Intra-arterial renal infusion of autologous mesenchymal stem cells for treatment of chronic kidney disease in cats: Phase I clinical trial.

BACKGROUND: There are no known treatments that halt or reverse chronic kidney disease (CKD) in cats. In rodent models, stem cell treatment has been associated with improvement in renal function parameters, especially when stem cells were delivered intra-arterially to the kidney. To date, only IV and intrarenal stem cell infusions have been studied in cats with CKD with no clinically relevant improvement noted.
OBJECTIVE: To assess the safety and feasibility of intra-arterial delivery of autologous mesenchymal stem cells (MSC) in stromal vascular fraction (SVF) to the kidney in cats with CKD. ANIMALS: Five client-owned domestic cats with International Renal Interest Society stage III CKD.
METHODS: Prospective cohort study (phase I clinical trial). Adipose tissue was harvested from study animals on day 0. On days 2 and 14, an infusion of MSC in SVF was administered into the renal artery via the femoral or carotid artery using fluoroscopic guidance. Serum creatinine and blood urea nitrogen concentration, plasma iohexol clearance, and quality of life assessments were monitored between days 0 and 90.
RESULTS: The procedure was performed successfully in all cats. No severe adverse events were observed in any cat during the study period. CONCLUSIONS AND CLINICAL IMPORTANCE: Intra-arterial infusion of MSC into the renal artery in CKD cats was feasible and safe within a 3-month postoperative period. Efficacy and long-term safety have yet to be established. This procedure requires careful technique and training.

acute kidney injury

chronic kidney disease

glomerular filtration rate

intra‐arterial

International Renal Interest Society

IV fluid therapy

mesenchymal stem cells

phosphate‐buffered saline

serum creatinine concentration

stromal vascular fraction

urine protein‐to‐creatinine ratio

INTRODUCTION

Chronic kidney disease (CKD) is a progressive degenerative disease that is common in cats. To date, no known treatments that stop disease progression or repair affected kidneys have been identified. Mesenchymal stem cells (MSC) are being explored as a treatment for CKD in people and animals. In addition to their well‐known ability to differentiate into specialized cells, MSC are capable of modulating inflammatory responses and mediating cell‐cell interactions to promote tissue repair.1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12 They have a “homing” ability in which they travel through the circulation to sites of inflammation, attracted by cytokines and chemokines released from injured tissue.13, 14, 15 They create anti‐inflammatory, pro‐angiogenic, anti‐fibrotic, and anti‐apoptotic effects within the local tissue environment by releasing growth factors, cytokines, and chemokines that provide trophic support for damaged tissue, which leads to regeneration and differentiation of injured cells and, ultimately, tissue recovery. 8, 15, 16, 17, 18

Numerous studies in rodent models of kidney disease have identified decreased inflammation, fibrosis, and glomerulosclerosis after MSC treatment.19, 20, 21, 22 In a meta‐analysis of 21 studies on MSC treatment in rat and mouse models of CKD and acute kidney injury (AKI), an overall decrease in serum creatinine concentration (SCr) was identified in treated animals as compared to controls, and intra‐arterial (IA) delivery was associated with a greater decrease in SCr than that seen with IV and intrarenal administration.23 Route of delivery may be critical for MSC treatment given what is known about stem cell behavior. In rodents, the majority of stem cells administered IV become trapped in the lungs because the pulmonary capillaries are the first to receive cells after injection.24 This “first‐pass effect,” plus the homing ability of MSC that draws them to various sites of inflammation in the body, may decrease the total number of MSC received by the kidneys. When compared with IV administration, IA delivery of MSC via the renal artery led to higher engraftment of cells in the kidneys of sheep after renal ischemia‐reperfusion injury.25 This may account for the difference in efficacy between IV and IA MSC infusion.

The few published studies on MSC treatment in cats with CKD have evaluated IV or intrarenal delivery of stem cells.26, 27, 28 Although safety of MSC treatment was demonstrated repeatedly, improvement in renal function was not a consistent finding. We therefore developed a 2‐phase study to assess the safety and efficacy of IA MSC treatment in cats with advanced CKD. The objective of phase I, the subject of the present study, was to assess the safety and feasibility of the procedure. We hypothesized that: (1) IA administration of autologous MSC to the kidney by selective catheterization of the renal artery would be possible by peripheral arterial access using fluoroscopic guidance and interventional radiology techniques and (2) IA autologous MSC delivery would be safe, with no major perioperative complications or clinically relevant worsening of kidney disease within 90 days of treatment. Upon successful completion of phase I, all study animals were automatically enrolled in phase II, the purpose of which was to determine the efficacy of IA MSC administration as treatment for CKD in cats.

