Late-onset thymidine kinase 2 deficiency: a review of 18 cases

We describe the phenotypic features of 16 Spanish and 2 US patients with mitochondrial myopathy due to mutations in the TK2 gene with the absence of clinical symptoms until the age of 12. The series include three pairs of siblings (P3-P4, P6-P10 and P14-P15). Partial data from five patients have been previously published elsewhere (P1, P5, P9 [7], P3 and P12 [10]).

The electronic records were reviewed to collect information about the age of onset, initial symptoms, severity, distribution and progression of the muscle weakness and extra-muscular symptoms. We gathered information from the latest neurological examination registered including, when available, the Muscle Research Council (MRC) scale to assess the muscle strength and the 6 min walk test (6MWT) for functional evaluation.

The latest value of the forced vital capacity (FVC) in seated and supine position, maximum inspiratory pressure (MIP), blood gas analysis, nocturnal ventilation (assessed with nocturnal pulse oximetry and/or capnography [11] and the need for mechanical ventilation (MV) type and hours of use were recorded.

CK (creatine kinase) and lactate levels were quantified in serum in basal conditions, at diagnosis. GDF-15 (growth/differentiation factor-15) levels were quantified in plasma samples using human GDF-15 quantitative ELISA kit (R&D Biosystems) according to the manufacturer’s instructions.

Muscle MRI was performed in 8 of the 18 patients. All of them were scanned in a 1.5 T MR scanner (Siemens). Lower limb axial T1-weighted sequences were used for morphological analysis and short-tau inversion recovery (STIR) sequences were examined to detect muscle edema. The muscle MRI studies were evaluated by the same neurologist (R F-T) with wide experience in neuromuscular disorders. The evaluator was blind regarding the clinical manifestations. He scored pelvic, thigh and lower leg muscles in axial T1-sequences with the semiquantitative Mercuri visual scale (MVS) modified by Fisher [12]: 0: Normal appearance; 1: Mild involvement, less than 30% of individual muscle volume; 2: Moderate involvement, 30–60% of individual muscle volumes; 3: Severe involvement, > 60% of individual muscle; 4: End stage, all the muscle is severely affected, replaced by increased density of connective tissue and fat, with only a rim of fascia and neurovascular structures distinguishable. We compared median value of muscle fatty replacement using the Wilcoxon-Mann-Whitney test. Statistical analyses were performed using IBM SPSS Statistics, V.22 (IBM, Armonk, New York, USA).

Exercise testing was performed in 5 patients on a cycle ergometer, following a ramp-like protocol (workload increases of 1 W every 6 s [averaging 10 W·min^− 1] starting from an initial load of 0 W, with a pedal cadence of 60–70 rpm throughout the test). Gas-exchange variables were collected breath-by-breath with an automated metabolic cart (Quark CPET, COSMED, Rome, Italy). The peak oxygen uptake (VO₂ peak) was computed as the highest value obtained for any 10-s period during the tests [13].

Muscle samples were obtained by open biopsy and processed following the standard procedures: Hematoxylin and eosin (H&E), modified Gomori trichrome, ATPase (adenosine triphosphatase), NADH (nicotinamide adenine dehydrogenase), SDH (succinate dehydrogenase), COX (cytochrome C oxidase), and COX-SDH stains were performed in all available samples. Respiratory chain enzyme activity levels were recorded when available.

Molecular diagnosis was performed either by direct Sanger sequencing of exons and intron/exon boundaries of the TK2 gene, or by customized next generation sequencing (NGS) panels. Patient’s skeletal muscle mtDNA deletions were investigated by long-range PCR (polymerase chain reaction) and/or Southern blot, and mtDNA copy number was assessed by quantitative PCR as previously described [10, 14].

The study was approved by the institutional review board of every centre and all patients signed an informed consent for the anonymous publication of this data.

We included 18 patients (6 male, 12 female). The mean age-at-onset was 31 years (range 12 to 60 years) with a mean age at diagnosis of 48.5 years (range 23 to 73 years) resulting in an average of 17.4 years between the onset of the disease until reaching a genetic diagnosis (range 1 to 44 years). The mean duration of the disease was 19.8 years (range 6 to 44 years). Four patients from the series were deceased, all of them due to respiratory insufficiency a mean of two decades after the onset.

