INTRODUCTION — Dialysis-related amyloidosis (DRA) is a disabling disease characterized by accumulation and tissue deposition of amyloid fibrils consisting of beta2-microglobulin (beta2-m) in the bone, periarticular structures, and viscera of patients with chronic kidney disease (CKD) [1-8]. Beta2-m is a component of the major histocompatibility complex that is present on cell surfaces and is normally cleared by glomerular filtration, with subsequent reabsorption and catabolism in proximal tubules. Clearance declines in patients with reduced kidney function, which leads to plasma accumulation and slow tissue deposition.
The prevalence of DRA has decreased with the use of high-flux biocompatible membranes, which provide better clearance of beta2-m and are less likely to induce reactive inflammation. (See 'Epidemiology and risk factors' below.)
However, even these more modern hemodialysis therapies can be associated with retention of beta2-m. An overview of DRA is presented in this topic review. Other bone diseases and other complications associated with kidney disease are discussed elsewhere.
●(See "Overview of chronic kidney disease-mineral and bone disorder (CKD-MBD)".)
●(See "Management of secondary hyperparathyroidism in adult dialysis patients".)
Other forms of amyloidosis are also discussed elsewhere.
●(See "Overview of amyloidosis".)
OVERVIEW — DRA is characterized by the tissue deposition of beta2-microglobulin (beta2-m) amyloid, particularly in bone, articular cartilage, synovium, muscle, tendons, and ligaments [9-11]. Beta2-m amyloid also deposits in other tissues, especially blood vessels [10,12] and the gastrointestinal tract [12]. In contrast to fragments of immunoglobulin light chains in primary amyloidosis and serum amyloid A in secondary amyloidosis, the amyloid protein in DRA is composed primarily of beta2-m [2,13,14].
Beta2-m amyloid deposits have been identified in almost every organ, except the brain, resulting in the following clinical presentations:
●Carpal tunnel syndrome (CTS)
●Scapulohumeral periarthritis
●Flexor tenosynovitis
●Destructive spondyloarthropathy
●Bone cysts
●Visceral involvement, particularly the gastrointestinal tract
The clinical presentation is discussed below. (See 'Clinical manifestations' below.)
PATHOGENESIS — Tissue deposition of beta2-microglobulin (beta2-m) is due to reduced clearance and, possibly, to increased production. The major underlying cause is the inability of patients with stage 5 chronic kidney disease (CKD) to adequately clear beta2-m, even with modern, high-flux hemodialysis and/or convective therapies. This is because continuous generation of beta2-m far exceeds its removal by these dialysis modalities. As an example, assuming steady beta2-m generation in a 70 kg anuric individual, net yearly beta2-m retention is 111, 97, 77, 53, and 51 grams with low-flux conventional hemodialysis, high-flux dialysis, short daily hemodialysis, nocturnal hemodialysis, and short daily hemofiltration, respectively [7].
In peritoneal dialysis, there is a small, daily removal of beta2-m due to the overall slow rate of convective transport and dialysate flow rate, despite the peritoneal membrane being highly permeable to small proteins. In one study, for example, clearance was significantly higher with high-flux hemodialysis versus peritoneal dialysis (29 versus 6 liters/week per 1.73 m2) [15]. However, residual kidney function is generally higher among patients undergoing peritoneal dialysis compared with hemodialysis; this may increase overall clearance of beta2-m. (See 'Lack of residual kidney function' below.)
In addition to decreased clearance of beta2-m, the dialysis procedure itself (particularly via exposure to bioincompatible membranes) may mildly stimulate intradialytic beta2-m production. This was suggested by in vitro experiments that demonstrated increased beta2-m production by cultured peripheral blood monocytes obtained from hemodialysis patients after dialysis with a cuprophane membrane but not by monocytes from patients dialyzed with the more biocompatible, non-complement-activating polymethylmethacrylate membrane [16].
Two mechanisms may contribute to the bioincompatible membrane-induced stimulation of beta2-m synthesis, including contact of the cells with the membrane and activation of late complement components [17]. It is also possible that dialysate contaminated with endotoxin stimulates the release of beta2-m from leukocytes and monocytes [18].
