“Volumetric Analysis in Autologous Fat Grafting to the Foot”

Abstract

Background:

Pedal fat grafting is a safe, minimally invasive approach to treat pedal fat pad atrophy. Prior randomized controlled trials demonstrate that the fat as measured directly under the metatarsal heads disappears between 2–6 months after fat grafting, despite patients having relief for 2 years. We aim to utilize MRI to further assess 3D volume of fat in the foot after autologous fat grafting to help explain the mechanism for improved pain.

Methods:

A prospective study was performed utilizing MRI before and at 6 months after pedal fat grafting to assess changes in the 3D morphology of the fat.

Results:

Seventeen patients (6 men and 11 women) underwent injections with a mean volume of 5.8 cc per foot. At 6 months, patients demonstrated increased tissue thickness (p = 0.008) and volume (p = 0.04). Improvements were seen in pain (p < 0.05) and activity (p < 0.05). Foot pressures and forces were significantly decreased and positively correlated with increased fat pad volume (p < 0.05).

Conclusions:

Pedal fat grafting significantly increases metatarsal fat pad volume. The distribution of the fat may contribute to lasting clinical relief in these patients.

INTRODUCTION

Fat pad atrophy of the forefoot is common, affecting approximately 30% of patients over the age of 60.¹ Although often age-related, it may be due to abnormal foot mechanics, obesity, steroid use, or collagen vascular disease.^2–4 Displacement or atrophy of the fat pad can lead to osseous prominences in the foot that may be seen with painful skin lesions. Fat pad atrophy may result in significant pain or compensatory gait leading to callous formation or ulceration. In sensate patients, the pain can lead to emotional and physical pain, leading to productivity and financial losses.^5–9

Various procedures have been described to augment atrophied fat pads; however, many of these techniques have failed to demonstrate sustained tissue thickness over time.^10–12 Our group previously published our results from the first clinical trial to apply standardized fat grafting techniques for pedal fat pad atrophy.^13,14 This study served as a proof-of-concept and highlighted the effectiveness of fat grafting in significantly improving pain and disability, while also decreasing foot pressures and forces at 2 years.

Our prior randomized, cross-over clinical trial demonstrated that patients treated with standard-of-care padding got worse, with decreased tissue thickness over time. In our fat grafting group, the tissue thickness under the metatarsal heads was measured by ultrasound and discovered to return to baseline thickness between 2 and 6 months, despite long lasting pain relief out to the 2-year end of study time point. We hypothesize that, despite our previous finding of decreasing tissue thickness over time, the volume of fat that is injected into the foot is retained at 6 months and likely redistributes around the metatarsal heads to support them long term. We utilized MRI to validate this hypothesis of volume retention, in comparison to our standard ultrasound measurements, to further understand changes that are occurring in the foot when autologous fat is used to treat plantar fat pad atrophy.

METHODS

Trial Design

17 adults who experienced pain from fat pad atrophy were recruited into an IRB-approved, prospective, case series (ClinicalTrials.gov identifier: NCT02638532). The overall purpose of this study was to utilize MRI to determine the 3D volumetric changes in the soft tissue of the foot after fat grafting. Inclusion criteria consisted of patients with foot pain under the head of the metatarsals, a diagnosis of fat pad atrophy by a foot and ankle specialist, and were 6 months post any surgical intervention or steroid injection to the foot. Exclusion criteria included patients with uncontrolled diabetes mellitus (HgA1c level > 7.0), open ulcerations, infection (including osteomyelitis), systemic disease that would render the fat harvest and injection procedure unsafe to the patient, pregnancy, known coagulopathy, and tobacco use within the last 12 months.

Medical, surgical, social, and activity histories were performed. Vitals including temperature, blood pressure, height, weight, and body mass index were performed. Any prior foot injury, surgery, or previous foot ulcerations were noted. A physical exam and complete foot exam were documented including vascular, neurologic, dermatologic, and gait evaluations.

Upon completion of the screening visit, phlebotomy was performed to assess serum complete blood count with differential, comprehensive chemistry panel, coagulation studies, erythrocyte sedimentation rate, albumin, and HgA1c. Standardized two-dimensional photos of the foot, including any lesion pattern were captured.

