Therapeutic Plasma Exchange: Complications and Management

Therapeutic Plasma Exchange: Complications and Management Michele H. Mokrzycki, MD, and Andre A. Kaplan, MD • Therapeutic plasma exchange is a treatment modality used in a variety of disease states, some of which are characterized by renal involvement (ie, Goodpasture's syndrome, multiple myeloma, cryoglobulinemia, and thrombotic thrombocytopenic purpura). To investigate the safety of this procedure we evaluated all patients receiving plasmapheresis at the University of Connecticut from January 1988 to June 1991. Sixty-eight adverse reactions occurred in 699 treatments, resulting in an incidence of 9.7%. The most frequent complications were symptoms of hypocalcemia, hypovolemia, and anaphylactoid reactions. The incidence of hypocalcemic symptoms was lowered with the prophylactic administration of calcium. Without calcium prophylaxis the incidence of symptoms was 9.1% (six in 66 treatments), whereas with calcium prophylaxis the incidence was reduced to 1% (six in 633 treatments) (P < 0.01). Treatments in which albumin was administered as volume replacement were associated with fewer adverse reactions when compared with those using fresh-frozen plasma (1.4% v 20%). Our experience, combined with the 15,658 procedures reported in the literature, reveals that serious complications do not commonly occur. These are characterized by cardiovascular events (O.2%), respiratory events (O.2%), and anaphylactoid reactions (O.25%). Hemorrhage and infection are rare, each occurring at a rate of 0.02%. Death was reported in eight of 15,658 procedures (O.05%). We conclude that therapeutic plasma exchange is relatively safe and alterations in plasma proteins generally are well tolerated. Prophylactic calcium administration lowers the incidence of hypocalcemic symptoms. Adverse reactions are associated more commonly with the administration of fresh-frozen plasma. © 1994 by the National Kidney Foundation, Inc. INDEX WORDS: Plasmapheresis; plasma exchange; calcium; citrate; albumin; plasma; immunoglobulins; coagulation factors; extracorporeal membrane compatibility; cholinesterase; kidney disease; myeloma; Goodpasture's disease; thrombotic thrombocytopenic purpura.

T

HERAPEUTIC plasma exchange is well established as an accepted treatment for many diseases, including some in which there is renal involvement (Goodpasture's syndrome, multiple myeloma, cryoglobulinemia, and thrombotic thrombocytopenic purpura [TTP]).1-6 Although commonly performed with centrifuge devices used in blood banking, an increasing number of these treatments are being provided by renal personnel using standard dialysis equipment and plasma-permeable filters.? Although several reviews have identified the potential risks of therapeutic plasmapheresis,8-10 there have been only a few large series that allow the clinician to assess the incidence of these complications. II - 18 Furthermore, although technical advances have in-

creased the ease and availability of this procedure, comprehensive guidelines for eliminating or minimizing the occurrence of complications have been lacking. Our purpose is to report the type and incidence of complications in our own series of 699 treatments and to compare them with those of previous reports. The treatment of complications and methods to reduce the risk of their occurrence are outlined. MATERIALS AND METHODS All patients who underwent plasmapheresis at the University of Connecticut Health Center between January 1988 to June 1991 were studied. Indications for plasmapheresis are listed

in Table I. Therapeutic plasma exchange (TPE) was performed most frequently in patients with neurologic diseases (26 of 50 patients), including chronic demyelinating polyneuropathy, myasthenia gravis, and Guillain-Barre syndrome. The connective tissue diseases were the second most common indication (13 patients) and consisted of systemic lupus erythematosus, cryoglobulinemia, and vasculitis. Ten patients were treated for hematologic indications (ie, myeloma, macroglobulinemia, idiopathic thrombocytopenic purpura, TTP, and autoimmune hemolytic anemia). One patient was receiving TPE for hypercholesterolemia and pruritus secondary to primary biliary cirrhosis. The diseases for which TPE was performed in our study are comparable to those in other recent series. 13 ,14,16,18 All procedures were done on an autopheresis C rotating membrane system modified to provide double-needle continuous flow operation, as previously described. 19,20 Venous access was obtained using a dual-lumen hemodialysis catheter inserted into either the femoral or subclavian vein, or by inserting

16-gauge needles into both antecubital veins or into an arteriovenous graft or fistula. On average, each treatment ex-

From the Department ofMedicine, Division ofNephrology, University of Connecticut School of Medicine, Farmington, CT.

Received April 2, 1993; accepted in revised form January 4, 1994. Dr Mokrzycki is currently at the Bronx Municipal Hospital Center, Albert Einstein College of Medicine, Bronx, NY. Address reprint requests to Andre A. Kaplan, MD, Division of Nephrology, MC 1405, University of Connecticut Health Center, Farmington, CT 06030. © 1994 by the National Kidney Foundation, Inc. 0272-6386/94/2306-0008$3.00/0

American Journal of Kidney Diseases, Vol 23, No 6 (June), 1994: pp 817-827

817

818

MOKRZYCKI AND KAPLAN

changed 3 L over a 70-minute period, yielding a plasma removal rate of 45 mL/min.21 Formula B citrate solution at a 6% or 8% volume ratio was used for anticoagulation and was infused at a rate of approximately 0.7 mg/kg/min. In 66 treatments calcium was given for hypocalcemic symptoms (ie, distal extremity paresthesias and perioral numbness). Calcium replacement was administered within 15 minutes of symptom onset as 10 to 20 mL of 10% calcium chloride solution. In 633 treatments prophylactic calcium was administered (10 mL of 10% calcium chloride solution) 15 minutes after the beginning of the treatment and was infused over a 15- to 30-minute period. In treatments lasting longer than I hour, an additional 10 mL of calcium was administered at the I-hour point. Replacement solution was in the form of 5% albumin for 612 treatments. In 45 of these treatments, partial replacement with fresh-frozen plasma (FFP), usually 400 to 500 mL, was administered at the end of the treatment to replenish coagulation factors. In 87 treatments, the entire plasma volume was in the form of FFP. Minimal symptoms were considered those of little clinical significance. Moderate symptoms were of clinical significance, but were easily treatable and without prolonged consequence. Serious symptoms were defined as those that were considered potentially life-threatening. The same two members of the hemodialysis/plasmapheresis nursing staff performed and recorded adverse reactions for all treatments. An exhaustive literature search was undertaken to identify all reported series of TPE in which the incidence of complications could be determined. The search was performed using the MEDLINE database and by manual examination of all relevant bibliographies.

