Targeted Therapeutics for Transthyretin Cardiac Amyloidosis

Abstract

Despite the availability of proven treatments for some patients with heart failure (HF), many patients—particularly those with HF and preserved ejection fraction (HFpEF)—remain difficult to treat, resulting in persistently high morbidity and mortality in the majority of HF patients. The lack of effective treatments, disappointing results of many HF randomized clinical trials (RCT), and variable treatment responses even for proven therapies are all possible reasons for the poor prognosis of HF patients. In the context of this backdrop, it is not surprising that there has been growing interest and enthusiasm for “precision medicine” in order to improve outcomes for patients who suffer from the HF syndrome.^{1, 2}

Notwithstanding the intuitive appeal of precision medicine for heterogeneous clinical syndromes such as HF, in reality achieving the goal of targeted therapeutics can be quite difficult. Indeed, there are those who believe that it is nothing more than hype that may never be realized due to a number of potential pitfalls.³ There is significant inter-individual variation in treatment responses, thereby making it very difficult to identify a “responder” phenotype in RCTs.^{3, 4} In addition, due to the biological complexity of clinical syndromes such as HF, biomarkers, genetic variants, or other diagnostic tests that purportedly identify patients as candidates for specific therapies may not be inaccurate, and may lead to inappropriate withholding or targeting of therapies. Furthermore, developing precision therapeutics typically requires understanding the molecular pathogenesis of disease, which is difficult for vaguely defined clinical syndromes such as HF in which access to diseased tissue is not easy. Finally, by targeting therapeutics, we reduce the pool of eligible patients for RCTs, potentially making it difficult to enroll patients into these studies.

Given the many potential pitfalls of precision medicine and targeted therapeutics, we may ask ourselves whether it is even worth trying to use novel methods^{2, 5} to classify HF into subgroups that have more homogeneous disease pathophysiology and may respond in a more consistent manner to specific treatments. Fortunately, recent data from transthyretin (TTR) amyloid cardiomyopathy (ATTR-CM) RCTs^6–8 provide compelling evidence that supports continued efforts for more precise classification of HF.

ATTR-CM is an increasingly recognized infiltrative cardiomyopathy that results from the dissociation of the normal tetrameric form of TTR—which causes the release of TTR monomers that are prone to misfolding, thereby leading to TTR amyloid fibril formation and deposition in the myocardium.⁹ ATTR-CM can be due (1) a mutation in the TTR gene (hereditary ATTR-CM), which causes a change in the amino acid sequence of TTR resulting in tetramer dissociation; or (2) incompletely understood age-related misfolding of TTR (wild-type ATTR-CM). Several therapeutics, including patisiran (an RNA interference [RNAi] therapeutic),⁷ inotersen (an anti-sense oligonucleotide inhibitor),⁸ and TTR stabilizers (tafamadis⁶ and AG10¹⁰) are currently in development for ATTR-CM and for ATTR-associated polyneuropathy, the other major manifestation of the hereditary form of the disease. Most ATTR-CM patients with HF have a relatively preserved ejection fraction (until late in the disease process), and typically fall under the umbrella of HFpEF, especially because these patients have increased LV wall thickness and are often elderly. Although still under-diagnosed, ATTR-CM is increasingly recognized as a cause of HFpEF especially due to advances in imaging with characteristic findings on echocardiography (speckle-tracking imaging demonstrating a relative sparing of longitudinal strain at the apex compared to the base); cardiac magnetic resonance (difficulty nulling the myocardium on delayed gadolinium enhancement imaging and high extracellular volume fraction on T1 mapping); and bone scintigraphy (e.g., elevated ratio of heart to contralateral lung uptake on 99m-technetium pyrophosphate scanning). However, the development of each of these imaging techniques for ATTR-CM was based at least initially on pathologic confirmation of TTR protein within amyloid deposits on endomyocardial biopsy samples. Thus, the identification of the ATTR-CM subgroup of HFpEF (particularly in patients with wild-type ATTR-CM or hereditary ATTR with predominant cardiac manifestations [e.g., the V122I mutation, which is present in 3–4% of individuals with African ancestry]) was based on pathologic tissue analysis of the primary diseased organ—the heart.

