet while scanning (∼5 seconds). The data were processed with the accompanying software (I-Ware Laboratory), which provides length, width, and height measures; circumferences; and some angles. More detailed information can be found in Supplement 1 (available in the online version of this article at http://onlinelibrary.wiley.com/journal/10.1002/(ISSN)2151-4658).

Dorsoplantar and sagittal weight-bearing radiographs were taken bilaterally. For the dorsoplantar view the subject stood on the film, which was placed flat on the floor. For the sagittal view the x-ray cassette was positioned vertically with 1 foot placed on each side of the film. This was done for both feet in random order. Focus-to-film distance, position, and angle of the central ray and exposure were consistent for all of the subjects (32, 33). Length, width, heights, and angles were measured. More detailed information can be found in Supplement 2 (available in the online version of this article at http://onlinelibrary.wiley.com/journal/10.1002/(ISSN)2151-4658). The distribution pattern of the present erosions was documented by a senior rheumatologist (DE).

The Dutch version of the FIS-RA (13) and the Dutch version of the FFI (26) were used as patient-reported outcome measures to assess the subjective consequences of the foot problems.

Statistical analysis.

Linear mixed models (IBM SPSS Statistics, version 19) were applied to compare the 3-D scanner data between the controls and the patients with and without erosions and to compare the radiograph data between the patients with and without erosions. Correction was done for sex, age, body weight and height, foot length, DAS28, disease duration, and the presence of tender or swollen foot joints or the presence or absence of erosions. Post hoc analyses were performed with Bonferroni correction. A univariate generalized linear model was used to analyze the group differences for the FIS-RA and the FFI. P values less than or equal to 0.05 were considered to be statistically significant.

RESULTS

Deformity, tender and swollen joints, and erosions.

Initially the foot digitizer data were analyzed for foot deformities and found that 41% of the patients showed >1 abnormal parameter, measured with the 3-D foot scanner. A parameter was considered “abnormal” when it showed a value outside the range of the cutoff values. The cutoff values were determined by calculating the 2-SD value on the histograms, made for each parameter obtained with the digitizer for the healthy group.

Looking at the disease characteristics, actual swollen and/or tender joints were present in 49% of patients (Table 1). A majority of the RA patients (62%) had signs of erosions at the feet, with varying distribution patterns and the forefoot being the most affected area. In 25% of the patients, the entire foot was affected by erosions. Fifty-four percent had erosions only in the forefoot and 1% only in the hindfoot. Four percent showed signs of erosions both in the forefoot and the midfoot, 15% both in the forefoot and the hindfoot, and 1% in the midfoot combined with the hindfoot.

The findings of the digitizer data were independent of sex, age, body weight, disease duration, DAS28, or the presence of tender or swollen foot joints or the presence of erosions.

3-D scanner parameters.

Ball girth circumference and height of top of ball girth were significantly larger for both subgroups of RA patients versus the controls, while the fibular instep length was significantly smaller for the patients. Foot width and hallux angle were significantly larger for the erosive group, whereas instep length and height of toe 1 were significantly smaller for the nonerosive group (Table 2 and Figure 1).

Table 2. Significant differences for the 3-dimensional scanner parameters between the controls and the patients with and without erosions*
Infoot parameterNo erosionsErosions
ControlsRA patientsPControlsRA patientsP
  • *

    Values are the mean ± SD. RA = rheumatoid arthritis; NS = nonsignificant.

Ball girth circumference, mm237.6 ± 15.07243.2 ± 12.060.021237.6 ± 15.07247.1 ± 17.460.002
Height of top of ball girth, mm40.0 ± 3.4741.9 ± 3.200.00240.0 ± 3.4741.5 ± 3.900.003
Foot width, mm98.4 ± 6.81100.9 ± 5.95NS98.4 ± 6.81103.1 ± 7.320.009
Instep length, mm186.5 ± 13.00182.0 ± 10.730.004186.5 ± 13.00183.1 ± 11.66NS
Fibular instep length, mm164.1 ± 11.63159.6 ± 9.930.03164.1 ± 11.63158.7 ± 11.150.024
Hallux angle, °11.0 ± 8.1411.4 ± 8.06NS11 ± 8.1414.6 ± 10.620.007
Height of toe 1, mm25.5 ± 2.8423.8 ± 3.700.02625.5 ± 2.8424 ± 3.71NS
Figure 1.

