Does microbiological contamination of homografts prior to decontamination affect the outcome after right ventricular outflow tract reconstruction?

Abstract

OBJECTIVES

Homografts are often in short supply. Today, European guidelines recommend that all tissues contaminated by any of 18 different bacteria and fungi be discarded before antibiotic decontamination has been conducted. The tissue bank in Lund uses more liberal protocols: It accepts all microbes prior to decontamination except multiresistant microbes and Pseudomonas species. The aim of this study was to analyse the effect of contamination on the long-term outcome and occurrence of endocarditis in recipients.

METHODS

Data were collected on homografts and on recipients of homografts in the right ventricular (RV) outflow tract who were operated on between 1995 and 2018 in Lund. The long-term outcome of recipients was analysed in relation to different types of contamination using Cox proportional hazard regression. The proportion of patients with endocarditis was analysed with the χ² test.

RESULTS

The study included 509 implanted homografts. Follow-up was a maximum of 24 years and 99% complete. A total of 156 (31%) homografts were contaminated prior to antibiotic decontamination. Homografts contaminated with low-risk microbes had the lowest reintervention rate, but there was no significant difference compared to no contamination [hazard ratio (HR) 1.1, 95% confidence interval (CI) 0.73–1.7] or contamination with high-risk microbes (HR 1.6, 95% CI 0.87–2.8) in the multivariable analysis. There was no significant difference in the proportion of cases of endocarditis during the follow-up period between recipients of homografts contaminated prior to decontamination and recipients of homografts with no contamination (P = 0.83).

CONCLUSIONS

Contamination of homograft tissue prior to decontamination did not show any significant effect on the long-term outcome or the occurrence of endocarditis after implantation in the RV outflow tract. Most contaminated homografts can be used safely after approved decontamination.

Homografts have been a well proven conduit for right ventricular outflow tract (RVOT) reconstruction since it was introduced by Ross in 1966 [1].

INTRODUCTION

Homografts have been a well proven conduit for right ventricular outflow tract (RVOT) reconstruction since it was introduced by Ross in 1966 [1]. Access to homografts is ensured by authorized tissue banks. European tissue establishments use guidelines developed by the European Directorate for the Quality of Medicine and Health Care (EDQM) [2]. Most guidelines are mainly based on experience, and the majority of tissue banks make local adaptions when developing their protocols. The result is a plethora of protocols affecting the number of homografts accepted as transplants [3, 4].

EDQM guidelines provide recommendations for discarding homografts because of microbiological contamination to avoid disease transmission to the recipient. The guidelines include a suggested list of microbes (Table 1) and advise that cultures that are positive for any of the microbes on the list, at any stage of the process (including pre-decontamination), should lead to disposal of all cardiovascular tissue from that donor [2]. Many tissue banks follow these recommendations, the result being that microbiological contamination is the main reason for discard of homografts [3, 4].

Table 1:

Contaminants that should result in tissue being discarded if detected at any stage of processing according to the European Directorate for the Quality of Medicine and Health Care [2]

High-risk microbes
Aspergillus spp.
Candida spp.
Clostridium spp. (notably C. perfringens or C. tetani)
Enterococcus spp.
Flavobacterium meningiosepticum
Klebsiella rhinoscleromatis
Listeria monocytogenes
Methicillin-resistant Staphylococcus aureus
Mucor spp.
Mycobacterium spp.
Neisseria gonorrhoeae
Nocardia spp.
Penicillium spp.
Pseudomonas aeruginosa or P. pseudomallei
Salmonella spp.
Shigella spp.
Streptococcus pyogenes (group A)
Other yeast and fungi

Studies have shown contamination rates of 14–84% in cardiovascular tissue prior to decontamination, with higher rates in homografts retrieved from non-heart beating donors (NHBDs) compared to multiorgan donors (MODs) [5–10]. The most common microbiological findings are low virulent skin flora such as coagulase-negative Staphylococcus and Cutibacterium acnes, followed by more virulent bacteria such as different species of Streptococcus, Staphylococcus aureus and Clostridium species [5, 7–11].

