Diagnosis and management of bronchiolitis obliterans syndrome following lung or hematopoietic cell transplantation

KEYWORDS:

• obliterative bronchiolitis
• bronchiolitis obliterans syndrome
• hematopoietic cell transplantation
• lung transplantation
• treatment

Introduction

Both allogeneic hematopoietic cell transplantation (HCT) and lung transplantation (LTX) can rescue patients from disorders with poor prognoses and allow improved quality of life and prolonged survival. However, patients who undergo HCT or LTX can develop lung function loss due to obliterative bronchiolitis (OB), which is characterized by small airway injury and inflammation that progresses to fibrosis and luminal obliteration that is irreversible. This post-transplant complication, which has been commonly referred to as bronchiolitis obliterans syndrome (BOS), is associated with high morbidity and mortality and usually leads to obstructive physiology with progressive lung function decline, impaired quality of life due to dyspnea and poor exercise tolerance, eventual need for chronic use of supplemental oxygen, and death due to respiratory failure.[1–3] Intensification of immunosuppression tends to provide only marginal benefit for recipients who develop BOS and significantly increases the risk of adverse events such as life-threatening infections. Therefore, more effective therapies are much needed to treat this inflammatory/fibrotic process that primarily affects small airway bronchioles.

Because OB is difficult to reliably detect without performing a surgical lung biopsy (e.g. via bronchoscopy with bronchoalveolar lavage and transbronchial lung biopsy), demonstration of a significant decline in forced expiratory volume in one second (FEV1) via pulmonary function testing (PFT) has been adopted as a surrogate marker for post-transplant lung function decline due to OB. A significant decline to ≤80% of stable post-transplant baseline FEV1 values that is not explained by potentially reversible, non-OB factors that can cause lung function impairment is considered diagnostic of BOS in LTX recipients.[3] Similarly, the diagnosis of BOS in HCT recipients requires a combination of airflow obstruction on PFT, evidence of air trapping by high-resolution computed tomography (HRCT) and/or increased residual lung volume, a non-OB manifestation of chronic graft-versus-host disease (cGVHD), and absence of an active pulmonary infection.[1,4] However, although histopathologic changes are similar for BOS in both HCT and LTX recipients, significant differences in incidence and prevalence, risk factors, pathophysiology, and management exist.

Do key differences exist between HCT-BOS versus LTX-BOS?

Loss of immunologic tolerance due to an allorejection response is commonly perceived as the major cause of BOS in both allogeneic HCT (HCT-BOS) and LTX (LTX-BOS). BOS in HCT recipients is a pulmonary manifestation of cGVHD that usually occurs within two years of transplant and is attributed to an immune response by grafted hematopoietic cells to antigens on bronchial epithelia.[1] The incidence of cGVHD (defined as any manifestation that is present/sustained ≥100 days post-transplant) ranges up to 80% and is affected by many factors (e.g. source of transplanted cells, donor type, recipient age, T-cell depletion). The prevalence of BOS in HCT recipients is, however, considerably lower than that of cGVHD without lung involvement and has been estimated to range from 5% to 14% using National Institutes of Health (NIH) criteria.[1,2,4] Median survival for patients who develop HCT-BOS has been estimated to be 13% at 5 years, while 5-year survival for recipients without this complication is generally considered to exceed 50–60%. In addition to the presence of cGVHD, risk factors linked to HCT-BOS have included exposure to specific chemotherapy agents, viral infection, and lower baseline FEV1 to SVC (slow vital capacity) ratio, but some suggested risk factors have not necessarily held up when multivariate analyses were performed.[2]