The role of reactive inflammation is suggested by observations that lesions in which beta2-m amyloid is deposited are associated with a marked influx of activated macrophages expressing cytokines, such as interleukin-1, tumor necrosis factor-alpha, and transforming growth factor-beta [19]. These macrophages appear to be unable to adequately phagocytose deposited beta2-m [20]. Thus, the destructive spondyloarthropathy may be mediated in part by amyloid deposition, an inability to adequately process such deposition, and reactive inflammation.
However, observational data suggest that intradialytic generation of beta2-m is not a significant underlying factor in DRA [21,22]:
●Serum levels of beta2-m are similar or higher in patients treated with peritoneal dialysis as with hemodialysis.
●DRA has developed in patients who have never undergone hemodialysis.
Although sustained increase in serum beta2-m concentration is a prerequisite for formation of beta2-m amyloid fibrils, the exact mechanism of amyloidogenesis in dialysis patients remains unclear. Studies have shown that elevation of circulating beta2-m levels is not the only cause of DRA [23], and other substances, such as glycosaminoglycans, proteoglycans, apolipoprotein E, and serum amyloid P component, may be involved with amyloidogenesis of patients with long-term dialysis treatment [24]. These molecules are thought to stabilize the amyloid fibrils and inhibit their depolymerization. Moreover, at a neutral pH, lysophospholipids and nonesterified fatty acids accelerate amyloid fibril formation from beta2-m monomer and extension of amyloid fibrils in vitro and may enhance the amyloid deposition in vivo [24,25].
Glycosylated beta2-m, a modified microglobulin resulting from the activity of 3-deoxyglucose, has been found in amyloid deposits as advanced glycation end products (AGEs) [26]. Since 3-deoxyglucose is present at increased levels in the serum of uremic and hemodialysis patients [27], the modification of beta2-m may occur more readily in renal failure. The presence of glycosylated beta2-m in amyloid deposits may further enhance the development of these lesions by both stimulating the secretion of cytokines and acting as a chemoattractant and an apoptosis-delaying agent for monocytes [28-30].
The pathogenic effects of glycosylated beta2-m might be prevented by the administration of aminoguanidine, an agent that inhibits advanced glycation. The potential utility of aminoguanidine in DRA was suggested by an in vitro study that incubated beta2-m and D-glucose in the presence or absence of the agent [31]. Aminoguanidine resulted in a significant inhibition of the formation of AGEs on beta2-m. However, aminoguanidine has not been approved for human use by regulatory agencies and is not available to be administered to patients.
Beta2-m may cause bone destruction by directly stimulating formation of osteoclasts [32].
EPIDEMIOLOGY AND RISK FACTORS — The exact prevalence of DRA is unknown since biopsy (which is the conclusive diagnostic test) is rarely performed [33] (see 'Diagnosis' below). The tissue deposition of amyloid that is detected histologically occurs much earlier than any clinical or radiographic manifestation of the illness. A prospective postmortem study found joint amyloid deposition in 21 percent of patients receiving hemodialysis for <2 years, in 50 percent at 4 to 7 years, in 90 percent at 7 to 13 years, and in 100 percent at >13 years [34]. By comparison, some centers have reported the clinical prevalence of the disease as 0 percent at 5 years, which increased to approximately 50 percent at 12 years and to almost 100 percent at 20 years [2]. However, these data were largely obtained during periods when hemodialysis was performed with low-flux, cellulose-derived dialysis membranes that were impermeable to beta2-microglobulin (beta2-m).
Globally, hemodialysis using low-flux dialysis membranes is being replaced by therapies that have a higher clearance of beta2-m, such as hemodialysis using high-flux dialyzer membranes or convective therapies [7]. This has resulted in a decline in the prevalence of DRA. As an example, in one study of over 200,000 patients from Japan, the incidence of new carpal tunnel syndrome ([CTS] a surrogate for DRA) declined by almost one-half between 1998 and 2010 [35]. However, DRA still remains prevalent, ranging from 21 to 28 percent in two studies, especially among patients with a long dialysis vintage [36-38].
The incidence of DRA in patients on peritoneal dialysis is even less clear since there are few patients who have been dialyzed for prolonged periods. Some observations suggest that peritoneal dialysis is associated with a similar risk of developing DRA as with hemodialysis, a finding that would be expected since there is limited removal of beta2-m by peritoneal dialysis [15].
