https://orcid.org/0000-0002-3330-8525
Introduction
Surgical treatment of fistulizing perianal Crohn’s disease (pCD) in patients with failed medical therapy is challenging with relative high recurrence and failure rates.1 The disease has a significant impairment on quality of life1 and carries a relatively high risk of proctectomy.2 The risk of developing a Crohn’s perianal fistula has not altered during the last 2 decades despite the increased prescription of biologics.3
The use of stem cell injection in or around the fistula has been shown to be safe and feasible—but with variable healing rates.4,5 Encouraging results have been reported with autologous adipose tissue graft combined with surgical closure of the internal fistula opening.6 The stromal vascular fraction (SVF), also referred to as adipose derived regenerative cells (ADRCs), is isolated from freshly harvested autologous adipose tissue (by liposuction). The SVF includes several cell types with regenerative potential, as demonstrated in experiments of erectile dysfunction following radical prostatectomy7 and breast cancer–related lymphedema.8
The aim of the present study was to evaluate the outcome and safety of treating fistulizing pCD with autologous fat graft enriched with ADRC. The primary outcome was the healing rate defined as absence of discharge and closure of external fistula opening by clinical examination. The secondary outcomes were time to healing (weeks), fecal incontinence measured by Wexner Fecal Incontinence Score,9 and radiological recurrence of fistula on magnetic resonance imaging (MRI) scanning.
Materials and Methods
The study was designed as a prospective single-center pilot study and was conducted between June 2018 and December 2020. The inclusion criterion was adult patient (18 years and older) with fistulizing pCD not responding to medical therapy for at least 6 months and a loose seton suture for at least 3 months. Exclusion criteria were multiple (2 or more tracts) fistulas, anal stenosis, suppuration around the fistula tract, a subcutaneous perianal fistula, active intestinal Crohn’s disease not in remission [Active Crohn’s disease not in remission is based on serum biomarkers and colonoscopy findings.], body mass index (BMI) <18.5, coagulopathy, previous radiotherapy to the abdomen and pelvis, any malignancy within 5, years and verified infection on screening test.
After patients gave informed consent, all included patients were scheduled for preoperative workup with MRI scanning of pelvis/rectum; serological screening test for syphilis, hepatitis, and HIV; and physical examination under general anesthesia including transanal ultrasound with anatomical mapping of the fistula tract. Patients’ current Crohn’s medical therapy was not altered or postponed during the period of treatment and follow-up.
Surgical Procedures and ADRC Preparation
All procedures were performed at the same day as ambulant surgery. The liposuction and ADRC isolation have been previously described in detail.7,8 Liposuction was followed by surgical debridement of the fistula tract and double-layered closure of the internal opening using absorbable suture material. The external opening was excised and left open for drainage. Then 30-50 milliliters of the freshly harvested lipoaspirate were injected around the entire length of fistula tract, taking care not to perforate the fistula tract (Figures 1A-E). The patient was observed at the recovery unit for 120-150 minutes. Meanwhile, ADRCs were isolated from the remaining lipoaspirate using an automated processing Celution 800/IV system (Lorem Cytori, San Diego, California, USA) according to the manufacturer’s instructions. The patient was then taken back to the operating theater, and 4mL of isolated ADRCs was injected and evenly distributed around the fistula tract corresponding to the locations of previous injections of lipoaspirate under mild sedation with propofol.
A, A high transsphincteric Crohn’s anal fistula on the right side. B, Liposuction from anterior abdomen using waterjet assisted liposuction (body-jet, Human med AG, Schwerin, Germany). C Lipoaspirate (upper layer) harvested with waterjet liposuction from the anterior abdomen. D, While the lipoaspirate left to sediment at the operating theater, the fistula is curated, and internal opening is closed with 2-layer absorbable sutures. E, The lipoaspirate is directly injected around the entire length of fistula tract. When the ADRC suspension is ready, it is injected into the same site of lipoaspirate injection. F, Complete healing at 6-month follow-up.
The patients were discharged the same day with a prescription of postoperative oral antibiotics (metronidazole 1500mg/day and cefuroxime 1000mg/day) for 7 days.
Follow-up
All included patients were scheduled for 3 follow-ups at 2 weeks, 3 months, and 6 months for clinical evaluation of healing (inspection, digital palpation, and anoscopy), as well as Wexner Fecal Incontinence Score. The MRI scanning was performed at 6-month follow-up for patients with clinical healing and no sign of recurrence at physical examination. Whenever there was a suspicion of recurrence or abscess formation, examination under general anesthesia was performed. Further clinical follow-up for healing and recurrence was done at 12 months for patients who achieved clinical and radiological healing at 6-month follow-up.
