Lung transplantation in 2025: a narrative review of progress, challenges, and the road ahead Authors: Xiaohan Jin, Jiaqi Li, Yaling Liu, Audrey Yaulap Li, Jier Ma, Linxi Liu, Zhiqiang Deng, Zengwei Yu, Xiangyun Zheng, Haoji Yan, Qiang Pu, Lunxu Liu, Masaaki Sato, Dong Tian Published: 2026 ABSTRACT Background and ObjectiveLung transplantation (LTx) remains the only definitive treatment for end-stage lung disease. Throughout 2025, the global transplant community has made notable progress in addressing persistent challenges in the field, ranging from policy-level improvements in graft allocation to deeper biological insights into post-transplant complications. This review aims to synthesize the pivotal literature published in 2025 across four core areas: optimizing allocation, advancing perioperative care, defining the mechanisms of graft injury, and exploring translational frontiers.MethodsWe conducted a structured literature search focused primarily on studies published in 2025 that highlight major advancements in LTx. In addition, the search was supplemented with key society guidelines, consensus statements, and relevant early 2026 publications. Selected articles were critically analyzed and categorized into clinical advancements, and basic/translational findings.Key Content and FindingsIn clinical research, the refinement of allocation strategies and the expansion of donor criteria, have effectively broadened access. Perioperative management has evolved with the integration of artificial intelligence to predict complications such as primary graft dysfunction (PGD) and chronic lung allograft dysfunction (CLAD). In basic and translational research, studies have dissected the molecular basis of complications. Additionally, new insights into the evolution of drug-resistant pathogens and macrophage plasticity have deepened our understanding of infection and rejection.ConclusionsIn 2025, significant progress has been achieved in the field of LTx research. Strategies for clinical donor allocation and perioperative management, along with the understanding of molecular mechanisms underlying complications, have also been greatly improved. These advancements are essential to further optimize outcomes and address the complex challenges in LTx. Introduction Lung transplantation (LTx) has been regarded as the only effective therapy for patients with advanced pulmonary disease (1,2). Despite advances in surgical techniques and perioperative management, the field is still facing several real-world challenges, including donor-organ shortage, increasing recipient complexity, primary graft dysfunction (PGD), chronic lung allograft dysfunction (CLAD), infection, rejection, and limited long-term graft survival. Recently, LTx has been increasingly challenged by the growing disparity between rising recipient acuity and limited donor-organ availability. A substantial proportion of recipients now present with advanced disease, substantial comorbidity burdens, frailty, pulmonary hypertension, or dependence on extracorporeal support (3,4). To reduce waitlist mortality and expand transplant access, programs have increasingly adopted broader geographic sharing, extended-criteria donor lungs, donation after circulatory death (DCD), normothermic regional perfusion (NRP), ex vivo lung perfusion (EVLP), and other donor-optimization strategies. These developments have intensified debates over organ allocation, particularly regarding how to balance medical urgency, equitable access, donor-organ utility, and acceptable post-transplant outcomes. Meanwhile, LTx research has expanded from traditional clinical outcome studies toward more mechanistic and translational approaches. Contemporary research is applying risk prediction models, artificial intelligence, donor and recipient phenotyping, immune profiling, microbiological surveillance, and multi-omics to better understand complications such as PGD, CLAD, infection, and other postoperative organ injuries. The literature published in 2025 highlights ongoing efforts by the global transplant community to tackle these multifaceted challenges, organ allocation strategies, perioperative management, and deeper molecular investigations into allograft dysfunction. Therefore, this narrative review aims to summarize the pivotal clinical, basic and translational studies in 2025. We focus on organ allocation and donor management, surgical and perioperative support, mechanisms of postoperative complications, and translational frontiers (Figure 1) (5). We present this article in accordance with the Narrative Review reporting checklist (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2026-0554/rc). Central infographic summarizing the key advancements in lung transplantation in 2025. Created with BioGDP.com (5). ALAD, acute lung allograft dysfunction; BLAD, baseline lung allograft dysfunction; CAS, Composite Allocation Score; CLAD, chronic lung allograft dysfunction; DCD, donation after circulatory death; ECMO, extracorporeal membrane oxygenation; EVLP, ex vivo lung perfusion; NRP, normothermic regional perfusion; PGD, primary graft dysfunction; TA-NRP, thoracoabdominal normothermic regional perfusion. Methods This narrative review is based on a structured literature search designed to highlight key breakthroughs in LTx throughout 2025. We used PubMed as our primary database because of its comprehensive coverage of clinically related transplantation literature and its practical suitability for a time-specific annual review. The detailed search terms, timeframe, and specific inclusion/exclusion criteria are summarized in Table 1. Two independent reviewers (J.L. and Y.L.) screened the articles, prioritizing clinical trials and translational studies with high relevance to allograft outcomes. Conflicts were resolved by a third author (X.J.). In addition to the primary database search, we complemented the database search by manually incorporating key society guidelines, consensus statements and foundational studies. Moreover, early 2026 publications reporting 2025 findings were included. Finally, a snowballing approach was used to identify additional relevant references. The detailed literature screening and selection process is visually summarized in a flowchart (Figure 2). Table 1The search strategy summaryItemsSpecificationDate of searchJanuary 19, 2026DatabasePubMedSearch terms used(“lung transplant” [MeSH Terms] OR “lung transplantation” [MeSH Terms] OR “lung transplant” [Title/Abstract] OR “lung transplantation” [Title/Abstract])TimeframeJanuary 1, 2025 to December 31, 2025Inclusion and exclusion criteriaInclusion criteria: clinical