Primary Graft Dysfunction After Lung Transplantation

Abstract

Primary graft dysfunction (PGD) is a clinical syndrome occurring within the first 72 hours after lung transplantation and is characterized clinically by progressive hypoxemia and radiographically by patchy alveolar infiltrates. Resulting from ischemia-reperfusion injury, PGD represents a complex interplay between donor and recipient immunologic factors, as well as acute inflammation leading to alveolar cell damage. In the long term, chronic inflammation invoked by PGD can contribute to the development of chronic lung allograft dysfunction, an important cause of late mortality after lung transplant. Recent work has aimed to identify risk factors for PGD, focusing on donor, recipient, and technical factors both inherent and potentially modifiable. While no PGD-specific therapy currently exists, supportive care remains paramount and early initiation of ECMO can improve outcomes in select patients. Initial success with ex vivo lung perfusion platforms has been observed with respect to decreasing PGD risk and increasing lung transplant volume; however, the impact on survival is not well delineated. This review will summarize the pathogenesis and clinical features of PGD, as well as highlight treatment strategies and emerging technologies to mitigate PGD risk in patients undergoing lung transplantation.

Introduction

Lung transplantation is the gold standard therapy for respiratory failure refractory to medical management. Primary graft dysfunction (PGD) is an important contributor to early mortality following lung transplantation and occurs in as many as 30-80% of lung transplant recipients in the first 48-72 hours following transplant.^[1-3] In the long term, PGD has also been implicated in the development of chronic lung allograft dysfunction (CLAD).^[4] This review will summarize our current understanding of PGD with a focus on clinical, operative, and physiologic characteristics; and will highlight future directions for prevention and treatment of PGD in the lung transplant population.

Definition

“PGD” historically referred to graft failure after lung transplantation and is a form of acute respiratory distress syndrome (ARDS). Prior to a standardized terminology, PGD was referred to by several other names: ischemia-reperfusion injury, non-cardiogenic pulmonary edema, early graft failure, and post-transplant ARDS among them.^[5] In an attempt to standardize language surrounding graft dysfunction following lung transplant, the International Society for Heart and Lung Transplantation (ISHLT) published diagnostic criteria for PGD in 2005.^[6] This consensus terminology allowed for greater recognition of PGD as a syndrome and further opportunities to investigate targeted therapies. This definition was further refined by the ISHLT in 2016^[7] to address limitations of the 2005 consensus with respect to ECMO, high flow oxygen, and the use of pulmonary vasodilators.^[1]

The 2005 ISHLT diagnostic criteria for PGD is presented in Table 1 and is based on both radiographic criteria (present of patchy alveolar infiltrates on chest radiographs) and the PaO2/FiO2 (P:F) ratio observed immediately post-operatively and again at 24-hour intervals for the first 72 hours following lung transplantation.^[6] The 2005 consensus criteria was updated in 2016 to reflect that patients receiving high flow oxygen should be graded in the same way as those requiring invasive mechanical ventilation; that use of pulmonary vasodilators should not affect grading; and that all patients requiring ECMO remained PGD grade 3.^{[1, 7]}

Table 1.

2005 ISHLT Diagnostic Criteria for Primary Graft Dysfunction

Grade at T0, T24, T48, T72	Radiographic infiltrates consistent with diffuse pulmonary edema	PaO2:FiO2* (P:F Ratio)
0	Absent	Any
1	Present	>300
2	Present	200-300
3	Present	<200

^*measured on FiO2 = 1.0 and PEEP 5 cm H₂O

Source: adapted from reference 6^[6]

Importantly, the diagnosis of PGD requires exclusion of mechanical, immune, and infectious etiologies for hypoxemia.^[2] In addition to chest radiograph opacities, computed tomography (CT) of the chest may reveal abnormalities including ground glass opacities, peribronchial and perivascular thickening, and/or reticular interstitial and airspace opacities occurring predominantly in the middle and lower lobes. ^[8-10] PGD typically begins on post-operative day (POD) 1, with clinical and radiographic signs peaking by POD3. The time course of resolution is variable and depends on a number of clinical factors, but typically improvement is seen by POD5 with resolution by POD10. ^[8-11] Of note, improvement in the P:F ratio may be seen prior to resolution of radiographic abnormalities, as there is often poor correlation between the radiograph and clinical status.^[2]

