Bank, Jonathan M.D.; Fuller, Sam M. M.D.; Henry, Ginard I. M.D.; Zachary, Lawrence S. M.D.
Raynaud phenomenon causes progressively decreased blood flow to the extremities, resulting from an imbalance between vasoconstriction and vasodilation. It is a common affliction, affecting 2 percent of the adult population, and is more common in women.1 Raynaud phenomenon can be divided into two subtypes: primary Raynaud phenomenon, which usually exists in isolation from systemic disease and in which the vasoconstriction results in cold attacks and pain that can be severe and debilitating; and secondary Raynaud phenomenon, which typically occurs in the setting of autoimmune diseases such as systemic sclerosis (scleroderma). The vascular imbalance ultimately leads to tissue ischemia, fibrosis, scarring, contractures, ulcerations, and even autoamputation in severe cases. Initial intervention is aimed at protection and cold avoidance. Many patients worsen and ultimately require medical treatment. Multiple treatment options have been attempted, including biofeedback, phosphodiesterase inhibitors, calcium channel inhibitors, and, recently, endothelin-1 receptor antagonists. Despite all of these options, a subset of patients develop severe, refractory symptoms that require more-invasive modalities.
Injection of neuromodulators such as botulinum toxin has been shown to be effective but short-lived,2 and can be an arduous undertaking for the patient and surgeon when performed in the clinic setting. Another option is surgical sympathectomy of the hands, which is not without its own inherent procedural risks,3 requires specialized training, and may not be long-lived with the regrowth of the stripped adventitia.
The intervention we propose is autologous fat grafting to hands affected by Raynaud phenomenon. The rationale stems from the clinical improvement witnessed in cosmetic rejuvenation of the hand, radiation dermatitis treated with fat grafting,4 and burn reconstruction. A published study on a radiation dermatitis murine model demonstrated regression of hyperpigmentation, ulcers, and fibrosis with histologic correlates.5 Recognizing that the pathophysiology in Raynaud phenomenon is different, we contend that the end result at the tissue level is similar and therefore suggest that this intervention may be of benefit.
PATIENTS AND METHODS
Indications included symptomatic Raynaud phenomenon with failure of medical management. The study was performed following approval by the University of Chicago Medical Center Internal Review Board (11-0383). Preauthorization was obtained from insurance companies in all cases.
We adapted a previously described method of cosmetic rejuvenation of the hand by means of fat grafting. Abdominal adipose donor sites were injected with approximately 200 ml of infiltration solution (50 ml of 1% lidocaine with epinephrine in 1 liter of normal saline). After waiting 15 minutes to allow the infiltrate to take effect, fat was harvested using either the Coleman catheter technique or the LipiVage (Genesis BioSystems, Lewisville, Texas) fat harvest system. The aspirate was left to decant, without the use of centrifugation. Our goal was to collect 30 ml of fat to be grafted, per recipient hand.
The fat grafting was performed using 3-ml syringes with 1-ml blunt cannulae after a stab incision was made with an 18-gauge needle. The graft was dispersed without prior infiltration of the hand, in 1-ml aliquots subcutaneously and in deeper planes, along the palmar, dorsal, and digital vessels as follows: 10 to 15 ml in the dorsum of the hand, 2 to 3 ml in the snuffbox, 1 to 2 ml in each dorsal webspace, 3 to 4 ml along the superficial palmar arch, 1 to 2 ml in volar webspaces 2 to 4, and 2 to 3 ml in the first webspace (Figs. 1 and 2). The incisions were closed with either absorbable suture or Octylseal (Medline Industries, Inc., Mundelein, Ill.) and Steri-Strips (3M, St. Paul, Minn.). Soft dressings were used for 5 days postoperatively.
Extremity perfusion was assessed preoperatively and after the intervention with the PeriCam PSI System (126.96.36.199; Perimed AB, Stockholm, Sweden), which is a rapid, noninvasive blood perfusion imager based on the laser speckle contrast analysis technology. PIMSoft software (188.8.131.5278; Perimed) was used to analyze the microcirculatory data.
Repeated measures analysis of variance was used to assess changes in laser speckle imaging measurements with treatment. Side (right versus left) and time (preoperatively versus postoperatively) were repeated factors. The average laser speckle contrast analysis measurements for the hand, the fingers, and palm were used in the analysis.
Between August of 2011 and November of 2012, we performed this procedure on 13 patients. Most of the patients were women (n = 12); nine had associated scleroderma, two had mixed connective tissue disorder, and two had primary Raynaud phenomenon (Table 1).