Risk factors for DRA include the following [39,40]:
●Increasing age and dialysis vintage
●Use of low-flux dialysis membranes (see 'Low-flux dialysis membrane' below)
●Use of bioincompatible dialysis membrane (see 'Bioincompatible dialysis membrane' below)
●Lack of residual kidney function (see 'Lack of residual kidney function' below)
Age and dialysis vintage — Age and dialysis vintage are closely associated with DRA [40]. Virtually all studies have demonstrated increasing prevalence with time on dialysis [2,7,37].
Low-flux dialysis membrane — DRA is more common among patients who undergo low-flux dialysis compared with high-flux dialysis [7,16,41,42]. This is suggested by the following studies:
●One multicenter study that included 221 patients who were receiving hemodialysis for more than five years found that, among patients >60 years of age at the start of dialysis, the likelihood of developing amyloid bone disease was five times higher with low-flux membranes compared with high-flux membranes (figure 1) [41].
●In a retrospective study, clinically evident DRA was compared among 89 patients with stage 5 chronic kidney disease (CKD) who were maintained on regular hemodialysis for at least 10 years and treated exclusively with low-flux, bioincompatible cellulose membranes; low-flux, intermediately biocompatible polysulfone (PMMA) membranes; or high-flux, highly biocompatible polysulphone (AN69) membranes [42]. Clinical symptoms of DRA were most pronounced in the low-flux, bioincompatible cellulose membrane group and least pronounced in the high-flux, highly biocompatible membrane group.
●One report compared patients treated with a polyamide high-flux membrane with those treated with low-flux dialyzers; those on the high-flux membrane had lower beta2-m concentrations [43].
●A cross-sectional study of 147 patients who had been undergoing hemodialysis for >10 years and had a confirmed diagnosis of CTS (which is one of the two most common manifestations of DRA) reported that the combined use of high-flux dialyzer membranes and ultrapure dialysis fluid resulted in delayed onset of CTS [37].
The mechanisms by which high-flux dialysis membranes decrease DRA include improved dialysis clearance of beta2-m [44-49] and, possibly, better preservation of residual kidney function, which improves overall clearance [37,50,51]. Clearance of beta2-m by low-flux dialysis membranes is poor because the molecular weight of beta2-m (11,800 Da) is greater than the cut-off level of the membrane porosity. The more permeable membranes have an intrinsically greater rate of convective transport (although diffusion is limited by the size of the beta2-m molecule) and also may bind beta2-m directly [44-49].
Bioincompatible dialysis membrane — DRA appears to be more common among patients who are dialyzed using a bioincompatible dialysis membrane than among those who are dialyzed using a highly biocompatible membrane [41,42,52]. As examples:
●In one study (cited above), in which DRA was compared among patients with stage 5 CKD who used low-flux, bioincompatible cellulose membranes and low-flux, intermediately biocompatible polysulphone (PMMA) membranes versus high-flux, highly biocompatible polysulphone (AN69) membranes, the biocompatibility of the dialysis membrane appeared to be an independent determinant of the risk for developing DRA [42].
●Among 159 new hemodialysis patients who were prospectively randomly assigned to either a low-flux biocompatible or low-flux bioincompatible membrane, although plasma beta2-m levels rose over time in both groups (since the rate of production remains greater than the rate of removal), the increase above baseline was less among patients treated with biocompatible membranes compared with those treated with bioincompatible membranes (2.6 versus 11.1 mg/L) [52]. This beneficial effect was independent of the influence of residual kidney function and increased progressively over time.
However, in some studies, the prevalence of DRA was not higher among patients treated with cuprophane hemodialysis membranes compared with those treated with AN69 membranes [53,54].
Lack of residual kidney function — The persistence of even a minimal degree of residual kidney function is sufficient to increase clearance and catabolism of beta2-m, thereby protecting against the development of DRA. In one study, the plasma beta2-m concentrations were twice as high in hemodialysis patients with a glomerular filtration rate (GFR) <1 mL/min than in those with a GFR of 4 to 5 mL/min [55]. As noted above, some studies have shown that high-flux membranes or ultrapure dialysis fluid preserve residual kidney function better than low-flux cellulosic membranes and facilitate better overall clearance of beta2-m [37,50,51].
CLINICAL MANIFESTATIONS
Symptoms — Patients with DRA most commonly complain of shoulder pain related to scapulohumeral periarthritis and rotator cuff infiltration by amyloid and of symptoms of carpal tunnel syndrome (CTS) [1-3,56].
