Anaesth. Intens. Care (1979), 7, 239
THROMBUS FORMATION ON CATHETERS IN NEW BORN LAMBS J.
F. HECKER*, G. C. FISKt, J. M. GUPTA:j:, N. ABRAHAMSt, R. A. COCKINGTON:j: AND B. R. LEWIst
University of New England, Armidale, N.S.W. and The Prince of Wales Hospital, Sydney SUMMARY
Catheters were inserted into the aorta and inferior vena cava of newborn lambs by cutdown on the femoral vessels and directly into the umbilical vein. The lambs were killed after six days and the amounts of thrombus surrounding the catheters were measured. Most catheters inserted into the umbilical vein looped within the liver and only a few passed directly into the portal vein or through the ductus venosus into the vena cava. On different types of catheters inserted via the femoral artery or femoral vein, there were significant differences both in the proportion with thrombus and the amount of thrombus. There were also significant differences in renal infarction caused by different types of catheters.
INTRODUCTION
There are numerous references in the literature to problems associated with thrombus formation on catheters inserted into umbilical vessels of newborn infants. Thrombus has been shown on umbilical artery catheters by pullout angiography and at autopsy (Cochran et al. 1971, Egan and Eitzman 1971, Goetzman et al. 1975, Gupta et al. 1968, Marsh et al. 1975, Neal et al. 1972, Olinski et al. 1975, Strauss et al. 1974, Symansky and Fox 1972, Tooley 1972, Wigger et al. 1970). In some babies embolized thrombus causing infarcts has been the probable cause of death (Tooley 1972, Wigger et al. 1970). Renal hyptertension (Bauer et al. 1975, Plumer et al. 1976) and paraplegia (Aziz and Robertson 1973, Krishnamoorthy et al. 1976) have been noted in others with umbilical artery catheters. Umbilical vein catheters also form thrombus (Symansky and Fox 1972, Oski et al. 1963, • Department of Physiology, University of New England, Armidale, N.S.W. t Department of Anaesthesia and Intensive Care, Prince of Wales Hospital, Randwick, N.S.W. t Prince of Wales Children's Hospital, Randwick, N.S.W. Address for reprints: Dr. 1. F. Hecker, Department of Physiology, University of New England, Armidale, N.S.W. 2351, Australia.
Anaesthesia and Intensive Care, Vol. VII, No. 3, August, 1979
Scott 1965) and have caused portal hypertension (Oski et al. 1963), hepatic infarction (Riggs et al. 1976, Scott 1965), and inferior vena cava obstruction (Tooley 1972) in babies. In nearly all of these references, the Argyle umbilical catheter was used and we suspect that this catheter is used world wide for most umbilical vessel catheterizations. At the Prince of Wales Children's Hospital, a series of autopsies was performed on babies who died after periods of respiratory distress and who had umbilical artery catheters inserted. Although death was usually due to the respiratory distress syndrome, the autopsies regularly showed thrombus surrounding catheters and some infarcted vessels. This thrombus appeared identical to that seen on catheters in experimental sheep and we were prompted to use sheep to test Argyle umbilical catheters in comparison with other catheters used in humans and animals (Hecker et al. 1976). That experiment indicated that the Argyle and the Pharma-Plast umbilical artery catheters were comparatively thrombogenic. Fetal and newborn lambs are used extensively as models for human disease (Hecker 1974). In this paper, we present the results of an
240
J. F. HECKER ET A L.
experiment in which we have used the newborn lamb as a model for testing thrombus formation subsequent to insertion of commercially available catheters.
MATERIALS AND METHODS
Merino lambs were separated from their mothers within 24 hours of birth in October, 1976, November, 1977, and September, 1978. Anaesthesia was induced with pentobarbital sodium and maintained when necessary with Ketamine. In 1976 it was intended that tubing would be inserted directly into an umbilical artery of each lamb but after several unsuccessful attempts to catheterize umbilical arteries this approach was abandoned and instead tubing was inserted into the aorta by cutdown into a femoral artery. In babies most complications of umbilical artery catheters are due to the intra-aortic portion and the only reported complications that potentially cannot be reproduced in this lamb model are those involving the leg. In 1976, tubing was also inserted into the umbilical vein of each lamb but for reasons given below this route was not used in later years and instead tubing was inserted by cutdown into the femoral vein on the side opposite to the catheterized femoral artery. The following tubings were used: Argyle umbilical catheter (Sherwood Medical Industries Inc., St. Louis, Mo., U.S.A.) , 5 F.G., Argyle umbilical catheter, 5 F.G., treated with TDMAC and heparin (Hecker et al. 1976), Pharma-Plast umbilical catheter (PharmaPlast (Australia) Pty. Ltd., Sydney, N.S.W., Australia), 5 F.G.*, Portex umbilical catheter (Portex Ltd., Hythe, Kent, England), 6 F.G., Dural Plastics polyvinyIchloride tubing (Dural Plastics Ltd., Dural, N.S.W., Australia), 1.5 mm outside diameter, Dow Coming radio-opaque silicone rubber tubing (Dow Coming Corp., Hemlock, Mi., U.S.A.), 1.5 mm outside diameter, Pharmaseal K 15 umbilical catheter (Pharmaseal Laboratories, Glendale, Calif., U.S.A. 91201), 5 F.G., Angiodan umbilical catheter (Surgimed Pty. Ltd., Totfield House, 13-15 Thompson St., Frankston, Vic. 3199). The Argyle TDMAC catheters were not tested in femoral veins as insufficient numbers
were available. Catheters were filled and inserted to a distance of 17 to 22 cm. This placed the tips in the thoracic cavity. On insertion, the exterior end of each catheter was closed and left under the skin to avoid infection. Benzathine penicillin was given at the time of surgery by intramuscular injection. No attempts were made to flush catheters. After surgery, lambs were fed reconstituted powdered milk three or four times each day in 1976 and 1977 but in 1978 were returned to their mothers. At six days after surgery, lambs were anaesthetized with pentobarbital sodium to which had been added 200 i.u. of heparin and were killed by exsanguination. Catheterized vessels were dissected and opened and thrombus associated with catheters or the vessel wall was photographed and weighed. Kidneys, liver, other abdominal organs, lungs and the contralateral iliac artery were inspected for signs of infarction. Data (mass of thrombus and area of renal infarcts) were subjected to analysis of variance after square root transformation to fulfil the assumptions of normality and homogeneity of variances underlying analysis of variance. The un transformed data was not normally distributed and had non-homogeneous variances whereas after square root transformation the distribution was normal (Chi square test) and variances were homogeneous (Bartlett's test) (Pollard 1977). Significant differences between types of tubing were identified with Duncan's multiple range test (Duncan 1955).
