The use of vascular access ports for blood collection in feline blood donors

Abstract

We investigated vascular access ports for feline blood donation. Eight cats were anesthetized for conventional blood collection by jugular venipuncture at the beginning and end of the study. In-between conventional collections, vascular access ports were used for collection with or without sedation every 6 to 8 wk for 6 mo. Ports remained functional except for one catheter breakage, but intermittent occlusions occurred. Systolic blood pressure was lower during conventional collection. Behavioral abnormalities occurred during 3 port collections. Packed red cells prepared from collected blood were stored at 4°C for 25 d and assessed for quality pre- and post-storage. With both collection methods, pH and glucose level declined, and potassium level, lactate dehydrogenase activity and osmotic fragility increased. There were no differences between methods in pre-storage albumin and HCO₃⁻ levels, and pre and post-storage hematocrit, lactate dehydrogenase activity, and glucose and potassium levels. Pre-storage pH and pCO₂ were higher with conventional collection, and pre- and post-storage osmotic fragility were greater with port collection. One port became infected, but all cultures of packed red cells were negative. Tissue inflammation was evident at port removal. In a second study of conventional collection in 6 cats, use of acepromazine in premedication did not exacerbate hypotension. The use of vascular access ports for feline blood donation is feasible, is associated with less hypotension, and may simplify donation, but red cell quality may decrease, and effects on donors must be considered.

Introduction

Blood donation in cats may be challenging due to intolerance of physical restraint and small size. Hypotension during donation is common due to the combined effects of blood loss and anesthesia (1,2). While blood bank phlebotomists regularly perform jugular venipuncture in cats with relative ease, phlebotomy for blood donation may prove difficult when donations are only occasionally performed. The need for anesthesia and potentially difficult venipuncture may increase donation time. This is particularly a problem when emergency transfusion is needed.

Over the past 30 y, biomedical researchers and physicians have used totally implantable subcutaneous vascular access ports (VAPs) to access the circulation without venipuncture or external catheters (3–9). These devices have been used in veterinary clinics to facilitate anticancer chemotherapy and multiple episodes of anesthesia for radiation therapy (10–12). The use of VAPs was recently reported for feline blood donation (13).

The objectives of our study were: 1) to determine the feasibility of, and complications associated with, the use of VAPs for blood collection in cats over a 6-month period; 2) to compare systolic blood pressure variation between cats undergoing conventional blood collection and vascular access port collection; and 3) to compare quality of packed red blood cells obtained by both collection methods, before and after storage.

Materials and methods

Study design

Two separate studies were performed. In the first study of a group of 8 cats, conventional blood collection (CON-C) was compared to vascular access port collection (VAP-C) with respect to changes in systolic blood pressure (SBP) during collection, and quality of packed red blood cells (pRBCs). In the second follow-up study of a group of 6 different cats, SBP was measured during CON-C where premedication with acepromazine (Atravet; Wyeth Animal Health, Guelph, Ontario) was not used.

Animals

A total of 14 common-sourced mixed-breed cats ranging in weight from 4.0 to 6.5 kg were studied. The 8 cats for the first study consisted of 5 neutered males and 3 spayed females. The 6 cats in the second study were all neutered males.

All cats were vaccinated against panleukopenia, rhinotracheitis, calicivirus (Protex-3, Intervet/Schering-Plough Animal Health, Kirkland, Quebec), and rabies (Imrab 3, Mérial Canada, Baie d’Urfé, Québec). Cats were confirmed to be healthy based on normal results of physical examination; complete blood (cell) count, serum biochemical profile, and urinalysis (Animal Health Laboratory, University of Guelph, Guelph, Ontario); and negative results of tests for endoparasites (AHL), feline leukemia virus, and feline immuno-deficiency virus by enzyme-linked immunosorbent assay (IDEXX Laboratories Canada, Toronto, Ontario), Mycoplasma haemofelis by polymerase chain reaction (Laboratory of Veterinary Diagnostic Medicine, University of Illinois, Urbana, Illinois, USA) and Bartonella henselea antibody titer (North Carolina State University Tick-Borne Disease Laboratory, College of Veterinary Medicine, Raleigh, North Carolina, USA). Cats received daily socialization, and were acclimatized to the research facilities for a minimum of 4 wk prior to any experimental procedures. All procedures met the guidelines set by the Canadian Council on Animal Care (Guide to the Care and Use of Experimental Animals, Volumes 1 and 2), were performed in accordance with the Animals for Research Act (Ontario, 1980), and were approved by the Animal Care Committee, University of Guelph. All cats were adopted into the community after being retired as blood donors, except for one cat that was euthanized for behavioral reasons.

