Recovery from profound acidosis (pH 6.685) in multi-organ dysfunction syndrome

Abstract

Acidosis is a common feature of patients referred to critical care from the emergency department. We present the case of a 49-year-old female with multi-organ dysfunction syndrome (MODS) and an arterial pH of 6.685 on arrival to the emergency department. This case is unique as the patient was in circulatory shock with MODS from rhabdomyolysis on arrival and had not suffered a cardiac arrest. We believe this to be the first reported case of full recovery from such an extreme metabolic disturbance in this context, and discuss the relevance of profound acidosis to early clinical decision-making.

Background

The normal pH of human arterial blood is accepted to be within the range of 7.35 to 7.45, with a tolerable range of 7.00 to 7.80.¹ It is not uncommon for patients to be referred to critical care with arterial pH of less than 7.00. Few reports exist on survival rates from these disturbances. Intense exercise has been shown to lower blood pH to 6.80 in healthy volunteers.² The lowest pathological pH to have been survived in the literature was that of pH 6.33 following cardiac arrest from drowning in Norway.³ Outcomes and patient groups reported have been diverse but there is an overall paucity of literature available surrounding profound acidosis, which may not reflect clinical experience. We present the first reported case of survival in an adult with circulatory shock and established MODS secondary to rhabdomyolysis, and their complete neurological recovery from a pH of 6.685 on arrival to hospital.

Case report

A 49-year-old female presented to our emergency department overnight in December 2017, having being found on the floor in her own home. She had last been seen by relatives two days prior. She had MODS on arrival with respiratory, cardiovascular, neurological and renal compromise. There was widespread mottling of the skin, non-invasive blood pressure was 58/36 (mean 43) mmHg, Glasgow Coma Scale was 9 (E3, V1, M5), and she had 4 mm pupils with sluggish responses to light bilaterally. Her background included Type 2 diabetes mellitus, hypertension, obstructive sleep apnoea, depression with psychosis and previous intentional overdoses of tablets. Regular medications comprised metformin, gliclazide, canagliflozin, quetiapine, venlafaxine, mirtazapine, zopiclone, senna and omeprazole. Her body mass index was 38.1 kg.m⁻² and she had been independent and living alone prior to admission.

Departmental arterial blood gas (ABG) measurement revealed a profound mixed acidosis with pH 6.685 (Figure 1). Laboratory tests revealed: haemoglobin 159 g.l⁻¹, urea 19.5 mmol.l⁻¹, creatinine 278 µmol.l⁻¹, serum phosphate 4.01 mmol.l⁻¹ and initial creatine kinase (CK) 3272 IU.l⁻¹. Urine dip revealed blood ++, protein ++, ketones ++. An electrocardiogram showed sinus rhythm at 76 beats per minute with a corrected QT interval of 434 ms; portable chest radiograph was unremarkable; and computer tomography (CT) imaging of the head revealed no acute pathology.

Figure 1.

Presenting ABG. Temperature 32℃, FiO₂ 0.80. Reference ranges are given in square brackets. c: calculated values; cHCO₃-(Pst)c: standard bicarbonate; cBase(Ecf)c: standard base excess.

Initial management included warmed crystalloid fluid resuscitation, empirical antibiotics, sodium bicarbonate and hydrocortisone. Haemodynamic stability was controlled with a peripheral adrenaline infusion until central venous access could be secured. The patient aspirated blood from spontaneous epistaxis soon after arrival. Following intubation and ventilation, there was an episode of extreme hypotension, responding to 2 min of chest compressions and bolused adrenaline. The patient was transferred to our intensive care unit (ICU) for ongoing resuscitation and commenced on central veno-venous haemofiltration immediately. Her APACHE II score was 25.0 and SOFA score 12.

