Time to let go of the biopsy? A new era in cardiac rejection monitoring?

Heart transplantation (HTx) is the optimal treatment for advanced heart failure in selected patients. Acute rejection is a potentially fatal complication. Endomyocardial biopsy (EMB) is still considered the gold standard for diagnosis of acute rejection according to current guidelines from the International Society of Heart and Lung Transplantation.¹

The article by Kieviet et al presents a 36-year, single-center experience encompassing 8,185 routine EMBs.² Over time, there was a notable reduction in the number of biopsies performed per patient. Most clinically relevant rejection episodes (≥2R) were identified within the first 6 months post-transplantation. Furthermore, the incidence of significant acute cellular rejection (ACR) decreased over the study period. Based on these findings, the authors rightfully question the necessity of routine EMB for all patients.

The lower ACR incidence may partly be related to the lower surveillance EMB frequency during the study period. However, the fact that outcomes improved while ACR risk declined indicates that some EMB-proven ACR episodes without hemodynamic compromise, symptoms, or changes in echocardiographic measurements and biomarkers are likely to resolve spontaneously or may even be classified as “false positive.”³ The finding that ACR was detected in only 0.8% of EMBs in the most recent era in the study by Kieviet et al is striking and raises the question if we should be doing routine screening for ACR at all, particularly beyond the first 3-6 months post-transplant? Clearly, the transplant society have made tremendous strides in ACR prevention, and perhaps we are now over-immunosuppressing our patients and placing them at risk of immunosuppressing-related complication.

Combined with the procedural risks (i.e., cardiac tamponade, tricuspid valve injury, and pneumothorax) and a notable interobserver variability in histopathological interpretation observation, the demonstrated low ACR risk challenges the paradigm of routine surveillance biopsies and suggests that an individualized, risk-stratified monitoring may be more appropriate. Despite the clear need to avoid unnecessary EMBs, no globally accepted guidelines for alternative monitoring strategies have yet been established. As a result, many European centers remain conservative, relying primarily on biopsy in combination with serological markers and cardiac imaging. Echocardiograms are widely available as a standard bedside method in many European centers. In contrast, in the United States, despite that relevant imaging with both advanced echocardiography and cardiac magnetic resonance is available in most transplant centers, there is an increasing adoption of commercially driven tests in rejection surveillance—such as gene expression profiling (GEP) for immune activation and donor-derived cell-free DNA (dd-cfDNA) for allograft injury. The increasing adoption of these non-invasive tests in the US likely reflects a combination of a wish for more granular data than just imaging as well as existing reimbursement structures. GEP (CareDx) quantifies 11 immune-related transcriptome expressions and offers a high negative predictive value (NPV) for ruling out ACR. GEP has been validated in 2 randomized clinical trials4, 5 and large prospective observational cohort studies (The OAR, SHORE, and CARGO II) as non-inferior to EMB for rejection surveillance. The main limitation of these studies, however, was a low number of biopsy-proven acute rejection (4%). Furthermore, the GEP test is affected by other factors leading to immune activation (i.e., steroids, infection, and leukopenia) and has not been validated for antibody-mediated rejection. Adoption of the GEP test in Europe is limited by cost considerations and establishing laboratory infrastructure for testing. However, GEP has received endorsement both in guidelines from the International Society of Heart and Lung Transplantation (2023) and the European Society of Transplantation (ESOT, 2024) as a reliable non-invasive diagnostic tool to rule out ACR in stable, low-risk heart transplant recipients, who are >55 days post transplantation.⁶

In the context of cellular injury or apoptosis, DNA fragments originating from donor cells are released into the recipient's bloodstream. The dd-cfDNA test quantifies these circulating fragments and thus provides a measure of allograft injury, although it may not distinguish between underlying causes. Commercial dd-cfDNA kits are now available for rejection screening in HTx patients (CareDX and Natera). However, to date, no randomized controlled trials have directly evaluated the noninferiority of dd-cfDNA-based monitoring compared to EMB-based surveillance for the detection of acute rejection. However, the clinical utility of dd-cfDNA has been supported by five prospective cohort studies conducted in North America (SHORE, GRAFT, D-OAR, and DEDUCE trials) and Spain (FreeDNA-CAR study). The dd-cfDNA test can be applied to both AMR and ACR monitoring with a >97% NPV for allograft rejection. Furthermore, a multicenter open-label Comparative Effectiveness Research study randomizing 250 patients to Prospera dd-cfDNA surveillance or EMB surveillance in a 2:1 ratio (the ACES-EMB study) is well on its way. However, until further evidence is present, only a weak recommendation has been given in the European Society of Transplantation guidelines for the use of dd-cfDNA.⁶

Replacing EMB with expensive noninvasive commercial tests such as GEP and dd-cfDNA for generalized use seems difficult to be implemented and requires reliable local laboratory infrastructure. The questions are whether generalized one-dimensional surveillance is needed at all and whether it can be replaced by a more easily accessible and rational strategy involving integrated clinical monitoring with more simple multimodal diagnostics.

Cardiac troponin is a key biomarker used in clinical cardiology to detect myocardial injury. Since myocyte injury is characteristic of moderate to severe ACR, elevated high-sensitivity troponin levels (hs-cTn) may be anticipated during rejection episodes. A systematic review by Fitzsimons et al⁷ showed that hs-cTn assays could have sufficient sensitivity (82%-100%) and NPV (97%-100%) to exclude ACR and limit the need for surveillance EMB. Natriuretic peptides, including B-type natriuretic peptide (BNP) and its inactive fragment NT-proBNP, are hormones released by the myocardium in response to atrial and ventricular wall stress. In HTx recipients, BNP levels have been found to correlate with higher grades of rejection and left ventricular dysfunction.⁸ Even though cardiac troponins and BNP levels may be raised during ACR, the pitfall is that they are too sensitive and not specific for rejection. Troponins and NT-proBNP are markedly elevated during the initial weeks following HTx but generally stabilize within the first 3 months. As such, the use of biomarkers in the ACR surveillance requires the establishment of reliable individual cut-off values. Furthermore, trends in serial measurements of these biomarkers or relative changes from baseline could provide a more reliable indication of ACR than isolated values, especially when combining with markers of immune activation or myocardial dysfunction.

Global longitudinal strain (GLS), measured by speckle tracking echocardiography, is a valuable tool for detecting early signs of graft dysfunction. It can identify myocardial dysfunction before other echocardiographic parameters show changes. Decreases in GLS magnitude over time are associated with moderate to severe acute rejection in HTx recipients.⁹ Combining serial measurements of simple biomarkers and echocardiography with GLS measurements have been shown to enhance diagnostic accuracy and offer a high NPV for ACR detection.³

Balancing immunosuppression on an individual level is key for long-term survival but requires reliable surveillance to establish how low we safely can go. The low ACR risk demonstrated in the study by Kieviet et al implies that the traditional approach of routine scheduled EMB is approaching obsolescence. Within the next years, we will have more exciting data on the robustness of GEP and dd-cfDNA. More data is needed to validate risk models combining trends in widely available cardiac biomarkers (i.e., hs-cTn and NT-proBNP) with functional imaging. Advances in emerging technologies, including artificial intelligence and machine learning, offer exciting opportunities to enhance the integration and clinical utility of such multimodal strategies.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.