Diagnosing hATTR Amyloidosis With Cardiomyopathy

John L. Berk, MD: Let’s circle back to Akshay at this point. You’ve heard from both Michael and Jim that nerve biopsies are a diagnostic tool and that there are other noninvasive methods for a diagnosis, such as the abdominal fat pad aspirate looking for congophilic material, sweat gland, and potentially salivary gland. Cardiomyopathy is something that’s present in at least 60% of the populations that have been enrolled for neuropathy trials, so we see a full spectrum of TTR [transthyretin amyloidosis] disease in these patients who are being studied. How does cardiomyopathy, as an avenue to diagnosis, help us through the difficult identification of patients?

Akshay S. Desai, MD, MPH: As you pointed out, as cardiologists, the way in which we encounter familial amyloid polyneuropathy patients, or hATTR [hereditary TTR amyloidosis] patients, is very different. We certainly don’t see patients in the very earliest phases of the disease. We see patients who have developed heart failure with some of the characteristic features of amyloid cardiac involvement, and those tend to be patients with unexplained increases in left ventricular wall thickness, particularly the significant appearance of left ventricular hypertrophy in the absence of a stimulus like long-standing hypertension. That heart failure, in the setting of preserved EF [ejection fraction] with a concentrically remodeled heart, often happens in the setting of discordant changes in electrical voltage on the ECTs [electroconvulsive therapies], so that there’s a mass de-voltage discrepancy that we see.

And so I think what we are often left encountering is patients with heart failure and normal ejection fraction with significant wall thickness that’s unanticipated in the context of the clinical history. We’re then searching out the adjuncts of carpal tunnel syndrome and peripheral neuropathy that then heightens suspicion for familial disease.

With regard to the diagnostic approach, we have then tried to increasingly go toward some sort of tissue confirmation to extract the patients with amyloid from the general population, which is quite large in heart failure with preserved ejection fraction. Many patients with hypertrophic cardiomyopathies or other familial infiltrative cardiomyopathies or storage diseases might mimic the clinical or echocardiographic features of patients with amyloid heart disease. So the tissue biopsy becomes pretty important. And as cardiologists, when we’re looking for tissue we’re often looking at endomyocardial biopsy as the approach.

In patients who have had noncardiac tissue biopsy that’s already confirmed the diagnosis—which would mean that it’s a patient coming from 1 of your clinics, for example, to us for evaluation of mild changes in wall thickening or incident heart failure—I think there the enthusiasm for an additional endomyocardial biopsy is less. Although as has been stated, just the fact that somebody has thick walls in the setting of a context that would suggest familial amyloid doesn’t mean that the cardiomyopathy is amyloid-related. So the confirmation would definitively require an endomyocardial biopsy. I think we’ve increasingly seen the value of nuclear medicine scans, particularly technetium pyrophosphate scans, which tend to light up in the heart in patients with TTR that is both wild type and familial as an intermediate between clinical assessment by echo and biopsy. If that scan is highly positive, then that would be quite suggestive.

I think that the cardiac diagnosis is suggested by the context and whatever extra cardiac tissue is available. But also, we see these patients in undifferentiated forms arriving as HFpEF [heart failure with preserved ejection fraction] patients with unexplained hypertrophy, and then are sort of asked to assess out the phenotype.

For cardiologists or people in the community who are seeing a lot of patients with heart failure and normal ejection fraction, I think the key is to have some high index of suspicion for amyloid, in general, so that it’s considered. And then to suspect the familial variance when some of the features that have been so eloquently described by others at the table are present, including peripheral neuropathy findings or autonomic dysfunction or evidence of those features that would suggest amyloid.

John L. Berk, MD: Jim and Michael, as neurologists, a patient who has a length-dependent sensory-motor process arrives on your doorstep. Or during evaluation at your institution, gets a technetium pyrophosphate scan for which the tracer preferentially binds to the heart and is consistent and considered diagnostic of TTR amyloid cardiomyopathy. Is that enough, or do you need a nerve biopsy in that setting as well?

P. James B. Dyck, MD: The answer is, I don’t know. Historically, I would say that was not enough, but I think our practice is evolving. I certainly would want the neuropathy to really seem like it was consistent with hATTR. I don’t think a skin punch biopsy or a fat aspirate is that invasive, so I would probably do those or 1 of those. I probably would not necessarily do nerve biopsies on such a patient.

John L. Berk, MD: Michael?

Michael J. Polydefkis, MD: Yeah, I agree completely.

John L. Berk, MD: So if the next phase III clinical trial was a trial of polyneuropathy and included technetium pyrophosphate scans as a diagnostic of TTR disease, you’re OK with that?

P. James B. Dyck, MD: I was involved in the patisiran study that didn’t require pathology, so yeah. Do I think it’s perfect? No. But again, this is an evolving field, and I think I’m OK with it.

John L. Berk, MD: Michael?

Michael J. Polydefkis, MD: They need to have neuropathy, and they need to have other causes of neuropathy excluded. But if those things are in place, yeah, I think it’s possible.