A nationwide trend analysis on the usage of endomyocardial biopsy

Abstract Background Endomyocardial biopsy (EMB) is a safe procedure performed in diagnostic work‐up of cardiac disease. Hypothesis Data regarding temporal trends of total numbers, characteristics, in‐hospital outcomes, and complications of patients undergoing EMB are sparse. Methods The nationwide German inpatient sample (2005–2019) was used for this analysis. Patient cases of EBM during the 5‐year cycles from 2005 to 2009, 2010 to 2014, and 2015 to 2019 were compared, and temporal trends regarding total numbers and presumable major and minor EMB‐associated complications were investigated. Results Overall, 67 745 EMB were performed in Germany 2005–2019. Total number of EMB increased from 3083 in 2005 to 5646 in 2019 (β 0.40 [95% confidence interval [CI] 0.37–0.43], p < .001). Among these EMB, 19 083 (28.2%) were performed during the period 2005–2009, 22 867 (33.7%) 2010–2014, and 25 795 (38.1%) between 2015 and 2019. The proportion of patients aged ≥70 years was highest 2015–2019 (2005–2009: 9.3%; 2010–2014: 13.8%; 2015–2019: 16.1%, p < .001) and the most aggravated comorbidity profile (Charlson Comorbidity Index 2.25 ± 1.93; 2.67 ± 2.14; 3.01 ± 2.29, p < .001) was also detected 2015–2019. Major complications occurred less often in the period 2015–2019 compared to 2005–2009 (odds ratio [OR] 0.921 [95% CI 0.893–0.950], p < .001), whereas minor complications were more frequently observed between 2015 and 2019 (OR 1.067 [95% CI 1.042–1.093], p < .001). While a decrease in major complications was detected irrespective of age, an increase in minor complications was identified only in patients between 30–59 years. Conclusions Annual numbers of EMB increased significantly in Germany 2005–2019. Patients who underwent EMB in recent years were older and showed an aggravated comorbidity profile accompanied by fewer major complications, underscoring safety of the procedure.


| INTRODUCTION
The etiology of heart failure (HF) is in the majority of cases multifactorial, 1 and the underlying pathologies cannot always be identified by the most frequently used diagnostic procedures such as echocardiography, coronary angiography, and cardiac magnetic resonance imaging. 2,3Considering recent progress and improvements in development of new therapies for specific myocardial diseases, an accurate diagnosis is mandatory to choose the most appropriate and beneficial therapy for the individual patient.Endomyocardial biopsy (EMB) allows detailed tissue characterization and has been demonstrated to improve diagnostic accuracy in many cases of unexplained HF. 2,4 Historically, the first percutaneous EMB was reported by Konno and Shakakibara in the year 1963. 5,6Subsequent modifications by Caves and Schultz 7 and the simultaneously developed long sheath technique, improved the feasibility and safety of the EMB technique. 8 2014, the radial approach was introduced with the primary aim to reduce bleeding complications. 9At the outset, EMB was mainly used to monitor graft rejection after cardiac transplantations. 8,10Today, the spectrum of indications to perform EMB has considerably expanded.The established indications were recently listed and summarized in a joint consensus document by the Heart Failure Association of the European Society of Cardiology, the Heart Failure Society of America, and the Japanese Heart Failure Society. 113][14][15] However, data regarding EMB-associated complications and outcomes are predominantly derived from single-center experiences and registries.
Therefore, the objective of the present study was to investigate temporal trends regarding total numbers and complications of EMB in the German nationwide inpatient sample comprising all hospitalizations between 2005 and 2019.

| MATERIAL AND METHODS
We analyzed all hospitalizations of patients undergoing EMB (OPScodes 1-497.0,1-497.1,1-497.In Germany, patients' diagnoses are documented based on the coding guidelines ICD-10-GM (International Classification of Diseases, 10th Revision with German Modification).[18][19] In the present study, we selected and included all hospitalizations of patients with performed EMB identified by the OPS-codes 1-497.0,We analyzed the study sample for temporal trends and differences regarding patient characteristics inclusive patients' comorbidity profile, additional treatments, and outcomes between the patients who underwent EMB during these three different 5-year cycles.

