Methylene blue for refractory shock in children: A systematic review and survey practice analysis

Abstract

Objective:

Shock refractory to fluid and catecholamine therapy has significant morbidity and mortality in children. The use of methylene blue (MB) to treat refractory shock in children is not well described. We aim to collect and summarize the literature, and to describe physicians’ practice patterns, regarding the use of MB to treat shock in children.

Design:

We conducted a systematic search of MEDLINE, Embase, PubMed, Web of Science and Cochrane for studies involving the use of MB for catecholamine refractory shock from database inception to 2019. Collected studies were analyzed qualitatively. To describe practice patterns of MB use, we electronically distributed to US-based pediatric critical care physicians. We assessed physician knowledge and experience with MB. Survey responses were quantitively and qualitatively evaluated.

Setting:

Pediatric critical and cardiac care units.

Patients or Subjects:

Patients less than or equal to 25 years old with refractory shock treated with MB and attending physicians.

Interventions:

None.

Measurements and Main Results:

1293 abstracts met search criteria, 139 articles underwent full text review and 24 studies were included. Studies investigated refractory shock induced by a variety of etiologies and found that MB was generally safe and increased mean arterial blood pressure. There is overall lack of studies, low number of study patients, and low quality of studies identified.

Our survey had a 22.5% response rate, representing 125 institutions. Similar proportions of physicians reported using (40%) or never even considering (43%) MB for shock. The most common reasons for not using MB were unfamiliarity with this drug, its proper dosing, and lack of evidentiary support.

Conclusions:

MB appears safe and may benefit children with refractory shock. There is a stark divide in familiarity and practice patterns regarding its use among physicians. Studies to formally assess safety and efficacy of MB in treating pediatric shock are warranted.

Introduction

Shock is a commonly encountered condition in pediatric emergency departments(1) and critical care units (PICUs)(2). Shock refractory to fluid resuscitation and vasopressor therapy is termed catecholamine-resistant shock(3). Nearly 30% of pediatric patients with shock will progress to catecholamine-resistant shock, which carries significant risk of death(4, 5). Current guidelines for treating catecholamine-resistant septic shock endorse hydrocortisone, vasopressin, or angiotensin and ultimately, extracorporeal membrane oxygenation (ECMO, 6). These therapies have shown mixed results in adults(7-10), and studies in children are even less conclusive(11-13). Given the high morbidity and mortality associated with catecholamine-refractory shock in infants and children, there is a need to identify effective therapies.

Catecholamine-resistant shock with pathologically decreased vascular tone is termed vasoplegic shock(14). Proposed diagnostic criteria for vasoplegic shock in adults include hypotension with preserved cardiac output (CO) and diminished systemic vascular resistance (SVR, 14). There is no consensus definition of this condition in children and its diagnosis may be challenging due to less frequent use of invasive hemodynamic monitoring(15) and poor correlation of clinical parameters with shock severity(16-18). However, recent advances in cardiac point of care ultrasound (POCUS) allow for rapid detection of normal biventricular function in hypotensive patients, mirroring the pathophysiology of vasoplegic shock, supporting the diagnosis of vasoplegic shock(19).

Methylene blue (MB) has shown promise in treating vasoplegic shock in adults(20). Initially synthesized as a blue dye, MB is generally safe for a variety of medical conditions(21). However, MB also inhibits guanylate cyclase (GC) thereby decreasing the production of cyclic guanylate monophosphate(cGMP). In endothelial cells, decreased cGMP leads to less nitric oxide (NO) production, promoting vascular tone(22). MB was first employed to treat shock in the late 1990s(23), and has since been shown to improve mean arterial blood pressure (MAP), increase SVR and lessen vasopressor requirements in adults with vasoplegic shock(20, 24, 25).

The safety and efficacy of MB for the treatment of shock in infants and children are less well described. Moreover, physician’s knowledge of MB as a therapy for shock as well as their patterns of use are not known. We aim to systematically identify and review all studies reporting the use of MB for shock in pediatrics and ascertain the current state of United States-based pediatric critical care physician knowledge and practice patterns.

