John H. Coakley

Persistent neuromuscular and neurophysiologic abnormalities in long-term survivors of prolonged critical illness

ObjectiveTo establish the prevalence, clinical characteristics, and electrophysiologic features of residual neuromuscular dysfunction after prolonged critical illness. DesignProspective follow-up study of survivors of prolonged critical illness. SettingA university hospital and two district general hospitals in the UK. PatientsThe study occurred for a period of 5 yrs. All patients during that time who were in the intensive care unit for >28 days were entered in the study. Measurements and Main ResultsAt follow-up, length of intensive care unit and hospital stay, duration of mechanical ventilation and admission Acute Physiology and Chronic Health Evaluation II score were recorded from the case notes. A clinical history was obtained, a Barthel Index disability score was calculated, and a full neurologic examination was performed. Nerve conduction studies, needle electromyography, single-fiber electromyography and thermal thresholds were performed. A total of 195 patients were identified. There were 86 survivors, of whom 47 could be contacted and 22 consented to be studied. The median time from intensive care unit discharge to follow-up was 43 months (range, 12–57 months). All gave a clear history of severe weakness and functional impairment after hospital discharge and, in all, recovery was prolonged. Motor or sensory deficits were present on clinical examination in 59% of the patients studied. Common peroneal nerve palsy was present in two patients. A total of 21 of 22 (95%) patients had electromyographic evidence of chronic partial denervation at follow-up, findings indicative of a preceding axonal neuropathy. The single-fiber electromyographic studies were also consistent with a preceding motor neuropathy. ConclusionSevere weakness requiring prolonged rehabilitation and abnormal clinical neurologic findings are extremely common in survivors of protracted critical illness. Neurophysiologic evidence of chronic partial denervation of muscle consistent with previous critical illness polyneuropathy is almost invariable and can be found up to 5 yrs after intensive care unit discharge in >90% of these long-stay patients. Evidence of myopathy is unusual. These findings have important implications for the management and rehabilitation of intensive care survivors.

Patterns of neurophysiological abnormality in prolonged critical illness.

ObjectiveTo describe the various patterns of neurophysiological abnormalities which may complicate prolonged critical illness and identify possible aetiological factors.DesignProspective case series of neurophysiological studies, severity of illness scores, organ failures, drug therapy and hospital outcome. Some patients also had muscle biopsies.SettingGeneral intensive care unit (ICU) in a University Hospital. Patients: Forty-four patients requiring intensive care unit stay of more than 7 days. The median age was 60 (range 27–84 years), APACHE II score 19 (range 8–33), organ failures 3 (range 1–6), and mortality was 23%.ResultsSeven patients had normal neurophysiology (group I), 4 had a predominantly sensory axonal neuropathy (group II), 11 had motor syndromes characterised by markedly reduced compound muscle action potentials and sensory action potentials in the normal range (group III) and 19 had combinations of motor and sensory abnormalities (group IV). Three patients had abnormal studies but could not be classified into the above groups (group V). All patients had normal nerve conduction velocities. Electromyography revealed evidence of denervation in five patients in group III and five in group IV. There was no obvious relationship between the pattern of neurophysiological abnormality and the APACHE II score, organ failure score, the presence of sepsis or the administration of muscle relaxants and steroids. A wide range of histological abnormalities was seen in the 24 patients who had a muscle biopsy; there was no clear relationship between these changes and the neurophysiological abnormalities, although histologically normal muscle was only found in patients with normal neurophysiology. Only three of the eight patients from group III in whom muscle biopsy was performed had histological changes compatible with myopathy.ConclusionsNeurophysiological abnormalities complicating critical illness can be broadly divided into three types — sensory abnormalities alone, a pure motor syndrome and a mixed motor and sensory disturbance. The motor syndrome could be explained by an abnormality in the most distal portion of the motor axon, at the neuromuscular junction or the motor end plate and, in some cases, by inexcitable muscle membranes or extreme loss of muscle bulk. The mixed motor and sensory disturbance which is characteristic of ‘critical illness polyneuropathy’ could be explained by a combination of the pure motor syndrome and the mild sensory neuropathy. More precise identification of the various neurophysiological abnormalities and aetiological factors may lead to further insights into the causes of neuromuscular weakness in the critically ill and ultimately to measures for their prevention and treatment.