| Study endpoints and in-hospital adverse events
The primary study outcomes were defined as presumable major as Obesity was defined according to the recommendations of the World Health Organization as a body mass index ≥30 kg/m². 20[23]

| Ethical aspects
By German law, approval by an ethical committee and patients' informed consent were not required for the epidemiological studies of the German nationwide inpatient sample since these study did not involve direct access of the study investigators on data of individual patients, but only on aggregated summarized data.

| Statistical methods
Temporal trends regarding annual and age-related hospitalizations of patients with EMB and temporal trends of treatments and adverse inhospital outcomes of patients who underwent EMB were calculated.
The highest number of EMB were performed in patients between the 5th and 7th decade of life, but decreased statistically with growing age (β per age-decade −1.44 [95% CI −1.45 to −1.42], p < .001)(Figure 1B).Indications to perform EMB and identified pathologies are illustrated in Figure 2. Cardiomyopathy was coded as the leading pathology in more than 50% of all EMB cases, followed by myocarditis.History of heart transplantation was the reason for performed EMB in approximately 1/5 of the cases.1).
Although the median age of patients was only slightly higher in the later 5-year cycles, the proportion of patients aged ≥70 years was approximately 7% higher in the last observational period (2015-2019) compared with the first observational period (2005-2009) (Table 1).
Patients undergoing EMB in later years (2015-2019) were more symptomatic as indicated by a higher NYHA-functional class.Importantly, patients who underwent an EMB in later 5-year cycles had an unfavorable comorbidity profile, reflected by a higher Charlson Comorbidity Index and an increasing prevalence of cardiovascular comorbidities (Table 1).In contrast, history of heart transplantation as the primary cause of EMB decreased from 28.5% of all EMB during the timeframe 2005-2009 to 17.7% during the period 2015-2019 (Table 1).
To analyze annual temporal trends, we calculated unadjusted linear regressions.While total annual numbers of major complications  1).3).