Materials and Methods

We followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (Supplementary Table 1). Our study protocol is registered in the PROSPERO database of systematic reviews, PROSPERO # 131363 (Supplementary Table 2). We used Yale’s MeSH Analyzer (26), refining the search strategy for the terms: methylene blue, vasoplegic shock (Supplementary Table 3) with scoping queries followed by search strategy refinement. All searches were limited to the English language. Searches in Embase and MEDLINE were limited to human. Additional articles were identified by examining other systematic reviews, bibliographies, discussion with experts, conference abstracts (Web of Science), and publicly available internet searches (Google Scholar).

On April 18, 2019 the authors searched PubMed (1966 to April 2019), Ovid MEDLINE All (1946 to April 17, 2019), Ovid Embase (1974 to 2019 April 17), Web of Science Core Collection (Thompson Reuters, all years), Cochrane Central Register of Controlled Trials (CENTRAL), and Cochrane Database of Systematic Reviews (The Cochrane Library Issue 4 of 12, April 2019). On July 8, 2019, the search was repeated. Literature identified from these searches were de-duplicated in EndNote X8 (Clarivate Analytics) and uploaded to Covidence.

Two independent reviewers completed title, abstract, and full text screening (Figure 1). The senior author mediated consensus meetings to resolve discrepancies. Studies were screened for meeting the inclusion criteria: (1) Observational case report, cohort, case control, or clinical trial studies; (2) studies involving patients less than 25 years old; and (3), studies that use MB specifically for the treatment of shock. Studies that did not specifically utilize MB specifically to treat refractory shock (i.e. shock due to documented methemoglobinemia) were excluded. Due to the limited number of manuscripts identified, peer reviewed conference abstracts were included.

Figure 1:

Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flow chart

The Newcastle Ottawa Scale was used to assess the quality of the large number of non-randomized studies in meta-analyses(27). A quantitative meta-analysis was not performed for several reasons. Relative estimates of effect could not be calculated for studies that lacked control groups. Studies included patients with a variety of etiologies of shock. Heterogeneity in the hemodynamic parameters and timing of their assessment relative to disease onset or MB administration. Therefore, all studies were qualitatively analyzed.

Reported vasoactive-inotropic scores (VIS) extracted from individual studies were assumed to be non-normally distributed and analyzed with a Mann Whitney test (GraphPad Software, 8.1.2, San Diego, California)

We conducted a cross-sectional, self-administered email survey of pediatric critical care physicians in the United States using Qualtrics (Qualtrics Company, Provo, Utah). The survey was conducted from April 23^rd, 2019 to May 24^th, 2019. Three weekly reminders were sent to non-responders. Yale University Human Investigations Committee approved the survey protocol and the implied consent of voluntary participation. Target respondents were board-certified pediatric critical care physicians practicing in North America. To generate the mailing list, hospitals with PICUs were identified through internet and listserv searches curated by the senior author. Electronic mail addresses were obtained from prior surveys and from hospital websites.

The survey (Supplemental Figure 1) was created after a literature review, pilot-testing and revision. Respondent demographic data were gathered. Respondents were requested to answer each survey item based on their personal opinions, experiences, and practice. Responses to clinical queries were gathered through both multiple-choice questions and free text answers. Multiple-choice responses were analyzed with descriptive statistics and reported as percentages of total responses. Qualitative analysis of free text answers elicited themes that are reported.

Results

A total of 2,008 citations were retrieved, pooled and de-duplicated to 1293. Of these, 139 articles met criteria for full text review, and 24 of these studies were included in the final analysis. We identified 17 case studies (Table 1), 4 case series, 2 retrospective cohort trials and a single randomized controlled trial (Table 2). Five identified studies were peer-reviewed conference abstracts, including 3 of the case reports and both of the retrospective cohort studies. The 17 case studies included catecholamine refractory shock induced by various etiologies such as sepsis, heart disease, organ transplantation, drug overdose, trauma and anaphylaxis.

Table 1.

