Determination of brain death in potential donors on ECMO support Authors: Jan Sönke Englbrecht, Jan Sackarnd, Felix Dinkel, Markus Holling Published: 2026 ABSTRACT BackgroundA recent update of the German directive on determining brain death introduced changes regarding the diagnostics in potential donors on extracorporeal membrane oxygenation (ECMO). Physicians familiar with ECMO must accompany apnea testing and cerebral blood flow-based methods are no longer permitted as ancillary tests in potential donors on venoarterial (VA) ECMO because they are considered insufficiently validated.ObjectiveThe aim was to identify challenges in determining brain death under VA-ECMO in the context of the updated directive.MethodsWe analyzed and compared the complete process of brain death (BD) determination before and after the directive update in all potential donors on VA-ECMO between 2021 and 2024 at the University Hospital Münster.ResultsA total of 8 potential organ donors on VA-ECMO were identified, BD was determined in seven donors and four donations were utilized (15% of all donors during the observation period). Apnea testing was performed safely in all cases, although PaCO2 varied depending on the sampling site. In three cases before the update, cerebral circulatory arrest was demonstrated. After the update, a flat-line EEG was obtained in four cases. In one case, EEG was inconclusive due to artefacts.ConclusionBoth apnea tests and cerebral blood flow-based methods were safely performed in donors on VA-ECMO. Electroencephalography (EEG) was only used after the directive update. Future revisions of the directive should provide precise protocols for apnea testing. Further evidence is desirable to reauthorize cerebral blood flow-based methods as ancillary tests. This could support higher organ donation rates in potential donors on VA-ECMO.Graphic abstract The new directive on the determination of brain death in Germany has introduced revised diagnostic requirements for potential donors on extracorporeal membrane oxygenation (ECMO). Physicians experienced in ECMO must be involved during apnea testing. Ancillary testing via methods assessing cerebral circulatory arrest is no longer permitted during venoarterial (VA)-ECMO support, as these procedures are deemed insufficiently validated. These limitations may lead to the exclusion of potential donors on ECMO, which is particularly concerning considering the ongoing organ shortage. Background The determination of brain death (BD) is a prerequisite for deceased organ donation in Germany [4]. The German Medical Association (BÄK) issues a legally binding directive on the procedure for determining BD (“Richtlinie irreversibler Hirnfunktionsausfall”, RL-IHA). This directive is regularly updated and outlines, among other things, the mandatory prerequisites on the part of the patient, the clinical examination, including apnea testing (with a target PaCO2 of ≥ 60 mm Hg) and proof of irreversibility. This can be done by repeating the clinical examination after a defined observation period, depending on the etiology and localization of the brain injury or ancillary testing. Ancillary testing can consist of using additional diagnostic modalities to assess the absence of cerebral blood flow (CBF) or electrocerebral activity [7]. The newest version of the RL-IHA came into force on 5 September 2022 and introduced some important changes regarding the determination of BD in patients requiring extracorporeal membrane oxygenation, either venovenous (VV) or venoarterial (VA) extracorporeal membrane oxygenation (ECMO). Performing an apnea test requires adherence to the same principles as in those patients not on ECMO but it must now be performed with the support of physicians who are familiar with ECMO treatment. Methods based on CBF are no longer approved as ancillary tests in patients on VA-ECMO support. According to the BÄK, these procedures are not sufficiently validated or not suitable for detecting cerebral circulatory arrest in patients on VA-ECMO [7]. Consequently, the determination of BD in potential donation after brain death (DBD) donors on ECMO is likely to have become more demanding, or in the worst case, these requirements may contribute to a donation not utilized [3]. Around 25% of all utilized DBD donors in Germany are recruited from patients with hypoxic brain injury [8]. It can be assumed that the use of VA-ECMO will continue to increase, especially as a treatment option for patients with prehospital cardiovascular arrest (extracorporeal cardiopulmonary resuscitation, ECPR) who have a high risk for hypoxic brain injury [1, 2, 13]. This suggests that a standardized protocol for apnea testing is desirable and validation of CBF-based ancillary tests in potential donors on VA-ECMO is important to enable these tests in future revisions of the RL-IHA. Otherwise, there is a risk that BD will not be determined in a potentially relevant number of otherwise eligible donors. Notably, Germany ranked 22nd for utilized donors compared to other European nations in 2023 and held the lowest position among Eurotransplant members [8, 