Conventional blood collection

The same method of conventional blood collection was used in both studies. One unit (52.2 mL) of blood was collected by left jugular venipuncture at the beginning and end of the first 6-month study period. Cats were fasted 12 h before collection, and sedated with a mixture of butorphanol (Torbugesic; Wyeth Animal Health), 0.2 mg/kg body weight (BW), and acepromazine (Atravet; Wyeth Animal Health), 0.05 mg/kg BW, IM, to allow placement of a 22-gauge catheter (BD Insyte-W; BD-Canada, Oakville, Ontario) into a cephalic vein. The cephalic catheter was used to administer ketamine (Ketalean; Bimeda-MTC Animal Health, Cambridge, Ontario), 6.0 to 10.0 mg/kg BW, and diazepam (Diazepam; Sabex International, Boucherville, Quebec), 0.3 to 0.5 mg/kg BW, given as a 1:1 volume mixture containing 50 mg ketamine and 2.5 mg diazepam per mL. Oxygen was delivered by face mask during collection and recovery. The venipuncture site was clipped and prepared with 7% chlorhexidine gluconate (Germi-Stat, Germiphene Corporation, Brantford, Ontario), 70% isopropyl alcohol (Commercial Alcohol, Brampton, Ontario), and a solution of 0.05% chlorhexidine gluconate, 0.5% cetrimide and 70% isopropyl alcohol (tincture of Savlon; Ontario Veterinary College Pharmacy, Guelph, Ontario). A 19-gauge winged infusion set (Butterfly-19, Hospira, Lake Forest, Illinois, USA) was used to collect blood into a 60-mL plastic syringe (Monoject; Covidien, Pointe-Claire, Quebec) to which had been added 7.8 mL of an anticoagulant-preservative solution (CPDA1, Cytosol Laboratories, Braintree, Massachusetts, USA). Saline (100 mL) (0.9% Sodium Chloride Injection; Baxter Healthcare Canada, Montreal, Quebec) was administered IV over 20 min at the end of blood collection or, alternatively, during collection if SBP dropped below 80 mmHg (see Blood pressure measurement). In the second study, blood was collected in an identical fashion except for the exclusion of acepromazine from the anesthesia protocol.

Vascular access ports

Surgical placement of VAPs

Five and 7 Fr silicone attachable catheters and corresponding 5 and 7 Fr silicone VAPs (Access Technologies, Skokie, Illinois, USA) were used. Size 5-Fr VAPs were placed in 6 cats and 7-Fr VAPs were placed in cats 6 and 8 because of their larger size. Non-coring Huber needles (Access Technologies) were used to invade the ports. The catheters were implanted in the right external jugular vein of anesthetized cats by two of the investigators (A.M.S. and I.A.) The cat was placed in left lateral recumbency and a first incision was made lateral to the right jugular vein. The vein was isolated over 2 cm and one simple ligature suture (Surgipro 4-0; United States Surgical Corporation, Norwalk, Connecticut, USA) was used to obliterate blood flow. Two other ligatures were preplaced distally to the suture and the vein was incised between the sutured and preplaced ligatures. While connected to the VAP, the catheter was inserted in the vein and threaded under fluoroscopic guidance to the junction of the right atrium and cranial vena cava. Catheter patency was ensured by withdrawing blood and then flushing the port and catheter with 0.5 mL of 200 U/mL heparinized 0.9% saline (prepared from Physiologic Saline, Vétoquinol Canada, Lavaltrie, Quebec, and Heparin LEO 1000 U/mL, Leo Pharma, Canada, Thornhill, Ontario), prior to suturing the catheter in place by tightening the preplaced ligatures. Minute manipulations of the catheter were often necessary to achieve easy blood sampling. A first catheter loop (approximately 2 cm in diameter) was formed and sutured laterally to the jugular vein, to minimize potential catheter kinking associated with positioning of the cat’s neck. The catheter was clamped and detached from the port. The cat was then placed in sternal recumbency. A 2-cm transverse incision was made cranially to the medial aspect of the right scapula, and the subcutaneous tissues were dissected to create a pocket for port placement approximately 1 cm caudal to the transverse incision. Forceps were inserted through the dorsal skin incision and advanced with blunt dissection to retrieve and clamp the loose end of the catheter at the level of the first incision. The forceps were pulled dorsally to tunnel the catheter to the placement site for the port. The port was placed in the subcutaneous pocket and sutured to the underlying fascia. A second, larger catheter loop (approximately 5 cm in diameter) was formed and sutured in a location cranial to the port, in order to minimize tension on the jugular vein associated with manipulation of the port. The port end of the catheter was transected to the appropriate length such that it could be attached to the port without tension or compression of the catheter, and it was then reattached to the port. The system was flushed with heparinized saline. Subcutaneous and intradermal sutures (Biosyn 4-0, United States Surgical Corporation) were placed to close subcutaneous tissues and skin incisions.

Maintenance of VAP patency

Patency of the VAP was maintained by flushing the port every 10 d. In preparation for flushing, the hair over the port was clipped and a topical anesthetic cream (EMLA Cream, AstraZeneca Canada, Mississauga, Ontario) was applied to the skin. Twenty minutes later, the skin was surgically prepared as previously described. A 20-gauge hypodermic needle was used to puncture the skin to facilitate passage of a 22-gauge Huber needle connected to an IV adapter (PRN Adapter, BD-Canada), which was inserted through the puncture hole and port diaphragm. Approximately 0.5 mL of a mixture of heparinized saline and blood was aspirated from the port and discarded, then 10 mL of 0.9% saline was infused into the port by pressing the syringe plunger in a pulsatile motion. Finally, the port and catheter were locked with 0.5 mL of heparinized saline (200 U/mL) by maintaining positive pressure on the plunger of the syringe while the Huber needle was removed. Flushing and locking of VAPs was discontinued after completion of the study prior to VAP removal, except in cat 1, where the VAP was maintained for an additional 11 mo.

Blood collection using VAPs

Blood was first collected through VAPs 4 wk after placement and then at 6, 8, and 8-week intervals. After a 12-hour fast, the VAP was accessed as previously described, and the heparin lock was aspirated if possible and discarded. If light sedation was needed, ketamine, 1.5 to 4.0 mg/kg BW, and diazepam, 0.075 to 0.20 mg/kg BW, given as 1:1 mixture, were administered through the VAP. A 19-gauge Butterfly was inserted into the PRN adapter, 52.2 mL of blood was collected, 100 mL of saline was infused, and the VAP was heparin-locked as previously described.