Pathogenesis

The pathogenesis of PGD is complex and represents an interplay between baseline donor and recipient factors, and a proinflammatory cascade that begins at the time of reperfusion. Previous work has established that PGD occurs in two phases: an early phase modulated by donor lung macrophages and lymphocytes, and a later phase regulated by recipient neutrophils and lymphocytes.^[2] In response to acute inflammation, reactive oxygen species (ROS) are generated and accelerate the pace of injury by disturbing cellular function, activating proteolytic enzymes, and increasing autophagy. ^[12-14] The transcriptional signature of this process of ischemia-reperfusion injury is the upregulation of genes involved in cell survival, cell surface signaling, and oxidative stress response.^[4] Histologically, the alveolar spaces in acute PGD demonstrate significant endothelial and epithelial injury, neutrophilic inflammation, and edema. ^[15] These effects carry over in the long-term, as the inflammation increases the immunogenicity of the allograft, as evidenced by the observation of increased levels of class II donor specific human leukocyte antigen (HLA) antibodies at 5 years post-transplant.^[16]

Significant interest in the immunologic signature of PGD has resulted in a body of evidence that proinflammatory factors including free radicals, reactive oxygen species, cytokines, and neutrophilic inflammation are critical to the development of PGD.^{[1, 2, 13-17]} The recipient inflammatory environment plays a crucial role in the development of PGD post-transplant. An important study using bronchoalveolar lavage (BAL) and tissue biopsy revealed that levels of IL-8 correlate with PGD severity and mortality in a dose-dependent manner.^[18] More recently, a single center analysis of patients with PGD post-transplant found differential perioperative expression of multiple cytokines including IL-8, G-CSF, PDGF-BB, and IL-1Ra, and a correlation of cytokine levels with grade 3 PGD at 72 hours.^[19] Neutrophilic activation has been identified as a unifying mechanism of PGD, and in animal models depletion of neutrophils prior to lung transplant attenuated PGD severity.^[17] Downstream effects of neutrophilic inflammation include ROS generation, activation of proteolytic enzymes, propagation of cellular death, and cytokine release, exacerbating injury.^{[20, 21]} Studies investigating genetic determinants of PGD have pointed to genes in inflammasome-mediated innate immune pathways and genes involved in oxidant stress regulation.^[22-24] Finally, donor genetic and immune status also plays a role, with increased donor cytokine levels and donor NOX3 polymorphisms shown to be associated with PGD risk.^{[23, 25]}

Outcomes

Prior to a consensus definition of PGD, a broad range of incidence of allograft dysfunction was reported, from 15-57% of all lung transplant recipients.^[26] In a more recent multicenter analysis of 1528 patients who underwent lung transplantation between 2011 and 2018, 25.7% of subjects experienced PGD.^[27] Moreover, annual incidence of PGD increased over the course of the study (14.3%-38.2%, p=0.0002), in tandem with changes in organ allocation that have prioritized sicker patients with higher median LAS scores (38.0-47.7, p=0.009).^{[27, 28]} Additionally, there remains significant variability by center and region, suggesting that center-specific protocols, experience, and technique may play a role in PGD and its subsequent management.^[2] A recent multicenter analysis of 7322 lung transplant recipients across 72 centers noted a near-linear relationship between center volume and odds of PGD (OR 0.94 for every 10 transplants, 95% CI 0.89-0.99) as well as a survival benefit for patients with PGD3 at high volume centers.^[29]