RESULTS

In a 42-month period, plasmapheresis was performed on 50 patients, for a total of699 treatments. Sixty-eight adverse reactions were reported (Table 2, column 8), resulting in a complication rate of 9.7%. Minimal reactions occurred in 38 treatments (5.4%) and were characterized by urticaria, paresthesias, nausea, dizziness, and leg cramps. Included in this group are 66 treatments performed without calcium prophylaxis, in which six patients had symptoms of hypocalcemia (9.1 %). In contrast, these symptoms were noted in only six of 633 treatments in which calcium was administered at the initiation of the treatment (0.95%) (P < 0.01 using chi-squared analysis). Prophylactic calcium administration was not associated with an increased risk of membrane thrombosis. Moderate symptoms occurred in 17 treatments (2.4%) and included hypotension, chest pain, and ventricular ectopy. All were of brief duration with no sequelae. Of seven episodes of hypotension, five were mild and responded rapidly to volume

Table 1. Indications for Plasmapheresis

Diagnosis

Chronic demyelinating polyneuropathy Myasthenia gravis Systemic lupus erythematosis Lupus nephritis Lupus anticoagulant Cerebritis Vasculitis Guillain-Barre syndrome Myeloma nephropathy Cryoglobulinemia Waldenstrom's macroglobulinemia/ hyperviscosity Primary central nervous system vasculitis Idiopathic thrombocytopenic purpura Thrombotic thrombocytopenic purpura Autoimmune hemolytic anemia Primary biliary cirrhosis/ hypercholesterolemia Total

No. of Treatments

No. of Patients

175 48 38 16 16 3 3 41 15 113

12 8 8 4 2

7

2

10

2

5

2

84 3

2

160

1

699

50

1 6 3 3

replacement; one patient with pretreatment hypotension required increased vasopressor support. One episode of hypotension was due to loss of venous access and resulted in an inability to replace 300 mL of removed plasma. Two patients experienced chest pain during the treatment; both episodes resolved quickly and no long-term clinical effect was observed. During one treatment a four-beat run of ventricular tachycardia was noted, for which the patient was asymptomatic and otherwise clinically stable. Routine electrocardiographic monitoring was not performed on most patients; therefore, the incidence of asymptomatic arrhythmia during plasma exchange cannot be assessed. There were five potentially serious events (0.7%), all of which were experienced by the same patient and were anaphylactoid reactions associated with FFP administration. The patient's underlying diagnosis was TTP, which was considered an absolute indication for plasma exchange. The patient's first episode was characterized by urticaria, hypotension, and wheezing, which responded to appropriate therapy. Subsequent episodes, occurring after premedication

COMPLICATIONS OF THERAPEUTIC PLASMA EXCHANGE

819

Table 2. Complications of Plasmapheresis Centrifugal System

Borberg"

Aufeuvreetal'2

Ziselman et al'3

No. of treatments Adverse reactions

205 13%

3.086 22%

1.389 1.6%

Mild Moderate Severe Deaths

o

4.8% 16.4% 0.6% 0.1%

0.4* 0.6% 0.5% 0.1%

Symptoms Urticaria Paresthesias Muscle cramps Dizziness Headache Nausea Hypotension Chest pain Dysrhythmia Anaphylactoid treatments Rigors Hyperthermia Bronchospasm Seizure Respiratory arrest/ pulmonary edema Myocardial ischemia Shock/MI Metabolic alkalosis DIC CNS ischemia Other Hepatitis Hemorrhage Hypoxemia Pulmonary embolism Access-related Thrombosis/hemorrhage Infection Pneumothorax Mechanical

11.2% 2%

o

1.5%

0.03% 0.5% 8.8% 1.0%

0.1% 0.4% 0.1%

Fabre et al 14

Rossi et al 15

Sprenger et al'6

Present Study

Sutton et al'·

578 25%

926 17.3%

306 4.2%

120 17.5%

699 9.7%

5.235 12%

} 23%

7.6% 6.5% 3.1%

o

5 12.5%

5.4% 2.4% 0.7%

9% 3% 0.5%

2.4% 1.7% 0.4%

3.7% } 2.5%

0.3% 0.6% 1.4% 1.3% 0.1% 0.7%

1.2% 1.5% 2.3% 0.2% 0.1% 0.5%

1.5%

o

o

12% 9%

2.4%

1% 0.2% 0.5%

5% 1% 1.6%

0.2%

0.7%

}0.7%

}5.3%

0.1% 0.03% 0.03% 0.03%

Both'·

Samtleben et al 17

4.2%

o o

o o

0.7%

8.3% 5%

2%

4.2%

0.03%

0.03% 0.3%

1.5%

Membrane-based System

o

o

1.1% 0.4%

0.1% 0.4%

0.8%

0.8%

0.2% 0.1% 0.1%

0.1% 0.3% 0.7%

0.2% 0.1% 0.1% 0.7% 0.3% 0.1% 0.4%

1.2%

4%

1.5%

0.02%

0.08%

Abbreviations: MI. myocardial infarction; DIC. disseminated intravascular coagulation; CNS. central nervous system.

with prednisone, ephedrine, and diphenhydramine, were less severe and were characterized by wheezing and urticaria, but without hypotension, thus permitting the treatments to be continued. No long-term sequelae were noted. Complications related to access occurred in eight treatments (1.1 %) and included four episodes of thrombosis, two episodes of catheter-related infection, one bleeding episode postcatheter removal, and one pneumothorax that occurred as a result of catheter insertion.