Patients who develop HF due to ATTR-CM have a high morbidity and mortality, with progressive decline in functional status and quality of life, and a high rate of hospitalization and premature death.⁹ The recently completed ATTR-ACT trial of the oral TTR stabilizer tafamadis demonstrated reduced all-cause mortality and cardiovascular hospitalizations, along with prevention of a rapid decline in 6-minute walk test distance and quality of life.⁶ In this issue of Circulation,¹¹ Solomon and colleagues report the results of a pre-specified subgroup analysis of the APOLLO trial⁷ of the RNAi therapeutic patisiran in patients with hereditary ATTR polyneuropathy. The pre-specified cardiac subgroup included enrolled patients who had increased LV wall thickness (≥13 mm) but no history of hypertension or aortic valve disease. The subgroup analysis demonstrated that patisiran, compared to placebo, reduced LV wall thickness, improved LV longitudinal strain, increased cardiac output, and lowered N-terminal B-type natriuretic peptide (NTproBNP) levels. The lowering of NTproBNP with patisiran was also seen in the overall APOLLO trial. Finally, there was a suggestion of improved cardiac outcomes in the patisiran group compared to placebo. These findings led the authors to conclude that patisiran may be helpful in halting the cardiac progression and thereby may lead to improved outcomes in patients with ATTR-CM.

Although the study findings are compelling, several questions remain. The APOLLO trial only enrolled patients with hereditary ATTR polyneuropathy. Although the cardiac subgroup was pre-specified, the findings were generated in a post-hoc, exploratory analysis, and the sample size was relatively small (90 patisiran-treated patients vs. 36 placebo-treated patients). In addition, the APOLLO trial was started prior to the broad recognition of bone scintigraphy as a diagnostic test for ATTR-CM. Future studies of hereditary ATTR polyneuropathy will have an easier time determining which of the enrolled patients truly have cardiac involvement via the use of scintigraphy for more accurate diagnosis of ATTR-CM. Furthermore, while a prior history of known cardiac involvement was not an exclusion criteria per se, those with New York Heart Association class III or greater symptoms were excluded. APOLLO also did not include the more prevalent wild-type form of ATTR-CM. Therefore, we do not know if patisiran is effective in hereditary ATTR-CM patients with more advanced cardiac disease or in wild-type ATTR-CM patients. As shown in the Table, which compares patient characteristics among the APOLLO cardiac subgroup,¹¹ the ATTR-ACT trial (which specifically enrolled ATTR-CM patients [both hereditary and wild-type]),⁶ and an observational study that included hereditary and wild-type ATTR-CM,¹² demonstrates why this distinction between patient entry criteria is important. The cardiac subgroup in APOLLO was younger and had much less severe HF compared to typical ATTR-CM patients; therefore the results of the APOLLO cardiac subgroup analysis may not be generalizable to typical ATTR-CM patients.

Table.

Comparison of Baseline Characteristics of the APOLLO Trial (Patisiran) vs. the ATTR-ACT trial (Tafamadis)

	APOLLO11 Cardiac Subgroup (n=126)	ATTR-ACT6 Tafamadis (N=264)	ATTR-ACT Placebo (N=177)	Quarta et al.12 ATTRm (n=36)	Quarta et al. ATTRwt (n=56)
Age, years*	61 (54–67)	75 (46–88)	74 (51–89)	62 (13)	76 (6)
Male sex	78%	91%	89%	75%	88%
Black race	2%	14%	15%	Not reported	Not reported
ATTRwt	0%	76%	76%	0%	100%
ATTRm	100%	24%	24%	100%	0%
Predominant genotypes in ATTRm patients	Val30Met, Ala97Ser, Thr60Ala	Val122Ile, Thr60Ala, Ile68Leu	Val122Ile, Thr60Ala, Ile68Leu	Ile68Leu, Glu89Glyn, Val122Ile	—
Hypertension	0%	55%	48%	36%	29%
NYHA class
Class I	40%	9%	7%	Not reported	Not reported
Class II	60%	61%	57%	Not reported	Not reported
Class III or IV**	0%	30%	36%	33%	32%
NTproBNP, pg/ml (median [IQR])	837 (292–2354)	2996 (1752–4862)	3161 (1864–4825)	1636 (770–3396)	3708 (1807–6068)
LV ejection fraction, % (mean [SD])	61 (10)	48 (10)	49 (10)	57 (13)	51 (11)
Septal wall thickness, mm***	17 (15–19)		16.2 (3.5)	16 (3)	17 (3)
LV global longitudinal strain, %***	−15.1 (−17.5, −12.6)		−9.4 (3.6)	−15 (4)	−11 (3)