PGD carries significant morbidity both early and late following transplant, with early complications including increased intensive care unit length of stay, more frequent use of ECMO, prolonged mechanical ventilation, and increased short-term mortality. ^{[1, 2]} In fact, in the first 30 days following lung transplantation, PGD accounts for 50% of all-cause mortality.^[2] Furthermore, it is well-established that PGD is associated with worse medium- and long-term survival; an effect that persists even when conditioned on 1-year survival. In an analysis of 1000 lung transplant recipients, Kreisel et. al. reported survival rates of 72.8, 43.9, and 18.7% at 1, 5, and 10 years post-operatively for patients with PGD.^[30]

PGD is linked to chronic lung allograft dysfunction (CLAD), the leading cause of late mortality of lung transplantation characterized by progressive loss of allograft function after transplant.^[31] In addition, severity of PGD is directly correlated to an increase in the relative risk of bronchiolitis obliterans syndrome (BOS).^[1] More recently, it has been suggested that PGD is associated with baseline lung allograft dysfunction (BLAD), an indicator of failure of normalization of lung function post-transplant. In a recent analysis, patients with PGD3 developed BLAD more frequently (58% vs 36%) and of more severe grades in a dose-dependent fashion related to highest PGD grade at 48-72 hours after transplant.^[31] Taken together, these findings reflect the consequences of PGD immediately post-operatively, as well as persisting into the long term.

While it is widely agreed upon that PGD affects morbidity and mortality following lung transplant, there are fewer studies that investigate the impact of PGD on patient quality of life or functional status. A link between poor quality of life and PGD is inferred based on the association of PGD with BOS and CLAD. With respect to functional status, patients with PGD have been shown to have shorter median walk distances on 6-minute walk tests and lower mean FEV1 than patients without PGD.^{[32, 33] [34]}

Risk Factors

There is substantial interest in identifying risk factors for the development of PGD, a task that was particularly challenging prior to the implementation of the 2005 consensus definition. Broadly speaking, risk factors for PGD can be divided into three categories: donor-specific, recipient-specific, and technical (related to aspects of procurement and/or implantation). Recognized risk factors are presented by category and summarized in Table 2.

Table 2.

Risk factors for PGD

Category	Risk Factors
Donor	Age >45 years, age <21 years African American race Female gender History of smoking Prolonged mechanical ventilation Aspiration Positive sputum culture Head trauma Hemodynamic instability after brain death
Recipient	BMI >30 Female gender Diagnosis of idiopathic pulmonary hypertension Diagnosis of secondary pulmonary hypertension Diagnosis of idiopathic pulmonary fibrosis Diagnosis of sarcoidosis Elevated pulmonary arterial pressure at time of surgery (post-reperfusion) Elevated pre-transplant levels of IL-10, IL-8, IL-6, and chemokine (CC-motif) ligand
Operative	Single lung transplantation Prolonged ischemic time Use of cardiopulmonary bypass Blood product transfusion >1L High FiO2 >0.4 at time of reperfusion Use of intracellular (hyperkalemic) type (Euro-Collins) preservation solution

Source: adapted from reference 2^[2]

Donor Risk Factors

Despite increasing demand for lung transplantation, overall donor lung utilization rates remain low, with significant declines noted in the wake of the COVID-19 pandemic.^[35] Nearly all donor lungs have some degree of injury related to the mechanism of death; this may include pulmonary emboli, edema, aspiration pneumonia, and/or trauma.^{[3, 36]} While there is renewed interest in the use of extended criteria donors to supplement the traditional donor pool, lack of consensus in patient-centered studies has limited widespread use.^{[37, 38]} In the future, ex-vivo lung perfusion may play a critical role in expanding donor lung supply by reversing donor conditions currently thought to preclude transplantation (e.g. pneumonia, pulmonary edema). In the interim, supply has increased through a more focused management of donors including low tidal volume ventilation, PEEP 8-10 cm H₂0, avoidance of volume overload, recruitment maneuvers, and bronchoscopic evaluation and treatment.

