Patients with a cardiac complete response in AL amyloidosis have survival rates similar to those of a matched general population Free

TO THE EDITOR:

Systemic light chain (AL) amyloidosis is a clonal disorder characterized by the excess production of free light chains, immunoglobulin fragments, with amyloidogenic properties.¹ The process of amyloid formation and steady accumulation in body tissues leads to organ dysfunction. The prognosis in AL amyloidosis depends on several factors, with the presence and severity of heart involvement being the most influential factors for survival.² Current therapeutic approaches aim at eradicating the underlying clone to interrupt the supply of the amyloidogenic light chains, allowing gradual recovery of the affected organ(s).³ Therapeutic approaches aimed at the existing amyloid deposits are being assessed in clinical trials to determine their efficacy.⁴ 

As current treatments are based on anticlonal plasma cell (or B-cell) therapies, response assessment has been primarily hematological (ie, stratifying response based on the degree of reduction in the circulating free light chains).⁵ The achievement of hematological complete response is associated with a higher frequency of organ response and longer survival.^5-9 However, hematological response does not measure the degree of improvement in organ function and, thus, alone is an insufficient measure of response. Natriuretic peptides are the standard measure of cardiac response, given wide access, low cost, reproducibility, and the survival advantage associated with a reduction in their level.^5,10 

The increased availability of effective therapies for AL amyloidosis led to improvement in the proportion and depth of cardiac responses. We recently published a multicenter study demonstrating the superiority of a 4-tier, natriuretic peptide–based, graded cardiac response over a binary response method.¹¹ This study included 651 patients with cardiac involvement evaluable for response. A greater reduction in natriuretic peptides was associated with longer survival. For patients achieving cardiac complete response (CarCR), defined as N-terminal pro B-type natriuretic peptide (NT-proBNP) ≤350 pg/mL or B-type natriuretic peptide (BNP) ≤80 pg/mL, the estimated 8-year survival was ≈90%. Given this excellent long-term outcome, we compared in this analysis the overall survival of CarCR patients with the survival of a sex-, age-, and country-matched control population.

Expected mortality rates were matched on the basis of sex, age, the year of initial therapy, and the country (1 × 1 sex-specific life tables at www.mortality.org). Relative survival analyses evaluated the observed vs expected survival of our CarCR cohort using the survival and relsurv packages in the R statistical program.^12-14 A 1-sample log-rank test compared the CarCR cohort overall survival with the matched population estimates. The institutional review board in the participating centers approved the study.

In this study across 6 countries, we identified 104 patients who achieved a CarCR as their best response to treatment (16% of the study population). The baseline characteristics of these patients are listed in the supplemental Table (available on the Blood website). The CarCR cohort was predominately male (57%) with a median age of 57 years. Hematological complete response (CR) was achieved in 64% of patients. The median baseline NT-proBNP and BNP were 1565 and 257 ng/L, respectively. At CarCR, the median NT-proBNP and BNP were 275 (interquartile range, 205-303) ng/L and 65 (interquartile range, 44-77) ng/L, respectively. In patients who had achieved a CarCR, the estimated 5-year survival rate was 93% (95% confidence interval, 87%-98%) (Figure 1A). The expected 5-year survival rate for this age-, sex-, and country-matched cohort was 95%, with a 1-sided log-rank test P value of .45. When CarCR patients were stratified by cardiac stage at diagnosis (Figure 1B) and best hematological response (Figure 1C), no difference in survival rate was noted between the subgroups and matched general population survival estimates. In contrast, the survival rate of non-CarCR patients (cardiac very good partial response [CarVGPR] or lower response) was inferior in cardiac stage III vs cardiac stage II and in those not attaining hematological CR compared with those in hematological CR (supplemental Figure 1A-B). Of the 104 CarCR patients, 60 had renal involvement assessable for renal response. Of those patients, 50 achieved renal CR (defined as 24-hour urine protein ≤200 mg) or renal very good partial response (VGPR) (defined as >60% reduction in 24-hour urine protein from baseline not meeting criteria of renal CR),¹⁵ which are 2 renal response categories that were shown collectively to have a superior survival over lower renal response.¹⁵ The survival rate of patients in CarCR with no coexisting renal involvement was similar to patients with CarCR with renal involvement who achieved renal CR/VGPR, whereas it was marginally lower in the small subgroup of patients in CarCR with coexisting renal involvement who attained less than renal VGPR, although statistical significance was not reached (5-year survival rate, 92%, 96%, and 77%, respectively; P = .25) (supplemental Figure 2). We have previously shown that the depth of cardiac response inversely correlates with the risk of cardiac progression.¹¹ Of the 100 CarCR patients with data available on cardiac progression, 10 experienced NT-proBNP/BNP progression. The survival rate of CarCR patients who experienced cardiac progression was lower than CarCR patients with no cardiac progression (5-year survival rate, 80% vs 95%), but statistical significance was not reached (P = .08) (supplemental Figure 3).

We then analyzed the interaction between the hematological and cardiac responses in the wider study cohort (CarCR and non-CarCR patients, n = 651). In Figure 2A, we stratified the survival in this cohort based on the best hematological response, irrespective of cardiac response, and compared it with the survival estimates of a matched general population. The survival was longer with deeper hematological response, with the longest survival seen in patients who achieved hematological CR. However, the survival of hematological CR patients was inferior to the expected survival of the matched general population (P < .0001). When we look at the survival of patients who achieved hematological CR as their best response stratified by their best cardiac response, the survival was longer with a deeper cardiac response. Only patients who in addition to achieving hematological CR also achieved CarCR had a similar survival to a matched general population (Figure 2B).

This analysis marks an important treatment milestone and serves as proof of concept that a complete reversal of cardiac damage is achievable in AL amyloidosis through suppression of the amyloid light chain. Remarkably, a complete cardiac response is associated with recovery of normal life expectancy, and CarCR should be considered a relevant end point in clinical trials. In our AL cardiac amyloidosis population, treated mainly with bortezomib-based regimens, only 16% of patients reached CarCR as their best response. Therefore, to raise the proportion of patients achieving CarCR, efforts should focus on increasing early diagnosis, as patients with less advanced cardiac stages are more likely to achieve CarCR.³ The quest for more effective therapeutic options should continue to further improve the response rate, as seen with the addition of daratumumab to bortezomib-based regimens.⁶ 

This study demonstrates that the hematological response is important as a means to increase the likelihood of cardiac response. However, the cardiac response has a greater impact on survival than the hematological response, strengthening the argument that the purpose of the hematological response is to produce an organ response and the organ response eventually translates into the survival advantage.

Limitations of this work include a small number of CarCR patients, which may be underpowered to differentiate survival from the general population and underpowered analysis for some subgroups of interest.

In conclusion, the achievement of CarCR in AL amyloidosis, based on the reduction in natriuretic peptide level, results in similar survival to the general population and indicates that complete reversal of cardiac damage is feasible with currently available therapies.