Keywords Polyneuropathies - Prealbumin - Amyloid Neuropathies, Familial - Amyloidosis, Familial
INTRODUCTION
Hereditary transthyretin amyloidosis (ATTRv) is an autosomal dominant disorder caused
by mutations in the transthyretin (TTR ) gene, which is a tetrameric transport protein produced mainly by the liver that
carries retinol and thyroxin throughout the body. The mutation provokes an amino acid
exchange, ensuing protein misfolding and tetramer dissociation. The deposition of
monomers will cause amyloid deposits mainly in the nerves, heart, and kidney.[1 ]
A severe, progressive, and disabling disease, ATTRv presents variable phenotypes depending
on the mutation, but it is classically divided into the polyneuropathy and cardiomyopathy
forms. The most common mutation in Brazil is V30M (p.Val50Met), which causes a peripheral
neuropathy-predominant phenotype (ATTRv-PN).[2 ] In the United States, the most common variant is Val122ile (p.Val142Ile), which
leads to a cardiomyopathy-predominant phenotype (ATTRv-CM) and is present in 3.5%
of African Americans.[3 ] Once TTR amyloid is formed, it cannot be cleared with the therapy currently available.
Therefore, establishing an early diagnosis and treatment are paramount.
For many years, liver transplantation (LT) was the only available treatment for ATTRv-PN;
however, nowadays, it is rarely recommended. Additionally, the costs associated with
LT are high, and the procedure carries significant morbidity and mortality. Over the
past 20 years, many new medications have been developed which are much safer than
LT. There are two main categories of drugs: the tetramer stabilizers and the gene
silencers. The stabilizers were the first to be developed, and they act by directly
binding to the tetramers with high affinity and preventing the monomers from dissociating.
The gene silencers act in the TTR mRNA, preventing it from being translated.[3 ]
Tafamidis is a kinetic stabilizer that binds to TTR with high affinity, inhibiting
its dissociation. It was the first drug approved for the treatment of ATTRv-PN in
Europe, South America, and Asia, and it acts by slowing down the progression of the
disease, but does not completely halt it.[4 ] It is very safe and has shown for more than10 years its efficacy and very low rate
of adverse events.[5 ] The drug has not been approved yet for ATTRv-PN by the US Food and Drug Administration
(FDA), although it has been recently approved to treat ATTRv cardiomyopathy. Tafamidis
was approved by the Brazilian Health Regulatory Agency (Agência Nacional de Vigilância
Sanitária, ANVISA) in 2017 and included in the National Therapeutic Form (Formulário
Terapêutico Nacional, FTN) of the Brazilian Unified Health System (Sistema Único de
Saúde, SUS) in 2018. We herein describe a 10-year-long real-world experience with
tafamidis in the Brazilian National Amyloid Referral Center (Centro de Estudos em
Paramiloidose Antônio Rodrigues de Mello, CEPARM).
METHODS
We retrospectively studied all patients with ATTRv-PN evaluated at our center, from
September 2011 to March 2022 (data cut-off), who were initiated on daily doses of
20 mg of tafamidi and had at least 1 follow up visit 6-months after baseline. All
patients signed an informed consent form, and the study was approved by the Review
Board at Universidade Federal do Rio de Janeiro. The patients were assessed by a neurologist,
a nephrologist, and a cardiologist at each visit, as per our center's protocol. All
patients were carriers of a known pathogenic variant in the TTR gene and symptomatic ATTRv-PN.
To be included in the present study, the patients had to present signs of peripheral
neuropathy on the neurologic examination or nerve conduction studies or SUDOSCAN (Impeto
Medical, Issy-les-Moulineaux, France) test, or positive tissue biopsy. The following
assessments were performed on each visit: Polyneuropathy Disability Score (PND); Neuropathy
Impairment Score (NIS); Karnofsky Performance Status (KPS); body mass index (BMI);
electrocardiogram (ECG); transthoracic echocardiogram; and conduction studies of the
right peroneal motor and sural nerves.