12]. The aim of the present study was to describe the proportion of potential DBD donors on VA-ECMO at our institution and the procedure and challenges of determining BD in this cohort before and after the update of the RL-IHA to add to the limited evidence. Methods This retrospective study was performed in accordance with the Declaration of Helsinki. The need for informed consent was waived by the local Ethics Committee of the University of Münster due to the retrospective analysis of routinely collected patient data. The study protocol was approved on 28 May 2025 (file number 2021-801-f-S). Based on the records of the in-house organ donation coordinator, all potential DBD donors for the period from January 2021 to December 2024 with the following characteristics were retrospectively included:the attending physicians considered BD to be highly probable,age ≥ 18 years,no refusal of donation known at the time of BD evaluation,VA-ECMO previously set up for cardiopulmonary support. the attending physicians considered BD to be highly probable, the attending physicians considered BD to be highly probable, age ≥ 18 years, age ≥ 18 years, no refusal of donation known at the time of BD evaluation, no refusal of donation known at the time of BD evaluation, VA-ECMO previously set up for cardiopulmonary support. VA-ECMO previously set up for cardiopulmonary support. For each identified case, the following parameters were assessed through review of medical files: patient demographics, clinical course, VA-ECMO characteristics, process of BD determination, ventilator and ECMO settings during apnea testing including results of arterial blood gas (ABG) sampling, methods of ancillary testing and outcome. The total number of potential and utilized DBD donors at our institution were identified using a computer program (“Transplant Check”), provided by the German organ procurement organization (OPO). This program identifies all in-hospital deaths of patients with brain injury from the patient data according to § 21 Hospital Remuneration Act [5]. The total number of ECMO-supported patients at our institution were extracted from internal quality reports. Statistics Statistical analysis was performed using SPSS (IBM Corporation, Armonk, NY, USA, version 28). Data were expressed as absolute (percentage of total) for categorical variables. Continuous variables were assessed for normality using the Shapiro-Wilk test. Due to the small sample size, all continuous variables were expressed as median with interquartile range (25; 75 percentiles). Results Of the potential DBD donors eight matched the inclusion criteria during the observation period and were all included in the analysis (Table 1). Of these, BD was successfully determined in seven cases and organs were procured in four. The first three cases occurred at a time when the current directive on the determination of BD had not yet come into force.Table 1Potential donation after brain death donors on venoarterial extracorporeal membrane oxygenationCaseAge (years)Reason for ICU admissionECMO after admission (days)BD after ECMO (days)Ancillary testConsent to donationDonation utilized/reason if not utilizedOrgans procured144OHCA01SPECTYesYesLungs, liver, kidneys273Hemorrhagic shock after cardiac valve surgery131SPECTYesNo (rejection by OPO)–361OHCA01SPECTYesNo (suspected tumor)–Directive update462OHCA03CT-A/EEGNoNo (inconclusive BD diagnosis and rejection from next kin)–542OHCA02EEGYesYesLiver, kidneys622IHCA after coronary artery dissection04EEGYesNo (rejection by public prosecutor)–732OHCA01EEGYesYesLiver, kidneys870IHCA after CABG23CT-A*/EEGYesYesLiver kidneysICU intensive care unit, ECMO extracorporeal membrane oxygenation, BD brain death, OHCA out-of-hospital cardiac arrest, IHCA in-hospital cardiac arrest, BD brain death, OPO organ procurement organization, CABG coronary artery bypass grafting, SPECT single-photon emission computed tomography, CT-A cerebral computed tomography angiography, EEG electroencephalography* Originally not intended as ancillary test Table 1Potential donation after brain death donors on venoarterial extracorporeal membrane oxygenationCaseAge (years)Reason for ICU admissionECMO after admission (days)BD after ECMO (days)Ancillary testConsent to donationDonation utilized/reason if not utilizedOrgans procured144OHCA01SPECTYesYesLungs, liver, kidneys273Hemorrhagic shock after cardiac valve surgery131SPECTYesNo (rejection by OPO)–361OHCA01SPECTYesNo (suspected tumor)–Directive update462OHCA03CT-A/EEGNoNo (inconclusive BD diagnosis and rejection from next kin)–542OHCA02EEGYesYesLiver, kidneys622IHCA after coronary artery dissection04EEGYesNo (rejection by public prosecutor)–732OHCA01EEGYesYesLiver, kidneys870IHCA after CABG23CT-A*/EEGYesYesLiver kidneysICU intensive care unit, ECMO extracorporeal membrane oxygenation, BD brain death, OHCA out-of-hospital cardiac arrest, IHCA in-hospital cardiac arrest, BD brain death, OPO organ procurement organization, CABG coronary artery bypass grafting, SPECT single-photon emission computed tomography, CT-A cerebral computed tomography angiography, EEG electroencephalography* Originally