Surgical removal of VAPs

All VAPs were removed from anesthetized cats 8 mo after placement (cats 2 to 8) or 17 months after placement (cat 1). A 2-cm incision was made parallel to the right jugular vein and fibrotic tissue was incised. The jugular vein was gently retracted and the retention ligatures removed. A new simple ligature (Surgipro 4-0) was preplaced around the patent portion of the vein and was tightened as the catheter was withdrawn. A 2-cm longitudinal incision was made cranial to the port, and the capsule encompassing the port and catheter was incised to allow removal of sutures. The catheter and port were removed with traction through the dorsal incision and both incisions were closed. The patency of VAPs was tested. Samples of the capsule and fluid surrounding the ports of cats 1 and 3 were submitted to the Animal Health Laboratory for histopathologic and cytologic evaluation, respectively, and for aerobic and anaerobic bacterial culture.

Blood pressure measurement

Systolic blood pressure was measured with an ultrasonic Doppler flow detector (Parks Medical Electronics, Aloha, Oregon, USA). Hair was clipped between the carpal and metacarpal pad over the median artery. Ultrasonic coupling gel (EcoGel 200, Eco-Med Pharmaceutical, Mississauga, Ontario) was placed on the concave surface of an 8-MHz Doppler probe and the probe was moved over the clipped area until a strong pulse signal could be heard. The probe was fixed in position with adhesive tape. An inflatable size 3 pediatric cuff (GE Health Care Canada, Mississauga, Ontario), with an approximate width of 30% to 40% of the limb circumference, was placed around the antebrachial region and inflated to a pressure 30 to 40 mmHg higher than that required to obliterate the pulse. The cuff was then slowly deflated and the SBP recorded as the pressure at which the first pulse signal became audible. A minimum of 3 measurements were obtained immediately prior to blood collection, then every 5 min during collection, again at the end of collection, and then 15 min post-collection.

Preparation and storage of packed red blood cells

Blood was transferred to a 100-mL polyolefin bag (Animal Blood Bank, Dixon, California, USA) by gentle injection, and centrifuged at 5000 × g for 8 min at 4°C (Sorvall RC3BP, Kendro Laboratory Products, Newton, Connecticut, USA). The plasma was expressed into a second bag using a plasma extractor (Plasma Extractor 4R4414, Fenwal, Lake Zurich, Illinois, USA). Ten mL of a red blood cell preservative solution (Optisol, Terumo Medical Corporation, Elkton, Massachusetts, USA) was added to, and gently mixed with, the pRBCs, which were then stored in an upright position at 4°C for 25 d.

In vitro assessment of red blood cell and plasma quality

Blood bags containing pRBCs were gently mixed by inversion and 3 mL of blood were collected aseptically from each bag on the day of preparation, and after 25 d of storage, for determination of hematocrit, glucose and potassium levels, lactate dehydrogenase (LDH) activity, pH, and red blood cell median osmotic fragility (MOF).

Hematocrit was determined using an automated hematology analyzer (ADVIA 120, Bayer Diagnostics Division, Toronto, Ontario). Glucose and potassium levels and LDH activity were determined using a biochemistry analyzer (Hitachi 911, Roche Diagnostics Canada, Laval, Quebec) on the supernatants of samples of pRBCs centrifuged at 3500 × g for 10 min. The pH and CO₂ levels were measured at 37°C, and HCO₃⁻ levels calculated, using a blood gasanalyzer (ABL-500 Radiometer Canada, London, Ontario).

Osmotic fragility was determined by assessing the degree of hemolysis generated when pRBCs were exposed to various concentrations of phosphate buffered saline (PBS) (14). A 10% PBS stock solution was prepared by mixing 180.00 g of sodium chloride (Fisher Scientific Company, Ottawa, Ontario), 27.31 g of dibasic sodium phosphate (Fisher Scientific), and 4.86 g of monobasic sodium phosphate (Fisher Scientific) in 2000 mL of deionized water. Prior to each experiment, deionized water was added to the stock solution to prepare serial dilutions of PBS of 0%, 0.20%, and from 0.4% to 0.85% in 0.05% increments. Three mL of each dilution were transferred to a set of 12 test tubes and 0.03 mL of pRBCs were added to each tube. Each tube was immediately covered with a self-sealing film (Parafilm M, SPI Supplies/Canada, Toronto, Ontario) and mixed by gentle inversion, and left at room temperature for 20 min. The tubes were remixed prior to being centrifuged at 2000 × g for 5 min (Adams Sero-Fuge, Clay-Adams, New York, New York, USA). The supernatant from each tube was transferred to a cuvette and read on a spectrophotometer (Shimadzu UV-1601PC, Shimadzu Scientific Instruments, Columbia, Massachusetts, USA) at a wavelength of 550 nm. The optical density (OD) was set at 0 using deionized water. The percent hemolysis for each supernatant was calculated as:

$$[({\text{OD}}_{\text{S}}-{\text{OD}}_{0.85\%})/({\text{OD}}_{0\%}-{\text{OD}}_{0.85\%})]\times 100$$

where: S refers to the test supernatant, and 0% and 0.85% refer to the respective concentrations of PBS. Percent hemolysis was plotted against PBS concentration, and the concentration at which 50% of red blood cells were lysed (median cell fragility) was extrapolated from the osmotic fragility curve.

