Introduction

Pulmonary hypertension (PH) represents a group of conditions which share the haemodynamic definition of a mean pulmonary artery pressure equal or greater than 25mmHg at rest. Acute pulmonary hypertension in critical care may often be secondary to acute respiratory failure, left heart failure and pulmonary thromboembolism or due to decompensated chronic pulmonary hypertension by concurrent comorbid conditions. Pulmonary hypertension can complicate perioperative management in patients admitted for elective surgical procedures to treat their underlying pulmonary hypertension (e.g. pulmonary endarterectomy, lung transplantation).

In the assessment of patients with PH, the pulmonary capillary wedge pressure (PCWP) is of importance in discriminating between pulmonary arterial (precapillary) and venous hypertension (postcapillary). Current guidelines use the cut-off of 15 mmHg to differentiate but it is important to acknowledge the preload dependent nature of the PCWP, which can be affected in particular by fluid status. Pulmonary hypertension is commonly thought of as a rare disease but in reality is commonly associated with varied pathophysiological processes. This is reflected in current classification systems, which highlight just how many diseases can result in PH. Evidence from primary care suggests that up to 13% of all-comers to open access breathlessness clinics may have echocardiographic evidence of PH, and many of these patients remain essentially undiagnosed either due to lack of awareness of the managing physicians or fatalistic attitudes about prognosis. A particular problem lies in the classification categories where treatment with targeting PH therapies has not been of proven benefit, especially groups 2 and 3, pulmonary venous hypertension and hypoxic lung disease. Additionally the current classification system does not consider causes of transient PH in acutely ill patients. There are therefore three distinct patient populations when considering managing PH in the ITU: (1) those with recognised PH, (2) those with occult or unrecognised PH and (3) those who develop PH as a consequence of their acute illness. Compounding this challenge is the paucity of evidence of how to manage any of these differing groups and often the same evidence-free treatment approaches are trialed regardless of aetiology.

Patients with PH most commonly die of right ventricular (RV) failure and its consequences, regardless of whether it is acute or chronic. Any rise in pulmonary vascular resistance (PVR) triggers RV dysfunction, and impaired RV adaptation to rapid increases in afterload results in ‘ventriculoarterial uncoupling’, RV distension, myocardial oxygen consumption/delivery imbalance and heart failure. It is important to recognise that the right and left ventricles (LV) are not separate organs and functionally impact on each other. Not only do they share an interventricular septum, but 25% of the cross-fibres in the RV are common to the LV. Significant RV–LV interdependence exists in the setting of RV pressure and/or volume overload, reduced RV stroke volume, reduced LV venous return leading to impaired LV filling contributing to the fall in cardiac output and heart failure. In acute causes of PH the RV is unprepared to cope with a resultant inability to generate very high pressures and there is an easily breached contractile reserve. In chronic PH the RV ability to ‘cope’ will partly rely on its coupling to the pulmonary circuit. This will be dependent on the degree of disease but also varies between disease processes. More explicitly, how the RV copes is not linearly linked to simple pressure and resistance. This was elegantly demonstrated in a mechanical animal model of chronic progressive RV pressure overload (pulmonary artery banding). Despite identical pressure profiles to an established model of angioproliferative pulmonary hypertension (sugen/hypoxia), the banding model does not develop RV failure but the sugen/hypoxia does. This is borne out clinically most notably in congenital heart disease where suprasystemic pressures can develop but long-term outcomes are generally better than for other forms of PH. Added to this is the recognition that although pulmonary arterial hypertension (PAH) traditionally is thought of as affecting the small precapillary vessels, there is heterogeneity of vascular compartments affected across the spectrum of causes of PH (involving both precapillary and postcapillary compartments). This heterogeneity may in part explain the difficulty translating the success of treating PAH with pulmonary vasodilators to most of the other causes. Therefore all ‘PH’ is not the same from a biomechanical or treatment response perspective.

In this chapter we will clarify what evidence-based medicine and expert consensus is available and the impact on management and treatment strategies in the intensive care unit (ICU). Explicitly we will discuss managing the patient with known PH, and how to approach patient populations at risk.

Supportive Care and General Management of PH in the ICU

Acutely unwell patients need to be moved to a critical care environment. Appropriate investigations and monitoring should aim to identify and reverse precipitating factors, especially those associated with a high mortality.