Case reports of the use of methylene blue to treat refractory shock in patients less than 25 years old. MB: methylene blue, B: bolus dose, I: infusion, BP; blood pressure, MB; methylene blue, CHD; congenital heart disease, CO; cardiac output, SVR; systemic vascular resistance, VIS; vasoactive-inotropic score, MAP; mean arterial blood pressure, CI; cardiac index, SvcO2; central venous oxygen saturation, CPB; cardiopulmonary bypass, N/s: not specified. (*) Indicates conference abstracts

Study, year, country	Patient age, sex and diagnosis	MB dose	VIS pre MB	VIS post MB	Major findings
Chan et al, 2018, Australia(28)	15 y/o male with shock due to intentional overdose	B: 1.5 mg/kg I: 1.5 mg/kg/hr for 12 hr, then 1 mg/kg/hr for 12 hr	1200 with metaraminol	402 by 6 hr	BP increased from 65/40 to 120/45 and CO was 8L/min and SVR 300 dyn×sec×cm−5 after MB. MB levels were followed and t1/2 was calculated to be 37.6 hr
Shen, 2018, United States (29)	17 y/o male with shock due to trauma	B:100 mg I: 500mg over 6 hr	n/s	0 by 6 hr	Clinical status improved, and epinephrine and norepinephrine were weaned off 6 hr after MB
Volpon et. al. 2018, Brazil(30)	4 y/o male with shock due to trauma	B: 0.5 mg/kg	80	40 by 2 hr	BP increased from 90/40 to 100s/50s and norepinephrine, epinephrine and dopamine weaned 2 hr after MB
*Cebula and Musfeldt, 2016, United States(31)	24 y/o male with shock due to intentional overdose	B: 1.5 mg/kg I: 1 mg/kg/hr	N/s	0 by 2 hr	MAPs improved from 40s to 60s and vasopressin, epinephrine, and phenylephrine weaned off 2 hr after MB
Lee 2016, United States(32)	7 y/o female with shock due to heart failure	B: 1.5 mg/kg	25	3 by 30 min	BP increased, and vasopressin and epinephrine were weaned 30 min after MB
Hershman et al. 2015, Israel(33)	19 y/o female with shock after kidney transplant	B: 1mg/kg	92 with phenylephrine	0 by 8 hr	BP normalized to 120/60, HR lowered to 100 bpm and norepinephrine, dopamine and phenylephrine weaned within 8 hr after MB
Rutledge et. al. 2015, United States(34)	22 m/o female with shock due to CHD and sepsis	B: 1 mg/kg I: 0.25 mg/kg/hr	520	0 by 14 h	Systolic BP rose 40%, diastolic BP rose 46% and dopamine, norepinephrine and vasopressin were weaned 14 h after MB
*Cheng et al.2012, United States(35)	18 y/o female with shock after liver transplant	B: 2 mg/kg	10	4 by n/s	BP increased from 72/25 to 115/55 and dopamine and epinephrine were weaned immediately after MB
*Ali and Lindley, 2012, United Kingdom(36)	23 y/o female with shock due to sepsis	B: 1 mg/kg I: 0.25 mg/kg/hr	N/s	N/s	HR, BP and CI increased, SVR increased from 200 to 800 dyn×sec×cm−5 and epinephrine and vasopressin weaned after MB.
Banille et al. 2011, Argentina(37)	15 m/o male with shock due to CHD	B: 1 mg/kg/ for 2 doses	24	N/s	HR improved, BP and SvcO2 increased and dopamine and norepinephrine were weaned 11 hr after MB
Lopez, 2011, Spain(38)	Newborn with shock due to sepsis	B: 1 mg/kg	100 with hydrocortisone	50 by 6hr	MAP increased from 28 to 35 mmHg and dopamine, dobutamine and norepinephrine weaned 6 hr after MB
Bhalla et al, 2011, United States(39)	5 y/o female with cardiogenic shock after CPB	B: 1 mg/kg	840	0 by 12 hr	MAP increased from 30 to 60 mmHg and epinephrine, norepinephrine, and vasopressin weaned 12 h after MB
Jang et al. 2011, United States (40)	25 y/o female with shock due to intentional overdose	B: 2 mg/kg I: 1 mg/kg/hr	32 with high dose insulin	n/s	CI was 5.1 L/min/m2 and SVR was 400 dyn×sec×cm−5 before MB. BP increased from 75/40 to 90/75 mmHg and HR decreased from 120 to 90 bpm after MB
Flynn and Sladen, 2009, United States(41)	14 y/o female with shock after lung transplant	B: 1.5 mg/kg	73	3.3 by 10 min	BP improved and norepinephrine, epinephrine, vasopressin and phenylephrine weaned after MB.
Rodrigues et al. 2007, Brazil(42)	23 y/o female with anaphylactic shock	B: 1.5 mg/kg I: 1mg/kg/hr for 1 hour	N/s	N/s	Reversal of shock, angioedema, urticaria and dyspnea with 5 minutes of MB
Taylor and Holtby, 2005, Canada(43)	10 y/o female with shock due to CHD and sepsis	B: 2 mg/kg twice I: 1 mg/kg/hr	60	n/s	MB was administered during CPB and vital signs remained stable
Pagni et al, 2000, United States(44)	20 y/o male with shock due to CHD	B: 2 mg/kg over 30 min	72.5	17.5	BP increased from 80/51 to 118/59 mmHg, SVR increased from 538 to 842 dyn×sec×cm−5, HR decreased from 130 to 110 bpm and CI decreased from 3.0 to 1.6 L/min/m2 1 hr after MB