MATERIALS AND METHODS

Animals

Client‐owned domestic cats with International Renal Interest Society (IRIS) stage III CKD (all substages), defined as stable disease with SCr between 2.9 and 5 mg/dL over a 3‐month period, were considered for enrollment. A full diagnostic evaluation including physical examination, medical history taken by 1 of the study's principal investigators (Allyson C. Berent or Catherine E. Langston), Doppler blood pressure, CBC, serum biochemistry, serum total T4 concentration, feline immunodeficiency virus antibody and feline leukemia virus antigen ELISA (SNAP FIV/FeLV Combo Test; IDEXX Laboratories, Inc, Westbrook, Maine), Toxoplasma titers, focused ultrasound examination of the kidneys and ureters including renal blood flow, abdominal radiographs, chest radiographs, echocardiogram (if abnormal heart sounds or rhythm were noted), urinalysis, urine bacterial culture, and urine protein‐to‐creatinine ratio (UP : C ratio) were performed on each animal before inclusion in the study. Exclusion criteria included history of urolithiasis, history of uremic crisis, anemia (PCV <25%), current positive urine bacterial culture, existence of clinically relevant comorbidities other than CKD, history of other experimental treatments for CKD, and history of corticosteroid, recombinant human erythropoietin, or calcitriol use. The first 6 patients with a complete diagnostic evaluation that met the above criteria were enrolled in phase I. At enrollment, each patient had a plasma iohexol clearance study performed. Starting at enrollment, all patients were on a consistent diet, medication, and supplement regimen through day 90. A renal specific diet was not mandatory because of prior failure of compliance.

Approval for the study protocol by the Institutional Animal Care and Use Committee and client consent were obtained before patient enrollment.

Adipose tissue harvesting

All cats were hospitalized for IV fluid therapy (IVF) before anesthesia for adipose tissue harvesting. Upon admission to the hospital on day −1, an IV catheter was placed, and IVF was administered to ensure euhydration before anesthesia. On day 0, all cats were placed under general anesthesia using commonly accepted protocols for patients with CKD. Standard anesthetic monitoring was performed, and mean arterial blood pressures were maintained >60 mm Hg. The abdomen was clipped and aseptically prepared for mini‐laparotomy. A 3‐ to 4‐cm midline incision was made on the ventral abdomen cranial to the umbilicus, and falciform adipose tissue with or without omental adipose tissue was removed using electrocautery, suture, or an electrothermal bipolar tissue sealing system (LigaSure; Covidien, Dublin, Ireland). At collection, adipose tissue was placed into a sterile 50 mL conical tube containing 15 mL of phosphate‐buffered saline (PBS) without calcium or magnesium, packaged in a validated, temperature‐controlled (2°C‐8°C) shipping container and shipped overnight to the stem cell processing laboratory (VetStem Biopharma, Poway, California) for MSC isolation, viability testing, and quantification. All cats were maintained on IVF postoperatively through day 3.

Adipose tissue processing and isolation of stromal vascular fraction

Adipose tissue was processed within 36 hours of collection. Tissue was washed with PBS, minced, and rewashed with PBS. The minced tissue was thoroughly mixed, and enzymatic digestion was achieved using a combination of collagenase and hyaluronidase at 37°C for 50 minutes with agitation. The mixture was centrifuged at 500‐700 relative centrifugal force for 15 minutes, and the cell pellet was resuspended in PBS 4 times. An aliquot of the final cell suspension was assessed for viability and total nucleated cell count using a nucleocounter (Mediatech). A target dose of 3 × 106 cells was sought with doses between 1.5 and 6 × 106 cells considered acceptable. Cells were brought to a final volume of 3 mL with PBS and loaded into a sterile syringe for injection that was shipped back to the hospital as described above and delivered on day 2.

