A complex obstetric case

Abstract

A 33-year-old primigravida presented with severe sepsis, severe pre-eclampsia, peripartum cardiomyopathy and the haemolysis, elevated liver enzymes and low platelets syndrome manifesting over the course of 24 h causing a diagnostic conundrum and a difficult sequence of physiological problems to overcome. We describe a previously unreported sequence of events involving a pre-eclamptic, septic parturient to improve anaesthetic and intensive care physician awareness of confounding factors that complicate assessment and management of these patients.

Background

Pre-eclampsia is a common and potentially fatal condition associated with the later stages of pregnancy. The pathophysiology is believed to relate to imperfect trophoblastic implantation of the placenta to the uterus. This results in endothelial dysfunction, not localised to the uteroplacental unit, causing maternal hypertension and proteinuria. A spectrum of illness is described ranging from mild through severe, qualified by diagnostic parameters or the presence of other organ involvement: namely eclamptic seizures; acute kidney injury; or haemolysis, elevated liver enzymes and low platelets (HELLP) syndrome. Again, the pathophysiology of these complications is well studied yet poorly defined with multiple theories – many of which centre on endothelial dysfunction and vasoconstriction. A summary of pre-eclampsia is available in multiple review articles. This report describes a near-fatal combination of complications of pregnancy.

Peripartum cardiomyopathy is an uncommon complication of pregnancy historically occurring in around 1:4000 deliveries. ¹ The incidence is believed to have been rising over the past decade due to increasing maternal age with rates approaching 1:2000 – such that the average delivery unit will be encountering several cases each year.^2,3 The diagnosis is based on exclusion of other causes of cardiomyopathy in the setting of late pregnancy. There is an increased risk of peripartum cardiomyopathy in pre-eclampsia and some papers describe it as a direct consequence thereof. The illness mandates a left ventricular ejection fraction (LVEF) of less than 45% developing in the last month of pregnancy or the first 5 weeks post-partum in the absence of other causes. The pathophysiology is unknown with multiple theories including it being the cardiac manifestation of pre-eclampsia or the consequence of an autoimmune myocarditis.^4,5 An increasing focus is the role of abnormal cleavage of prolactin in the near-term parturient. Genetically modified mice producing an atypical 16 kDa prolactin fragment show altered vascular endothelium leading to myocardial dysfunction, remodelling and symptomatic peripartum cardiomyopathy. ⁶ The same raised prolactin derivative has been found in women with peripartum cardiomyopathy and has opened up a therapeutic target through the use of bromocriptine (a dopamine agonist that reduces prolactin production). ⁷ Managing one of these conditions can be challenging but a patient with a combination of these conflicting physiological changes causes difficult diagnostic and therapeutic dilemmas.

Case report

A 33-year-old primigravida with an in vitro fertilisation pregnancy using a donor egg presented at Term + 7 with a 2-week history of coryzal symptoms and pyrexia. Her past medical history consisted of a previous deep vein thrombosis whilst taking the oral contraceptive pill for which she had been prophylactically anticoagulated during pregnancy.

On initial presentation, our patient had evidence of moderate pre-eclampsia (blood pressure (BP) 150/80 mmHg; urinary protein-creatinine ratio (PCR) 216) with a pyrexia (temperature 38.7℃) and raised inflammatory markers (C-reactive protein (CRP) 109 mg/dl; white cell count 12.9 × 10⁹/L). She was admitted to the labour ward for observation, antibiotics (IV Cefuroxime (1.5 g), Metronidazole (500 mg) 8-hourly as per obstetric guidelines), and induction of labour (Propess 10 mg). Throughout her induction, she was tachycardic and tachypnoeic with a productive cough and pyrexia, whilst her systolic BP remained roughly between 140 and 150 mmHg.

Pre-eclampsia protocols mandate strict fluid monitoring (with fluid restriction in the case of severe pre-eclampsia – whose criteria our patient did not fulfil), whilst sepsis protocols advocate targeted fluid challenges. Following the departmental sepsis algorithm, our patient received 11 L of IV crystalloid over 96 h and IV antibiotics for 4 days. She failed induction of labour and underwent a Category 3 lower segment caesarean section (LSCS) under combined spinal-epidural delivering a healthy baby boy on the fifth day.

Following delivery, she was transferred to obstetric recovery where she deteriorated significantly over the ensuing 24 h, becoming increasingly tachypnoeic (RR 30–40 breaths per minute) and tachycardic (HR 130–150 beats per minute (bpm)) with an increasing oxygen requirement. BP remained around 150/95 mmHg. She was reviewed multiple times by both medical, obstetric and anaesthetic teams. Diagnoses considered at the time included possible pulmonary embolus (PE) and worsening of her chest sepsis. A computed tomography pulmonary angiogram (CTPA) showed bilateral consolidation and significant interstitial oedema with no PE. Following this, she was admitted to the intensive care unit (ICU) due to severe hypoxaemia.