APOLLO = A Phase 3 Multicenter, Multinational, Randomized, Double-blind, Placebo-controlled Study to Evaluate the Efficacy and Safety of Patisiran (ALN-TTR02) in Transthyretin-Mediated Polyneuropathy; ATTR-ACT = Transthyretin Amyloid Cardiomyopathy Tafamidis Study; ATTRm = hereditary (mutant) transthyretin amyloidosis; ATTRwt = wild-type transthyretin amyloidosis; NYHA = New York Heart Association; NTproBNP = N-terminal B-type natriuretic peptide; IQR = interquartile range; LV = left ventricular; SD = standard deviation.

^*Reported as median (25^th-75^th percentile) in the APOLLO and ATTR-ACT trials, and mean (standard deviation [SD]) in the study by Quarta et al.

^**NYHA class III and IV patients were excluded from the APOLLO trial, and NYHA class IV patients were excluded from the ATTR-ACT trial.

^***Reported as median (25^th-75^th percentile) in the APOLLO trial, and mean (SD) in the ATTR-ACT trial and the study by Quarta et al.

Despite the encouraging results of the APOLLO cardiac subgroup analysis, a dedicated phase 3 RCT in ATTR-CM will be necessary to truly determine whether patisiran (or newer generation RNAi therapeutics) are effective in ATTR-CM. Indeed, a subcutaneous formulation of an RNAi (resuviran) was associated with an increased mortality compared to placebo in hereditary ATTR-CM patients enrolled in the Phase 3 ENDEAVOUR randomized controlled trial,¹³ which underscores the need to specifically test novel therapeutic classes in ATTR-CM patients prior to their use even if they are effective in ATTR-associated polyneuropathy. Other unknowns include (1) whether combination therapy with an RNAi (to reduce production of TTR) and a TTR stabilizer (to reduce the rate-limiting step of fibril formation) will further improve cardiac structure/function, symptoms, and outcomes in ATTR-CM; and (2) the effect of patisiran in patients with TTR mutations that were underrepresented in APOLLO but are relatively common (e.g., V122I).

Despite these limitations, the development of ATTR-CM as an increasingly well-recognized cause of HFpEF, along with the encouraging results of recent RCTs, teaches us several lessons. First, although HF—particularly HFpEF—is a heterogeneous syndrome, careful determination of its etiology via tissue sampling of the heart and other affected organs, may be extremely useful in identifying subgroups with a distinct molecular basis. Prior failed HFpEF RCTs have no doubt included some ATTR-CM patients and may have also included other HFpEF subtypes that require specific therapies that differ from those tested in these RCTs. Second, we should not give up hope on precision medicine but rather continue to strive to develop novel techniques—clinical, molecular, imaging, statistical—to identify treatable subgroups and then investigate them further to determine their molecular basis. Finally, earlier recognition of ATTR-CM is essential. In APOLLO, as shown in the Table, patients in the cardiac subgroup who appeared to respond to patisirian had less advanced disease compared to patients with typical ATTR-CM.^{6, 12} In addition, in ATTR-ACT, the NYHA class II patients appeared to respond the best to tafamadis.⁶ Systematic, automated screening of repositories of echocardiograms and electrocardiograms using deep machine learning and computer vision techniques;¹⁴ screening at-risk patients with bone scintigraphy⁹ or ATTR-specific biomarkers¹⁵ or genetic testing may allow for earlier identification of patients with ATTR-CM.

In conclusion, in a pre-specified cardiac subgroup analysis of the APOLLO trial, patisiran was associated with improvements in cardiac structure/function, NTproBNP, and outcomes. Although not definitive, these findings are encouraging for further development of patisiran for ATTR-CM. Potentially even more intriguing, however, is the possibility that the results of trials such as APOLLO and ATTR-ACT foretell a future of more successful applications of precision medicine to HF syndromes.