Shoulder pain typically is localized to the anterolateral aspect and is usually bilateral. Abduction may elicit pain, and patients may have limited range of motion, which worsens when patients are in a supine position, particularly at night or when undergoing dialysis treatments. This pain often improves when the patient sits or stands up [33,57].
Beta2-microglobulin (beta2-m) amyloid deposition in the carpal tunnel causes hand pain and numbness and dysesthesias in the distribution of the median nerve (see "Carpal tunnel syndrome: Clinical manifestations and diagnosis", section on 'Clinical features'). CTS symptoms tend to occur more commonly on the side of the longest-functioning vascular access, but bilateral symptoms can occur among patients who have only one vascular access [33,58]. Carpal tunnel symptoms may worsen during dialysis due in part to steal syndrome [33,59].
Patients may experience neck pain related to involvement of the cervical spine by a destructive spondyloarthropathy [60]. The loss of the intervertebral disk space with erosion of the adjacent vertebral endplates may result in this aseptic process being confused with infectious diskitis. However, patients who have radiographic evidence of cervical spine involvement usually experience no related symptoms [61]. Very infrequently, in patients who have undergone dialysis for 20 years or longer, epidural deposition of beta2-m amyloid may compress the spinal cord, causing quadriparesis or quadriplegia [62-65].
The bone cysts of beta2-m amyloidosis may be distinguished from the brown tumors of hyperparathyroidism by their rapid rate of enlargement. These occur most commonly in the carpal bones but may also occur in the femoral neck, phalanges of the hands, humeral head, acetabulum, tibial plateau, and distal radius. Patients with bone cysts may present with pathologic fractures, particularly of the femoral neck [66].
Symptoms of visceral involvement are uncommon, but those that do occur are most often related to gastrointestinal involvement [33]. The colon may be most frequently involved [67-69]. Other sites of involvement include the tongue, esophagus, stomach, and small intestine. Patients may describe difficulty macroglossia, difficulty swallowing, or pain with swallowing [70,71]. Gastrointestinal bleeding, ischemic or infarcted bowel with perforation, and pseudo-obstruction with gastric or colonic dilation have all been described [72-78].
Cardiac, pulmonary, and cutaneous involvement with DRA have been described infrequently [56,70,72,79]. Cardiac deposition of beta2-m amyloid may cause congestive heart failure or mitral regurgitation. Beta2-m amyloid deposition in rectal mucosa, liver, spleen, and blood vessels is uncommon.
Physical examination — Careful physical examination of the patient may suggest the diagnosis of DRA. The shoulders of patients with significant scapulohumeral periarthritis appear hypertrophied because of thickening of and/or deposition of amyloid between muscles and tendons of the rotator cuff ("shoulder pad sign") (picture 1) [80,81]. The coracoacromial ligament and bicipital tendon may be tender to palpation.
CTS is suggested by weakness and possibly atrophy of the muscles of the thenar eminence and by decreased sensation in the thumb and index, middle, and ring fingers. These sensory deficits typically spare the thenar eminence. (See "Carpal tunnel syndrome: Clinical manifestations and diagnosis", section on 'Examination'.)
Irreducible flexion contractures of the fingers may be present among patients who have amyloid deposition along flexor tendons [81]. Often, these amyloid deposits result in soft tissue fullness in the palm; when the fingers are extended, the flexor tendons become prominent, creating a pathognomonic "guitar string" sign [82]. The "guitar string" sign in a patient with either shoulder pain or evidence of CTS strongly suggests DRA. Flexion contractures of the fingers together with atrophy of the thenar eminence have also been called "amyloid hand" (picture 1) [2].
The plasma beta2-m concentration is increased among patients with stage 5 chronic kidney disease (CKD) treated with dialysis, with concentrations ranging from 30 to 50 mg/L, much higher than the normal values between 0.8 and 3 mg/L [83]. In the absence of characteristic clinical features, elevated beta2-m levels alone do not establish the diagnosis, since they may be seen in patients without (and/or prior to) DRA. However, mean cumulative predialysis serum beta2-m levels were shown to be associated with all-cause mortality in a secondary analysis of the Hemodialysis (HEMO) Study such that optimal survival was in patients with beta2-m levels <27.5 mg/L [84].