studies, translational research, and preclinical investigations (including animal models and mechanistic studies) regarding lung transplantation, focusing on key topics such as immunomodulation, ischemia-reperfusion injury, and donor lung assessment; English-language articles. Exclusion criteria: studies focusing on non-pulmonary solid organ transplantation, basic research lacking relevance to allograft outcomes or lung physiology, case reports, or reviews/editorials without original dataSelection processTwo authors (J.L. and Y.L.) independently conducted the selection. Conflicts were resolved by a third author (X.J.)Additional considerationsFoundational historical studies and highly relevant early-2026 articles were manually added for comprehensive coverage and comparison to highlight 2025 advancements. A snowballing method was applied to identify additional studies The search strategy summary Flowchart of the literature search and selection process. Clinical research progress Optimization of organ allocation The Composite Allocation Score (CAS) represents a major policy shift in the United States intended to improve allocation equity and system transparency (6,7). Since transitioning from lung allocation score (LAS) in March 2023, early registry data indicate that CAS has largely achieved its primary goals of reducing waitlist mortality and increasing overall transplantation rates without compromising short-term post-transplant survival (6,8). Moreover, the CAS era is characterized by broader geographic sharing, evidenced by longer transport distances and ischemic times, alongside a concomitant increase in the utilization of DCD and NRP allografts (8). While this policy has successfully narrowed racial gaps, particularly improving access for Hispanic candidates (9), structural disadvantages persist for candidates with short stature (<162 cm), female gender, and blood type O, who continue to face lower transplantation probabilities or higher waitlist removal risks (8,10-12). In the future, continuous algorithmic refinement and targeted policy evaluations are essential to minimize these persistent gaps. The CAS system heavily weighs medical urgency, leading to expedited access for critically ill candidates. For candidates requiring extracorporeal membrane oxygenation (ECMO) bridge-to-transplant, early evidence suggested that CAS has effectively reduced waitlist duration and expedited organ access (10). However, the long-term implications of this urgency-driven prioritization remain uncertain. A systematic review and meta-analysis of urgent LTx indicated that such procedures are historically associated with inferior long-term survival compared to standard operations. Notably, studies included in this meta-analysis were conducted under the LAS era or in non-CAS allocation systems (13). Moreover, worse post-transplant outcomes in urgent recipients may reflect substantial baseline risk, including greater pre-transplant physiologic instability, ECMO dependence, infectious complications, and center-level variability, rather than the allocation strategy alone. Therefore, whether the current CAS framework can mitigate the historically observed survival disadvantage associated with urgent LTx remains to be determined, underscoring the need for CAS-specific longitudinal evaluations of both waitlist and post-transplant outcomes. Donor shortage Donor shortage remains the principal barrier to the advancement of LTx, necessitating novel strategies to expand donor eligibility criteria and minimize viable graft discard through optimized preservation and precise selection (14). Recent evidence supports that carefully selected lungs from donors aged 80 to 94 years yielded acceptable outcomes, with 1- and 3-year survival rates of approximately 84% and 70%, respectively (15). Moreover, Nakajima and colleagues examined adult-to-adult living-donor lobar LTx and found that despite recipients having more severe preoperative conditions, early postoperative outcomes and long-term survival were comparable to conventional brain-dead donor transplants (16). To reduce the discard of viable lungs, objective assessment tools are replacing subjective judgment (2). The LUNDON scoring system has emerged as a promising standard for donor evaluation (17). A subsequent multicenter study confirmed the applicability of this tool across diverse populations, demonstrating its utility in standardizing donor lung evaluation and improving both transplant efficiency and success rates (18). Furthermore, a visual scoring system based on chest computed tomography (CT) was developed and achieved a lobar utilization rate of 79.22% among brain-dead donors, thereby providing a quantifiable radiological basis for precise donor lung selection (19). Preservation and perfusion technologies Innovations in preservation have proven critical for graft utility (20). A multicenter study demonstrated that donor lung preservation at 10 ℃ for 24 hours is non-inferior to traditional static cold storage (0–4 ℃) (21). Additionally, Cenik and colleagues compared the dynamic temperature profiles of lungs during static ice storage versus controlled hypothermic storage throughout the transplantation process, revealing no statistically significant differences in postoperative prognostic indicators between the two groups (22). EVLP has emerged as a representative strategy for donor lung assessment and functional repair (23,24). In recent years, EVLP has evolved into portable platforms such as the Organ Care System (OCS), which enables normothermic blood-based perfusion during transport (25). Notably, it is confirmed that extended criteria donor recipients managed with the OCS Lung system achieved favorable long-term outcomes (26). The 2025 North American Expert Consensus recommends EVLP for clinical evaluation of lungs with questionable quality or marginal status, noting that grafts passing EVLP assessment demonstrate quality comparable to standard criteria lungs (24,27). Despite these encouraging developments, a large-scale analysis of United States registry data revealed that while EVLP facilitates transplantation, recipients in the EVLP cohort experienced higher 30-day mortality, prolonged postoperative length of stay, and increased requirements for mechanical ventilation at 72 hours compared to the non-EVLP cohort (28). Furthermore, several studies have suggested that LTx using EVLP-treated DCD donors may be associated with increased mortality and complication risks (3,29). However, a critical evaluation of these registry data suggests that this increased mortality is unlikely to be primarily driven by the EVLP technology. In