Table 2:

Non-case reports of the use of methylene blue to treat refractory shock in patients less than 25 years old. MB: methylene blue, B: bolus dose, I: infusion, ECMO; extracorporeal membrane oxygenation, MB; methylene blue, CO; cardiac output, BP; blood pressure, MAP; mean arterial blood pressure, HR; heart rate, CVP; central venous pressure, CPB; cardiopulmonary bypass, CHD; congenital heart disease, SVRI; systemic vascular resistance index, SBP; systolic blood pressure, DBP; diastolic blood pressure (*) Indicates conference abstracts

Study, year, country	Design	Critically ill patients	Control patients	MB Dose	Major findings
*Scheffer et al. 2018, United States(45)	Retrospective cohort	28	0	Not specified	Cohort of patient with vasoplegic shock from cardiac failure (n=7), CPB (n=16), ECMO decannulation (n=2) and sepsis (n=3). MB administration resulted in MAP increase of 0.9 mmHg/hr
Abdelazim et al. 2016, Egypt(46)	Randomized controlled trial	20	20	B: 1.5 mg/kg	Series of infants and children with vasoplegic shock coming off CPB for repair of CHD. Randomized to Methylene Blue or Norepinephrine. BP, SVRI and CVP increased and HR and CO decreased significantly after MB
Hassan et al. 2014, Egypt(47)	Case series	20	0	B: 1.5mg/kg	Series of patients with norepinephrine refractory shock coming off CPB. Administration of MB was associated with decreases in HR (13 bpm), CI (0.8L/min/m2) and norepinephrine infusion (0.57 to 0.11 mcg/Kg/min) and increases in MAP (17 mmHg) and SVRI (791 dyn×sec×cm−5×m−2)
*Banille et al. 2011, Argentina(48)	Retrospective cohort	35 with 5 receiving MB	0	Not specified	Study of hemodynamic profiles of children with septic shock. Improvement in 5 patients is noted after MB administration.
Oberpaur et al. 1997, Chile(49)	Case series	5	0	B: 2 mg/kg	Series of patient with septic shock. After bolus of MB, SBP increased by 61%, DBP increased 94%, and MAP increased 77% after MB. Repeat MB bolus dosing did not improve patient status
Evora 1997, Brazil(50)	Case series	2	0	B: 1.5 mg/kg	Case series of 2 children with anaphylactic shock. Immediate clinical improvement with MB
Driscoll et al. 1996, United States(51)	Case series	5	0	B: 1mg/kg and unspecified	Case series of neonates with septic shock. BP increased by 33% and inotropic and vasopressor agents weaned after MB. Repeat MB bolus dosing did improve patient status.

The quality of evidence for using MB in pediatric patients with catecholamine-refractory shock is poor. Quality assessment scores for the collected literature were low, with a mean of 2.5 out of possible 10 (Supplementary Table 4). Besides the single clinical trial, all other studies were retrospective observational trials or case reports that lacked appropriate control groups. The collected studies had low numbers of subjects, heterogeneous diagnoses and inconsistent reporting of hemodynamic data. Our search revealed, a total of only 102 patients 25 years or younger are reported to date to have had received methylene blue for refractory shock.