TABLE 1
Numhers of catheters with thrombus ill arteries alld veills of lambs The Chi square test for arterial + venous catheters was calculated without data for Argyle TDMAC. Tubing Portex Angiodan Dow Coming Dural Pharmaseal Argyle Pharma-Plast Argyle TOMe Total
X2
Arterial
Venous
5/12
4/8
3 ti ti
ti
7 11 10 9 57;96 17.58*
5 7 7 7 8 44/56 8.48 NS
Arterial venous
+
9/20 9 11 13 14 18 18 82/140 19.28i"
= not significant = p < 0.05 'i' = p < 0.01
NS
* • This catheter is no longer produced.
Anaesthesia a",/ Intemire Care, Vol. 1'11, No. 3, August, 1979
241
THROMBUS FORMATION ON CATHETERS
aorta but was found on some catheters to cover the entire length. The distal end of the external iliac artery was taken as a reference point and the mass of thrombus on the 15 cm length cephalad to this point was recorded. Results are shown in Table 2. The Pharma-Plast and Argyle catheters had significantly more thrombus than did other types.
RESULTS
Arterial catheters: Of 96 arterial catheters (12 of each type), 57 were associated with thrombus (Table 1). There were significant differences between types of catheter in the proportion with thrombus (Table 1). On catheters with thrombus, the thrombus was most common distal to the bifurcation of the
TABLE
2
Mass of thrombus on catheters inserted into the aorta via the femoral artery and the vena cava via the femoral vein of new born lambs. Results are for the 15 cm cephalad to the distal end of the external iliac artery for 12 lambs and to where the femoral vein enters the abdominal cavity for 8 lambs. Lambs were killed at 6 days. The analysis of variance was done afrer square root transformation. Significant differences between tubing are indicated by different letters. *p < 0.05, tp < 0.01, tp < 0.001. Catheter
Femoral artery (12 lambs) (mean ± S.E.)
Portex Angiodan Dow Coming Dural Pharmaseal Argyle TDMAC Argyle Pharma-Plast
27 32 50 20 22 44 85 100
ANALYSIS OF VARIANCE Source d.f. mean square Tubing 7 12.99 Vessel Residual 88 3.78
F 3.44t
± ± ± ± ± ± ± ±
Femoral vein (8 lambs) (mean ± S.E.)
14 mg a a 29 a 25 a 11 a 10 ab 16 b 22 b 27 d.f. 6
44 60 37 57 139
mean square F 2.35* 75.28
49
32.07
TABLE 3
Portex Angiodan Dow Coming Dural Pharmaseal Argyle Pharma-Plast Argyle TDMAC
Number with Renal Infarction
Area of Renal Infarction mm2 (mean ± S.E.)
Oa
0/12 4
69.7 0.2 1.5 41.1
I 1
6
Chi2
o o o
= 359;
Oa Oa
± 37.0b ± 0.2a ± 0.5a ± 28.8b
oa
F
16 mg a ab 36 a 19 ab 16 ab 68
168 ± 58 141 ± 33
Numbers of animals with renal infarction and surface areas of renal infarction in lambs in which catheters were inserted into the aorta The value of F was from analysis of variance which was done after square root transformation. Significant differences between catheters are indicated by different letters. 0 P < 0.(01) Catheters
± ± ± ± ±
= 4.24t
The only signs of infarction were in kidneys and these infarcts were seen mainly with two types of catheter (Table 3). Differences Anaesthesia and Intensive Care, Vol. VII, No. 3, August, 1979
b b
Femoral artery + vein (20 lambs) (mean ± S.E.) 37 ± 15 mga a 43 ± 21 a 45 ± 16 a 35 ± 10 a 69 ± 30 116 ± 21 118 ± 27
F 6 1 129
mean square 135.67 467.75 24.96
b
b
d.f. 5.44t 18.74;
between catheters in the number of animals with infarcts and in surface areas of infarcts were significant. Umbilical vein catheters: In only 10 of 63 lambs used in 1976 did the umbilical vein catheter pass directly through the ductus venosus and into the inferior vena cava. In seven others, catheters went into the portal vein and then into the intestinal venous trunk. The remaining 46 umbilical vein catheters looped or doubled back in the umbilical vein, ductus venosus or branch of the portal vein supplying the ventral lobe of the liver, apparently because the tips had first lodged in small branohes of the portal vein supplying the ventral lobe of the liver. Every umbilical vein catheter was associated with thrombus for part of its length. Most was in intrahepatic veins. The extrahepatic portion of the portal vein and the intestinal trunk were virtually free of thrombus and only in one lamb was thrombus found around a catheter in the vena cava. Because of the doubling back
