Circulating donor-derived cell-free DNA as a marker for rejection after lung transplantation

Objective Recently, circulating donor-derive cell free DNA (dd-cfDNA) has gained growing attention in the field of solid organ transplantation. The aim of the study was to analyze circulating dd-cfDNA levels in graft rejection, ACR and AMR separately for each rejection type compared with non-rejection, and assessed the diagnostic potential of dd-cfDNA levels in predicting graft rejection after lung transplantation. Methods A systematic search for relevant articles was conducted on Medline, Web of Science, China National Knowledge Infrastructure (CNKI), and Wanfang databases without restriction of languages. The search date ended on June 1, 2023. STATA software was used to analyze the difference between graft rejection, ACR, AMR and stable controls, and evaluate the diagnostic performance of circulating dd-cfDNA in detecting graft rejection. Results The results indicated that circulating dd-cfDNA levels in graft rejection, ACR, and AMR were significantly higher than non-rejection (graft rejection: SMD=1.78, 95% CI: 1.31-2.25, I2 = 88.6%, P< 0.001; ACR: SMD=1.03, 95% CI: 0.47-1.59, I2 = 89.0%, P < 0.001; AMR: SMD= 1.78, 95% CI: 1.20-2.35, I2 = 89.8%, P < 0.001). Circulating dd-cfDNA levels distinguished graft rejection from non-rejection with a pooled sensitivity of 0.87 (95% CI: 0.80-0.92) and a pooled specificity of 0.82 (95% CI: 0.76-0.86). The corresponding SROC yield an AUROC of 0.90 (95% CI: 0.87-0.93). Conclusion Circulating dd-cfDNA could be used as a non-invasive biomarker to distinguish the patients with graft rejection from normal stable controls. Systematic Review Registration https://www.crd.york.ac.uk/prospero/, identifier CRD42023440467.


Introduction
Lung transplantation has become the live-saving treatment for patients with end-stage pulmonary disease.Dramatic advancements have been made in surgery, medications, and postoperative management, and increased the overall survival rate to 85% and 59% at the end of the first and five years of transplantation, respectively (1).However, graft rejection, including acute cellular rejection (ACR) and antibody-mediated rejection (AMR), remains a substantial cause of morbidity and mortality after lung transplantation.ACR, as one of the most common complications, is detected in 27% of lung transplant recipients (LTRs) within the first year, and significantly related to chronic allograft dysfunction (2,3).Moreover, due to the absence of the standardized diagnostic and treatment approaches, it is estimated that the incidence rate of AMR ranges from 4% to >50% in LTRs (4).AMR, characterized by donor-specific antibodies (DSA), innate immune infiltration, and evidence of complement activation, has been consistently identified as a contributor to morbidity, chronic lung allograft dysfunction (CLAD), and graft failure (5,6).To date, invasive biopsy is the golden standard for diagnosis of graft rejection, performed either when rejection is suspected or for surveillance (7).However, invasive graft biopsy is difficult to practice for most LTRs in clinical settings.Apart from the possibility of sampling error, high economic cost, and physician dependence, a potential risk of complications has prevented invasive biopsy from becoming a routine screening approach (8,9).Thus, a non-invasive, relatively objective, and sensitive biomarker is urgently needed for rejection diagnosis after transplantation.
Donor-derive cell free DNA (dd-cfDNA) refers to the free DNA released by damaged cells of graft tissue after transplantation.Recently, circulating dd-cfDNA has gained growing attention in the field of solid organ transplantation.The dd-cfDNA examination takes advantage of sensitivity and specificity of genome sequencing and the wide genomic difference between donors and recipients to quantify the circulating dd-cfDNA (10).The assay methods of dd-cfDNA includes quantitative polymerase chain reaction, digital droplet PCR, shotgun sequencing, and HLA alleles-mismatches.Xiao H, et al. summarized that dd-cfDNA can be a helpful marker for the diagnosis of AMR in transplantation recipients suspected of renal dysfunction (11).Wijtvliet VPWM, et al. showed that dd-cfDNA may be a useful marker for AMR diagnosis, but probably not for ACR (12).Mounting evidence has suggested that dd-cfDNA was a potential non-invasive marker for allograft injury and rejection and could be used for graft rejection surveillance by the International Society for Heart and Lung Transplantation (ISHLT) (13-15).To the best our knowledge, no relevant systematic review and meta-analysis evaluated the circulating dd-cfDNA levels in graft rejection, ACR, and AMR, as well as the diagnostic value in graft rejection after lung transplantation.
To date, the clinical detection of circulating dd-cfDNA levels in lung transplantation has been documented in more than 10 studies, focusing on ACR, AMR or both.However, the data on the circulating dd-cfDNA levels in graft rejection after lung transplantation are conflicting.We therefore conducted a systematic review and metaanalysis to analyze circulating dd-cfDNA levels in graft rejection, ACR and AMR separately for each rejection type compared with non-rejection, and assessed the diagnostic potential of dd-cfDNA in predicting graft rejection after lung transplantation.

