A retrospective cohort study of idiopathic diaphragmatic palsy: a diagnostic triad, natural history and prognosis

Discussion

Depending on the disease severity, patients with diaphragmatic palsy present with a range of symptoms, including dyspnoea on exertion, reduced exercise tolerance, orthopnoea and dyspnoea or orthopnoea that may be attributed to comorbid conditions such as obesity or cardiopulmonary diseases [7, 11, 16, 17]. When their dyspnoea becomes physically limiting, they usually present to respiratory physicians [18]. Diaphragmatic palsy is not uncommon but there can often be a delay in diagnosis because lung fields are clear on radiological imaging, provided there is no other coexisting pathology. Fluoroscopy can be misleading, as can ultrasound imaging, particularly in cases of BDP. There is often a delay in requesting respiratory muscle strength tests and patients are not always able to perform supine VC manoeuvres. Transdiaphragmatic pressure measurements or electrical/magnetic stimulation of phrenic nerve requires considerable expertise and sophisticated equipment, which are not widely available [9].

The percentage drop in VC from sitting to supine is regarded as a screening test for diaphragmatic palsy with normal values ranging from 3% to 9%, whereas a drop in VC of 10–20% in the supine position is suggestive of UDP [7, 16]. A postural drop in VC from sitting/upright to supine position is generally used to characterise the severity of diaphragmatic palsy [18]. The gravitational shifts in abdominal contents that displace the diaphragm cephalad, a shift in blood volume into the thorax and the mechanical disadvantage of intercostal muscles in the supine position cause this reduction in VC [19]. A fall in VC from sitting to supine of ≥25% indicates diaphragm weakness, with the degree of decline in VC percentage from sitting to supine correlating with the severity of dyspnoea [8, 17].

However, supine measures of VC are hard to perform in those with severe diaphragmatic palsy who are unable to lie flat or are wheelchair-bound. Critically ill patients are also unable to perform supine and seated manoeuvres. Tests of respiratory muscle function that include MEP, MIP and SNIP manoeuvres can easily be performed with patients in the seated position. Our findings support the proposal of Koo et al. [12] that MEP/MIP ≥3 has the potential to be used as a screening tool for diaphragmatic palsy. Our cut-off of 2.7 would provide higher sensitivity than a cut-off of 3.0, and can be extended to MEP/SNIP ratio, but this needs to be confirmed by prospective studies. Screening for diaphragmatic palsy at an earlier stage could lead to more timely referrals to a specialist centre for further management of the condition. It is important to note that MIP and SNIP are not specific to the diaphragm, but include inspiratory force generated by intercostal and accessory muscles. Some of the changes seen in MEP/MIP and MEP/SNIP ratio could be explained by training up of intercostal and accessory muscles, rather than recovery in the strength of the diaphragm itself [20].

Our patients all had diaphragmatic palsy of significant severity to require NIV, even although the diaphragmatic palsy was apparently only unilateral. With the use of NIV, the majority of patients with diaphragmatic palsy experience improvement in their clinical symptoms. Use of NIV allows time for recovery of diaphragmatic strength to occur. Spontaneous recovery is well documented in subjects with diaphragmatic palsy [21], usually occurring within 1 year of diagnosis [22]. The majority of our subjects continued to require NIV 1 year after diagnosis, despite showing some improvement in diaphragmatic function (based on a smaller change in VC from sitting to supine and a reduction in MEP/MIP and MEP/SNIP ratios). Diaphragmatic palsy is well recognised after cardiothoracic surgery, due to injury to the phrenic nerve. Asymptomatic subjects with post-surgical diaphragmatic palsy do not require treatment, but those with symptoms will need NIV. In our cohort, an improvement in clinical symptoms was noted in all patients with post-surgical diaphragmatic palsy following treatment with NIV, but they did not recover fully.

The role of surgical plication of the diaphragm in diaphragmatic palsy remains unclear, but it should probably be delayed until at least a year after diagnosis, in order to allow time for the possibility of spontaneous recovery [23, 24]. Close observation during the first year after diagnosis of diaphragmatic palsy is important in order to assess whether there has been spontaneous recovery, in which case NIV may no longer be required. The patient's circumstances may have changed such that NIV is no longer appropriate, as was the case in several of our subjects. Regular measurement of MEP/SNIP or MEP/MIP ratios would reveal whether there is deterioration of diaphragm strength, or the development of expiratory muscle weakness, prompting careful neurological re-evaluation. None of our patients developed a generalised muscle disorder over the study period, although this may manifest itself until several years after the diagnosis of diaphragmatic palsy. For unexplained orthopnoea with normal chest imaging, a selective reduction in inspiratory pressure (MEP/MIP and/or MEP/SNIP ratio ≥2.7) will be useful for diagnosing isolated BDP. Conversely, reduction in expiratory and inspiratory pressures to similar values would suggest a generalised neurological condition such as motor neurone disease or a myopathy. In patients with sequential diaphragmatic palsy, one hemidiaphragm may be more significantly affected than the other, leading to a radiographic appearance of unilateral palsy. If they demonstrate a >25% drop in VC with MEP/MIP (or MEP/SNIP) ratio ≥2.7, they are likely to have BDP and are more likely to benefit from nocturnal NIV [20]. This highlights the need for checking supine VC and respiratory muscle strength if patients’ symptoms are disproportionate to imaging.