Plasma (3 mL) was collected from each plasma bag at the time of collection for determination of albumin concentration using a biochemistry analyzer (Hitachi 911).

Microbiologic assessment of packed red blood cells

Samples (1 mL) were obtained from each bag of pRBCs on day 0 and day 25 and were injected aseptically into trypticase soy broth (BBL Septi-Chek TBS, BD-Canada), incubated for 7 d, and subcultured aerobically and anaerobically 1 and 5 d after inoculation. Samples from the VAPs’ heparin locks were collected for aerobic and anaerobic culture (BBL Culture Swab and BBL Port-A-Cul, BD-Canada) at the time of the first VAP-C and whenever a cat became febrile, lethargic, or inappetent, or began to scratch at an incision site.

Statistical analysis

Statistical analysis was performed using a computerized program (PC SAS, SAS Institute, Cary, North Carolina, USA). All values for SBP and quality assessment of pRBCs were reviewed for normality and the variables examined for homogeneity of variance (Bonferroni adjustment), skewness, and kurtosis. Non-normally distributed variables were log transformed to improve normality and then parallel analyses were performed with transformed variables. Non-transformed and transformed variables were then evaluated for differences over time and between collection methods by analysis of variance (ANOVA) for repeated measures. Plasma albumin values were compared with a 2-tailed paired t-test. A P-value ≤ 0.05 was considered significant.

Results

Complications of VAPs

Complications of VAPs are summarized in Table I. Seromas at incision sites of cats 4 and 5, and right forelimb edema of cat 5, occurred within 1 wk, and resolved spontaneously within 3 wk, after surgery. Mild excoriations from scratching at incision sites occurred at 10 and 45 d after surgery in cat 4, and at 80 and 90 d after surgery in cats 5 and 7, respectively.

Table I.

Complications associated with totally implantable subcutaneous vascular access ports and their use for blood collection in 8 feline blood donors over a 6-month period

Cat	1	2	3	4	5	6	7	8
Post-operative complications
Dehiscence dorsal incision						✓
Seroma dorsal incision				✓		✓
Seroma cervical incision					✓
Right forelimb edema					✓
Chronic complications
Catheter-associated sepsis, cellulitis, dermatitis						✓
Mild excoriations from self-trauma				✓	✓		✓
Catheter fracture					✓
Blood collection complications
Sampling occlusion		✓	✓	✓	✓		✓	✓
Sedation required		✓	✓	✓	✓		✓	✓
Vomiting							✓
Defecation		✓
Vocalization		✓						✓
Collapse								✓

The most severe VAP complication occurred in cat 6, where the dorsal incision dehisced 1 d after surgery. The wound was lavaged with 0.9% saline and closed with surgical staples (AutoSuture, Pointe-Claire, Quebec). The cat developed pyrexia (41°C), diarrhea, and splenomegaly 35 d after surgery. Samples from the heparin lock and from a blood sample obtained through the VAP yielded heavy growths of Enterobacter cloacae and Klebsiella oxytoca, sensitive to enrofloxacin. Enrofloxacin (Baytril Injectable Solution, Bayer, Toronto, Ontario), 5 mg/kg BW, was administered through a cephalic catheter every 24 h for 4 d. An “antibiotic lock” was also prepared for the VAP by mixing 10 mg (0.2 mL) of enrofloxacin (50 mg/mL) with 0.1 mL of saline and 0.1 mL of heparin (1000 U/mL) and injecting this mixture into the VAP as previously described for the heparin lock. The antibiotic lock was changed daily for 4 consecutive days. Thereafter, 50 mg (8 mg/kg BW) enrofloxacin was administered orally q24h for another 15 d. Clinical signs resolved. Cultures of the VAP lock performed during, and at 10 and 20 d after treatment with enrofloxacin; blood cultures obtained from CON-C and VAP-C pRBCs as part of the study; and cultures of the VAP lock and portions of the catheter at the time of VAP removal were all negative. However, severe moist interscapular dermatitis (characterized by erythema, excoriations, and exudation), presumed cellulitis, and pyrexia (40.6°C) developed 15 mo after VAP removal, which resolved with topical administration of 0.05% chlorhexidine and 1% hydro-cortisone (Dermacool-HC, Allerderm/Virbac, Fort Worth, Texas, USA) and enrofloxacin 50 mg PO q24h for 21 d. Severe episodes of interscapular dermatitis recurred 18 and 22 mo post-VAP removal, which resolved with the topical treatment alone. Additional episodes prompted en-bloc resection of subcutaneous tissue in the area 4 y after VAP removal. At the time writing, severe moist dermatitis has not recurred, but excoriations still occur. There have been no clinical signs of retinopathy.

A thick, white, smooth capsule encompassing the port and the catheter was observed at removal of all VAPs. Three to 10 mL of serosanguinous fluid were present between the VAP and its capsule in 5 cats. Histopathology of samples of the capsules performed in cats 1 and 3 revealed a proliferation of fibroblasts with diffuse infiltration by moderate numbers of neutrophils and large numbers of macrophages. Cytology of the serosanguinous fluid revealed a high density of well-preserved neutrophils. Aerobic and anaerobic culture of capsule tissue and fluid was negative. Five out of eight VAPs contained a lining of red-brown material that extended the length of the catheter but occluded < 1/4 of its diameter. All catheters were patent.