Table 2 shows data collected from 21 hands of 13 patients (five patients had a procedure performed on one hand, five patients had procedures on both hands with no repeated procedures, two patients had procedures on both hands with a repeated procedure needed on one of the hands, and one patient had procedures on both hands with repeated procedures needed on both). Repeated procedures were needed on four hands on three patients; duration of time between initial procedure and repeated procedure ranged from 9 to 15 months. Median age at the first procedure was 46 years (range, 36 to 68 years). Median duration of Raynaud phenomenon was 10 years (range, 2 to 35 years). For the 21 hands, the median total fat injected at the first procedure was 26 ml (range, 11 to 30 ml). Eleven of 21 hands (52 percent) had prior sympathectomy, and 20 of 21 hands (95 percent) had prior Botox. Median duration of follow-up for the 13 patients (from the date of the initial procedure to the last follow-up) was 17.9 months (range, 9.4 to 24.3 months).
Sixteen of the 21 hands (76 percent) had LipiVage during the initial procedure. Of the five hands that did not have LipiVage during the initial procedure, one had LipiVage during a repeated procedure. Four of the 21 hands (19 percent) had concurrent single-digit sympathectomy during the initial procedure (for critical ischemia). Of the 17 hands that did not have a concurrent sympathectomy during the initial procedure, one had a concurrent sympathectomy during a repeated procedure.
All analyses assume that two hands from the same patient are independent. For those hands with a repeated procedure, the pretreatment values (pain, functional level) used were those from the initial procedure, and the posttreatment values used were those from the repeated procedure.
Based on a per-hand analysis (n = 21), there was a significant improvement in pain after treatment (from 6.86 ± 1.98 to 2.38 ± 2.13; p < 0.001, Wilcoxon signed rank test). Eighteen of 21 had a decrease in pain; the remainder had no change in pain (Table 2).
There was also a significant improvement in number of cold attacks (p < 0.001, Wilcoxon signed rank test); 20 of 21 had a decreased number of attacks. There was a significant improvement in duration of cold attacks (p < 0.001), with 16 of 21 experiencing a decrease in duration. There was a significant improvement in severity of cold attacks (p < 0.001); 21 of 21 had decreased severity.
Twelve of 21 hands (57 percent) had ulcers before treatment; five of 21 hands (24 percent) had ulcers after treatment (p = 0.016, McNemar exact test for comparison of paired proportions). Seven of 17 hands (41 percent) that had LipiVage had ulcer improvement and one of four hands (25 percent) that did not have LipiVage had ulcer improvement (p = 1.00, Fisher’s exact test). One of five hands (20 percent) that had concurrent single-digit sympathectomy had ulcer improvement, whereas seven of 16 hands (44 percent) that did not have concurrent sympathectomy had ulcer improvement (p = 0.61) (Figs. 3 and 4).
In an attempt to stratify patients into severity groups, we devised a scoring system incorporating several parameters. The total score was calculated as the sum of “functional group” (0 = full function, 1 = somewhat limited hand function, or 2 = severely affected hand function), plus the presence of ulcers (0 = absent and 1 = present) plus pain (1 to 3 = 0, 4 to 6 = 1, and 7 to 10 = 2), plus the number of cold attacks per day multiplied by the duration of the attacks in minutes (0 to 60 minutes = 0, 61 to 150 minutes = 1, ≥151 minutes = 2). Thus, the total score could range from 0 to 7. The average pretreatment score was 4.81 ± 1.25 and the average posttreatment score was 1.57 ± 1.16, which represents a significant decrease (p < 0.001, Wilcoxon signed rank test for comparison of pretreatment versus posttreatment scores).
Examining whether the use of LipiVage impacted outcomes–with regard to improvement in overall perceived hand warmth, number of cold attacks per day, duration of cold attacks, and change in patient-reported functional capacity–none was found to have changed in a statistically significant manner. Reduction in pain score came closest to statistical significance, with a reduction of 4.88 points with LipiVage versus 2.75 with decanting (p = 0.13).
Ten of the 13 patients commented on the following statement: “I would recommend the treatment I have received so far to a friend with similar problems.” Eight of 10 patients (80 percent) responded “strongly agree” to this statement. One patient (10 percent) responded “agree,” and one patient responded, “neutral.” Of note, one patient changed her response from “agree,” to “neutral,” to ultimately “strongly agree” at longer follow-up appointments. On further follow-up, 12 of 13 patients stated they would recommend the procedure to a person with similar issues.