not intended as ancillary test Potential donation after brain death donors on venoarterial extracorporeal membrane oxygenation ICU intensive care unit, ECMO extracorporeal membrane oxygenation, BD brain death, OHCA out-of-hospital cardiac arrest, IHCA in-hospital cardiac arrest, BD brain death, OPO organ procurement organization, CABG coronary artery bypass grafting, SPECT single-photon emission computed tomography, CT-A cerebral computed tomography angiography, EEG electroencephalography * Originally not intended as ancillary test Table 2 shows the characteristics of VA-ECMO support and hemodynamics during apnea testing. In two cases (cases 2 and 8), the respirator was disconnected during apnea testing. In cases 5 and 8 no ABG sample was taken during apnea testing, as the first ABG sample after start of apnea test already verified a PaCO2 ≥ 60 mm Hg.Table 2Extracorporeal membrane oxygenation parameters, apnea testing procedure, and haemodynamic statusCaseVA-ECMO cannulationECMO settings during apnea testingABG sampling sideMAD (mm Hg)Apnea test duration (min)OxygenationPreDuringEndECMO blood flow (l/min)1Right CFAFiO2 Resp. (%)1.01.01.02.5Right brachial6714Right CFVFiO2 ECMO (%)0.60.60.62Right CFAFiO2 Resp. (%)1.0–*–*2.96Right radial6926Right CFVFiO2 ECMO (%)0.40.40.43Right CFAFiO2 Resp. (%)0.51.01.03.52Right radial6812Left CFVFiO2 ECMO (%)0.70.70.74Right CFAFiO2 Resp. (%)0.70.70.73.5Right radial7117Right CFVFiO2 ECMO (%)0.60.70.75Right CFAFiO2 Resp. (%)0.5**1.03.55Left radial982Right CFVFiO2 ECMO (%)0.70.76Right CFAFiO2 Resp. (%)0.51.01.03.77Right brachial6211Right CFVFiO2 ECMO (%)0.550.70.77Left CFAFiO2 Resp. (%)0.51.01.04.34Left brachial7816Right CFVFiO2 ECMO (%)0.30.40.48Left CFAFiO2 Resp. (%)0.5**–*2.55Right radial7112Left CFVFiO2 ECMO (%)0.71.0Left radialVA-ECMO venoarterial extracorporeal membrane oxygenation, FiO2 resp (%) fraction of inspired oxygen, ABG arterial blood gas, MAD mean arterial pressure, CFA common femoral artery, CFV common femoral vein* Respirator disconnected during apnea testing; ** no arterial blood gas sample during apnea testing Table 2Extracorporeal membrane oxygenation parameters, apnea testing procedure, and haemodynamic statusCaseVA-ECMO cannulationECMO settings during apnea testingABG sampling sideMAD (mm Hg)Apnea test duration (min)OxygenationPreDuringEndECMO blood flow (l/min)1Right CFAFiO2 Resp. (%)1.01.01.02.5Right brachial6714Right CFVFiO2 ECMO (%)0.60.60.62Right CFAFiO2 Resp. (%)1.0–*–*2.96Right radial6926Right CFVFiO2 ECMO (%)0.40.40.43Right CFAFiO2 Resp. (%)0.51.01.03.52Right radial6812Left CFVFiO2 ECMO (%)0.70.70.74Right CFAFiO2 Resp. (%)0.70.70.73.5Right radial7117Right CFVFiO2 ECMO (%)0.60.70.75Right CFAFiO2 Resp. (%)0.5**1.03.55Left radial982Right CFVFiO2 ECMO (%)0.70.76Right CFAFiO2 Resp. (%)0.51.01.03.77Right brachial6211Right CFVFiO2 ECMO (%)0.550.70.77Left CFAFiO2 Resp. (%)0.51.01.04.34Left brachial7816Right CFVFiO2 ECMO (%)0.30.40.48Left CFAFiO2 Resp. (%)0.5**–*2.55Right radial7112Left CFVFiO2 ECMO (%)0.71.0Left radialVA-ECMO venoarterial extracorporeal membrane oxygenation, FiO2 resp (%) fraction of inspired oxygen, ABG arterial blood gas, MAD mean arterial pressure, CFA common femoral artery, CFV common femoral vein* Respirator disconnected during apnea testing; ** no arterial blood gas sample during apnea testing Extracorporeal membrane oxygenation parameters, apnea testing procedure, and haemodynamic status VA-ECMO venoarterial extracorporeal membrane oxygenation, FiO2 resp (%) fraction of inspired oxygen, ABG arterial blood gas, MAD mean arterial pressure, CFA common femoral artery, CFV common femoral vein * Respirator disconnected during apnea testing; ** no arterial blood gas sample during apnea testing The median ECMO sweep gas flow was reduced from 3.75 l/min (IQR 2.38: 5.50) before to 1.00 l/min (IQR 0.86–1.50) during and 1.00 l/min (IQR 1.00–1.00) at the end of apnea testing. The median PaCO2 increased accordingly from 41.0 mmHg (IQR 39.22–44.35 to 52.25 mmHg (IQR 47.38–57.00 during and 64.60 mmHg (IQR 61.53–73.50) at the end (with all cases reaching a PaCO2 of ≥ 60 mmHg). The median PaO2 changed from 262 mmHg (IQR 187–341) before to 341 mmHg (IQR 182–411) during and 162 mmHg (IQR 75–307) at the end of apnea testing and median pH from 7.36 (IQR 7.29–7.39) to 7.29 (IQR 7.19–7.33) and 7.21 (IQR 7.11–7.23) (Fig. 1). The median duration of apnea testing was 13 minutes (IQR 11.5–16.5). No apnea test was discontinued due to cardiopulmonary instability.Fig. 1a Extracorporeal membrane oxygenation (ECMO) sweep gas flow, b pH, c PaO2 and d PaCO2 measured at three time points before (Pre), during (During), and at the end (End) of apnea testing. Boxplots show the median and interquartile range with whiskers representing the full data range excluding outliers. Individual values are shown as black dots. Superimposed dot and line plots indicate group means with 95% confidence intervals a Extracorporeal membrane oxygenation (ECMO) sweep gas flow, b pH, c PaO2 and d PaCO2 measured at three time points before (Pre), during (During), and at the end (End) of apnea testing. Boxplots show the median