Prognosis and risk assessment in AL amyloidosis - There and back again

The presentation and course of AL amyloidosis is characterized by significant heterogeneity, depending on the organ tropism of the amyloidogenic light chain, the severity of organ dysfunction and the characteristics of the plasma cell clone that produces the toxic light chains. Despite multiorgan involvement, it is the amyloidotic involvement of the heart, and the resulting cardiac dysfunction, that dominates prognosis and determines short and intermediate term outcome. Current treatment approaches for AL amyloidosis aim to eliminate the toxic light chains, and a clonal response is followed by organ response and improvement in survival and organ function (Kastritis & Dimopoulos, 2016; Palladini & Merlini, 2016; Wechalekar et al, 2016). Treatment decisions in AL amyloidosis depend on the assessment of prognosis at presentation, and the identification of features of advanced risk. Accurate assessment of prognosis, based on objective, standardized and validated criteria, is crucial for informative treatment decisions. Unlike most diseases, patients with dismal prognosis are less likely to receive aggressive therapies, due to the high risk of detrimental complications and early mortality (Kastritis & Dimopoulos, 2016; Palladini & Merlini, 2016; Wechalekar et al, 2016). Given that the degree of cardiac dysfunction determines prognosis, cardiac biomarkers have been used to formulate an objective, sensitive, reproducible and easily available staging system for patients with AL amyloidosis (Dispenzieri et al, 2004). This system contains only two cardiac biomarkers [N-terminal prohormone of brain natriuretic peptide (NTproBNP) and troponins] and defines three stages: a unique staging system for a ‘haematological malignancy’. The addition of free light chains levels improves prognostic ability (Kumar et al, 2012) but still, cardiobiomarker levels offer the most important prognostic information: among patients with both biomarkers elevated (stage 3), those with NTproBNP levels ≥8500 pg/ml constitute a very high risk group with very dismal prognosis (Wechalekar et al, 2013). There is however, a continuing effort to identify new markers associated with prognosis that could further refine risk stratification, mostly concerning evaluation of heart involvement and its consequences. Imaging studies using cardiac echocardiography are used for initial evaluation and more sophisticated technology and analytic tools have been investigated, but have not been validated extensively, are operator dependent and not widely available. Cardiac magnetic resonance imaging is a valuable diagnostic tool and adds prognostic information, especially when new techniques of image acquisition and analysis are used; however, it is an expensive tool with limited availability (Boynton et al, 2016). New tracers may further improve the diagnostic ability of nuclear cardiac imaging but their prognostic performance is under investigation (Van Der Gucht et al, 2016). Alternatively, new biomarkers, which are standardized, widely available and at relatively low cost, seem also to provide added prognostic value, especially in identifying patients at higher risk (Kastritis et al, 2016). In order to assess the effectiveness of a therapy on the plasma cell clone and its impact on organ function, both haematological (Palladini et al, 2012) and organ response criteria (Palladini et al, 2012, 2014) have been developed. Assessment of cardiac response is based on NTproBNP levels, and it has been shown repeatedly in different series that reduction of NTproBNP levels is strongly associated with a major improvement in survival (Palladini et al, 2012), despite some limitations in specific patient groups. Similarly, for assessment of kidney response, biomarkers (estimated glomerular filtration rate and proteinuria) have been proposed and are systematically used to define organ prognosis and response to therapy (Palladini et al, 2014). Thus, we can claim that we can assess prognosis and follow a complicated disease, such as AL amyloidosis, based on reproducible, easily available and sensitive biomarkers (Merlini et al, 2016). In contrast, there is a limited role for imaging studies in assessing response to therapy, since significant changes may take months or years to become evident with current imaging tools, such as echocardiography. But, are these biomarkers and their changes associated with real and significant clinical outcomes? Solid data support that reduction of NTproBNP is associated with improvement in functional class [as represented by New York Heart Association (NYHA) staging for congestive heart failure](Palladini et al, 2012) and with better prognosis. Furthermore, how is this translating into an objective functional measure? NYHA class is a very rough measure of functional capacity, and reduction of NTproBNP and improvement NYHA class are strongly correlated in AL amyloidosis. In this issue, Pulido et al (2017), from the Boston University group, evaluated the six-minute walk test (6MWT) as a tool for an objective evaluation of functional exercise capacity in patients with AL amyloidosis. Surprisingly, while 6MWT has been widely used for the evaluation of patients with cardiopulmonary diseases, both for prognostic evaluation and assessment of response to medical intervention, there is no data regarding its usefulness for evaluating patients with AL amyloidosis. In this retrospective study, 6MWT distance was associated with heart involvement, NYHA class, BNP levels and overall survival. This test does not require sophisticated technology, essentially has no cost and offers prognostic information. However, there are difficulties in standardization and it is affected by several other conditions (for example peripheral neuropathy, orthopaedic problems, age, gender, body mass index, etc.), which can severely affect the applicability, validity and reproducibility of the test. Nevertheless, and despite these limitations and the lack of prospective data in patients with AL amyloidosis, 6MWT has been included as a measure of efficacy in phase 3 registration clinical trials in AL amyloidosis. But, why should we use a test like 6MWT and not just biomarkers, which are reproducible and validated in AL amyloidosis? Why move from a ‘biomarker-based’ assessment back to a ‘clinical/functional measure’ assessment, which is affected by many non-cardiopulmonary conditions and other factors not related to the patient? It is notable that the competent authorities have been reluctant to accept biomarkers alone as endpoints for the assessment of the efficacy of a therapy, especially one that is aiming to improve organ function, despite the amount of data on the reliability of biomarkers. This issue is becoming critical as we are witnessing a paradigm shift in the treatment of AL amyloidosis: monoclonal antibodies targeting the amyloid deposits may improve the organ function of patients with AL amyloidosis by facilitating macrophage-dependent amyloid fibre degradation (Richards et al, 2015; Gertz et al, 2016). The assessment of the effect of these new drugs is reflected on ‘biomarker responses’, such as NTproBNP and proteinuria reductions (Gertz et al, 2016); phase 3 studies are ongoing and, if positive, may ultimately lead to the approval of ‘amyloid targeting’ therapies for the first time. In our clinical practice, we continue to assess the prognosis of patients with AL amyloidosis based on biomarkers that include NTproBNP, troponins, eGFR, proteinuria and free light chains, along with our clinical judgment. The use of these biomarkers has improved substantially our approach to the treatment of patients with AL amyloidosis, tailoring therapy and allowing early evaluation of response. Meanwhile, we continue to search for new biomarkers that could further improve the accuracy of assessment of prognosis and response. EK and MAD reviewed the literature and authored the manuscript.