| DISCUSSION
The recommendations regarding the indications of EMB and the selection of patients for EMB were updated several times during the last decades.Nowadays, EMB is an established invasive procedure in daily cardiological routine. 24,255][26][27] Improvements in EMB equipment and techniques, in combination with significant progress regarding the histological and immune-histochemical examinations, analysis, and validity of the EMB specimens and samples, resulted in a significant improvement in diagnostic precision and usefulness of EMB. 3,25Previously published studies have shown that EMB is a widely safe procedure, and the changes overtime in interventional approach regarding localization of EMB (right vs. left ventricle) as well as procedure access (radial vs. femoral) seem not to have a significant impact on adverse events in patients undergoing EMB in large volume centers. 14,28Nevertheless, data of large nationwide studies including low volume as well as mid-and high-volume centers in respect to total annual EMB numbers are sparse; especially, for central Europe these data are missing.Thus, our study aimed to close this gap and to illustrate trends in EMB usage, rates of EMB-related and nonrelated adverse in-hospital events, and patients' accompanied treatment approaches.
The results of the present study can be summarized as follows: (I) Overall, more than 67 000 EMB were performed in Germany In accordance with our study results, national trend analyses of the United States of America (US) observed a steady increase of the annual numbers of the in-patient EMB procedures overtime. 29milarly to the US, patients who underwent EMB were predominantly of male sex, 29 whereas the median age of German patients with EMB was slightly higher than those of the US. 29In Germany, most patients who underwent EMB were frequently aged between the 5th and 7th age-decade.
5][26][27] In this context, patients who underwent EMB later during the 15-year observational period were more often aged ≥70 years and showed an aggravated comorbid profile.In line with the suggestion that primarily the change in guideline recommendations 3,[24][25][26][27] might explain this shift, the proportion of EMB in patients with unclear acute cardiac disease manifestations increased.In contrast, the proportion of planned control EMB in patients with a history of heart transplantation decreased.Remarkably, it is well known that both complication and case-fatality rates were substantially higher in EMB procedures in patients with native hearts (and acute and/or unknown cardiac diseases) compared to heart-transplanted patients with control EMB. 29,30In this context, we observed a substantial increase in the total numbers of major and minor complications and annual numbers of in-hospital case-fatality with patients' age.][15]25 Despite an annually increasing vulnerability of the patient group (with older age and aggravated comorbidity profile) undergoing EMB in combination with a decreasing proportion of planned control EMB in patients with a history of heart transplantation, the technical progress with improvements mainly driven by the use of smaller and more flexible bioptomes result in lower complications rates and might play key roles for this improvement of safety outcomes. 13,313][34][35] This is in line with previously published studies, demonstrating that higher EMBprocedure volumes were associated with lower complication rates after/peri-procedural. 3,12,25,35However, the annual number of EMB per operator, which is required to maintain the procedural skills, is still under debate.Recently, it has been suggested that the procedures per operator should exceed at least 20 procedures per year, but in other recommendations of the appropriate medical societies at least 50 procedures per operator per year. 3,25markably, despite MACCE rate increase in our study overtime primarily driven by the increase of ischemic stroke, the in-hospital case-fatality revealed unchanged after adjustment for age, sex, and comorbidities.While the increase in myocardial infarction and ischemic stroke overtime might be partly attributed to the higher age and aggravated comorbid profile of the patients undergoing EMB in later years, the increasing frequency of ischemic stroke might also be explained by higher numbers of left ventricular biopsies.Although it has been reported that both left ventricular EMB and right ventricular EMB are safe procedures with similar major complication rates if performed by experienced interventionalists, 13,14 ischemic stroke was reported as major complication in 0.3%-0.7% of studies in high-volume centers when left ventricular EMB was performed. 13,14 contrast, our real-world data of all-comers included in the German nationwide inpatient demonstrated a prevalence of ischemic strokes in patients who underwent EMB 2015-2019 of 1.5% (of all hospitalizations with performed EMB procedures in Germany).The high prevalence of ischemic stroke about EMB might be, on the one hand, driven by increasing numbers of left ventricular EMB, but on the other hand, driven by increasing patient age.In addition, the differences regarding the prevalence of the detected peri-procedural ischemic strokes might be the result of higher complication rates in hospitals with low EMB numbers since the German inpatient sample included all hospitalizations of patients with EMB regardless of EMB center volume, which is an important difference regarding studies of single high-volume centers.
Remarkably, we detected distinctly higher proportions regarding the combined outcomes of major and minor complications based on the present definitions of our study than in already published studies. 13,14The higher rates of major and minor complications as combined outcomes in our study compared to other studies might be primarily attributed to differences in the definitions of the combined outcomes minor and major complications in the different studies.
Additionally, due to the nature of the administrative data structure of the German nationwide inpatient sample, we could not verify that all of the peri-procedural adverse in-hospital outcomes and treatments were directly EMB related, but might also be the result of aggravation of the underlying disease, especially the performed interventional and surgical treatment, which were included in the combined outcomes of major and minor complications. 13,14,31Nevertheless, the primary objective of the present study focusing on the  .016organ transplant scandal with stagnating cardiac transplant numbers in later years. 36,37nce EMB is in most hospitals only performed in a small number of patients with the exception of high-volume centers, 3,25,38 analyses of large cohorts including patients with EMB treated in high-, mid-, and low-volume centers and especially of nationwide inpatient samples are of outstanding interest to identify important trends on performed procedures, patient characteristics and complications of patients who had undergo EMB. 16,39| LIMITATIONS Fourth, we could not distinguish between left and right ventricular EMB as well as ultrasound-guided versus fluoroscopy-guided EMB.

TAVR
Fifth, the exact timing and course of adverse events/complications during hospitalization (i.e., whether it was present on admission or a complication during the hospital stay) could not be determined, whereby EMB is a selective procedure that will in the vast majority, only be performed in stable settings and not in acute emergencies during adverse events.Thus, adverse in-hospital events are presumable in the vast majority peri-procedural complications.

| CONCLUSIONS
Annual total numbers of EMB increased significantly in Germany between 2005 and 2019.Although patients with EMB in later years were in median older with an aggravated and more severe comorbidity profile, major complications decreased during the observational period.
2) in Germany during the observational period between the years 2005 and 2019 (source: RDC of the Federal Statistical Office and the Statistical Offices of the federal states, diagnosis-related groups [DRG] Statistics 2005-2019, and own calculations).