and interquartile range with whiskers representing the full data range excluding outliers. Individual values are shown as black dots. Superimposed dot and line plots indicate group means with 95% confidence intervals Description of cases Case 1. A 44-year-old male with out of hospital cardiac arrest (OHCA) due to ST-elevation myocardial infarction (STEMI). The patient was admitted to our institution under ongoing mechanical cardiopulmonary resuscitation (CPR), and VA-ECMO was initiated immediately upon arrival. On the next day, severe hypoxic brain damage was suspected. BD was determined after an ancillary test using single-photon emission computed tomography (SPECT), and organ procurement was subsequently performed. Case 2. A 73-year-old female admitted to our institution for elective cardiac valve surgery. The patient developed respiratory insufficiency due to a newly diagnosed coronavirus disease 2019 (COVID-19) 4 days after surgery, 5 days later in-hospital cardiovascular arrest (IHCA) occurred and VA-ECMO support was established during manual CPR. After initial stabilization, the clinical examination revealed suspected severe hypoxic brain injury and BD was determined after an ancillary test using single photon emission computed tomography (SPECT). The German OPO rejected donation due to COVID-19 and concomitant multiorgan failure. Case 3. A 61-year-old male with OHCA due to STEMI. Admission to our institution under mechanical CPR. ECPR via VA-ECMO was initiated immediately upon arrival. The next day severe hypoxic brain damage was suspected based on the clinical examination and BD was determined after an ancillary test using SPECT. During the following evaluation, a newly diagnosed tumor prevented organ donation. Case 4. A 62-year-old male with OHCA due to STEMI. Admission to our institution under mechanical CPR with immediate establishment of VA-ECMO (ECPR) on arrival. On day 3 after admission, clinical examination revealed suspected BD. Ancillary test using cerebral computed tomography angiography (CT-A) showed complete cessation of intracerebral blood flow. The results of CT‑A did not provide sufficient evidence of irreversibility, in accordance with the then newly introduced RL-IHA and two subsequent electroencephalography (EEG) examinations could not be used as evidence of BD due to pronounced artefacts. The next of kin subsequently declined to consent to organ donation. Case 5. A 42-year-old female with OHCA of unknown origin. Admission to our institution under mechanical CPR with immediate establishment of VA-ECMO (ECPR) on arrival. Due to hypoxic-ischemic cerebral edema, a decompressive craniectomy was performed on the same day. Nevertheless, intracranial pressure increased and 2 days after admission BD was determined using EEG as an ancillary test and organs were procured. Case 6. A 22-year-old female with dissection of coronary arteries of unknown origin and cardiovascular arrest during coronary angiography in a remote hospital. On-site implantation of VA-ECMO (ECPR) and transfer to our institution. In the following days, progressive hypoxic brain injury was suspected and BD was determined 4 days after admission using EEG as an ancillary test. The public prosecutor did not authorize the body for organ donation due to potentially non-natural cause of death. Case 7. A 32-year-old female presented with OHCA of unknown origin. She was admitted to our institution under ongoing mechanical CPR and VA-ECMO (ECPR) was initiated immediately upon arrival. A CT scan performed to determine the underlying etiology revealed supratentorial perfusion arrest, BD was confirmed the following day using EEG as an ancillary test and organ procurement was subsequently performed. Case 8. A 70-year-old male with acute coronary syndrome and immediate coronary artery bypass grafting surgery after admission. The patient developed ventricular fibrillation 2 days later and VA-ECMO was established under ongoing manual CPR. On the same day a CT scan (originally not intended as ancillary test) showed a cerebral edema with complete cessation of intracranial perfusion. The next day, BD was determined using EEG as an ancillary test and organs were procured. Several cases provide valuable insights into the diagnosis of BD on VA-ECMO. Two sampling sides (right and left radial artery) were used for simultaneously obtained ABG analysis during apnea test in case 8. Prior to the apnea test, PaO2 (right/left) was 181/186 mm Hg and the corresponding PaCO2 (right/left) was 44.2/40.6 mm Hg. At the end of the apnea test, PaO2 (right/left) had decreased to 65/69 mm Hg, while PaCO2 (right/left) had increased to 66.2/67.2 mm Hg. In cases 1–4 and 8, CBF-based methods were used, and all showed cessation of intracranial perfusion (exemplary SPECT in case 1 and CT‑A in case 4 (Fig. 2)). In case 4, artefacts interfered with the diagnosis of an isoelectric EEG (Fig. 3). In case 8, two ancillary tests showed complete loss of brain function (CT‑A and isoelectric EEG).Fig. 2a Single-photon emission computed tomography showing absence of radiotracer uptake in both supratentorial and infratentorial