Cells for future treatments were prepared for cryopreservation using a proprietary solution with 10% dimethyl sulfoxide and gradually brought to a temperature of ≤−130°C. Relative to the first dose of stromal vascular fraction (SVF), a larger number of cells from each patient were cryopreserved to account for anticipated cell loss during freezing, thawing, and washing of cells. Cells were stored in vapor‐phase liquid nitrogen. Doses of MSC were recovered from cryopreservation using a proprietary solution containing PBS, autologous serum, and Iscove's Modified Dulbecco's Medium (Hyclone Laboratories, Logan, Utah). Once recovered, an aliquot was assessed for viability and total nucleated cell count using a nucleocounter. The same target dose and acceptable dose range were used. Recovered cells were brought to a final volume of 3 mL with the recovery solution and shipped back to the hospital as described above for injection on day 14.

Intra‐arterial MSC infusion

On day 2, all patients were placed under general anesthesia, and MSC were injected into either the right or left renal artery via a femoral or carotid arterial cut‐down approach. The side of injection chosen was either the larger of the 2 kidneys (if >25% difference in sagittal measurement was noted on ultrasound examination) or decided by coin toss (heads = left kidney, tails = right kidney). The site of IA access (femoral or carotid artery) was chosen based on clinician preference. The second treatment was administered in an artery not utilized for the first treatment (eg, carotid then femoral, femoral then carotid, left then right femoral).

For the femoral approach, the inguinal region was clipped and aseptically prepared. Using sterile technique, a 2‐cm incision was made over the femoral artery. The femoral artery was isolated and ligated distally using 3‐0 polydioxanone suture. A 22‐gauge IV catheter was used to puncture the femoral artery. A 0.018″ angle tipped hydrophilic guide wire (Weasel Wire; Infiniti Medical, LLC, Menlo Park, California) was advanced through the 22‐gauge catheter to the external iliac artery and the descending aorta. After removing the 22‐gauge IV catheter over the guide wire, a 4 French micropuncture set (Infiniti Medical, LLC, Menlo Park, California) was used to dilate the vessel and used as a 4 Fr sheath by adapting a Tuohy borst adaptor (Cook Medical, Bloomington, Indiana) over the hub. A 2.4 Fr microcatheter (Infiniti Medical, LLC, Menlo Park, California) was placed through the sheath over the 0.018″ wire or the wire was exchanged with a 0.014″ guide wire (Infiniti Medical, LLC, Menlo Park, California) depending on the internal diameter of the catheter being used. The catheter‐guide wire combination was used to select the appropriate renal artery. A volume of 0.25‐0.5 mL of 50% dilution contrast solution (1 : 1 iohexol : sterile saline) was injected through the catheter into the renal artery to obtain a renal angiogram using digital subtraction angiography in order to document proper catheter location and good renal blood flow (Figure 1).

Figure 1

Fluoroscopy images of 2 cats during intra‐arterial mesenchymal stem cell infusion. The cats are in dorsal recumbency. The top of each image is cranial. A, Femoral arterial access with a guide wire (white arrow) in the right renal artery and a 3 Fr catheter (black arrow) being advanced over the wire into the renal artery from the aorta. B, The catheter advanced over the wire into the renal artery (black arrow). C, Angiogram using digital subtraction angiography in the right renal artery. D, Carotid arterial access with a microcatheter (black arrow) in the aorta, entering the left renal artery during a renal arteriogram

Evaluation of renal arterial anatomy and filling was performed. Any contrast remaining in the catheter then was withdrawn and discarded, and the catheter was irrigated with 1 mL of saline. A 3‐mL aliquot of MSC solution was injected into the catheter and renal artery over 3 minutes using an 18‐μm Hemo‐Nate filter (Utah Medical Products, Inc, Midvale, Utah) that previously was tested by the stem cell company and shown to cause no loss of viable cells when compared to control syringes without filters (unpublished data). After infusion, another 1.5 mL of saline was infused through the filter slowly to move cells out of the catheter. Finally, 0.5 mL of the 50% contrast mixture was infused into the renal artery to confirm arterial patency. Once patency was confirmed, the catheter and sheath were removed and the femoral artery was ligated. The incision was closed with 3‐0 poliglecaprone 25 suture material.

