Can Thyroid Hormone Strengthen the Diaphragm in Nonthyroidal Illness?

A significant number of critically ill patients requiring mechanical ventilation have diaphragmatic dysfunction, defined as impaired force generation, with some patients developing it within 24 h after intubation.¹ Diaphragm dysfunction is associated not only with difficulty in liberation from mechanical ventilation, but with increased mortality as well.² This is widely viewed as a serious problem. A number of causal mechanisms have been proposed, including muscle inactivity, excessive diaphragmatic effort, systemic inflammation, oxidative stress, proteolysis, mitochondrial dysfunction, altered signaling pathways, adverse drug effects, and a disrupted metabolic, electrolyte, nutritional, and hormonal milieu.³

It is reasonable to hypothesize that treatments aimed at these mechanisms could improve respiratory muscle function in the short-term and lead to more rapid liberation from the ventilator, shortened ICU stay, reduced mortality, or other, relevant patient-centered outcomes. Thus far, trials have examined phrenic nerve stimulation to prevent muscle atrophy, antioxidant therapy to reduce oxidative stress, anti-inflammatory agents to block the harmful effects from interleukins and other pro-inflammatory mediators, and anti-proteolysis agents.^4–7 Optimal nutrition and positive protein balance have been shown to decrease diaphragm wasting.^8,9 Inspiratory muscle strength training, usually produced using inspiratory threshold loading, may be the intervention best supported by clinical evidence.¹⁰ Yet, despite these studies, none has proved sufficiently effective for clinical implementation. Aside from mechanism-based therapy, pharmacologic treatments could improve diaphragm function sufficiently to impact the duration of ventilation. In this regard, digoxin,¹¹ theophylline,¹² and levosimendan¹³ have long been known to have salutary effects on the diaphragm.

The challenging fact is that many studies hint at beneficial diaphragm effects, yet no treatment has been shown to improve patient-centered outcomes. Very large trials would be required to show the latter, so perhaps it makes sense to reserve clinical trials for only those treatments that are based on a strong pathophysiological connection to diaphragm dysfunction or those that demonstrate large effects on surrogate measures, such as force generation or neuromuscular efficiency.

Several lines of evidence implicate primary hypothyroidism in diaphragmatic weakness. Maximum inspiratory pressure (P_Imax) and maximum expiratory pressure (P_Emax) are negatively correlated with thyroid-stimulating hormone levels.¹⁴ In a similar vein, T3 levels are positively correlated with both P_Imax and P_Emax. Three months of thyroid hormone replacement produces a statistically significant increase in P_Imax, P_Emax, and vital capacity.¹⁵ In patients with hypothyroidism, transdiaphragmatic pressure (P_di) can be reduced, and severe myopathy has been reported.¹⁶ Nonthyroidal illness syndrome (NTIS) is commonly encountered in critical illness and is associated with mortality and the duration of mechanical ventilation.¹⁷ The contribution of NTIS to respiratory failure and whether thyroid hormone replacement improves respiratory muscle function remains unknown, and thyroid hormone replacement in critically ill patients with NTIS has not been shown to improve outcomes.^18,19 The paper by Bello and colleagues²⁰ in this issue of the Journal takes nonthyroidal illness, common in the ICU and potentially linked to respiratory muscle function, and uses diaphragmatic neuromuscular efficiency (NME) as a surrogate end point, seeking to show whether a larger clinical trial is worthwhile.

The investigators enrolled 15 mechanically ventilated subjects with NTIS who failed a spontaneous breathing trial after at least 48 h of ventilation. All subjects had an esophageal catheter to measure the electrical activity of the diaphragm (EA_di). During a brief end-expiratory occlusion, the fall in pressure at the airway opening was measured and assumed to mirror the P_di. NME was calculated as the ratio of the peak negative fall in airway opening pressure and the corresponding EA_di peak.²⁰ If thyroid replacement were to have beneficial effects, one would expect treatment to raise NME (ie, more pressure is produced for similar neural stimulus). In addition, NME was used as an index to convert the full EA_di signal into P_di, allowing the estimation of secondary outcomes including the pressure-time product of the diaphragm per breath (PTP_di/breath) and per minute (PTP_di/min). During the study period, subjects were ventilated using pressure support ventilation guided by neurally adjusted ventilatory assistance and targeted to tidal volumes of 6–8 mL/kg. All other ventilator settings were left unchanged throughout the study.

Thyroid hormone replacement commenced with an intravenous bolus of 0.4 μg/kg T3, then a subsequent 24-h infusion of 0.6 μg/kg. NME was calculated before T3 infusion, then at 3, 6, and 24 h. No significant differences were seen in NME (the primary outcome measure), EA_di, or PTP_di after T3 supplementation, despite normalization of circulating T3 levels. The authors concluded that thyroid hormone replacement did not benefit critically ill subjects with NTIS.²⁰ These results do not support further evaluation of the role of thyroid replacement for patients with NTIS and add to other work suggesting that NTIS may be adaptive, rather than pathological.

A strength of the investigation by Bello et al²⁰ relates to how the study population was selected. Identifying subjects already failing a spontaneous breathing trial seems likely to narrow the population to a subset at risk for prolonged ventilation and for whom diaphragm-targeted therapies could be beneficial. Moreover, NME seems a sensible surrogate metric, so that this negative study should leave little enthusiasm for further study of thyroid replacement in this population.

There are several important limitations of this study. First, the sample size is small, so that modest effects could be missed and large effects could be confounded. Second, NME was estimated for only 24 h after treatment; more delayed effects are possible but would not have been detected due to study design. A further limitation is that NME may not capture fully a hypothetically beneficial impact of thyroid replacement. This is one of the problems in using and choosing surrogate measures of efficacy. Finally, using the fall in airway opening pressure during a brief end-expiratory occlusion, rather than estimating the P_di with an esophageal balloon, could introduce error.

In conclusion, the results of this study suggest that there is little promise in pursuing thyroid replacement for NTIS with a goal of hastening ventilator liberation. As with several other putative therapies, including some with demonstrated impact on short-term measures of strength or fatigue resistance, translation to clinically meaningful benefits seems a remote dream. The study by Bello and colleagues²⁰ points in a helpful direction, however, by showing how to select a high-impact study population and by identifying a broadly applicable surrogate measure. Even when future studies identify candidate treatments, large clinical trials showing improved, patient-centered outcomes (eg, increased ventilator-free days, decreased length of ICU stay, reduced mortality) are needed before these interventions can be introduced into routine care.

• Copyright © 2019 by Daedalus Enterprises