In 87 treatments, plasma volume was replaced entirely with FFP, 84 for the treatment ofTTP. In an additional 45 treatments, 400 to 500 mL of FFP was administered at the end of the exchange to minimize the risk of bleeding. Twenty-six reactions were associated with the use of FFP; 21 of these were mild or moderate in nature (ie, mild urticaria). Five reactions were severe with anaphylactoid symptoms, ie, urticaria, wheezing, or hypotension (as described above as "serious events"). Twenty-five events, including the five serious reactions, oc-

820

MOKRZYCKI AND KAPLAN

curred in the same patient. Of 132 treatments in which FFP was administered, 106 were without adverse effect. DISCUSSION

Plasmapheresis can be performed safely, and complications can be avoided when alterations in plasma components are anticipated. To properly comprehend these changes several different aspects of the treatment must be considered. Hypocalcemia Symptoms of hypocalcemia (ie, perioral and distal extremity paresthesias) are the result of citrate infusion, either as the treatment's anticoagulant or in FFP administered as the replacement fluid. The incidence of hypocalcemic symptoms in previous reports is 1.5% to 9% (Table 2) and is similar to our overall experience of 1.7%. In agreement with a previous report by Buskard et aI, we found that the incidence of paresthesias was significantly reduced when calcium was administered prophylactically.22 These findings are in contrast to those of Sutton et aI, who reported no difference in symptomatology with or without calcium supplementation. 18 In Sutton et aI's report, however, the timing of calcium replacement is not reported, ie, prophylactically or as a treatment for ongoing symptomatology. These investigators also noted that hypocalcemic symptoms are more frequent in treatments in which plasma, as opposed to albumin, was used as replacement (4.8% v 1.8%).18 This result is most likely attributable to the large amounts of citrate contained in FFP (approximately 14% by volume). In addition to paresthesias, Hester et al demonstrated that citrate-induced hypocalcemia is often associated with prolongation of the QT interval on electrocardiogram, thus potentially increasing the risk of cardiac arrhythmia. 23 Olson et al have demonstrated that a 0.09-second prolongation of the QT interval is expected when ionized calcium levels are reduced by 35%.24 To avoid this complication, calcium replacement and control of the citrate infusion rate to between 1.0 and 1.8 mg/kg/min have been recommended. 23 As previously stated, the average citrate infusion rate in our series was 0.7 mg/kg/min.

Silberstein et al studied calcium homeostasis during plasma exchange and reported an 18% posttreatment decrease in the serum ionized calcium. 25 These investigators also reported that nterminal parathyroid hormone levels were increased 280% from baseline (mid-pheresis) in treatments without calcium supplementation compared with 160% in treatments using calcium replacement. Silberstein et al concluded that increases in parathyroid hormone and in nephrogenous cyclic adenosine monophosphate confirm an endogenous compensatory response to calcium removal during TPE and argued that calcium administration was unnecessary, as symptomatology was minor. In their study, however, symptoms of hypocalcemia were treated by decreasing the rate of exchange and, as has been recommended by Huestis, decreasing citrate to blood ratio and supplementing with heparin. 25 .26 In patients who receive multiple treatments for prolonged periods, significant loss of calcium may occur. This is supported by calcium kinetic studies in which a net loss of approximately 150 mg of calcium per treatment was found in patients not receiving calcium compared with an average 90-mg positive balance in patients who were given calcium. 26 Calcium supplementation is beneficial to the maintenance of calcium balance when albumin is used as the replacement fluid, particularly in patients undergoing multiple treatments. Its prophylactic administration decreases the incidence of hypocalcemic symptoms, allowing for uninterrupted plasma exchange and the use of adequate anticoagulation. Coagulation Abnormalities Clinical bleeding associated with plasmapheresis is rarely reported. In our study, one hemorrhagic episode was experienced after a largebore femoral vein catheter was removed soon after the patient had received three consecutive daily treatments. Following this episode, we initiated a protocol of substituting 2 U of FFP (400 to 500 mL) at the end of each exchange after which removal of a large-diameter catheter was anticipated. No further hemorrhagic complications were noted. Rossi et al reported one episode of hematemesis and one episode of epistaxis. 15 Sutton et al reported bleeding from a previous catheter site during one treatment. 18 In the re-

COMPLICATIONS OF THERAPEUTIC PLASMA EXCHANGE

maining six series reviewed, no bleeding complications were reported (Table 2). When albumin is used as the replacement fluid there is a depletion of all coagulation factors, including fibrinogen and antithrombin III (AT-III). Coagulation factors have been studied by several groupS.28-30 Chimside et al reported the following changes in coagulation factors immediately following a single plasma exchange, recorded as the percentage of initial plasma concentration: fibrinogen, 20%; prothrombin, 40%; factor V, 42%; factor VII, 47%; factor VIII, 50%; factor IX, 57%; factor X, 32%; and AT-III, 42%.28 After a single plasma exchange, recovery of coagulation factors is biphasic, characterized by a rapid initial increase up to 4 hours postpheresis and followed by a slower increase 4 to 24 hours postpheresis. 28 Twenty-four hours after treatment, fibrinogen levels are approximately 50% and AT-III levels are 85% of initial levels; both require 48 to 72 hours for complete recovery.28 One day following treatment, the prothrombin level is 75% and factor X is 30% of the original level; all other coagulation factors completely recover to normal values. Prothrombin time increases approximately 30% and partial thromboplastin time doubles immediately posttreatment. 25 Partial thromboplastin time and thrombin time are back to normal range 4 hours postpheresis, and prothrombin time normalizes in 24 hours28 When multiple treatments are performed over a short period (ie, three or more treatments per week), the depletion in clotting factors is more pronounced and may require several days for spontaneous recovery,z7-29 Under these conditions, the risks of hemorrhage can be minimized with the replacement of 2 U (400 to 500 mL) of FFP given toward the end of the procedure. Decreases in platelet count depend on the method of treatment. Whereas centrifugal methods may be associated with a 50% decrease, membrane plasma separation causes a more modest decrease in platelet count (15%).27,30-32 Thrombocytopenia may result from a loss of platelets in the discarded plasma or via filter thrombosis. Hematocrit may decrease by 10% after each treatment, even in the absence of any extracorporeal losses or hemolysis. This phenomenon may be due to intravascular dilution related to

821

the use of relatively hyperoncotic replacement solutions. 27 ,33,34 Lowering of AT-III levels and other inhibitors of coagulation may create a hypercoaguable state. Sultan et al reported two cases of thrombosis in which postpheresis depletion of AT-III was found and discussed the increased risk of thrombosis due to the removal of circulating anticoagulants. 35 In review of our data, four treatments (0.6%) were complicated by thrombosis associated with intravascular catheters. Pulmonary embolism (0.02% to 0.1 %), cerebral ischemia (0.02% to 0.1 %), and myocardial infarction (0.06% to 0.14%) have been reported infrequently,13,36 and an association with low levels of AT-III is speculative. Injection