Blood collection

All CON-C were performed uneventfully. Considering all collections together, the CON-C times, from venipuncture to withdrawal of the needle, ranged from 5 to 9 min.

With respect to VAP-C, all cats tolerated needle insertion into their VAPs; however, sedation was subsequently needed in 5 cats to complete VAP-C (see Table I). The VAP-C times, from needle insertion into the port-to-needle withdrawal, ranged from 3 to 7 min for the first collections. This rapid collection was associated with behavioral abnormalities. Cat 7 vomited within 5 min of completing the VAP-C. Cat 2 vocalized and defecated; SBP was > 100 mmHg and heart rate > 100 bpm when signs occurred. Cat 8 vocalized loudly and collapsed with marked pallor, heart rate of 160 bpm, dilated pupils, and tachypnea associated with open-mouth breathing. All signs resolved within 10 s. In both cases the signs appeared after loss of 25 to 35 mL of blood. Mild sedation (using ketamine and diazepam, as previously described) was subsequently administered to these cats before blood collection, and collection was then taken over a minimum of 10 min in all cats in an effort to minimize complications. No other episodes of abnormal behavior occurred.

Sampling occlusion, defined as one-way occlusion, where blood could not be easily withdrawn but where saline could be easily injected, first occurred in 4 cats at the time of the first collection and in 2 other cats at the time of the last collection. All occlusions resolved with injecting 3–5 mL of saline and/or manipulation of the cats’ necks.

The collection process became subjectively easier as the investigators gained experience in maneuvering the VAPs. Blood collection through VAPs was facilitated by withdrawing the blood using minimal, pulsatile, negative pressure. Even the easiest collection could be interrupted by the use of excessive aspiration; this was not recorded as a sampling occlusion. The longest VAP-C was 35 min.

All VAPs remained functional over the 6-month period except in cat 5, where catheter fracture precluded the last VAP-C, 206 d postoperatively. This was suspected when administration of ketamine and diazepam through the VAP did not procure typical sedation. Fracture was confirmed radiographically following injection of 3 mL of a 1/10 dilution of iohexol (240 mg I/mL) (Omnipaque; GE Health Care Canada) in saline into the VAP.

The VAP in cat 1 was functional at 17 mo when removed.

Blood pressure measurement

Changes in SBP are summarized in Table II. There was no difference in mean SBP at the beginning of collection between cats undergoing CON-C and VAP-C. Differences were found between mean SBP prior to collection and mean SBP nadir for both CON-C and VAP-C; however, CON-C resulted in a lower nadir SBP and greater drop in SBP than VAP-C. This difference persisted when SBP was compared between CON-C and VAP-C using only the cats that were sedated for VAP-C. There was no difference in SBP values between the 8 cats that received acepromazine for CON-C and 6 different cats in the study where acepromazine was not used. There was no difference (P = 0.27) for the mean time at which the SBP nadir occurred for CON-C (5.3 ± 2.1 minutes) and VAP-C (5.1 ± 1.7 minutes).

Table II.

Systolic blood pressure measured [no adjustment for underestimation of value (20,21)] with an ultrasonic Doppler flow detector in feline blood donors undergoing blood collection via conventional jugular venipuncture and via a vascular access port. In the first study 8 cats were anesthetized for conventional collection with butorphanol, acepromazine, ketamine and diazepam. Six of these 8 cats were sedated for vascular access port collection with ketamine and diazepam, while 2 cats were unsedated. In a second study 6 different cats were anesthetized for conventional blood collection with butorphanol, ketamine, and diazepam, but omitting acepromazine. Values are reported as mean ± standard deviation (range)

Collection method	SBPt0 (mmHg)	SBPmin (mmHg)	SBPt0-min (mmHg)	P-value: t0-min
Conventional collection — with acepromazine (main study, 8 cats)	(108 ± 17 (80–140)	53 ± 25 (20–100)	56 ± 29 (20–100)	0.000003
VAP collection — sedated and unsedated cats (main study, 8 cats)	132 ± 25 (100–190)	112 ± 21 (85–160)	24 ± 17 (0–55)	0.009
P-value: conventional versus VAP (main study, 8 cats)	0.06	0.003	0.03
VAP collection — sedated cats only (main study, 6 cats)	134 ± 27 (100–190)	111 ± 22 (85–160)	26 ± 16 (0–55)	0.007
P-value: conventional versus VAP-sedated cats (main study, 6 cats)	0.07	0.003	0.03
Conventional collection — without acepromazine (follow-up study of 6 cats)	113 ±18 (90–135)	52 ±22 (20–75)	61 ± 27 (30–100)	0.003
P-value: acepromazine versus no acepromazine (main study 8 cats versus follow-up study 6 cats)	0.65	0.99	0.81

VAP — Vascular access port; SBP_t0 — systolic blood pressure at beginning of collection; SBP_min — systolic blood pressure nadir; SBP_t0-min — drop in systolic blood pressure from beginning of collection to nadir.

In vitro assessment of red blood cell and plasma quality

Results of assessment of blood quality are summarized in Table III. There was no difference in hematocrit between collection methods, and hematocrit did not change over the storage period. There was no difference in plasma albumin level between collection methods.

Table III.