Materials and methods
This systematic review and meta-analysis was performed following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA 2020) guidelines.Quality assessment of Diagnostic Accuracy Studies-2 (QUADAS-2) tool and Newcastle Ottawa scale (NOS) tool were used to assess the quality of included studies, respectively.The protocol was registered in the International Prospective Register of Systematic Reviews (PROSPERO; CRD 42023440467).

Literature search
A systematic search for relevant articles was conducted on Medline, Web of Science, China National Knowledge Infrastructure (CNKI), and Wanfang databases without restriction of languages.No time restrictions were applied regarding the start date of publication, and the search ended on June 1, 2023.The search strategy was conducted using the following keywords: ("donor" OR "tissue donors") AND ("cell-free DNA" OR "cell-free nucleic acids") AND ("rejection") AND ("lung transplantation" OR "lung grafting" OR "pulmonary transplant" OR "pulmonary allograft") OR "lung transplant" OR "lung allograft").Two reviewers (LB AND LYH) screened the titles and abstracts of articles in the initial search to identify the relevant publications.Then, full texts were obtained and reviewed in detail.

Inclusion and exclusion criteria
Articles were considered eligible if they met all of the following inclusion and exclusion criteria.The inclusion criteria included: (1) case-control studies or cohort studies; (2) studies conducted on lung transplantation, including graft rejection or ACR or AMR cases; (3) studies assessing the circulating dd-cfDNA levels, or studies including sufficient data to calculate the number of true positive (TP), true negative (TN), false positive (FP), and false negative (FN) values.Moreover, studies that met the following criteria were excluded: (1) duplicate publications; (2) review articles, editorials, comments, and conference proceedings; (3) animal experiments; (4) essential data for pooled analysis or quality assessment were not retrievable.
Two reviewers (LB and LYH) independently selected and crosschecked the articles.Any discrepancies were resolved by discussion.

Data extraction and quality assessment
The following data were extracted from the selected articles: first author's name, publication year, country, study design, sample number, rejection type, control type, assay method, dd-cfDNA levels, thresholds, the area under the receiver operating characteristic curve (AUROC), and its sensitivity, specificity, TP, FP, FN, and TN.
The case-control study or cohort study was evaluated according to the NOS tool containing eight items that could be classified into three main categories (selection, comparability, and exposure/ outcome evaluation) and assigned on a score of 0-9.In the NOS assessment, scores of 7-9, 4-6, and 0-3 were considered "high quality", "moderate quality", and "low quality", respectively.The quality of diagnostic studies was evaluated using QUADAS-2 tool, which consisted of seven items covering case selection, index test, golden standard, and flow and timing.The total score ≥4 (full score= 7) indicated that the quality of the study was high.

Statistical analysis
STATA software (version 15.0, Stata Corp) was used to analyze the original data.We performed the meta-analyses of circulating dd-cfDNA levels in patients with graft rejection, ACR, and AMR in comparison with normal stable cases.Due to different assay methods of dd-cfDNA levels, standard mean difference (SMD) along with 95% confidence interval (CI) was used to compare the difference.If the original articles only provide median and interquartile ranges (IQR), the methods suggested by Luo DH, et al. and Wan X, et al. were used to convert them into means and standard deviations (SDs) (16,17).Given the variability in the sample characteristics, diagnostic criteria, and dd-cfDNA assay methods between studies, the chi-square Q test and I 2 statistics were used to determine the heterogeneity of eligible studies.If I 2 > 50% or P < 0.05, we considered heterogeneity to be significant and adopted a random-effects model, otherwise, a fixed-effects model was adopted.Publication bias was assessed through funnel plot asymmetry and Begg's test among the included studies.Sensitivity analysis was performed to assess the impact of single study on the overall result by serial removal of a specific study.
STATA software was used to estimate the diagnostic test accuracy of circulating dd-cfDNA in graft rejection by calculating the pooled ROC area of all studies.We performed pooled analysis of sensitivity (SEN), specificity (SPE), positive likelihood ratio (PLR), negative likelihood ratio (NLR), diagnostic odds ratio (DOR), and area under the summary receiver operating characteristic curve (AUSROC), along with the corresponding 95% CI.Deek's funnel plot was used to assess publication bias among studies.Fagan nomogram can show the correlation between prior probability, likelihood ratio and posttest probability.The more the difference between pre-test and posttest probability, the more important the biomarker is, which can provide a reference for clinical application evaluation (18).A P value < 0.05 was considered statistically significant.
Begg's test was used to explore the publication bias of the included studies in our meta-analysis, and a visual funnel plot was plotted to identify the publication bias (Figure S1).The funnel plots showed symmetrical features which indicated the absence of publication bias, identified by Begg's tests (graft rejection: P=0.533; ACR: P=0.851; AMR: P=0.764).Moreover, sensitivity analysis which removed each study in turn showed that the overall result was stable and no study could influence the final result in this metaanalysis (Figure S2).