When albumin is used as the replacement fluid, removal of immunoglobulins and complement may, theoretically, predispose patients undergoing plasmapheresis to high rates of infection. One plasma volume exchange will result in a 60% reduction in serum immunoglobulin levels and a net 20% reduction in total body immunoglobulin storesy,31,33 Multiple treatments over short periods, especially when associated with immunosuppressive agents, will yield a substantial decrease in immunoglobulin levels that may persist for several weeks. 35 ,37,38 Although concentrations of C3 and C4 may be depleted by a series of daily treatments, these levels rebound within several days. CH50 is unaffected by repetitive exchanges. 37 Therapeutic plasma exchange with FFP replacement would not be expected to deplete immunoglobulin or complement levels. The true incidence of infection in patients undergoing TPE is controversial. In a retrospective study, Wing et al compared the incidence of infection in patients with rapidly progressive glomerulonephritis who received standard therapy (steroids and cytotoxic agents) with or without plasma exchange. 39 A significantly higher occurrence of opportunistic infections was found among the apheresis-treated group (five of eight patients [63%]) compared with the control group (two of21 patients [9.5%]) (P = 0.016). Some of the control cases, however, were taken from retrospective review, and two of the five patients who developed an infectious complication in the apheresis-treated group had severe granulocyto-

822

MOKRZYCKI AND KAPLAN Table 3. Management Strategies to Avoid Complications of Plasmapheresis Complications

Hypocalcemia Hemorrhage Sensitivity to replacement fluids

Thrombocytopenia Volume-related hypotension Infection postpheresis Hypokalemia Membrane biocompatibility Hypothermia

Management

Prophylactic calcium administration (10 mL of 10% CaCI 2 infused over 15 to 30 min)27 Partial FFP replacement in patients at high risk for hemorrhage Evaluate coagulation parameters before catheter removal Consider diagnostic evaluation (ie, anti-lgA antibody, anti-ethylene oxide antibody, anti-human serum albumin antibody, endotoxin assay, and bacterial cultures of replacement fluid, etc)69 Premedication regimen for sensitized individuals: (1) prednisone 50 mg orally 13 hr, 7 hr, and 1 hr pretreatment; (2) diphenhydramine 50 mg orally 1 hr pretreatment, and (3) ephedrine 25 mg orally 1 hr pretreatment and before pheresis69 Consider membrane plasma separation 31 Consider continuous flow separation with matched input and output Consider increasing protein concentration of replacement fluid Infusion of intravenous immunoglobulin (100 to 400 mg/kg) Ensure a potassium concentration of 4 mmol/L in the replacement solution 16 Change membrane or consider centrifugal method of plasma separation 83 Warm replacement fluids

penia, which was not likely to be the result of the plasma exchange procedure per se. These investigators also reviewed previous reports of infections occurring in patients with rapidly progressive glomerulonephritis treated with TPE and immunosuppression. 39 Nine episodes of infection were found in 34 patients receiving treatment, four resulting in death.40-48 Two of the nine patients with infections were granulocytopenic. 40,46 In contrast, in a prospective randomized trial, Pohl et al studied 86 patients with lupus nephritis receiving cyclophosphamide and steroids, with or without pheresis. 49 These investigators found no increase in the rate of infections or in infection-related deaths in the apheresis-treated group. In patients treated with TPE, the infection rate was 1.22 infections per 100 weeks with three deaths, compared with 1.15 infections per 100 weeks and four deaths in the control group. Data from patients with myasthenia gravis suggest a lower incidence of infection than in patients with renal disease. 50-54 Thirty-six patients with myasthenia gravis received plasma exchange in addition to prednisone and azathioprine, and were followed for a mean period of9 months. Of these, only one patient developed an infectious complication. 50-53 This was a chronic ventilator-dependent patient who developed "marrow failure" and died with septicemia. 52 The etiology of "marrow failure" was not elucidated by the investigators, but may have been due to the immunosuppressive drugs the patient had been re-

ceiving. Granulocytopenia would have put this patient at high risk for infection, independent of the plasmapheresis therapy. We report two episodes of infection (0.3% incidence) in the immediate postpheresis period. One episode of catheter-related infection occurred in the patient described earlier, who had been receiving prednisone for TTP. It is unlikely that immunoglobulin levels were significantly lowered as the patient was undergoing weekly plasma exchange with FFP as replacement fluid. The second episode of catheter infection occurred in a patient who was receiving daily plasma exchange for TTP. Once again, replacement fluid was FFP and immunoglobulin depletion would not be expected. Both episodes of infection responded to intravenous antibiotics and catheter removal. The other reviews had no reports of infection (Table 2). The above results do not rule out the possibility that immunoglobulin or complement depletion may impair a patient's ability to combat an ongoing infection. Thus, if a severe infection develops in the immediate postpheresis period, a reasonable approach would be a single infusion of immunoglobulin (100 to 400 mg/kg intravenously) (Table 3), similar to the replacement dose recommended in patients with hypogammaglobulinemia. 55 ,56 Risk of viral transmission during plasma exchange is directly related to replacement with FFP. Albumin and immunoglobulin prepara-

COMPLICATIONS OF THERAPEUTIC PLASMA EXCHANGE

tions are treated with heat, inactivating hepatitis and the human immunodeficiency virus. 56 ,57 In our review, Fabre et al report one case of latent non-A/non-B hepatitis infection 14 and no viral infections were reported in the other series (Table 2). Transmission of the human immunodeficiency virus through therapeutic plasmapheresis is unlikely and would be anticipated as the result of infected FFP. 58 The current incidence of transfusion-acquired hepatitis B is approximately 0.0005% per unit 59,60; that of hepatitis C is 0.03% per unit60,61 and that of the human immunodeficiency virus is approximately 0.0004% per unit transfused. 60,62-65 It should be noted that during a single plasma volume exchange with FFP (approximately 3 L), 10 to 15 U, obtained from an equal number of donors, are used. Reactions to Fresh-Frozen Plasma and Albumin In our study, 567 treatments were performed with albumin as the sole replacement, while 132 treatments used at least some amount ofFFP (87 treatments were entirely with FFP and 45 were treated with partial FFP replacement). The incidence of adverse reactions in the albumin replacement group was 1.4%, compared with 20% in the group receiving FFP. In previous reports, FFP was associated with an incidence of adverse reactions ranging between 0.02% and 21 %. These were anaphylactoid in nature and were characterized by fever, rigors, urticaria, wheezing, and hypotension. 66,67 Forty-two deaths associated with therapeutic plasmapheresis have been compiled by Huestis, at least 30 of which were associated with FFP replacement. 68 In five other cases, replacement solution was uncertain and in the remaining six cases, albumin or plasma protein fraction (PPF) was used as replacement. The major causes of death were cardiovascular, respiratory, and anaphylactic. Unfortunately, specific details about these deaths and their temporal relationship to the plasmapheresis procedure are not provided. In his review, Huestis calculated an estimated three deaths per 10,000 procedures. 68 In 1980, Aufeuvre et al compiled data from 2,769 plasma exchange procedures via a questionnaire and added this information to their original study of 3,431 treatments. 12,36 The second study focused exclusively on serious com-

