Calcium phosphate crystal occlusion of central venous catheters used for total parenteral nutrition in infants and children: Prevention and treatment

Calcium P h o s p h a t e Crystal Occlusion of Central Venous Catheters Used for Total Parenteral Nutrition in Infants and Children: Prevention and Treatment By Charles W. Breaux, Jr, Debra Duke, Keith E. Georgeson, and Jose R. Mestre Birmingham, A l a b a m a 9 Calcium phosphate crystal occlusion is a complication occasionally encountered with long-term indwelling Silastic central venous catheters used for total parenteral nutrition (TPN) in-infants and children. These occluded catheters are usually treated by removal. We have successfully treated six patients who experienced seven episodes of calcium phosphate crystal central venous c a t h e t e r occlusion by irrigating their catheters w i t h a hydrochloric (HCl) acid heparin solution. Although temporary febrile reactions occurred in three cases (42%), no serious complications were encountered. An average of 46 catheter-days per patient episode were preserved. By paying close attention t o t h e calcium and phosphate concentrations in a patient's TPN solution, the clinician can minimize the risk of calcium phosphate precipitation. If central venous cathet e r occlusion does occur due to precipitation of calcium phosphate crystals, then HCI-heparin irrigation is a safe and effective method for salvaging such catheters. 9 19B7 by Grune & Stratton, Inc.

INDEX WORDS: Central venous catheter, complications; total parenteral nutrition (TPN}, complications; calcium phosphate crystal precipitation.

NFANTS AND CHILDREN who are unable to

tolerate adequate enteral feedings frequently Irequire total parenteral nutrition (TPN) administered

through a long-term indwelling Silastic central venous catheter. Occasionally, these catheters become occluded, most often by thrombosis but sometimes by calcium phosphate crystal precipitation. ~'4 Thrombolytic therapy has been used successfully to salvage thrombosed catheters in pediatric patients, s'6 but premature removal of the catheter is usually employed for those occluded by calcium phosphate crystals. 2"3 Salvage of these catheters is preferable to premature removal and replacement. In this paper, we review our experience with six patients successfully treated by catheter irrigation with a hydrochloric (HCI) acid heparin solution for seven separate episodes of calcium phosphate crystal catheter occlusion between June 1985 and January 1986. In addition, factors affecting calcium phosphate precipitation are discussed, and methods for preventing calcium phosphate precipitation are presented. MATERIALS A N D METHODS Our six patients included three with short bowel syndrome following massive small bowel resection, two for necrotizing enterocolitis, and one for midgut volvulus. Two patients had failureto-thrive; intestinal lymphangiectasia was the cause in one patient Journal of Pediatric Surgery, Vo122, No 9 (September), 1987: pp 829-832

and intrahepatic biliary hypoplasia in the other. The remaining patient had intestinal pseudoobstruction following gastroschisis repair. The seven Silastic central venous lines involved were Broviac catheters (Davol Evermed, Kirkland, WA); two were the 3 Fr neonatal-size, and five were the 4 Fr infant-size (internal diameters 0.5 and 0.7 mm, respectively). All catheters were placed under sterile conditions in the operating room with fluoroscopic confirmation of their positions. The saphenous vein in the groin was used for venous access in three cases, the common facial vein in two, and the middle thyroid and external jugular veins in one each. The catheters were tunneled subcutaneously to emerge through the skin of the anterior chest wall. The mean patient age at time of insertion was 9.8 months (range 6 weeks to 20 months). In every episode, catheter occlusion occurred during infusion of the TPN solution and was first manifested by resistance to infusion through the catheter. At the time of diagnosis, the mean patient age was 11.1 months (range 9 weeks to 22 months). The average prior indwelling time of their catheters was 38 days (range 8 to 112 days). The characteristics of the TPN involved in each episode are presented in Table 1. All patients were receiving single-bottle TPN solutions in which the calcium and phosphate were mixed in the same bottle. The calcium used in the TPN solution was the gluconate salt, and the phosphate was either the potassium or sodium salt. Addition of the phosphate salt preceded that of the calcium gluco9hate during TPN solution preparation. Lipid emulsion was either placed in the same bottle as the TPN solution or piggybacked into the central venous line during TPN infusion. All catheter occlusions were associated with a recent increase in the calcium or phosphate dose. Following each episode, the orders were rewritten to decrease the calcium or phosphate dose, or both. Catheter phlebography was performed on the first two patients of this series. In the first patient, only a trickle of contrast could be forced through the catheter with great pressure. Total resistance to injection of contrast with only partial opacification of the catheter was noted in the second patient. In two episodes, an attempt was first made to open the catheter with urokinase using the protocol of Curnow et al.6 These attempts were unsuccessful in both instances. In the other episodes, the cause of the occlusion was diagnosed by the circumstances surrounding the event and the characteristics of the TPN solution involved. Urokinase was not used, and our HCl-heparin protocol was instituted without delay. The HCl-heparin solution was prepared by mixing 0.1N HCI acid