Significant differences for the 3-dimensional scanner parameters between the controls and the patients with and without erosions. RA = rheumatoid arthritis; 1 = ball girth circumference; 2 = height of top of ball girth; 3 = foot width; 4 = instep length; 5 = fibular instep length; 6 = hallux angle; 7 = height of toe 1; 8 = height of toe 5.

Height of top of ball girth was significantly larger and instep length and height of sphyrion fibulare were significantly smaller for the patients with and without swollen and/or tender joints in the feet compared to the controls. Height of both toes 1 and 5 was significantly lower for the patient group with swollen and/or tender joints (Table 3).

Table 3. Significant differences for the 3-dimensional scanner parameters between the controls and the patients with and without swollen and/or tender joints*
Infoot parameterNo swollen and/or tender jointsSwollen and/or tender joints
ControlsRA patientsPControlsRA patientsP
  • *

    Values are the mean ± SD. RA = rheumatoid arthritis; NS = nonsignificant.

Height of top of ball girth, mm39.7 ± 3.4742.0 ± 3.210.00339.7 ± 3.4742.2 ± 4.360.002
Instep length, mm187.6 ± 13.00181.5 ± 9.980.013187.6 ± 13.00181.0 ± 13.900.037
Height of sphyrion fibulare, mm58.8 ± 5.5358.1 ± 5.580.01358.8 ± 5.5356.2 ± 7.970.002
Height of toe 1, mm26.2 ± 2.8423.4 ± 2.45NS26.2 ± 2.8422.6 ± 4.780.012
Height of toe 5, mm20.6 ± 2.6918.6 ± 2.89NS20.6 ± 2.6917.8 ± 4.800.027

The FFI (pain and disability subscale) and the FIS-RA (impairment/footwear and activities/participation subscales) both revealed significantly higher scores for the patient group versus the controls (P ≤ 0.0001).

3-D scanner parameters in patients with and without erosions.

The radiographic data showed significant differences between the patient groups with and without erosions in the feet. The navicular height was lower (P ≤ 0.01) and the medial arch angle (P ≤ 0.05) and the hallux angle, measured at the soft tissues (P ≤ 0.05), were bigger for the patients with erosions (Table 4 and Figure 2).

Table 4. Significant differences for the radiographic parameters between the patients with and without erosions*
 No erosionsErosionsP
  • *

    Values are the mean ± SD.

Navicular height, mm45.2 ± 6.7642.3 ± 7.120.004
Medial arch angle, °118.1 ± 6.92120.1 ± 7.950.02
Hallux angle (soft tissues), °16.0 ± 9.8917.0 ± 10.370.03
Figure 2.

Significant differences for the radiographic parameters between the patients with and without erosions. RA = rheumatoid arthritis; 1 = navicular height; 2 = medial arch angle; 3 = hallux angle measured at the soft tissues.

For the digitizer data, only the height of toe 5 differed significantly (P ≤ 0.05) between the patient groups with and without erosions in the feet, being higher for the erosive group (Figure 1).

There were no differences in the FFI and FIS-RA scores in those patients with and without erosive disease.

3-D scanner parameters in patients with and without actual swollen and/or tender joints in the feet.

There were no significant differences between the radiographic data of the patient group with and the group without swollen and/or tender joints in the feet. Likewise, no differences could be demonstrated for the 3-D scanner data.

The questionnaires revealed significantly higher scores for the pain subscale of the FFI (P ≤ 0.05) and for both the impairment/footwear (P ≤ 0.0001) and activities/participation subscales (P ≤ 0.01) of the FIS-RA for the patients with swollen and/or tender joints in the feet. There was no difference between the 2 patient groups for the disability subscale of the FFI.

DISCUSSION

The aim of this study was to determine if a foot digitizer is a valuable tool for daily clinical routine to examine the feet in patients with RA. The results demonstrated that significant differences in foot measures, as registered with the 3-D scanner, could be found between controls and RA patients, even in the absence of erosions. Since no significant differences were observed between the patients with and the patients without actual swollen and/or tender joints for these parameters, the differences could not be attributed only to the swelling caused by inflammation.