DIAGNOSIS — DRA is usually suspected in the dialysis patient with characteristic clinical and/or radiographic features. However, tissue deposition of beta2-microglobulin (beta 2-m) amyloid occurs years before clinical manifestations are evident [34].
As stated in the Kidney Disease Outcomes Quality Initiative (KDOQI) guidelines, biopsy remains the "gold standard" for the diagnosis of beta2-m amyloidosis [5]. The amyloid found in the bone cysts and synovial tissue is similar to other forms of amyloid in its staining properties, with Congo red, and in exhibiting apple-green birefringence under polarized light. In contrast to fragments of immunoglobulin light chains in primary amyloidosis and serum amyloid A in secondary amyloidosis, the amyloid protein in DRA is composed primarily of beta2-m [2,13,14]. (See "Overview of amyloidosis".)
However, because biopsy is rarely performed, the diagnosis in suspected dialysis patients is usually clinical and relies upon the combination of typical clinical features and characteristic radiographic findings of multiple, rapidly enlarging bone cysts that enlarge over time (image 1) [2,7]. Measurement of serum beta2-m is not helpful for the diagnosis of DRA, since levels are elevated among dialysis patients in the absence of DRA [83,85].
The following are some of the imaging modalities used and the characteristic findings with these modalities in DRA [86,87]:
●Conventional radiography, which often is used for screening or initial diagnosis, reveals radiolucent lesions or bone cysts, frequently with thin sclerotic margins. The lesions tend to enlarge more rapidly than the brown tumors of hyperparathyroidism and usually increase in number over time (image 2 and image 1).
●Computed tomography (CT) and magnetic resonance imaging (MRI) are particularly useful for detection of lesions too small to be seen with plain radiographs and for lesions of the axial skeleton [86,87]. One study found supraspinatus and/or subscapularis tendon thickening in chronic dialysis recipients, as determined by MRI [88]. MRI and CT may also be useful for detection of soft tissue deposits. However, the administration of gadolinium during MRI caused an often severe and disabling disease called nephrogenic systemic fibrosis (NSF) among patients with moderate to severe kidney disease, particularly those requiring dialysis. As a result, imaging with gadolinium-based contrast agents should be avoided, if possible, in dialysis patients. (See "Nephrogenic systemic fibrosis/nephrogenic fibrosing dermopathy in advanced kidney disease".)
●Ultrasonography may be of use in DRA [80,89-91]. It typically demonstrates increased rotator cuff thickness, deposits with increased echogenicity between the muscles and tendons of the rotator cuff [80], and a thickened synovial sheath of the long head of the biceps [89].
●Scintigraphy with radiolabeled beta2-m or serum amyloid P component has been performed but has not been consistently useful as the results depend upon which specific protein is radiolabeled [92-95].
●Positron emission tomography with fluorodeoxyglucose (PET-FDG) may aid in the diagnosis of DRA by detecting inflammation at characteristic sites of beta2-m amyloid deposition [96].
Abdominal fat pad aspiration is not useful to diagnose DRA.
DIFFERENTIAL DIAGNOSIS — Other conditions, besides beta2-m amyloidosis, may cause shoulder pain in patients with chronic kidney disease (CKD), especially in those who are younger than 40 years or who have received dialysis treatment for fewer than five years:
●Calcific periarthritis, caused by hydroxyapatite crystal deposition in periarticular tissue, may cause shoulder pain with associated soft tissue swelling, especially when the calcium x phosphorus product is >75 mg2/dL2. The presence of radiopaque periarticular calcifications on plain radiographs differentiates calcific periarthritis from the shoulder periarthritis observed in beta2-m amyloidosis. (See "Basic calcium phosphate (BCP) crystal-associated calcific periarthritis (tendinopathy)".)
●Subacromial bursitis, bicipital tendinitis, and supraspinatus tendinitis may develop with overuse of the shoulder. These conditions often present with severe shoulder pain at rest that is exacerbated by movement of the affected shoulder and may awaken the individual from sleep. Reflex sympathetic dystrophy syndrome also may present with shoulder pain and limited motion, often associated with or followed by pain, swelling, and stiffness of the ipsilateral hand and wrist with signs and symptoms of vasomotor instability. However, in these conditions, the shoulder does not appear hypertrophied with the appearance of a "shoulder pad."