Statistical Analysis
Continuous and categorical variables were analyzed using descriptive statistical tests (t test, Wilcoxon rank-sum test [Mann-Whitney], and Pearson correlation [r]) when appropriate. All P values below .05 were considered significant. Stata statistical software package version 16.0 and GraphPad Prism 9 were used.
Ethical Considerations
The study was conducted in accordance with the rules of the Helsinki declaration. The study was approved by the local research ethics committee (S-20170140). The ADRC isolation was performed at an authorized tissue establishment (Authorization no. 29035, Department of Clinical Biochemistry and Pharmacology, Odense University Hospital). The study was registered at ClinicalTrials.org (identification number NCT03466515).
Results
Demography and Surgical Data
Of the 76 patients with fistulizing pCD evaluated for surgical treatment during the study period, 13 patients were found eligible for the study. One patient was excluded due to exacerbation of Crohn’s disease and weight loss (BMI below 18.5) despite intensive medical treatment. Twelve patients received the intended treatment and completed the observation period.
Table 1 shows the demographic characteristics of the study population and the results of cell analysis. All but 3 patients received adjuvant medical therapy. The mean length of fistula was 4.5cm (95% confidence interval [CI], 3.2–5.8), all fistulas were high transsphincteric, and there were 2 external openings in 2 patients. One patient had a diverting colostomy. The mean volume of lipoaspirate injected was 40mL (95% CI, 31–49.6). The average number of injected ADRC cells was 37.6 million (95% CI, 29.3–45.8).
Study population.
| Patients’ Demographic Characteristics (N=12) | |
|---|---|
| Gender | Male=3, Female=9 |
| Mean age in years (range) | 33 (22–51) |
| Mean BMI (range) | 27.9 (20.07–37.03) |
| Smoking habits | |
| Smokers=3, Quit=3, Non-smokers=6 | |
| Alcohol consumptiona | |
| Within recommended | 11 |
| Above recommended | 1 |
| Co-morbidityb | 1 |
| Mean duration of fistula in months (range) | 42.2 (8–132) |
| Mean duration of Crohn’s disease in years (range) | 8.3 (1–23) |
| Medical therapy | 9 |
| Anti-TNF alone | 6 |
| Anti-TNF and Azathioprine | 3 |
| Mean duration of medical therapy in months (range) | |
| Anti-TNF | 25 (2–108) |
| Azathioprine | 108 (48–168) |
| Previous surgery | Anal (2), intestinal (4) |
| ADRC characteristics | |
| Mean lipoaspirate total volume (95% CI) | 263mL (220–307) |
| Nucleated ADRC/g fat tissue (95% CI) | 2.59×105 (1.93–3.24) |
| Mean lipoaspirate for ADRC isolation (95% CI) | 233mL (174–292) |
| Mean volume of lipoaspirate injected was (95% CI) | 40mL (31–49.6) |
| Average number of injected ADRC cells (95% CI) | 37.6×106(29.3–45.8) |
| ADRC subpopulation | |
| CD34 mean (95% CI):58.5% (55.8–61.3) | CD31 mean, 30.5% (95% CI, 0.5–60.6) |
| CD90 mean (95% CI):53.9% (39.1–68.6) | CD73 mean, 39.5% (95% CI, 24.4–54.9) |
| Erythrocytes %:5.8–57.8% | |
| Patients’ Demographic Characteristics (N=12) | |
|---|---|
| Gender | Male=3, Female=9 |
| Mean age in years (range) | 33 (22–51) |
| Mean BMI (range) | 27.9 (20.07–37.03) |
| Smoking habits | |
| Smokers=3, Quit=3, Non-smokers=6 | |
| Alcohol consumptiona | |
| Within recommended | 11 |
| Above recommended | 1 |
| Co-morbidityb | 1 |
| Mean duration of fistula in months (range) | 42.2 (8–132) |
| Mean duration of Crohn’s disease in years (range) | 8.3 (1–23) |
| Medical therapy | 9 |
| Anti-TNF alone | 6 |
| Anti-TNF and Azathioprine | 3 |
| Mean duration of medical therapy in months (range) | |
| Anti-TNF | 25 (2–108) |
| Azathioprine | 108 (48–168) |
| Previous surgery | Anal (2), intestinal (4) |
| ADRC characteristics | |
| Mean lipoaspirate total volume (95% CI) | 263mL (220–307) |
| Nucleated ADRC/g fat tissue (95% CI) | 2.59×105 (1.93–3.24) |
| Mean lipoaspirate for ADRC isolation (95% CI) | 233mL (174–292) |
| Mean volume of lipoaspirate injected was (95% CI) | 40mL (31–49.6) |
| Average number of injected ADRC cells (95% CI) | 37.6×106(29.3–45.8) |
| ADRC subpopulation | |
| CD34 mean (95% CI):58.5% (55.8–61.3) | CD31 mean, 30.5% (95% CI, 0.5–60.6) |
| CD90 mean (95% CI):53.9% (39.1–68.6) | CD73 mean, 39.5% (95% CI, 24.4–54.9) |
| Erythrocytes %:5.8–57.8% | |
Abbreviations: BMI, Body Mass Index; ADRC, adipose-derived regenerative cells; CI, confidence interval.