real-world clinical practice, EVLP is preferentially applied to marginal grafts and allocated to higher-acuity recipients, introducing substantial selection bias into unadjusted comparisons. Consistent with this interpretation, an adjusted analysis revealed that when rigorously controlling for confounding variables such as recipient acuity and ischemic duration, EVLP demonstrates no independent association with severe PGD or inferior survival (30). Therefore, these adverse trends are likely to stem from higher recipient acuity and the prolonged ischemic time, rather than the perfusion technology itself. Nevertheless, residual confounding in retrospective cohorts cannot be fully excluded, and prospective data are warranted to definitively clarify the independent impact of EVLP on post-transplant outcomes. Beyond EVLP, significant strides have been made in in situ donor lung repair. Thoracoabdominal NRP (TA-NRP) has emerged as a promising strategy that utilizes an extracorporeal circuit to restore blood flow to thoracic and abdominal organs immediately following circulatory death (31). Vandendriessche and colleagues compared two TA-NRP initiation protocols: full TA-NRP initiated via central cannulation 15 minutes after circulatory arrest versus initial isolated abdominal NRP with cardiac drainage, converting to full TA-NRP after 75 minutes. Both protocols maintained satisfactory lung function, suggesting potential for expanding the donor pool through this technology (32). ECMO ECMO serves as a critical modality for bridging end-stage lung disease patients to transplantation and providing perioperative support (33). Recent evidence has indicated that veno-arterial (VA) ECMO achieves superior pulmonary circulatory unloading compared with veno-venous (VV) ECMO, leading to more stable intraoperative hemodynamics and improved early outcomes (34). Furthermore, regarding cannulation strategies, a retrospective study suggested that peripheral cannulation in VA-ECMO was associated with worse prognosis than central cannulation (35). Beyond the optimization of ECMO configuration, management strategies during the bridging period are also of importance. Conventional cannulated ECMO necessitates invasive mechanical ventilation and sedation, increasing the risk of ventilator-associated pneumonia, ventilator-induced lung injury, and diaphragmatic dysfunction (36). To address this limitation, Zhou and colleagues confirmed non-intubated ECMO strategies as superior alternatives, with reduced stroke risk, shorter hospital stays, and better short-term survival (37). Meanwhile, for patients requiring VV-ECMO, dynamic monitoring of lung compliance has emerged as a key prognostic tool, with higher compliance correlating with reduced postoperative complications and better survival (38). Consistent with these findings, the International Society of Heart and Lung Transplantation (ISHLT) consensus statement advocates for early, indication-driven extracorporeal life support, preferring VV-ECMO for isolated respiratory failure while reserving VA-ECMO for cardiac dysfunction (39). This consensus further underscores the importance of individualized strategies, including cannulation choice and early mobilization, to optimize outcomes in this high-risk population. PGD PGD remains a leading cause of early morbidity and mortality following LTx (40). A recent meta-analysis comprising nearly 60,000 patients highlighted a complex interplay of risk factors, identifying donor age, smoking history, recipient body mass index (BMI), and pulmonary hypertension as significant predictors of PGD (41). Analysis from a national donor management registry revealed that specific critical care endpoints, such as donor serum pH and cocaine use, are predictive of PGD grade 3, whereas adherence to predefined donor management goals regarding the PaO2/FiO2 (P/F) ratio significantly reduces this risk (42). Furthermore, adjusted donor lung weight, serving for extravascular lung water and edema, has been identified as an independent risk factor, with high-weight lungs correlating with higher PGD incidence and prolonged intensive care unit (ICU) stays (43). The recipient’s preoperative cardiac function exerts a profound influence on graft performance, often beyond standard metrics. While left ventricular diastolic dysfunction (LVDD) has been independently associated with a twofold increase in PGD risk, recent evidence suggests that traditional echocardiographic markers may lack sensitivity (44). A study focusing on left atrial mechanics demonstrated that reduced left atrial compliance, assessed via left atrial strain, predicts PGD even in patients with normal systolic and diastolic function (45). Additionally, elevated recipient central venous pressure continues to be a notable risk factor for severe PGD (42). Given the complex risk factors for PGD and the limited predictive power of traditional clinical measures, artificial intelligence has shown promise for better risk assessment and early detection. Fessler and colleagues reported an intraoperative XGBoost model utilizing variables such as blood lactate and total lung capacity mismatch, which achieved high predictive accuracy for PGD (46). Similarly, random forest models have demonstrated optimal performance not only in predicting PGD but also in stratifying risks for postoperative ECMO and ventilator duration (47). Moreover, the fusion of donor lung CT scans with clinical data using deep learning approaches has shown superior predictive capability compared to clinical data alone (48). Despite these encouraging advances, the widespread clinical integration of these models remains hindered by a lack of robust external validation and unproven generalizability across diverse institutional cohorts. Graft dysfunction phenotypes CLAD CLAD represents the primary limitation to long-term survival following LTx, manifesting chiefly as bronchiolitis obliterans syndrome (BOS) or restrictive allograft syndrome (RAS). While spirometric monitoring remains the standard, recent data indicate that even a single decline in forced expiratory volume in the first second (FEV1) of ≥10% carries a significant positive predictive value for incipient CLAD, suggesting that current diagnostic thresholds might benefit from refinement (49). Additionally, automated CT image processing has emerged as a powerful tool, with specific biomarkers such as air trapping and airway wall thickening demonstrating high accuracy in diagnosing CLAD and differentiating between BOS and RAS phenotypes (50). The pathogenesis of CLAD is driven by a complex interplay of immunological and infectious