Multiple organ dysfunction syndrome (MODS) is a well-described and common condition in critically ill children (1), including patients with infectious and non-infectious systemic inflammatory response syndrome and those with heart disease. It has been reported in as much as 84.6% of patients at the time of their intensive care unit admission (2); hospital mortality remains high and is progressively worse depending on the number of failing organs (3–5). In addition, a distinct subset of very sick patients presents with clinical manifestations of Thrombocytopenia-associated multi-organ failure (TAMOF) (6). In this condition, patients develop a secondary thrombotic microangiopathy associated with a decreased activity of a protease (disintegrin and metalloproteinase with thrombospondin motifs 13, ADAMTS-13), leading to increased circulating ultra-large von Willebrand factor units (vWF), platelet overconsumption and organ failure secondary to vWF-rich microvascular thromboses. These patients are at exceptionally high risk of death and have been previously reported in the pediatric population (7–9).

Given the dire prognosis of TAMOF, additional therapeutic resources aiming to improve survival include extracorporeal membrane oxygenation (ECMO) and therapeutic plasma exchange (TPE). The use of TPE has been described as adjuvant therapy for critically ill patients presenting with TAMOF in small randomized controlled trials and case series (7–9), and it is hypothesized that its benefit may be related to removal of cytotoxins, dysregulated cytokines, and restoration of deficient or depleted humoral products including ADAMTS-13.

Recently, Kawai et al. reported a single-center case series of patients with a clinical diagnosis of TAMOF due to sepsis that underwent combined ECMO and TPE therapy. This strategy portends the theoretical advantage of providing sustained cardiopulmonary support while enhancing organ recovery. Interestingly, these patients had a higher than expected survival (9).

In children with critical medical or perioperative heart disease, low cardiac output syndrome (LCOS) is a common occurrence, presenting in as much as 23.8% of infants postoperative of congenital heart disease (10). It has been described as a condition associated with intrinsic heart disease, myocardial ischemia from aortic cross-clamp and the effects of cardiopulmonary bypass (CPB), resulting in a decreased systemic perfusion state as well as several adverse physiologic and immunologic effects (11, 12). When left untreated, a varying degree of extra-cardiac organ failure may complicate the clinical picture resulting in cardiac arrest, the need for ECMO, a prolonged respiratory failure, and an increased mortality (13).

The use of ECMO to provide mechanical cardiopulmonary support for children with critical cardiopulmonary failure is a well-established rescue therapy for children with critical medical or postoperative heart disease (14). The overall reported survival rate to hospital discharge for pediatric cardiac patients supported by ECMO ranges from 38 to 73% with good functional results (15, 16). However, a lower survival rate has been associated in patients supported with this therapy whenever used in the presence of MODS. The international Extracorporeal Life Support Organization report has published survival rates of 19% for neonatal cardiac patients that underwent dialysis during ECMO, and as low as 17% for those with concomitant DIC (17). Some of these patients may share physiologic, immune response, and hematologic features with non-cardiac patients with TAMOF and it is unknown whether they could benefit with a similar approach. In fact, a decreased ADAMTS-13 activity has been reported in cardiac patients after heart surgery with CPB (18).

We have previously reported our institutional outcomes with the use of ECMO in children with critical heart disease, including an intermediate survival rate of 66% with most survivors having normal or mild neurodevelopmental deficit at follow-up (16). In this study, the use of TPE was identified as an independent variable significantly associated with an increased risk for mortality. However, no further analysis was done for this subgroup of patients. During the last few years, we developed a systematic approach for children with critical heart disease with LCOS and a clinical diagnosis of TAMOF while supported by ECMO, including the use of TPE. To our knowledge, its use in this population has not been previously characterized and we describe here our institutional experience as we wanted to inform ourselves in anticipation of a potential prospective randomized controlled trial.

