Summary

This article discusses the last five presentations at the Symposium (those of Drs. Hamilton and Adkinson, Butcher and colleagues, Walker and coworkers, Dolovich and colleagues, and Ward and co‐workers) and makes some overall observations regarding hypersensitivity‐like reactions in patients on hemodialysis. Drs. Hamilton and Adkinson provided an excellent review of the basic mechanisms of acute allergic reactions. They postulate that allergic reactions in patients on hemodialysis could be related to any or all of the following mechanisms: (a) an immunoglobulin (1g)E or IgG immune response in previously exposed patients to substances in extracorporeal blood circuits, or from cross‐reactive antibodies in previously unexposed patients; (b) direct activation of mast cells and basophils by cationic polysaccharides and peptides in the extracorporeal circuit; (c) an idiosyncratic mechanism such as can occur with physical exercise, psychological stress, and aspirin intolerance; (d) anaphylatoxin formation secondary to the direct stimulation of complement via the alternative pathway. All nephrologists are aware that leukopenia occurs early in dialysis. This is probably related to activation of complement via the alternative pathway. The complement activation is less with certain membranes and reuse. Nevertheless, most patients do not exhibit an allergic reaction. Thus, footprints in the sand (blood) may have no direct clinical correlation with the clinical findings of allergic reactions. Dr. Butcher and co‐workers Reported on immunological studies of hollow‐fiber dialyzer extracts. Diisocyanates, which are used in the manufacturing process of hemodialyzers, have been known to induce asthma in occupationally exposed workers. These investigators were unable to detect free isocyanates in any of the fiber or dialyzer samples, or in dioxane‐rinsed materials. Cotton dusts have been shown to contain antigenic material. Cotton linters are used to produce hollow fibers for hemodialysis. Dr. Butcher and coworkers suggest that the Lirnulus amebocyte lysate‐reactive material (LAL‐RM) found in the hollow fibers probably has its origin in the cotton linters. The LAL‐RM contained at least four components with antigenic activity. Extended immunization of rabbits was necessary to produce LAL‐RM‐specific antibodies. They assert that further studies are indicated before any conclusions can be made. Dr. Walker and co‐workers described the hemodynamic manifestations of blood‐dialyzer interactions in their sheep model. Blood previously in contact with various membranes produced acute pulmonary hypertension with associated myocardial ischemia and arrythmias. This vascular response preceded maximal hypoxia, coincided with the period of neutropenia, and was dependent on the surface area of the dialyzer and dose of reactant. These hemodynamic responses were minimal with polyacrylonitrile and cellulose acetate membranes. Dr. Dolovich and co‐workers presented data on the role of ethylene oxide in the genesis of the hypersensitivity‐like reaction. The sera from 25 reacting patients were used in a radioallergosorbent test (RAST) for ethylene oxide. The test was positive for all of these sera. Inhibition studies confirmed the specificity of the antibodies to ethylene oxide. Their demonstration of IgE antibodies to a substance to which there is exposure would indicate a risk of allergic reaction. However, they also indicate that some people with IgE antibodies fail to react to these same substances. These investigators demonstrated a decreased responsiveness to histamine and codeine (stimulants of mast cell mediator release) in patients with chronic renal failure. It seems probable that prior exposure to ethylene oxide or chronic exposure at various dose levels could explain some of the divergent clinical results. Dr. Ward and co‐workers presented further information on the role of dialyzer contaminants in the allergic epiphenomena of hemodialysis. They found that Cuphrophan hollow‐fiber dialyzers obtained from industry contained 660 ± 85 mg of watersoluble material. These contaminants included glycerol, carbohydrates, LAL‐RM, particulates, and one unknown. Patients were dialyzed for 6 weeks with each of four dialyzer preparations: (a) Freon washed, prepared according to standard clinical practice; (b) Freon washed, with recirculation of the blood‐side prime through a 0.45‐μm filter for 10 min to remove particulates; (c) Freon washed, with I‐L saline extraction of the bloodside; (d) isopropyl alcohol washed, prepared according to standard clinical practice. They found that dialyzer preparation had no effect on predialysis eosinophil counts or IgE levels. All patients demonstrated similar transient leukopenia and complement activation. They question if either water‐soluble extractables or particulates play any role in the hypersensitivity‐like reaction in patients on hemodialysis. They also suggest that although complement activation may predispose patients to an allergic‐type reaction, a further agent may be required to initiate the clinical features of the reaction. It is apparent that the data on hypersensitivity‐type reactions in patients undergoing hemodialysis are very incomplete and subject to Reporting biases. There are those who would suggest that the incidence of reactions is much higher than the 35 per 1,000,000 dialyzers sold, as Reported by Dr. Villarroe1 of the U.S. Food and Drug Administration (FDA). He can hardly