J. F. HECKER ET AL.
242
of the majority of catheters, mass of thrombus on umbilical vein catheters is not presented. Femoral vein catheters: Results for individual catheters are shown in Tables 1 and 2. The Argyle and Pharma-Plast catheters had significantly more thrombus than did the Portex umbilical catheter or the Dow Coming tubing. Combined data: Results from the femoral artery and femoral vein ca~heters (less results for Argyle TDMAC) have been combined in Tables 1 and 2. Argyle and Pharma-Plast catheters had significantly more thrombus than did the other types but there were no significant differences between the Argyle and PharmaPlast catheters or between the other catheters. DISCUSSION
The Pharma-Plast catheter is no longer produced but it was included in this experiment because it has been used extensively in some Australian hospitals. Previously we reported phlebitis when this catheter was inserted into veins of adult sheep (Hecker et al. 1976). Both F.G. 5 and F.G. 6 sizes were used in the adult sheep experiment and a review of data subsequent to writing that paper showed that phlebitis had occurred only with the F.G. 6 catheters. In this experiment, only F.G. 5 catheters were used and no phlebitis was seen. We have seen only one reference to the use of the newborn lamb for umbilical catheterization. This reported that the tip of an umbilical vein catheter entered the substance of the liver (De Yore et al. 1976). From our results, it is obvious that vena caval catheterization via the umbilical vein should be done in lambs only with the aid of X-ray fluoroscopy. Difficulty in passing umbilical vein catheters through the ductus venosus of babies has been reported (Oski et al. 1963, Peck 1967, Rosen and Reich 1970, Symansky and Fox 1972). In 1976, we also cannulated some lambs which were killed after three days. Most catheters at this time were free from thrombus. The mean mass on tubings at three days was 11 mg while that on similar tubings at six days was 54 mg. It is likely that thrombus formation is a gradual process and that amounts would continue to increase if catheters were left in for longer periods. There was considerable variation in the, amounts of thrombus that formed. This considerable variation is present also with catheters in veins of adult sheep (Hecker et al. 1976), and is almost certainly present with catheters in humans. Because of such variation,
it is necessary to ~tudy reasonable numbers if differences in thrombogenicity are to be identified. Significant differences were identified in this experiment, the Argyle and Pharma-Plast catheters being the most thrombogenic in both arteries and veins. One difference between the present results and those of our previous experiment (Hecker et al. 1976) is that Pharma-Plast catheters were significantly more thrombogenic in veins of adult sheep than Argyle catheters. The reason for this difference is not apparent but it may be due to differences between batches of Pharma-Plast catheters. Silicone rubber was significantly less thrombogenic, both in this experiment and in our previous experiment, than the Argyle catheters. Boros et al. (1975), by "pullout" angiography, compared in babies a silicone rubber catheter with the Argyle catheter and obtained a similar result to ours. The silicone rubber used by Boros et al. (1975) and the silicone rubber used in these experiments were supplied by different companies and were relatively thromboresistant. However we believe that not all brands of silicone rubber are thromboresistant as oreliminary tests in veins of adult sheep on tubing used for human ventriculo-atrial shunts have shown it to be more thrombogenic. The finding that some tubings cause greater renal infarction than others is interesting. The infarction is due to thrombus which has dislodged from catheters and the presence of renal infarction with some types of catheters (Table 3) suggests that they are more thrombogenic than values in Table 2 would indicate. In experiments in which catheters were inserted into the aorta of adult sheep, we have also found that some plastics produce more renal infarcts than others. T'he TDMAC process was developed in the United States by Battelle Laboratories for incorporating a surface layer of heparin on plastics. We used some TDMAC treated Argyle catheters in a prospective blind trial in umbilical arteries of new born babies suffering from respiratory distress and obtained adequate data from 16 babies with TDMAC treated and 21 with untreated Argyle catheters. There was no significant difference in the incidence of complications such as blocked catheters. Single shot withdrawal angiograms in seven babies with treated and 10 with untreated catheters showed no definite evidence of thrombus. Four babies which received treated and four which Ana~sth~sia
and Tnte",i.'e Care. Vol. Vll, No. 3. AlIglIIt. 1979
THROMBUS FORMATION ON CATHETERS