The patients were divided into the late-onset (LO) ATTRv-PN group, with age at symptom
onset ≥ 50 years, and the early-onset (EO) ATTRvPN) group, with age at symptom onset < 50
years. We classified them as tafamidis non-responders if they presented an increase
on the NIS of 7 or more points per year since the start of the treatment (tafamidis
day 1, D1).[6 ]
[7 ] Interventricular septum hypertrophy was defined as > 12 mm (ATTRv-CM).[3 ]
Statistical analysis
The categorical variables were expressed as numbers and percentages, and the differences
between the groups were compared using the Fisher's Chi-squared (χ2 ) exact test. The continuous variables were expressed as median and interquartile
range (IQR) values, and the comparison between the groups was performed through Kruskal-Wallis
test. Pairs matched by baseline and the last follow-up matched were tested with the
Wilcoxon-rank signed test. The JMP (SAS institute, Cary, NC, USA) software, version
17, was used for all statistical analysis. All the tests were two-sided, and p -values < 0.05 were considered significant.
RESULTS
Of the 45 patients who started the treatment with tafamidis during the study period,
33 were included in the present study. All patients were carriers of the V30M mutation
of ATTRv-PN. The sample was composed of 18 (55%) female patients, and the EO-ATTRv-PN
group contained 20 (61%) subjects. The median age at baseline was of 40 (IQR: 36–68)
years ([Table 1 ]). The Median baseline NIS was of 10 (IQR: 6–24) points, and the median BMI, of 26
(23–28) kg/m2 . In total, 17 (59%) patients had an abnormal ECG and 3 presented ATTRv cardiomyopathy
(all LO). The mean time from baseline to the last follow-up was of 3.4 years. Besides
age at symptom onset, there were no significant differences in terms of demographic,
neurologic or cardiac features between the EO- and LO-ATTRv-PN patients. A total of
3 patients had used inotersen previously (3–5 years before initiating tafamidis, from
2015 to 2019) and changed to tafamidis due to side effects from inotersen (thrombocytopenia:
n = 2; and glomerulonephritis: n = 1).
Table 1
Demographic, clinical, and electrodiagnostic findings at tafamidis day 1 (baseline)
Total
EO-V30M patients
(N = 20)
LO-V30M patients
(N = 13)
p -value
Female sex: n (%)
18 (55)
11(55)
7 (54)
1.0
Age (years): median (IQR)
40 (36–68)
36 (30–39)
65 (62–68)
< 0.0001
Disease duration (years): median (IQR)
3 (1–4)
2.5 (1–4)
3 (2.5–4.5)
0.3512
NIS: median (IQR)
10 (6–24)
8 (2–17)
14 (7–29)
0.1269
BMI: median (IQR)
26 (23–28)
25 (22–27)
26 (25–29)
0.1846
KPS: median (IQR)
90 (80–90)
90 (80–90)
90 (80–90)
0.4
PND: n (%)
I
27 (82)
17 (85)
10 (77)
0.5874
II
5 (15)
3 (15)
2 (15)
IIIa
1 (3)
0
1 (8)
Sural amplitude (µV): n = 29
2.9 (0–7)
3.6 (0–9.7)
0 (0–3.8)
0.0666
Peroneal amplitude (mV): n = 29
2.4 (0.5–4.8)
3.1 (0.7–5)
1.1 (0–4)
0.2142
Abnormal ECG:n (%) – n = 29
17 (59)
9 (53)
8 (67)
0.7032
Abnormal IVS:n (%) – n = 25
3 (8)
0
3 (30)
0.0522
Abbreviations: BMI, body mass index; ECG, electrocardiogram; EO, early-onset; IQR,
interquartile range; IVS, interventricular septum; LO, late-onset; KPS, Karnofsky
Performance Status; NIS, Neuropathy Impairment Score; PND, Polyneuropathy Disability
Score.
