Table 1.
Demographic data of carriers and patients with familial amyloid polyneuropathy.
Fig 1.
Stimulus-response properties of the ulnar nerve.
A and B, Stimulus-response curves from the ulnar nerve of carriers (blue open squares), patients (orange open triangles) and two groups of age-matched controls, NC1 (black filled squares) for carriers and NC2 (gray filled triangles) for patients, are shown [The lines connecting the data points are drawn by hand]. The stimulant current to obtain 50% of CMAP (stimulus for 50% depolarization) in carriers was significantly higher for carriers (7.02±0.56 mA) than for NC1 (5.24±0.29 mA) (p = 0.0096). C, The motor threshold in patients was not significantly increased, as compared to NC2 (6.05±0.43 mA for patients v.s 5.53±0.33 mA for NC2, p = 0.33). In sensory axons, the stimulus for 50% depolarization was significantly increased in both carriers (5.5±0.52 mA for carriers v.s. 3.9±0.19 mA for NC1, p = 0.0035) and patients (6.34±0.79 mA for patients v.s. 4.74±0.31 mA for NC2, p = 0.04). D, There are no significant differences in peak responses of both CMAP and SNAP between carriers and NC1. However, the CMAP (3.4±0.6 mV) and SNAP (15.9±2.2 μV) for patients are significantly smaller than that of NC2 (11±0.4 mV and 56.6±4 μV, both p < 0.0001). *: p<0.05, **: p <0.01, ***: p <0.001, ****: p <0.0001.
Fig 2.
Strength-duration properties of the ulnar nerve.
A and B, The threshold charge-stimulus width curves for motor and sensory axons are shown (the meaning of the symbols are the same as in Fig 1). The slope of the line represents the rheobase and the absolute value of x-intercept is equivalent to the strength duration time constant (SDTC). The lines are linear regression fits to the data points. In sensory axons, the SDTC is significantly decreased in carriers (0.46±0.03 ms for carriers v.s. 0.55±0.02 ms for NC1, p = 0.03), but no significant difference between patients (0.49±0.05 ms) and NC2 (0.51±0.02 ms, p = 0.39). On the other hand, there is no significant change of SDTC in motor axons. C and D, In ulnar motor axons of carriers, the threshold in very short duration (0.2 ms) (15.02±1.39 mA for carriers vs. 11.13±0.60 mA for NC1, p = 0.015) and the rheobase (4.92±0.42 mA for carriers vs. 3.58±0.21 mA for NC1, p = 0.0082) are significantly increased. There is no significant change in both threshold in short stimulus width (0.2 ms) and rheobase in motor axons of patients. In sensory of axons of carriers, the increases of both threshold in short stimulus width (0.1ms) (17.76±1.98 mA for carriers vs. 13.26±0.89 mA for NC1, p = 0.025) and rheobase (2.84±0.41mA for carriers vs. 1.75±0.11 mA for NC1, p = 0.0058) are significant. *: p<0.05, **: p <0.01.
Fig 3.
Recovery cycle analysis and regression models demonstrating the refractoriness and relatively refractory period (RRP) against age.
A, Compared to NC1 and NC2 (the meaning of symbols are the same as in Fig 1), the refractoriness (at 2.5 ms, 91.8±28.0% for patient vs. 27.1±4.5% for NC2, p = 0.007) and relatively refractory period (RRP) (3.59±0.18% for patient vs. 3.04±0.07% for NC2, p = 0.0083) are significantly elevated in motor axons of patients, but not in carriers. The late subexcitability is also decreased in patients (9.9±0.6% for patients vs. 12.3±0.9% for NC2). [The lines connecting the data points are drawn by hand.] B, In the sensory axons, the refractoriness in patients is not markedly increased until interstimulus interval was 2 ms (129.2±20.6% for patients vs. 62.6±5.6% for NC2, p<0.01). The superexcitability is also significantly increased in patients (-25.7±2.6% for patients vs. -18.2±1% NC1, p = 0.011) [The lines are drawn by connecting the mean value of each spot.]. C and D, The refractoriness and RRP are increased by age and the evolution of disease (from carrier to patients). The gray band is the 95% confidence interval for controls. The quadratic regression model (Y = 72.04+5.03*X+0.082*X2) fits the data with a R2 value (0.2159), demonstrating the correlation between the refractoriness and progress (severity) of the disease. On the other hand, the correlation between RRP and progress of disease is fitted by the linear regression model (Y = 0.02563*X + 1.892) with a R2 value of 0.257.
Table 2.
Comparison of nerve excitability properties of ulnar nerves among controls, carriers and patients with familial amyloid polyneuropathy.
Fig 4.
Threshold electrotonus (TE) and I/V relationship of the ulnar nerve.