823

plications and therefore the data were excluded from Table 2. Aufeuvre et al reported 33 serious accidents and seven deaths in 6,200 treatments. Causes of death included nonhemodynamic pulmonary edema (FFP replacement), cardiac dysrhythmia, hemodynamic pulmonary edema, and pulmonary embolism. When the data from Aufeuvre et aI's second study are included, a total of eight deaths in 15,658 treatments is tabulated (0.05%). Human serum albumin consists of 96% albumin and trace amounts of alpha and beta globulins. Five percent albumin is isosmotic to plasma, contains no preservatives, and is characterized by a sodium level of approximately 145 ± 15 mmol/L and a potassium level lower than 2 mmol/L. 57 Albumin has been found to be a safe product and anaphylactoid reactions occur rarely, with a reported incidence between 0.01%66 and 4.0%.67 Potential reasons for anaphylactoid reactions are (1) the presence of anti-immunoglobulin A (IgA) antibodies in a patient who is IgA deficient and receiving IgA-containing fluids (ie, FFP, immunoglobulins), (2) contamination with bacteria and bacterial endotoxin or pyrogens, (3) the presence of a prekallikrein activator and bradykinin, and (4) the formation of antibodies to polymerized albumin created by heat treatment or stabilization with sodium caprylate. 57,69 Although two cases of severe anaphylactoid reactions have been associated with albumin replacement, the six deaths reported by Huestis as being associated with albumin may have been due to the use of PPF, since he combined albumin and PPF into one groUp.66,68,70 Plasma protein fraction contains approximately 87% albumin and 13% alpha and beta globulins, and is easier and less costly to prepare than albumin. Plasma protein fraction has been associated with hypotensive episodes and circulatory collapse, possibly due to the presence of prekallikrein activator and bradykinin. 57 In patients who demonstrate a particular sensitivity to FFP or albumin, we have found a premedication regimen of ephedrine, prednisone, and diphenhydramine to be effective in minimizing symptomatology; thereafter, further diagnostic evaluation should be considered (Table 3).69 Electrolyte, Vitamin, and Drug Removal As opposed to FFP, albumin is essentially devoid of potassium , and a 25% reduction in serum

824

potassium levels may occur in the immediate postpheresis period. 71 With the most modern techniques, plasma exchange may be very rapid (2 to 3 L/hr) and the risk of hypokalemic arrhythmia during pheresis and in the immediate postpheresis period can be reduced by ensuring that the replacement solution contains approximately 4 mmol/L of potassium. (We add 4 mmol potassium to each liter of 5% albumin.) Metabolic alkalosis may result from large amounts of infused citrate. Formula B citrate solution, which was used in our series, is an isosmotic solution and contains 73 mmol/L of citrate, which, when metabolized, will yield 219 mmol/L of bicarbonate. Formula A citrate solution is a hyperosmotic solution containing 252 mmol/L of sodium and its 112 mmol/L of citrate will yield 336 mmol/L of bicarbonate. Fresh-frozen plasma contains approximately 14% citrate solution by volume. In most patients, postpheresis bicarbonate levels are unchanged. 71 In patients with renal failure severe alkalemia may result from repeated treatments, especially when FFP is used as the replacement. 72 Immediately postpheresis, decreases in blood concentrations of vitamins B 12, B6, A, C, and E and beta-carotene (24% to 48% reduction) have been noted; however, these rebound to pretreatment levels within 24 hours. 73 Folate, thiamine, nicotinate, biotin, riboflavin, and pantothenate are not significantly altered by plasma exchange. All albumin solutions are contaminated with between 4 to 24 Jlmol/L of aluminum. 74.75 Repetitive plasma exchange with albumin may result in significant accumulation of aluminum. This is most likely to occur in patients with severe renal insufficiency, in whom 60% to 70% of infused aluminum is retained. 75 Nonetheless, bone deposition of aluminum has been reported in a patient with normal renal function. 74,75 During plasma exchange, alterations in plasma drug levels are most dependent on the percentage of protein binding and the volume of distribution. 27 ,76 Prednisone and prednisolone are minimally removed by plasma exchange, and supplemental dosing following the treatment has been found to be unnecessary.77 One can anticipate minimal removal of cyclophosphamide with TPE, as the percentage of protein binding is low (12%) and the volume of distribution is relatively large (0.8 L/kg).76 Azathioprine, how-

MOKRZYCKI AND KAPLAN

ever, is approximately 30% protein bound, with a volume of distribution of 0.6 L/kg, suggesting that moderate amounts may be removed during pheresis. The above recommendations notwithstanding, drug kinetics in any given individual may vary due to concurrent illness. When possible, all medications should be administered after a given plasma exchange.

Hypovolemia The incidence of hypotension in our study was 1.4%, with an overall incidence of 1.7% in the series reviewed (Table 2). Hypotension may occur

during TPE for a variety of reasons, including vasovagal episodes, hypo-oncotic fluid replacement, delayed or inadequate volume replacement, anaphylaxis, cardiac arrhythmia, and cardiovascular collapse. Discontinuous flow plasma exchange systems may be prone to a higher incidence of hypotensive episodes due to intermittent hypovolemia. The use of hypo-oncotic replacement solutions may also increase the risk of hypotensive events. In some institutions, albumin replacement solution is prepared by dilution with an electrolyte solution to achieve a concentration of 3.5% albumin. In most patients, this solution is clearly hypo-oncotic to the patient's plasma78 and may predispose them to hypotension, even when a policy of 1: 1 volume replacement is rigorously followed. We use an undiluted 5% albumin solution that is slightly hyperoncotic to the patient's own plasma.

