Treatment of AA (secondary) amyloidosis

INTRODUCTION — AA (secondary) amyloidosis is a disorder characterized by the extracellular tissue deposition of fibrils composed of fragments of serum amyloid A protein (SAA), an acute phase reactant. AA amyloidosis may complicate a number of chronic inflammatory conditions, including rheumatoid arthritis (RA), juvenile idiopathic arthritis, ankylosing spondylitis (AS), inflammatory bowel disease, familial periodic fever syndromes, chronic infections, and certain neoplasms (table 1) [1,2].

The treatment of AA amyloidosis is presented here. The pathogenesis, clinical manifestations, major causes, and approach to diagnosis of AA amyloidosis are presented separately, as are the clinical manifestations and treatment of the musculoskeletal and renal manifestations of amyloid diseases. (See "Pathogenesis of AA amyloidosis" and "Overview of amyloidosis", section on 'Clinical manifestations' and "Causes and diagnosis of AA amyloidosis and relation to rheumatic diseases" and "Musculoskeletal manifestations of amyloidosis" and "Renal amyloidosis".)

PROGNOSIS — If untreated, AA amyloidosis is a serious disease with a significant mortality due to end-stage kidney disease, infection, heart failure, bowel perforation, or gastrointestinal bleeding [3-5]. Patients with persistently high circulating levels of serum amyloid A protein (SAA) are at particular risk of these complications [6,7]. A progressive improvement in survival rates has been reported among patients with AA amyloidosis, likely reflecting, in part, improved treatment strategies for associated inflammatory disorders, as well as earlier detection [2,3,8,9].

In a review of 374 cases of AA amyloidosis, for example, median survival from diagnosis was 133 months, with older age, reduced serum albumin concentration, end-stage renal failure at baseline, and the degree by which the SAA concentration was elevated during follow-up being of poorer prognostic significance [3].

In a review of 48 cases of AA amyloidosis undergoing kidney transplantation, recurrence rate in those with persistent SAA elevations was 3 times those with the lowest mean SAA levels achieved by suppression post-transplant [10].

Importance of control of the underlying disease — The fibril precursor in AA amyloid, SAA, is a normal plasma protein that is produced by hepatocytes as part of the physiologic acute phase response. A chronic inflammatory state leads to sustained high levels of the acute phase proteins [11]. Successful treatment of the underlying inflammatory process by, for example, surgical resection of the focus of infection or the tumor, cytotoxic agents and/or biologic agents in rheumatoid arthritis (RA), or antibiotics with chronic infection, results in reduced hepatic production of the acute phase response and a fall in circulating SAA down to normal healthy levels. Over time, this can lead to stabilization of or improvement in renal function, to reduction in urinary protein excretion, and to partial resolution of amyloid deposits (as assessed in studies by scintigraphy with radiolabeled serum amyloid P component [SAP]) [12-18]. This was illustrated in a report of 80 patients with AA amyloid, mostly due to juvenile idiopathic arthritis or to RA, who were prospectively followed for a median of four years; in this series, the systemic amyloid load was assessed by yearly SAP scintigraphy [19]. The following findings were noted:

●Forty-two patients had median serum SAA concentration within the reference range (<10 mg/L); amyloid deposits regressed in 25 and stabilized in 14. Among patients with renal disease at baseline, proteinuria typically fell, while the serum creatinine concentration was either stable or improved. Two patients who required dialysis remained on dialysis.

●The outcomes were variable in the other patients. However, among those in whom the serum SAA was persistently above 50 mg/L, the amyloid load usually increased, and organ function deteriorated. In one patient who underwent kidney transplantation, proteinuria and renal amyloid deposits recurred within 36 months.

●There was a variable relation between changes in amyloid load and changes in organ function. In some patients, for example, renal function was stable despite a reduction in amyloid load, while others had improved renal function despite a stable amyloid load.

●The estimated 10-year survival was much higher in the patients with median serum SAA below 10 mg/L (90 versus 40 percent in those with higher median values), with 60 percent regression as assessed by I-123-labeled SAP scintigraphy, if this level was maintained [2,3,19].

Similar findings have been noted in other uncontrolled series [12,13,16-18,20-22]. As examples:

●In a 10- to 20-year follow-up of 16 adults given alkylating agents because of AA amyloid due to rheumatic disease (RA, ankylosing spondylitis, or juvenile chronic polyarthritis), treated patients had substantial preservation of renal function compared with 48 untreated controls [12].

●In three patients with Crohn disease who underwent kidney transplantation because of renal amyloidosis, immunosuppressive therapy led to remission of the inflammatory bowel disease [13]. Serial SAP scintigraphy demonstrated regression of amyloid deposits in two patients and stabilization of amyloid in the third.

