Fabry's disease is an X-linked disorder characterized by deficiency of the lysosomal enzyme α-galactosidase A, resulting in progressive pathologic intracellular glycosphingolipid deposition in different tissues, including the heart. Cardiac involvement is frequent, with left ventricular (LV) hypertrophy being the most common finding, which may mimic the morphologic and clinical features of hypertrophic cardiomyopathy (HCM) [1–3]. Fabry's disease has been reported in 3% of men with LV hypertrophy [1] and in up to 6% of men [2] and 12% of women [3] with late-onset HCM. Moreover, cardiac involvement can be the sole manifestation in female carriers and in male patients with specific gene mutations [1], in whom the nearly normal enzymatic activity and the lack of systemic manifestations make its identification more difficult.

The diagnosis of Fabry's disease has relevant therapeutic implications because enzyme replacement [4] and enzyme enhancement [5] therapy have been revealed to be effective in treating the disease. This now-treatable disease, however, is frequently undiagnosed [1]. To detect patients with Fabry's disease, measurement of α-galactosidase A activity in peripheral blood in patients with LV hypertrophy has been proposed; however, this assessment may be unreliable in female carriers, who frequently experience clinical manifestations of the disease but have normal enzymatic activity [3]. Unfortunately, other single noninvasive tools, such as clinical characteristics or findings at ECG or echocardiography, cannot provide a reliable and definite diagnosis of Fabry's disease. Growing evidence suggests that cardiac MRI with delayed gadolinium enhancement may identify areas of myocardial damage in HCM [6] and in Fabry's disease [7]. A recent study [8] has shown that delayed enhanced cardiac MRI may be useful in the noninvasive recognition of myocardial fibrosis in the setting of cardiac involvement of Fabry's disease. However, the effectiveness of delayed enhanced cardiac MRI in differentiating Fabry's disease from HCM remains unknown.

In our study, we compared the delayed enhanced cardiac MRI features of patients with Fabry's disease–related and of those with non-Fabry's disease–related symmetric HCM to identify noninvasive imaging hallmarks in patients affected by Fabry's disease.

The patients' characteristics and the LV systolic function parameters are shown in Table 1. No significant difference was observed between the two groups for ejection fraction (Fabry's disease, 70.9 ± 7.8%; symmetric HC, 71 ± 9.7%; p = not significant), stroke volume (Fabry's disease, 81.2 ± 15.8 mL; symmetric HCM, 86.4 ± 31.6 mL; p = not significant), cardiac output (Fabry's disease, 5.1 ± 0.9 L/min; symmetric HCM, 6 ± 2 L/min; p = not significant), end-diastolic volume (Fabry's disease, 114.8 ± 21.2 mL; symmetric HCM, 122.2 ± 43 mL; p = not significant), end-systolic volume (Fabry's disease, 33.4 ± 12.6 mL; symmetric HCM, 35 ± 16.6 mL; p = not significant), and heart rate (Fabry's disease, 63.2 ± 6.2 beats per minute [bpm]; symmetric HCM, 71.5 ± 13.8 bpm; p = not significant). The symmetric HCM patients had a significantly higher mass than Fabry's disease patients (Fabry's disease, 184.8 ± 53.6 g; symmetric HCM, 216.2 ± 80.1 g; p < 0.05).

Delayed enhancement presence, percentage, location, and distribution patterns are shown in Table 2. Ten of 13 (77%) patients with Fabry's disease and nine of 10 (90%) patients with symmetric HCM showed areas of delayed enhancement in the LV wall. Of the 10 Fabry's disease patients with delayed enhancement, seven had thick mesocardial striae of delayed enhancement and 8.4 ± 4.8% of the LV mass was involved. The remaining three of 10 Fabry's disease patients with delayed enhancement had small intramyocardial foci of delayed enhancement and a less severe myocardial involvement, 1.2 ± 0.4% of the LV mass (Table 2).

As shown in Table 2, the 10 Fabry's disease patients presenting areas of delayed enhancement showed a consistently seen distribution and pattern of delayed enhancement. In particular, in these patients the delayed enhancement was always located in the inferolateral wall of the LV, with an isolated basal involvement in three patients and with an extension to the mid cavity segments in the remaining seven patients. A delayed enhancement pattern was also consistently seen; in fact, the distribution was mesocardial in all patients with Fabry's disease, with at least a thin portion of subendocardial and subepicardial layer always spared. Only two patients with Fabry's disease presented a second site of delayed enhancement; in fact, striae of delayed enhancement were evident in the apical anterior and apical lateral walls in one patient, and small foci of delayed enhancement were seen in the mid ventricular inferior LV wall in another patient.

In the Fabry's disease group, the percentage of myocardium involved was not significantly different between the men and women (men, 5.2 ± 2.9%; women, 7.3 ± 7.0%; p = not significant). In the Fabry's disease patients with delayed enhancement, the inferolateral LV wall was involved with a mesocardial distribution in all the cases; the subendocardial layer of the LV wall was always spared (Figs. 1A, 1B, 1C, 1D, 2A, and 2B).