Recognized donor specific risk factors for the development of PGD relate to both the manner of death and donor management preceding organ procurement. Inherent risk factors related to the donor include increased age (>45), tobacco use (smoking history >20 pack years), chronic alcohol use, female gender, and African American race. ^{[2, 12, 39]} Recent evidence suggests that increasing organ age augments the damage related to ischemia due to impaired tolerance against damaging stress.^[12] In a similar way, prior donor tobacco or alcohol use is thought to contribute to the development of PGD by alterations in the redox state.^[4] Infection also remains of significant concern, as donor culture positivity is associated with a threefold increase in the likelihood of PGD.^[40]

Recipient Risk Factors

Recipient risk factors implicated in the development of PGD include female gender, African-American race, obesity (BMI >30 kg/m²), pre-operative right ventricular dysfunction, and higher pulmonary artery pressures post-reperfusion (mean >25, systolic >41).^{[41, 42]} Importantly, there is no association between recipient age and PGD.^{[43, 44]} Recipient diagnoses consistently shown to increase the risk of PGD include pulmonary arterial hypertension, idiopathic pulmonary fibrosis, and sarcoidosis.^{[43, 45]} Preliminary studies have also suggested that PGD is more frequently observed in patients who underwent lung transplant for COVID-19 related lung fibrosis (70% vs 20.8%); however, generalization of this finding is limited by small sample size.^[46] In immunologic studies, higher pre-transplant levels of IL-10, IL-8, IL-6, and chemokine (CC-motif) ligand were associated with PGD,^[47] suggesting that a heightened immunologic state may play a detrimental role in propagating post-operative allograft inflammation. Finally, practice-focused studies have suggested low PEEP (5 cm H₂0) and calculation of tidal volumes on the basis of donor lung size (rather than recipient size decrease the risk of PGD in the early post-operative period.^[48]

Technical Risk Factors

Technical risk factors for PGD relate to mechanism of procurement and preservation, as well as conduct of implant operation. It cannot be emphasized enough that procurement practices play a crucial role in PGD risk and in this regard, standardization of practices is crucial. The standard approach to procurement includes a low potassium, acellular preservation fluid with antegrade and retrograde flushing of pulmonary vasculature, FiO2 of 30-50%, maintenance of total lung capacity near 50%, and hypothermic organ preservation at 4-9 degrees Celsius with or without adjuncts such as prostaglandin or heparin.^{[3, 49]} That said, normothermic ex-vivo lung preservation has more recently been trialed with initial promising success and observed improvement in severe PGD observed at 72 hours post-transplant.^[50] However, additional studies are required to determine whether this observation translates into survival advantage in the long term.

Following procurement, conduct of the implant operation is also a significant contributor to PGD risk, with the major contributing variables including use of mechanical circulatory support, ischemic time, and type of transplant (single vs bilateral). ^[13] Cardiopulmonary bypass (CPB) has consistently been shown to increase PGD risk: in a recent multicenter analysis of 852 patients, patients undergoing lung transplant with CPB were found to have the highest risk of PGD compared to transplants performed without CPB (OR 4.24) or with VA ECMO (OR 1.89).^[51] Hemodynamic instability during transplant has been linked to increased PGD risk: in a single institution study, intraoperative blood lactate level <2.6 had a high negative predictive value for the subsequent development of PGD. ^[52] Finally, while this remains controversial, in previous analyses, single lung transplant was associated with increased risk of PGD compared to patients who underwent bilateral transplantation.^[43]

Treatment

Treatment for PGD is largely supportive and is often approached in a manner similar to patients with ARDS. Important strategies include lung protective ventilation, restriction of excess fluid administration, early mobilization, and treatment of underlying infection. By contrast, PGD treatment also includes avoidance of proning, less frequent paralytic use, and early initiation of ECMO. Traditional pharmacologic therapies used for ARDS such as inhaled beta-2 agonists, renin-angiotensin-aldosterone system inhibitors, and antioxidants are less effective.^[41] Recent attention has been turned to donor-based lung protective ventilation, which mitigates both PGD risk and the adverse outcomes associated with donor-recipient size mismatch.^[53] Calculation of tidal volume on the basis of donor (rather than recipient) size has consistently been shown to decrease the risk of severe PGD at 48-72 hours post-transplant, and is also associated with decreased 1 year mortality.^{[48, 53]}

