Roller Pump Induced Tubing Wear: Another Argument in Favor of Arterial Line Filtration

____________ _ Pump head tubing was studied to determine if internal wear occurred during routine cardiopulmonary bypass. Several segments of silicone rubber tubing were examined in a scanning electron microscope following clinical perfusion. In each case, the proper degree of occlusion had been set just prior to beginning bypass. The roller pump was a six-inch dual roller type, and the tubing was 3/8 inch ID with 3/32 inch wall thickness. Blood flow rates ranged from 3.8 to 5.2liters per minute, and duration of the pump runs ranged from 35 to 220 minutes. Evidence of tubing wear on the lumenal surfaces was observed and was related to time and flow rates. Two grooves opposite one another and corresponding to the location of maximum flexure were present in every sample examined. Alterations in the tubing surface were seen in the areas adjacent to the grooves. Craters ranging from <10 tim to >50 tim along with some degree of smoothing of the normal surface texture were seen. Presumably, some spallation had occurred, and particles had been pumped downstream. Use of an arterial line filter with this type of tubing is recommended. Introduction, _____________ _ The possibility of outright pump head tubing failure rarely concerns perfusionists today, although there have Presented at the 17th International AmSECT Conference Houston Texas, 5-8 March, 1979. ' Address correspondence to: M. Kurusz, Division of Cardiovascular and Thoractc Surgery, M-603 John Sealy Hospital, The University of Texas Medtcal Branch, Galveston, Texas 77550 been reports of this rare accident occurring during cardiopulmonary bypass.l.2 A less dramatic type of tubing failure occurs during perfusion when silicone rubber is used in a roller pump head despite the proper degree of occlusion being set and the rollers being properly balanced within the manufacturer's tolerances. This type of failure appears to be time and flow rate dependent and consists of abrasion and erosion of the lumenal surface of silicone rubber tubing during roller compression during pumping. As a result, another source of microemboli is produced within the extracorporeal circuit. In 197 5, Hubbard and coworkers3 filtered fluid pumped in a mock circulatory loop and demonstrated spallated particles with light microscopy to point out the superiority of a constrained vortex centrifugal pump. This paper presents additional information obtained by scanning electron microscopy (SEM) of clinical segments of pump head tubing to illustrate the nature and extent of tubing wear that occurs when this type of tubing is used in a roller pump. SEM has been used previously to characterize blood surfaces found in the typical extracorporeal circuit,4 but this is the first report of the application of SEM to segments of roller pump tubing that have been used clinically. Materials and Methods ________ _ Medical grade, close tolerance silicone rubber tubing* measuring 3/8 inch internal diameter with a 3/32 inch wall thickness was used during clinical cardio* Medical Engineering Corporation, Racine, Wisconsin 53404 v Sarns, Inc., Ann Arbor, Michigan 48103 Volume 12, Number 2, 1980 The Journal of Extra-Corporeal Technology 49 pulmonary bypass in a six-inch dual roller pumpv' for arterial infusion of blood. Immediately prior to bypass, the degree of occlusion was carefully set to be just under occlusive.5·6 Both roller-to-housing tolerances were checked prior to surgery, and both were found to be nearly identical and within the pump manufacturer's tolerance of .0015 inches. (Personal communication, Mr. Herb Hammond, Sarns, Inc.) Following cardiopulmonary bypass, the arterial pump head tubing was carefully removed from the roller pump in the same direction as it was used during the perfusion. Blood was drained from the tubing, and a 2-3 centimeter segment of whole tubing was cut and removed from the center portion of the pump head segment using a clean # 21 scalpel blade. The distal end was notched to note the direction of roller movement. Because no attempt was being made to study blood/tubing interaction, the segment was immediately rinsed in flowing tap water and then ultrasonically cleaned in detergent# for approximately five