None of the patients suffered major complications such as severe infection, loss of digits, or vascular or tendon injury. One patient had a recurrence of ulcerations and increase in the severity of her cold hands 2 months after her initial grafting, and elected to undergo reoperation. The second procedure was complicated by acute carpal tunnel syndrome that required surgical decompression.
Two patients had minor complications: one patient had cellulitis at the fat harvest donor site treated with oral antibiotics and one patient had transient digital numbness. Despite having improvements in all parameters measured, four patients required further interventions at 3 to 6 months after fat grafting [two patients (three hands) underwent botulinum toxin injections, and two other patients had single-digit sympathectomies].
On laser speckle imaging, five patients showed improvement in perfusion unit counts (Fig. 5 and Table 3). Four of these five improved clinically as well. However, the laser imaging improvement was not statistically significant. Six hands had decreased laser speckle imaging measurements, yet four of those actually had improved symptomatology. There were 11 hands from nine patients used in the analysis. Using a repeated-measures analysis of variance with side (right, left) and time (preoperative, postoperative) as repeated factors, there was no significant change with intervention for fingers (p = 0.67), palm (p = 0.97), or hand (p = 0.74) (Fig. 6). All analyses presented here are based on the average measure for hand, fingers, or palm. Also, analyses were performed using natural log-transformed data because of evidence of nonnormality, but data are presented (summary statistics and plots) on the raw scale for ease of interpretation (Table 4).
Current pathophysiologic understanding of Raynaud phenomenon attributes the eventual digital ischemia to an imbalance between vasoconstriction and vasodilation.1 Primary Raynaud phenomenon is thought to be a benign, reversible vasospasm, whereas patients with systemic sclerosis have structural problems in addition to the functional abnormality. Patients without critical digital ischemia or ulceration are treated initially with general protective measures. Progression warrants medical interventions, typically with calcium channel blockers as a first-line agent. These may be supplemented or replaced with angiotensin-receptor blockers, angiotensin-converting enzyme inhibitors, alpha-blockers, selective serotonin reuptake inhibitors, phosphodiesterase inhibitors, or topical nitrates, with or without aspirin or clopidogrel. Intravenous prostanoids are used in cases of severe symptoms, such as critical ischemia and digital ulcerations. Endothelin-1 receptor antagonists are now approved for use in advanced disease. However, many patients will require surgical intervention in the form of débridement of necrotic and infected digits and even amputation. Digital palmar sympathectomy has a role in prevention of progression to amputation, although the reported data are few and this procedure requires specialized care not widely available. Case series of botulinum toxin have been reported but have yet to be thoroughly investigated.
Descriptions of fat grafting to the dorsum of the hands can be found in the dermatology literature from the 1980s onward.6–8 Early techniques used placement of harvested fat subcutaneously with closed manipulation to spread the graft to the desired areas of the hand and digits. The original intent of this procedure was to obtain a more youthful appearance by camouflaging prominent veins and tendons. Coleman9 refined structural fat grafting by injecting small parcels of centrifuged fat (2 of 4 ml total) with blunt cannulas through six or seven incisions on the hand (a total of 20 to 30 ml), into the subdermal plane (superficial to the superficial veins of the hand). Multiple passes of the injection cannulas and digital massage allow even distribution of the fat. Coleman reports acceptable long-term results with softening of the color and definition of the veins and tendons. However, it remains unclear how much fat volume is retained over an extended period; it is generally presumed that 40 to 60 percent of the injected fat is absorbed. Giunta et al., in 2010, demonstrated that 69 percent of the injected volume remained at a 6-month three-dimensional laser scanning imaging follow-up.10
Several reports have shown clinical improvement in radiation-induced fibrosis following fat injection in the context of breast cancer reconstruction. The precise mechanism remains unclear: it is proposed to be related to adipose-derived stem cells present in the stromal vascular fraction of the fat graft. These cells may stimulate growth factors, such as vascular endothelial growth factor, that promote proliferation of endothelial cells and inhibit their apoptosis, ultimately leading to angiogenesis to improve perfusion.11 The presence of adipose-derived stem cells may also interfere with overactive scarring cascades. Proteins such as transforming growth factor-β play a role in the progression of wound healing and fibrosis through the Smad signaling cascade.12 By altering an activated profibrotic response, adipose-derived stem cells may serve to improve wound healing by interfering with this pathway. This may lead to accelerated wound healing with less fibrosis. Further research needs to be performed to better elucidate that exact mechanism of action.