For the carotid approach, the patient was placed in dorsal recumbency. The neck was extended, clipped, and aseptically prepared. A 3‐cm incision was made in the right jugular groove between the jugular vein and trachea. The carotid sheath was identified, and the carotid artery was isolated. The cranial aspect was ligated using 3‐0 polydioxanone suture, and an 18‐gauge IV catheter was used to cannulate the artery. A 0.035″ angle tipped hydrophilic guide wire (Weasel Wire; Infiniti Medical, LLC, Menlo Park, California) was advanced through the catheter into the external carotid artery, brachiocephalic trunk, and descending aorta. The 18‐gauge catheter was removed over the wire, and a 4 Fr vascular access sheath (Infiniti Medical, LLC, Menlo Park, California) was advanced over the guide wire and sutured to the skin. The dilator then was removed. A microcatheter over a 0.018″ or 0.014″ guide wire was advanced through the sheath and down the descending aorta. The guide wire‐catheter combination was used to cannulate the selected renal artery, and the MSC infusion technique was performed as described above. Once infusion was complete, the carotid artery was ligated, and the incision was closed using 3‐0 poliglecaprone 25 suture material. An MSC infusion was repeated on day 14 with cryopreserved autologous MSC resuspended in recovery solution using a different artery.

During all procedures, standard anesthetic monitoring was performed. After anesthetic recovery from both infusion procedures, animals were hospitalized for 24 hours on IVF. Buprenorphine 0.01 mg/kg was administered IV every 6‐8 hours for pain management until discharge from the hospital.

Patient follow‐up

At each study visit on days 0, 2, 14, 30, 60, and 90, all cats underwent physical examination. On days 30, 60, and 90, CBC, serum biochemistry, urinalysis, urine culture, UP : C ratio, and blood pressure measurement were performed. Iohexol clearance was determined on day 90.

Owners were asked to monitor their cats at home between study visits. They completed evaluation forms at enrollment and on days 14, 30, 60, and 90 on which they assigned their cats a score from 1 to 5, where 1 indicating normal behavior (no problems) and 5 indicating a severe problem, for each of the following categories: perceived abdominal pain, appetite, weight loss, lethargy, weakness, water consumption, urination, vomiting, energy level, overall condition, and quality of life.

Statistical analysis

Given the small sample size, only descriptive statistics were calculated. Data was considered nonparametric, and summary statistics are presented as medians and ranges.

RESULTS

Six cats were enrolled in phase 1. The medical record for 1 cat was lost, and this animal was excluded from analysis, leaving 5 cats that completed phase I. Breeds represented included Domestic Shorthair (3), Birman (1), and Norwegian Forest Cat (1). The median age was 14 years (range, 8.5‐17 years). Three cats were neutered males; 2 were spayed females.

All cats were diagnosed with IRIS stage III CKD >3 months before the start of the study. Azotemia and decreased iohexol clearance were noted in all cats at enrollment (median blood urea nitrogen concentration, 46 mg/dL; range, 41‐88 mg/dL; median SCr, 3.3 mg/dL; range, 2.9‐4.6 mg/dL; median iohexol clearance, 0.666 mL/min/kg; range, 0.579‐1.041 mL/min/kg). Four of the 5 cats had comorbidities including historical proteinuria controlled with benazepril, periodontal disease, historical hypertension controlled with amlodipine, uncharacterized hepatopathy, and osteoarthritis. Abdominal ultrasound examination showed right renal pelvic dilatation (5.2 mm) without ureteral dilatation in 1 cat. An antegrade pyelogram performed on this cat at the time of adipose tissue harvest was normal with no evidence of ureteral obstruction. Additional patient information can be found in supplemental information.

On day 0, adipose tissue was harvested as described above. Omental as well as falciform adipose tissue were obtained from 1 cat because of the small volume of falciform tissue available, ventral SC and falciform adipose tissue were obtained from 1 cat because of an excessive amount of available SC tissue, and falciform adipose tissue alone was harvested from the remaining cats. Procedural data including amount of adipose tissue harvested, anesthetic time, and procedure time are shown in Table 1. For 1 cat, recorded anesthesia and procedure times for the adipose tissue harvest were lost during transition from paper to electronic medical records during the study period. All cats recovered from anesthesia uneventfully.

Table 1

Procedural statistics

Procedural statistics	Median (range)
Adipose tissue harvest
Amount collected (g)	35.1 (31.5‐54.9)
Anesthetic time (min)	58 (50‐70)a
Procedure time (min)	26 (20‐35)a , b
1st MSC infusion
Number of cells injected (×106)	2.4 (1.5‐3.5)
Anesthetic time (min)	60 (45‐100)a
Procedure time (min)	45 (34‐65)a
2nd MSC infusion
Number of cells injected (×106)	3.7 (3.4‐5.7)
Anesthetic time (min)	70 (60‐95)a
Procedure time (min)	43 (35‐80)a

Data based on 4 animals.