From the Departments of Surgery and Gastroenterology, The Children's Hospital of Alabama, and the Departments of Surgery and Pediatrics, The University of Alabama School of Medicine, Birmingham. Address reprint requests to Keith E. Georgeson, MD, Department of Surgery, The Children's Hospital of Alabama, 1600 7th Ave S, Birmingham, AL 35233. 9 1987 by Grune & Stratton, Inc. 0022-3468/87/2209-0012503.00/0 829

830

BREAUX ET AL

Table 1. TPN Characteristics Episodes 1 Setting

Type Flow rate (mL/h) Dextrose (% or g/dLI Amino acids Brand (% or g/dLI Calcium (mEq/L) Phosphate (retool/L)

2

3

4

5

6

7

Hospital Hospital Home Home Home Hospital Hospital Cycled Constant Cycled Cycled Cycled Cycled Cycled 35 4 75 38 36 60 21 23

25

20

20

25

25

30

T 2.0 8.0 16.6

F 2.5 12.0 8.0

F 2.4 9.8 19.5

F 3.9 18.4 23.8

F 3.9 12.8 25.8

F 4.1 13.5 17.3

F 2.9 12.6 17.3

Abbreviations: T, TrophAmine; F, FreAmine III.

(i mL/kg) with 10 mL heparin (10 U/mL). Usually, the catheter was so occludedthat an appreciableamountof this solutioncouldnot be infused. For this reason, 0.2 mL of the HCl-heparin solution was drawn up in a tuberculin syringe and used to irrigate the catheter with a gentle to-and-fromotion for two mintues. The catheter was then capped and the process repeated hourlyuntil patency was restored as determinedby the ability to aspirate blood freely. Once the catheter was open, either the TPN infusionwas reestablishedor a heparin lockwas placed. RESULTS

All seven episodes of catheter occlusion were successfully treated with the HCl-heparin therapy. Five catheters opened on the initial irrigation, one on the second, and one on the third. The duration of catheter occlusion averaged 29.6 hours (range 5.3 to 91.3 hours). Febrile reactions occurred in three episodes (42%), beginning 0.2, 5, and 5.8 hours after reestablishment of catheter patency and persisting for 8, 24, and 36 hours. Blood and urine cultures were negative in all three instances. Two patients received parenteral methicillin and gentamiein empirically for three days until documentation of sterile cultures; the other patient received no specific treatment. No other complications were noted. At the time of this writing, 321 total catheter days have been preserved, averaging 46 catheter days per patient episode (range 11 to 126 catheter days). Only one catheter remains in place. Five lines were removed, four for sepsis and one for thrombosis resistant to urokinase therapy. One catheter was accidentally pulled out by an infant on home T P N during a dressing change. DISCUSSION