All studies reported improvement in patients’ MAP and ability to wean vasoactive and inotropic support. To standardize this observation, VIS were calculated in 13 of the 17 case reports (76%, Table 2). Adjunctive therapies for shock that are not part of the VIS are listed in Table 1. The median VIS before MB in the 13 reported cases was 73 (interquartile range, IQR, 310, 28.5). The median VIS post-MB administration, calculated in 10 patients, was 3.15 (IQR, 34.4, 0), significantly lower than pre-MB (p=0.0007). Timepoints for VIS calculation ranged from 10 minutes to 14 hours after MB administration.

SVR was measured in 6 of the 24 studies. Only 2 studies reported measurements before and after MB, demonstrating an increase in SVR (200 to 800 and 538 to 842 dyn×sec/cm⁵/m² respectively)(36, 44). Children with catecholamine-refractory shock coming off cardiopulmonary bypass reported significant decreases in heart rate (HR, 165 to 152 bpm), and cardiac index (CI, 7.09 to 6.28 L/min/m²) and increases in MAP (40 to 57 mmHg), central venous pressure (CVP, 5 to 7.8 mmHg), mean pulmonary artery pressure (22 to 25 mmHg) and SVR index (SVRI, 1307 to 2098 dyn×sec/cm⁵/m²) after MB treatment(47). Similarly, children coming off cardiopulmonary bypass with vasoplegic shock randomized to MB had significantly lower HR (161 to 151 bpm) and CI (6.2 to 6 L/min/m²), and increased MAP (43 to 56 mmHg), CVP, (5.9 to 7.2 mmHg) and SVRI (45 to 989 45 dyn×sec/cm⁵/m²) than those treated with norepinephrine(46). In this study, vasoplegic shock was defined as MAP <50 mmHg, SVR <800 dyn×s/cm⁵ or SVRI <1500 dyn×s/cm⁵/m² and CI >2.5 L/min/m² and hemodynamic parameters were obtained with echocardiography and pulsed-wave doppler measurements. Overall mortality in this study was too low for statistical comparisons(46).

A single bolus dose, ranging from 0.5 to 2 mg/kg, was used in all studies. Two studies reported repeated boluses of MB, with only one showing benefit. A third of the studies reported a MB infusion, ranging in dosing from 0.25 to 1.5 mg/kg/hr for a total treatment time of 6 to 12 hours. One study reported potential side effects in a patient with polysubstance intoxication with serotonin syndrome 6 hours after infusion of MB and 12.5 hours after ingestion(28).

A total of 403 of 1780 pediatric critical care physicians across the United States participated in the survey (23% response rate) with 384 of them responding to all items (96% completion rate). Respondents represented 125 separate institutions from 46 states. The majority (93%) worked in an academic hospital setting, 52% in a mixed medical surgical setting, with 11% working in an exclusively cardiac unit. Most respondents finished their critical care fellowship between 4-14 years ago (Table 3).

Table 3:

Demographic characteristics of survey respondents

Characteristic	Responses (%)
Year of pediatric critical care fellowship completion	N = 384
2015 or more recently	92 (24)
2005-2015	158 (41)
1995-2005	79 (21)
1985-1995	48 (13)
Before 1985	7 (2)
Description of unit	N = 385
Pediatric medical-surgical ICU	198 (51)
Pediatric cardiac ICU	41 (11)
Mixed pediatric medical-surgical and cardiac ICU	146 (38)
Description of practice environment	N = 384
Academic hospital	365 (93)
Community hospital	22 (6)
Other	6 (2)
How much clinical service time	N = 385
Less than 25%	26 (7)
25-50%	82 (21)
50-75%	117 (30)
More than 75%	160 (42)

Regarding the use of MB for refractory shock, 40% of physicians responded affirmatively, 17% confirmed having considered its use, but the plurality of 43% had never considered using it (Table 4). Of the participants who had used MB, the clinical scenarios most commonly described in free text responses were catecholamine resistant vasodilatory shock (49%) or unspecified catecholamine resistant hypotensive shock (46%). MB use for shock after cardiopulmonary bypass surgery was specifically cited by 9% of responders. Of physicians who have used MB, 37% report difficulty determining appropriate dosing.