Includes antegrade pyelogram performed in 1 cat.

Intra‐arterial MSC infusion was performed on days 2 and 14 as described above. Immediately before the first infusion in 1 cat, a small filling defect was observed in 1 branch of the renal artery on the renal angiogram. All arterial branches to the kidney were still patent and perfusing the entire kidney. This filling defect remained in place after stem cell infusion and did not worsen. Ultrasound examination of renal blood flow to that kidney the next morning identified no visible abnormalities on color flow Doppler. Procedural statistics including the number of cells infused, anesthetic time, and procedure time are shown in Table 1. The higher number of cells administered during the second treatment was the result of a higher than expected viable cell yield after cryopreservation. For 1 cat, recorded anesthesia and procedure times for both MSC injections were lost as a result of transition from paper to electronic medical records. All cats recovered from anesthesia uneventfully and survived both procedures.

Blood urea nitrogen concentration, SCr, and plasma iohexol clearance results from enrollment through day 90 are shown in Figure 2. Between day 0 and day 90, the median change in SCr was +0.5 mg/dL (range, −0.2 to +1 mg/dL). A net decrease in SCr was observed in 2 cats (−0.2 mg/dL), and a net increase was seen in 3 cats (+0.5‐1 mg/dL). The maximum percent increase in SCr noted between any 2 time points in a single patient was 27%, which represents an absolute increase of 1.1 mg/dL between day 0 and 30 (Patient 4). The maximum increase in SCr between 2 consecutive time points was 0.7 mg/dL or a 16% increase between day 14 and 30 (Patient 4). Summary statistics of this data are shown in Table 2.

Figure 2

Trends in serum creatinine, blood urea nitrogen (BUN), and plasma iohexol clearance during phase I

Table 2

Summary statistics of BUN, creatinine, and iohexol clearance at each study time point presented as median (range)

	Day 0	Day 14	Day 30	Day 60	Day 90
BUN (mg/dL)	41 (24‐74)	44 (36‐70)	47 (33‐62)a	47 (38‐58)	46 (33‐54)
SCr (mg/dL)	3.3 (2.3‐4.1)	3.1 (2.2‐4.5)	2.9 (2.6‐3.8)b	3.3 (2.5‐4.7)	3.7 (2.7‐5.1)
Iohexol clearance (mL/min/kg)	0.666 (0.579‐1.041)				0.697 (0.447‐1.102)

Denotes data from 4 animals due to lost medical records.

Denotes data from 3 animals due to lost medical records.

Owners reported no adverse events or decline in overall condition during the 3‐month study period. Owners of 2 cats noted improvement in overall condition with enrollment scores of 2 improved to 1. In 1 cat, improvement in body weight was observed with an enrollment score of 3 improved to 1. One cat developed a mildly decreased appetite at day 60 that persisted (score change from 1 to 2), and water consumption scores improved in 2 cats by 1 point indicating decreased polydipsia. The owner of the cat with osteoarthritis reported dramatic improvement in energy and activity levels over the 90‐day period.

Complications were noted in 2 of 10 infusion procedures. One cat, from which SC adipose tissue was harvested, developed moderate bruising and discomfort around the ventral abdominal incision. One cat developed Horner's syndrome after the carotid artery was accessed for MSC infusion; signs resolved within 3 days.

All cats were alive at the conclusion of phase 1 and were enrolled in phase II.

DISCUSSION

The objective of this phase I clinical trial was to assess the feasibility and safety of fluoroscopy‐guided IA delivery of autologous MSC to the kidney via percutaneous access to the renal artery in cats with IRIS stage III CKD. The procedure was performed successfully in all 5 cats, and only 2 minor, self‐limiting perioperative complications were observed. These outcomes indicate that the procedure itself is both feasible and safe. However, the effect of MSC infusion on the health of the kidneys over time remains unclear.

