Acta Med Scand 207: 455-460, 1980

Hemoperfusion with Amberlite Resin in the Treatment of Self-Poisoning

Andrew Heath, Krister Delin, Elisabeth Eden, Eric Mhrtensson, Dag Selander, Ingemar Wickstrom and Jar1 Ahlmen

From the Department of Anesthesiology and intensive Cure and the Deportment of Nephrology, Sahlgrensku Hospitul, und Psychiatric Department ill, Lillhagens Hospital,

University of Gothenbrirg , Cotlienburg. Sweden

ABSTRACT. Ten patients with various intoxications were treated with resin hemoperfusion. Three of four patients with grade 1V coma due to tricyclic antide- pressant (TCA) poisoning could be extubated during or on termination of hemoperfusion. Clearance val- ues of 135-185 and 190-200 ml/min were obtained for amitriptyline and nortriptyline, respectively. One pa- tient with severe chloral hydrate poisoning could be extubated after less than one hour's hemoperfusion. A clearance of 140 ml/min was obtained in a uremic patient with AV block I1 due to digitoxin intoxication. Four patients with mushroom poisoning were treat- ed with combined hemoperfusion-hemodialysis. A transient fall in platelet count was seen in all patients. Resin hemoperfusion is of definite value in selected, severe cases of self-poisoning with psychotropic drugs such as TCA and possibly in cases of mushroom poisoning where the prognosis with hemodialysis and supportive therapy is doubtful.

Key words: intoxication. hemoperfusion, resin. Acta Med Scand 207: 455, 1980.

Self-poisoning is a common cause of admission to hospital, especially in large cities. Many of these patients require intensive care. In a select group of severe cases it may be advantageous to remove the drug as quickly as possible; for example, if drug- induced cardiac arrhythmias are life-threatening, if there is a risk of permanent kidney or brain damage or if prolonged coma indicates tracheostomy .

Forced diuresis, peritoneal dialysis and hemo- dialysis are of limited help in most instances of self-poisoning with psychotropic drugs such as tricyclic antidepressants (TCA) (2, 13. 17). Hemo- perfusion with a charcoal filter, first introduced in 1964 (22), is more effective than hemodialysis and has been proven effective in several poisonings,

including barbiturate, glutethimide. methaqualone, salicylate, paracetamol (8) and digitalis poisoning (4, 18), but reports differ as to its value in TCA poisoning (3 , 7).

The Amberlite XAD-4 non-ionic resin was first introduced by Rosenbaum et al. (14) in 1970 as an alternative to the charcoal filter. It has been proven effective in a number of cases of self-poisoning but has been little used in TCA or mushroom poisoning.

METHODS AND MATERIAL The hemoperfusion system consisted of a 650 g Amber- lite" resin column in a Lexan cartridge (Extracorporeal. USA).

Standard blood tubings (Gambro, Sweden) employed in hemodialysis were used to connect the components in the circuit. The treatment was controlled with a blood monitor device (Gambro) for measurement of blood pres- sures on the inflow and outflow sides of the cartridge. blood flow and inhibition of air embolization by an air detector (Fig. I ) .

The system was perfused with 3-4 I heparinized iso- tonic saline, 2500 IU heparin/l. The priming volume for the system is approximately 550 mi. During perfusion, parenteral heparin was given continuously at a rate of 1500-2000 IU/hour.

Both femoral veins were catheterized percutaneously. using a Cobe catheter with a tapered end and an inner dia- meter of 0.80 mm. One catheter was used as the inflow (arterial) line to the cartridge, and the other as the outflow (venous) line. In two patients, however, Scribner shunts were used for access to blood. After catheterization, 5 OOO IU heparin were given. Perfusion was started about 10 min later at a flow of 150-300 ml/min, and continued until a satisfactory clinical response had been obtained or up to 4 hours.

Before starting and after perfusion, hemoglobin. hematocrit and platelet counts were estimated as well as serum electrolyte, total protein, calcium, glucose. creatinine. bilirubin, alkaline phosphatase, aspartate trans- aminase and lactate dehydrogenase concentrations. Dur-

456 A . Henth et NI.

HEPARIN

P

T -+OUTFLOW I

E N

+-INFLOW T

WOOL FILTER

Fig. I . The hemoperfusion circuit.

ing hemoperfusion, hemoglobin, platelet count and acti- vated partial thromboplastin time were measured fre- quently. Plasma concentrations of TCA were analyzed before, during and after perfusion by a gas chromato- graphic method (12) in the TCA-intoxicated patients. From two of the four TCA cases, samples were taken on each side of the cartridge so that clearance could be calculated. Plasma concentrations of digitoxin (case 6) were measured using a radioactive immunoassay technique (Diagnostic Products Corporation, USA).