Blood quality parameters after collection and after storage for 25 days at 4°C for 8 cats where blood collection was performed via conventional jugular venipuncture and via a vascular access port. All parameters except albumin were determined on packed red blood cells. Values are reported as mean ± standard deviation (range)

Parameter	Collection method	Pre-storage	Post-storage	Change over storage period	P-value: Pre-storage versus post-storage	P-value: Pre-storage VAP versus post-storage C
Hct (L/L)	C	0.52 ± 0.04 (0.43 to 0.55)	0.52 ± 0.09 (0.39 to 0.62)	0.00 ± 0.07 (−0.12 to 0.08)	0.84	0.44
	VAP	0.55 ± 0.03 (0.50 to 0.57)	0.48 ± 0.08 (0.32 to 0.55)	−0.03 ± 0.04 (−.09 to −0.01)	0.22
	P-value: C versus VAP	0.28	0.31	0.38
Plasma albumin (mmol/L)	C	26 ± 3 (23 to 30)	ND	ND		ND
	VAP	29 ±2 (26 to 30)	ND	ND
	P-value: C versus VAP	0.15
Glucose (mmol/L)	C	35.9 ± 10.0 (26.3 to 46.8)	30.5 ± 8.6 (21.4 to 45.0)	−2.9 ± 14.1 (−18.9 to 16.9)	0.33	0.12
	VAP	41.4 ± 1.5 (39.9 to 43.8)	23.8 ± 6.0 (15.0 to 31.6)	−17.3 ± 3.6 (−21.7 to −11.7)	0.05
	P-value: C versus VAP	0.32	0.19	0.30
pH	C	6.56 ± 0.07 (6.44 to 6.64)	6.35 ± 0.09 (6.22 to 6.46)	−0.21 ± 0.10 (−0.36 to −0.13)	0.03	0.01
	VAP	6.77 ± 0.07 (6.72 to 6.89)	6.36 ± 0.16 (6.13 to 6.54)	−0.33 ± 0.13 (−0.54 to −0.18)	0.01
	P-value: C versus VAP	0.03	0.80	0.30
pCO2 (mmHg)	C	70 ±19 (51 to 98)	ND	ND		ND
	VAP	43 ± 7 (34 to 54)	ND	ND
	P-value: C versus VAP	0.01
HCO3− (mmol/L)	C	5.2 ± 1.4 (3.5 to 6.9)	ND	ND		ND
	VAP	5.6 ± 0.5 (4.9 to 6.2)	ND	ND
	P-value: C versus VAP	0.70
Potassium (mmol/L)	C	1.9 ± 0.6 (1.1 to 2.8)	4.9 ± 0.4 (4.4 to 5.6)	3.0 ±0.7 (2.2 to 4.0)	0.0007	0.0008
	VAP	2.2 ± 0.4 (1.8 to 2.6)	5.1 ± 0.4 (4.6 to 5.8)	2.9 ±0.3 (2.4 to 3.1)	0.001
	P-value: C versus VAP	0.28	0.47	0.38
LDH activity (U/L)	C	193 ± 99 (73 to 351)	5309 ±2145 (3660 to 9670)	5321 ±2176 (3309 to 9597)	0.03	0.03
	VAP	182 ± 70 (95 to 251)	4316 ± 2209 (2022 to 8670)	3174 ± 1233 (1807 to 4740)	0.05
	P-value: C versus VAP	0.99	0.38	0.17
Median osmotic fragility (% NaCl)	C	0.54 ± 0.03 (0.49 to 0.57)	0.56 ±0.04 (0.48 to 0.60)	0.02 ±0.05 (−0.07 to 0.01)	0.15	0.18
	VAP	0.58 ± 0.02 (0.54 to 0.61)	0.61 ± 0.02 (0.59 to 0.64)	0.04 ± 0.03 (0.01 to 0.09)	0.03
	P-value: C versus VAP	0.02	0.004	0.50

C — conventional blood collection; VAP – vascular access port; Hct — hematocrit; ND — not determined; LDH — lactate dehydrogenase; HCO₃⁻ — bicarbonate; pCO₂ — partial pressure carbon dioxide; Median osmotic fragility — saline concentration resulting in lysis of 50% of red blood cells.

There was no difference in glucose level between the 2 collection methods. Glucose level fell during storage in both groups, although the change for the CON-C was not significant. The degree of glucose depletion during storage was not different between collection methods.

Prior to storage, pRBCs obtained by VAP-C had a higher pH than those obtained by CON-C, but there was no difference in post-storage pH between collection methods or in the degree of decrease in pH between methods. Level of CO₂ in pRBCs after collection was higher for CON-C than for VAP-C, but the HCO₃⁻ level did not differ between collection methods.

There were no differences in potassium or LDH levels between the 2 collection methods. Values for both parameters increased during storage and there were no differences in the degree of potassium and LDH release during storage between collection methods.

Median osmotic fragility was greater with VAP-C than CON-C. The MOF increased in both groups, but the change was not significant with CON-C. There was no difference between collection methods with respect to the change in MOF over the storage period.

The above parameters were also compared for pre-storage VAP-C pRBCs and post-storage CON-C pRBCs. The CON-C post-storage pRBCs had higher potassium and LDH levels, and lower pH, than VAP-C pRBCs, but there was no difference in MOF.

All aerobic and anaerobic cultures of CON-C and VAP-C donations were negative.

Discussion

In this study we investigated the use of VAPs for blood collection in cats in an effort to avoid anesthesia and facilitate donation. Consistent with a previous study (13), our study demonstrated that long-term use of VAPs in feline blood donors is technically feasible.