Physiological stability was achieved with normalisation of blood pH on day 3 (Figure 2). CK peaked at 41747 IU.l⁻¹ on day 4. There was no evidence of compartment syndrome. Causes of high anion gap metabolic acidosis were investigated. Lactate normalised within 6 h. A course of fomepizole was given whilst awaiting toxic alcohol screen results, which were normal. Paracetamol and salicylate levels were within normal limits. The patient had been taking metformin for diabetes mellitus but there was no evidence from the blister pack of medications that an overdose had been ingested. Microbiological samples revealed no positive cultures and no infective or obstructive focus was identified on CT imaging of the head, thorax, abdomen or pelvis. The kidneys and urinary tract were reported to be normal. Bedside FICE on presentation revealed good left ventricular function, no regional wall motion abnormality, and a non-collapsing inferior vena cava. Formal echocardiography on day 3 showed normal left ventricular size and function, with an estimated ejection fraction of 60–65% and functionally normal valves.

Figure 2.

pH(T) resolution over time. pH(T), rather than uncorrected pH, is documented in this ICU.

The patient recovered on respiratory, cardiovascular and renal support over the following days. Renal replacement therapy was provided using heparinised central veno-venous haemofiltration, followed by haemodialysis on day 6 due to recurrent extra-corporeal circuit failure due to thrombosis. The patient failed multiple sedation holds due to agitation and underwent a percutaneous tracheostomy on day 13. By day 30 she had weaned respiratory support and was decannulated. Following step down to the respiratory ward on day 36, the patient had made a complete neurological recovery to baseline, and no longer required renal replacement therapy. She had no recollection of the events leading to her collapse at home. She was discharged home from hospital on day 53.

Blood gas analysis was carried out using the Radiometer ABL800 FLEX blood gas analyser in the Resuscitation area of the emergency department at Manchester Royal Infirmary. Routine calibration, servicing and quality checks are carried out regularly. The pH measurable range for this machine was 6.300–8.000 as specified in the product manual.⁴ Temperature-corrected values were obtained using the device’s built-in algorithm.

Discussion

This patient presented with a mixed respiratory and metabolic acidosis, but the predominant pathology was likely to have been metabolic. The cause of the initial collapse was not established. Possible causes of this patient’s high anion gap metabolic acidosis include acute kidney injury and lactic acidosis from tissue hypoperfusion (Type A) or metformin ingestion (Type B). Cohen and Woods⁵ classified lactic acidosis by aetiology. Type A covers tissue hypoxia and hypoperfusion. Type B lactic acidosis is due to other mechanisms including overproduction and under-utilisation of lactate. These are subclassified: diseases such as diabetes and liver disease; drug-related; and hereditary causes.⁵ Although the lactate was high, above 5 mmol.l⁻¹, it is unlikely to have been the sole cause of this metabolic derangement. There was also no evidence of tablet overdose. Sepsis on presentation was also possible but not confirmed via microbiological cultures. It was thought that rhabdomyolysis secondary to a long lie was a key component to this patient’s organ dysfunction. The ketonuria is suggestive of euglycaemic diabetic ketoacidosis secondary to canagliflozin, a sodium-glucose cotransporter-2 inhibitor.

Possible limitations to this interpretation include errors in ABG sampling and measurement. However, follow-up samples and machine checks support that measurements were accurate. There is debate about the validity of temperature correction for clinical use and whether the ideal pH for cellular function shifts alongside hypothermia.⁶ Our patient was hypothermic, with a temperature of 32℃, which may have limited harm from the degree of acidosis observed.

Case reports exist of patients with good outcomes following pH measurement lower than that presented above. Cases were identified using a literature search for titles containing combinations of the following keywords: profound, extreme, severe, acidosis, acidaemia, pH, recovery. They are discussed below in ascending order of pH. However, we argue that the nature of this patient’s presentation and likely diagnosis differs to these and is therefore notable. Our patient was in established MODS with circulatory shock on arrival to hospital and had not suffered a cardiac arrest prior to blood sampling. Most reports volunteered little detail surrounding blood sampling and whether or not values were temperature-corrected.