brain regions, consistent with cerebral circulatory arrest. Additional thoracoabdominal imaging to confirm physiological radiotracer distribution as in vivo quality control. b CT angiography of the head. The image shows global cerebral edema with complete loss of intracranial pressure compensation and absence of contrast enhancement in the intracranial segments of the supra-aortic arteries (left). Contrast enhancement of the extracranial segments of the supra-aortic arteries is preserved (arrow, right)Fig. 3EEG for confirmation of brain death. A representative EEG segment is shown in bipolar montage according to the German directive for the determination of brain death, without (a) and with (b) filtering of excessive 50 Hz artefact. After applying a 50 Hz filter, residual activity exceeding 2 μV is detectable in leads C4–O2 and T4–O2, most likely representing remnants of 50 Hz activity due to incomplete filtering (differential diagnosis: true cerebral activity). Electrode placement according to the international 10–20 system. Fp = frontopolar, C = central, O = occipital, T = temporal a Single-photon emission computed tomography showing absence of radiotracer uptake in both supratentorial and infratentorial brain regions, consistent with cerebral circulatory arrest. Additional thoracoabdominal imaging to confirm physiological radiotracer distribution as in vivo quality control. b CT angiography of the head. The image shows global cerebral edema with complete loss of intracranial pressure compensation and absence of contrast enhancement in the intracranial segments of the supra-aortic arteries (left). Contrast enhancement of the extracranial segments of the supra-aortic arteries is preserved (arrow, right) EEG for confirmation of brain death. A representative EEG segment is shown in bipolar montage according to the German directive for the determination of brain death, without (a) and with (b) filtering of excessive 50 Hz artefact. After applying a 50 Hz filter, residual activity exceeding 2 μV is detectable in leads C4–O2 and T4–O2, most likely representing remnants of 50 Hz activity due to incomplete filtering (differential diagnosis: true cerebral activity). Electrode placement according to the international 10–20 system. Fp = frontopolar, C = central, O = occipital, T = temporal Proportion of DBD donors on VA-ECMO among all donors A total of 1163 patients with brain injury who died during the hospital stay were identified by Transplant Check during the observational period at our institution. After excluding cases with a documented reason against donation, 479 potential DBD donors remained where BD was determined in 35 cases and 26 DBD donors were utilized, including the 4 donors (15%) on VA-ECMO (Table 3).Table 3Proportion of potential and utilized donors among all deceased with a diagnosed brain injuryCohortn (%)Deceased with brain damage1163 (100)No mechanical ventilation during hospital stay262 (22.5)Absolute contraindication for donation156 (13.4)A medical condition that prevented donation126 (10.8)Refused consent116 (10.0)No relevant brain damage24 (2.1)Potential DBD donors479 (41.2)BD confirmed35 (7.3)Utilized DBD donors26 (5.4)BD brain death, DBD donation after brain death Table 3Proportion of potential and utilized donors among all deceased with a diagnosed brain injuryCohortn (%)Deceased with brain damage1163 (100)No mechanical ventilation during hospital stay262 (22.5)Absolute contraindication for donation156 (13.4)A medical condition that prevented donation126 (10.8)Refused consent116 (10.0)No relevant brain damage24 (2.1)Potential DBD donors479 (41.2)BD confirmed35 (7.3)Utilized DBD donors26 (5.4)BD brain death, DBD donation after brain death Proportion of potential and utilized donors among all deceased with a diagnosed brain injury BD brain death, DBD donation after brain death Between 2021 and 2024, a total of 623 patients received ECMO support at our institution. Of these, 256 patients were treated with VV-ECMO, while 367 received VA-ECMO support (including venoarteriovenous configurations). Among the 367 patients with VA-ECMO, 117 were managed in a medical intensive care unit (i.e., nonsurgical patients). Within this subgroup, ECPR was the primary indication in 65 cases (56%). This cohort included the four patients (6%) from our analysis who subsequently became utilized DBD donors. Discussion In this analysis of BD determination in potential DBD donors on VA-ECMO, eight cases were identified where BD was determined in seven and organs were procured in four cases. Apnea testing was safely performed in all cases but ABG analysis revealed considerable differences depending on the sampling site in one case. The CBF-based methods for ancillary testing were used before the RL-IHA update and demonstrated complete cessation of intracerebral perfusion. An EEG was used after the update and was inconclusive in one case. The potential for organ donation in DBD donors on ECMO Although organ donation following ECMO treatment has been increasingly reported in international studies, reliable national data from Germany are currently lacking. The annual