received untreated catheters died and in each of these groups two babies had the catheter in situ at death while the catheter had been removed before death in two. Each of the eight babies had thrombus present in the internal iliac artery. Various in vitro tests have indicated that the TDMAC process increases thromboresistance (Bruck 1974). The above small experiment with babies, our previous experiment with adult sheep (Hecker et al. 1976) and this experiment suggest that the TDMAC treatment of Argyle catheters is of limited value. Though treatment of plastics with TDMAC may reduce 1'hrombogenicity, it would seem preferable to use an inherently less thrombogenic plastic than the polyvinylchloride from which the Argyle catheter is made. This may then be treated with TDMAC. Such plastics are available as the Portex catheter (made also from polyvinylchloride) did perform well in lambs and in veins of adult sheep (unpublished results for the Portex catheter). ACKNOWLEDGEMENTS
We wish to thank Sherwood Inds., Portex Ltd. and Dow Coming Corporation for supplying catheters and tubing, Battelle Laboratories for applying the TDMAC process and the National Health and Medical Research Council for financial support. The project would not have been possible but for the help of the several volunteer lamb feeders. REFERENCES
Aziz, E. M., and Robertson, A F. (1973): "Paraplegia: A complication of umbilical artery catheterization", I. Pediatr., 82, 105l. Bauer, S. B., Feldman, S. M., Gelliss, S. S., and Retik, A. B. (1975): "Neonatal hypertension: A complication of umbilical-artery catheterization", New En%!l. I. Med., 293, 1032. Boros, S. J., Thomoson, T. R .. Reynolds, J. W., Jarvis, C. W., and Williams, H. J. (1975): "Reduced thrombus formation with silicone elastomer (Silastic) umbilical artery catheters", Pediatrics, 56, 98l. Bruck, S. D. (1974): "Blood Compatible Synthetic Polymers: An Introduction", Thomas, Springfield, I11. Cochran, W. D., Davis, H. T., and Smith, C. A. (1968): "Advantages and complications of umbilical arterv catheterization in the newborn", Pediatrics, 42. 769. De Vore, J. S .. Wheeler, A. S., and Latyshevsky, H. (1976): "Misplacement of an umbilical vein catheter". Anesthesiology. 44, 451. Duncan, D. B. (1955): "Multiple range and multiple F tests", Biometrics, 11, 1. Egan, E. A, and Eitzman. D. V. (1971): "Umbilical vessel catheterization", Am. I. Dis. Child, 121, 213. Ford. K. T., Teplick, S. K., and Cl ark, R. E. (1974): "Renal artery embolism causing neonatal hyperAnaesthesia and Intensive Care. Vol. VIl. No. 3, August, 1979
243
tension: A complication of umbilical artery catheterization", Radiology, 113, 169. Goetzman, B. W., Stadalnik, R. c., Bogren, H. G., Blankenship, W. J. Ideda, R. M., and Thayer, J. (1975): "Thrombotic complications of umbilical artery catheters; A clinical and radiographic study", Pediatrics, 56, 374. Gupta, J. M., Robertson, N. R. C., and Wigglesworth, J. S. (1968): "Umbilical artery catheterization in the new born", Arch. Dis. Child, 43, 382. Hecker, J. F. (1974): "Experimental Surgery on Small Ruminants", Butterworths Ltd., London. Hecker, J. F., Fisk, G. C., and Farrell, P. C. (1976): "Measurement of thrombus formation on intravascular catheters", Anaesth. In tens Care, 4, 225. Krishnamoorthy, K. S., Fernandez, R. J., Todres, I. D., and DeLong, G. R. (1976): "Paraplegia associated with umbilical artery catheterization in the newborn", Pediatrics, 58, 443. Marsh, J. L., King, W., Barrett, C., and Fonkalsrud, E. W. (1975): "Serious complications after umbilical artery catheterization for neonatal monitoring", Arch. Surg., 110, 1203. Neal, W. A, Reynolds. J. W., Jarvis, C. W., and Williams, H. J. (1972): "Umbilical artery catheterization: Demonstration of arterial thrombus by aortography", Pediatrics, 50, 6. Olinskv, A, Aitken, F. G., and Isdale, J. M. (1975): "Thrombus formation after umbilical arterial catheterization", South Afr. Med. I., 49, 1467. Oski, F. A. Allen, D. M., and Diamond, L. K. (1963): "Portal hypertension - a complication of umbilical vein catheterization", Pediatrics, 31, 297. Peck, D. R., and Lowman, R. M. (1967): "Roentgen aspects of umbilical vascular catheterization in the newborn", Radiology, 89, 874. Plumer. L. B .. Kaplan, G. W., and Mendoza, S. A. (1976): "Hypertension in infants - a complication of umbilical arterial catheterization", I. Pediatr., 89, 802. Pollard, J. H. (1977): "A Handbook of Numerical and Statistical Techniques with Examples Mainly from the Life Sciences", Cambridge University Pre~s, Cambridge. Riggs. T .. Hirschfeld, S., Borkat, G., and Liebman, J. (1976): "Inferior vena cava obstruction secondary to indwelling venous catheters: Two cases", Pediatrics, 58. 446. Rosen, M. S., and Reich, S. B. (1970): "Umbilical venous catheterization in the newborn: Identification of correct positioning", Radiology, 95, 335. Scott. J. M. (1965): "Iatrogenic lesions in babies following umbilical vein catheterization", Arch. Dis. Child, 40. 426. Strauss, A W., Escobedo, M., and Goldring, D. (1974): "Continuous monitoring of arterial oxygen tension in the newborn infant", I. Pediatr., 85, 254. Symansky, M. R., and Fox, H. A (1972): "Umbilical vessel catheterization: Indications, management, and evaluation of the technique", I. Pediatr., 80, 820. Tooley, W. H. (1972): "What is the risk of an umbilical artery catheter?", Pediatrics, 50, 1. Tyson. J. E., deSa, D. J., and Moore, S. (1976): "Thromboatheromatous complications of umbilical arterial catheterization in the newborn period: CHnico-pathological study", Arch. Dis. Child, 51, 744. Wigger, H. J .. Bransilver, B. R., and Blanc, W. A. (1970): "Thromboses due to catheterization in infants and children", I. Pediatr., 76, 1.