Preliminary observations on the neuromuscular abnormalities in patients with organ failure and sepsis

ObjectiveTo estimate the incidence and nature of neuromuscular abnormalities in a representative group of ITU patients.DesignProspective sequential study.SettingTeaching hospital ITU.Patients23 patients who eventually stayed >7 days on ITU who had no contraindication to muscle biopsy and whose relatives gave informed consent.Measurements and resultsMuscle histopathology, neurophysiological studies, record of all drugs administered, APACHE II score, organ system failure score, presence or absence of sepsis, clinical evaluation of neuromuscular problems, time to hospital discharge. Heterogeneous neuromuscular abnormalities were present in 22 out of 23 patients studied and included axonal neuropathy, denervation, generalised fibre atrophy, non-specific myopathy and necrotising myopathy.ConclusionNeuromuscular abnormalities are almost invariable in longstay intensive care patients and the resulting weakness may seriously delay hospital discharge. Various abnormalities were seen but no obvious aetiological factors were identified. The origin of the abnormalities is probably multifactorial.

Administration of human recombinant insulin-like growth factor-I in critically ill patients

OBJECTIVES To study the pharmacokinetics of a single subcutaneous dose of recombinant human insulin-like growth factor-I (IGF-I) in patients with systemic inflammatory response syndrome in the intensive care unit (ICU). To evaluate the effects of exogenous recombinant human IGF-I on circulating concentrations of IGF-I binding protein-1 (IGFBP-1), IGF-I binding protein-3 (IGFBP-3), and growth hormone in the critically ill patient; to assess the safety of the subcutaneous administration of 40 microg/kg of recombinant human IGF-I in these patients; and to investigate any effect this dose might have on nitrogen balance, creatinine clearance, and serum electrolyte and lipid concentrations. DESIGN Open-labeled, noncontrolled, prospective, single-dose study of eight fully evaluable ICU patients with systemic inflammatory response syndrome. SETTING ICUs in a teaching hospital and a linked district general hospital in England. PATIENTS Nine patients were examined, eight of whom were fully evaluable. INTERVENTIONS Subcutaneous administration of 40 microg/kg of recombinant human IGF-I. MEASUREMENTS AND MAIN RESULTS Blood samples were taken 24 hrs before the subcutaneous injection of 40 microg/kg of recombinant human IGF-I, and for 48 hrs thereafter. Urine was collected throughout this period. Serum concentrations of IGF-I, IGFBP-1, IGFBP-3, growth hormone, and insulin were measured by radioimmunoassay. IGF-I concentrations (median and range) increased significantly above baseline values (35 ng/mL [20 to 144]) from 15 mins (p < .02) until 10 hrs (p < .02) after injection of recombinant human IGF-I. Peak IGF-I concentrations were sustained from 2 hrs (90.5 ng/mL [23 to 228]) to 5 hrs (88.5 ng/mL [29 to 300]). By 24 hrs, circulating IGF-I concentrations had returned to baseline values. Baseline IGF-I concentrations were extremely low, and although peak values were three times greater, these values only approached the fifth percentile of defined reference ranges for normal values. Compared with values in less seriously ill patients, maximum IGF-I concentrations were reached earlier, the elimination half-life was shorter, clearance was more rapid, and the apparent volume of distribution was similar. IGFBP-3 concentrations also increased after recombinant human IGF-I injection, and at 3 to 4 hrs were significantly elevated, from 30 mins (p = .04) to 8 hrs (p = .04). There was marked between-patient variability in changes in circulating IGF-I, and IGFBP-1, and IGFBP-3 concentrations. More severely ill patients had the lowest circulating IGF-I concentrations and the least response to exogenous recombinant human IGF-I. Elevated baseline circulating growth hormone concentrations (2.3 ng/mL, range 0.8 to 4 [5.1 mU/L, 1.5 to 8]) were significantly depressed from 4 hrs (0.5 ng/mL, 0.5 to 1.5 [1 mU/L, 1 to 3], p = .01) to 6 hrs (0.8 ng/mL, 0.5 to 4 [1.5 mU/L, 1 to 8], p = .02) after recombinant human IGF-I administration. CONCLUSION We observed no adverse effects (e.g., hypoglycemia) that could be attributed to recombinant human IGF-I therapy.