The first symptom was muscle limb weakness in 10/18 (55.6%), eyelid ptosis in 6/18 (33%) (two patients also presented ophthalmoparesis), and respiratory insufficiency in 2/18 (11.1%). All patients developed muscle weakness during the evolution of the disease, 17/18 showing proximal and distal limb muscle weakness, 1/18 with only distal limb weakness, and 16/18 axial involvement. It is noteworthy that neck flexor weakness was clearly more severe than limb weakness (mean, 2.14 on the MRC scale).

The following muscle groups were the most frequently affected, in a symmetrical manner: shoulder abductor (mean, 4 on the MRC scale), hip flexor (mean, 3.75 on the MRC scale) and hip extensor (mean, 3.87 for both on the MRC scale) and finger extensor muscles (mean, 4.14 on the MRC scale). Four patients (22%) lost the ability to walk without support. Facial musculature was symmetrically affected in 17 patients (94.4%), with predominance of the orbicular oculis muscle. 16/18 of the patients (88.9%) also had symmetrical eyelid ptosis of variable severity, with this being the first symptom in 6 patients (33.3%). Six of them required surgical blepharoplasty due to vision impairment. Nine patients had CPEO.

The majority (11/18) had difficulty in swallowing, which resulted in severe weight loss and/or detriment to the safety of oral feeding in 6 cases, requiring percutaneous gastrostomy tube in 5 cases (27.8%) on average 19.6 years after the onset of the disease (ranging from 12 to 28 years).

Other clinical manifestations included sensory axonal polyneuropathy (7/18;38.9%), neurosensory hearing loss (3/18;16.6%) and dysphonia due to vocal cord palsy (2/18;11.1%). No patient had cardiomyopathy.

FVC at diagnosis from the total cohort was 55.4% (ranging from 17 to 103) with a mean decrease of FVC in supine position of 8% (ranging from 0 to 14), and a mean MIP of 36.8% (ranging from 20 to 101%), independent of the associated muscle symptoms. From a respiratory perspective, the high frequency of complications should be noted, with need for non-invasive MV in 12/18 patients (66.6%). The mean use of the MV was 11.6 h per day (ranging from 8 to 24 h). Eight out of the 12 patients with MV (66.6%) presented with acute respiratory insufficiency following a routine upper respiratory infection as the first manifestation of the disease. None of these cases had any prior respiratory symptoms; however, once detected, they required MV due to hypercapnia secondary to alveolar hypoventilation. Although limb muscle weakness and/or eyelid ptosis were already present at the onset of respiratory insufficiency, those neuromuscular symptoms had not prompted a neurology consultation in any of the eight patients. Thus, the respiratory involvement resulted in the diagnosis of an underlying myopathy in these patients; the mean FVC was of 40.8% (range from 28 to 58) at the time of diagnosis. Of the six patients who did not needed MV, all showed evidence of respiratory muscle weakness on functional tests, although only one of them (P8) reported respiratory symptoms (orthopnoea), suggesting diaphragmatic weakness. This patient displayed ptosis and CPEO at the age of 50 associated with moderate axial and proximal limb muscle weakness (4 on the MRC scale). Strikingly, although the functional respiratory tests and nocturnal pulse oximetry were normal (FVC seated 103%, FVC decubitus 100% and MIP 101%) nocturnal transcutaneous capnography revealed high mean levels of carbon dioxide (CO₂, mean of 48 mmHg, with a maximum peak of 54 mmHg).

Four patients died of respiratory insufficiency at mean age of 56 years (ranging from 40 to 68), and a mean of 24 years after the onset of their initial symptoms (ranging from 17 to 35).

94.4% of patients had increased variable serum CK levels ranging from 190 to 2435 UI/l (normal levels < 170 UI/l)), and 16.7% showed levels 10-fold above the upper normal limit. Serum lactate levels were measured in basal conditions in 12 of the 18 cases. Of these, only three (25%) displayed slightly increased levels (1.4-2x above the upper normal limit).