THROMBUS FORMATION ON CATHETERS IN NEW BORN LAMBS J.
F. HECKER*, G. C. FISKt, J. M. GUPTA:j:, N. ABRAHAMSt, R. A. COCKINGTON:j: AND B. R. LEWIst
University of New England, Armidale, N.S.W. and The Prince of Wales Hospital, Sydney SUMMARY
Catheters were inserted into the aorta and inferior vena cava of newborn lambs by cutdown on the femoral vessels and directly into the umbilical vein. The lambs were killed after six days and the amounts of thrombus surrounding the catheters were measured. Most catheters inserted into the umbilical vein looped within the liver and only a few passed directly into the portal vein or through the ductus venosus into the vena cava. On different types of catheters inserted via the femoral artery or femoral vein, there were significant differences both in the proportion with thrombus and the amount of thrombus. There were also significant differences in renal infarction caused by different types of catheters.
INTRODUCTION
There are numerous references in the literature to problems associated with thrombus formation on catheters inserted into umbilical vessels of newborn infants. Thrombus has been shown on umbilical artery catheters by pullout angiography and at autopsy (Cochran et al. 1971, Egan and Eitzman 1971, Goetzman et al. 1975, Gupta et al. 1968, Marsh et al. 1975, Neal et al. 1972, Olinski et al. 1975, Strauss et al. 1974, Symansky and Fox 1972, Tooley 1972, Wigger et al. 1970). In some babies embolized thrombus causing infarcts has been the probable cause of death (Tooley 1972, Wigger et al. 1970). Renal hyptertension (Bauer et al. 1975, Plumer et al. 1976) and paraplegia (Aziz and Robertson 1973, Krishnamoorthy et al. 1976) have been noted in others with umbilical artery catheters. Umbilical vein catheters also form thrombus (Symansky and Fox 1972, Oski et al. 1963, • Department of Physiology, University of New England, Armidale, N.S.W. t Department of Anaesthesia and Intensive Care, Prince of Wales Hospital, Randwick, N.S.W. t Prince of Wales Children's Hospital, Randwick, N.S.W. Address for reprints: Dr. 1. F. Hecker, Department of Physiology, University of New England, Armidale, N.S.W. 2351, Australia.
Anaesthesia and Intensive Care, Vol. VII, No. 3, August, 1979
Scott 1965) and have caused portal hypertension (Oski et al. 1963), hepatic infarction (Riggs et al. 1976, Scott 1965), and inferior vena cava obstruction (Tooley 1972) in babies. In nearly all of these references, the Argyle umbilical catheter was used and we suspect that this catheter is used world wide for most umbilical vessel catheterizations. At the Prince of Wales Children's Hospital, a series of autopsies was performed on babies who died after periods of respiratory distress and who had umbilical artery catheters inserted. Although death was usually due to the respiratory distress syndrome, the autopsies regularly showed thrombus surrounding catheters and some infarcted vessels. This thrombus appeared identical to that seen on catheters in experimental sheep and we were prompted to use sheep to test Argyle umbilical catheters in comparison with other catheters used in humans and animals (Hecker et al. 1976). That experiment indicated that the Argyle and the Pharma-Plast umbilical artery catheters were comparatively thrombogenic. Fetal and newborn lambs are used extensively as models for human disease (Hecker 1974). In this paper, we present the results of an
240
J. F. HECKER ET A L.
experiment in which we have used the newborn lamb as a model for testing thrombus formation subsequent to insertion of commercially available catheters.
MATERIALS AND METHODS
Merino lambs were separated from their mothers within 24 hours of birth in October, 1976, November, 1977, and September, 1978. Anaesthesia was induced with pentobarbital sodium and maintained when necessary with Ketamine. In 1976 it was intended that tubing would be inserted directly into an umbilical artery of each lamb but after several unsuccessful attempts to catheterize umbilical arteries this approach was abandoned and instead tubing was inserted into the aorta by cutdown into a femoral artery. In babies most complications of umbilical artery catheters are due to the intra-aortic portion and the only reported complications that potentially cannot be reproduced in this lamb model are those involving the leg. In 1976, tubing was also inserted into the umbilical vein of each lamb but for reasons given below this route was not used in later years and instead tubing was inserted by cutdown into the femoral vein on the side opposite to the catheterized femoral artery. The following tubings were used: Argyle umbilical catheter (Sherwood Medical Industries Inc., St. Louis, Mo., U.S.A.) , 5 F.G., Argyle umbilical catheter, 5 F.G., treated with TDMAC and heparin (Hecker et al. 1976), Pharma-Plast umbilical catheter (PharmaPlast (Australia) Pty. Ltd., Sydney, N.S.W., Australia), 5 F.G.*, Portex umbilical catheter (Portex Ltd., Hythe, Kent, England), 6 F.G., Dural Plastics polyvinyIchloride tubing (Dural Plastics Ltd., Dural, N.S.W., Australia), 1.5 mm outside diameter, Dow Coming radio-opaque silicone rubber tubing (Dow Coming Corp., Hemlock, Mi., U.S.A.), 1.5 mm outside diameter, Pharmaseal K 15 umbilical catheter (Pharmaseal Laboratories, Glendale, Calif., U.S.A. 91201), 5 F.G., Angiodan umbilical catheter (Surgimed Pty. Ltd., Totfield House, 13-15 Thompson St., Frankston, Vic. 3199). The Argyle TDMAC catheters were not tested in femoral veins as insufficient numbers
were available. Catheters were filled and inserted to a distance of 17 to 22 cm. This placed the tips in the thoracic cavity. On insertion, the exterior end of each catheter was closed and left under the skin to avoid infection. Benzathine penicillin was given at the time of surgery by intramuscular injection. No attempts were made to flush catheters. After surgery, lambs were fed reconstituted powdered milk three or four times each day in 1976 and 1977 but in 1978 were returned to their mothers. At six days after surgery, lambs were anaesthetized with pentobarbital sodium to which had been added 200 i.u. of heparin and were killed by exsanguination. Catheterized vessels were dissected and opened and thrombus associated with catheters or the vessel wall was photographed and weighed. Kidneys, liver, other abdominal organs, lungs and the contralateral iliac artery were inspected for signs of infarction. Data (mass of thrombus and area of renal infarcts) were subjected to analysis of variance after square root transformation to fulfil the assumptions of normality and homogeneity of variances underlying analysis of variance. The un transformed data was not normally distributed and had non-homogeneous variances whereas after square root transformation the distribution was normal (Chi square test) and variances were homogeneous (Bartlett's test) (Pollard 1977). Significant differences between types of tubing were identified with Duncan's multiple range test (Duncan 1955).