On admission to ICU, she was in pulmonary oedema, requiring 15 L/min O₂ therapy via non-rebreather mask to maintain SpO₂ > 94% with a respiratory rate of 40 breaths per minute. Her heart rate was 140 bpm and BP 192/114 mmHg – a diagnosis of severe pre-eclampsia was made. Her Glasgow Coma Scale score was 15 throughout. Considering our patient did not have a PE but had hypertension with consolidation; pre-eclampsia and sepsis alone did not adequately account for the clinical severity of her pulmonary oedema as manifested by severe hypoxia and copious pink frothy sputum. On this basis, an echocardiogram was requested. However, due to her rapid deterioration it could not be performed immediately.

Knowing a diagnosis of severe pre-eclampsia at this time, management should involve BP control and MgSO₄ therapy. However, given the possible need for intubation and the potential cardiovascular changes anticipated at induction, these were withheld and she was trialled on continuous positive airway pressure ventilation (CPAP + 10 cm H₂O, FiO₂ 1.0).

Unfortunately, she deteriorated with a HR of 180 bpm (sinus rhythm), BP of 190/130 mmHg and falling peripheral oxygen saturations to 70–80% despite the aforementioned CPAP settings. She underwent peri-arrest (secondary to hypoxaemia) intubation (Propofol 50 mg, Fentanyl 150 mcg, Rocuronium 70 mg, Propofol 50 mg) with a brief (pulseless electrical activity) cardiac arrest on induction. Return of spontaneous circulation was quickly obtained (<40 s) through advanced life support protocols with 1 mg adrenaline given. Persistent hypotension followed, resistant to noradrenaline (0.3 mcg/kg/min) and intravascular volume resuscitation.

A transthoracic echocardiogram (TTE) was performed to assess ventricular function, which showed an equally dilated LV/RV with global severe systolic impairment and LVEF < 12%, implicating cardiomyopathy as a significant contributor to her deterioration.

Oesophageal Doppler monitoring was used to titrate further inotropic support sequentially: adrenaline (0.1 mcg/kg/min), dobutamine (15 mcg/kg/min) and enoximone (5 mcg/kg/min), the emphasis on inotropy. Optimised treatment achieved stroke volume 27 ml, corrected flow time 300 ms and a cardiac output of 3.7 L/min. Early liaison with a tertiary cardiothoracic centre prompted intra-aortic balloon pump insertion and transfer to regional heart transplant centre for further management.

During the above 12-h period, the patient’s platelet count fell from 172 to 48 × 10⁹/L, whilst hepatic transaminases rose from 89 to 470 IU/L with development of coagulopathy (International Normalised Ratio 1.6) and resistant hypoglycaemia. There was a mild anaemia (haemoglobin concentration of 11.2 g/dl) and a partial HELLP syndrome was diagnosed. ⁸ Supportive management was initiated with emphasis placed on the optimisation of cardiac output. Once her haemodynamic status had stabilised, she was commenced on an MgSO₄ infusion in an attempt to mitigate her pre-eclampsia (4 g IV bolus followed by 1 g/h infusion for 24 h).

Following arrival at the cardiac centre, she was commenced on bromocriptine. Other cardiovascular support was slowly weaned. Her liver function deteriorated for the first 36 h with transaminases peaking > 1000 IU/L before slowly improving. Inotropes were weaned and LVEF recovered to 20% by the end of 2 weeks. She was then extubated and discharged to the medical ward and subsequently home approximately 5 weeks after her initial hospital presentation. Two months after discharge, follow-up TTE showed a LVEF of 68% with normal sized chambers.

Discussion

The septic, pre-eclamptic patient

Sepsis and pre-eclampsia are both common enough pathologies that the two co-present. The pre-eclamptic exists in a state of hypertensive vasoconstriction with endothelial dysfunction predisposing to renal proteinuria and peripheral oedema. As a consequence of this, patients tend to be intravascularly fluid deplete. The septic patient also experiences endothelial aberrance but develops hypotension due to fluid deficit and cytokine-mediated vasodilatation. The common theme of these two syndromes is an intravascularly hypovolaemic patient. Cardiovascular collapse can be seen on induction of anaesthesia both due to further vasodilatation secondary to the centrally mediated actions of anaesthetic drugs combined with the abrupt reduction in venous return seen in the transition to positive pressure ventilation. In this combination of contrasting physiologies, how does one interpret a reasonably normotensive patient – are either both, or neither, pathologies at work?