Fagan nomogram
To further explore the potential value of circulating dd-cfDNA level for graft rejection in lung transplantation, we built a Fagan nomogram.ISHLT reported that 28% of LTRs experience at least one episode of treated ACR within the first year (37).Assuming a 28% pre-test probability of developing graft rejection, we used circulating dd-cfDNA levels to predict a 68% post-probability of developing graft rejection and a 7% risk of not developing graft rejection (Figure 5).The Fagan nomogram showed that the postprobability was increased by 40% in patients with positive pre-test and decreased by 21% in patients with negative pre-test.Our findings suggest that circulating dd-cfDNA was a clinically beneficial diagnostic marker in predicting graft rejection after lung transplantation.

Discussion
Rejection remains a significant problem following lung transplantation.In spite of advances in surgical techniques and medication, long term outcomes after lung transplantation remains worse compared with other solid organ transplantation (38).The clinical data indicated that 28% of adult recipients experienced at least one episode of acute rejection during the first year after lung transplantation (37).Currently, the biopsy with detailed pathology is regarded as the gold standard for making diagnosis based on histologic lesions associated with acute rejection.However, frequent The flow chart of literature search.
coexistence of rejection and infection, disagreement between different pathologists, and invasive complication limited the accuracy, timeliness, and repeatability.The value of dd-cfDNA as a non-invasive biomarker for lung transplantation has been investigated in some studies.Moreover, most studies were published in the last three years, underlining the emerging interest in circulating dd-cfDNA detection as a monitor of graft rejection.However, does dd-cfDNA play a role in the current clinical paradigm for the monitoring of underlying graft rejection?In this review and meta-analysis, we found that circulating dd-cfDNA levels in lung transplantation patients with graft rejection, ACR, and AMR were higher than patients without rejection or stable patients, and circulating dd-cfDNA was a noninvasive potential marker in diagnosis of graft rejection after lung transplantatiopn.
The dd-cfDNA, detected in the blood of transplant recipients, provided huge potential as a sensitive, non-invasive, cost-effective biomarker for long-term monitoring of graft health, based on the fact that organ transplantations are also genome transplants.The plasma dd-cfDNA levels were generally low in stable transplant recipients, and dd-cfDNA levels increased due to tissue injury after solid organ transplantation (39).Snyder TM, et al. found that dd-cfDNA levels were less than 1% at stable patients, and increased to 3-4% during rejection episode after heart transplantation (40).In    However, not all injurious processes may release substantial cell lysis or cause the dysregulation leading to an elevation in dd-cfDNA (44).Khush KK, et al. reported that dd-cfDNA level with AMR was no difference with stable patients after lung transplantation (24).value of dd-cfDNA of included studies, and found that dd-cfDNA fraction levels were significantly higher in graft rejection, not only in ACR, but also in AMR in this meta-analysis.The release of cell-free DNA into the blood is influenced by many physiological and pathological factors, such as physical and psychological stress, inflammation, and autoimmunity (46).Remarkably, dd-cfDNA was elevated in the presence of infection, in keeping with the notion that dd-cfDNA could predict graft damage.Compared with asymptomatic respiratory tract infections, symptomatic respiratory tract infections were also more likely to have high %ddcfDNA (36).A multicenter prospective cohort study of Genomic Research Alliance for Transplantation found that dd-cfDNA is a marker of pathogen associated allograft injury and may detect subclinical injury (26).In addition, very few patients underwent a prolonged decline in dd-cfDNA levels within the first 10 days after kidney transplantation, graft rejection might be suspected incorrectly due to elevated dd-cfDNA levels (47).Hence, we performed a synthesis of the currently available knowledge of circulating dd-cfDNA in relation to rejection diagnosis.The results indicated that circulating dd-cfDNA has a high accuracy in the diagnosis of graft rejection with pooled sensitivity and pooled specificity of 0.87 and 0.82, respectively.The PLR and NLR were 4.7 and 0.16, which indicated that probability of positive dd-cfDNA in graft rejection was 4.7 times higher than normal stable patients, and normal stable patients with negative results have about one-sixth chance of developing graft rejection.The DOR value was 29, indicating that circulating dd-cfDNA could effectively distinguish graft rejection from normal stable patients.Based on the Fagan's nomogram, if the pre-test probability was 28%, the post-test probability of dd-cfDNA's positive diagnostic result for graft rejection was 68%.In addition, the SROC curve showed that corresponding AUC was 0.90 (0.87-0.93), which was much higher than the common diagnostic criteria (>0.8).Above results indicated that circulating dd-cfDNA has a promising diagnostic performance in predicting graft rejection.However, the heterogeneity in these studies was considerable.The clinical characteristics of patients, methodological differences, cutoff values, and sample preparation protocols may lead to hidden bias in the pooled estimation.In view of no sufficient studies to perform subgroup analysis, in depth analysis of heterogeneity was not possible.With the advances in laboratory technology and knowledge, the subgroup analysis of circulating dd-cfDNA levels is needed in the future.
The present study has some limitations.First, the sample size was relatively small, and there are not many studies related to circulating dd-cfDNA levels in graft rejection in recent years.Second, the diagnostic criteria were not standardized in each study.Third, it was difficult to obtain the raw data for some included studies, which restricts us to investigate the diagnostic value of dd-cfDNA absolute quantification.Forth, owing to the different measurement methods, the circulating dd-cfDNA cut-off values used among the included studies were not completely consistent, and different cut-off values would affect the accuracy of the combined results to a certain extent.Most importantly, sources of heterogeneity in the results should still be considered carefully.Nonetheless, our study clearly showed that circulating dd-cfDNA was a promising marker in diagnosis of graft rejection, despite the fact that the meta-analysis has the limitations mentioned.