A and C, Compared to NC2, the threshold elevation in hyperpolarizing TE of motor axons is reduced in patients (TEh(10-20ms) and TEh(20-40ms), -65.2±2.0 mV and -80.4±2.6 mV for patients vs. -71.6±1.0 mV and -89.2±1.3 mV for NC2, p = 0.0043 and 0.0056, respectively). The recovery from hyperpolarization (TEh(slope 101-140ms)) is slightly less steep in patients (1.56±0.1 for patients vs. 1.78±0.04 for NC2, p = 0.049). However, it is slightly steeper in carriers (1.86±0.07 for carriers v.s. 1.72±0.04 for NC1, p = 0.039). In carriers, the threshold reduction is reduced in depolarizing TE (TEd(10-20ms) and TEd(90-100ms), 63.3±0.9 mV and 42.6±1.4 mV for carriers vs. 65.7±0.7 mV and 45.5±0.6 mV for NC1, p = 0.037 and 0.07, respectively). B and D, Similar to findings in motor axons, the slope for recovery from hyperpolarization, TEh(slope 101-140ms) is also significantly decreased in sensory axons (1.87±0.13 for patient vs. 2.41±0.06 for NC2, p = 0.002), along with significantly prolonged time to overshoot (49.0±5.0 ms for patients vs. 37.2±1.9 ms for NC2, p = 0.024). Threshold reduction in depolarizing TE in sensory axons is increased in patients (TEd(10-20ms), 75.3±4.1 for patient vs. 60.5±0.8 for NC2, p = 0.001). E and F, After 200 ms of preconditioning depolarization, there are significantly less threshold reduction in motor axons (47.6±1.4 for carrier vs. 51.6±0.6 for NC1, p = 0.014; 47.5±1.9 for patient vs. 51.4±1.1 for NC2, p = 0.024). The hyperpolarizing I/V slope is increased in both motor (0.451±0.039 for patient vs. 0.352±0.012 for NC2, p = 0.012) and sensory axons (0.446±0.03 for patient vs. 0.347±0.015 for NC2, p = 0.02) of patients. The lines in parts A to F are all drawn by hand.
Table 3.
Regression analysis to identify the relationship between the nerve excitability parameters and NDS as well as ONLS.
Fig 5.
Computer modelling of the properties of nerve excitability (threshold electrotonus and recovery cycle).
A-D, The results from nerve excitability test of motor and sensory axons from NC2 (the meaning of symbols are the same as in Fig 1) and patients. E and F, The gray lines are the simulated excitability curves for NC2, and the orange lines are for patients. For motor axons, the nodal sodium permeability is reduced from 4.1 to 3.75 (cm-3 x 109) to simulate the test results shown in A and B. Based on the motor nerve NET data from patients, we identified three parameters with little discrepancy that are the leak current conductance (1.38 to 2.7 nS), hyperpolarization-activated conductance (6.05 to 7.85 nS) and decrease of nodal transient sodium current permeability (4.1 to 3.75 cm-3 x 109). Best fits with changes in these parameters can reduce the discrepancy of 21.7%, 20.9% and 19.8%, respectively. Reduction in sodium currents permeability (4.1 to 3.75 cm-3 x 109) causes highest increase of threshold (7.3%), which is the characteristic feature in motor NET findings of patients. In contrast, increase of leak current or hyperpolarization-activated conductance would cause a prominent reduction of threshold elevation after hyperpolarization in I/V curve, which is not found in our patients (Data not shown). G and H, For sensory axon simulation, increase of capacitance upon internodal membrane from 0.196 to 0.273 nF is used to simulate the tested results in C and D. Two parameters in computer models may be changed to fit the results from sensory NET findings from patients. They are the pumping currents and internodal capacitance. With 8.5% reduction of transient sodium channel permeability, increase of the nodal and internodal pumping currents from 11.8 to 21.7 pA and of the internodal capacitance from 0.196 to 0.273 nF could reduce the discrepancy of 68.4% and 38.7%, respectively. The increase in pumping currents, however, is physiologically difficult to envisage in our case (See Discussion). The findings with decrease of the TEh(slope 101-140ms) and protracted the time to overshoot in hyperpolarization TE as shown in patients NET findings (in part C) can be replicated in the model with increased internodal capacitance. Similarly, the modeling with increased internodal capacitance also well describes the increase of refractoriness and superexcitability from patients (in part D). The computer simulation demonstrates the likelihood of early changes in nodal sodium conductance and internodal capacitance, but would by no means rule out concomitant minor alterations in the other axonal membrane conductances. The error bars indicate the standard error of mean (S.E.M.). The lines in part-A to -D are drawn by hand.