A case of prolonged cardiac arrest in Norway, with pH 6.33, received cardiopulmonary resuscitation before transfer to hospital.³ Cases of poisoning with metformin (pH 6.38) and ethylene glycol (pH 6.46) did not involve cardiovascular failure.^7,8 Two cases demonstrated survival in infants after pH 6.48 and 6.49 from haemorrhage and salicylate intoxication respectively but are less applicable to the adult population and, furthermore, neurological outcome in these cases is unknown.^9,10 A patient with lactic acidosis and pH 6.48 from septic shock suffered a cardiac arrest prior to sampling after being relatively stable on presentation.¹¹ An adolescent patient with isoniazid toxicity and seizures was found to have pH 6.49.¹² These derangements are likely to have been contributed to by significant Type A lactic acidosis.

One patient with MODS and pH 6.5 survived their illness but was discharged from hospital with significant morbidity to a nursing facility.¹³ There are cases of survival after low pH (6.527 and 6.57) in severe haemorrhage.^14,15 The early clinical course is unclear in a case with pH 6.54 from metformin on a background of chronic kidney disease, which may confound the acute presentation.¹⁶ There is limited information available surrounding a patient with pH 6.55 from strychnine-induced rhabdomyolysis, a rare aetiology.¹⁷ MODS from exercise- and supplement-induced rhabdomyolysis resulted in pH 6.64 in one case. That patient had relative circulatory stability.¹⁸ Alcohol was implicated in another patient’s acidosis with pH 6.67.¹⁹ The cause, in addition to magnitude, of acidosis is thought to contribute heavily to prognosis.

Although they form an interesting study of physiological extremes, how useful are such measurements in prognostication? There is little evidence in this population. A retrospective study found that patients with extreme acidosis (pH < 7.00) on admission to the ICU had better outcomes than expected.²⁰ One study of out-of-hospital cardiac arrest patients found that pH < 6.85 was not associated with survival to hospital discharge.²¹ In another, patients receiving CPR only made a favourable neurological recovery if presenting pH was ≥ 6.8.²² It is possible that using extreme acidosis as a bedside marker of prognosis may be unreliable.

There is controversy surrounding the relationship between intracellular and extracellular pH and its clinical implications. The use of sodium bicarbonate to reduce acidosis has been examined recently in a multi-centre randomised controlled trial. It had no effect on mortality but there may have been benefits in reducing or delaying the requirement for renal replacement therapy.²³ This patient received sodium bicarbonate but it is unlikely to have affected her outcome. There is suggestion that low bicarbonate levels enhance cells’ ability to resist acidification.²⁴ Therefore the converse may be true and arterial pH may not be an accurate representation of disease severity. However, it is possible that there are some benefits of increasing it to ‘normal’ levels.

Cases of profound acidosis may be more common than the current literature reflects. This could be due to the association with death often being verified, leading to publication bias. False assumption of laboratory or sampling error may also be responsible.³ The initial question we faced in this case was whether ICU admission would be futile. This patient’s unexpected survival and complete neurological recovery illustrate that presenting pH is unreliable in guiding prognostication when faced with a referral to intensive care. Although survival following such extreme physiological derangements may be rare, prompt resuscitation may still be of benefit to such patients and lead to favourable outcomes.

Conclusion

We believe this to be the first published case of an adult making a full neurological recovery from rhabdomyolysis with established MODS including circulatory failure and pH 6.685 on arrival to hospital without cardiac arrest. The differences in presentation and aetiology to existing reports are relevant to clinical practice. This outcome was unexpected and raises the question of whether or not profound acidosis is useful as an indicator of poor prognosis. The authors suggest that, based on current evidence, it should not prejudice our decision whether or not to resuscitate or admit to the ICU. This case supports previous reports that early, effective resuscitation can lead to good outcomes. We welcome further case reports and research into the area of profound acidosis relevant to clinical practice and prognostication.

Consent

Written consent for publication was obtained from the patient.