The recommended weekly alcohol consumption by the Danish health council.
Only 1 patient had significant comorbidity (impaired renal function).
Study population.
| Patients’ Demographic Characteristics (N=12) | |
|---|---|
| Gender | Male=3, Female=9 |
| Mean age in years (range) | 33 (22–51) |
| Mean BMI (range) | 27.9 (20.07–37.03) |
| Smoking habits | |
| Smokers=3, Quit=3, Non-smokers=6 | |
| Alcohol consumptiona | |
| Within recommended | 11 |
| Above recommended | 1 |
| Co-morbidityb | 1 |
| Mean duration of fistula in months (range) | 42.2 (8–132) |
| Mean duration of Crohn’s disease in years (range) | 8.3 (1–23) |
| Medical therapy | 9 |
| Anti-TNF alone | 6 |
| Anti-TNF and Azathioprine | 3 |
| Mean duration of medical therapy in months (range) | |
| Anti-TNF | 25 (2–108) |
| Azathioprine | 108 (48–168) |
| Previous surgery | Anal (2), intestinal (4) |
| ADRC characteristics | |
| Mean lipoaspirate total volume (95% CI) | 263mL (220–307) |
| Nucleated ADRC/g fat tissue (95% CI) | 2.59×105 (1.93–3.24) |
| Mean lipoaspirate for ADRC isolation (95% CI) | 233mL (174–292) |
| Mean volume of lipoaspirate injected was (95% CI) | 40mL (31–49.6) |
| Average number of injected ADRC cells (95% CI) | 37.6×106(29.3–45.8) |
| ADRC subpopulation | |
| CD34 mean (95% CI):58.5% (55.8–61.3) | CD31 mean, 30.5% (95% CI, 0.5–60.6) |
| CD90 mean (95% CI):53.9% (39.1–68.6) | CD73 mean, 39.5% (95% CI, 24.4–54.9) |
| Erythrocytes %:5.8–57.8% | |
| Patients’ Demographic Characteristics (N=12) | |
|---|---|
| Gender | Male=3, Female=9 |
| Mean age in years (range) | 33 (22–51) |
| Mean BMI (range) | 27.9 (20.07–37.03) |
| Smoking habits | |
| Smokers=3, Quit=3, Non-smokers=6 | |
| Alcohol consumptiona | |
| Within recommended | 11 |
| Above recommended | 1 |
| Co-morbidityb | 1 |
| Mean duration of fistula in months (range) | 42.2 (8–132) |
| Mean duration of Crohn’s disease in years (range) | 8.3 (1–23) |
| Medical therapy | 9 |
| Anti-TNF alone | 6 |
| Anti-TNF and Azathioprine | 3 |
| Mean duration of medical therapy in months (range) | |
| Anti-TNF | 25 (2–108) |
| Azathioprine | 108 (48–168) |
| Previous surgery | Anal (2), intestinal (4) |
| ADRC characteristics | |
| Mean lipoaspirate total volume (95% CI) | 263mL (220–307) |
| Nucleated ADRC/g fat tissue (95% CI) | 2.59×105 (1.93–3.24) |
| Mean lipoaspirate for ADRC isolation (95% CI) | 233mL (174–292) |
| Mean volume of lipoaspirate injected was (95% CI) | 40mL (31–49.6) |
| Average number of injected ADRC cells (95% CI) | 37.6×106(29.3–45.8) |
| ADRC subpopulation | |
| CD34 mean (95% CI):58.5% (55.8–61.3) | CD31 mean, 30.5% (95% CI, 0.5–60.6) |
| CD90 mean (95% CI):53.9% (39.1–68.6) | CD73 mean, 39.5% (95% CI, 24.4–54.9) |
| Erythrocytes %:5.8–57.8% | |
Abbreviations: BMI, Body Mass Index; ADRC, adipose-derived regenerative cells; CI, confidence interval.