insults. A nested case-control study identified that RAS is strongly associated with persistent donor-specific antibodies (DSA), particularly against HLA-DQ and HLA-DR loci, as well as a history of severe infections including Pseudomonas, Aspergillus, and cytomegalovirus (CMV) (51). This is corroborated by a prospective multicenter study where late onset of DSA and CMV infection were independent predictors for probable CLAD, a precursor state that significantly increases the risk of graft loss (52). Furthermore, the role of respiratory viruses has been highlighted by findings that even asymptomatic human coronavirus (HCoV) infections are linked to a 2.5-fold increased risk of subsequent CLAD (53). Moreover, the overall lung microbiome burden, specifically high bacterial density in bronchoalveolar lavage (BAL) fluid, has been correlated with CLAD development, suggesting that microbial dysbiosis acts as a driver of allograft dysfunction (54). In terms of management, strategies are evolving from broad immunosuppression to targeted prevention and therapy. A systematic review supports the prophylactic use of azithromycin, which significantly reduces the risk of CLAD onset and improves CLAD-free survival (55). Regarding maintenance immunosuppression, registry analysis indicates that tacrolimus-based regimens are superior to cyclosporine in preventing CLAD and extending graft survival (56). For patients with refractory rejection, extracorporeal photopheresis has been validated in a randomized controlled trial as an effective add-on therapy, significantly reducing the rate of CLAD and acute cellular rejection (57). However, for those with advanced, end-stage CLAD, therapeutic options are limited. While retransplantation can offer survival benefits in carefully selected patients, it is associated with high post-operative mortality, emphasizing the need for early referral and optimal candidate selection (58). Acute lung allograft dysfunction (ALAD) and baseline lung allograft dysfunction (BLAD) While CLAD remains the primary barrier to long-term survival, the diagnostic landscape has evolved to distinguish between acute reversible events, termed ALAD, and the fundamental failure to achieve peak lung function, known as BLAD. According to the framework proposed by Verleden and Bos, ALAD represents an acute, potentially reversible ≥10% decline in FEV1, typically driven by infection or rejection, whereas BLAD, defined in bilateral recipients as the failure to achieve 80% of predicted FEV1 or forced vital capacity (FVC), reflects a permanent limitation on allograft capacity rather than a secondary decline (59). Current research identifies BLAD as a distinct, high-risk phenotype affecting up to 60% of recipients and serving as an independent predictor of mortality, particularly in single-lung transplant cohorts where functional recovery is rare (60,61). Pathologically, BLAD appears to be driven by unrecognized, subclinical alloimmune injury. It is strongly associated with the development of DSA even in the absence of clinical antibody-mediated rejection (AMR) (62), displays molecular gene expression profiles of T-cell mediated rejection often missed by standard histology (63), and correlates with high HLA eplet mismatch burdens compounded by insufficient immunosuppression (64). Beyond immunological factors, anatomical constraints, such as donor-recipient size mismatch, native lung hyperinflation in single-lung recipients (65), and persistent pleural effusions requiring thoracocentesis (66), play a critical role in limiting peak graft function. In the search for predictive tools, while traditional inflammatory markers like the neutrophil-to-lymphocyte ratio (NLR) have shown limited utility (67), emerging biomarkers such as donor-derived cell-free DNA (dd-cfDNA) have demonstrated significant potential (68). Elevated levels of dd-cfDNA can accurately discriminate acute rejection and functional decline from stable grafts (69), while early absolute elevations specifically correlate with the diagnosis of BLAD (70), supporting its utility in early risk stratification. Infection LTx recipients face substantially elevated infection risks attributable to high-intensity maintenance immunosuppression and the allograft’s direct environmental exposure (71). Viral pathogens remain a primary concern (72). CMV reactivation, often driven by sustained immunosuppression, is a known cause of end-organ disease. Recent applications of metagenomic next-generation sequencing have revealed a higher incidence of CMV viremia than previously reported; notably, this viremia is an independent risk factor for ICU mortality and is significantly associated with inferior 90-day overall survival (73). Similarly, recipients are highly vulnerable to respiratory syncytial virus. Although vaccination can elicit CD4+ T-cell responses in the majority of patients, humoral immunity remains impaired, with detectable antibody responses observed in only 40% of vaccinated recipients (74). Fungal and bacterial infections also present critical challenges. Invasive aspergillosis (IA) is most prevalent during the first three months post-transplant, with isolated tracheobronchial aspergillosis being the predominant phenotype. This early-onset IA correlates with poor prognosis, though BAL galactomannan testing has demonstrated high diagnostic utility (75). Regarding bacterial complications, nosocomial Acinetobacter baumannii infection significantly compromises cumulative survival from 30 days to 3 years post-transplant. Key independent risk factors include preoperative hospitalization/ICU stay, absence of prior anticoagulant use, and positive donor lung cultures (76). While therapeutic management is complex, eravacycline has shown efficacy against carbapenem-resistant A. baumannii (77). Finally, vigilant management of antifungal regimens is essential, with voriconazole-induced hepatotoxicity necessitating rigorous therapeutic drug monitoring to optimize risk stratification and safety (78). Acute kidney injury (AKI) AKI represents a common complication that profoundly impacts postoperative outcomes and survival (79). A recent retrospective study identified BMI, diabetes, plasma transfusion, operative duration, and postoperative ECMO as independent risk factors for AKI. Based on these variables, the investigators developed and calibrated a risk scoring model that effectively predicts AKI risk within 48 hours post-surgery, facilitating early identification and intervention for high-risk patients (80). Furthermore, AKI requiring dialysis is associated with elevated all-cause mortality risk, as well as increased 30-day