TABLE 1
Numhers of catheters with thrombus ill arteries alld veills of lambs The Chi square test for arterial + venous catheters was calculated without data for Argyle TDMAC. Tubing Portex Angiodan Dow Coming Dural Pharmaseal Argyle Pharma-Plast Argyle TOMe Total
X2
Arterial
Venous
5/12
4/8
3 ti ti
ti
7 11 10 9 57;96 17.58*
5 7 7 7 8 44/56 8.48 NS
Arterial venous
+
9/20 9 11 13 14 18 18 82/140 19.28i"
= not significant = p < 0.05 'i' = p < 0.01
NS
* • This catheter is no longer produced.
Anaesthesia a",/ Intemire Care, Vol. 1'11, No. 3, August, 1979
241
THROMBUS FORMATION ON CATHETERS
aorta but was found on some catheters to cover the entire length. The distal end of the external iliac artery was taken as a reference point and the mass of thrombus on the 15 cm length cephalad to this point was recorded. Results are shown in Table 2. The Pharma-Plast and Argyle catheters had significantly more thrombus than did other types.
RESULTS
Arterial catheters: Of 96 arterial catheters (12 of each type), 57 were associated with thrombus (Table 1). There were significant differences between types of catheter in the proportion with thrombus (Table 1). On catheters with thrombus, the thrombus was most common distal to the bifurcation of the
TABLE
2
Mass of thrombus on catheters inserted into the aorta via the femoral artery and the vena cava via the femoral vein of new born lambs. Results are for the 15 cm cephalad to the distal end of the external iliac artery for 12 lambs and to where the femoral vein enters the abdominal cavity for 8 lambs. Lambs were killed at 6 days. The analysis of variance was done afrer square root transformation. Significant differences between tubing are indicated by different letters. *p < 0.05, tp < 0.01, tp < 0.001. Catheter
Femoral artery (12 lambs) (mean ± S.E.)
Portex Angiodan Dow Coming Dural Pharmaseal Argyle TDMAC Argyle Pharma-Plast
27 32 50 20 22 44 85 100
ANALYSIS OF VARIANCE Source d.f. mean square Tubing 7 12.99 Vessel Residual 88 3.78
F 3.44t
± ± ± ± ± ± ± ±
Femoral vein (8 lambs) (mean ± S.E.)
14 mg a a 29 a 25 a 11 a 10 ab 16 b 22 b 27 d.f. 6
44 60 37 57 139
mean square F 2.35* 75.28
49
32.07
TABLE 3
Portex Angiodan Dow Coming Dural Pharmaseal Argyle Pharma-Plast Argyle TDMAC
Number with Renal Infarction
Area of Renal Infarction mm2 (mean ± S.E.)
Oa
0/12 4
69.7 0.2 1.5 41.1
I 1
6
Chi2
o o o
= 359;
Oa Oa
± 37.0b ± 0.2a ± 0.5a ± 28.8b
oa
F
16 mg a ab 36 a 19 ab 16 ab 68
168 ± 58 141 ± 33
Numbers of animals with renal infarction and surface areas of renal infarction in lambs in which catheters were inserted into the aorta The value of F was from analysis of variance which was done after square root transformation. Significant differences between catheters are indicated by different letters. 0 P < 0.(01) Catheters
± ± ± ± ±
= 4.24t
The only signs of infarction were in kidneys and these infarcts were seen mainly with two types of catheter (Table 3). Differences Anaesthesia and Intensive Care, Vol. VII, No. 3, August, 1979
b b
Femoral artery + vein (20 lambs) (mean ± S.E.) 37 ± 15 mga a 43 ± 21 a 45 ± 16 a 35 ± 10 a 69 ± 30 116 ± 21 118 ± 27
F 6 1 129
mean square 135.67 467.75 24.96
b
b
d.f. 5.44t 18.74;
between catheters in the number of animals with infarcts and in surface areas of infarcts were significant. Umbilical vein catheters: In only 10 of 63 lambs used in 1976 did the umbilical vein catheter pass directly through the ductus venosus and into the inferior vena cava. In seven others, catheters went into the portal vein and then into the intestinal venous trunk. The remaining 46 umbilical vein catheters looped or doubled back in the umbilical vein, ductus venosus or branch of the portal vein supplying the ventral lobe of the liver, apparently because the tips had first lodged in small branohes of the portal vein supplying the ventral lobe of the liver. Every umbilical vein catheter was associated with thrombus for part of its length. Most was in intrahepatic veins. The extrahepatic portion of the portal vein and the intestinal trunk were virtually free of thrombus and only in one lamb was thrombus found around a catheter in the vena cava. Because of the doubling back