Sequential changes in in vivo muscle and liver protein synthesis and plasma and tissue glutamine levels in sepsis in the rat

We have investigated sequential changes in skeletal muscle and hepatic protein synthesis following sepsis, and their relationship to changes in circulating and tissue glutamine concentrations. Male Wistar rats underwent caecal ligation and puncture (CLP) or sham operation, with starvation, and were killed 24, 72 or 96 h later. A group of non-operated animals were killed at the time of surgery. Protein synthesis was determined using a flooding dose of L-[4-(3)H] phenylalanine, and glutamine concentrations were measured by an enzymic fluorimetric assay. Protein synthesis in gastrocnemius muscle fell in all groups. Gastrocnemius total protein content was reduced after CLP and at 72 and 96 h after sham operation. After CLP, protein synthesis was lower at 24 h, and total protein content was lower at 72 and 96 h, than in sham-operated animals. CLP was associated with increased liver protein synthesis at all time points, whereas there was no change after sham operation. Liver protein content did not change after CLP, but was lower at 72 and 96 h after sham operation than in non-operated animals. Plasma glutamine concentrations were reduced at 24 h after sham operation, and at 72 and 96 h after CLP. Muscle glutamine concentrations were reduced in all groups, with the decrease being greater following CLP than after sham operation. In the liver, glutamine concentrations were unchanged after CLP, but increased after sham operation. In rats with sepsis, decreases in muscle protein synthesis and content are associated with markedly reduced muscle glutamine concentrations. Plasma glutamine concentrations are initially maintained, but fall later. In liver, protein synthesis is increased, while glutamine concentrations are preserved. These results support a peripheral-to-splanchnic glutamine flux in sepsis.

Liver albumin synthesis in sepsis in the rat: influence of parenteral nutrition, glutamine and growth hormone

The effect of sepsis on liver synthesis of albumin remains controversial, with studies in man suggesting that synthesis increases, whereas in animals increased, decreased and unaltered synthesis have been reported. To reconcile these conflicting data, total and relative albumin synthesis was measured in rats 24 h after caecal ligation and puncture (CLP) by immunoprecipitation of albumin following a flooding dose of L-[4-3H]phenylalanine. Following CLP, animals were starved for 18 h and then received intravenous infusions of saline or parenteral nutrition (PN) with or without glutamine for 6 h. In animals receiving PN, parenteral injections of growth hormone (GH) or saline vehicle were also administered. Fractional rate of liver total protein synthesis was elevated and total albumin synthesis rate was reduced in all CLP groups when compared with non-operated animals. Total albumin synthesis was also lower in all animals receiving PN than those receiving saline alone, although these differences did not attain statistical significance, except for the group receiving PN+GH. Relative albumin synthesis was also reduced after CLP, and was significantly lower in animals receiving PN than in those receiving saline alone. These findings suggest that in sepsis hepatic protein synthesis is reprioritized away from the production of albumin towards the production of acute-phase proteins and that this change is not influenced by the provision of nutritional support, glutamine or the administration of GH.

The role of insulin, growth hormone and IGF-I as anabolic agents in the critically ill

ConclusionsThere is a considerable amount of data to suggest that changes in insulin, GH and IGF-I are permissive to protein catabolism in critically ill patients. All three hormones can improve nitrogen economy in the catabolic patient. However, sicker patients are more resistant to the anabolic actions of GH and insulin has only a minor effect. Animal work suggests IGF-I may be a potent anabolic agent and preliminary data in man suggests that it may be a practical therapy. Future work will be required to define the exactrole for each hormone, but it is likely that endocrine therapy of the critically ill will play a major role in nutritional supplementation.

Effect of growth hormone on muscle and liver protein synthesis in septic rats receiving glutamine-enriched parenteral nutrition.