•(See "Biceps tendinopathy and tendon rupture".)
•(See "Biceps tendinopathy and tendon rupture".)
●Septic arthritis of the shoulder, most commonly caused by Staphylococcus aureus in patients receiving hemodialysis, typically causes severe shoulder pain at rest or upon any shoulder movement. This must be diagnosed by shoulder arthrocentesis performed under fluoroscopic guidance with culture of the aspirated synovial fluid.(See "Septic arthritis in adults".)
●Patients with CKD and renal osteodystrophy may develop shoulder pain due to pathological humeral fractures, which can be diagnosed on plain radiographs. (See "Evaluation of renal osteodystrophy".)
Carpal tunnel syndrome also may be caused by conditions other than beta2-m amyloidosis in patients with CKD. Patients with peripheral neuropathy may experience symptoms of carpal tunnel syndrome. Carpal tunnel syndrome symptoms may be induced by prolonged immobility of the wrist in other than a neutral position, such as during hemodialysis treatments, causing median nerve compression. Median nerve ischemia in the arm where vascular access has been established may be exacerbated by a fistula-induced arterial steal phenomenon during hemodialysis treatment. (See "Carpal tunnel syndrome: Clinical manifestations and diagnosis".)
Shoulder pain and carpal tunnel syndrome may be associated with the underlying disease process that resulted in renal failure, such as the scapulohumeral periarthritis and carpal tunnel syndrome that occurs in diabetic cheiroarthropathy. When renal failure has resulted from primary (AL) amyloidosis, shoulder periarthritis and carpal tunnel syndrome may develop and can be distinguished from those of beta2-m amyloidosis only by histologic identification of the amyloid subunit protein in tissue. Patients with systemic lupus erythematosus may develop shoulder joint inflammation or avascular necrosis of the humeral head causing shoulder pain and also may develop carpal tunnel syndrome due to median nerve compression by wrists synovitis. Patients with scleroderma may experience pain and restricted shoulder motion due to fibrosis of the periarticular skin [97].
TREATMENT
Overview — There is no specific medical treatment for DRA. However, removal of significant amounts of beta2-microglobulin (beta2-m) may prevent or slow progression of the disease. This can be accomplished best by kidney transplantation. However, hemodialysis with a biocompatible high-flux dialysis membrane, hemodiafiltration, the use of ultrapure dialysate, or a beta2-m adsorbent column may also result in lowering the level of beta2-m.
Kidney transplantation is considered the definitive treatment for DRA in patients with end-stage kidney disease (ESKD). Thus, for patients with DRA who are appropriate candidates, we suggest kidney transplantation since it provides the most effective reduction of beta2-m levels (see "Kidney transplantation in adults: Patient survival after kidney transplantation"). Successful kidney transplantation reduces plasma beta2-m levels to normal, and joint pains usually resolve soon after the renal allograft has begun to function [5,98-101]. Over time, after transplantation, the amyloid deposits may also regress. One study, for example, found regression of articular amyloid deposition in eight of nine patients with DRA, as detected by scanning with radiolabeled serum amyloid P component approximately five years after successful kidney transplantation [101]. By contrast, bone cysts resolve slowly after transplantation, and the destructive spondyloarthropathy continues to progress; symptoms generally recur quickly in patients who restart dialysis after renal graft failure [100-102].
However, many ESKD patients are not appropriate candidates for transplantation, because of comorbid conditions or advanced age. In addition, many patients who are appropriate candidates for transplantation do not have an available living donor and must wait for months to years for a deceased-donor kidney. For such patients, dialytic therapy should be optimized to provide maximum reduction of beta2-m. The effective removal of beta2-m by dialytic therapy requires the use of newer and more permeable dialysis membranes (for patients on hemodialysis) since cuprophan membranes are impermeable to beta2-m.
Moreover, the dialysis modality, duration, and frequency also influence the clearance of beta2-m.
There are few studies examining the relative efficacy of specific therapies in patients with established DRA. Dialytic clearance of beta2-m is dependent upon membrane pore size. Cellulose membranes are impermeable to beta2-m while its clearance is increased with high-flux membranes and convective transport techniques. Recommendations for patients on hemodialysis and peritoneal dialysis are discussed below.