Criteria for hemoperfusion in cases of drug-induced self-poisoning were either respiratory depression or se- vere arrhythmias, such as rnultifocal ventricular extrasys- tole or ventricular fibrillation. Gastric aspiration was car- ried out in all patients, irrespective of the time of inges- tion. after ensuring a patent airway. When necessary, as- sisted ventilation was given as well as sodium bicarbonate if acidosis was suspected.

Hemoperfusion was combined with hemodialysis in four cases of mushroom poisoning, where the mushroom taken was suspected of being highly toxic. In two of these cases, Cortinarius speciosissirnus had been identified be- fore admission to this hospital.

A neurological assessment was made of the patient's level of consciousness according to Arieff and Friedman (1 ) . reflex activity, and pupil response, both before and during hemoperfusion. Likewise, the pulse rate, blood pressure, respiration rate and response to voice and pain were registered.

Ten patients (7 males and 3 females), with a mean age of 37 years (range 23471, received a total of 14 hemoperfu- sion treatments.

RESULTS

Ingestion of TCA was the reason for hemoperfusion in 4 patients, chloral hydrate and digitoxin in one patient each and ingestion of mushroom in 4 pa- tients.

The clinical response of each patient is sum- marized in Table I together with relevant changes in blood chemistry. Three TCA patients were extu-

bated during or on termination of hemoperfusion. One patient (no. 5 ) . with chlorimiprarnine poison- ing. who was hemoperfused for only one hour could be extubated after a further 12 hours.

Plasma drug concentrations of amitriptyline, nor- triptyline (a metabolite of amitriptyline) and digi- toxin are given in Table I I . Clearance at a blood flow rate of 200 mllmin was 135-185 mI/min for amitriptyline and 190-200 ml/min for nortriptyline. Extraction ratios were 0.65-0.93 for amitriptyline and 0.95-1 .00 for nortriptyline.

A patient with severe chloral hydrate intoxication required repeated cardioversion for VF and could be extubated after 50 min of hemoperfusion. This patient had grade 1V coma at the start of hemoper- fusion and was awake four hours later when the hemoperfusion was terminated.

A patient with uremia (no. 6) was scheduled for a kidney transplant, but showed an AV block 11 and had a PQ interval of 0.30 sec prior to operation. caused by an overdose of digitoxin. He was treated with combined hemoperfusion-hemodialysis. Digi- toxin clearance was 140 ml/min at a blood flow rate of 240 ml/min. This patient had a normal ECG with a PQ time of 0.20 sec on the day after dialysis-perfusion, and received a kidney transplant as planned. In this case we had problems with coagulation in the hernoperfusion filter and the hemoperfusion was discontinued after 35 min. This may explain why there was only a small reduction of the initial digitoxin concentration, from 0.024 pmol/l before treatment to 0.020 pmoI/l on the day after hemodialysis-hemoperfusion.

In 4 of the 14 hemoperfusions, variations in blood flow during perfusion and coagulation in the filter ruled out a reliable estimate of the amount of drug removed.

Four patients were treated for mushroom poi- soning. At follow-up, three have normal serum creatinine values, while the final prognosis concern- ing kidney function of the fourth patient is still uncertain. After mushroom ingestion, one patient was first treated with hemoperfusion and on the following day with hemodialysis because of clinical- ly suspected ingestion of Amanita phalloides: analysis of the stomach content revealed later that this patient had eaten the non-toxic Macrolepiota rhacodes. Two patients each who had ingested Cor- tinarius speciosissimus received three hemoperfu- sions and hernodialyses in which the perfusion car- tridge was serially coupled to a dialyser (area 1.0

Hemoperfirsion treutment qf'into.rictitions 457

Table I . Dotcrils of'potionts r i n d resporise to treritnient

Pat. Age no. ( y . )

I 44

z 34 3 23 4 31

5 33 6 44 7 24

8 47

9 44 10 41

Interval be- tween start of

Sex Poisoning Condition perfusion and extubation

d Chloral hydrate Grade IV coma, repeated 3 h 40 rnin 50 min

d Amitriptyline Grade IV coma 1 h30min 1 h P Amitriptyline Grade IV coma 35 min I h 6 Trimipramine Grade IV coma, sinus 4 h 13 min 4 h I8 rnin

tachycardia d Chlorimipramine propiomazine Grade IV coma 60 rnin I2h 6 Digitoxin Uremia, AV block 11, awake 35 min -