In contrast to HCT, BOS in LTX recipients is attributed to a host-versus-graft reaction to lung antigens, affects ≥50% of patients who survive beyond 5 years post-LTX, and is the leading cause of death for those who survive beyond one year post-LTX.[3] A more recent examination of LTX-BOS reveals that different phenotypes exist,[5] and use of the term chronic lung allograft dysfunction (CLAD) has been proposed as an overarching term for chronic lung allograft rejection that encompasses the subsets/phenotypes of (i) obstructive BOS, which typically lacks significant lung parenchymal changes on thoracic imaging and is characterized by obstructive physiology, and (ii) restrictive allograft syndrome (RAS), which is characterized by a restrictive pattern of lung physiologic impairment accompanied by parenchymal infiltrates (both entities were historically identified as BOS). Recent investigations indicate that RAS has a worse prognosis than the more prevalent obstructive BOS phenotype, and responses to therapy may also differ.[6] Risk factors associated with the development of LTX-BOS include primary graft dysfunction, acute rejection, persistent neutrophilia in bronchoalveolar lavage (BAL) sampling, lymphocytic bronchiolitis, antibody-mediated rejection, gastroesophageal reflux (GER), infections (viral, bacterial, fungal), autoimmunity, and exposure to air pollution.[3] Additionally, other factors (e.g. anastomosis dysfunction, allograft inflammatory or infectious complications, external allograft compression, primary disease recurrence, vascular obstruction) can affect allograft function, may coexist with lesions of OB, and can contribute to CLAD. Finally, some explanted allografts that met criteria for a diagnosis of BOS were found to have relatively rare lesions of OB but were characterized by considerable vascular remodeling and interstitial fibrosis,[7] which illustrates the need for alternative terminology (e.g. CLAD and CLAD phenotypes) that can supplant BOS as an overarching term for LTX-associated chronic allograft dysfunction.

How is BOS diagnosed?

Updated consensus criteria for the diagnosis of HCT-BOS include (i) FEV1 to vital capacity ratio <0.7, (ii) FEV1 < 75% predicted with ≥10% decline over <2 years, (iii) adequate exclusion of the presence of respiratory tract infection, and (iv) one of two supporting features (either HRCT evidence of small airway disease or increased residual lung volume on PFT).[4] Screening PFTs (spirometry and lung volumes) are recommended at day 100 post-HCT, at 1 year post-HCT, when GVHD is initially diagnosed, and at 6-month intervals after a GVHD diagnosis is established.[4] Additionally, because GVHD may occasionally affect only the lung, a lung biopsy is recommended if extra-pulmonary manifestations of GVHD are not detected. An algorithm to diagnose HCT-BOS has been suggested.[1]

Because criteria for LTX-BOS pertain to a lung allograft that must recover from initial peri-operative allograft injury and stabilize (which is usually accompanied by progressive improvement in spirometric values until a plateau is reached) the diagnosis of BOS (or CLAD) is made when persistent decline in FEV1 to ≤80% of a previously stable, maximal FEV1 value (an average of the 2 best post-LTX values) occurs, other possible causes are excluded, and the onset of lung function decline is delayed (generally beyond 3 months post-LTX).[3] Other testing, such as bronchoscopy with BAL and transbronchial lung biopsy and HRCT imaging, is needed to rule out other entities (e.g. restrictive syndromes or infection that may masquerade as BOS) as well as provide supportive evidence (e.g. air-trapping on HRCT), and an algorithm to facilitate obtaining a definitive diagnosis of LTX-BOS has been suggested.[3] Pulmonary function testing is typically obtained within 2–4 weeks following LTX when the patient has recovered from surgery and then periodically (e.g. every 3 months once stabilized over the first 12–24 months post-LTX) per transplant center protocols for surveillance.

How can BOS be treated?

Although intensified immunosuppression has been used to treat HCT- and LTX-BOS, this approach has had negligible impact on disease course and can substantially increase the risk of serious toxicity, especially the risk of serious and fatal opportunistic infections. While initial, intense corticosteroid therapy can significantly benefit HCT recipients with acute onset of GVHD or restrictive, late-onset, non-infectious pulmonary complications (organizing pneumonia, late interstitial pneumonia), such therapy does not appear to provide sustained benefit for BOS associated with cGVHD.[1,2,4,8] In LTX-BOS, although switching from cyclosporine A to tacrolimus as the calcineurin inhibitor component of maintenance immunosuppression may stabilize LTX-associated BOS for some recipients [3] and high-dose corticosteroid pulse therapy is routinely used to treat episodes of acute rejection, corticosteroids are not an effective therapy for treating LTX-BOS, and the use of chronic higher doses of corticosteroids to treat BOS is not recommended.[3] Novel, non-corticosteroid, treatment strategies are needed to arrest the underlying processes that lead to progressive loss of lung function when LTX-BOS is detected.