Conclusion
Generally, our meta-analysis suggested that circulating dd-cfDNA was higher in patients with graft rejection, ACR, and AMR.Circulating dd-cfDNA could be used as a non-invasive biomarker to distinguish the patients with graft rejection from normal stable controls.Further evidence is required to explore the diagnostic accuracy of circulating dd-cfDNA in each type of graft rejection, thus paving the way for clinical application of circulating dd-cfDNA detection.

FIGURE 1
FIGURE 1 lung transplantation, similar results were found that dd-cfDNA had a strong concordance with clinical indicators of rejection (33).In liver transplantation, dd-cell free DNA was elevated significantly after engrafting, followed by a steady decrease to 10% of total cfDNA after 1 week(41).Moreover, dd-cfDNA accounts for <0.5% of total plasma cell free DNA in renal transplantation recipients without allograft injury, and normal dd-cfDNA could reduce the probability of detecting rejection (42).Early diagnosis of subclinical rejection might improve the clinical prognosis.Agbor-Rnoh S, et al.
reported allograft injury detected by dd-cfDNA preceded clinical AMR diagnosis by a median of 2.8 months (31).The result from a total of 1,092 adult kidney transplant recipients across 7 transplant centers showed that the elevation of dd-cfDNA was significantly correlated with clinical and subclinical graft rejection(43).

2 3 Forest
FIGURE 2 Quality assessment of the included literature.(A) QUADAS-2 summary plot of bias assessment.(B) QUADAS-2 bar plot of the individual risk of bias domains and applicability concerns.

4
FIGURE 4 Performance of circulating dd-cfDNA in graft rejection detection.(A) forest plots of pooled diagnostic performance for sensitivity and specificity of circulating dd-cfDNA in graft rejection detection; (B) summary receiver operating characteristic (SROC) plot of circulating dd-cfDNA in graft rejection detection; (C) Deeks' funnel plot for publication bias.

FIGURE 5 Fagan
FIGURE 5Fagan nomogram of circulating dd-cfDNA for the diagnosis of graft rejection.

TABLE 1
Characteristics of included studies in graft rejection, ACR, and AMR.

TABLE 2
Characteristics of included studies for diagnostic performance.

TABLE 3
Quality scores of cohort studies using Newcastle-Ottawa Scale.
Q1. Representativeness of the exposed individuals; Q2.Selection of the non-exposed cohort; Q3.Ascertainment of exposure; Q4.Demonstration that outcome of interest was not present at start of study; Q5.Comparability on the basis of the design or analysis; Q6.Assessment of outcome; Q7.Adequate follow-up duration; Q8.Adequacy of follow up of cohorts.