We present a series of 5 novel FA2H-mutated patients without evidence of cerebral iron accumulation. They showed a progressive spastic-ataxia syndrome without dystonia; disease onset was in adolescence in family 1 (mean 14 yr) and in childhood in family 2 (mean, 4 years).
Patients from family 1 presented with the full spectrum of FA2H-related MRI alterations, but iron deposition (Fig. 1). In family 2, isolated WMH without iron deposition was a constant finding, with additional cerebellar atrophy in 1 subject.
Our 5 patients presented with symmetrical T2-weighted and fluid attenuated inversion recovery WMH, predominantly in the posterior, but also in the anterior periventricular regions; sometimes WMH extended to the centrum semiovale and to the subcortical and perirolandic regions; no tract-specific pattern of white matter involvement was observed. Given that FA2H is primarily involved in myelin metabolism and given that WMH could be an early and isolated finding, as in family 2, these alterations are probably a direct consequence of the disease.
None of our patients showed pallidal iron accumulation on T2-weighted sequences; T2* and SWI sequences were additionally performed in patient 1, from family 1, and T2* imaging was also available for patient 2, from family 2. Although T2* and SWI sequences are more sensitive in the detection of cerebral iron deposition, pallidal iron accumulation in NBIA is often also detected as a symmetrical, homogeneous hypointensity on T2-weighted sequences. In previously reported SPG35 cases,[1-4, 8-13] cerebral iron deposition was almost always examined with T2-weighted sequences; in the sole patient with iron deposition and available T2* sequences, the hypointensity was already evident on T2-weighted images.
Although SPG35 was included among the causes of NBIA, the exact frequency of cerebral iron accumulation is still unknown. We therefore performed a review of the literature,[1-4, 8-13] and, after inclusion of the 5 patients described in the present study, a total of 33 FA2H-mutated patients were available for comparison of cerebral MRI data (Table 1): iron accumulation was definitively present in only 11 of 33 patients (33%; stated as mild in 2 of 11), was present with borderline features in 2 of 33 (6%),[1, 9] and absent or not mentioned in 20 of 33 (61%); pallidal iron deposition was more frequently reported in patients with longer disease duration (Table 1).
Notably, according to the literature, other neuroimaging findings were more constantly present: WMH (30 of 33 patients; 91%), cerebellar atrophy (24 of 32; 75%), thin CC (22 of 32; 69%), and brainstem atrophy (15 of 23; 65%); finally, cortical atrophy was reported in 13 of 23 patients (56%). One patient had normal cerebral MRI. Isolated cerebral iron accumulation was never detected in FA2H-mutated patients.
The two families reported on here were identified from a group of patients mainly selected on the basis of the presence of ataxic signs and absence of predominant extrapyramidal presentation; this could constitute a bias with respect to the presence of iron deposition. However, no FA2H mutations were detected in two different cohorts of 43 and 46 PANK2- and PLA2G6-negative NBIA patients, defined by the presence of movement disorders and MRI evidence of iron deposition in the basal ganglia, confirming that FA2H mutations are a rare cause of NBIA.[3, 14] On the contrary, mutations in PLA2G6, also causing cerebellar atrophy, WMH, and basal ganglia iron accumulation, were detected in 20 of 105 NBIA patients (19%) of one of the two cohorts, and pallidal iron accumulation was detected in 50% to 80% of the PLA2G6-mutated patients.
Literature-reported FA2H-mutated patients showed a well-defined progression of WMH and brain atrophy according to disease evolution.[4, 8, 9] The inconsistency of iron accumulation was explained as a tardive finding, possibly appearing mainly in later disease stages.However, in the 2 patients from family 1, subsequent MRI showed a clear progression of brain atrophy and WMH and no iron deposition in T2 and T2* sequences, even after 18 years of disease duration, clearly indicating that at least some patients are not prone to develop iron deposition. Of note, FA2H is primarily and directly involved in metabolism of the myelin; the pathophysiological mechanism of iron accumulation is less clear and may be a possible indirect consequence.
In conclusion cerebral iron accumulation is a very inconstant finding in SPG35, particularly in the initial and intermediate stages of the disease. Its absence should not discourage one from evoking this diagnosis. Notably, iron deposition does not seem to be always dependent on disease duration.