The aim of this study was to investigate 2 aspects of microbiological contamination. The primary objective was to determine if contamination prior to decontamination has any negative effects on homograft tissue by affecting its long-term durability. The secondary objective was to evaluate the risk of endocarditis after implanting a homograft that has been contaminated at any point during the processing of tissue compared to a homograft that was not contaminated during processing.

METHODS

The regional ethical review board in Lund, Sweden approved the study (Dnr 2017/133). Data were collected retrospectively.

Homografts

Data were collected on homografts from the tissue bank in Lund between 1995 and 2018. The maximal donor age was 65 years, but a few older donors were accepted after their medical history was closely reviewed. Maximal ischaemic time was 48 h, with a recommended maximum of 6 h of warm ischaemic time. Donor types were MOD, domino donors (donation of the explanted heart after a heart transplant) and NHBD. Tissue from MOD and domino donors is retrieved in operating theatres by a transplant surgeon or a cardiothoracic surgeon. Tissue from NHBD is retrieved from departments of forensic medicine, in the autopsy room in the clinical pathology department or in morgues under sterile conditions, by specially trained personnel in teams of 2–3 persons.

Antibiotic decontamination and microbiological cultures

Tissues are prepared in a laminar flow cabinet under sterile conditions. Since 2010, the laminar flow cabinet has been placed in a specially designed clean room. After preparation, the homograft is placed in an antibiotic cocktail of Medium199 Earle’s salt (200 ml), vancomycin (100 mg), gentamicin (106 mg) and amphotericin B (50 mg). Earlier, both colistin and ketoconazole were included in the cocktail, but they were excluded in 2013 and 2011, respectively. The recommended time in the antibiotic solution is 24–48 h at 4°C, with a maximum of 72 h.

Four cultures are collected during the processing of the tissue:

1. Tissue culture at preparation, prior to antibiotic decontamination.

2. Tissue culture prior to cryopreservation (after decontamination). The tissue is rinsed twice and then put in a cryoprotectant solution for 90 min prior to culture collection.

3. Culture from the cryoprotectant solution, after the homograft has been placed there for 90 min.

4. Tissue culture after thawing and right before the homograft is implanted, collected in the operating room.

For details on culture methods, please see Supplementary Material, Methods 1.

The growth of methicillin-resistant S. aureus, vancomycin-resistant Enterococcus, extended spectrum beta-lactamase or Pseudomonas in the culture medium leads to disposal of tissue at any stage in the procedure. All other microbes are accepted for further evaluation if the culture is positive prior to antibiotic decontamination only, meaning that the tissue bank in Lund has less strict recommendations than the EDQM guidelines (Table 1). If cultures are positive after antibiotic decontamination, the tissue is discarded. If the culture collected at the implantation is positive, the recipient will be closely monitored for signs of infection postoperatively and treated with antibiotics accordingly.

Recipients

This study includes recipients of a homograft from the tissue bank in Lund, implanted in the RVOT and operated on between 1995 and 2018 in the paediatric cardiac surgery unit or the adult cardiothoracic department in Lund. Data were collected from medical records, the Swedish Registry for Congenital Heart Disease and the cause of death register of the Swedish National Board of Health and Welfare. Follow-up started at implantation and finished 31 December 2019. The end point was defined as homograft failure requiring reintervention (surgical or endovascular) or death. Recipients who died of non-homograft-related events were censored at the time of their death. Some surgical interventions were considered non-homograft related but related rather to the complexity of the heart defect. These recipients were censored at their intervention date; the interventions included heart transplants, total cavopulmonary connection, the Glenn procedure and reimplantation of the autograft in the pulmonary position after early failure in the aortic position after a Ross procedure.

Follow-up was conducted at the recipient’s domicile hospital. Decisions on reintervention were made on an individual basis in a multidisciplinary conference. Indications for discussion were decreased physical capacity, progressive right ventricular (RV) dilatation, RV end diastolic volume >150 ml/m² body surface area, RV ejection fraction <45%, pulmonary stenosis with peak gradient >50 mmHg or >4 m/s on Doppler recording, pulmonary regurgitation fraction >40%, tricuspid regurgitation or ventricular arrhythmias.