Azithromycin (AZI), a neo-macrolide antibiotic with both antibacterial and immunomodulatory properties,[9] has been used to treat BOS in both HCT and LTX recipients, and a growing body of clinical research shows that a substantial number of patients may respond to AZI or AZI-based combination therapy. Norman et al. [10] reported a prednisone-sparing effect of combination therapy with inhaled fluticasone propionate, AZI, and montelukast (FAM) for a case series of eight HCT-BOS recipients compared to 14 historical controls (inhaled fluticasone and montelukast have also been suggested to have potential treatment benefit for BOS in both HCT and/or LTX recipients on the basis of small case series investigations). Williams et al. [11] subsequently reported that FAM therapy for a larger cohort of 36 patients enrolled in a single-arm, open-label, multicenter clinical trial was associated with a significantly reduced likelihood of treatment failure (defined as ≥10% decline in FEV1 at 3 months) for the FAM cohort versus historical controls and allowed a substantial reduction in corticosteroid dosing. Additional multi-center, prospective, randomized, placebo-controlled clinical trials (RCTs) are currently underway to better evaluate the efficacy of AZI for HCT-BOS, but results are not yet available.

Because CLAD/BOS is a much more prevalent complication following LTX as compared to HCT-associated BOS, a much larger body of literature reporting clinical research trial outcomes has been published that supports a beneficial effect of AZI monotherapy for a substantial number of recipients who develop a decline in lung function that met criteria for LTX-BOS. Because several reports have consistently indicated that AZI can benefit approximately 30–40% of LTX recipients who develop allograft dysfunction that meets BOS criteria (especially patients with BAL neutrophilia and those with earlier disease stage), a clinical practice guideline conditional recommendation was made (although based upon very low-quality evidence due to lack of well-powered RCT data) to treat a decline in FEV1 consistent with the onset of BOS with AZI.[3] A more recently published, placebo-controlled RCT of AZI for later onset LTX-BOS also supports clinical benefit for a substantial number of patients treated with AZI,[12] and AZI therapy for LTX-BOS has been associated with relatively few adverse events.[3] It should be noted that newer views on classification of CLAD suggest that the term azithromycin-responsive allograft dysfunction may be a more precise and preferred term for graft dysfunction that reverses or at least improves significantly in response to AZI therapy.[5]

Prophylactic AZI has also been evaluated as an intervention to prevent BOS in both LTX and HCT recipients. Vos et al. [13] published a placebo-controlled RCT that demonstrated significantly improved, BOS-free survival and significantly better preservation of FEV1 in the AZI treatment arm when AZI started prior to hospital discharge following LTX. Additionally, longer term analysis of the treatment arms from this same RCT showed that prophylactic AZI was associated with significantly improved CLAD-free survival, pulmonary function, and functional exercise capacity versus the placebo group.[14] In contrast, a study of prophylactic AZI for allogeneic HCT recipients did not demonstrate an ability to prevent BOS,[15] but this study employed a non-randomized, retrospective design, and significant differences in baseline characteristics were present between the AZI and placebo groups.

Other therapies may benefit HCT and/or LTX recipients who develop BOS. A substantial number of patients with end-stage lung disease have excessive GER, and new onset GER can develop following LTX. Because pathologic GER has been identified as a risk factor for developing LTX-BOS and anti-reflux surgery (which has been associated with improvement in FEV1 in a number of observational studies) can be performed with relative safety in LTX recipients who develop FEV1 decline consistent with the diagnosis of BOS, consideration for referral to an experienced surgeon for anti-reflux surgery is suggested, although this is a conditional recommendation based upon very low quality evidence.[3] In contrast to LTX, GER has not been identified as a risk factor for HCT-BOS.