Optimal fluid balance is paramount, as both hypovolaemia and hypervolaemia may be detrimental to cardiac output. In hypervolaemic patients, reducing RV preload through diuresis will improve RV function, RV-LV interdependence and LV diastolic compliance. Patients with decompensated RV failure develop secondary hyperaldosteronism and addition of an aldosterone antagonist, such as spironolactone, to a loop diuretic can be useful. In the critically ill patient with PH and right heart failure it is a constant and delicate balancing act getting the fluid balance right. The most common mistake is overfilling patients who are already fluid overloaded in an attempt to bolster systemic blood pressure. In the context of right heart failure it is likely that patients will be overfilled and therefore maintaining a negative balance is critical in the majority of cases. Underfilling, however, will affect LV output and systemic blood pressure. We are therefore often left in a situation considering vasopressors whilst trying to offload. Control of pulmonary artery pressure could be improved by treatment with high flow oxygen as a selective pulmonary vasodilator in patients with pulmonary hypertension, regardless of the underlying diagnosis, baseline oxygenation or RV function. Anticoagulation may be desirable due to the low cardiac output states usually seen, and should be considered. There is however no evidence basis for this. Mechanical ventilatory support may be considered depending on the underlying pathology, in particular if there is an identifiable and reversible cause of the PH. If the patient is a potential transplant candidate this should be considered. ECMO is largely currently restricted to bridging patients to transplantation; however its role may evolve.

Treatment of Underlying Causes

The underlying disease process should be improved or optimised. For example, with left heart failure, therapy needs to be optimised, ventilation in lung diseases needs to be optimised and antibiotics used if an infective process is suspected. Specific attention should be paid to avoiding respiratory acidosis and metabolic acidosis. This can be limited by maximisation of recruitment of lung alveoli and optimisation of ventilation-perfusion matching by patient positioning and optimisation of ventilation and positive end expiratory pressure. All attempts should be made to reduce PVR in PAH patients presenting with heart failure. An excellent example of successful management of patients in the ICU with significant pulmonary hypertension is chronic thromboembolic pulmonary hypertension (CTEPH) patients treated with pulmonary endarterectomy (PEA). This represents the largest cohort of PH patients successfully managed in the ICU. Mortality rates in high-volume centres are now down as low as 2–5%. Despite this, these patients are often not considered when assessing the evidence on how to manage patients with established PH. This is understandable as the majority of patients have significant early reduction in pulmonary pressure and resistance, and therefore are not comparable from this perspective to other groups of PH patients. The clearest lesson from PEA is that patients with PH can be successfully managed in the ICU but most notably when the cause of the increased resistance is treatable. In patients where the cause of the PH is unknown or unclear, consideration should be given to rarer causes of PH and investigations tailored to identify potentially reversible causes. The current classification of PH aetiology is delineated in Table 31.1.

To Monitor or Not To Monitor?

The standard assessment of non-invasive haemodynamic and renal, hepatic and neurological function will provide direct and surrogate information on cardiac status, but beyond this there is no good evidence for or against additional monitoring. Echocardiography will potentially be diagnostically useful and can give extra information such as the presence of pericardial effusions but has no clear role in monitoring. This is largely due to the complicated physical nature of the RV, which makes bedside echo assessment of function crude at best. Invasive cardiac monitoring is no longer in vogue in the ICU having proven of no benefit in the management of critically ill general ICU patients or in high-risk postoperative cohorts. There is a lack of trials looking specifically at PH and therefore any advice is based on theory and opinion. However, most expert reviews still recommend invasive monitoring by pulmonary artery catheter (PAC) of PA pressure, cardiac output and PCWP, though not unambiguously. There are non-invasive methods for measuring cardiac output on its own, such as bioimpedance, bioreactance and pulse wave analysis, though again there are no data strongly supporting a role in guiding treatment decisions. Monitoring and basic investigations are outlined in Table 31.2.

Resuscitation In and Outside of the ICU

Patients with PH have been demonstrated to have very poor outcomes in a large but retrospective analysis of tertiary care centres. This included all comers but reflected the nature of specialist centres with 49% of the patients reported having a diagnosis of idiopathic PAH and the majority of the rest being either group 1 PAH or CTEPH. Only eight patients (6% of the survey) survived to 90 days without significant neurological damage. This was despite the very high proportion of 63% of the patients already being located in the ICU. Of these patients, all but one had an identifiable reversible cause for their arrest such as vasovagal reactions, digitalis toxicity or pericardial tamponade. The mean pulmonary vascular resistance in this cohort was very high at 1694 dyne s/cm⁵ and it is probably extremely difficult to achieve effective pulmonary blood flow and left ventricular filling with chest compressions in this context. Therefore in the ICU it is appropriate in all PH patients, and specifically those with PAH, to address early the plans for escalation in the event of an arrest, with a clear understanding that in the absence of reversible causes, success is unlikely.