Early initiation of ECMO is associated with better survival for patients with PGD, and is not infrequently encountered in this population. Multiple cannulation strategies exist, and advances in technique and patient care have led to recent improvement in outcomes for these patients.^[54] A recent single institution analysis of patients requiring ECMO post-lung transplant found 90-day, 1 year, and 5 year survival rates of 67.3, 50, and 31.5% in this population. Notably, higher in-house and 3 year mortality rates were observed in patients who were placed on ECMO more than 48 hours post-transplant compared to those in whom ECMO was initiated early, further emphasizing that early recognition of need for ECMO is critical to improved outcomes.^[55]

Ex-vivo Lung Perfusion

Ex-vivo lung perfusion (EVLP) is a promising technology that may address many shortcomings of traditional lung donation, including low overall utilization rates of available donor lungs.^[35] In the United States, two systems have achieved FDA approval: the Organ Care System (OCS Lung, Transmedics, Andover MA) and the XPS XVIVO Perfusion AB system (XVIVO Perfusion, Go-theborg, Sweden) and have been evaluated both for the use of traditional and extended criteria lungs.^[56] EVLP platforms have the potential to increase lung transplant volume by affording the opportunity to rehabilitate donor lungs, more closely evaluate allograft function prior to implant, administer antibiotics pre-operatively, and optimize allograft function.^[57] Early studies from centers using EVLP have demonstrated increased transplantation volume of up to 33%.^[58] The data regarding PGD is less clear: in the largest cohort study to date, there was no difference in the incidence of severe PGD at 72 hours between EVLP and control lungs.^[59] However, more recent studies have noted decreased PGD rates compared to traditional cold storage (17.7% vs. 29.7%).^[14] What is unknown is whether a potential benefit in PGD translates into improvements in long-term survival. Currently available data suggests that despite a decrease in PGD with EVLP, short- and long-term survival and functional outcomes are similar to traditional transplantation. ^{[14, 59, 60]} Thus, while tremendous improvements in EVLP technology have amplified its use and illuminated its potential benefits, future research and perseverance are critical to realizing the potential of the EVLP platform.

Conclusion

Primary graft dysfunction remains an important contributor to both early- and late-term mortality following lung transplantation and is a consequence of the deleterious effects of ischemia-reperfusion injury occurring at the time of implantation. PGD represents a complex interplay between donor and recipient immune responses and leads to a cascade of acute and chronic inflammation that contributes to chronic lung allograft dysfunction. The consensus definition of PGD has allowed for a closer study of incidence, identification of risk factors, and realization of potential targets for therapeutic intervention. Despite this, no specific therapy exists for PGD, and care remains supportive, including ventilatory support and early ECMO for severe cases. EVLP platforms hold substantial promise as a means by which to increase lung transplant volume and potentially mitigate PGD risk. Future studies are required to further investigate this promising technology, as well as to develop other therapies targeting the underlying pathophysiology of PGD.

Key points.

• Primary graft dysfunction is an important contributor to early mortality after lung transplantation and is linked to the development of chronic lung allograft dysfunction (CLAD) in the long term.

• PGD is characterized clinically by the presence of patchy alveolar infiltrates on chest radiograph and PaO₂:FiO₂ ratio <300 and is graded based on severity at several time points following implant.

• PGD results from ischemia-reperfusion injury and a subsequent inflammatory cascade leading to endothelial and epithelial damage of alveolar cells.

• Treatment for PGD is largely supportive, and early initiation of mechanical support may provide a survival benefit for patients with severe PGD.

• Ex vivo lung perfusion is an important technology that may mitigate PGD risk and increase lung transplant volume, but future studies are needed to determine its impact on survival.