minutes. Following this cleaning to remove blood, the segment was again rinsed in flowing tap water and blown dry with compressed air. It was then immediately covered to prevent surface contamination by air-born particles. At no time was the lumenal surface to be examined touched with any instrument. In preparation for SEM, the segments were transected with a clean # 21 scalpel blade leaving approximately one centimeter segments. The ends were discarded, and the center segment was split longitudinally into quarters. One quarter segment exhibiting groove formation visible to the naked eye was chosen for SEM examination. This groove corresponded to one of the two areas of maximal flexure caused by roller compression of the tubing against the pump housing. Segments for examination were then mounted on 1.5 centimeter aluminum stubs with quick-drying epoxy cement and sputter-coated with gold-palladium in a Hummer IJ+ apparatus to make them conductive. They were coated for one minute at 5 rnA current which deposited a very thin layer of metal approximately 100 A in thickness. A stub-to-tubing trail of silver paint ensured conductivity between the sample surface and the stub. Following application of the gold palladium, the tubing samples were immediately examined in an lSI Super III% scanning electron micro# Liqui-Nox, Alconox, Inc., New York, New York 10003 +Technics, Inc., Alexandria, Va. 22310 % International Scientific Instruments, Inc., Mountainview, Ga. 94043 scope at 15 kV and 45°-50° tilt from the horizontal to enhance surface topographical features. Control samples of tubing were prepared in a similar fashion as clinical tubing samples including the ultrasonic cleaning. Results _______________ _ Figures One and Two are SEM micrographs of the lumenal surface of unused silicone rubber tubing as received in a sterile custom tubing pack.¢ The bore of the tubing runs vertically in the micrograph, and the surface features consist of gentle undulations with occasional smooth-edged blebs and pits. On a blood cellular level, such a surface would be termed fairly smooth. This surface texture is produced by the tubing manufacturing process as silicone rubber tubing is extruded and cured. (Personal communication, Mr. Tom Brown, Dow Corning Corp.) Clinically used tubing samples (Figs. 3-11) showed evidence of tubing wear on the lumenal surfaces. This wear appeared to be related to bypass time and flow rate, in that, an increased number of roller strokes on a tubing segment coincided with a more altered surface. Determination of the number of strokes a particular tubing segment received was calculated by multiplying the average number of revolutions per minute (RPM) times two (for each roller) times the pump time in minutes. Typically, a longitudinal groove was seen in every sample examined. Figure 3 is a low magnification view of this groove, and it corresponds to the location of maximum tubing flexure caused by roller compression. Figure 4 is a higher magnification of the same sample and shows blebs arising from the tubing material itself. Surrounding areas of tubing are relatively smooth with occasional shallow, smooth-edged craters. Figure 5 shows groove formation and bleb formation running parallel to the groove on a tubing segment that was used clinically for 121 minutes. Figure 6 is a higher magnification view of the blebs which are clearly shown to be part of the tubing. Small pits ( <5 ,urn) are also present. Relatively short pump runs also produced tubing alterations, although, to a lesser degree. Figure 7 is a sample from a 35 minute pump run, and the typical crater and bleb formation is present. Further SEM evidence that the lumenal surface of tubing not only is damaged but can erode potentially dangerous m10 Cobe Laboratories, Inc., Lakewood, Co. 80215 50 The Journal of Extra-Corporeal Technology Volume 12, Number 2, 1980 FIGURE 1. Control silicone rubber tubing. Medical grade, close tolerance; 3Js inch internal diameter and 3h2 inch wall thickness. Bore of tubing runs vertically in the micrograph. Bar= I 00 Jlm. FIGURE 2. Control silicone rubber tubing at a higher magnification than Figure 1. Note the generally smooth, undulating surface. The occasionally adherent white particles were presumably deposited during specimen