Intervention

Surgical consent was provided and surgical procedures were performed in the UPMC Aesthetic Plastic Surgery Center in Pittsburgh, PA, USA. Subjects received local anesthesia (lidocaine 1% with epinephrine 1:100,000) at the site of aspiration of the fat grafts (abdomen, thighs, or flanks) and a tumescent solution (500 milliliters normal saline, 10 milliliters 2% lidocaine, 1 milliliter 1:1000 epinephrine) was injected into the harvest site. A tibial nerve block and forefoot Mayo block was performed with a 50:50 mixture of 2% lidocaine 0.5% bupivacaine without epinephrine. A blunt tip multi-hole hollow cannula was used to aspirate approximately 50–100 milliliters of fat tissue through a stab incision made with a #11 blade. Liposuction was performed under a low, consistent negative pressure using 10-milliliter syringes to limit trauma to the adipocytes. Incisions for donor sites were closed with benzoin and Steristrips™.

A standard Coleman technique was used to process the fat, where the harvested fat graft was placed in centrifugation at 3000 rpm for three minutes.¹⁵ The resultant fat was decanted, oil was wicked using absorbent gauze, and the high-density fraction (bottom-most 1 milliliter of each 10-milliliter syringe) was transferred to 1-milliliter syringes for injection into the foot. An 18-gauge needle was used to make an entry site between the first and second toe and the fourth and fifth toe on the plantar aspect of the foot, allowing for a cross-hatch injection pattern. Occasionally injections were performed from the dorsal aspect. A 0.9-millimeter blunt cannula was used to inject the 1-milliliter syringes of fat into the foot.

Post-operatively, subjects walked out of the clinic in comfortable sneakers with padded insoles, allowing for offloading of the fat grafted region. Subjects were instructed to limit strenuous activity for 4–6 weeks including barefoot walking. All procedures were performed by the same surgeon (J.A.G.).

No subjects received additional treatment of fat grafting or underwent any other surgical intervention to the foot during the clinical trial.

Measurement of Tissue Thickness

Ultrasound [Terason Ultrasound Imaging System, Version 4.7.6, Burlington, MA 01803 USA] was used to measure plantar tissue thickness under each metatarsal head. Ultrasound was performed by the same clinician at every visit, excluding the 2-week post-operative visit.

MRI Analysis

Initial pre-operative MR imaging was performed at 1.5T using T1-weighted spin echo, T2-weighted fast spin echo with or without fat suppression and STIR sequences in sagittal, coronal and axial planes. Three-dimensional volumetric reformatted imaging was performed using Vitrea (Vital Images, Minnetonka, MN, USA) with surface rendering, as well as emphasis on the underlying forefoot fat pads. Post-operative imaging was performed at a ~6-month interval with similar imaging parameters.

Pre- and post-operative imaging included characterization of the location and appearance of the adipose tissue in the plantar forefoot surrounding the 2^nd through 4^th metatarsal heads. Overall percentage change in fat pad volume was measured.

Additionally, attention was paid to the presence of subchondral bone marrow edema both pre- and post-operatively. Metatarsophalangeal joint angles were also measured and compared. All pre- and post-operative imaging was reviewed by a musculoskeletal-trained radiologist with 7 years of experience.

Measurement of Pain and Disability

Foot pain and subject disability was measured by the Manchester Foot and Disability Index (MFDI), a validated assessment of the foot that includes components of pain, function, appearance, and work/leisure activities.¹⁶ Additionally, ability to perform activities of daily living (ADLs), as well as sports-related activities, was assessed by the Foot and Ankle Ability Measure (FAAM).¹⁷ The questionnaires were administered at every visit, excluding the operative visit and 2-week post-operative visit.

Measurement of Stance and Gait Force and Pressure

The Tekscan HR Mat™ pressure measurement system and Research Foot Module [South Boston, MA] was used to obtain pedobaragraphic data to obtain baseline plantar foot forces and pressures. Subjects were weight-calibrated to measure forces and pressures applied while standing, then re-calibrated for walking. Standing measurements were captured from an average of 150 seconds. Walking measurements were captured from an average of a minimum of three passes for each foot at a self-selected speed. The pedobarograph was performed at every visit, excluding the operative visit and 2-week post-operative visit.

Statistical Analyses

Statistical analyses were performed with IBM Corporation SPSS Statistics for Windows Version 24.0 [IBM Corporation, Armonk, NY, USA]. Normality tests of Kolmogorov-Smirnov and Shapiro-Wilk were conducted. Paired t-tests were run to determine difference in means between parametric data and Wilcoxon rank-sum tests were used to evaluate differences in medians for non-parametric data. Tests were two-sided and significance was set to the level of p<0.05, 0.01, or 0.001, as indicated. All outlier data (2ϭ) were removed prior to analyses. Data for injected feet only were evaluated, to avoid diluting the results with unaffected foot measurements. Correlations were determined by Pearson’s coefficient at a confidence level of p<0.05. All data are presented as mean ± standard error of the mean.