The distinction between thrombosis or calcium phosphate crystal precipitation as the cause of central venous catheter occlusion is usually easy to make clinically. Thrombosis generally happens over a period of days and is heralded by progressive difficulty in aspirating blood through the catheter. Calcium phos-

phate crystal occlusion usually occurs in a single day over one infusion cycle of TPN, is manifested by progressive resistance to infusion, and is associated with T P N solutions containing incompatible calcium and phosphate concentrations, often recently increased. If catheter phlebography is performed, it reveals pathology at the tip of the catheter in the case of thrombosis (either a fibrin sleeve or a mural thrombus), whereas with calcium phosphate crystal occlusion, the obstruction occurs more proximally. Early in our experience, we tried to inject half of the prepared HCl-heparin solution into the catheter. However, we have found that most catheters occluded by calcium phosphate crystals are so plugged that it is not possible to infuse this much solution. We now use a gentle to-and-fro irrigation of the catheter with 0.2 mL of the solution in a tuberculin syringe. This procedure has worked well, and in addition, has limited the amount of HCl-heparin solution received by the patient once the catheter is opened. The occurrence of a febrile reaction may relate to the volume of solution infused as two of the three fevers seen in our patients occurred in those patients receiving larger amounts of HCl-heparin solution (4 and 17 mL v 0.6 mL). The solubility or precipitation of calcium phosphate in T P N solutions is determined by the following factors: (1) Calcium and phosphate concentrations2'4'7"2: Higher concentrations of these two components in a given T P N solution increase the likelihood of precipitation. (2) pH2"l~ The solubility of calcium phosphate is pH-dependent. At lower pHs, the soluble monobasic form predominates while at higher pHs, the insoluble dibasic one is the preponderate form. (3) Salt form of calcium9: Calcium chloride dissociates to a greater degree in solution than does calcium gluconate. The concentration of free calcium ion available for precipitation is greater with the chloride than the gluconate salt. (4) Order of mixingT'8: Addition to the T P N solution of the phosphate before the calcium salt increases the concentrations of the two additives that can be compatibly admixed. (5) Amino acid composition and concentration 2"1~ Different brands of amino acid solutions have different pHs. In addition, increasing the concentration of amino acids in a T P N solution decreases the pH, allowing higher concentrations of calcium and phosphate to be added. (6) Dextrose concentrations~~ The higher the concentration ofdextrose, the lower the pH of the T P N solution and consequently the higher the solubility of calcium phosphate.

CALCIUM PHOSPHATE OCCLUSION OF TPN CATHETERS

831

X

\

20

v

E E 9-I u r

\

4

15 "".

6

5

10

5

110

1'5

20

2'5

Phosphate (mmollL) Fig 1. Calcium and phosphate concentrations of the TPN solutions involved in the episodes of central venous catheter occlusion with superimposed calcium phosphate precipitation curves adapted from Eggert et al. I~ Concentrations to the right of the curves result in calcium phosphate precipitation. The solid line represents TPN solutions with FreAmine III 2% and dextrose 20% (pH 6.3) and the dashed line FreAmine III 4% and dextrose 25% (pH 6.4}.

(7) Lipid administrationt~ Simultaneous administration of lipid emulsion through the central venous line increases the pH of the TPN solution and thus the risk of calcium phosphate precipitation. (8) Temperature3'9't~ The degree of calcium salt dissociation is greater at higher temperatures providing more free calcium ion to complex with phosphate. (9) Time after preparationt~ The longer the mixed TPN solution stands, the more likely calcium phosphate precipitation is to occur. (10) Permeability of catheter and infusion rater: The Silastic wall of the central venous catheter is permeable to gases. At slow infusion rates, water vapor may pass through the catheter wall and be replaced by air causing supersaturation of calcium phosphate. Precipitation of the calcium phosphate in the central venous catheter rather than in the TPN solution bottle is especially influenced by the piggybacking of lipid emulsion into the line, the higher body temperature to