The recommended weekly alcohol consumption by the Danish health council.
Only 1 patient had significant comorbidity (impaired renal function).
ADRC Characterization
The cell yields were neither affected by patients’ gender (P = .911) nor by medical therapy (P = .552) in the present study. Medical therapy did not have an impact on any of the analyzed cell parameters that all were within normal range. A negative correlation was observed between the percentage of erythrocytes and percentage of colony forming units-fibroblasts (CFU-F) (P = .0088) and CD31-CD34+ stromal cells (P = .0209), suggesting that the total yield of stromal cells may be affected by hemorrhage. As for all other cell parameters, the number of injected nucleated, nonhematopoietic CD31-CD34+ ADRCs was not significantly different between responders and nonresponders.
Recurrence and Healing
Fistula recurrence occurred in 3 patients (2 on combined azathioprine and antitumor necrosis factor [anti-TNF] treatment and 1 on anti-TNF alone) within the 2-weeks of follow-up, mainly due to failure of the closure of the internal opening. Nine patients (75%) had a complete closure of the fistula tract and were free of symptoms at 6-months (Figure 1F) and 1-year follow-up. Apart from the recurrences, there were no serious surgical complications (Calvien-Dindo grade 3 or above) observed in any cases. Wound healing (excised external fistula opening) was achieved in 8 of the 12 patients (67%) at 12-weeks of follow-up. The complete healing of the fistula was confirmed by MRI scanning at 6-months of follow-up in 8 of the 9 patients, and the last patient had regression of the fistula without fluid collection.
Fecal Incontinence
The mean Wexner Fecal Incontinence Score was reduced by more than 50% at 6-month follow-up (1.4) compared to baseline measurement (5.4). (See Figure 2).
Mean profile of differences in Wexner Fecal incontinence Score over the observation period.
Discussion
The present study showed 75% clinical healing and 67% radiological healing of the fistula at 6-month follow-up and no clinical sign of recurrence at 1-year follow-up. Similar results were shown by Serrero et al10 using a comparable method, with clinical response of 80% and fistula healing confirmed by MRI scanning in 60% at 48-week follow-up. Apart from minor differences in ADRC isolation and characterization, both studies applied the same treatment principle with comparable outcomes, suggesting that the method is robust. A negative correlation between hemorrhage due to liposuction and the yield of CD34+ cells was shown in the present study, suggesting that a regenerative effect cannot be ascribed to this population alone; although it should be interpreted with caution due to the restricted number of participants.
Unaltered Crohn’s medical therapy during the study might have a positive influence on the results of the study due to continuous long effect of medical therapy. A similar positive influence could also be attributed to the meticulous surgical closure of the internal opening, inasmuch as there was special attention given to avoid disruption of the fistula tract during fat and ADRC injection. The improvement in Wexner score can be explained by healing of the fistula, leading to cessation of secretion and the need of diapers.
The commercially available product Cx601(darvadstrocel) has been shown to have clinical and radiological remission in 56% after 1 year compared with 35% with placebo in patients with Crohn’s disease and a single fistula tract.11 The present study strongly supports an alternative approach with the advantage of using autologous, ADRC-enriched, freshly collected lipoaspirate injection.
The major limitations of the study are the small population of selected patients, lack of control group, and a short observation time. Further evaluation of the results of the present study in a larger setting in a randomized trial is highly recommended,
Conclusions
Autologous lipoaspirate enriched with ADRCs can be safely used in the treatment of high anal fistula in patients with Crohn’s disease with high rate of success.
Acknowledgments
Authors thank Tina Kjærgaard Andersen and Tonja Lyngse Jørgensen (LMCC, Odense University Hospital) for excellent technical assistance in relation to ADRC isolation and characterization.
Funding
This research was funded by The University of Southern Denmark (16/45767), the Region of Southern Denmark (15/50967), Surgical Department A, Odense University Hospital, Odense University Hospital Foundation, the Danish Crohn Colitis Association CFF (Forskningstøtte 2016), and the Fionia Foundation (January 22, 2016).
Conflicts of Interest
S.P.S. is owner and CEO of Blue Cell Therapeutics. The other authors declare that they have no conflicts of interest.
References