Introduction of new therapies and treatment strategies led to an increased importance of imaging techniques. The trend to treat RA earlier and more aggressively causes the need to identify early changes. Radiography was long the mainstay for imaging patients with RA, but because ultrasound and magnetic resonance imaging (MRI) are sensitive for the detection of synovitis, they became superior. As described elsewhere, both have their advantages and disadvantages (34, 35). A digitizer can be an additional outcome measure used for objective followup of feet with RA. It is a noninvasive imaging technique without ionizing radiation that provides a global, multiplanar picture of the foot in a functional, loaded condition. The 3-D foot scanner can be an additional tool to objectively quantify changes in foot structure, whereas ultrasound and MRI focus more on the inflammation. Its diagnostic utility is the possibility to depict early changes in the RA foot in an outpatient clinic. Although scanning is done in a loaded and therefore functional position, the system is not dynamic. Although having the advantage of being highly accurate, it is a surface scanning technique, i.e., a proxy to direct measurement of bone and joint alignment. It also has to be recognized that local soft tissue swelling, general foot/ankle edema swelling, and obesity can influence the results.

Even though all of the participants included belonged to the same age range (between 30 and 80 years), the average age turned out to be a difference of ∼10 years. Therefore, corrections for age were applied in the statistical analysis to exclude the influence of age on the findings.

The presence of swollen and/or tender joints was scored separately. Since the DAS28 is the most common index to describe disease activity, it was also chosen for this study to ease comparison with other studies. However, as mentioned earlier, the 28 joint count omits the ankles and feet. The RAI includes the ankles and feet but clusters the joints, and is possibly subjective and more complex due to the graded evaluation of the severity of joint tenderness. Recently, simplified versions of the original DAS are validated. They include the feet and ankles but use a yes/no answer to score tenderness in the joints instead of the 0–3 score used previously (17). Unfortunately, data were collected before publication of these new DAS scores. Therefore, the ankle and foot joints were scored for the presence/absence of swelling or tenderness as a separate index. This score was used as a correction factor when comparing the patients with versus without erosions. However, the 3-D measurements made with the digitizer did not show any difference between the patient group with and the group without swollen and/or tender joints in the feet. This might be because the digitizer not only reflects the influence of the swelling, but it is an accumulation of all of the influences on the foot, such as a changed bony configuration and changes due to the ligamentous laxity.

The patient cohort in this study might differ from those included in other studies about foot deformities in RA. Most often the patients are selected based on the presence of foot deformities or pain and symptoms. In this study design, patients were included in a consecutive manner when presenting for their normal followup visit, irrespective of foot symptoms/deformities being present or not. This resulted in 62% having signs of erosions at the feet and 49% having swollen and/or tender foot joints.

Forefoot erosions were present in 98% of the patients in the erosive group, with 54% showing erosions only in the forefoot. Likewise, most differences found with the foot digitizer are located in the forefoot. This might reflect the fact that the forefoot is the most frequently affected area of the foot (21, 36–39).

The FFI and the FIS-RA revealed significant differences between the controls and the RA patients, indicating the subjective burden caused by the feet affected by the disease. The presence or absence of erosions is not reflected in the results, since no differences were found for the questionnaires between the erosive and nonerosive groups. The presence of swollen and/or tender joints, representing arthritis activity, leads to higher scores. This suggests a higher subjective impact on the group with higher disease activity.

Future work should focus on early RA before erosions are present to confirm that early signs of foot involvement can be detected with a foot digitizer. Research should also compare the 3-D scanner with ultrasound and MRI to determine its relative value as an outcome parameter. Future work should also examine if early detection of foot deformities by means of a 3-D scanner can prevent secondary mechanical strain induced by these structural deformities, if this prompts early treatment with customized insoles or shoe wear and the impact of early treatment modalities.

The results of this study highlight the impact of RA on foot structure, even in the absence of clinical signs of swelling or radiographic erosions. The foot digitizer offers a valuable tool to screen for such foot deformities before the presence of erosions.

AUTHOR CONTRIBUTIONS

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Elewaut had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. De Mits, Mielants, De Clercq, Woodburn, Roosen, Elewaut.

Acquisition of data. De Mits, Mielants, De Clercq, Woodburn, Roosen, Elewaut.

Analysis and interpretation of data. De Mits, Mielants, De Clercq, Woodburn, Roosen, Elewaut.

Ancillary