Duration of perfusion

ventricular fibrillation

6 Cortinarius speciosissimus Nausea, single vomiting, 4 h + 4 h - abdominal pain, awake +4 h"

awake +4 h" P Cortinarius speciosissimus Nausea, abdominal pain, 4 h + 4 h -

P Macrolepiota rhacodes Diarrhea, awake 40 min" - d Unspecified mushroom Awake 3 h -

" Combined with hemodialysis on each occasion. '' Hemodialysis on the following day.

m"). One patient with acute anuria clinically sug- gesting intoxication with Cortinarius species was also treated with a serially coupled perfusion car- tridge and a dialyser. There were no significant changes in the serum concentrations of electrolytes, calcium, glucose, bilirubin, protein, alkaline phos- phatase. ALAT or ASAT before and efter hemo- perfusion in any patient. No hemodynamic changes occurred during hemoperfusion. A moderate fall in total platelet count was seen in all patients, as well as a fall in hemoglobin concentration. The fall in platelet count was pronounced in cases with clot- ting complications in the filter. One patient devel- oped a minor hemorrhage at the site of catheteriza- tion in the left groin. This complication was related to heparin administration and was controlled by pro- tamin. All patients made a full recovery. Four of the

14 hemoperfusion treatments given were discon- tinued because of coagulation in the filter.

DISCUSSION

Since 1964, both charcoal and resin columns have been shown to be effective in the treatment of in toxications with many drugs, including barbiturates, glutethimide, paracetamol, methaqualone, salicyl- ate and diazepam ( 5 . 18). Trafford et al. (20) used resin hemoperfusion in one case each of chlorimi- pramine and amitriptyline poisoning. Their clear- ance rates for amitriptyline and nortriptyline agree with the present data (Table 111) and with those recently published by others (3). Our extraction ratios for amitriptyline (0 ,65493) agree with that found by Topf et al. (19), who also used a resin

Table I I . Plasrnn clrirg concentrations, clerironce wlires and extrnction ratios

Max. plasma Inflow outflow Extrac- Pat. conc. conc. conc. Blood flow Clearance tion no. Drug (pmolll) (pmol/l) (pmolll) (ml/min) (ml/min) ratio

2 Amitriptyline 4.4 2.3 0.8 200 135 0.65

3 Amitriptyline 4.24 2.29 0. I6 200 I85 0.93

6 Digitoxin" 0.024 0.018 0.0075 240 I40 0.58

" Together with hemodialysis.

Nortriptyline 0.4 0.0 200 200 I .oo

Nortriptyline - 2.37 0.11 200 190 0.95

458 A . Heath et a/.

Table 111. Si imrnmy qfpiiblished clewtince mid other cltitci oti hetizoper:fiisioii of' TCA poisoiii i ig

R = resin, C = charcoal

Drug Duration of Flow Clearance Extraction hemoperfusion (mllmin) (mllmin 1 ratio Reference

C hlorimipramine Chlorimipramine Trimipramine Arnitriptyline Nortriptyline Amitriptyline Nortript yline Amitriptyline Nortripty line Amitriptyline Amitript yline Nortriptyline

3 h R I h R 4 h 13 min R 1 h 30 min R

35 min R

2 h 30 min R

4 h R 4 h C 3 h C

300 200 200 200

200

300

200 200 200

- 185 190 I35 200 210 280 I77 100-1 12.50 14-58

- 0.93 0.95 0.65 1 .OO

- 0.83-0.96 0,504.56 0.07-0.29

Trafford et al. (20) Present report Present report Present report

Present report

Trafford et al. (20)

Topf et al. (191 Diaz-Buzo et al. ( 3 ) lversen et al. ( 7 )

column. It appears that nortriptyline is more easily cleared with resin hemoperfusion than amitriptyline and that nortriptyline is more easily cleared with resin hemoperfusion than charcoal hemoperfusion with reported clearance rates as low as 1 4 % ml/ min at a flow of 200 ml/min. With resin hemoperfu- sion an effective clearance of TCA is correlated to a change in the patient's clinical condition, even after shorter periods of perfusion. This may suggest that these drugs move quickly between plasma and the body's extravasal compartments, and are not avidly bound to extravasal tissues (3). This may in turn explain the slight increase in plasma drug concen- trations following termination of hemoperfusion in case 2 (Fig. 2).