Table 4:

Survey responses. See Supplementary Figure 1 for survey flow and nested questions.

Question	Responses (%)
Have you used MB to treat refractory shock?	N = 401
Yes	162 (40.4)
In your experience, does MB administration generally lead to clinical improvement?	N = 157
Yes	54 (34.4)
No	20 (12.7)
Unsure	55 (35.0)
Other	28 (17.8)
No	239 (59.6)
Have you considered using MB to treat refractory shock?	N = 237
Yes	66 (27.9)
No	171 (72.2)
Have you ever had trouble with, or been unsure of, pediatric dosing of MB?	N = 390
Yes	137 (35)
No	253 (65)
Would you consider randomizing this patient for a study in which they receive MB or a placebo?	N = 386
Yes	224 (58.0)
No	26 (6.7)
Need more information	136 (35.2)

A third (34%) of physicians who reported having used MB, noted a clinical improvement, while 13% denied improvement and 35% were unsure. The remaining 18% chose “other” and provided free text descriptions of MB efficacy that were qualitatively analyzed. Nearly half of the free text responses reported sustained or transient improvement after MB administration (Supplemental Table 5). Other free text responses described ambiguous reactions, improvements that did not influence survival, or binary responses in which some patients dramatically improved while others did not.

Among the 17% of survey participants who had considered but not used MB, reasons for its consideration were similar to those that endorsed its use: 51% described a clinical scenario of “refractory shock”, with 36% more specifically referring to vasodilatory shock (Supplemental Table 5). Common reasons for not using MB after consideration were that it was not well studied (56%); patient status improved (42%), and respondents were unsure of pediatric dosing (24%).

Finally, 43% of physicians reported never considered using MB to treat shock. Common reasons cited for not considering MB were that it was not well studied (71%), uncertainty of pediatric dosing (38%), and lack of approval(19%). Free text responses for not considering MB included: not being aware of this drug as a therapeutic option, never previously seen it used, or simply not thinking of it (Supplemental Table 5).

To assess equipoise on the use of MB, a clinical scenario of a 14-year-old female in catecholamine-resistant septic shock was presented. Most respondents (58%) replied that they would be comfortable randomizing the above patient to receive either MB or placebo, 7% declined, and 35% needed more information before deciding, including evidence of or treatment for adrenal insufficiency. Free text responses also indicated a small proportion of physicians (7%) who would not randomize because of their strong belief of efficacy of this treatment.

Discussion

Our dual investigations reveal that MB is a frequently used, yet poorly studied, medication to treat catecholamine-resistant shock in children and infants. Our survey discovered that slightly more than half of pediatric critical care medicine physicians in the United States have used or have considered using MB to treat shock, while the remainder have never considered its use. Our survey reveals that a minimum of 15% of PICU physicians in the United States have used or considered using MB to treat refractory shock. This surprising prevalence is in stark contrast to the paucity of supporting evidence identified in our literature review. Despite the scant evidence collected, several important tentative conclusions may be drawn on the safety, efficacy and utility of MB in children and neonates with catecholamine resistant shock.

MB has been shown to be safe for multiple uses in non-critically ill children, such as the treatment of methemoglobinemia and during surgical procedures(21). Common side effects of therapy are self-limited blue-green cutaneous staining and urine discoloration(22). Serious complications of MB use have included serotonin syndrome(52), hyperbilirubinemia and hemolytic anemia both in neonates (53) and in those with glucose-6-phosphate dehydrogenase deficiency(22). The occurrence of serotonin syndrome after MB administration in a patient with poly-substance abuse is consistent with known interactions between MB and mono-amine oxidase inhibitors(46)Even when given at high doses, MB may be well tolerated without major side effects(54). A practical limitation to its use is the decrease in pulse oximetry measurements (with a stable PaO₂)(55).. In our survey, less than 3% of physicians who had used MB mentioned safety concerns in free text responses. In adults, usual recommended intravenous dosing of MB is 2 mg/kg bolus, sometimes followed by a 12-hour infusion at 0.5 mg/kg/hr(56). Our collected data suggest that these doses of MB are likely to be safe in critically ill children and infants.