reports published by the German OPO do not provide specific figures on utilized organ donations from ECMO-supported patients. As hypoxic brain injury is the cause of BD in roughly 25% of DBD donors in Germany, a significant share of these may have been on ECMO support following cardiac arrest (ECPR) but no reliable data are available [8]. A study from the USA reported 419 VA-ECMO patients presenting with refractory OHCA over 7 consecutive years and organs from 38 patients were procured. The number of organ donors and successful transplants correlated positively with the years, indicating an increase of donations from patients on ECMO-support [13]. Another retrospective analysis showed that among 141 potential organ donors on ECMO, 33% proceeded to donation [6]. A systematic review indicated that organ transplantation from donors on ECMO support at the time of death showed high graft and recipient survival [22]. These studies suggest that a significant share of patients on ECMO support, especially after cardiac arrest, can become eligible DBD donors and the number is likely increasing. Apnea test in donors on VA-ECMO The apnea test represents the final component of the clinical assessment for the diagnosis of BD and must also be performed in patients receiving ECMO support [7]. Our results demonstrate that with careful adjustment of ECMO sweep gas flow, a PaCO2 ≥ 60 mm Hg can be safely achieved without risk of desaturation; however, several specific considerations apply in the context of VA-ECMO as it involves two distinct circulatory systems where oxygenation and decarboxylation occur: one via the patient’s lungs and the other via the extracorporeal oxygenator of the VA-ECMO. Oxygenated blood from the ECMO circuit enters the aorta in a retrograde fashion. In contrast, blood that is oxygenated through the pulmonary circulation is ejected anterogradely into the ascending aorta by the left ventricle. The location of this mixing point is influenced by native cardiac output and ECMO blood flow. As a result, different values of PaO2 and PaCO2 can occur at different sites within the patient’s circulation, a phenomenon known as Harlequin syndrome. Consequently, different gas tensions can be measured depending on where blood is sampled [11]. The goal of the apnea test is to enable the PaCO2 to increase to levels that would normally maximally stimulate the respiratory centers in a functioning medulla. Because gas tensions cannot be measured directly within the medulla, ABG sampling from a site that reliably reflects central arterial gas tensions is required. In the presence of the Harlequin syndrome, a single sampling site may not accurately represent the gas tensions perfusing the brainstem and therefore may not reliably reflect the PaCO2 at the level of the medulla. Therefore, physicians performing the apnea test must be familiar with the effects of VA-ECMO on systemic and cerebral blood gas distribution, which vary according to the cannulation site, extracorporeal blood flow and the presence and magnitude of intrinsic cardiac output. Only with such expertise can a reliable ABG sampling site be identified that accurately reflects the gas tensions at the level of the medulla [18]. Otherwise, it is possible that the PaCO2 has not reached the required threshold there without this being detected. The RL-IHA mandates that a PaCO2 ≥ 60 mm Hg must be achieved during the apnea test, even in patients on ECMO but without distinguishing between ECMO configurations or specifying the ABG sampling side. Other sources suggest ABG sampling from a distal arterial side and additionally from the ECMO circuit post-oxygenator [11] or bilaterally from both right and left radial arteries [25]. The World Brain Death Project, a consensus statement by an international, multidisciplinary panel on the determination of BD, also clearly recommends that PaCO2 should be measured both from a distal arterial site and post-oxygenator in potential donors on VA-ECMO in case of intrinsic cardiac output [9]. The RL-IHA grants physicians considerable discretion in this decision-making process, merely stating that physicians experienced in ECMO treatment must perform the assessment. While this approach offers the advantage of allowing physicians to individually select the most appropriate ABG sampling site for the specific clinical context, the absence of an explicit recommendation to perform the analysis at two separate sampling sites, at least in cases of intrinsic cardiac output, can result in nonstandardized and potentially contestable findings. This degree of procedural flexibility is particularly striking giving that, in contrast, explicit prohibitions are defined for ancillary tests, leaving no room for individual clinical judgement. Specifically, the use of methods aimed at demonstrating cerebral circulatory arrest to prove irreversibility is explicitly not permitted in the current version of the RL-IHA [7]. Proof of irreversibility in donors on VA-ECMO According to the RL-IHA, irreversibility of BD diagnosis must be confirmed following the initial clinical examination. This is achieved (in