J. F. HECKER ET AL.
242
of the majority of catheters, mass of thrombus on umbilical vein catheters is not presented. Femoral vein catheters: Results for individual catheters are shown in Tables 1 and 2. The Argyle and Pharma-Plast catheters had significantly more thrombus than did the Portex umbilical catheter or the Dow Coming tubing. Combined data: Results from the femoral artery and femoral vein ca~heters (less results for Argyle TDMAC) have been combined in Tables 1 and 2. Argyle and Pharma-Plast catheters had significantly more thrombus than did the other types but there were no significant differences between the Argyle and PharmaPlast catheters or between the other catheters. DISCUSSION
The Pharma-Plast catheter is no longer produced but it was included in this experiment because it has been used extensively in some Australian hospitals. Previously we reported phlebitis when this catheter was inserted into veins of adult sheep (Hecker et al. 1976). Both F.G. 5 and F.G. 6 sizes were used in the adult sheep experiment and a review of data subsequent to writing that paper showed that phlebitis had occurred only with the F.G. 6 catheters. In this experiment, only F.G. 5 catheters were used and no phlebitis was seen. We have seen only one reference to the use of the newborn lamb for umbilical catheterization. This reported that the tip of an umbilical vein catheter entered the substance of the liver (De Yore et al. 1976). From our results, it is obvious that vena caval catheterization via the umbilical vein should be done in lambs only with the aid of X-ray fluoroscopy. Difficulty in passing umbilical vein catheters through the ductus venosus of babies has been reported (Oski et al. 1963, Peck 1967, Rosen and Reich 1970, Symansky and Fox 1972). In 1976, we also cannulated some lambs which were killed after three days. Most catheters at this time were free from thrombus. The mean mass on tubings at three days was 11 mg while that on similar tubings at six days was 54 mg. It is likely that thrombus formation is a gradual process and that amounts would continue to increase if catheters were left in for longer periods. There was considerable variation in the, amounts of thrombus that formed. This considerable variation is present also with catheters in veins of adult sheep (Hecker et al. 1976), and is almost certainly present with catheters in humans. Because of such variation,
it is necessary to ~tudy reasonable numbers if differences in thrombogenicity are to be identified. Significant differences were identified in this experiment, the Argyle and Pharma-Plast catheters being the most thrombogenic in both arteries and veins. One difference between the present results and those of our previous experiment (Hecker et al. 1976) is that Pharma-Plast catheters were significantly more thrombogenic in veins of adult sheep than Argyle catheters. The reason for this difference is not apparent but it may be due to differences between batches of Pharma-Plast catheters. Silicone rubber was significantly less thrombogenic, both in this experiment and in our previous experiment, than the Argyle catheters. Boros et al. (1975), by "pullout" angiography, compared in babies a silicone rubber catheter with the Argyle catheter and obtained a similar result to ours. The silicone rubber used by Boros et al. (1975) and the silicone rubber used in these experiments were supplied by different companies and were relatively thromboresistant. However we believe that not all brands of silicone rubber are thromboresistant as oreliminary tests in veins of adult sheep on tubing used for human ventriculo-atrial shunts have shown it to be more thrombogenic. The finding that some tubings cause greater renal infarction than others is interesting. The infarction is due to thrombus which has dislodged from catheters and the presence of renal infarction with some types of catheters (Table 3) suggests that they are more thrombogenic than values in Table 2 would indicate. In experiments in which catheters were inserted into the aorta of adult sheep, we have also found that some plastics produce more renal infarcts than others. T'he TDMAC process was developed in the United States by Battelle Laboratories for incorporating a surface layer of heparin on plastics. We used some TDMAC treated Argyle catheters in a prospective blind trial in umbilical arteries of new born babies suffering from respiratory distress and obtained adequate data from 16 babies with TDMAC treated and 21 with untreated Argyle catheters. There was no significant difference in the incidence of complications such as blocked catheters. Single shot withdrawal angiograms in seven babies with treated and 10 with untreated catheters showed no definite evidence of thrombus. Four babies which received treated and four which Ana~sth~sia
and Tnte",i.'e Care. Vol. Vll, No. 3. AlIglIIt. 1979
THROMBUS FORMATION ON CATHETERS