The efficacy of MB in the treatment of shock is less clear. In each study, treatment with MB increased MAPs and decreased need for vasoactive and inotropic support. Less commonly, MB increased SVR and CVP concurrently decreasing CI and HR. These pediatric results do not match adult reports showing significant increases in MAP, SVR and left ventricular stroke work after MB, with minimal changes in cardiac output or filling pressures(23, 56-58). Collectively, adult data suggest that MB acts as catecholamine-independent, vaso-constrictive and positive inotropic agent. Our review suggests that in children, MB acts to restore vascular tone with insufficient evidence of inotropic effects.

In our survey, just under half of the 40% of respondents who have used MB observed at least transient clinical benefit. Similarly, free text responses regarding the efficacy of MB reveal a dichotomous clinical effect, in which some patients dramatically improved, and others showed no response. This observation may reflect different stages of the pathophysiologic disruptions associated with the shock state and it relation to the inhibition of NO secretion mediated by MB. The single clinical trial we cited demonstrated increased MAP and SVR in response to MB after cardiopulmonary bypass, an intervention known to alter NO activity(59). Increasing NO activity in infants with congenital heart disease after bypass is known to result in decreased MAP and SVR(60). Characterizing patients with pathologically increased NO activity may identify a subpopulation of critically ill children who might particularly benefit from MB. Since measuring NO levels is challenging, POCUS assessment of normal cardiac function may provide a surrogate measure of the clinical consequences of NO over-activity (i.e. hypotension due pathologic vasodilation).

In our survey, many physicians endorsed using MB in the setting of vasoplegia with normal biventricular function (Supplemental Table 4). This practice pattern is consistent with the findings of our systematic review in which more than a third (9 out of 24) of studies confirmed normal cardiac function prior to administering MB. Reports in the adult literature cite a similar approach, indicating that practioners feel MB is more effective in patients with hypotensive shock with normal CI(61-63). The increasing use of POCUS by pediatric intensivists facilitates the real-time determination of cardiac function that could identify patients who most likely to benefit from MB(64, 65). MB therapy in this setting of catecholamine resistant vasoplegic shock and preserved cardiac function could be considered a “last resource intervention”(49).

There are limitations to this investigation. The lack of relevant, high quality studies precluded a quantitative meta-analysis. Our literature search is subject to publication bias. Although we attempted to identify all pertinent studies, it is possible that qualifying studies were inadvertently omitted from this systematic review. Our practice survey is subject to recall and responder bias, as people who have use MB, particularly with perceived success, might be more likely to respond. This may partly explain the polarity between respondents who have either used MB or not even entertained its use. However, our survey did have a high response and completion rates and included the experience and opinions of physicians from a majority of PICUs across the United States. Although slightly more physicians worked in an academic center and cardiac units, our responses generally agree with previously reported demographics of the pediatric critical care workforce(66).

Conclusions

Use of MB to treat catecholamine-refractory hypotensive shock in children and infants is surprisingly common although evidentiary support is weak. There is a stark divide in practice patterns regarding its use in refractory shock among physicians, with similar proportions of physicians having used MB as having never considered its use. MB appears to be generally safe in children and infants in shock, but its efficacy remains ill-defined. A major effect of MB appears to be increases in vascular tone, perhaps by decreasing NO activity, although pediatric evidence to support this is lacking. Patients with catecholamine-refractory hypotensive shock and normal ventricular function may particularly benefit from MB. Additional studies to formally assess safety and efficacy of MB in refractory pediatric shock are warranted.

Abbreviations:

MB

methylene blue

PICU

pediatric critical care unit

ECMO

extracorporeal membrane oxygenation

CO

cardiac output

POCUS

point of care ultrasound

SVR

systemic vascular resistance

GC

guanylate cyclase

NO

nitric oxide

cGMP

cyclic guanylate monophosphate

MAP

mean arterial blood pressure

VIS

vasoactive-inotropic score

BP

blood pressure

CI

cardiac index

SvcO2

central venous oxygen saturation

HR

heart rate

CVP

central venous pressure

CPB

cardiopulmonary bypass

SVRI

systemic vascular resistance index

SBP

systolic blood pressure

DBP

diastolic blood pressure