●In four patients with systemic AA amyloidosis complicating Castleman disease (angiofollicular lymph node hyperplasia), resection of the tumors may have led to clinical regression of the amyloid, which was also corroborated by SAP scintigraphy [15,23].

TREATMENT OF AA AMYLOIDOSIS — As illustrated by the preceding observations, the preferred therapy of AA amyloidosis is control of the underlying inflammatory disease and thus near complete suppression of serum amyloid A protein (SAA) production. Treatment needs to be specific for the underlying disease; for example, colchicine is the accepted therapy for AA amyloidosis complicating familial Mediterranean fever (FMF), and agents active against proinflammatory cytokines (interleukin [IL] 1 beta, tumor necrosis factor [TNF] alpha, IL-6) have been shown to have efficacy in some cases of AA amyloidosis due to rheumatic disorders and to hereditary autoinflammatory disease. Eprodisate, a glycosaminoglycan mimetic that was developed to interfere with fibril formation in AA amyloidosis, was withdrawn from clinical development after phase 3 studies failed to show benefit. Agents that enhance clearance of amyloid from tissue are undergoing clinical studies. The efficacy of dialysis and kidney transplantation in those patients who progress to end-stage kidney disease is discussed separately. (See "Renal amyloidosis".)

Colchicine — Colchicine has become accepted prophylactic therapy to prevent AA amyloidosis in FMF, where doses of 0.6 to 1.2 mg/day in adults markedly reduced the frequency of attacks of abdominal pain, diminished the incidence of clinical renal disease (including prevention of recurrence in the kidney transplant), and stabilized the glomerular filtration rate in patients with mild proteinuria. In the large series reported from Tel Hashomer Hospital in Israel, the rate of development of proteinuria at 9 to 11 years was 1.7 in 960 compliers versus 49 percent in 54 non-compliers [24]. In 86 patients who already had nonnephrotic proteinuria prior to therapy, colchicine led to resolution of proteinuria in five and to stabilization in 68. (See "Management of familial Mediterranean fever".)

Among FMF patients with nephrotic syndrome due to AA amyloid, prevention of disease progression and a reduction in protein excretion can be achieved. However, a higher colchicine dose of 1.5 to 2.0 mg/day appears to be required, and therapy should be instituted before the plasma creatinine concentration reaches 1.5 mg/dL (132 micromol/L) [25,26]. Protein excretion in responders can be reduced to below 1 to 2 grams/day with this regimen; the benefit is gradual, occurring over a one- to two-year period [25]. A small number of patients do, however, appear to be resistant to the effects of colchicine, possibly reflecting genetic subsets of underlying disease or variable pharmacokinetics of the drug and anti-IL-1 therapies are the second-line treatment of choice [27,28].

Colchicine is not likely to be effective in salvaging renal function in FMF patients who already have chronic renal failure, since irreversible glomerular injury is probably present. However, it can prevent recurrent disease in the transplant and prevent progressive amyloid deposition in other sites [26]. The optimal colchicine dose in this setting is 1.5 to 2.0 mg/day; lower doses are less predictably effective. (See "Renal amyloidosis".)

Anecdotal successes with chronic colchicine therapy (0.6 mg twice daily) have also been reported in patients with AA amyloid due to inflammatory bowel disease, in patients with Behçet syndrome, and in intravenous drug users with suppurative skin lesions. Proteinuria due to renal amyloidosis may be markedly reduced, and renal function may remain stable over long periods of time [29-32]. However, repeat renal biopsy in one such patient did not demonstrate a reduction in total amyloid deposition despite almost complete resolution of proteinuria [31].

Dimethylsulfoxide — Dimethylsulfoxide (DMSO) may have anti-amyloid activity in several forms of the disorder; it has been shown to be active in murine models of AA amyloidosis and in case reports of patients with AA amyloid due to rheumatoid arthritis (RA) or to Crohn disease [33-35]. Use of DMSO is limited, however, by its acrid odor and, because of its potent solvent activity, by the difficulty attaining purity for pharmaceutical usage [36].

Cytotoxic and immunosuppressive agents — Azathioprine, chlorambucil, methotrexate, and cyclophosphamide have been shown to be helpful in AA amyloidosis complicating treatment-responsive inflammatory diseases in individual case reports and small series. As examples:

●A small retrospective study reported that cyclophosphamide may provide a significant survival benefit in patients with RA and renal AA amyloid [22]. In this study of 15 French patients, of whom six received monthly cyclophosphamide following confirmation of renal involvement, the survival of patients treated with pulse cyclophosphamide was longer than that of those administered non-alkylating drug regimens (mean survival of 165 and 46 months, respectively) [22]. Trends toward decreased proteinuria and maintenance of renal function were also noted with cyclophosphamide.