Endocarditis during the follow-up period was defined as ‘definite’ or ‘possible’ according to the Duke criteria [12]. Positive blood cultures from the endocarditis diagnosis were compared to cultures from the homograft during preparation, to evaluate possible transmission.

Statistical analyses

Continuous variables are presented as median with interquartile range and total range. Categorical variables are presented with absolute numbers and percentages. Differences between proportions were analysed with the χ² test. The confidence level was set at 95%. There was no correction for multiple testing. Study objectives were defined prior to the statistical analyses.

Freedom from reintervention was analysed with univariable and multivariable Cox proportional hazard regression. In categorical variables, the group with the lowest proportion of reinterventions was used as the reference group. Risk factors analysed in the univariable model were chosen because of clinical importance and include microbiological contamination, homograft type, homograft size, donor age, donor type, anatomical position at implantation, recipient age and era of surgery. Proportional hazard assumptions were checked with graphical inspection of Kaplan–Meier curves and by checking Schoenfeld residuals after univariable analysis. Continuous variables were temporarily transformed into categorical variables for proportional assumptions analysis but were kept as continuous variables in the regression models.

Microbiological contamination was analysed as a categorical variable, with groups defined as ‘no contamination’, ‘contamination with low-risk microbes’ and ‘contamination with high-risk microbes’. High-risk microbes were defined as microbes from the EDQM table (Table 1). All other microbes were defined as low-risk. Homografts that were contaminated with both low and high-risk microbes were defined as high-risk. In donor types, NHBD were divided into NHBD with up to 24 h and >24 h of ischaemic time, corresponding to different cut-offs set by tissue banks [3]. The surgical eras were divided into 3 groups with equally sized timespans.

All variables were considered for inclusion in the multivariable model. To avoid multicollinearity, correlation between continuous variables was checked with Pearson’s test of correlation.

A backward stepwise Cox proportional hazards model was performed, including all variables except the variable of interest (microbiological contamination). Variables with a P-value <0.1 were kept in the model. The variable of interest was included in the final model.

Analyses were performed with Stata Statistical Software: Release 15, 2017 (StataCorp LLC, College Station, TX, USA).

RESULTS

Donors and homografts

There were 2860 homografts collected, with 1273 (45%) discards. Only 37 (2.9%) discards were due to contamination. Four homografts were discarded due to positive cultures prior to decontamination (3 cases of Pseudomonas spp. and 1 case of extensive growth identified in the histopathological biopsy).

Donor characteristics for homografts implanted in the RVOT are summarized in Table 2. For more details on donor characteristics on all homografts and in different contamination groups (see Supplementary Material, Table S1).

Table 2:

Donor and homograft characteristics for homografts implanted in the right ventricular outflow tract

	n = 509
	Median/n	IQR/%	Range
Donor age (years)	29	15–48	0–67
Donor gender
Male	292	57
Female	213	42
Missing	4	0.79
Donor type
Multi	151	30
Domino	63	12
NHBD (1–24 h)	112	22
NHBD (>24 h)	183	36
Homograft type
PV	417	82
AV	92	18
Homograft size (mm)	22	18–24	7–30
Microbiological contamination pre-decontamination
No	351	69
Low-risk microbe	123	24
High-risk microbe	35	6.8

AV: aortic valve; IQR: interquartile range; NHBD: non-heart beating donor; PV: pulmonary valve.

Contamination

The proportion of contaminated homografts during different steps of the collection process is described in Fig. 1. Of 690 contaminated homografts, 499 (72%) were contaminated with a single microbe; 191 (18%) were contaminated with 2 or more microbes; and 140 (20%) homografts were contaminated with high-risk microbes. Contamination was most frequent in NHBD (44%) compared to MOD (20%) and domino donors (17%). The difference was significant (P < 0.001). There was no significant difference in the contamination rate when comparing ischaemic times of 1–24 h (42%) and >24 h (45%) in NHBD (P = 0.32). There was no significant difference in the contamination rate between donor types after decontamination (2.4% in NHBD, 1.3% in MOD, 1.9% in domino donors; P = 0.40).