received untreated catheters died and in each of these groups two babies had the catheter in situ at death while the catheter had been removed before death in two. Each of the eight babies had thrombus present in the internal iliac artery. Various in vitro tests have indicated that the TDMAC process increases thromboresistance (Bruck 1974). The above small experiment with babies, our previous experiment with adult sheep (Hecker et al. 1976) and this experiment suggest that the TDMAC treatment of Argyle catheters is of limited value. Though treatment of plastics with TDMAC may reduce 1'hrombogenicity, it would seem preferable to use an inherently less thrombogenic plastic than the polyvinylchloride from which the Argyle catheter is made. This may then be treated with TDMAC. Such plastics are available as the Portex catheter (made also from polyvinylchloride) did perform well in lambs and in veins of adult sheep (unpublished results for the Portex catheter). ACKNOWLEDGEMENTS
We wish to thank Sherwood Inds., Portex Ltd. and Dow Coming Corporation for supplying catheters and tubing, Battelle Laboratories for applying the TDMAC process and the National Health and Medical Research Council for financial support. The project would not have been possible but for the help of the several volunteer lamb feeders. REFERENCES
Aziz, E. M., and Robertson, A F. (1973): "Paraplegia: A complication of umbilical artery catheterization", I. Pediatr., 82, 105l. Bauer, S. B., Feldman, S. M., Gelliss, S. S., and Retik, A. B. (1975): "Neonatal hypertension: A complication of umbilical-artery catheterization", New En%!l. I. Med., 293, 1032. Boros, S. J., Thomoson, T. R .. Reynolds, J. W., Jarvis, C. W., and Williams, H. J. (1975): "Reduced thrombus formation with silicone elastomer (Silastic) umbilical artery catheters", Pediatrics, 56, 98l. Bruck, S. D. (1974): "Blood Compatible Synthetic Polymers: An Introduction", Thomas, Springfield, I11. Cochran, W. D., Davis, H. T., and Smith, C. A. (1968): "Advantages and complications of umbilical arterv catheterization in the newborn", Pediatrics, 42. 769. De Vore, J. S .. Wheeler, A. S., and Latyshevsky, H. (1976): "Misplacement of an umbilical vein catheter". Anesthesiology. 44, 451. Duncan, D. B. (1955): "Multiple range and multiple F tests", Biometrics, 11, 1. Egan, E. A, and Eitzman. D. V. (1971): "Umbilical vessel catheterization", Am. I. Dis. Child, 121, 213. Ford. K. T., Teplick, S. K., and Cl ark, R. E. (1974): "Renal artery embolism causing neonatal hyperAnaesthesia and Intensive Care. Vol. VIl. No. 3, August, 1979
243
tension: A complication of umbilical artery catheterization", Radiology, 113, 169. Goetzman, B. W., Stadalnik, R. c., Bogren, H. G., Blankenship, W. J. Ideda, R. M., and Thayer, J. (1975): "Thrombotic complications of umbilical artery catheters; A clinical and radiographic study", Pediatrics, 56, 374. Gupta, J. M., Robertson, N. R. C., and Wigglesworth, J. S. (1968): "Umbilical artery catheterization in the new born", Arch. Dis. Child, 43, 382. Hecker, J. F. (1974): "Experimental Surgery on Small Ruminants", Butterworths Ltd., London. Hecker, J. F., Fisk, G. C., and Farrell, P. C. (1976): "Measurement of thrombus formation on intravascular catheters", Anaesth. In tens Care, 4, 225. Krishnamoorthy, K. S., Fernandez, R. J., Todres, I. D., and DeLong, G. R. (1976): "Paraplegia associated with umbilical artery catheterization in the newborn", Pediatrics, 58, 443. Marsh, J. L., King, W., Barrett, C., and Fonkalsrud, E. W. (1975): "Serious complications after umbilical artery catheterization for neonatal monitoring", Arch. Surg., 110, 1203. Neal, W. A, Reynolds. J. W., Jarvis, C. W., and Williams, H. J. (1972): "Umbilical artery catheterization: Demonstration of arterial thrombus by aortography", Pediatrics, 50, 6. Olinskv, A, Aitken, F. G., and Isdale, J. M. (1975): "Thrombus formation after umbilical arterial catheterization", South Afr. Med. I., 49, 1467. Oski, F. A. Allen, D. M., and Diamond, L. K. (1963): "Portal hypertension - a complication of umbilical vein catheterization", Pediatrics, 31, 297. Peck, D. R., and Lowman, R. M. (1967): "Roentgen aspects of umbilical vascular catheterization in the newborn", Radiology, 89, 874. Plumer. L. B .. Kaplan, G. W., and Mendoza, S. A. (1976): "Hypertension in infants - a complication of umbilical arterial catheterization", I. Pediatr., 89, 802. Pollard, J. H. (1977): "A Handbook of Numerical and Statistical Techniques with Examples Mainly from the Life Sciences", Cambridge University Pre~s, Cambridge. Riggs. T .. Hirschfeld, S., Borkat, G., and Liebman, J. (1976): "Inferior vena cava obstruction secondary to indwelling venous catheters: Two cases", Pediatrics, 58. 446. Rosen, M. S., and Reich, S. B. (1970): "Umbilical venous catheterization in the newborn: Identification of correct positioning", Radiology, 95, 335. Scott. J. M. (1965): "Iatrogenic lesions in babies following umbilical vein catheterization", Arch. Dis. Child, 40. 426. Strauss, A W., Escobedo, M., and Goldring, D. (1974): "Continuous monitoring of arterial oxygen tension in the newborn infant", I. Pediatr., 85, 254. Symansky, M. R., and Fox, H. A (1972): "Umbilical vessel catheterization: Indications, management, and evaluation of the technique", I. Pediatr., 80, 820. Tooley, W. H. (1972): "What is the risk of an umbilical artery catheter?", Pediatrics, 50, 1. Tyson. J. E., deSa, D. J., and Moore, S. (1976): "Thromboatheromatous complications of umbilical arterial catheterization in the newborn period: CHnico-pathological study", Arch. Dis. Child, 51, 744. Wigger, H. J .. Bransilver, B. R., and Blanc, W. A. (1970): "Thromboses due to catheterization in infants and children", I. Pediatr., 76, 1.