In non-Fabry's disease–related symmetric HCM, the areas of delayed enhancement had variable locations (Figs. 3A, 3B, 4A, and 4B). As shown in Table 2, the most frequently involved segment (six of nine patients) was the inferior LV wall. Only one patient with symmetric HCM showed isolated LV lateral wall involvement, but with a subendocardial distribution. The distribution pattern of delayed enhancement in patients with symmetric HCM was variable as well. In four patients the pattern was subendocardial, in four it was mesocardial, and in one the anterior papillary muscle was involved. The percentage of myocardium involved was not significantly different between the groups (Fabry's disease, 6.2 ± 5.2%; symmetric HCM, 8.8 ± 8.9%; p = not significant).

Although some reports in the literature show cardiac delayed enhancement in patients with Fabry's disease and HCM [6–8], no systematic study of the differential diagnosis based on delayed enhancement of these two forms of myocardial hypertrophy is available. The correct diagnosis is important because recent advances in the treatment of Fabry's disease may offer stabilization and reversal of the cardiovascular manifestations [4, 5]. Fabry's disease is usually indistinguishable from symmetric HCM in all noninvasive and invasive cardiac examinations except endomyocardial biopsy; moreover, in most cases, the clinical features are similar [3].

Our study documented, for the first time, that with delayed enhanced cardiac MRI it is possible to differentiate patients with symmetric HCM from patients with Fabry's disease. As we have already pointed out, our data are important because they may help to differentiate between the typical forms of concentric Fabry's disease and HCM; furthermore, we believe that our findings could also be useful in the evaluation of the asymmetric forms. In fact, patients with Fabry's disease can have asymmetric septal hypertrophy or an eccentric pattern of hypertrophy that is indistinguishable from the asymmetric forms of HCM. It has also been shown [2] that patients with Fabry's disease may have an outflow gradient in association with concentric hypertrophy that is typical of HCM. Another interesting aspect of our results is the presence of delayed enhancement with specific distribution in Fabry's disease in heterozygotic women, in whom cardiac hypertrophy could be the only manifestation of the disease.

The data concerning cardiac delayed enhancement in patients with Fabry's disease presented in this article are consistent with those reported by Moon et al. [7]. In fact, the mean percentage of involved myocardium(6.2% ± 5.2%) of Fabry's disease in our study was similar to that found in their study (6.7% ± 5.1%) [7]. Moreover, in our patients as well as in theirs, delayed enhancement occurred mostly in the basal inferolateral LV wall. In particular, patients with Fabry's disease in our group always showed involvement of the inferolateral LV wall with a mesocardial distribution. In patients with symmetric HCM, the basal lateral segments may be involved. In fact, in our series, three patients with symmetric HCM had basal inferolateral involvement; however, in two patients this involvement was associated with large areas of delayed enhancement located in different segments; and in the remaining patient, the delayed enhancement involved the subendocardium, which was always spared in patients with Fabry's disease.

Delayed enhancement represents areas of myocardial collagen scarring both in Fabry's disease and in HCM, as shown in previous studies [6, 8] with a whole-heart histologic validation. The causes of the different scar distributions between Fabry's disease and HCM patients are not clear, but probably they reflect a different pathophysiologic origin of fibrosis. In fact, histologic evidence suggests that in HCM the areas of fibrosis are the final sign of intramyocardial ischemic events; for Fabry's disease, it has been hypothesized that cardiac involvement might predispose to subclinical myocarditis leading to tissue injury, or it could impair myocardial resistance to physical stress that is most prominent in the inferolateral wall of the LV [6, 8].

Other authors [10] recently showed a different MR approach in the evaluation of cardiac involvement in Fabry's disease, adopting a black blood multiecho multishot sequence. They showed that the myocardial T2 relaxation time is prolonged in patients with Fabry's disease compared with that in hypertrophic patients. They pointed out that this technique was useful for the characterization of myocardial tissue in Fabry's disease; such interesting results suggest that this sequence could be complementary to the delayed enhanced technique used in our study to improve the ability of MRI to characterize cardiac involvement in Fabry's disease. The combination of the two techniques should be confirmed prospectively in a large group of patients.

One limitation of our study is that we have not included other specific cardiomyopathies that may be associated with LV hypertrophy, such as sarcoidosis or amyloidosis. The number of patients affected by these diseases that were proven at endomyocardial biopsy in our MR database in the time interval of the study was too low to allow any relevant comparison. Besides, as shown by Mahrholdt et al.[11], it is possible to find specific location and distribution patterns of delayed enhancement in other cardiomyopathies also. Recent articles [12, 13] have shown that cardiac delayed enhancement has a specific distribution in patients with histologically proven cardiac amyloidosis [12] and with cardiac sarcoidosis [13], suggesting a peculiar pattern for these two diseases and making the differential diagnosis possible on the basis of the distribution of delayed enhancement.

In conclusion, the differentiation of LV hypertrophy caused by Fabry's disease or nonobstructive symmetric HCM can be difficult. No single finding in clinical characteristics, ECG, or echocardiographic evaluations is completely diagnostically accurate. The evaluation of the myocardial location and distribution patterns of delayed enhancement helps in the identification of the two causes of LV hypertrophy. The differential diagnosis of the two diseases has relevant clinical implications because the commonly adopted pharmacologic therapy in HCM may be contraindicated in patients with Fabry's disease. In addition, enzyme therapy has proven to be effective in male and even in female heterozygotes.

Address correspondence to F. De Cobelli ([email protected]).

This is a Web exclusive article.