Figure 1:

Flow chart of homograft destiny, showing proportion of contaminated homografts at different steps in the collection process. RVOT: right ventricular outflow tract.

Of all the negative cultures (n = 1537) prior to decontamination, there were 16 (1.0%) positive cultures after decontamination. Three homografts were positive in the culture retrieved at implantation (see below for further details).

The types of microbes pre- and post-decontamination are presented in Table 3. Some species are only defined by gram-positive status and shape (rod, cocci); these could not be further defined in the cultures.

Table 3:

Microbes before and after antibiotic decontamination

	All collected homografts with culture growth
	Prior to decontamination		Post-decontamination
	n = 690	%	n = 33	%
Cutibacterium acnes	247	36	6	18
Alpha streptococcus	212	31	5	15
Coagulase-negative staphylococci	128	19	1	3.0
Clostridium spp.	61	8.8	2	6.1
Staphylococcus aureus	43	6.2
Corynebacterium spp.	31	4.5	4	12
G+ cocci	27	3.9
Enterococcus spp.	21	3.0
Streptococcus spp.	18	2.6
Group B streptococcus	14	2.0
Candida spp.	9	1.3	5	15
Pseudomonas aeruginosa	9	1.3	2	6.1
G− rod	6	0.87	2	6.1
Klebsiella oxytoca	5	0.72
G+ rod	3	0.43	1	3.0
Neisseria spp.	2	0.29
Aspergillus	2	0.29	2	6.1
Serratia spp.	2	0.29	1	3.0
Stenotrophomonas maltophilia	2	0.29	2	6.1
Flavobacterium spp.	1	0.14
Othera	91	13
Missing	5	0.72	2	6.1

Some tissues are contaminated with more than 1 microbe.

^aMicrobes constituting <2% of contamination, only occurring pre-decontamination and classified as low-risk.

Recipients

A total of 461 recipients received 535 homografts in the RVOT. Nineteen recipients were excluded because they moved abroad shortly after the operation. A total of 442 recipients who received 509 homografts in the RVOT were included in the analysis. Recipient characteristics are presented in Table 4. Median follow-up was 9.9 years (4.7–14.6; range 0.17–24 years). Follow-up was 99% complete of the 509 implanted homografts, 156 (31%) were contaminated prior to decontamination. Three (0.59%) homografts had a positive culture at implantation.

Table 4:

Recipients of homografts in the right ventricular outflow tract between 1995 and 2018

	All (n = 509)		No contamination (n = 351)		Low-risk contamination (n = 123)		High-risk contamination (n = 35)
	Median	IQR	Median	IQR	Median	IQR	Median	IQR
Recipient age (years)	12.6	4.9–24	12.0	5.0–22	14.6	7.9–27	4.6	1.4–17
Diagnosis	n	%	n	%	n	%	n	%
PA, VSD	18	3.5	12	3.4	3	2.4	3	8.6
PA, VSD, MAPCA	23	4.5	15	4.3	5	4.1	3	8.6
TGA, VSD, Pulmonary stenosis	9	1.8	5	1.4	2	1.6	2	5.7
Truncus arteriosus	34	6.7	25	7.1	3	2.4	6	17
PA, IVS	12	2.4	11	3.1			1	2.9
Conduit exchange	154	30	105	30	41	33	8	23
Tetralogy of Fallota	118	23	86	25	31	25	1	2.9
Ross	81	16	51	15	26	21	4	11
Otherb	60	12	41	12	12	9.8	7	20
Gender
Male	300	59	210	60	77	63	13	37
Anatomical position
Anatomical	286	56	200	57	74	60	12	34
Extra-anatomical	223	44	151	43	49	40	23	66
Conduit number
First	355	70	246	70	82	67	27	77
Second	118	23	83	24	29	24	6	17
Third	30	5.9	21	6.0	7	5.7	2	5.7
Fourth and more	6	1.2	1	0.28	5	4.1
Year of surgery
1995–2002	143	28	80	23	43	35	20	57
2003–2010	237	47	179	51	48	39	10	29
2011–2018	129	25	92	26	32	26	5	14

^aRecipients previously corrected for tetralogy of Fallot who developed postoperative pulmonary insufficiency or stenosis.