●Similar results were confirmed in a cohort study reported from Japan [17]. Prospective studies are required to properly assess the role and toxicity of this agent in this setting.

Anticytokine therapy — Increasing use of biologics with activities against proinflammatory cytokines (TNF-alpha, IL-1, and IL-6) for diseases such as RA, psoriatic arthritis, and ankylosing spondylitis (AS) may reduce the risk of development of AA amyloid as well as treating existing amyloidosis.

The majority of published experience that has shown efficacy has involved patients with RA and has reported on the first-generation TNF-alpha antagonists (ie, etanercept, infliximab, and adalimumab) [37-40]. In one large study involving 86 patients with RA complicated by AA amyloidosis, etanercept was superior to cyclophosphamide independent of the SAA 1.3 allele [41]. Less information is available regarding the newer TNF-alpha antagonists (certolizumab pegol and golimumab).

The effectiveness of the TNF antagonists appears to be directly related to their ability to control the underlying inflammatory disorder and may be enhanced by pulse therapy with glucocorticoids for induction [42]. This experience has been extended to AA amyloidosis complicating AS and Crohn disease, with regression of tissue amyloid occurring as early as three months, documented by gastrointestinal biopsies in some instances [43-45]. Efficacy of biologic response modifiers (BRMs) may be responsible in part for the marked decline in the incidence of renal replacement therapy for amyloidosis associated with inflammatory rheumatic diseases reported to nationwide registries in Finland during the period 1999 to 2008 [46].

Efficacy of the rapidly acting IL-1 receptor antagonist (IL-1ra) anakinra has been established for the 20 to 30 percent of cases of "idiopathic" AA amyloidosis reported in large series [47] and is effective in retarding recurrence of AA amyloid in combination with colchicine after kidney transplantation [48].

Since 2006, with the first report of the effectiveness of a humanized anti-IL-6 receptor antibody (tocilizumab) for the treatment of AA amyloidosis complicating juvenile chronic polyarthritis, there has been a large experience with the use of this biologic agent to improve AA amyloid complicating adult RA, as well as isolated case reports in patients with ankylosing spondylitis, Behçet syndrome, polyarteritis nodosa, Castleman disease, Sweet syndrome, and Crohn disease [49-54]. Tocilizumab has been used effectively in patients with RA and renal insufficiency due to AA amyloid [55], with biologic activity manifesting as a dramatic and rapid drop in C-reactive protein (CRP) and SAA levels to normal compared with more traditional treatments, such as methotrexate or TNF-alpha antagonists, with a significantly lower incidence of progression to hemodialysis over time [56]. In a retrospective review of 42 patients with RA and AA amyloidosis treated with tocilizumab or TNF antagonists, the former was more efficient in normalizing SAA levels, and was a more effective strategy for treating AA amyloidosis [57]. By contrast, although TNF inhibitors may reduce SAA levels in clinical practice, complete normalization is rare [52].

Only limited information is available regarding the efficacy of other biologics, such as abatacept and rituximab, which are effective in reducing disease activity but do not have direct anticytokine activity, in reversing the development of AA amyloid [58,59]. A possible therapeutic effect for IL-17 inhibition with regard to extrahepatic SAA synthesis in inflammatory rheumatic diseases has yet to be explored [60].

The impact of anticytokine therapy in preventing AA amyloid in hereditary autoinflammatory diseases, such as FMF, TNF receptor-associated periodic syndrome (TRAPS), or Muckle-Wells disease, is increasingly clear. AA amyloidosis may complicate conditions related to pyrin mutations (in particular, homozygosity for the M694V mutation in MEFV), with, and, very rarely, without the phenotype of FMF, TRAPS, the full spectrum of cryopyrinopathies (familial cold autoinflammatory syndrome, which is also termed familial cold urticaria; Muckle-Wells syndrome; and neonatal onset multisystem inflammatory disorder), and mevalonate kinase deficiency (also known as hyperimmunoglobulinemia D syndrome), with varying incidences depending upon disease and ethnic background [61].