preparation and represent air-born contaminants. Bar= 10 Jlm. Volume 12, Number 2, 1980 The Journal of Extra-Corporeal Technology 51 FIGURE 3. Clinically used silicone rubber tubing. Pump time 78 minutes; average flow rate 5250 mljminute ( 190 RPM) or approximately 29,640 roller strokes. Note the distinct groove formation and areas of increased surface roughness adjacent to the groove. Craters are up to 25 JLm in width. Bar = I 00 JLm. FIGURE 4. Clinically used silicone rubber tubing. Same sample as Figure 3 but at a higher magnification. Note craters adjacent to the groove and pitting. Bar = 20 JLm. 52 The Journal of Extra-Corporeal Technology Volume 12, Number 2, 1980 FIGURE 5. Clinically used silicone rubber tubing. Pump time 121 minutes: average flow rate 4500 ml/minute ( 165 RPM) or approximately 39,930 roller strokes. Note material ncar the groove and normal silicone rubber surface on either side of the groove. Bar= 50 J.Lm. FIGURE 6. Clinically used silicone rubber tubing. Same sample as Figure 5 but at a higher magnification. Note multiple bleb formation running parallel to the groove. Bar = 20 J.Lm. Volume 12, Number 2, 1980 The Journal of Extra-Corporeal Technology 53 FIGURE 7. Clinically used silicone rubber tubing. Pump time 35 minutes; average flow rate 4250 ml/minute (!55 RPM) or approximately I 0,850 roller strokes. Note blebs near large (50 /-Lm), shallow crater and smaller rough-edged craters to the left. Bar= 20 /-Lm. FIGUREs. Clinically used silicone rubber tubing. Pump time 220 minutes; average flow rate 3850 mljminute (140 RPM) or 61,600 roller strokes. Note disruption in the tubing surface and a particle partially detached from the tubing. Bar= 10 /-Lm. 54 The Journal of Extra-Corporeal Technology Volume 12, Number 2, 1980 FIGURE 9. Clinically used silicone rubber tubing. Pump time 165 minutes; average flow rate 4650 ml/minute ( 170 RPM) or 56,100 roller strokes. Note the typical groove formation and surface irregularities adjacent to the groove Bar = I 00 ,urn. FIGURE 10. Clinically used silicone rubber tubing. Same sample as Figure 9 but at a higher magnification. Note jagged tears and pitting on the surface of the tubing. Bar= 10 ,urn. Volume 12, Number 2, 1980 The Journal of Extra-Corporeal Technology 55 FIGURE 11. Clinically used silicone rubber tubing. Same sample as Figures 9 and I 0 but at a higher magnification. Note partially eroded particle within a crater that measures approximately 12.5 ttm in diameter. Bar = I 0 ttm. croemboli is shown in Figure 8. This sample was taken from a tubing segment that was used clinically for 220 minutes. The center of the micrograph shows a relatively large particle (approximately 20J.tm) of silicone tubing partially eroded from the tubing surface. One side of the particle is continuous with the surrounding surface, so, presumably, it was not deposited during preparation, but instead emanated from the tubing material. The area around this particle is smoother than the control silicone rubber tubing. The last three micrographs show the effects of roller damage to tubing that was used clinically for 165 minutes. Figure 9 is an overall, low magnification view of the characteristic groove formation. As in previous samples, the area on either side of the groove shows signs of crater formation and pitting. In areas away from the groove the tubing surface is normal. Figure l 0 shows severe pitting of the tubing surface. Roughedged craters have formed in the area immediately adjacent to the groove. Figure II shows a particle from the same sample that was nearly detached from the bulk tubing. The diameter of the particle is approximately 12.5 Jlm. Smaller pits are seen immediately below the particle. As these SEM micrographs demonstrate, changes in the normal surface features of silicone rubber tubing employed in a dual roller pump at clinical flow rates routinely occur. These changes appear to be time and flow rate dependent. That is, with longer pump runs or. increased flow rates, the tubing damage is more severe. The two predominant features are pitting or crater formation and bleb formation. Both types of surface changes can presumably lead to spallation of the tubing material into the flowing blood. The largest craters consistently observed were in the range of 20 to 30