and 1-year mortality rates compared with AKI not requiring renal replacement therapy (81). Pediatric LTx Pediatric LTx has evolved largely independently since its inception in 1987, yet for decades, management relied on extrapolations from adult guidelines (82). Addressing this gap, ISHLT released its first consensus statement on pediatric candidates in 2025. This landmark guideline shifts the paradigm from adult-centric protocols to a holistic pediatric framework, prioritizing age-specific physiological, psychological, and social contexts while advocating for early referral and the rigorous optimization of modifiable risk factors (83). Perioperative and postoperative management in this population presents unique challenges. Regarding intraoperative care, recent evidence suggests that strict restrictive fluid strategies may not necessarily reduce adverse outcomes; however, this lack of benefit does not endorse fluid overload, emphasizing the need for a balanced approach to volume administration (84). Postoperatively, complications often diverge from adult patterns. Notably, scoliosis occurs substantially more frequently in pediatric recipients, where asymmetric lung volumes, specifically unilateral atelectasis, can accelerate spinal deformity, necessitating routine spinal surveillance and timely intervention (85). Pediatric interstitial lung disease has been validated as an increasingly important indication, demonstrating meaningful survival benefits despite severe pre-transplant disease burdens (86). Furthermore, to address global donor shortages, a multicenter analysis has confirmed that living-donor lobar transplantation yields survival and functional outcomes comparable to deceased-donor grafts, underscoring its role as a critical alternative for expanding pediatric transplant access (87). Basic/translational research progress Infection LTx recipients face a significantly elevated risk of respiratory infections due to a confluence of factors, including long-term high-intensity immunosuppression, continuous exposure of the allograft to the external environment, impaired mucociliary clearance, and bronchial anastomotic ischemia (88). Beyond direct morbidity, infections serve as critical drivers of inflammation, remodeling the local immune microenvironment and influencing longitudinal graft outcomes, ranging from acute rejection to CLAD (89). Resistance mechanisms and evolutionary pressure Driven by chronic exposure to broad-spectrum antibiotics and the intensive care environment, pathogens within LTx recipients face immense selective pressure, leading to complex and dynamic resistance patterns (89). Recent genomic surveillance has identified carbapenem-resistant Klebsiella pneumoniae carrying the plasmid-mediated tmexCD-toprJ operon in transplant recipients, which coexists with other resistance genes to enhance multidrug resistance (MDR) and transmissibility (90). Similarly, under the specific therapeutic pressure of ceftazidime-avibactam, K. pneumoniae has been observed to rapidly acquire high-level resistance through parallel mechanisms of blaKPC gene point mutations and gene dosage expansion (91). In cases of chronic Acinetobacter baumannii infection, exposure to colistin has been shown to drive the co-evolution of resistance to both colistin and cefiderocol, resulting in the intra-host coexistence of subpopulations with varying susceptibility profiles (92). This suggests that long-term antimicrobial pressure drives pathogens toward a pan-drug-resistant phenotype while maintaining complex population structures that complicate therapeutic eradication. Pathogen-graft immune interaction Pseudomonas aeruginosa infection directly promotes AMR (93,94). Using a murine LTx model, investigators demonstrated that P. aeruginosa directly damages the allograft via type III secretion system-associated virulence factors, subsequently triggering B-cell activation and AMR. Crucially, the study established that blocking or deleting the chemokine receptor CXCR3 on B-cells significantly attenuates this process, offering a potential therapeutic target to decouple infection from alloimmune rejection (94). Host factors shaping heterogeneity The trajectory of infection evolution, susceptibility, and treatment response is largely shaped by the host’s unique physiological state and immune microenvironment, presenting highly individualized characteristics (74). A study tracking three recipients sharing the same donor revealed that an initially identical CMV strain evolved completely different resistance mutation profiles and drug susceptibility trajectories depending on the specific host’s drug exposure and immune background (95). Regarding susceptibility to opportunistic fungal infections, investigations into alveolar macrophage function have demonstrated that cellular ferroptosis significantly impairs the core ability of these cells to phagocytose and kill Aspergillus spores, with clinical data correlating high iron load and stress markers to infection risk independent of systemic immunosuppression (96). Moreover, the evaluation of vaccine-induced protection in this population requires a redefinition of relevant immune markers. A study on the respiratory syncytial virus prefusion F vaccine in LTx recipients observed that while all vaccinated individuals induced specific CD4+ T-cell responses, only approximately 40% generated detectable antibody responses (74). This discrepancy suggests that traditional immunogenicity evaluation systems centered on antibody titers may severely underestimate the true protective efficacy of vaccines in transplant recipients. Immune regulation and allograft rejection Cellular functional remodeling Macrophages, as the core of the innate system, have become a focal point for understanding plasticity in rejection (97,98). Recent epigenetic studies reveal that targeting the EZH2-STAT1 axis can reprogram macrophages, inhibiting their pro-inflammatory M1 polarization and attenuating fibrosis associated with obliterative bronchiolitis (99). Similarly, pharmacological intervention with decitabine has been shown to reduce acute rejection by inhibiting host macrophage recruitment and inducing a low-inflammatory phenotype (100). Spatial transcriptomics further identified the macrophage-to-myofibroblast transition as a critical driver of fibrosis (101). In the adaptive immune landscape, T-cell responses represent another critical front, balancing tolerance-inducing regulatory T cells (Tregs) against rejection-mediating effector T cells (102). Investigations into tissue-resident immunity have identified a population of amphiregulin-producing Foxp3+ cells enriched in bronchus-associated lymphoid tissue, which suppress chronic type 2 inflammation and pathological remodeling following viral infection (103). Mechanistically, local IL-33 signaling via ST2 activates this population to attenuate antibody-mediated and acute cellular rejection significantly, even in the absence of exogenous Treg supplementation (104). Conversely, T-cell receptor (TCR) sequencing revealed that the specific enrichment of donor-reactive tissue-resident memory T cells correlates with early rejection and lung function decline (105). Molecular networks and genetic basis Building upon cellular processes, molecular networks that orchestrate immune cell function represent a second layer of immune regulatory control. A large-scale analysis of LTx recipients revealed that approximately one-third of the population carries a functional C3 gene polymorphism (R102G). This variant impairs complement regulation and enhances activation, leading to the accumulation of memory B cells and antibody-secreting cells, which significantly correlates with the risk of CLAD (106). Epigenetic modifications further mediate the dynamic plasticity of immune regulation. In addition to the EZH2-STAT1 axis mentioned above (99), HMGA1 overexpression in obliterative bronchiolitis enhances chromatin accessibility at SP1 and DOCK4 loci, driving TGF-β1-mediated epithelial-mesenchymal transition (107). Single-cell transcriptomics further characterized CLAD-associated subsets, including fibrotic type II alveolar epithelial cells engaging in abnormal HGF-MET and PDGF crosstalk (108). Proteomic analysis indicated that eosinophilic inflammation shares key markers (galectin-3, WARS1) with acute cellular rejection, with 83.3% of affected recipients progressing to CLAD (109). Furthermore, alemtuzumab induction was found to persistently alter the proteomic profile of immunoglobulin subunits and complement factors at 12 months post-transplant (110). Monitoring and intervention Regarding monitoring, novel biomarkers for specific rejection phenotypes are emerging. A non-invasive platform based on circulating B-cell small extracellular vesicles (EVs) has demonstrated that activation markers (HLA-DR, CD38, BAFF-R) are significantly upregulated during AMR (111). For chronic rejection prediction, levels of liver kinase B1 in circulating small EVs were found to decrease significantly 6 months prior to the clinical diagnosis of CLAD in cystic fibrosis recipients (112). Additionally, elevated IL-17A levels in BAL fluid have been associated with acute cellular rejection, implicating the Th17 axis in specific rejection patterns (113). Regarding intervention, transbronchial delivery of AAV9-PD-L1 during static cold storage, combined with postoperative CTLA-4 Ig, significantly reduces acute rejection scores (114). Moreover, decitabine has been confirmed to promote Treg expansion and induce anti-inflammatory CD8+ T-cell phenotypes (100,115). Similarly, combining EVLP with Treg infusion effectively expands intragraft Tregs and mitigates acute rejection (116). Donor-targeted interventions, such as corticosteroid pretreatment, have also been shown to improve early graft immune status by targeting alveolar macrophages and downregulating T-cell organism chemokines (117). CLAD Early identification and risk stratification Since functional decline typically lags behind underlying molecular and immunological alterations, liquid biopsy techniques targeting EVs, serum proteomics, antibody profiles, or immune cell phenotypes have emerged as pivotal tools for capturing early warning signals and achieving pre-emptive risk stratification (118). In a cohort of 78 LTx recipients followed for three years, high levels of plasma HLA-G-bearing EVs (>21.3 ng/mL) at 12 months post-transplant were effectively predictive of favorable outcomes, significantly associating with a reduced risk of CLAD, BOS, and graft failure (119). Meanwhile, the presence of specific IgG autoantibodies combined with de novo DSA at 6 months could strongly predict CLAD onset (120). Proteomic screening further reveals that osteoprotegerin and inflammation-related proteins show differential expression prior to clinical diagnosis (121). Notably, a higher frequency of post-CMV expanded CD57 + KLRG1 + TEMRA cells correlates with reduced CLAD risk, suggesting a protective immune phenotype (122). Beyond blood-based markers, a 31-gene classifier derived from BAL transcriptomics effectively identifies genomic late acute cellular rejection as a precursor to CLAD (123). Additionally, morphometric analysis of surveillance transbronchial biopsies (TBBs) has demonstrated that alveolar septal widening serves as a preclinical risk factor for CLAD and is associated with poor survival (124). Concurrently, proteomic exploration of TBB samples revealed that patients who later developed CLAD exhibited early alterations in innate immunity and extracellular matrix remodeling proteins, highlighting the potential of tissue-based proteomic panels for early detection (125). Immune mechanisms and fibrosis evolution Multiplex imaging mass cytometry reveals significant enrichment of cytotoxic T cells, γδ T cells, plasma cells, and M2-like macrophages in CLAD tissues, with compositional divergence between BOS and RAS phenotypes (126). Mechanistically, complement regulation defects involving functional C3 polymorphisms promote memory B cell accumulation and DSA production, accelerating pathology (106). The link between PGD and CLAD might be mediated by downregulated LKB1 in plasma small EVs, which activates the LKB1-EMT axis to drive fibrosis (127). Furthermore, single-cell comparisons between CLAD-BOS and graft-versus-host disease-associated BOS identify a shared pathogenic network characterized by the expansion of CD8+ tissue-resident memory T cells (TRM) and dysregulated macrophage-stromal interactions (128). Intervention and translation Emerging therapeutic strategies target specific immune, fibrotic, and mechanotransduction pathways. Murine CLAD models reveal a critical therapeutic window for CTLA4-Ig, wherein immediate treatment markedly mitigates fibrosis and T-cell infiltration; conversely, delayed administration cannot rescue Treg depletion (129). Targeting tissue-resident cells, blockade of the NKG2D receptor, upregulated on cytotoxic CD8+ TRMs, successfully mitigates airway remodeling (130). Finally, mechanotransduction represents another critical pathway: inhibition of the mechanosensitive ion channel Piezo1 attenuates calcium influx and YAP nuclear translocation, thereby driving EMT and fibrosis (131). Ischemia-reperfusion injury (IRI) IRI