ObjectiveAdministration of recombinant human growth hormone (rhGH) to critically ill adults in an attempt to attenuate catabolism was associated with increased morbidity and mortality. Possible explanations included inhibition of glutamine release from skeletal muscle and consequent restriction of splanchnic glutamine supply. In this study, we examined the effects of rhGH on plasma glutamine levels and on muscle and liver glutamine concentrations and protein synthesis rates in sepsis. We investigated the possibility that administration of supplemental glutamine might ameliorate any adverse effects of rhGH. DesignProspective study in rats rendered septic by cecal ligation and puncture. SettingUniversity hospital laboratory. SubjectsA total of 78 male Wistar rats in six groups. InterventionsAnimals received 6-hr tail vein infusions, commencing 18 hrs after cecal ligation and puncture, of either (a) 0.9% sodium chloride, (b) a standard parenteral nutrition (PN) solution without glutamine, or (c) an isocaloric, isonitrogenous PN solution with glutamine. PN groups received 400 &mgr;g rhGH or equivolume 0.9% sodium chloride vehicle in a divided subcutaneous and intravenous dose at PN commencement. Sacrifice was at the end of the infusion period. A further group was unoperated and uninfused and killed at 24 hrs as baseline controls. Measurements and Main ResultsGlutamine concentrations were measured by fluorometry. Protein synthesis in muscle and liver was measured by a “flooding-dose” technique employing l-[4-H]phenylalanine. Plasma glutamine was increased after cecal ligation and puncture except in the saline and glutamine with rhGH animals. Muscle glutamine was reduced after cecal ligation and puncture and was significantly lower in animals receiving standard PN with rhGH vs. saline alone. Liver glutamine was increased in animals receiving saline and those receiving standard PN with rhGH. PN, with or without glutamine, increased muscle protein synthesis, and the administration of rhGH tended to further increase this effect. Neither PN, glutamine, nor rhGH had an effect on the increased liver protein synthesis characteristic of sepsis. ConclusionsIn sepsis, increased muscle protein synthesis with PN and rhGH administration is not associated with increased muscle glutamine levels. Administration of rhGH does not result in reduced liver glutamine levels or rates of hepatic protein synthesis. PN containing glutamine was no more efficacious than standard PN at increasing muscle protein synthesis.

Influence of starvation, surgery, and sepsis on cardiac protein synthesis in rats: effects of parenteral nutrition, glutamine, and growth hormone.

The effect of sepsis on the rate of protein synthesis in the heart is poorly described. We have investigated changes in protein synthesis in the ventricles of the heart over time after cecal ligation and puncture (CLP) in rats in comparison with sham-operated and unoperated animals (ad libitum). All operated animals were starved from the time of surgery to the time of sacrifice. When operated animals were compared with ad libitum animals, ventricular weight and ventricular protein, and DNA and RNA contents were unchanged at 24 h, but were invariably reduced at 72 and 96 h. Fractional rate of protein synthesis (FSR), RNA activity, and cellular efficiency were reduced at 24 h and further reduced at 72 and 96 h. There were no differences, however, between septic and sham-operated animals. Eighteen hours after CLP, additional groups of rats were infused intravenously with 0.9% sodium chloride, parenteral nutrition (PN), or PN with glutamine, and were given a single dose of 400 &mgr;g recombinant human growth hormone (rhGH) or an equal volume of 0.9% sodium chloride. FSR was higher in animals given PN when compared with those given 0.9% sodium chloride only, and did not differ from FSR measured in unoperated animals. There was no additional benefit from the acute administration of either glutamine-enriched PN or rhGH. These results indicate that ventricular protein synthesis is markedly reduced by surgery and starvation, but that superimposed sepsis does not further influence these changes. PN can prevent the fall in cardiac protein synthesis associated with starvation, surgery, and sepsis, but neither glutamine nor rhGH produced any additional benefit.

Reversible tetraplegia due to polyneuropathy in a diabetic patient with hyperosmolar non-ketotic coma.

Abstract Critical illness polyneuromypathy has not previously been reported as a complication of diabetic coma. We describe a patient with hyperosmolar non-ketotic coma (HONK) complicating gram-negative sepsis in whom persistent coma and profound tetraplegia caused considerable concern. Although, initially, it was feared that the patient had suffered a central neurological complication such as stroke or cerebral oedema, a diagnosis of critical illness motor syndrome (CIMS) was subsequently confirmed neurophysiologically. Profound limb weakness associated with HONK is not necessarily due to a catastrophic cerebral event, rather it may be a result of CIMS, which has an excellent prognosis for full neurological recovery.