When calculating our clearance values, however, a small error was not accounted for: the blood flow was measured only from the thoroughly calibrated blood pump. However, the calculated extraction factor is independent of the blood flow rate at the moment when the blood tests are performed. Thus, this factor is an expression of the efficiency of the filter.

Polyresin hemoperfusion is also an effective method of removing chloral hydrate-which, as far as we are aware, has not been reported before (pa- tient I ) . The protein binding of about 40% of the rapidly formed trichloroethanol might indicate that hemoperfusion is more effective than hemodialysis ( 16) in severe chloral hydrate poisonings.

At present there are two active ways of treat- ing digitalis intoxication. One is the antibody tech- nique described by Smith et al. (IS), the other is hernoperfusion. Charcoal hemoperfusion is effec-

tive in the treatment of digitalis intoxication. Tobin and Mookerjee (18) have observed digoxin clear- ances of 60-100 ml/min in four patients, i.e. 5-10 times higher than those reported for hemodialysis.

4wv,,v.3-2 1 COMA GRADE

+ ~ / ~ ~ H o V R S siart Finish

HEMOPERF

Fig. 2. Concentrations of TCA and clinical course in patient 2. AMIT = amitriptyline, NORT = nortriptyline.

Hemoperfusion

Blood pressure

Pulse Hemoglobin (gll) Erytrocyte volume fraction (a)

Thrombocytes (l0"I)

Temperature ("C)

(mmHg)

Start 90min

lSOll10 120 152

46

269

35.3

Finish + I20 80

I I8

37

80

35.4

6 h 8 h I I I20 I20 80 90

I I5

37

80 140

36.3

Hemoperfirsio)i treatmenf O f i n ~ x i r ~ i t i o ~ i s 459

and Gilfrich et al. (4) reported that 2.13 mg of the drug were removed by means of a charcoal column from a patient with severe digitoxin poisoning. Resin perfusion is also an effective method of elimi- nating digitoxin, with a clearance of 140 ml/min for a cardiac glycoside with a much longer half-life than digoxin. However, the distribution volume for the drug is so large that even high clearance values do not necessarily result in the removal of large quanti- ties from extravasal tissues. In spite of this, our pa- tient (no. 6) improved with a shortened PQ interval. This might in part be explained by the uremic state, which includes a disturbed drug-protein binding in blood, and a greater proportion of "free" digitoxin compared to a non-azotemic state. A higher degree of non-protein-bound digitoxin might also explain the somewhat surprising cardiac toxicity at a nearly optimal therapeutic level of digitoxin concentration in serum.

Hemoperfusion may be of value in the treatment of poisoning with cytotoxic agents. Nielsen ( 1 1 ) reported recently the use of a charcoal filter in a case of insecticide (oxychlordane) poisoning.

Unfortunately, it is not yet possible to quantify mushroom toxins, and therefore it is difficult to evaluate the effect of hemoperfusion on mushroom intoxication. However. there is substantial evi- dence to support the use of hemoperfusion in the treatment of Amanita phalloides poisoning. A re- cent report (21) describes 8 cases of A. phalloides poisoning. Seven patients treated with hemoperfu- sion at one hospital made a rapid recovery. The eighth patient treated with exchange transfusion and peritoneal dialysis at another hospital remained in hepatic coma for 8 days and left the hospital after 3 months. Nine cases of Cortinarius speciosissimus poisoning have been followed up in Finland (9). Four of the six patients who required hemo- dialysis have received kidney transplants. One of the remaining three patients is healthy and two are relatively healthy. In our two patients with Cortinarius specioissimus poisoning, serum creat- inine values have returned to normal. Circum- stantial evidence thus suggests that hemoperfusion is of value in Cortinarius speciosissimus poisoning.

The role of hemoperfusion in drug intoxication has recently been questioned (10) since hospital mortality due to severe self-poisoning is low in most centres, less than 1 % in our previous study (6). This low mortality is a result of aggressive support- ive therapy, which has been the mainstay of our

treatment since the 50s. We believe that hemoper- fusion has a definite place in the therapy of drug poisoning. A small proportion of patients with TCA poisonings are critically i l l on arrival at hospital. and hospital mortality still remains high among these cases. The importance of shortening the most severe period in the course of an intoxication can hardly be exaggerated. The costs per day for nurs- ing a patient in an intensive care unit are also much higher than for care in an ordinary medical ward. I t is, however, difficult to predict on admission which patients are going to need prolonged intensive care. As soon as the course of the poisoning becomes clear, the value of hemoperfusion should be con- sidered.

