Scientific and Medical Aspects of Apheresis: Issues and Evidence 3 ● Scientific and Medical Aspects of Apheresis: Issues and Evidence

3 Scientific and Medical Aspects of Apheresis: Issues and Evidence 3 ● Scientific and Medical Aspects of Apheresis: Issues and Evidence Various types of apheresis procedures have paucity of high-quality research, conclusions been performed on a clinical basis for many years, about the safety, efficacy, and effectiveness of but the number of patients and types of diseases apheresis are necessarily limited, although some treated have risen significantly in the last 5 years. tentative conclusions and directions for treatment This increase is partially due to increased under- can be discerned. standing of the disease and partially due to engi- The present chapter analyzes the methodolog- neering advances in equipment technologies. By ical problems in conducting apheresis research and almost any standard, treatment by apheresis is still examines available evidence of the safety, ef- in relatively early stages of development—there ficacy, and effectiveness of apheresis. Following are no ideal protocols based on a thorough un- a discussion of methodological issues, several derstanding of reasons for its efficacy. Never- major reviews of apheresis research will be sum- theless, there is an increasing flow of clinical data, marized and evaluated. This chapter will further sometimes describing dramatic patient improve- include the findings of a primary literature review ment, supporting the view that apheresis is a and assessment of apheresis in the treatment of rapidly emerging technology with significant three diseases—namely, hemolytic uremic syn- promise (117). Such evidence of treatment effec- drome, acquired Factor-VIII inhibitor, and Guil- tiveness’ is even today, however, often based on lain-Barré syndrome— where preliminary reports unsystematically collected data. Because of the and evidence have been “promising” in utilizing apheresis as a therapeutic approach (57). (A full discussion of these findings can be found in apps. *Ef&tiveness is the health benefit as measured under average conditions of use. Efficacy is the health benefit as measured under con- B, C, and D.) Present and future research direc- trolled conditions such as those in a randomized clinical trial (104). tions for apheresis will be considered last. METHODOLOGICAL ISSUES An assessment of any medical technology de- To be valid, and to permit generalizations to pends, in part, on the development of a strategy be drawn, there must be clarity about what is be- for identifying technologies to be evaluated, and ing tested, what is being compared, which sub- on the development of clear-cut standards for the ject populations are involved in the research, and quality of the evidence that should be considered what is being measured. Operationally, these four (104,147). Proper research methods, as a result, factors refer to treatment design, research design, become essential to the evaluation of a technol- patient selection, and outcomes (102,104). ogy. Careful and systematic investigations are the essential ingredients in establishing that observed Treatment Design effects are due to the medical intervention. Poorly and haphazardly conducted research studies are Treatment design involves the extent to which plagued with problems of validity and general- clarity about the “active ingredients” of the pro- izability, and these same issues continue to hinder cedure being tested can be achieved. Questions attempts to perform assessments based on such to be answered include whether the procedure in- research (85). volves a single treatment, a combination of treat- 25 26 ● Health Technology Case Studies 23: The Safety, Efficacy, and Cost Effectiveness of Therapeutic Apheresis ments, or a combination of treatment and non- Even if standardized protocols could be devel- treatment factors. Often, because apheresis pro- oped, however, it maybe difficult or undesirable cedures involve a complex interplay of many fac- to administer them. This is particularly prob- tors (i.e., are “multivariant”), resulting research lematic if, for research purposes, assignment to is confounded by inability to separate effects one group or another is required. Use of sham (85,117). The extent to which researchers can treatment in control groups, for example, could measure the impact of any one component of the very well cause this group of patients to suffer procedure is limited when all patients receive or some of the side effects of apheresis, raising the have access to multiple components concurrent- ethical question of subjecting them to a potentially ly. Clarity of design is essential to being able to harmful technique. (See the next section, “Safe- attribute outcomes to particular treatments or ty: A Review of the Evidence, ” for a discussion packages of treatments. of the safety and risk issues of apheresis.) Another obvious ethical concern is whether treatment can Because it is an experimental therapy, the use be denied patients in near-fatal, disease states in of apheresis has not been standardized. Protocols which apheresis has served as the treatment of last in various studies have varied considerably. Var- resort. A third issue is the difficulty of setting up iables include type of replacement fluid, patient a controlled trial for some rare autoimmune dis- selection criteria, other medications, extended eases such as Goodpasture’s syndrome, which respirator and intensive care therapy, and inten- strikes only 2 out of 100,000 people in the United sity of plasma exchange (i.e., frequency and States every year (22,34). Even with autoimmune volume exchanged in each treatment). Many dif- diseases of more common occurrence, such as sys- ferent protocols have been used for apheresis, temic lupus erythematosus, presentation of disease even in the treatment of a single disease, so that symptoms can occur with such broad variety that variation in procedures undoubtedly has led to setting up controlled trials for these conditions can variation in results (117). These variations make become equally difficult (49). it difficult if not impossible to achieve some level of comparison between studies. A last treatment design problem has to do with possible placebo effects of the therapy itself. For For example, apheresis is often used as an “ad- example, among the several explanations dis- juvant” or auxiliary therapy to immunosuppres- cussed in the literature for improvement of pa- sive since drug therapy is required to inhibit the tients undergoing apheresis was the possible rebound reaction (see ch. 2). Although apheresis psychotherapeutic effects of such therapy. Few is used as an adjuvant therapy to anti-inflamma- studies have involved double blind protocols (with tory, immunosuppressive, or cytotoxic drugs, this sham apheresis) which are necessary to eliminate fact should not be viewed as a threat to its validi- the possibility of “placebo improvements” (85, ty: any improvement in the course of disease 117,138). would not be attributable to the pharmacological agents alone, but rather to the combined (or synergistic) effects of apheresis and drug therapy. Research Design There could be a validity problem, however, with the application of the treatment when the con- A valid research design, perhaps most impor- comitant drug therapy varies across studies. When tantly, requires systematic comparison. At min- there is differential improvement by type of drug imum, these comparisons involve the same group used, the integrity of the definition of treatment of patients measured before and after treatment; is called into question. Even though treatments optimally, they involve two or more randomly are presented in the literature in a similar fashion, assigned groups tested before and after treatment they may, in fact, operate quite differently. It may (147). The latter design is usually called a true ex- be the case that the combined (or synergistic) ef- periment (25,122) or, in health care research, a fects of apheresis and drug therapy may vary ac- randomized clinical trial (RCT). The advantage cording to the strength of the drug and the fre- of this design, in comparison to nonrandom selec- quency with which it is administered (85). tion design, is that differences in outcomes can Ch. 3—Scientific and Medical Aspects of Apheresis: Issues and Evidence ● 27 be attributed more confidently to the treatment, prognosis, differing remittive drug regimens), the rather than preexisting differences in the sample generalizability of the research findings is limited populations tested (102,104). and selection bias is bound to occur (102,104). Evaluating existing research on apheresis ther- Perhaps the most severe sampling problem in apy poses difficulties in any attempt to draw valid apheresis studies stems from the use of the therapy conclusions. Other than references to prior treat- as a last resort, i.e., for the “worst cases.” Typical- ment regimens, comparative data on treatment ly, apheresis therapy has been initiated when pa- groups are typically not available. The great ma- tients diagnosed with a specific disease do not re- jority of the reported studies are case reports spond to other conventional therapies, including without any concurrent control groups, blinding, drug therapies and other forms of dialysis such randomization, or other techniques used in con- as hemodialysis or peritoneal dialysis. The ap- trolled clinical trials. plication of apheresis in the most severe cases of rheumatoid arthritis with multiple complications, Because of operational and ethical difficulties for example, has

Load more