GDF-15, a biomarker identified in the analysis of transcriptomic profiling of TK2 deficient human skeletal muscle [15], has been proven useful in the diagnosis of mitochondrial myopathies [16], being especially increased in patients with mitochondrial TK2 deficiency [17]. Serum levels of GDF-15 were increased in 5 out of 5 cases analysed (100%), ranging from 1529 to 2438 pg/mL (2113 pg/mL ± 462, mean ± standard deviation, upper limit of normal =550 pg/mL) [16].

It was performed in 8 patients. Mean age at muscle MRI was 46.4 years old (range: 23–73). Mean disease duration at the time of the scan was 18 years (range 10–31). The most severely affected muscles in axial T1-weighted sequences were the gluteus maximus, semitendinosus, sartorius and gastrocnemius medialis (median MVS: 3). Of these, only the gluteus maximus and sartorius were affected in all patients. Apart from the later, gluteus medius, adductor magnus and semitendinosus were also moderately affected in the thighs and gastrocnemius lateralis in the legs (median MVS: 2). No muscle fat infiltration was observed in obturator, quadratus femoris, extensoris digitorum and tibialis posterior (Fig. 1). The fat replacement followed a diffuse pattern and no focal areas of fat infiltration were detected. We did not observe statistical differences regarding asymmetric involvement. STIR sequence was normal in all patients.

In addition to weakness, one of the most frequent clinical manifestations in the mitochondrial myopathies is poor exercise capacity [18].The latter is reflected by low levels of VO₂peak or by poor muscle-oxygen extraction (as assessed with near-infrared spectroscopy) during graded cycle-ergometer/treadmill testing [19]. Aerobic exercise testing was performed on a cycle ergometer in five patients. The mean ± SD VO₂ peak obtained was 14.8 ± 3.2 mL/kg^− 1/min^− 1, with normal consumption values of 40.0 ± 9.5 mL/kg^− 1/min^− 1 [20].

Muscle biopsies were performed in 16 patients, 11 were available to re-analysis. The morphological study revealed numerous ragged-red fibers in 100% of the biopsies, which were hyper-reactive with SDH reaction and usually COX-deficient. COX-deficient fibers accounted for approximately 5–15% of all fibers. Frequently these muscles also showed dystrophic features with frequent necrotic fibers, some with phagocytosis, and increased endomysial connective tissue (present in 7 out of 11 biopsies revised). Marked type I fiber predominance was also observed in 2 patients (Fig. 2). These findings differ from the usual pattern displayed in other mitochondrial myopathies, where the typical signs of mitochondrial proliferation and dysfunction are not associated with other relevant changes in muscle histology structure [21]. We have results of the analysis of the enzymatic activity of respiratory chain complexes of 10 patients. Only in half of them a reduction in the activity of one or more enzymatic complexes were identified (Table 2).

All patients harbored biallelic mutations in the TK2 gene (Ref.Seq. NM_004614.4) (Table 2). Most patients (16/18;88.9%) were homozygous. All mutations were previously reported [7, 8], with the in-frame deletion p.Lys202del (c.604_606AAGdel) being the most frequent (16/36 alleles; 44.4%), followed by the missense mutation p.Thr108Met (c.323C > T) (12/36;27.8%). Additionally, three missense mutations were identified in 3 patients: p.Arg130Trp (c.388C > T), p.Ala139Thr (c.415G > A), and p.Tyr208Cys (c.623A > G). Finally, one patient harbored a frameshift mutation p.Asp157Valfs*11(c.469-470insTGGG) in compound heterozygosis with a splice site mutation c.156 + 6 T > G. Genetic data from patients P1, P2, P5, P9 and P12, were previously reported [7, 10]. Muscle mtDNA copy-number was studied in 9 patients and severe mtDNA depletion was detected in only two (17% of residual mtDNA in P1 and 19% of residual mtDNA in P6). Fourteen out of 14 patients (100%) showed the presence of multiple mtDNA deletions in muscle.