Miscellaneous Complications Two reports of apneic events have been reported following plasma exchange and were due to abnormally low posttreatment levels of plasma cholinesterase. 37,79 Succinylcholine, a commonly used anesthetic agent that was used in these cases, is metabolized by cholinesterase. Cholinesterase levels are reduced by 50% immediately after a single treatment. 80 Levels less than 30% normal (approximately <1,000 V/L) are likely to be associated with decreased metabolism of succinylcholine. 81 ,82 While FFP contains normal levels of cholinesterase, albumin and PPF contain little or none of this enzyme. Thus, cholinesterase depletion is expected only when albumin or PPF is used as replacement. Anesthetic agents dependent on serum cholinesterase for their metabolism should be used with caution immediately post-

COMPLICATIONS OF THERAPEUTIC PLASMA EXCHANGE

plasma exchange, especially after a series of daily treatments. Repletion of cholinesterase with FFP may be a reasonable approach for treatments in the immediate preoperative period. Adverse reactions (hypotension, dyspnea, and chest pain) may occur secondary to complementmediated membrane bioincompatibility, similar to those described during hemodialysis. 83 Anaphylactoid symptoms may also occur due to ethylene oxide sensitivity, which is used as a sterilizing agent. 84 The incidence of filter-related leukocytopenia, thrombocytopenia, and hypocomplementemia is reduced with more biocompatible membranes and reactions to ethylene oxide can be avoided with adequate priming of the filter. 85 Despite concerns regarding hemolysis, especially with membrane-based techniques, our review identified only one episode of hemolysis that was related to inappropriately hypotonic priming solution. Chills and other symptoms of hypothermia may be experienced due to inadequately warmed replacement fluid. ACKNOWLEDGMENT The authors thank Judith Reardon, RN, and Judith Sevigny, RN, for their skill in performing the plasmapheresis treatments and for their contribution in compiling these data.

REFERENCES I. Consensus Conference: The utility of therapeutic plasmapheresis for neurologic disorders. JAMA 253:819-825, 1985 2. Clinical Applications Committee, American Society for Apheresis: Clinical applications of therapeutic apheresis. J Clin Apheresis 3:i-vi, 1-92, 1986 3. D'Amico G, Colasanti G, Ferrario F, Sinico RA: Renal involvement in essential mixed cryoglobulinemia. Kidney Int 35:1004-1014, 1986 4. Zucchelli P, Pasquali S, Cagnoli L, Ferrari G: Controlled plasma exchange trial in acute renal failure due to multiple myeloma. Kidney Int 33:1175-1180, 1988 5. Rock GA, Shumak KH, Buskard NA, Blanchette VS, Kelton JG, Nair RC, SpasoffRA, and the Canadian Apheresis Study Group: Comparison of plasma exchange with plasma infusion in the treatment of TTP. N Engl J Med 325:393397, 1991 6. Consensus Conference: The utility of therapeutic plasmapheresis for neurologic disorders. JAMA 256:1333-1337, 1986 7. Gerhardt RE, Ntosos KA, Koethe JD, Lodge S, Wolf CJ: Acute plasma separation with hemodialysis equipment. J Am Soc NephroI2:1455-1458, 1992 8. Isbister JP: The risk/benefit equation for therapeutic plasma exchange, in Nydegger UE (ed): Therapeutic Hemapheresis in the 1990s. Basel, Switzerland, Karger AG, 1990, pp 10-30

825 9. Hazards of Apheresis. Lancet 2:1025-1026, 1982 (editorial) 10. Westphal RG: Complications of hemapheresis, in Westphal RG, Kasprisin DO (eds): Current Status of Hemapheresis: Indications, Technology and Complications. Arlington, VA, American Association of Blood Banks, 1987, pp 87-104 II. Borberg H: Problems of plasma exchange therapy, in Gurland HJ, Heinze V, Lee HA (eds): Therapeutic Plasma Exchange. New York, NY, Springer-Verlag, 1980, pp 191201 12. Aufeuvre JP, Morin-Hertel F, Cohen-Solal M, Lefloch A, Baudelot J: Hazards of plasma exchange, in Sieberth HG (ed): Plasma Exchange. Stuttgart, Germany, FK Schattauer Verlag, 1980, pp 149-157 13. Ziselman EM, Bongiovanni MB, Wurzel HA: The complications of therapeutic plasma exchange. Vox Sang 46: 270-276, 1984 14. Fabre M, Andreu G, Mannoni P: Some biological modifications and clinical hazards observed during plasma exchanges, in Seiberth HG (ed): Plasma Exchange. Stuttgart, Germany, FK Schattauer Verlag, 1980, pp 143-148 15. Rossi PL, Cecchini L, Minichella G, De Rosa G, Alfano G, Pieralla L, Testa A, Candido A, Vittorio M, Mango G: Comparison of the side effects of therapeutic cytapheresis and those of other types of hemapheresis. Haematologica 76:7580, 1991 (suppl I) 16. Sprenger KBG, Rasche H, Franz HE: Membrane plasma separation: Complications and monitoring. Artif Organs 8:360-363, 1984 17. Samtleben W, Hillebrand G, Krumme D, Gurland HJ: Membrane plasma separation: Clinical experience with more than 120 plasma exchanges, in Sieverth HG (ed): Plasma Exchange. Stuttgart, Germany, FK Schattauer Verlag, 1980, pp 23-27 18. Sutton DMC, Nair RC, Rock G, and the Canadian Apheresis Study Group: Complications of plasma exchange. Transfusion 29:124-127,1989 19. Kaplan AA, Halley SE: Evaluation of a rotating filter for use with therapeutic plasma exchange. ASAIO Trans 34: 274-276, 1988 20. Kaplan AA, Halley SE, Reardon J, Sevigny J: One year's experience using a rotating filter for therapeutic plasma exchange. ASAIO Trans 35:262-264, 1989 21. Kaplan AA, Reardon J, Sevigny J: Continuing evaluation of a rotating filter for therapeutic plasma exchange. Blood Purif9:31-32, 1991 (abstr) 22. Buskard NA, Varghese Z, Wills MR: Correction ofhypocalcemic symptoms during plasma exchange. Lancet 2:344345, 1976 23. Hester JP, McCullough J, Mishler JM, Szymanski 10: Dosage regimens for citrate anticoagulants. J Clin Apheresis 1:149-157, 1983 24. Olson PR, Cox C, McCullough J: Laboratory and clinical effects of the infusion of ACD solution during plateletpheresis. Vox Sang 33:79-87, 1977 25. Silberstein LE, Naryshkin S, Haddad JJ, Strauss JF: Calcium homeostasis during therapeutic plasma exchange. Transfusion 26:151-155, 1986 26. Huestis DW: Complications of therapeutic apheresis, in Pinada AA, Valbonesis M, Diggs JC (eds): Therapeutic