Blood collections of 52.2 mL via VAPs were successful in 97% of attempts. Sampling occlusion was the most common complication. Potential causes of sampling occlusion include catheter kinking, lodging of the catheter tip against the wall of the right atrium or cranial vena cava, thrombosis, and collapse of the catheter wall under negative pressure. Sampling occlusions could be corrected by repositioning the cats’ necks and/or flushing the VAPs with saline. Resolution of occlusions through repositioning suggests that catheter kinking and/or lodging of the catheter tip against the wall of the right atrium or cranial vena cava were often responsible. These are not common causes of VAP occlusion in humans (6), but the catheter insertion site is in comparatively close proximity to the port site in humans, and catheter loops are not used. Similarly, positional occlusions were not reported in the previous study of VAPs in feline donors, where the port was in a cervical location and catheter loops were not used, but 6 of 20 collections required a second flush to establish VAP patency (13). It is possible that the catheter loops used in our study, although intended to prevent catheter kinking, may have had the opposite effect.

Sampling occlusion caused by thrombosis is less likely to be positional, but was another possibility where saline flush resolved occlusion. Catheter thrombosis is a major complication in human medicine (6), but there was no gross evidence of this in our study. None of the VAPs contained thrombi when removed, even though the VAPs in seven cats had not been flushed for 8 wk at the time of removal. It is possible, however, that thrombi were dislodged during catheter removal. The material observed lining 5 of the catheters was not analyzed. Although blood cultures obtained through the catheters were negative, it is possible that the material was a biofilm, a microbial community enveloped by extracellular biopolymers produced by the microbial cells, adhering to the interface of a liquid and a surface (15). Such biofilms are frequent complications of intravascular catheters in humans (15).

Collection through VAPs was readily interrupted by excessive negative pressure. This may have been due to collapse of the catheter wall, promoting lodging of the catheter-tip against the right atrium or cranial vena cava, or aspiration of a thrombus into the catheter tip.

Sampling occlusion occurred with a higher frequency in our study than in other studies of VAPs in cats (3,8,9,13,16). The stringency with which even the slightest difficulty with blood withdrawal (not attributed to negative pressure) was recorded as a sampling occlusion may account for the high occlusion rate. Description of VAP patency is limited in some previous studies to the VAP being functional or not. For example, non-correctable sampling occlusion of 3 of 35 VAPs was previously reported in cats (8). However, partially occluded VAPs from which blood could be obtained after corrective measures were still considered functional and were not recorded as occlusions. Duration of VAP implantation probably did not have a major role in the development of sampling occlusion in our study as most occlusion-prone VAPs were noted at the first collection. Similarly, in a previous study all occlusions developed within a week of VAP implantation (8).

There were several dermatologic complications. Post-surgical seromas occurred in 3 cats. Pruritus, self-inflicted excoriations and moist dermatitis also occurred in these 3 cats, as well as in a fourth cat. Surprisingly, most of the lesions developed 45 to 90 d postoperatively. At the time of writing, cat 6 continues to have recurrence of pruritus and erythema at the site of port implantation despite en-bloc excision of the area’s subcutaneous tissue 4 y previously. Seroma formation after VAP placement is rare in humans (6), but was reported as a common complication of VAP placement in dogs when the interscapular space and the jugular vein were used for port implantation and catheter insertion, respectively (5). Less subcutaneous dissection is needed in humans, and this may partially account for the difference (6). Seroma formation and inflammatory response rates were lower in the previous study of VAPs in feline donors, which may reflect the cervical location of the port and/or the smaller number of cats studied (13). Given that persistent inflammation has been identified as risk factor for development of fibrosarcoma, and a fibrosarcoma has recently been reported with microchip implantation (17,18), it is recommended that cats with VAPs implanted for longer duration be assessed for persistent inflammation surrounding the port and that cats that have had VAPs have long-term follow-up to determine if these devices increase the risk for cancer.

Breakage of cat 5’s catheter 206 days post-implantation was the only instance of device failure. Fracture occurred at the proximal end of the catheter, at the point corresponding to the distal end of the pin by which the catheter is attached to the port. Attachable catheters have the advantage of allowing trimming at the proximal end rather than at the tip, but the connection is a point of mechanical weakness. A reinforcing silicone sleeve that can be slid over the port-catheter junction is now available (Access Technologies). This cat’s VAP had been prone to sampling occlusion and collection was not expected to occur without prior saline flush. Fortunately, peri-portal administration of ketamine-diazepam only caused prolonged drug effect, but the incident emphasizes that catheter fracture should always be considered when there is sampling occlusion, and VAP integrity should be verified to minimize the risks of extravasation.

Hypotension, defined as a SBP < 80 mmHg, is considered to be an important complication of feline blood donation (1,2). Ultrasonic Doppler is an acceptable method for comparing SBP in cats (19), and was used in our study because of anticipated problems with repeated arterial catheterization. The Doppler method may underestimate SBP in cats by a mean of +14 to +26 mmHg (20,21). If a correction factor of + 26 mmHg is applied to SBP_min values in Table II, SBP_min estimates during CON-C would range from 46 to 126 mmHg, suggesting that true episodes of hypotension occurred.

In contrast, hypotension did not occur during VAP-C, even when mild sedation was needed. It may be argued that the use of acepromazine in the CON-C group biased results in favor of VAP-C. Acepromazine is routinely used in CON-C at our institution to facilitate peripheral vein catheterization and decrease vomiting and dysphoria upon recovery from anesthesia with ketamine. In an effort to determine the contributory effect of acepromazine on SBP, a second study was performed on 6 blood donors. No differences were found in SBP values compared to the 8 study cats that received acepromazine. The relatively low dose of acepromazine was therefore not considered to be the sole factor in the anesthesia protocol accounting for the differences in SBP between cats undergoing CON-C versus VAP-C.