Sepsis remains a significant cause of maternal mortality in the CEMACH/MBRACE reports. It is possible that the hypertension of pre-eclampsia can compensate for what would otherwise be sepsis-induced hypotension. In mild cases, adverse consequences of this may not be seen but at the severe end of both spectra, there is potential for significant organ dysfunction, albeit with a seemingly normal blood pressure.

Whilst early management of sepsis should continue as per the Surviving Sepsis Guidelines, the liberal administration of fluid may be poorly tolerated in pre-eclampsia. Lack of experience and confidence in the management of these combined conditions can lead to suboptimal management of both and patient deterioration as a consequence. There is a good argument to be made for early ICU input for invasive monitoring to optimise cardiac output.

Peripartum cardiomyopathy

The presence of a dilated, cardiomyopathic heart in a previously healthy obstetric patient is, thankfully, uncommon. Differentials for this form of acute heart failure include septic and peripartum cardiomyopathies.

Septic cardiomyopathy is not uncommon but characteristically forms a part of multi-organ dysfunction seen in advanced sepsis in a patient who is usually on multiple vasopressors/inotropes, more associated with Gram negative sepsis. Management often follows a sequence of preload optimisation, inotropic improvement of contractility and as a later line, inodilation. Source control should be achieved.

Peripartum cardiomyopathy is a rare complication of pregnancy and is thought to be slightly more common in pre-eclampsia (although may simply be more symptomatic). Presentation is that of heart failure with symptoms including low BP and a tendency to both peripheral and pulmonary oedema. Management aims to optimise contractility whilst minimising further myocardial damage.

Hypotension may be poorly tolerated by the uteroplacental unit and augmentation of left ventricular function should be the primary goal.

Cardiomyopathy in the pre-eclamptic patient

A number of case reports describe the antepartum diagnosis of peripartum cardiomyopathy and the management through late pregnancy and labour. There is no significant treatment difference between this and any acute heart failure other than the avoidance of teratogenic drugs pre-delivery. The role for bromocriptine is increasing to ameliorate prolactin-related pathology and should be discussed promptly with local cardiac centres. Mechanical left ventricular support should be employed early as recovery may be slow and can be comorbidity-dependent. As a consequence of the vasoconstriction associated with pre-eclampsia, this subgroup of patients typically presents with the extremes of low ejection fractions resulting from the increased workload placed on an already failing ventricle. However, providing the patient survives the initial period, there is subsequently a rapid improvement in function following delivery once pre-eclampsia starts to resolve as subsidence of the vasoconstriction augments LV recovery making a strong case for pre-emptive delivery.

As large studies are lacking, figures defining overall prognosis are variable and based on small case-series. Approximately half of patients will make a functional recovery back to baseline although show inducible cardiac dysfunction; a quarter of patients will go on to live in a state of pharmacologically managed heart failure but be reasonably asymptomatic; whilst the final quarter develop long-term symptomatic heart failure and may require cardiac transplantation if they survive the initial illness.^9–11 It is likely that a lot of milder cases go undiagnosed.

Diagnosing cardiomyopathy in the presence of pre-eclampsia

Pre-eclampsia is associated with poor tolerance of fluid overload and pulmonary oedema is well described. In our patient’s case, the combination of chest sepsis and pre-eclampsia explained her symptoms through the early stages and a hunt for further diagnoses was, not inappropriately, delayed. With retrospect, her deteriorations both occurred following periods of supination, specifically her LSCS and CTPA.

The role of B-natriuretic peptide (BNP) has been suggested as a screening technique and its levels in normal pregnancy should not be significantly elevated (up to twice normal). ¹² In the absence of pre-eclampsia, a BNP level of < 100 pg/ml has a negative predictive value of 100%. ¹³ However, BNP levels rise significantly in pre-eclampsia making its interpretation in the investigation of cardiomyopathy less useful and warrants discussion with cardiologists. ¹⁴

The definitive investigation is echocardiography although this may be difficult in this group of patients: the gravid uterus renders transthoracic imaging significantly more uncomfortable for the parturient whilst impairing image quality; a transoesophageal echo is not a low-risk intervention in late pregnancy with high regurgitation and aspiration risk, not to mention difficulties with positioning.