patients older than 2 years) either by a second clinical examination after a defined observation period (12 h in cases of primary supratentorial lesions and 72 h in secondary lesions) or by ancillary testing without a waiting period. In cases of infratentorial lesions, ancillary testing is mandatory. In clinical practice, ancillary testing is the preferred method in more than 80% of BD diagnoses in Germany, presumably for organizational reasons or due to concerns about potential deterioration of organ function due to a long waiting period, especially after secondary lesions [10]. Consistent with this practice, ancillary testing was performed in all cases analyzed in our study; however, the BÄK restricts the use of ancillary testing in potential donors on VA-ECMO. It justifies the prohibition of CBF-based methods with insufficient validation in potential donors on VA-ECMO [7]. Concerns have been raised, among others, because nonpulsatility can influence the autoregulation of cerebral blood flow [21, 25] and competing contrast media flows may occur with preserved pulsatility [23]. Moreover, it remains to be clarified whether the stipulated minimum MAD of ≥ 60 mm Hg during CT‑A also constitutes the applicable threshold on VA-ECMO support [25]. The RL-IHA defines standardized CT‑A acquisition parameters, including contrast medium volume, injection rate and the delay between contrast enhancement in the common carotid artery and initiation of the spiral scan; however, these specifications may be suboptimal in donors on VA-ECMO as extracorporeal circulation may significantly affect contrast transit times and effective distribution volume. The use of SPECT evaluates cerebral tissue perfusion using radiotracers that are metabolically taken up, remaining at an almost unchanged concentration for several hours. In contrast to CT‑A it does rely on tracking the passage of a dynamic contrast agent bolus. In addition, thoracic and abdominal scintigraphy must be performed as an in vivo quality control to confirm the physiological distribution of the radiopharmaceutical agent. The use of ECMO is thought to have little impact on the kinetics of the radiopharmaceutical agent [23] but in principle, altered pharmacokinetics might be expected under ECMO [16]. Cases 4 and 8 provide important information that may help to address these concerns. In both cases, CT‑A demonstrated cessation of cerebral perfusion and EEG demonstrated isoelectric EEG (albeit with artefacts in case 4). Each method thus confirmed the other in demonstrating complete loss of brain function. Additionally, cases 1–3 demonstrated cerebral perfusion arrest using SPECT. This indicates that with a thorough approach, CBF-based methods can be safely performed for BD determination in donors on ECMO support. As previously noted, additional studies are warranted to investigate cerebral contrast medium distribution on VA-ECMO support. Such investigations can help determine whether CT‑A acquisition parameters require adjustment to account for the altered hemodynamic conditions associated with extracorporeal circulation. Regarding SPECT, given the inclusion of in vivo quality control procedures, it remains unclear why non-pulsatile flow and extracorporeal circulation should limit the diagnostic reliability. To our knowledge, no published data exist that demonstrate a reduced diagnostic validity in this context. Accordingly, as far as we know, CBF-based methods are not explicitly excluded in any international guidelines on BD determination. Only guidelines from the United Kingdom recommend using CT‑A for this purpose only on a case-by-case basis when no alternatives are available [24]. The RL-IHA also mandates ancillary testing based on CBF-methods whenever the results of the clinical examination are inconclusive. In patients supported by VA-ECMO, this implies that the determination of BD is not possible if, for instance, the apnea test cannot be performed, a circumstance reported in approximately 22% of potential donors on ECMO [18]. Consequently, organ donation would, by definition, not be possible in this situation. The available data on ancillary testing via electrophysiological examinations during VA-ECMO support are generally limited [15]. Although the RL-IHA does not specify particular restrictions on the use of EEG as an ancillary test in patients on VA-ECMO support, making it potentially suitable in this setting, EEG recordings in the intensive care unit are frequently subject to artefacts [26]. Additionally, artefacts generated by ECMO devices have also been reported, which complicates the use of EEG in patients on ECMO [14]. Our cases support these anticipated challenges, suggesting that EEG may not be the most suitable method for determining BD in potential donors on ECMO due to possible artefacts [19]. Potential of DBD donors on VA-ECMO support In this study 4 potential donors on VA-ECMO became utilized donors, representing 15% of all utilized donors during the observation period. Notably, in each of these cases, ECPR was the indication for initiation of VA-ECMO support, representing 6% of all cases with ECPR