826 Hemapheresis. Milan, Italy, Wichtig Editore, 1986, pp 179186 27. Kaplan AA, Halley SE: Plasma exchange with a rotating filter. Kidney Int 38: 160-166, 1990 28. Chirnside A, Urbaniak SJ, Prowse CV, Keller AJ: Coagulation abnormalities following intensive plasma exchange on the cell separator. Br J HaematoI48:627-634, 1981 29. Gelabert A, Puig L, Maragall S, Monteagudo J, Castillo R: Coagulation alterations during massive plasmapheresis, in Sieverth HG (ed): Plasma Exchange. Stuttgart, Germany, FK Schattauer Verlag, 1980, pp 71-75 30. Flaum MA, Cuneo RA, Appelbaum FR, Disseroth AB, Engel WK, Gralnick HR: The hemostatic imbalance of plasma exchange transfusion. Blood 54:694-702, 1979 31. Keller AJ, Chirnside A, Urbaniak SJ: Coagulation abnormalities produced by plasma exchange on the cell separator with special reference to fibrinogen and platelet levels. Br J HaematoI42:593-603, 1979 32. Gurland HJ, Lysaght MJ, Samtleben W, Schmidt B: A comparison of centrifugal and membrane based apheresis formats. Int J ArtifOrgans 7:35-38, 1984 33. Chopek M, McCullough J: Protein and biochemical changes during plasma exchange, in Ulmas J, Berkman E (eds): Therapeutic Hemapheresis: A Technical Workshop. Washington, DC, American Association of Blood Banks, 1980, pp 13-52 34. Wood L, Jacobs P: The effect of serial therapeutic plasmapheresis on platelet count, coagulation factors, plasma immunoglobulin and complement levels. J Clin Apheresis 3: 124-128, 1986 35. Sultan Y, Bussel A, Maisonneuve P, Sitty X, Gajdos P: Potential danger of thrombosis after plasma exchange in the treatment of patients with immune disease. Transfusion 19:588-593, 1979 36. Aufeuvre JP, Mortin-Hertel F, Cohen-Solal M, Lefloch A, Baudelot J: Clinical tolerance and hazards of plasma exchanges: A study of 6200 plasma exchanges in 1033 patients, in Beyer JH, Burberg H, Fuchs C, Nagel GA (eds): Plasmapheresis in Immunology and Oncology. Basel, Switzerland, Karger, 1982, pp 65-77 37. Keller AJ, Urbaniak SJ: Intensive plasma exchange on the cell separator: Effects on serum immunoglobulins and complement components. Br J Haematol 38:531-540, 1978 38. Kaplan AA: Towards a rational prescription of plasma exchange: The kinetics of immunoglobulin removal. Semin Dial 5:227-229, 1992 (editorial) 39. Wing EJ, Bruns FJ, Fraley OS, Segel DP, Adler S: Infectious complications with plasmapheresis in rapidly progressively glomerulonephritis. JAMA 244:2423-2426, 1980 40. Lockwood CM, Pinching AJ, Sweny PM, Rees AJ, Pussell B, Uff J, Peters OK: Plasma exchange and immunosuppression in the treatment of fulminating immune-complex crescentic nephritis. Lancet 1:63-67, 1977 41. Rossen RD, Hersh EM, Sharp JT, McCredie KB, Gyorkey F, Suki WN, Eknoyan G, Reisberg MA: Effect of plasma exchange on circulating immune complexes and antibody formation in patients treated with cyclophosphamide and prednisone. Am J Med 63:674-682, 1977 42. Lockwood CM, Pearson TA, Rees AJ, Evans OJ, Peters OK, Wilson CB: Immunosuppression and plasma-exchange