Report data he does not have. It is extremely interesting that his data reveal that 51% of the reactions occurred when a patient was exposed to a particular model of dialyzer for the first time, 14% after either the second or third exposure, 10% during the fourth to tenth exposure, and 25% after a patient was exposed to a given dialyzer for several months. This could suggest gradual sensitization, variable dose(s) of the sensitizing agent(s), or variable patient response. It seems clear from Dr. Villarroel's presentation that the rate of reactions is significantly decreased when industry flushes the hollow‐fiber dialyzers with isopropyl alcohol rather than Freon. Furthermore, he Reported that no reactions have occurred in patients undergoing hemodialysis with a reused dialyzer. There are anecdotal Reports that bleach or peroxide is an important ingredient of the reuse process if reactions are going to be prevented. Currently, it is clinical practice in some hemodialysis centers to reprocess a new dialyzer before using it to treat a patient. This could lead to another data bias in evaluating the epidemiology of hypersensitivity‐like reactions. In the United States a large proportion of dialyzers are of the hollow‐fiber type. The author has never seen a clinically significant allergic‐type reaction on a coil or flat‐plate dialyzer of the magnitude seen with hollow‐fiber dialyzers, although they have been Reported from the United Kingdom. It would seem easier for industry to clean a flat sheet of membrane rather than a hollow fiber. In addition, a potting compound has been introduced into the manufacturing process of hollow‐fiber dialyzers. To the best of the author's knowledge, there is only one flat‐plate dialyzer with potting mix, and in that one there is only minimal blood exposure. It seems clear that ethylene oxide can cause allergic‐type reactions and that there are multiple sources of sensitization even prior to initiation of hemodialysis. Dr. Henne and co‐workers Reported the deaeration and water extraction kinetics of ethylene oxide and its reaction products in hollow‐fiber dialyzers. Deaeration of the membrane is fairly rapid after a storage time of 10 days. It is this author's understanding that industry has a usual quarantine period of 10–14 days. However, deaeration of the potting compound took >60 days. For lesser periods of time, ethylene oxide was still detectable in the milligram range and could not be eliminated via the usual rinsing procedures. After changing the method of sterilization from ethylene oxide to γ‐irradiation, eosinophilia was Reported to improve significantly. Patients with eosinophilia had higher IgE levels and a slightly higher incidence of hypersensitivity‐like reactions. These observations must give us serious concern. Dr. Pearson and colleagues Reported two reactions with γ‐irradiated hollow‐fiber dialyzers. With the experience of allergic‐like reactions in repeated plasma donors, the blood tubing sets, saline lines, fistula needles, and syringes used in hemodialysis could also be a source of ethylene oxide when used with γ‐irradiated dialyzers. The safe limits of ethylene oxide residuals have not been determined, and there is no reliable test for low levels. The slow diffusion of other substances out of the potting compound must also be carefully evaluated. From the FDA data, it appears that ˜60% of the allergic‐like reactions have occurred when the hollow‐fiber dialyzer has been used on a patient after improper rinsing techniques. Industry will be reevaluating its washing recommendations. For example, would it be better to recirculate saline in a dialyzer with negative pressure and then flush with IL saline, rather than the converse? Ultimately, it seems that industry must sell a cleaner dialyzer or else the user will have to clean it better by the methodology currently being employed in reuse procedures. Patients on maintenance hemodialysis experience a lot of chronic phenomena. They are being stimulated continually, and the level of stimulation is constantly changing. With all of these stimuli, it is hard to precisely sort out what is causing the clinical symptoms on a given day. It is probable that hypersensitivity‐like reactions are multifactorial. To accumulate scientific data, it is necessary to develop a systematic approach to evaluating these reactions. A body of experts from the fields of nephrology, immunology, biochemistry, etc., is needed to develop a protocol to be followed when hypersensitivity‐like reactions occur. A central group of experts to perform the more difficult laboratory tests (e.g., RAST) is required. In the absence of such preparations, no good scientific data will be accumulated because nephrologists will need to deal with the medical emergency at hand. Other members of the staff could be trained to obtain the appropriate blood determinations. Patients with low‐level stimulation, such as those that sneeze or have rhinorrhea or conjunctivitis with a new dialyzer, would also be a good group to study. It is apparent that a better Reporting system for hypersensitivity‐like reactions is also needed. The repeated exposure of all of our hemodialysis patients to complement activation must be bad. It can result in decreased phagocytosis and decreased chemotaxis. This is important because infection is a problem in dialysis patients. However, the relationship of this chronic phenomenon to the acute hypersensitivity‐like reaction remains unclear.