Because airway fibrosis is involved in the pathogenesis of OB, anti-fibrotic agents may be useful to treat BOS, and both pirfenidone and nintedanib have been recently approved for treatment of idiopathic pulmonary fibrosis. Pre-clinical studies suggest that pirfenidone may be useful to prevent or ameliorate OB associated with LTX, and recent case reports suggest that pirfenidone may be useful to treat both obstructive and restrictive forms of LTX-associated CLAD.[16]

Extracorporeal photophoresis (ECP) has been associated with significant benefit for the treatment of acute GVHD in HCT recipients and has also been associated with significant responses even in steroid-refractory acute GVHD, cGVHD, and BOS.[17,18] Although clinical investigations of ECP for HCT-BOS are fairly limited, a more substantial experience with ECP has been reported for LTX-BOS.[3,19] These studies have shown a significant lessening of FEV1 decline, and Greer et al. [19] found significantly improved progression-free survival in a LTX recipient cohort that had been refractory to first-line therapy with AZI, and subgroup analysis showed that patients with BAL neutrophilia were more likely to respond to ECP.

Primary lung transplantation for HCT-BOS or lung retransplantation for LTX-BOS are potential therapies for patients with progressive loss of lung function that is refractory to pharmacologic or other therapies, and satisfactory outcomes have been reported for both HCT recipients and for patients with refractory pulmonary dysfunction following LTX. Post-LTX survival is higher for recipients retransplanted for BOS versus other indications,[3] and a recent, multi-center study suggests that LTX recipients with the RAS CLAD phenotype have a worse retransplant outcome than those with the obstructive BOS CLAD phenotype.[20]

Summary and conclusion

Although delayed loss of lung function following allogeneic HCT or LTX has been termed BOS and viewed as predominantly driven by the development of OB, our understanding, classification system, and terminology/definitions for pulmonary dysfunction following HCT and, to a greater degree, persistent decline in allograft function following initially successful LTX, are in flux. Nonetheless, some treatments, such as AZI therapy alone or the combination therapy, FAM, appear to benefit a substantial number of patients, and prophylactic administration of AZI may decrease the risk of developing LTX-associated BOS. An evolving clinical experience with ECP suggests that a significant, beneficial effect for at least a subset of HCT or LTX recipients who develop BOS may exist. Anti-fibrotic therapy (e.g. pirfenidone) may also prove beneficial, but additional research is needed to determine the efficacy of new therapies and to identify disease phenotypes that respond to specific interventions. Additionally, the ability to attain an early diagnosis of BOS is likely to play a key role in improving responses to new therapies for BOS. However, the net survival benefit for patients with this complication (BOS/CLAD) must be weighed against the medical risks to patients associated with these therapies as well as the potential economic consequences to both patients and their health care systems. There is an urgent need to develop truly effective therapies to prevent and/or treat these life-threatening, devastating complications of HCT and LTX, and adequately powered RCTs to support the efficacy of novel agents and interventions represent the best approach for demonstrating true therapeutic efficacy.

Declaration of interests

K.C. Meyer has served on a Clinical Advisory Board for InterMune and on an Adjudication Committee for Medimmune, and he has received research funding from Abbott, Actelion, Altana, American College of Physicians, American Lung Association of Wisconsin, Amgen, Asthmatx, Bayer, Boehringer-Ingelheim, Bristol Meyers Squibb, Chiron, Cystic Fibrosis Foundation, Discovery Labs, DuPont Merck, Fibrogen, Genentech, Gilead, GlaxoSmithKline, Inspire. InterMune, Johnson & Johnson, National Institutes of Health, Novartis, Nycomed, Parion, Pfizer, Pharmaxis, PreAnalytiX, Roche, Ross, Vertex and Wyeth. The author has no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.