The bulk of early experience has focused on the central role of the NALP3 inflammasome, with mutations resulting in constitutive overproduction of IL-1 beta in the cryopyrinopathies, and on the efficacy of anakinra in attenuating both disease and resulting amyloid [62-64]. Similar experience has accrued with the use of the long-acting anti-IL-1 monoclonal antibody canakinumab before and after kidney transplantation [65,66]. As yet, less information is available regarding the potential efficacy for AA amyloidosis of other synthetic agents (eg, caspase 1 inhibitors) or alternative anticytokines (eg, IL-1 TRAP or IL-18 inhibitors) that may be effective in reducing cryopyrinopathy disease manifestations [67]. With regard to TRAPS and FMF, both etanercept and anakinra have been reported as effective treatments for disease manifestations and for lowering of SAA levels, although anti-IL-1 treatments are increasingly favored [68,69]. However, similar efficacy to that seen with etanercept has not been reported with use of other TNF inhibitors, infliximab and adalimumab, in patients with TRAPS [61], and whether these agents are effective for preventing or reversing AA amyloidosis has not been described. IL-6 receptor blockade with tocilizumab may act downstream of inflammasome activation and has been reported to be an effective therapy for AA amyloidosis complicating both FMF and TRAPS; in a series of 20 patients with autoinflammatory disease and AA amyloidosis, including four patients with kidney transplants, a rapid drop in SAA levels to normal occurred within 10 days and was sustained through 23 months, with stabilization or regression of amyloid on SAP scintigraphy [70].

Anticytokine therapy remains an option for patients presenting with renal insufficiency who undergo kidney transplantation as prophylaxis to prevent recurrence of AA amyloid in the allograft; in particular, IL-1 inhibitors may be especially beneficial in this setting [69,71]. In addition, the calcineurin inhibitor FK506 (tacrolimus), which is commonly used to prevent allograft rejection, may also slow progression of AA amyloid, based on animal studies [72]. Tocilizumab was also shown to be beneficial for AA amyloidosis in a family with the SAA1.1 haplotype linked to a polymorphism in the SAA promoter [73].

Demonstration of the efficacy of available anticytokines holds the promise that novel monoclonal antibodies and low molecular weight (mw) inhibitors in phase 2/3 trials may also prove useful for the inhibition of the deposition phase of SAA amyloid [74]. In particular, tofacitinib, which is approved for the treatment of RA, has potential for the inhibition of both IL-6 and SAA in AA amyloidosis [75]. Identification of the central role of activation of the inflammasome in the generation of proinflammatory cytokines (eg IL-1beta, IL-6, IL-18) as a key pathogenic event in some autoinflammatory diseases that may be associated with the development of AA amyloid is also an active area of drug development [76,77]. The role of SAA as a signal for activation of the NLRP3 inflammasome appears to be complicated. Delipidated SAA stimulates IL-1 secretion through a mechanism dependent on NLRP3 expression and caspase-1 activity, an effect abrogated by binding to high-density lipoprotein [78]; a direct effect of SAA oligomers on the inflammasome analogous to that shown for other forms of amyloid has been postulated, but not yet shown [79].

Binding to cofactors and peptidic inhibitors — New classes of agents may eventually prove efficacious by interfering with fibril formation [80,81]. Low mw anionic sulfonates or sulfates, as well as low mw heparins, impede AA fibril formation in vitro and attenuate the development of murine AA amyloid [82,83]. This hypothesis was tested in a randomized, multicenter international phase II/III trial of the first of these compounds (eprodisate) for the treatment of 150 patients with AA amyloid [84]. In this study, eprodisate showed a clinical benefit in delaying the decline in renal function for AA amyloidosis patients, defined as a reduction in risk and delayed time to doubling of serum creatinine, to 50 percent decrease in creatinine clearance, or to progression to dialysis, end-stage kidney disease, or death (from all causes). Unfortunately, a second international phase 3 randomized trial (ClinicalTrials.gov identifier: NCT012157470) failed to confirm benefit, and drug development has been terminated. An alternative strategy is the design of low mw peptides complementary to SAA domains critical for self-aggregation in order to inhibit fibrillogenesis [85].

Clearance of amyloid deposits from tissue — Two approaches to the use of immunotherapy as a mechanism to promote clearance of existing amyloid and thereby treat patients with significant organ-system compromise have reached the stage of clinical trials.