At the last follow-up, the BMI, neurological impairment, and level of disability slightly
worsened compared with baseline; however, the sural nerve's sensory amplitude, and
the peroneal nerve's compound muscle action potential (CMAP) amplitude remained stable
([Figure 1 ]). This same result was observed across EO- and LO-ATTRv-PN patients. Overall, 25
(75.8%) patients were considered responders, and 8 (24.2%), non-responders ([Figure 2 ]). Similar results were observed among LO-ATTRv-PN patients (n = 13; 10 responders and 3 nonresponders), patients who previously used inotersen
(n = 3; 2 responders and 1 nonresponder), and those who presented cardiomyopathy (n = 3; 3 responders). The median rate of NIS progression was of 0 points per year in
responders and of 11.3 per year in nonresponders. There were 10 nonresponders who
worsened the PND stage in at least 1 point compared with baseline ([Figure 2 ]). Furthermore, 8 outpatients discontinued tafamidis between 2019 and 2021 to join
the Helios[8 ] and Neuro-Transform[9 ] trials. Comparing the baseline characteristics of responders and non-responders,
there were no differences between the groups ([Table 2 ]).
Abbreviations: NIS, Neuropathy Impairment Score; KPS, Karnofsky Performance Status;
BMI, body mass index.
Figure 1 Matched comparisons of neurological and functional outcomes from baseline (tafamidis
day 1, D1) to the last follow-up.
Abbreviation: PND, Polyneuropathy Disability Score.
Figure 2 Progression of disability at the last follow-up and response to tafamidis.
Table 2
Comparison of baseline characteristics of tafamidis in responders versus nonresponders
Responders
(N = 25)
Non-responders
(N = 8)
p -value
Female sex: n (%)
15 (60)
3 (38)
0.4184
Age at tafamidis initiation (years)
41 (36–63)
36 (30–70)
0.8003
LO-V30M patients: n (%)
10 (40)
3 (38)
1.0
Disease duration (years): median (IQR)
3 (1–4)
3 (1–5)
0.8985
NIS: median (IQR)
10 (4–26)
15 (9–24)
0.3326
BMI: median (IQR)
26 (24–29)
26 (19–27)
0.3890
KPS: median (IQR)
90 (80–90)
85 (70–90)
0.6029
PND: n (%)
I
20 (80)
7 (88)
0.7234
II
4 (16)
1 (12)
IIIa
1(4)
Sural amplitude (µV): n = 29
3.5 (0–8.1)
1.2 (0–3.7)
0.3017
Peroneal amplitude (mV): n = 29
2.8 (0.5–5.3)
1(0.3–3.8)
0.3731
Abnormal ECG: n (%) – n = 29
14 (61)
3 (50)
0.6693
Abnormal IVS: n (%) – n = 25
3 (15)
0
1.0
Abbreviations: BMI, body mass index; ECG, electrocardiogram; IQR, interquartile range;
IVS, interventricular septum; LO, late-onset; KPS, Karnofsky Performance Status; NIS,
Neuropathy Impairment Score; PND, Polyneuropathy Disability Score.
No patient stopped tafamidis to undergo LT, neither were there deaths during the study
period. No adverse event nor serious adverse events were considered related to tafamidis.
DISCUSSION
The current study showed that the efficacy and safety of tafamidis reported in clinical
trials is expandable to the Brazilian real-world scenario. Most patients benefited
from the therapy, most remained in the same PND stage, there was no significant neurophysiological
worsening of the peripheral neuropathy, and the KPS, BMI, and NIS only worsened slightly
over a long period of time. Tafamidis is the only ATTRv disease-modifying therapy
available in the SUS. The drugs inotersen and patisiran were recently approved by
ANVISA, but the high costs make them a limited option to our ATTRv-PN population,
as private health insurers do not cover these medications, and the SUS cannot afford
them.
A tafamidis phase-II/-III randomized, double-blinded, placebo-controlled clinical
trial (FX-005) showed a greater proportion of NIS-Lower Limbs (NIS-LL) responders
among the tafamidis patients compared to those who received placebo (60.0 versus 38.1%
respectively; p = 0.041) and better quality of life in the tafamidis group (Least-Square Norfolk
Quality of Life-Diabetic Neuropathy total score: 8.8-point difference; p = 0.045) using efficacy-evaluable analysis.[10 ] This trial lasted for 18 months and was followed by an open label extension study
(FX-006), which confirmed that the benefits were sustained for 30 months, and that
early initiation of tafamidis was associated with better responses and outcomes.[11 ] Other studies[12 ] demonstrated that this drug caused a long-term (6 years) delay in neurological and
nutritional status deterioration in V30M patients.