The study was performed in the cardiac intensive care unit of Children’s Hospital of Pittsburgh of UPMC, a tertiary academic center with an active heart–lung transplantation and mechanical circulatory support program. A detailed description of our program as well as short and intermediate outcomes in our patients supported by cardiac ECMO has been previously published (16).

In children with heart disease and critical cardiopulmonary failure, ECMO support remains the most important rescue therapy. However, LCOS is commonly present as a complication and in a distinct subset of patients their clinical presentation is consistent with TAMOF. They may benefit from additional therapy with TPE and these two therapies are well tolerated together (9).

In our group of patients, we observed a significant improvement in MOFI and platelet count over time after initiation of TPE therapy as revealed by our time-effect analysis, although further study is needed to determine whether this is truly additive over the benefits of ECMO alone. Furthermore, the rapid improvement in MOFI and platelet count in our patients with early TPE therapy resembles previous reports suggesting that early initiation of TPE may be more beneficial. Kawai et al. described a trend toward a decreased ECMO run duration when TPE was initiated early, nonetheless the association was limited due to their small cohort of patients (9). Also, Akca et al. reported an increased mortality in adult patients with thrombocytopenia when low platelet count persisted at day 14 post ICU admission (19). Alternatively, patients with higher MOFI scores (sicker patients) at the time of ECMO cannulation may have warranted an earlier intervention. These findings have potential therapeutic implications, but the descriptive nature of our study does not allow us to establish whether TPE is truly beneficial or a bystander to other interventions like ECMO or ultrafiltration (a less expensive potential alternative for some patients), among others.

A few additional observations are also worth mentioning. Overall survival for this cohort of patients was higher than what would be expected given their serious clinical condition. This could be explained based on a difference in the distribution of several features between our patient group and those in previous reports on overall survival for children with cardiac disease supported by ECMO and specifically related to surgical status, univentricular vs two-ventricle physiology, diagnostic group category, or reason for ECMO cannulation. Indeed, hospital survival for patients with myocarditis (mostly two-ventricle physiology) is the highest among cardiac patients receiving ECMO (20). On the opposite side, patients with congenital heart disease have exhibited a comparatively low survival (15, 17). Whether these patients would further benefit with TPE, and which group would benefit the most, are important questions that need an answer. Particularly, our patient cohort included an important proportion of them with two-ventricle physiology (72.3%) as well as those categorized in the heart failure group (36.6%), and both subgroups presented a significantly higher hospital survival rate when compared with the ones with univentricular physiology or congenital heart disease subgroups, respectively. These findings are in alignment with a previous report from our team, reporting a higher survival for patients with two-ventricle physiology compared with those with a single ventricle (16). Finally, the significantly lower hospital survival in our patients cannulated in the OR compared with previous reports from our center and others (16, 21) provides support to the concept of a higher risk for patients undergoing a surgical procedure with unmitigated multi-organ failure.

Therapeutic plasma exchange was successfully used in children with cardiac disease and a clinical diagnosis of TAMOF while supported by ECMO. The overall survival to hospital discharge was 53.7%. This therapy when used with ECMO appeared to be associated with significant organ failure recovery, and earlier initiation after ECMO cannulation was associated with a greater initial improvement in organ dysfunction. These data support the need for a larger, multicenter prospective RCT to identify overall benefit and indications, best timing of therapy start and which population subgroups are most likely to exhibit a positive response.

Presented at the 8th World Congress of the World Federation of Pediatric Intensive & Critical Care Societies (PICC 2016), Toronto, ON, Canada, on June 4–8, 2016.

Conception and design: MC, AL-M, and RM. Analysis and interpretation: MC, AL-M, LS, MS, AA, VM, and RM. Manuscript writing: MC, AL-M, LS, MS, and AA. Critical revision of manuscript: MC, AL-M, LS, MS, AA, VM, RM. Data collection: MC. Provision of materials, patients, or resources: MS, VM, and RM. Statistical expertise: AA. Literature search: MC, AL-M, LS, and RM.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.