1 -
497.1, 1-497.2during the observational period 2005-2019 in Germany.The identified and included hospitalization cases of this 15-year observational period were subdivided according to their year of hospitalization in three 5-year cycles: the first period includes the years 2005-2009, the second period the years 2010-2014 and the third period comprises the years 2015-2019.
Linear regressions were used to assess trends overtime, and the results are shown as beta (β) with corresponding 95% confidence intervals (CI).For further temporal comparisons, we subdivided the 15-year observational period of the present study into three different 5-year cycles comprising the years 2005-2009, 2010-2014, and 2015-2019 and compared the three periods.Descriptive statistical comparisons of patients who underwent EMB during the three different 5-year cycles were calculated as absolute numbers and corresponding percentages.We tested the comparison of the three different 5-year cycles for statistical difference with the help of the Kruskal-Wallis Test.In addition, we analyzed differences between the first and last 5-year cycle with logistic regressions.For this purpose, we compared the last 5-year cycle including the years 2015-2019 with the first 5-year cycle 2005-2009 (defined as the reference).We adjusted the logistic regressions used for analyzing (I) associations of being hospitalized in a later 5-year cycle with the different outcomes and (II) associations of being hospitalized in a later 5-year cycle with the usage of treatments for age, sex, obesity, cancer, HF, coronary artery disease, hyperlipidaemia, chronic obstructive pulmonary disease, essential arterial hypertension, renal insufficiency (glomerular filtration rate [GFR] < 60 mL/min/1.73m²), diabetes mellitus, and atrial fibrillation/flutter.These comparisons between hospitalizations during the first versus last investigated 5-year cycle computed by adjusted logistic regressions were performed for all patients with EMB regardless of age and additionally in the age subgroups <30 years, 30-59 years, and ≥60 years.Statistical significance was presumed for p < .05(two-sided).Statistical analyses were performed with the software SPSS ® (version 20.0; SPSS Inc.).

F I G U R E 2
Pathologies associated with endomyocardial biopsy.T A B L E 1 Baseline characteristics, medical history, presentation, and outcomes of the 67 745 patients with EMB stratified for the three analyzed timeframes(2005-2009, 2010-2014, and 2015-2019).
between 2005 and 2019.Annual total numbers of EMB continuously increased during this investigated timeframe.(II) Most patients who had to undergo EMB were most frequently aged between the 5th and 7th age-decade of life.(III) Patients who underwent EMB in the later 5-year cycles were older and sicker, reflected by a higher dyspnoea level in coprevalence with an aggravated comorbidity profile.(IV) Major complications decreased from 2005-2009 to 2015-2019, whereas prevalence of minor complications increased during the same period.While the decrease in major complications was regardless of age, the increase of minor complications was observed in patients aged 30-59 years but not in the younger and older age group.(V) While MACCE rate increased overtime, driven by an increase of ischemic stroke, in-hospital case fatality remained unchanged.(VI) Total numbers of major complications and minor complications, as well as annual numbers of in-hospital case-fatality, increased with patients' age.(VII) Since the annual numbers of EMB procedures, which were performed to establish the diagnosis of different cardiac disorders, increased overtime, the proportion of annual EMB procedures in patients with a history of heart transplantation as a primary cause of EMB decreased from 2005 to 2019.
age shift and aggravated comorbid profile of the patients undergoing EMB into account, major complications decreased from the timeframe 2005-2009 in comparison to the period 2015-2019.In contrast, the prevalence of minor complications increased during the same period.While the decrease in major complications was seen regardless of age and therefore, in all age groups, the increase regarding the occurrence of minor complications was observed in patients aged 30-59 years only.
identification of time trends regarding outcomes and safety of the EMB interventions with decreasing major complications in recent years could be impressively established.In this context, it must be mentioned that increasing numbers of EMB performed in patients with unclear cardiac diseases in combination with low left-ventricular output lead to a higher number of implanted ventricular assist devices in the patients who underwent EMB during the hospitalization 2010-2019 compared to those treated between 2005 and 2009.In contrast, heart transplantation surgeries were less often performed during later years in patients who underwent EMB, which might be explained by the German T A B L E 3 Comparison of first and last investigated 5-year cycles (years 2005-2009 vs. 2015-2019) of patients with myocardial biopsy (univariate and multivariate logistic regression model) stratified for age groups <30 years, 30-59 years, and ≥60 years.
Certain limitations regarding our study require consideration: First, the study results are based on ICD and OPS discharge codes of inpatients.This might lead to incomplete data based on underreporting/under-coding.Second, data about the administration of medications are not available in the data set of the Federal Statistical Office of Germany.Third, we could not provide follow-up data.
Comparison of first and last investigated 5-year cycles (years 2005-2009 vs. 2015-2019) of patients with myocardial biopsy (univariate and multivariate logistic regression model).
a Adjusted for age