at our institution. As emerging evidence suggests improved outcomes after ECPR in selected patient populations, the use of this modality is likely to increase [17]. This development can, in turn, lead to an increasing number of potential donors after ECPR via VA-ECMO; however, data on the current incidence and utilization of ECPR in Germany remain limited. Results from a pilot phase of the German Cardiac Arrest Registry identified 61 patients with ECPR among 559 patients treated for OHCA over a 16-month period across 15 centers. The report, however, does not provide any information about the incidence of BD and organ donation in these patients. Notably, anoxic brain injury was the leading cause of death in 38% of all cases, suggesting a relevant proportion of potential DBD donors [20]. Limitations The comparatively high proportion of organ donors on VA-ECMO support among all donors may also reflect the large number of ECPR procedures performed at our institution. Consequently, the generalizability of these findings to overall potential donor numbers on ECMO across Germany is limited. We did not address the potential impact of VA-ECMO treatment on pharmacokinetics and pharmacodynamics, which may be relevant when evaluating the residual effects of sedative and analgesic agents that must be fully metabolized or eliminated prior to testing for the diagnosis of BD. Moreover, we were unable to assess the influence of VA-ECMO support on other ancillary testing modalities, such as transcranial Doppler ultrasound, brainstem auditory evoked potentials, or somatosensory evoked potentials. The duration of apnea testing was estimated based on the time points of ABG sampling results, which may not accurately reflect the actual time interval from the beginning to the end of the apnea test. Conclusion Adult DBD donors on VA-ECMO accounted for 15% of all donors at our institution.With appropriate ECMO settings, the apnea test can be performed safely.Varying gas tensions may be present during apnea testing depending on the ABG sampling site.Describing suitable ABG sampling sites during apnea testing depending on ECMO configurations and intrinsic cardiac output should be considered for inclusion in future revisions of the RL-IHA.CT‑A during VA-ECMO support is technically challenging, as extracorporeal circulation may reduce contrast opacification despite preserved cerebral perfusion, whereas SPECT appears less susceptible to these effects when assessing the absence of CBF.Given the overall low donation numbers, it appears imperative to generate further evidence to not exclude ancillary testing methods without clear justification, as such exclusion could further compromise the availability of eligible donors.The lack of national data in Germany underlines the need for further studies to better understand the feasibility, challenges and frequency of organ donation in ECMO-supported patients. Adult DBD donors on VA-ECMO accounted for 15% of all donors at our institution. Adult DBD donors on VA-ECMO accounted for 15% of all donors at our institution. With appropriate ECMO settings, the apnea test can be performed safely. With appropriate ECMO settings, the apnea test can be performed safely. Varying gas tensions may be present during apnea testing depending on the ABG sampling site. Varying gas tensions may be present during apnea testing depending on the ABG sampling site. Describing suitable ABG sampling sites during apnea testing depending on ECMO configurations and intrinsic cardiac output should be considered for inclusion in future revisions of the RL-IHA. Describing suitable ABG sampling sites during apnea testing depending on ECMO configurations and intrinsic cardiac output should be considered for inclusion in future revisions of the RL-IHA. CT‑A during VA-ECMO support is technically challenging, as extracorporeal circulation may reduce contrast opacification despite preserved cerebral perfusion, whereas SPECT appears less susceptible to these effects when assessing the absence of CBF. CT‑A during VA-ECMO support is technically challenging, as extracorporeal circulation may reduce contrast opacification despite preserved cerebral perfusion, whereas SPECT appears less susceptible to these effects when assessing the absence of CBF. Given the overall low donation numbers, it appears imperative to generate further evidence to not exclude ancillary testing methods without clear justification, as such exclusion could further compromise the availability of eligible donors. Given the overall low donation numbers, it appears imperative to generate further evidence to not exclude ancillary testing methods without clear justification, as such exclusion could further compromise the availability of eligible donors. The lack of national data in Germany underlines the need for further studies to better understand the feasibility, challenges and frequency of organ donation in ECMO-supported patients. The lack of national data in Germany underlines the need for further studies to better understand the feasibility, challenges and frequency of organ donation in ECMO-supported patients.