Hemodialysis — For all hemodialysis patients with DRA, we use a highly biocompatible, high-flux membrane. This recommendation is consistent with 2003 Kidney Disease Outcomes Quality Initiative (KDOQI) guidelines [5]. Most [7,16,37,41,42,44-51], but not all [53,54,103] studies have found that beta2-m level is lower among patients dialyzed with high-flux biocompatible membranes. (See 'Low-flux dialysis membrane' above and 'Bioincompatible dialysis membrane' above.)
There is limited evidence that the use of super-flux membranes may improve beta2-m clearance even more than with high-flux membranes [104]. However, there are insufficient data upon which to recommend their use at this time.
It should be recognized that, despite substantial clearance, there is still substantial retention of beta2-m with conventional three times weekly high-flux hemodialysis [7,21]. This is because slow, intercompartmental mass transfer limits beta2-m removal and a 30 percent rebound in its plasma concentrations occurs within 90 minutes after a four-hour post-dilution hemodiafiltration [105]. Dialysis duration and frequency are important determinants of beta2-m clearance [106]. Among hemodialysis patients with DRA, we increase the dialysis duration and/or frequency. Increasing the weekly treatment time decreases beta2-m concentrations [106].
Nocturnal or short daily hemodialysis may be better than conventional three times weekly hemodialysis for reducing beta2-m. Nocturnal hemodialysis (performed for eight hours, six nights per week, at low blood- and dialysate-flow rates) is associated with significantly higher clearances of beta2-m (585 versus 127 mg/week) and a greater reduction in plasma beta2-m concentrations (39 versus 21 percent) compared with conventional hemodialysis (performed for four hours, three times per week) [107]. (See "Technical aspects of nocturnal hemodialysis".)
Short daily hemodialysis also appears to be associated with enhanced middle-molecule removal compared with that observed with conventional hemodialysis. (See "Short daily hemodialysis".)
The role of other dialytic therapies, such as hemofiltration and hemodiafiltration, for the prevention and/or treatment of DRA is unclear. Significant removal of beta2-m occurs with hemofiltration and hemodiafiltration, although plasma beta2-m levels remain high with these modalities [108-112]. Some [113-115], though not all [116], studies have suggested that the development of clinically significant carpal tunnel syndrome (CTS) occurs less frequently in patients treated with hemodiafiltration and hemofiltration. On the other hand, some clinicians feel that hemofiltration and hemodiafiltration do not provide a clear benefit versus that associated with biocompatible high-flux hemodialysis performed at high blood flow rates [86]. Until more data are available, we do not recommend such therapies solely for the treatment of DRA.
Peritoneal dialysis — Although the peritoneal membrane is highly permeable to small proteins and has the highest biocompatibility, only a small amount of beta2-m is removed daily with peritoneal dialysis because of slow convective transport and dialysate flow rate. Clearance of beta2-m in peritoneal dialysis patients depends mainly upon residual kidney function; thus, increasing peritoneal dialysis dose does not necessarily compensate for declining residual kidney function [117]. Some studies suggested that the incidence of DRA is equal in hemodialysis and peritoneal dialysis patients, while others suggest that its incidence and progression are greater in hemodialysis patients [93,118]. For example, an earlier study compared the incidence of CTS in populations of patients undergoing continuous ambulatory peritoneal dialysis (CAPD) with that in patients receiving hemodialysis populations. One-hundred fifty-one patients (90 patients on hemodialysis and 61 patients on CAPD) were evaluated by questionnaire, physical examination, and nerve conduction studies. Eight of 57 CAPD and 15 of 83 hemodialysis patients had CTS. The authors concluded that CTS occurs with equal frequency in CAPD and hemodialysis populations [119]. However, a more recent study found significantly higher clearance of beta2-m with high-flux hemodialysis than with peritoneal dialysis (29 versus 6 liters/week per 1.73 m2) [34]. It is important to point out that such comparative studies have been limited by the shorter average duration of peritoneal dialysis. Apart from kidney transplantation, the most definitive treatment for DRA, there are insufficient data to recommend a change in the dialysis modality among patients suffering from DRA.
Other therapies — Limited data suggest that use of a beta2-m adsorption column, which is available for clinical use in Japan, Europe, and the United States, in patients with DRA is associated with increased beta2-m removal and better improvement in patients' symptoms versus dialysis alone in those with DRA [120-123]. A nationwide survey of 345 Japanese patients dialyzed at 138 institutions who used this column found that use of beta2-m column was associated with improvement of DRA symptoms in over 85 percent of patients [124]. However, hypotension and anemia have limited use of this treatment in some patients, and its role remains undefined.