serves as the fundamental pathological basis of PGD following LTx, driven by an interconnected cascade of oxidative stress, mitochondrial injury, innate immune activation, endothelial barrier compromise, and programmed cell death in its various forms (132). Innate immunity The amplification of inflammation during early reperfusion relies heavily on signal communication between innate immune cells and their interaction with the endothelium (132). Mechanistic studies have revealed that macrophage-derived EVs can exacerbate tissue injury by carrying EMAP-II to promote neutrophil extracellular trap formation, suggesting that targeting this signaling axis could interrupt the inflammatory cascade (133). Conversely, histidine-rich glycoprotein improves oxygenation by inhibiting HMGB1-mediated inflammation (134). Furthermore, the nuclear receptor NR4A1 was found to be significantly upregulated during early reperfusion, with donor-specific Nr4a1 deletion shown to improve graft function and attenuate endothelial leakage (135). Pharmacologically, it is reported that in a murine LTx model, the administration of a viral inhibitory peptide for TLR4 significantly attenuated inflammatory responses and improved early lung function (136). Additionally, the mitochondrial complex I in CD177+ neutrophils has been identified as a critical checkpoint for neutrophil activation, with metabolic targeting of this complex offering a potential strategy to mitigate pulmonary IRI (137). Programmed cell death Multiple forms of regulated cell death have been shown to be closely associated with the development of IRI. Ferroptosis is a form of cell death driven by iron-dependent lipid peroxidation (138). By performing single-cell RNA sequencing of lung tissues from patients after LTx, Kollareth’s team (139) identified differential expression of ferroptosis-related genes in post-transplant samples and further demonstrated that specialized pro-resolving mediators, such as lipoxin A4, exert protective effects against lung IRI by inhibiting ferroptosis. In animal studies, both porcine and murine models have confirmed that ferroptosis represents a key pathological component in the development of lung IRI (140,141). Meanwhile, cuproptosis involves impaired copper homeostasis that induces ferredoxin 1-mediated lipoylated proteins accumulation, mitochondrial respiratory dysfunction, and depletion of iron-sulfur cluster proteins, ultimately leading to proteotoxic stress. Experimental studies have demonstrated that cuproptosis is involved in lung IRI, and that protective effect of dexmedetomidine against lung IRI is partially mediated through inhibition of cuproptosis (142). Additionally, disulfidptosis depends on extensive depletion of reduced nicotinamide adenine dinucleotide phosphate under glucose-deficient conditions, leading to abnormal accumulation of disulfides (143). Using machine-learning models, SLC7A11 and LRPPRC genes were identified as specific predictive biomarkers for IRI, and olanzapine and vortioxetine were further identified as potential therapeutic agents (144). Mitochondrial homeostasis Mitochondrial dysfunction is widely regarded as an initiating event in IRI (137). In cold ischemia/reperfusion models, the administration of 4-octyl itaconate was shown to activate the Nrf2 pathway, reducing reactive oxygen species (ROS) levels and inhibiting apoptosis, thereby restoring cellular homeostasis (145). Similarly, the mitochondrial-derived peptide MOTS-c was found to mimic the effects of remote ischemic preconditioning to improve endothelial barrier function, a protective effect that was significantly attenuated under Nrf2-deficient conditions (146). Direct metabolic intervention has also proven effective. In post-transplant IRI models, supplementation with β-nicotinamide mononucleotide mitigated edema, inflammation, and oxidative stress, likely through the SIRT1/NF-κB pathway (147). Additionally, nebulized hucMSC-derived EVs overexpressing miR-146a are effectively encytosed by alveolar macrophages, yielding superior IRI attenuation compared to standard EVs (148). Multi-omics analysis With the advent of single-cell sequencing and multi-omics integration, IRI research is shifting from single-pathway analysis to holistic network characterization. Recent investigations utilizing single-cell and bioinformatic analyses have characterized cell states associated with reperfusion, revealing progressive metabolic inhibition in epithelial cells accompanied by signatures of cuproptosis activation (149). Machine learning approaches applied to EVLP and IRI datasets have further screened mitochondrial-related candidates, identifying the suppression of mitochondrial modules such as oxidative phosphorylation assembly and mtRNA metabolism as key features of injury (150). ScRNA-seq has identified lactylation-associated molecular subtypes of IRI (151) and highlighted alveolar macrophage phenotypic remodeling as a critical driver in large animal models (140). Furthermore, network pharmacology and in vivo validation demonstrate that Baicalein attenuates IRI by inhibiting the AGE/RAGE axis and downstream endoplasmic reticulum stress (132). EVLP optimization Functional repair strategies within EVLP systems have advanced through physiological stress conditioning and circulatory refinement. A rat model of warm ischemia demonstrated that transient controlled heat stress (41.5 ℃ for 30 minutes) successfully induced an endogenous heat shock response, significantly increasing heat shock protein 70 expression while improving static compliance and reducing edema (152). Subsequent optimization revealed that an EVLP duration of 4.5 hours (with 180 minutes of recovery following heat stress) maximized the upregulation of anti-apoptotic and antioxidant proteins compared to shorter or longer durations (153). Enhancing anatomical perfusion, the retention of bronchial artery circulation in a dual-circulation EVLP model was found to increase airway perfusion and reduce bronchial epithelial apoptosis compared to traditional pulmonary-only perfusion (154). Modifications to the perfusate composition and pharmacological interventions within the EVLP circuit further augment donor protection. The introduction of hepatocyte-conditioned perfusate significantly reduced advanced glycation end-products (AGEs), inhibited NF-κB signaling, and decreased matrix metalloproteinase-2 activity, suggesting that hepatic factors facilitate metabolic clearance and endothelial protection (155). Addressing economic and biocompatibility factors, a modified gelatin-based perfusate demonstrated efficacy in maintaining lung function during 