MOKRZYCKI AND KAPLAN

in the treatment of Goodpasture's syndrome. Lancet 1:711715, 1976 43. Johnson JP, Whitman W, Briggs WA, Wilson CB: Plasmapheresis and immunosuppressive agents in antibasement membrane antibody-induced Goodpasture's syndrome. Am J Med 64:354-359, 1978 44. Depner T A, Chafin ME, Wilson CB, Gulyassy PF: Plasmapheresis for severe Goodpasture's syndrome. Kidney Int 8:409, 1976 (abstr) 45. Lang CH, Brown DC, Staley N, Johnson G, Ma KQ, Border WA, Dalmasso AP: Goodpasture's syndrome treated with immunosuppression and plasma exchange. Arch Intern Med 137:1076-1078, 1977 46. Rosenblatt SG, Knight W, Bannayan GA, Wilson CB, Stein JH: Treatment of Goodpasture's syndrome with plasmapheresis. Am J Med 66:689-696, 1979 47. Walker RG, D'Apice AJF, Becker GJ, Kincaid-Smith P, Craswell PW: Plasmapheresis in Goodpasture's syndrome with renal failure. Med J Aust 1:875-879, 1979 48. McKenzie PE, Taylor AE, Woodroffe AJ, Seymour AE, Chan YL, Clarkson AR: Plasmapheresis in glomerulonephritis. Clin Nephrol 12:97-108, 1979 49. Pohl MA, Lan SP, Berl T, and the Lupus Nephritis Collaborative Study Group: Plasmapheresis does not increase the risk for infection in immunosuppressed patients with severe lupus nephritis. Ann Intern Med 114:924-929, 1991 50. Pinching AJ, Peters OK, Davis IN: Remission of myasthenia gravis following plasma-exchange. Lancet 2:13731376, 1976 51. Dau PC, Lindstrom JM, Cassel CK, Denys EH, Shev EE, Spitter LE: Plasmapheresis and immunosuppressive drug therapy in myasthenia gravis. N Engl J Med 297:1134-1140, 1977 52. Behan PO, Shakir RA, Simpson JA, Burnett AK, Allan TL, Haase G: Plasma-exchange combined with immunosuppressive therapy in myasthenia gravis. Lancet 2:438-440, 1979 53. Newsom-Davis J, Wilson SG, Vincent A, Ward CD: Long-term effects of repeated plasma exchange in myasthenia gravis. Lancet 1:464-468, 1979 54. Winklestein A, Volkin RL, Starz TW, Maxwell NG, Spero JA: The effects of plasma exchange on immunologic factors. Clin Res 27:691, 1979 (abstr) 55. Consensus on IVIG. Lancet 1:470-472, 1990 56. Haas A: Use of intravenous immunoglobulin in immunoregulatory disorders, in Stiehm ER, moderator. Intravenous immunoglobulins as therapeutic agents. Ann Intern Med 107:367-382, 1987 57. Finlayson JS: Albumin products. Semin Thromb Hemost 6:85-120, 1980 58. Kiprov 0, Simpson 0, Romanick-Schmiedl S, Lippert R, Spira T, Busch 0: Risk of AIDS-related virus (human immunodeficiency virus) transmission through apheresis procedures. J Clin Apheresis 3:143-146, 1987 59. Centers for Disease Control: Public health service interagency guidelines for screening donors of blood, plasma, organs, tissues, and semen for evidence of hepatitis B and hepatitis C. Morb Mortal Wkly Rep 40: 1-17, 1991 60. The risk of transfusion-transmitted infection. N Engl J Med 327:419-421, 1992 (editorial) 61. Alter HJ, Prince AM: Transfusion-associated non-A, non-B hepatitis: An assessment of the causative agent and its clinical impact. Transfusion Med Rev 2:288-293, 1988

COMPLICATIONS OF THERAPEUTIC PLASMA EXCHANGE 62. Cohen N, Munoz A, Nelson K, Ness P: Transmission of retroviruses by blood transfusion. N Eng! J Med 321: 1338, 1989 (letter) 63. Ward JW, Bush TJ, Perkins JA, Lieb LE, Allen JR, Goldfinger D, Samson S, Pepkowitz SH, Fernando LP, HoIland PV, Kleinman SH, Grindon AJ, Garner JL, Rutherford GW, Holmberg SD: The natural history oftransfusion-associated infection with human immunodeficiency virus. Factors influencing the rate of progression to disease. N Eng! J Med 321 :947-952, 1989 64. Busch M, Eble B, Heilbron D, Vyas G: Transmission of retroviruses by blood transfusion. N Eng! J Med 322:850851, 1989 (letter) 65. Cumming PD, Wallace EL, Schorr 18, Dodd RY: Exposure of patients to human immunodeficiency virus through the transfusion of blood components that test antibody-negative. N Eng! J Med 321:941-946, 1989 66. Ring J, Messmer K: Incidence and severity of anaphylactoid reactions to colloid volume substitutes. Lancet 1:466469, 1977 67. Bambauer R, Jutzler GA, Albrecht D, Keller HE, Kohler M: Indications of plasmapheresis and selection of different substitution solutions. Biomater Artif Cells Artif Organs 17:9-27,1989 68. Huestis DW: Mortality in therapeutic haemapheresis. Lancet 1: 1043, 1983 (letter) 69. Apter AJ, Kaplan AA: An approach to immunologic reactions with plasma exchange. J Allergy C1in Immunol 90: 119-124, 1992 70. Stafford CT, Lobel SA, Fruge BC, Moffitt JE, HoffRG, Fadel HE: Anaphylaxis to human serum albumin. Ann Allergy 61 :85-88, 1988 71. Orlin JB, Berkman EM: Partial plasma exchange using albumin replacement: Removal and recovery of normal plasma constituents. Blood 56: 1055- 1059, 1980 72. Pearl RG, Rosenthal MH: Metabolic alkalosis due to plasmapheresis. Am J Med 79:391-393, 1985

827 73. Reddi A, Frank 0, DeAngelis B, Jain R, Bashruddin I, Lasker N, Baker H: Vitamin status in patients undergoing sing!e or multiple plasmapheresis. J Am Coli Nutr 6:485-489, 1987 74. Mousson C, Charhon SA, Ammar M, Accominotti M, Rifle G: Aluminum bone deposits in normal renal function patients after long-term treatment by plasma exchange. Int J ArtifOrgans 12:664-667, 1989 75. Milliner DS, Shinaberger JH, Shurman P, Coburn JW: Inadvertent aluminum administration during plasma exchange due to aluminum contamination of albumin replacement solutions. N Engl J Med 312:165-167, 1985 76. Jones JV: The effect of plasmapheresis on therapeutic drugs. Dial Transplant 14:225-226, 1985 77. Stigelman WH, Henry DH, Talbert RL, Townsend RJ: Removal of prednisone and prednisolone by plasma exchange. C1in Pharmacol 3:402-407, 1984 78. Landis E, Pappenheimer JR: Circulation, in Hamilton WF, Dow P (eds): Handbook of Physiology, vol 2. Washington, DC, American Physiologic Society, 1950, pp 972-974 79. MacDonald R, Robinson A: Suxamethonium apnea associated with plasmapheresis. Anaesthesia 35:198-201, 1980 80. Wood GJ, Hall GM: Plasmapheresis and plasma cholinesterase. Br J Anaesth 50:945-948, 1978 81. Bowen RA: Anaesthesia in operations for the relief of hypertension. Anaesthesia 15:3-10, 1960 82. McCaul K, Robinson GD: Suxamethonium "extension" by tetrahydroaminoacrine. Br J Anaesth 34:536-542, 1962 83. Jorstad S: Biocompatibility of different hemodialysis and plasmapheresis membranes. Blood Purif 5: 123-137, 1987 84. Nicholls AJ, Platts MM: Anaphylactoid reactions due to haemodialysis, haemofiltration or membrane plasma separation. BMJ 285:1607-1609, 1982 85. Aeschbacher B, Haeverli A, Nydegger DE: Donor safety and plasma quality in automated plasmapheresis. Vox Sang 57:104-111, 1989