K . Komeswaran et al .: J Extra Corpor Technol 2024 , 56 , 114 – 119 117
and ECMO ( from 19 to 13 and then to 12 ). Patients PIP decreased after IVA by only 2 but within the first 24 h of ECMOcannulationby10 ( fromameanof36to26 ).
Discussion
The use of IVA as a “ rescue therapy ” for SA is utilized in our institution . Six out of 7 patients with SA refractory to standard medical therapy , received IVA , however , 5 ( 83 %) of those were subsequently escalated to ECMO . Although our case number is small , the sequence of rescue therapy would indicate that there is not a clear advantage to using IVA prior to ECMO in severe SA .
The use of ECMO overall , including for SA , is increasing [ 13 ]. Reports from ELSO from 1986 to 2007 indicate a significant increase in the use of ECMO for pediatric respiratory failure and 66 % of patients cannulated for SA specifically occurred after 2002 . This trend is coinciding with an increased use of VV ECMO [ 14 ]. In 2015 , 47 % of pediatric ECMO was utilized for respiratory failure with the best survival to hospital discharge in patients with SA ( 92 % survival ) [ 15 , 16 ]. Our institution ’ s experience was in line with this trend as 86 % of patients with severe SA refractory to standard medical therapy required ECMO cannulation and 83 % of those patients survived .
Reported data on the use of IVA for SA is even more limited and has not shown clear benefitoverECMOasarescue therapy for this population . IVAs are known to be potent bronchodilators and are effective in improving oxygenation , lowering CO 2 levels and possibly decreasing ventilator-induced lung injury ( VILI ) in SA [ 17 ]. The CO 2 and ventilatory parameters that would contribute to VILI in our patient population were more markedly improved after ECMO cannulation . This was likely secondary to sweep gas initiation and change in ventilatory settings based on the common method of “ rest settings ” and not necessarily a function of alleviation of bronchospasm . Therefore , true beneficial conclusions cannot be made [ 18 ]. We also do not have data on plateau pressures which would indicate true alveolar harm in this physiology . It is also important to take caution to not decrease the CO 2 in severe hypercarbia rapidly to avoid rapid changes in cerebral blood flow [ 19 ]. The debatable question that arises is not which rescue therapy is most beneficial , but which is less harmful .
In one of the largest single-center studies conducted by Hebbar et al on IVA use in patients with SA , 8 out of 13 patients who received IVA were still escalated to ECMO . The patients who received both IVA and ECMO as rescue therapies had longer hospital LOS , longer ventilator hours after decannulation , and more hospital charges compared to those on IVA alone [ 16 ]. The median time on ECMO was only 95 h . In the larger ELSO SA patient population , the median hours on ECMO were even less ( 91 h ) [ 16 ]. While both ECMO and IVA have significant side effect risks , the most concerning ECMO risks are all decreased with VV cannulation [ 20 ]. Although our median time on ECMO in the patients who received both IVA and ECMO was longer at 168 h , the combined benefit ofVV > VAECMO plus short run times would indicate ECMO is a beneficial rescue therapy for refractory SA . In addition , it is worth noting that in our institution , all patients are cannulated percutaneously , further reducing complications related to cut-down cannulation technique . Comparing these studies directly is difficult but the trends could indicate that patients who are sick enough to receive IVA and then ECMO have increased ventilator days and hours on ECMO due to severity of illness . It is also plausible that delaying ECMO and using IVA first could have contributed to increased patient morbidities .
The risks with IVA , while less studied , include profound hypotension caused by a drop in systemic vascular resistance , nephrotoxicity , carbon monoxide toxicity , and cognitive deficits [ 21 , 22 ]. Perhaps the most important risk from IVA is the risk of long-term neurotoxicity , particularly in the developing brains of children who would be exposed to much longer durations of inhaled anesthetics than typical of an operating room procedure [ 23 , 24 ]. While the neurodevelopmental impact of early exposure to general anesthesia ( GA ) in the pediatric population is still poorly understood , in vitro and in vivo studies have consistently shown that exposure produces dose-dependent and developmental age-dependent effects on various neuronal transmission systems [ 25 ]. The Food and Drug Administration warning for risk of neurodevelopmental effects due to anesthesia is for > 3 h [ 26 ]. There is an increased risk for neurodevelopmental deficits in young children (< 4 years ) especially [ 27 , 28 ]. The median age in our cohort was 5 years with ranges from 2 to 8 years , a stage of developmental vulnerability . The median hours on IVA were 50 h which puts them at significantly increased risk of neurodevelopmental side effects . Neurodevelopmental outcomes after ECMO use in patients are as low as 4 % in VV patients and correlated with time spent on ECMO [ 29 ].
Based on this cohort of patients and known literature on rescue therapy for SA it is not possible to say there is a clear advantage of one over another , however , it is possible that the risk of side effects is lower in patients who are candidates for VV ECMO [ 30 ]. The data on increased neurological side effects and neurodevelopmental outcomes from increased exposure to IVA is concerning and should be considered when making decisions on rescue therapy in SA . This is especially a concern when it seems most patients who are sick enough to receive IVA are escalated to ECMO regardless based on our cohort as well as prior publications . Lastly , it should be mentioned that the definition of refractory SA requiring rescue therapy is also not clearly defined . Reports show that hypercarbia as high as 500 is shown to not cause harm [ 31 , 32 ]. The decision to utilize either rescue therapy discussed here is subjective in nature and often guided by toxicity to the medical interventions or physician anecdotal experience . Despite elevation in CO 2 , often the pH is what guides next steps in rescue therapy and our initial mean pH was 7.08 .
This review is limited in its conclusions due to the retrospective nature of the data collection and small sample size . We also were not able to identify the denominator of patients who met the definition of SA receiving standard medical therapy in the ICU to know the exact prevalence of the need for rescue therapy at our institution . Also due to the retrospective nature , timing of blood gases and ventilator data were not standardized , and long-term follow-up and neurodevelopmental outcomes are unknown . Minimal data were missing in these five patients presented . It was also not possible to know the medical decision tree framework of each intensivist when