4-hour EVLP while avoiding the inflammatory risks associated with high-concentration xenogeneic proteins (156). Pharmacologically, the novel S1P receptor 3 antagonist, PICCS-3, improved endothelial and bronchial epithelial cell survival during preservation by inhibiting apoptotic pathways, effectively reducing IRIs in transplant models (157). Multiple graft options for LTx Researchers have overcome the key immunological barriers in pig-to-human lung xenotransplantation (158). The world’s first 6-gene-edited pig-to-human LTx demonstrated genetic modification can effectively prevent hyperacute rejection, and the graft can maintain short-term function and vitality under multiple immunosuppression (159). Cell- and organoid-based therapies have increasingly become a research focus in recent years. KRT5(+)/KRT8(+) transitional cells derived from alveolar type II cells play a central role in pulmonary fibrosis and bronchiolization of the alveoli. Studies have shown that strategies such as pre-culturing cells into three-dimensional organoids and reducing immune clearance of transplanted cells can improve engraftment efficiency, providing a novel approach for replacing abnormal transitional cells and treating pulmonary fibrosis (160). Similarly, airway basal stem cells (BSCs) are capable of repairing epithelial injury; however, their clinical application still faces challenges related to immune rejection and technical limitations (161). Animal models refinement Suzuki and colleagues have developed a mouse-based decellularized lung platform that offers a cost-effective and technically feasible framework for optimizing cell-seeding protocols and assessing post-transplant perfusion (162). Moreover, recent studies have refined rat LTx techniques through optimized cuff methods and hilar anastomosis for both left and right lung procedures (163,164). These modifications significantly reduced ischemic times and procedural complexity, achieving a 100% operative success rate. Discussion In 2025, the field of LTx achieved significant milestones across both clinical and basic/translational domains. Clinically, advances were primarily reflected in allocation policy reform, donor expansion, preservation technologies, perioperative support, rejection surveillance, infection control, and complication management. In parallel, basic and translational studies provided deeper mechanistic insights into IRI, infection-associated allo-immunity, immune remodeling, and CLAD. However, the field remains characterized by a persistent gap between substantial improvements in access and perioperative care and the more limited progress in long-term graft and patient survival. This imbalance highlights a central challenge in contemporary LTx: although early outcomes have markedly improved, durable immunological tolerance and long-term graft function remain unresolved. Clinical research has shifted from expanding donor availability and improving surgical success toward optimizing resource allocation in the past years. The transition from geographically constrained LAS systems to the CAS system represents a major step toward improving equity, although the balance between medical urgency and post-transplant utility remains imperfect and controversies still exist. At the same time, technologies such as EVLP have enabled the safe utilization of marginal donors and enhanced surgical flexibility, particularly in the context of long-distance organ sharing. Nevertheless, these advances primarily target early-phase outcomes and have yet to translate into substantial improvements in long-term survival. In basic and translational research, recent advances reflect a shift from studying isolated mechanisms to understanding integrated disease networks. Mechanistic studies have revealed the roles of epithelial injury, immune cell reprogramming, and regulated cell death pathways, providing a more comprehensive framework for disease progression and offering potential noninvasive biomarkers for early detection. The application of frontier technologies, including multi-omics and machine learning, has further accelerated discovery; however, these findings require robust clinical validation. Despite the growing convergence between clinical and basic research around shared targets such as EVLP and CLAD, the translation of mechanistic insights into effective clinical interventions remains limited. To translate recent advances into durable clinical benefit, several priorities should guide future practice and research. Clinically, transplantation centers should adopt rigorous, data-informed candidate stratification to balance waitlist urgency with expected post-transplant benefit. Perioperative and post-transplant surveillance protocols should be standardized, the donor pool should be cautiously expanded through evidence-based use of marginal donors and EVLP, and noninvasive monitoring tools should be incorporated following adequate validation. From a research perspective, three major gaps require priority attention. First, noninvasive biomarkers need to be validated for the early detection of rejection and infection. Second, mechanistic discoveries, including regulated cell death pathways and immunometabolic remodeling, should be translated into targeted therapeutics, potentially delivered during ex vivo perfusion. Third, risk stratification tools based on artificial intelligence require robust external validation and prospective clinical testing. Addressing these gaps will be essential to ensure that innovations in LTx improve not only access to transplantation but also durable long-term outcomes. Overall, the future of LTx will depend not only on further incremental improvements, but also on the field’s ability to convert mechanistic insight into clinically actionable strategies that enable durable, long-term graft and patient survival. Conclusions In 2025, LTx continued to advance across organ allocation, donor expansion, preservation technologies, perioperative management, and mechanistic understanding of graft dysfunction. Advances in artificial intelligence, multi-omics, immune profiling, and noninvasive biomarkers have provided new opportunities for risk prediction, early diagnosis, and targeted intervention. However, PGD, CLAD, infection, rejection, and limited long-term graft survival remain major challenges. Future efforts should focus on validating emerging tools and translating mechanistic discoveries into clinically actionable strategies to improve durable graft and patient outcomes. Supplementary The article’s supplementary files as Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2026-0554/prf