Doxycycline has been shown to modulate the formation of beta2-m fibrils in vitro [125]. In one report, among three long-term dialysis patients with severe DRA-induced articular impairment and pain, treatment with doxycycline 100 mg daily was associated with a reduction in joint pain and improvement in passive and active range of motion, despite no change in the amyloid deposits [126].
Surgery and analgesia — Apart from enhancing beta2-m clearance by dialysis or kidney transplantation, treatment of DRA is otherwise palliative. The most debilitating aspects of DRA are CTS, paraparesis due to epidural beta2-m amyloid deposition, and the propensity to pathologic fractures from bone cysts in anatomic sites of significant weight bearing (such as the femoral neck). Analgesics help with periarticular and bone pain.
Since DRA is a progressive disease, early surgical correction of CTS is warranted [127]. No data have been published comparing the outcome of surgical procedures to treat DRA with that of the same procedures in patients without DRA. However, based upon clinical experience, we suggest that the following surgical interventions may be beneficial:
●Arthroscopic or open surgery of the shoulder with removal of synovium infiltrated by amyloid often provides dramatic pain relief [128].
●Carpal tunnel surgery should include debridement of the hypertrophied synovium that is infiltrated by beta2-m amyloid, rather than just transection of the transverse carpal ligament, to more effectively relieve median nerve compression. However, despite surgery, CTS typically recurs within two years and requires multiple repeat surgical decompression procedures over time [33].
●Curettage and bone grafting of amyloid cysts in the femoral neck has been successful in relieving hip pain [129] and may prevent pathologic fractures. The grafts were successfully incorporated into the bone defects.
●Replacement of a diseased joint with a prosthesis must be considered on an individual basis; when performed, this modality can relieve pain and restore lost mobility [33]. Pathologic fractures of the femoral neck, occurring in bone compromised by beta 2-m amyloid deposition, should be treated with total joint arthroplasty, rather than by internal fixation, because of the poor quality of the bone.
SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Dialysis".)
SUMMARY AND RECOMMENDATIONS
●Among patients with stage 5 chronic kidney disease (CKD), dialysis-related amyloidosis (DRA) is a disorder caused by the inability to clear beta2-microglobulin (beta2-m), resulting in the deposition of beta2-m amyloid fibrils in bones, joints, and other soft tissues. The incidence of DRA is now much lower than had been reported previously, a trend that correlates with the increased use of high-flux biocompatible dialyzers with enhanced clearance of beta2-m. (See 'Introduction' above and 'Overview' above.)
●The major clinical manifestations of DRA are the characteristic triad of scapulohumeral periarthritis, carpal tunnel syndrome (CTS), and flexor tenosynovitis of the hand; rapidly enlarging bone cysts; destructive spondyloarthropathy; and pathologic fractures. The amyloid found in involved tissue is similar to other forms of amyloid in its staining properties, with Congo red, and in exhibiting apple-green birefringence under polarized light. Plain radiography often reveals bone cysts, especially near joints. The diagnosis of DRA is based upon the presence of typical clinical features with characteristic imaging findings. (See 'Clinical manifestations' above and 'Diagnosis' above.)
●Once DRA has developed, patients should be dialyzed with biocompatible, high-flux hemodialysis membranes, which are the standard of therapy. In addition, altering the duration, frequency, and/or type of dialysis treatment may provide some benefit. Kidney transplantation is the only modality that appears to slow or halt the progression of DRA. In patients with CTS, early surgical correction is warranted. Surgical treatment of diseased bones and joints with modalities such as open surgery, curettage, and total joint arthroplasty may result in dramatic pain relief, particularly in cases where debility or pain cannot be controlled with analgesics or other nonsurgical modalities. Although a beta 2-m adsorption column appeared to improve symptoms of DRA in Japanese dialysis patients, randomized, controlled trials are needed to substantiate the benefit of this modality. (See 'Treatment' above.)
ACKNOWLEDGMENTS — The UpToDate editorial staff acknowledges Robert E Cronin, MD, and William L Henrich, MD, MACP, who contributed to earlier versions of this topic review.