All cats remained in stage III for the 3‐month study period except for 1 cat in which the SCr increased from 4.1 mg/dL at enrollment to 5.1 mg/dL at day 90, changing that cat to IRIS stage IV CKD. Serum creatinine concentration increased in 3 cats, the median SCr increased from 3.3 to 3.7 mg/dL, and the median change in SCr was +0.5 mg/dL between days 0 and 90. Given interindividual variability in CKD progression, and the absence of disease‐matched control animals in phase I, it is difficult to interpret whether worsening of azotemia was the result of MSC treatment or natural progression of disease. In our experience, the magnitude of the increase in median SCr is larger than expected for disease progression in a 3‐month time period, but this result may be skewed as a consequence of small sample size. Further investigation involving a larger sample size, disease‐matched placebo and control groups, and a longer follow‐up period is needed to assess the long‐term effect of IA MSC treatment in CKD patients and currently is underway in phase II of the study.

Clinical improvement after treatment was reported by some owners. Such changes may have resulted from MSC treatment, may reflect waxing and waning signs of disease, may indicate a placebo effect, or may be some combination of these. Follow‐up studies that are blinded, randomized, and placebo‐controlled are needed to decrease bias in subjective scoring by owners and allow for better interpretation of treatment outcomes. Phase II was designed in this manner and, hopefully, will achieve these objectives.

Interestingly, 1 of 3 cats with a net increase in SCr had a net increase in iohexol clearance. This finding was unexpected given that SCr and iohexol clearance are inversely related markers of glomerular filtration rate (GFR). Such discordant results may represent error in measurement, the effect of extrarenal factors on 1 or both variables, or physiologic variation. In a study on IV MSC treatment in CKD cats, discordant results also were reported between different measures of GFR, specifically, SCr and nuclear scintigraphy.28 The investigators discussed variability in GFR as a possible explanation for the discrepancy, referring to previous studies in which day‐to‐day variability in GFR, as assessed by nuclear scintigraphy, by as much as 25% and 28% was shown in healthy dogs and CKD cats, respectively.27, 28, 29 Another study found similar variability in GFR in cats using iohexol clearance.30 In that study, variability in GFR as well as SCr was assessed.30 Comparison of measurements taken 6 months apart indicated greater variability in GFR (approximately 29% versus 20%) but lower variability in SCr (6.8% versus 8.8%) in azotemic cats relative to nonazotemic cats.30 This finding suggests that SCr may be the more reliable marker of GFR in cats with CKD. Additional studies on variability in SCr and measured GFR are needed to assess the stability of these variables in both normal and CKD cats to determine which is the more dependable marker of disease progression.

Adipose‐derived MSC were chosen for this study for several reasons. An in vitro study of feline stem cell growth and behavior characteristics indicated that MSC of adipose and bone marrow origin were similar in phenotype and differentiation ability with adipose‐derived cells having even faster, more efficient replication potential.31 Adipose‐derived MSC also have been shown to express comparable to higher levels of paracrine factors relative to those from bone marrow, which is important given current understanding of the role of MSC in treating disease.32 Based on these findings, the abundance of adipose tissue stores in most domestic cats, and the ease of harvesting adipose tissue as compared to bone marrow in cats, adipose‐derived MSC seemed to be the more suitable option. In practice, adipose tissue was easy to obtain from all study animals despite having to sample >1 site in some animals.

Although the IA procedure was successful in all cats, it entails specific challenges that would limit its use among veterinarians. In general, fluoroscopy‐guided IA procedures require training in interventional techniques using fluoroscopy and highly specialized equipment. We have extensive experience in this area, and still, there was difficulty accessing the renal artery in some cats because of the small size and inherently narrow vasculature of the feline renal artery despite the use of microcatheters and wires. Additionally, variation in patient size made prediction of equipment needs such as microcatheters challenging and expensive.

Another limiting factor associated with IA MSC administration is the need for general anesthesia. Patients must be anesthetized for each treatment as well as for the initial adipose tissue harvesting when autologous cells are used. Given the likelihood of anesthesia‐related hypotension, especially in geriatric cats, there is potential risk for renal injury in addition to that associated with the procedure itself. In our study, IVF was administered before and after each anesthetic period, and anesthetic events were relatively short in duration. No evidence of AKI was observed in any cat after anesthesia in phase I. Perioperative fluid administration therefore is recommended with this procedure.

Finally, although the IA infusion procedure is minimally invasive, it does require a small 1‐1.5 cm skin incision to access, isolate, and safely enter a major peripheral artery. In small patients such as cats, sheath and catheter exchanges can result in substantial blood loss. The artery then must be ligated after treatment, making future use of that specific artery more difficult, especially the carotid artery. In dogs, the femoral artery is relatively easy to access again after ligation, and each femoral artery can be used 2‐6 times, in our experience. Doing so is more difficult in cats because of the small size of the femoral artery, which necessitates access at a more proximal location each time.