A recently published murine model study demonstrated regression of radiation-induced ulcers and gradual resolution of hyperpigmentation, and histologic correlates were observed with improvements in the inflammatory, microvascular, and fibrotic characteristics of the treated tissues. Effects were observed in both the acute and chronic stages of radiodermatitis. Chronic radiodermatitis is characterized by fibrotic changes, ulceration, and necrosis (endpoints similar to those in advanced Raynaud phenomenon, albeit of different cause). Sultan et al.5 provided histologic evidence of slowing of progression of fibrosis in early chronic radiodermatitis following fat injection. The mechanism is as of yet unclear. In vitro studies are underway to determine the potential role of repletion of stem cells that have been shown to be susceptible to radiation injury.
Multiple studies across numerous fields of medicine seek to use stem cells of various origins to treat ischemia models (in limbs,13 neural tissue, myocardium,14 and kidney).15 To our knowledge, application of fat grafting to treat ischemic complications of Raynaud phenomenon has not been described. Although the effects of stem cells on ischemic tissue remain hypothetical and the mechanisms to explain positive results have not been elucidated, we believe that applying a safe and well-recognized procedure may have beneficial effects on the ischemic digits in the setting of advanced Raynaud phenomenon.
Results from this study are promising. A majority of the patients experienced some improvement in their symptoms. This is particularly evident from our patients’ reports demonstrating that 92.3 percent of the patients (12 of 13) would “agree” or “strongly agree” with recommending this treatment to a friend with similar problems. This illustrates the high level of satisfaction achieved through fat grafting. In particular, for the patient who changed their response from “agree,” to “neutral,” to ultimately “strongly agree” with recommending this procedure, the changes in response may indicate the long-lasting and enduring improvement in symptoms and improvement in vascularity. It is also worthwhile to note that no patient disagreed or strongly disagreed with recommending this procedure to a friend with similar problems.
Decreases in pain and cold attacks were observed following fat grafting in many patients. Limitations of our study include the inconsistent laser speckle imaging assessments, which demonstrated perfusion values that were at times different than clinical findings. This may be attributable to the variable room temperatures when the test was performed, or the seasonal climate differences at the time of testing. Laser speckle contrast analysis is a relatively new modality, and previous studies have found support in its ability to quantify microvasculature values and changes.16 However, normal values of perfusion units per anatomical site have yet to be established or validated. The laser imaging measurements obtained here may serve as a baseline for future investigation. A more precise modality will need to be identified to properly assess whether there is an increase in perfusion following fat grafting that correlates with the clinical examination findings. In addition, the scoring system used here has not been validated, although it is similar to the Raynaud Classification Score,17 and was devised in a retrospective fashion. We intend to prospectively assess the parameters examined in this study to further evaluate the immediate and long-lasting effects of fat grafting to the hands in Raynaud phenomenon patients.
Fat grafting in patients with Raynaud phenomenon of varying degrees provides a durable clinical improvement in the majority of treated patients. It appears to be a safe and relatively straightforward intervention. The precise mechanism has yet to be elucidated–perhaps it involves neoangiogenesis and decreased scar promoted by transferred stem cells. We believe that in scleroderma patients, fat grafting at the very least provides advantageous “padding” to the severely sclerotic digits and relieves some of these patients’ symptoms. We believe that there is a continued role for fat grafting in advanced, refractory Raynaud phenomenon, and are working toward further characterizing and assessing this treatment option.
This information provided by Dr. Raymond Janevicius is intended to provide coding guidance.
20926 Fat graft
* Fat grafting is reported with code 20926.
* Code 20926 is global and includes:
* Harvest of the fat graft material by any method (e.g., syringe, suction assisted lipectomy)
* Processing of fat graft material
* Infection of fat graft into the recipient site
* Bandaging of the donor and recipient sites
* Code 20926 is not volume dependent. Thus 20926 is reported once no matter how much fat is injected.
* Report 20926 once for all injections in each hand. It is not to be reported for each individual injection.
* If both hands are injected, report
* Even though the procedures are “bilateral” clinically, 20926 is not considered a “bilateral code,” so do not append modifier 50, as it is only appropriate for certain CPT codes.
* This procedure should be pre-authorized in writing with the insurance company prior to surgery. Code 20926 is often considered a cosmetic code. The pre-authorization letter should clearly explain that this is a reconstructive procedure and explain its indications, including appropriate ICD-9 codes, such as:
443.0 Raynaud’s phenomenon
883.1 Wound of finger, complicated
882.1 Wound of hand, complicated
729.5 Pain in extremity