The more severe hypotension observed with CON-C may have been due to anesthesia-induced impairment in restorative reflexes necessary to compensate for blood loss (22). However, sedation was also given for many of the VAP-C, and it is also possible that the more rapid blood collection with CON-C (< 10 min) compared to VAP-C (> 10 min for 75% of the collections) was responsible for hypotension. With both collection methods the SBP nadir occurred at a mean of 5 min. It may be that acute blood loss during the first 5 min with both collection methods was sufficient to trigger restorative reflexes, and that the more severe hypotension associated with CON-C reflected more severe volume depletion during that period. A more recent study of SBP changes during blood collection of 50 mL by jugular venipuncture using sevoflurane anesthesia documented less hypotension, although SBP was measured by a different method, and average collection times were probably longer than with CON-C in our study (2).

Vocalization, and syncope or defecation, occurred after blood loss of 25 to 35 mL [corresponding to 10% to 15% total blood volume (1)] during the first VAP-Cs in 2 cats. Heart rate and SBP were normal at the time, but the episodes were of such short duration that transient absolute bradycardia and/or hypotension may not have been detected. Vocalization and defecation were also considered to be consistent with anxiety-induced behaviors, as adrenergic response, which occurs in response to substantial blood loss, and anxiety, are closely related (23,24). The collapse in one cat was also potentially a sign of vasovagal syncope (25,26). We postulated that reducing the acuteness of blood loss or treatment with anxiolytic drugs would prevent recurrence (27,28). From then on, collections were performed no faster than over 10 min for all cats, and cats 2 and 8 were administered low doses of ketamine and diazepam through the VAPs prior to each collection. Four other cats were also sedated to facilitate cervical manipulation; therefore, the relative benefits of slower collection and sedation are not known.

The use of a VAP could potentially increase RBC damage during collection. This was assessed by measuring physical and metabolic properties of pRBCs. Feline pRBCs are considered suitable for transfusion for up to 35 d when stored in a nutrient solution (29), but there are no reports of RBC properties during such storage. A storage period of 25 d was chosen as reasonably long period where pRBCs would by consensus be considered suitable for transfusion. Comparison of pre and post-storage glucose, pH, potassium, and LDH levels and MOF revealed changes characteristic of the “storage lesion” in other species in both the CON-C and VAP-C pRBCs (30,31). There were no differences between CON-C and VAP-C pRBCs except for pH and MOF. The difference in pre-storage pH between collection methods was attributed to the higher pCO₂ in blood obtained by CON-C, which was likely a consequence of drug-induced hypoventilation. The higher initial pH in VAP-C pRBCs did not, however, affect drop in pH during storage, which is due to lactic acid accumulation from glycolysis (30,31).

The VAP-C pRBCs were more fragile than CON-C pRBCs, both before and after storage, although MOF increased by the same proportion during storage in both groups. This is consistent with previous studies of VAPs in cats and humans that suggest that RBC passage through the system causes some membrane damage (16,32). However, in the previous study of VAPs for feline blood donation, there was no difference in RBC morphology between the 2 collection methods, although blood had not been separated into plasma and pRBCs (13). Osmotic fragility correlates with post-transfusion RBC viability in humans, but the importance of the difference between collection methods seen in our study is not known, and increased osmotic fragility may normalize upon transfusion (14,30,33). It is also not known if syringe-collection, as opposed to gravity-collection, exacerbated VAP-induced RBC damage.

One of the postulated advantages of VAP-C is that it may be performed rapidly without donor anesthesia, allowing for collection on an as-needed basis rather than for blood banking. Blood quality was therefore compared on relevant parameters between pre-storage VAP-C pRBCs and post-storage CON-C pRBCs. Results suggests that fresh pRBCs obtained by VAP-C are at least as suitable for transfusion as pRBCs obtained by CON-C and stored for 25 d.

None of the CON-C or VAP-C pRBCs had positive pre or post-storage microbiological cultures, supporting the consensus that open collection systems in cats may yield blood products without significant bacterial contamination (1,13,34). In the previous report, 4 of 20 VAP-C units of blood were contaminated, although the VAPs used were not considered to be infected (13). The presence of an implant is normally a criterion for donor exclusion (1), and this is a particularly important consideration when the implant is repeatedly invaded through the skin for blood collection. One VAP became infected in our study. It is likely that contamination of the peri-portal tissues occurred during post-operative dehiscence, and then VAP contamination occurred during access. The VAP infection responded well to systemic and local antibiotic therapy with enrofloxacin. It is acknowledged that the parenteral preparation of enrofloxacin is not approved for intravenous use, nor for use in cats, and that use of the drug beyond standard dose recommendations increases the risk for retinopathy (35). The drug was chosen as it allowed for an antibiotic-lock and continuity of therapy from an injectable to an oral preparation, and retinal toxicity was not widely recognized at the time the study was conducted. Although clinical signs of sepsis resolved and culture-negative blood products were subsequently collected, the recurrence of skin lesions suggests that peri-portal infection had not been eradicated. Furthermore, catheter tips were not cultured upon removal, nor were catheters examined for infection by more sensitive methods such as electron microscopy (15), so catheter infection may have been underdiagnosed. One VAP became infected in the previous report, which prompted removal (13). The results of both studies emphasize that VAPs and VAP-collected blood products must be closely monitored microbiologically.