^bIncludes absent pulmonary valve syndrome, isolated pulmonary insufficiency or stenosis, congenital corrected transposition and double-outlet right ventricle with associated pulmonary insufficiency or stenosis.

IQR: interquartile range; IVS: intact ventricular septum; MAPCA: major aortopulmonary collateral artery; PA: pulmonary atresia; TGA: transposition of the great arteries; VSD: ventricular septal defect.

The first case was a 1-year-old boy undergoing the Ross procedure. The tissue culture was positive for a Gram-negative rod (all other cultures were negative). Postoperatively, the patient had prolonged antibiotics due to pleural effusions. He was discharged after 9 days. At the end point, 5 years had passed, and the homograft showed good function.

The second case was a 12-year-old girl undergoing conduit replacement. The tissue culture was positive for Candida albicans (cultures prior to decontamination were positive for C. acnes). Her postoperative recovery was normal. The homograft was explanted after 13 years due to stenosis.

The third case was a 13-year-old boy undergoing conduit replacement. The tissue culture was positive for Aerococcus viridans (cultures prior to decontamination were positive for Streptococcus anginosus). During the postoperative period, antibiotic treatment was prolonged due to high levels of C-reactive protein (maximum 340), fever and diarrhoea. Blood cultures were negative, but faeces cultures were positive for Clostridium difficile. He was discharged after 12 days without signs of infection, and the homograft culture was not ready until after discharge. After consulting an infection specialist, he received 14 days of ampicillin orally as a safety measure. The patient recovered well. At the end point, 3 years had passed, and the homograft had mild signs of conduit obstruction without indication for reintervention.

Endocarditis

There were 18 (3.5%) cases of endocarditis during the follow-up period (Table 5). Six (33%) cases required surgical intervention. There was no statistical difference in the proportion of cases of endocarditis between recipients of a homograft with no contamination during the tissue processes and the recipients of a homograft that was contaminated prior to antibiotic decontamination (P = 0.83).

Table 5:

Cases of endocarditis during follow-up after homograft implantation in the right ventricular outflow tract

	No contamination			Contamination prior to decontamination
	n = 351a	%	Median time to onset (years)	n = 158b	%	Median time to onset (years)
No endocarditis	336	97		151	96
Endocarditis	12	3.4	8.4 (5.7–13)	6	3.8	5.2 (2.2–13)
Possiblec	6	1.7	5.7 (1.1–11)	4	2.5	5.2 (2.3–10)
Definitec	6	1.7	9.9 (7.0–15)	2	1.3	10 (2.1–19)

^aThree cases lost to follow-up.

^bOne case lost to follow-up.

^cAccording to the Duke criteria [12].

All 6 cases of endocarditis in a homograft that had been contaminated prior to decontamination were closely reviewed. The first case of endocarditis within this group occurred 2 years and 2 months after the implant. Two cases were from homografts contaminated with high-risk microbes prior to decontamination (Clostridium sp. and Enterococcus sp.). Cultures existing at the time of the endocarditis diagnosis showed no growth of these bacteria. One case of endocarditis showed similarities between homograft tissue cultures at preparation and blood cultures at endocarditis diagnosis. The patient was a 17-year-old boy who received a homograft in the RVOT due to pulmonary stenosis and developed endocarditis at 2 years and 3 months after surgery. The homograft culture showed growth of alpha-haemolytic streptococci prior to antibiotic decontamination; all follow-up cultures were negative. Blood culture at endocarditis diagnosis showed growth of alpha-haemolytic streptococci as well. The patient was treated successfully with antibiotics. At the end point, 5 years had passed since the implant, and the homograft was still functioning well.