which the line is subject relative to ambient temperature, the permeability of the catheter, and the infusion rate. Ideally, calcium phosphate crystal occlusion of central venous catheters should not occur. Various aids are available to assist the clinician in writing TPN orders or the pharmacist .in checking them to avoid calcium phosphate precipitation: (I) Precipitation curves for calcium phosphate have been generated for TPN solutions containing different amino acid solutions, eg, by Eggert et al to for Aminosyn and FreAmine III (American McGaw, Irvine, CA), by Poole et al tl for Aminosyn, and by Fitzgerald and MacKay t2 for TrophAmine. After plotting the calcium and phosphate concentrations of the TPN solutions of our patients on the appropriate graphs of Eggert et al, t~ all but one fall within the zone of precipitation or on the borderline (Fig 1). (2) The calcium times phosphate (Ca x PO4) solubility product has been recommended by Knight et al 4 as a useful clinical aid to avoid precipitation. They suggest that TPN solutions with Ca x PO4 products >75 mmol2/L2 be avoided for central TPN (and > 100 mmol/L2 for peripheral TPN). Our Ca x PO4 products were 66, 48, 96, 219, 165, 117, and 109 mmol/L2 for episodes one to seven, respectively. (3) Alexander and Arena 13have developed a pocket computer program for predicting calcium phosphate precipitation, based on an equation describing the precipitation curve that they generated. The clinician should write TPN orders carefully to minimize the risk of calcium phosphate precipitation. Multiple factors affect calcium phosphate solubility, and slight changes may promote precipitation in solutions with marginally compatible concentrations of calcium and phosphate. Central venous catheters can occlude in such situations. We have presented a method of salvaging calcium phosphate crystal occluded catheters, avoiding their premature removal. ACKNOWLEDGMENT

We are grateful to Isaac Wornom, MD, Charles Bruce, RPh, Kelly Rooker, RPh, and Jackson Como, PharmD for their assistance.

REFERENCES 1. Pomerance HH, Rader RE: Crystal formation: A new complication of total parenteral nutrition. Pediatrics 52:864-866, 1973 2. Knight PJ, Buchanan S, Clatworthy W Jr: Calcium and phosphate requirements of preterm infants who require prolonged hyperalimentation. JAMA 243:1244-1246, 1980 3. Robinson LA, Wright BT: Central venous catheter occlusion caused by body-heat-mediated calcium phosphate precipitation. Am J Hosp Pharm 39:120-121, 1982 4. Knight P, Heer D, Abdenour G: CaxP and Ca/P in the parenteral feeding of preterm infants. JPEN 7:110-114, 1983

5. Winthrop AL, Wesson DE: Urokinase in the treatment of occluded central venous catheters in children. J Pediatr Surg 19:536538, 1984 6. Curnow A, Idowu J, Behrens E, et al: Urokinase therapy for Silastic catheter-induced intravascular thrombi in infants and children. Arch Surg 120:i237-1240, 1985 7. Kobayashi NH, King JC: Compatibility of common additives in protein hydrolysate]dextrose solutions. Am J Hosp Pharm 34:589-594, 1977 8. Schuetz DH, King JC: Compatibility and stability of electro

832

lytes, vitamins and antibiotics in combination with 8% amino acids solution. Am J Hosp Pharm 35:33-44, 1978 9. Henry RS, Jurgens RW Jr, Sturgeon R, et al: Compatibility of calcium chloride and calcium gluconate with sodium phosphate in a mixed TPN solution. Am J Hosp Pharm 37:673-674, 1980 10. Eggert LD, Rusho WJ, MacKay MW, et al: Calcium and phosphorus compatibility in parenteral nutrition solutions for neonates. Am J Hosp Pharm 39:49-53, 1982

BREAUX ET AL

11. Poole RL, Rupp CA, Kerner JA Jr: Calcium and phosphorus in neonatal parenteral nutrition solutions. JPEN 7:358-360, 1983 12. Fitzgerald KA, MacKay MW: Calcium and phosphate solubility in neonatal parenteral nutrient solutions containing TrophAminc. Am J Hosp Pharm 43:88-93, 1986 13. Alexander SR, Arena R: Predicting calcium phosphate precipitation in premature infant parenteral nutrition solutions. Hosp Pharm 20:656-658, 1985