A novel bis (proline) compound (CPHPC) has been developed that binds with high affinity to serum amyloid P component (SAP), binding to amyloid fibrils, resulting in rapid clearance of SAP by the liver and depleting its serum level by approximately 90 percent. In a preliminary report, this resulted in disappearance of deposits in three types of systemic amyloidosis (AL, ATTR, and AA), as assessed by SAP scintigraphy [86]. This effect was sustained in follow-up studies [87]; however, approximately 20 percent of P-component remained associated with the amyloid deposits, which were only slowly resorbed. This limitation was addressed initially in a murine model for AA amyloid, in which SAP depletion following CPHPC was then followed with treatment using polyclonal anti-SAP, resulting in a rapid complement-dependent, macrophage-derived giant cell reaction that appeared to clear visceral amyloid deposits [88]. Administration of a humanized monoclonal anti-SAP antibody following CPHPC treatment has been shown to trigger clearance of amyloid deposits from the liver, kidney, and other tissues in an early phase trial in patients with systemic amyloidosis, including a patient with AA amyloidosis, as well as patients with AL and hereditary forms of amyloidosis [89]. This effect has subsequently been updated to include patients with cardiac amyloidosis [90] preparatory to a phase 3 multicenter trial. This approach may provide a new and generic therapeutic approach to the treatment of systemic and possibly localized forms of amyloidosis.

An alternative approach is the use of monoclonal anti-AA antibodies as agents for the in vivo imaging of AA amyloidosis and potential mobilization of amyloid deposits from tissue [91]. One of these antibodies directed to an epitope derived from a cleavage site of SAA was found to recognize a cryptic epitope shared with human light chain (AL) amyloid [92]. Extended efficacy for AA amyloidosis remains to be established.

Investigational approaches — Accumulating knowledge regarding the structural basis of fibrillogenesis has stimulated alternative approaches to generic therapy for the amyloidoses. Sometimes referred to as akin to putting "sand in the (beta) sheets," they include the following:

●Design of or search for ligands that stabilize the native conformation of subunit proteins

●Capping and stabilization of prefibrillar aggregates

●Ligands or monoclonal antibodies specific for lipid-free SAA oligomers

●Development of drugs that inhibit the adoption of the beta-pleated sheet configuration (either generically or specifically to the type of amyloid)

●Use of chaperones to reverse aggregation

●Alterations of the milieu that is responsible for off-pathway aggregation

●Development of other methods to stimulate resorption of amyloid from tissue

Animal models — In addition to the use of low mw sulfates such as eprodisate, other studies have examined the potential efficacy of novel compounds that affect glycosaminoglycan biosynthesis and that retard the development of AA amyloid in experimental models [93]. Colchicine, dimethylsulfoxide, inhibition of interactions between SAA and the receptor for advanced glycation end-products (RAGE), fenofibrate, FK506, and lovastatin have also been shown to have efficacy in retarding the deposition phase of murine AA amyloid [72,94,95]. In a mouse model of accelerated AA amyloidogenesis, alkylating agents (chlorambucil and cyclophosphamide) were found to be more active than methotrexate and azathioprine in suppressing amyloidogenesis [96]. Proton pump inhibitors had an antiinflammatory effect and caused regression of AA amyloid in a murine knock-in model for cryopyrinopathy [97].

A novel transgenic murine model for AA amyloid may provide an in vivo system in which the kinetics of amyloid regression and the efficacy of novel therapeutics might be tested [98,99].

Murine studies have validated treatment approaches that knock down SAA using a low mw inhibitor of IL-6 [100], induction of macrophage apoptosis using clodronate-containing liposomes [101] for the attenuation of the deposition phase, and the triggering of a complement-dependent, macrophage-derived giant cell reaction by anti-human SAP antibodies for the rapid removal of massive established visceral amyloid without adverse effects [88].

SUMMARY AND RECOMMENDATIONS

●Complications and mortality risk – In untreated patients, AA (secondary) amyloidosis carries a significant risk of mortality due to end-stage kidney disease, infection, heart failure, bowel perforation, or gastrointestinal bleeding. Patients with persistently high circulating levels of SAA are at particular risk of these complications of disease. Other adverse prognostic signs included older age, reduced serum albumin concentration, and end-stage kidney failure at baseline. (See 'Prognosis' above.)

●Potential benefits of treatment and prevention – Successful treatment of the underlying inflammatory process can lead to stabilization of or improvement in renal function, to reduction in protein excretion, and to partial resolution of amyloid deposits. (See 'Importance of control of the underlying disease' above.)

●Therapeutic interventions – The preferred therapy of AA amyloid is control of the underlying inflammatory disease. Colchicine has become accepted therapy for prophylaxis and treatment of familial Mediterranean fever (FMF), and anti-proinflammatory cytokines (interleukin [IL] 1 beta, tumor necrosis factor [TNF] alpha, IL-6) have been shown to have efficacy in some cases of AA amyloidosis due to rheumatic disorders and to hereditary autoinflammatory disease. Agents under development interfere with fibril formation. (See 'Treatment of AA amyloidosis' above.)