Mortality

Twenty-four (5.4%) recipients died during the follow-up period. There were 14 cardiac events, 8 non-cardiac events, 1 postoperative complication and 1 homograft-related complication, with rupture of a pseudoaneurysm adjacent to the suture line. The case was closely reviewed, but there were no abnormalities in the collection or processing of the homograft.

Reintervention

There were 136 (27%) homograft-related reinterventions during follow-up. Median time to reintervention was 7.5 years (3.3–10, range 0.26–23). Of 136 reinterventions, 109 (80%) were open heart operations with conduit replacement, 26 (19%) were endovascular interventions and 1 (0.74%) was open heart surgery with other interventions. Eight reinterventions were considered non-homograft related, including 5 cases having a heart transplant, 1 case with a total cavopulmonary connection, 1 case with a Glenn procedure and 1 case with autograft replacement in the RVOT after failure in the aortic position.

Freedom from reintervention was 98% [95% confidence interval (CI) 96–99%] at 1 year, 89% (95% CI 86–92%) at 5 years, 76% (95% CI 71–80%) at 10 years and 57% (95% CI 50–64%) at 20 years.

Pearson’s test of correlation showed the highest correlation between homograft size and donor age. Donor age was excluded from the multivariable model to avoid collinearity.

Results from the univariable and backward stepwise multivariable Cox proportional hazard regression are shown in Table 6. All variables fulfilled the proportional hazard assumption with proportional curves in the Kaplan–Meier graphs and P-value >0.05 when analysing Schoenfeld residuals. The univariable model shows significant differences in reintervention rate in homografts contaminated with a high-risk microbe compared to the reference group (P < 0.001). The result could not be reproduced in the multivariable model.

Table 6:

Univariable and multivariable analysis of potential risk factors in relation to time to reintervention after homograft implantation in the right ventricular outflow tract

	Univariable analysis			Multivariable analysis
	HR	95% CI	P-value	HR	95% CI	P-value
Microbiological contamination pre-decontamination
No	1.3	0.82–1.9	0.28	1.1	0.73–1.7	0.58
Low-risk microbe	1.0			1.0
High-risk microbe	2.8	1.6–5.1	<0.001	1.6	0.87–2.8	0.13
Homograft type
Pulmonary	1.0
Aortic	4.1	2.9–5.8	<0.001	1.4	0.96–2.1	0.081
Homograft size	0.80	0.77–0.83	<0.001	0.88	0.83–0.93	<0.001
Donor age	0.94	0.93–0.95	<0.001
Donor type
MOD	1.0
Domino donor	1.9	1.0–3.6	0.041
NHBD 1–24 h	2.9	1.7–4.9	<0.001
NHBD >24 h	2.0	1.2–3.3	0.007
Anatomical position
Anatomical	1.0
Non-anatomical	4.1	2.8–6.0	<0.001
Recipient age	0.90	0.88–0.92	<0.001	0.96	0.93–0.98	0.002
Era of surgery
1995–2002	1.1	0.61–2.1	0.676
2003–2010	0.86	0.47–1.6	0.638
2011–2018	1.0

CI: confidence interval; HR: hazard ratio; MOD: multiorgan donor; NHBD: non-heart-beating donor.

DISCUSSION

This study is the first to investigate the impact of microbiological contamination in relation to long-term outcome in the homograft recipients and the risk of endocarditis. The proportion of contamination was 31% in all homografts prior to decontamination, similar to the proportion in other tissue banks that show both higher and lower contamination rates [7, 9, 11, 13].

NHBD had the highest contamination rate (44%), but there was no significant difference between ischaemic times within the group. Other tissue banks have described contamination rates of 33–58% in NHBD with a maximum of 24 h ischaemic time [7, 8, 13]. We have not found any evidence that contamination rate increases if the cold ischaemic time is prolonged in NHBD [7, 14].

Three homografts showed microbiological growth from tissue cultures collected at implantation, most likely contaminated during unpacking in the operating theatre. All recipients recovered well, with no signs of postoperative endocarditis. Jashari et al. [15] described 3 cases of microbiological contamination of tissue at implantation when investigating homografts from the European Homograft Bank, with no complications due to infection postoperatively.