Hemoperfusion is of definite value if. for exam- ple, the patient is spared a tracheostomy or if some other disease complicates the clinical course. I t may be of particular value in certain mushroom poisonings for which the prognosis is doubtful with hemodialysis and supportive therapy alone.

I .

2.

3.

4.

5 .

6.

7.

8.

9. 10.

1 1 .

12.

REFERENCES Arieff, A. & Friedman, E.: Coma following non-nar- cotic drug overdose: management of 208 adult pa- tients. Am J Sci 266: 405. 1973. Asbach, H. W.. Mohring. K.. Holz. F.. Schuler, H. W., Herman, B. & Faigle, J. W.: Haemodialysis in imipramide poisoning? An experimental study. Klin Wochenschr 54: 83, 1976. Diaz-Buzo, J . . Farmer, C. D. & Chandler, J . T.: Hemoperfusion in the treatment of amitriptyline poisoning. Trans Am SOC Artif Intern Organs 24: 699. 1978. Gilfich, H., Kasper. W., Meinertz. T., Okonek, S. & Bark, R.: Treatment of massive digitoxin overdose by charcoal hemoperfusion and cholestyramine. Lancet 1:505, 1978. Goulding, R.: Experience with hemoperfusion in drug abuse. Kidney In1 (Suppl) 7: 338, 1976. Heath, A. & Selander, D.: Self-poisoning treated in the ICU. Acta Med Scand 206: 51. 1979. Iversen, B., Willasen, Y. & Bakke, 0.: Charcoal hemoperfusion in nortriptyline poisoning. Lancet I : 388, 1978. Koffler, A., Bernstein, M.. La Sette. A. & Massry. S.: Fixed-bed charcoal hemoperfusion. Arch Intern Med 138: 1691, 1978. Kuhlback, B.: Personal communication. 1979. Lorch, J. A. & Garella, S.: Hemoperfusion to treat intoxications. Ann Intern Med 91: 301, 1979. Nielsen, E. F.: Charcoal hemoperfusion for insec- ticide overdose. Lancet 1: 506, 1978. Nyberg, G. & Mlrtensson, E.: Quantitative analysis of tricyclic antidepressants in serum from psychiatric patients. J Chromatogr 143:491. 1971.

A< 111 ,\led S c w d 207

460 A . Heath et (11.

13. Oreopoulos, D. G. & La], S.: Recovery from massive amitriptyline overdosage. Lancet 2: 221, 1%8.

14. Rosenbaum, J., Winsten, S. , Kramer, M. S., Moros, J. & Raja, R.: Resin hemoperfusion in the treatment of drug intoxication. Trans Am SOC Artif Intern Organs 16: 134, 1970.

IS. Smith, T. W., Haber, E. & Yeatman, L.: Reversal of advanced digoxin intoxication with fragments of dig- oxin specific antibodies. N Engl J Med 294:797, 1976.

16. Stalker, N., Gambertoglio, J . , Fukumitsu, C., Naughton, J. & Benet, L.: Acute massive chloral hydrate intoxication treated with hemodialysis: A clinical pharmacokinetic analysis. J Clin Pharmacol 18: 136, 1978.

17. Steel, C. M., ODuffy, J. & Brown, S . S.: Clinical effects and treatment of imipramide and amitriptyline poisoning in children. Br Med J 3: 663, 1976.

18. Tobin. M. & Mookerjee. B.: Charcoal hemoperfusion in digitalis intoxication. Dialysis and Transplantation 7:614, 1978.

19. Topf. G., Schiiltz. W. & Bartels. 0 . : XAD-4- Hamoperfusion bei Amitriptyline-Intoxikation. Pro- ceedings of Arbeitstagung iiber Hamoperfusion. 9.6.1978. p. 73. Braun-Fraba 1978.

20. Trafford, J., Jones, R., Evans. R.. Sharp, P.. Sharpstone, P. & Cook, J.: Hemoperfusion with R- 004 Amberlite resin for treating acute poisoning. Br Med J 2: 1453, 1977.

21. Wauters, J . , Rossel, C. & Farquet. J.: Amanita phal- loides poisoning treated by early charcoal hemoperfu- sion. Br Med J 2: 1465, 1978.

22. Yatzidis. H.: A convenient hemoperfusion microap- paratus over charcoal for the treatment of endogenous and exogenous intoxications. Its use as an artificial kidney. Proc Eur Dial Transplant Assoc I : 83. 1964.