(H)ELLP syndrome

In the critically ill patient, multi-organ failure can involve both the liver and bone marrow. Severe sepsis is a common cause of this. A complication of pre-eclampsia includes HELLP syndrome but this diagnosis needs to be made with consideration to other causes of the involved triad. A review article of HELLP syndrome and its imitators provides a fascinating summary of the similarities and differences between these. ⁸ Preservation of renal function in our patient counts against sepsis being the cause of the liver dysfunction together with a normal central venous pressure refuting a cardiac component to this. Minimal anaemia may occur in up to a third of HELLP patients (as in our case). The antecedence of the transaminitis in relation to cardiovascular deterioration implies it worsened concomitantly with the pre-eclampsia rather than being a consequence of multi-organ failure.

Combined pathologies

The interaction between the vasoconstriction of pre-eclampsia, the vasodilatation of sepsis and peripartum cardiomyopathy will be unpredictable and dependent on the severity of each component. Ultimately, it will make clinical measurement less reliable in the assessment of each one – leading to a dependence on biochemical and radiological investigations. In our patient, the presence of chest sepsis with pre-eclampsia was considered sufficient explanation for her symptoms until she failed to respond to vasopressors/fluids. BP maintenance through her early illness led to an assumption that her sepsis was mild despite a CRP stubbornly between 100 and 150 mg/dL on antibiotics. Postoperatively, this rose to > 300 mg/dL, but it is not possible to say whether this was related purely to sepsis or to what degree there was post-surgical change involved.

The profound hypertension that she developed in extremis is another confounder that could lead one to discount cardiac failure and sepsis as co-existent pathologies. This may represent the desperate catecholaminergic response to her severe hypoxaemia compounding the decompensation of her impaired LV and further exacerbating her pulmonary oedema. We recognise that reconciling the severe hypertension in the face of a poor LV is subject to conjecture but the final spiral of symptoms occurred over a short space of time and may have represented the last reserves of her physiological capabilities (the documented LVEF of 12% was not until after her cardiac arrest and intubation – by which time, there may have been some superadded myocardial stunning).

With advances in fertility treatment and congenital heart disease survival rates, it is likely that an ever more complex maternal population will present with cardiomyopathy or pre-eclampsia, with or without sepsis. This case emphasises the importance of close liaison between obstetrics and ICU services, as advocated by recent Royal College meetings, and hopefully reminds others of the difficulties encountered when managing patients with opposing pathological processes.

Conclusion

In our case, the hypertension of pre-eclampsia reduced during the induction of labour. Whether this was an improvement of her pre-eclampsia, progression of sepsis or myocardial depression secondary to cardiomyopathy is an irreconcilable academic query. That said, its distinction could have impacted management in so far as improving early care.

Consensus amongst experts is that this patient had a peripartum cardiomyopathy whose LV failure will have been exacerbated by the increased afterload seen in pre-eclampsia. The balance struck between these two pathologies, together with her ongoing sepsis meant that the clinical manifestation was an isolated tachycardia and tachypnoea, without gross haemodynamic compromise until her complete cardiovascular collapse.

The physiological combination of pre-eclampsia and sepsis offset each other clinically leading to a delay in diagnosis. The fallout from these both compounded and was compounded by the presence of undiagnosed peripartum cardiomyopathy.

This patient reflects the most severe of interactions between opposing pathologies. Whilst sequential management of organ failure in all conditions remains supportive, the discrimination between deteriorating pre-eclampsia and worsening sepsis, with or without cardiomyopathy, is an important one to make to guide further management.

Summary

The potential for continuing deterioration of cardiomyopathy in the progressing pregnancy requires consideration of delivery of the foetus and early use of mechanical LV support given the high morbidity and mortality found in severe cardiomyopaths. Where possible, this should be done at a specialist centre.

Bromocriptine is an emerging potential treatment for peripartum cardiomyopathy.

Pre-eclampsia exacerbates a failing ventricle but its resolution accelerates symptomatic recovery in peripartum cardiomyopathy.

The hypotension of severe sepsis may be masked by pre-eclampsia, whereas the hypertension of severe pre-eclampsia may be suppressed in septic shock. This may be compounded by cardiomyopathy.

Cardiovascular collapse in a term parturient requires a structured approach to the potential precipitant issues. A large amount of distinction can be achieved through echocardiography looking at:

Collapsed inferior vena cava (IVC), kissing ventricle in preload depletion;

Dilated right ventricle (RV); collapsed left ventricle (LV) in pulmonary artery obstruction (either mechanical or physiological);

Dilated RV and LV with reduced contractility in cardiomyopathy.

Multidisciplinary management by cardiology, obstetricians and intensive care should be assumed mandatory in the care of these patients outside of tertiary cardiac units and early interhospital liaison should be considered.

Consent

Published with generous written consent of the patient.