RESULTS

Participant Characteristics

17 fat pad atrophy patients (11 female, 6 male) were enrolled into the study and retained throughout the 6 months to study completion. Average age at screening was 57.5 ± 2.8 years, average BMI was 26.0 ± 1.5 kg/m². Neither age or gender were found to have any correlation to any outcome. BMI was negatively correlated to volume increase [p=0.043, r=−0.52] (i.e., with higher BMIs, a lower increase in volume was measured) and reported pain outcomes [p=0.033, r=−0.394] (i.e., with higher BMIs, a lower improvement in pain was reported).

Etiologies for fat pad atrophy include failed neuroma surgery, prior foot surgery, steroid injections, and overuse. Fat pad atrophy was diagnosed in 31 feet where 15 subjects underwent bilateral fat grafting injections and 2 subjects underwent fat grafting in only the one injured foot. An average of 5.8 ± 0.4 cc of fat were injected into the left foot and average of 5.8 ± 0.4 cc of fat injected into the right.

Subjects experienced post-operative bruising of the donor site and feet, soreness, and pain. No patients experienced infection, hematoma, seroma, or oil cysts. No peri-operative antibiotics or narcotics were utilized. No serious adverse events or unanticipated events occurred.

Tissue Thickness Outcomes

Figure 1 displays fat pad thickness measured over time. Fat pad thickness is lowest at baseline (p<0.0001), highest immediately post-operatively (p<0.0001), then, at 2-months post-operative, fat pad thickness decreases and settles to the same thickness measured at 6-month post-operative (p=0.144), remaining thicker than at baseline (p=0.008).

Figure 1.

Fat pad thickness measured over time by ultrasound.

MRI Analysis

Figure 2 highlights volumetric measurements obtained at baseline and 6 months post-operatively. There was no difference in means between L3 or R3 fat pad cross-sectional area at baseline and 6 months post-operative (p>0.05). L234 volume and R234 volume were both significantly higher at 6 months post-operative than at baseline (p=0.04). Fat volume measured under 3 and 234 were found to be positively correlated [p=0.001, r=0.567]. Figure 3 demonstrates the fat pad as seen on MRI.

Figure 2.

MRI volumetric measurements obtained at baseline and 6 months post-operatively.

Figure 3.

(a) Selection of the fat pad volume on the sagittal T1 sequence for measurement. (b) 3D model showing the actual volume of post-procedural fat on the background of the patient’s soft tissues.

Only 4 of our 17 patients (23.5 %) showed any signs of marrow edema on either their pre- or post-operative MRI. In one patient the edema improved post-operatively; in another it was unchanged; and in the final two it was actually a new finding post-operatively.

Metatarsophalangeal joint angles on MRI at baseline averaged 154.0 ± 1.7 and, at 6-month post-operative averaged 155.0 ± 1.6, with no significant difference in means between the time points (p=0.819) [Figure 4].

Figure 4.

MRI angle measurements obtained at baseline and 6 months post-operatively, boxes represent medians, lines as ranges.

Pain and Disability Outcomes

MFDI scores were surveyed at each time point where a lower score indicated improved functionality, pain, appearance, and work/leisure activities; in addition to FAAM scores where a higher score indicated improved ability to perform ADLs and sports.

Function scores were lowest (most improved) at 6 months post-operative (p<0.05) [Figure 5]; pain scores were lowest (most improved) at 2 ad 6 months post-operative (p<0.05); no changes in appearance were revealed throughout the study (p>0.05); work leisure scores were lowest (most improved) at 6 months post-operative (p<0.05). Both ADL and sports scores were highest (most improved) at 6 months post-operative (p<0.05).

Figure 5.

Pain and disability outcome scores.

Pearson’s correlations indicated that improved functionality via the MFDI was directly correlated to improved pain, appearance, work leisure, ADLs, and sports (p<0.05), and improved pain scores were directly correlated with improved functionality, appearance, work leisure and ADLs (p<0.01). Increase in pain scores directly correlated to worsened reported work/leisure and function outcomes in both groups (p<0.05).