Some of the specific challenges described above are avoided with IV and intrarenal MSC delivery. However, unlike these methods of delivery, IA infusion ensures direct delivery of MSC to the entire kidney by renal vascular perfusion, and, theoretically, more even distribution of MSC throughout the renal parenchyma through first pass engraftment. Intra‐arterial delivery also avoids entrapment of MSC in peripheral capillary beds, effectively increasing the number of MSC delivered to the kidney itself. Hence, there are clear advantages to the IA approach, but these benefits must be weighed against risk of hemorrhage, impaired circulation, embolization, multiple anesthetic events, expense, and expertise.

Our study had several limitations. First, stem cells were delivered in the form of SVF, which consists of a heterogenous population of cells, including MSC, pericytes, endothelial progenitor cells, macrophages, T‐regulatory cells, smooth muscle cells, and preadipocytes isolated from adipose tissue by enzymatic digestion.33 A nucleocounter was used to determine the number and viability of cells (both fresh cells and cells recovered from cryopreservation) in each aliquot of autologous SVF delivered to study animals, but no characterization of cells (such as detection of cell surface markers) was performed. Hence, the actual number of MSC delivered to each patient is unknown. This makes assessment of MSC treatment in our patients difficult because we cannot assert that any change or absence of change in renal function is attributed to MSC alone.

Additionally, the doses of SVF (ie, number of cells) infused during the 2 treatments were different, with a larger dose being given on day 14. Previous data from the stem cell company showed that a minimum of 65%, maximum of 79%, and average of 74% of all cells recovered from cryopreservation remain viable and remain so for at least 2 days in the controlled shipping container used in our study. In anticipation of cell losses, a larger number of cells than the fresh SVF dose were cryopreserved. Our viable cell yield was higher than expected for most cats, which resulted in a second dose that was 1 to 2.5 times the first. Despite the dose escalation, no major complications were observed after the second treatment. Additional studies in which MSC are culture‐expanded, characterized, and administered at various doses are needed to more definitively associate study results with MSC alone as well as with MSC dose. Radiolabeling of MSC and tracking their migration with nuclear scintigraphy as recently has been done in cats and horses also could be considered to assess MSC engraftment in the kidney and association with clinical outcome, but renal clearance of the radiolabel may complicate the interpretation of results.34, 35

The source of adipose tissue used for SVF in our study varied among patients. Multiple studies in dogs and people have shown differences in MSC yield as well as replication rate and differentiation ability of MSC depending on the source of adipose tissue (eg, omental versus SC).36, 37, 38, 39 Thus, functional variation may exist among MSC derived from different adipose sites and may affect their regenerative and reparative capacities. Future studies of this nature are needed in cats in which the source of adipose‐derived MSC is kept consistent to determine whether adipose origin is an important variable.

Over the course of our study, the hospital medical record system changed from paper to electronic, and the diagnostic laboratory changed. Unfortunately, this change precluded us from finding some valuable patient data. Our sample size also was small, which further limited data available for analysis.

Despite these limitations, our results indicate that selective IA delivery of autologous MSC in SVF into the renal artery of cats with IRIS stage III CKD is feasible. The procedure appears to be safe in the short‐term, but long‐term safety and efficacy currently are unknown. This procedure is technically difficult to perform because of the small size of cats and only should be performed by individuals trained in interventional techniques once a clear indication for use of this treatment is established. Our small patient population in phase I precludes us from drawing conclusions about the efficacy of IA MSC treatment for cats with CKD. However, efficacy of this treatment currently is being investigated in a phase II study.

CONFLICT OF INTEREST DECLARATION

Authors declare no conflicts of interest.

OFF‐LABEL ANTIMICROBIAL DECLARATION

Authors declare no off‐label use of antimicrobials.

INSTITUTIONAL ANIMAL CARE AND USE COMMITTEE (IACUC) OR OTHER APPROVAL DECLARATION

There was IACUC approval granted from the Animal Medical Center.

HUMAN ETHICS APPROVAL DECLARATION

Authors declare human ethics approval was not needed for this study.

Appendix S1: Supporting information

Click here for additional data file.