Real-world studies have shown mixed results regarding tafamidis efficacy around the
globe. In 252 Portuguese subjects (V30M in 92.5% of the cases and mean age of 40.4
years), tafamidis significantly slowed disease progression during the 2 years of follow-up,
when compared with nonrandomized untreated subjects.[13 ] Patients on tafamidis presented slower progression on the NIS-LL and its subscales,
as well as lower levels of deterioration in the Norfolk Quality of Life scale. There
were no significant differences in terms of BMI and KPS score between the groups.[13 ] In a large Portuguese natural history study,[5 ] patients with EO stage-I V30M ATTRv-PN on tafamidis presented approximately 91%
of reduction in the mortality risk compared with untreated patients.
Monteiro et al.[14 ] retrospectively studied 210 patients with V30M ATTR with predominantly neuropathic
phenotype treated with tafamidis for 18 to 66 months. The patients were classified
by an expert as responders, partial responders, and non-responders. In total, 72 patients
were classified as responders (34%; no disease progression, NIS change from baseline
≤ 0). Non-responders (61; 29%) and partial responders (76; 36%) presented worsening
of sensory, motor and autonomic neuropathy (non-responders: average NIS increase of
5.9/year; partial responders: 1.8/year). Milder disease severity, female sex, and
native higher levels of tetrameric TTR concentration at the start of the treatment
were the most relevant good predictors of response.[14 ]
In contrast, in non-endemic regions, real-world studies have shown that tafamidis
is not similarly effective. A recent German single center study[15 ] with a mean follow-up of 34 months demonstrated that 56% of the patients (36/64;
most carriers of variants other than V30M) on tafamidis worsened at least one stage
in the PND scale at the last follow-up. In an Italian study[16 ] (n = 61; most carriers of variants other than V30M) that included mostly LO-ATTRV-PN
patients, only 1/3 subjects appeared to benefit from tafamidis. Another study[17 ] with non-V30M LO-ATTRv-PN patients from France also observed that tafamidis was
only effective in slowing disease progression in 1/3 patients after 3 years of follow-up.
The response to tafamidis in ATTRv-PN patients appears to be different in endemic
versus non-endemic regions, which is likely explained by different genotypes, time
from symptom onset to diagnosis, and neuropathy severity at drug initiation.
In the present study, 75% of the patients responded to tafamidis. Only 10 patients
presented progression of ther PND stage at the last follow-up. The median rate of
NIS progression per year among the non-responders was of 11.3, which is similar to
the expected rate of progression in untreated patients (11.7–14.3/year).[18 ]
[19 ] Given the difficult access to the very expensive gene therapy in Brazil, we advocate
the use of this simple classification of responders and non-responders based on NIS
progression of 7 or more points per year from tafamidis D1 after also considering
the cardiac, renal, and nutritional aspects of the disease. In V30M ATTRv-PN patients
who do not respond to tafamidis, we recommend switching to one of the gene-silencing
drugs if available, but LT is still an option for nonresponding V30M patients.
We acknowledge several limitations to the current study. First, the small sample size
may have influenced the results and precluded multivariate analyses. Second, half
of the patients only had 1 follow-up visit at least 6 months after tafamidis initiation.
Third, our center is located in an upper-middle income country, and not all patients
underwent nerve conduction studies at the last follow-up visit given the real-world
scenario. Furthermore, part of the patients were evaluated during the coronavirus
disease 2019 (COVID-19) pandemic. Finally, the present study included only V30M ATTRv-PN
patients; therefore, the efficacy of tafamidis in Brazilian non-V30M patients is still
unknown.
In conclusion, Tafamidis proved to be a safe and effective disease-modifying therapy
in ATTRv-PN, and it has the benefit of being easy to manage by patients and doctors.
It has the most extensive accumulated experience of all ATTRv disease-modifying treatments,
which is important, as patients may have to take the medication for decades. Tafamidis
is also effective and has been approved for ATTRv-CM treatment. The gene-silencing
drugs appear to be more potent, but more expensive, and with more adverse effects.
For being safe, cheaper, and effective, tafamidis still has an important role in the
treatment of ATTRv-PN, especially in very early stages of EO V30M patients.
Bibliographical Record