Stance and Gait Force and Pressure Outcomes

Standing force was lowest at 6-month post-operative (p<0.05) [Figure 6] and standing pressure was lowest at 2 and 6 months post-operative (p<0.05). Walking forces and pressures were both highest at baseline (p<0.01), then decreased throughout the study.

Figure 6.

Stance and gait force and pressure outcomes represented in box-and-whisker plots.

Decrease in walking forces and pressures were found to be positively correlated with increase in fat pad volume [p=0.013, r=0.441; p=0.032, r=0.382 respectively]. Improvements in reported work leisure and activities of daily living scores were found to be positively correlated with decrease in standing forces [p=0.024, r=−0.431; p=0.043, r=0.362 respectively] and pressures [p=0.003, r=−0.562; p=0.009, r=0.465 respectively].

DISCUSSION

While our group’s prior work has already demonstrated the effectiveness of fat grafting in significantly improving foot-related pain and disability, this study is the first to show a significant increase in retained volume over time. This confirms our hypothesis that, while plantar tissue thickness as directly measured under the metatarsal head may decrease over time, injected fat is retained and remains supportive while redistributing around the metatarsal heads. Three-dimensional imaging is better than ultrasound at assessing volume retention in the ball of the foot.

An interesting finding in our study is that higher body mass index correlated to lesser improvement in fat pad volume and pain reduction. Despite padding patients’ insoles and instructing them to avoid strenuous activity postoperatively, it is challenging to totally offload the grafted region in these subjects. Thus, the greater a patient’s weight, the more pressure that will be applied to the grafted fat, and this may lead to a less-favorable outcome. An alternative hypothesis is that the quality of fat in high body mass patients may not be as favorable. While this should not necessarily exclude higher body mass patients from undergoing pedal fat grafting, this is something upon which they should be counseled preoperatively. It is also worth noting that all patients in our study were either non- or well-controlled diabetics without sequelae such as peripheral neuropathy. The application of pedal fat grafting to diabetics with peripheral neuropathy remains unstudied, but we hypothesize that this would lead to greater difficulty with offloading the grafts and less-favorable outcomes.

While we hypothesized that patients with palpable metatarsal pain would show evidence of underlying bone marrow edema that could be improved upon by pedal fat grafting, we did not observe this to a significant extent. Thus, bone marrow edema does not appear to reliably correlate with foot pain in our patients.

Although several patients in our study with hammertoes appeared to have less toe deformity following pedal fat grafting, analysis of the metatarsophalangeal angle failed to reveal any significant change in toe position postoperatively. Hammertoes cause a retrograde force leading to greater plantar flexion and prominence of the metatarsal heads at the ball of the foot with fat pad displacement just behind the toes. Therefore, it may be ideal to address osseous deformities (i.e. bunions, hammertoes, etc.) in conjunction with fat grafting for these patients.

A limitation of this study is the 6-month follow-up, which was primarily due to funding of the study. While our study shows a significant increase in volume at 6 months, we also observed increased tissue thickness at this time point, and it remains to be seen how volume might change at a longer time point. Further MRI imaging at 1- or even 2-year time will help answer the question of whether volume is retained and correlated with improved pain. Another limitation is the small sample size of our cohort. A larger sample may reveal more significant information regarding bone marrow edema, metatarsal angle, and volume over time. Forefoot fat pad atrophy is often bilateral. We decided to offer the procedure on both feet if patients had pain in both feet to avoid the need for a secondary operation. Our outcomes may have been different if we had just operated on one foot at a time, allowing for better pressure relief of the operated foot and potentially better recovery and fat take.

Another possible explanation for the long-term improvement in patient outcomes following pedal fat grafting relates to the quality, rather than the quantity, of fat in the metatarsal fat pads. Studies have shown that the biomechanical properties of plantar soft tissues change with advancing age, with increasing fat pad stiffness and reduced compliance likely contributing to the development of foot-related problems.^18–20 Thus, perhaps a qualitative biomechanical change in the metatarsal fat pad following pedal fat grafting could account for long-term symptomatic improvement. We are currently investigating the stem cell characteristics of the fat used in our foot fat grafting cases and aim to correlate it to clinical findings in the near future.

CONCLUSION

Pedal fat grafting leads to a significantly increased volume in the metatarsal fat pads seen at 6 months, which correlates with improved function and pain, along with decreased foot forces and pressures. Further studies will show whether this lasts at longer time points, or if the qualitative rather than quantitative changes in the fat pads are contributory as well.