There were 18 cases of endocarditis during the follow-up period, with no difference in proportion when comparing homografts that were contaminated or not prior to antibiotic decontamination. One case showed the same kind of microbe in the blood culture at endocarditis diagnosis as in the culture from homograft tissue prior to decontamination (alpha-haemolytic streptococci). Considering the disease-free interval (over 2 years) and the presence of bacteria that are highly sensitive to the antibiotics included in our antibiotic cocktail, transmission from the homograft seems unlikely.

There are few reports in the literature about complications due to microbiological contamination of tissue. In 1998, one recipient developed C. albicans endocarditis 16 days after homograft implantation that was judged to originate from the homograft [16]. Wang et al. [17] summarized reported cases of allograft infections in the USA from 2001 to 2004, describing 35 reported homograft infections, with 9 deaths. We have not been able to find more information on these cases, and the authors emphasize that no verification was made to ensure that the homografts were actually the origin of the infection. They state that infections arising from allograft contamination could be difficult to differentiate from ‘normal’ postoperative infection, a conclusion made by Eastlund as well [18].

Results from the univariable model show no difference in long-term outcome when comparing different contamination groups. Homografts contaminated with high-risk microbes show a significantly higher reintervention rate in the univariable model, but the result could not be reproduced in the multivariable analysis.

The criteria for discarding tissue due to contamination need to be further investigated. EDQM guidelines if strictly followed lead to a high proportion of discards due to contamination of tissue. The guidelines are based mainly on expert opinions and experience, which is evident from the fact that the references cited in this chapter only describe contamination rates and types in cardiovascular tissue and evaluate the effect on the contamination rates of different decontamination regimes [5, 19]. Given the available data, homograft infection due to microbiological transmission from tissue seems to be rare. EDQM guidelines should be used as a valuable tool when developing protocols, but local circumstances have to be considered as well.

Suspected homograft infections should be reported and documented, and thorough evaluation of homograft tissue should be conducted. Further research should focus on finding the best decontamination technique and culture methods to ensure the safety of implanted homografts.

Limitation

A limitation of this study is the risk of false negative culture results, which could be affected by different culturing methods in tissue banks or by residual antibiotics in the tissue [20, 21]. The sensitivity of our methods has not been evaluated in comparison to those of other tissue banks, which decreases the generalizability of our result. Other limitations are the retrospective data collection method and the heterogeneity of the group in terms of age and diagnostic severity.

CONCLUSION

Contamination of tissue prior to decontamination does not seem to affect the long-term outcome or the occurrence of endocarditis in recipients. Transmission of microbes during homograft implantation seems to be rare, but both under- and over-reporting may be present due to diagnostic difficulties.

Many tissue banks report contamination as the main reason for discarding tissue. Our results indicate that no harm was done by using homografts that were contaminated prior to decontamination if adequate cultures after decontamination were negative. We agree with the recommendation that any tissue that is contaminated post-decontamination should always be discarded.

SUPPLEMENTARY MATERIAL

Supplementary material is available at ICVTS online.

Funding

This study was supported by grants from the Swedish Fund for Congenital Heart Defects and The Swedish Odd Fellow Jubilee Foundation.

Conflict of interest: none declared.

Author contributions

Ida Axelsson: Conceptualization; Data curation; Formal analysis; Funding acquisition; Methodology; Project administration; Resources; Writing—original draft; Writing—review & editing. Torsten Malm: Conceptualization; Funding acquisition; Methodology; Project administration; Supervision; Writing—review & editing. Johan Nilsson: Formal analysis; Methodology; Software; Supervision; Validation; Writing—review & editing.

Reviewer information

Interactive CardioVascular and Thoracic Surgery thanks Duccio Di Carlo, Hans-Heiner Kramer and the other, anonymous reviewer(s) for their contribution to the peer review process of this article.

ABBREVIATIONS

CI

Confidence interval

EDQM

European Directorate for the Quality of Medicine and Health Care

HR

Hazard ratio

MOD

Multiorgan donor

NHBD

Non-heart beating donor

RV

Right ventricular

RVOT

Right ventricular outflow tract