Expert Review of Respiratory Medicine Downloaded from informahealthcare.com by Nyu Medical Center on 05/30/14 For personal use only.
Review
Lung transplantation in patients with cystic fibrosis: special focus to infection and comorbidities Expert Rev. Respir. Med. 8(3), 315–326 (2014)
Daniel J Dorgan and Denis Hadjiliadis* Division of Pulmonary, Allergy and Critical Care, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA 19104, USA *Author for correspondence: Tel.: +1 215 662 4202 Fax: +1 215 614 0869 [email protected]
Despite advances in medical care, patients with cystic fibrosis still face limited life expectancy. The most common cause of death remains respiratory failure. End-stage cystic fibrosis can be treated with lung transplantation and is the third most common reason for which the procedure is performed. Outcomes for cystic fibrosis are better than most other lung diseases, but remain limited (5-year survival 60%). For patients with advanced disease lung transplantation appears to improve survival. Outcomes for patients with Burkholderia cepacia remain poor, although they are better for patients with certain genomovars. Controversy exists about Mycobacterium abscessus infection and appropriateness for transplant. More information is also becoming available for comorbidities, including diabetes and pulmonary hypertension among others. Extra-corporeal membrane oxygenation is used more frequently for end-stage disease as a bridge to lung transplantation and will likely be used more in the future. KEYWORDS: Burkholderia cepacia • cystic fibrosis • cystic fibrosis-related diabetes • extracorporeal membrane oxygenation • lung transplantation • Mycobacterium abscessus
Cystic fibrosis (CF) is the most common lifelimiting genetic disease among Caucasians [1]. The primary cause of death is progressive lung disease due to recurrent infections and destruction of pulmonary tissue. Lung transplantation is often performed in CF patients with endstage lung disease. CF is the third most common indication for lung transplantation, accounting for 16.7% of all lung transplants and 26.3% of bilateral lung transplants performed from January 1995 to June 2011 [2]. Based on registry data from the International Society for Heart and Lung Transplantation (ISHLT), survival after lung transplantation is greater for CF than for other diagnoses, although this comparison is not adjusted for recipient age, bilateral versus single lung transplant or other clinical factors. Overall, median survival after lung transplantation is 7.5 years for CF, compared with 5.5 years for all lung transplants and 6.7 years for all bilateral lung transplants. Combining all diagnoses, posttransplant survival was 88% at 3 months, 79% at 1 year, 64% at 3 years, 53% at 5 years and 30% at 10 years as of the 2012 ISHLT registry informahealthcare.com
10.1586/17476348.2014.899906
report. For CF, post-transplant survival was 90% at 3 months, 83% at 1 year, 69% at 3 years, 60% at 5 years and 43% at 10 years [2]. This article reviews factors that have been studied as potential predictors of outcomes in lung transplantation for CF. However, given prior controversy regarding survival benefit with lung transplantation in CF, we begin with a discussion of studies that have addressed the impact of transplant on survival in patients with advanced lung disease due to CF. Survival benefit of lung transplantation in CF Early studies
Several studies have cast doubt on whether lung transplantation confers a survival benefit for patients with CF, with results suggesting that lung transplantation does not improve survival among CF patients with end-stage lung disease [3–5]. One criticism of these studies is that they model survival using Cox proportional hazards with lung transplantation as the only time-dependent variable [6]. By including pre-transplant clinical data from
Ó 2014 Informa UK Ltd
ISSN 1747-6348
315
Expert Review of Respiratory Medicine Downloaded from informahealthcare.com by Nyu Medical Center on 05/30/14 For personal use only.
Review
Dorgan & Hadjiliadis
only a single point in time, the models could underestimate the increasing mortality risk over time due to progression of disease without lung transplantation, thereby creating bias when comparing the risk of mortality without transplant with the risk of mortality after lung transplantation. Perhaps more important, these studies were performed when priority for allocation of lungs for transplantation was based on time on the waiting list rather than on clinical characteristics, which could impact the potential for benefit or harm from lung transplantation. That is, an allocation system based on time on the waiting list preferentially provides transplants to those who are able to remain stable long enough to receive a transplant (and therefore may have a lower instantaneous risk of death without a transplant), while rapidly declining patients who potentially could derive the most benefit from transplantation may not have been able to accrue adequate time on the waiting list to receive a transplant before dying. Older studies that have been done using a priority system for lung allocation have found a survival advantage with lung transplantation in CF [7]. On the other hand, such studies are also biased because survivors to transplant have inherent differences from patients who die; these differences likely cannot all be controlled for, thereby favoring transplant. The controversy was even greater for adolescents with CF when a well publicized study suggested that lung transplant might harm the majority of that group [4]. Aside from the factors noted above, one criticism of this paper was the inclusion of data from all eras, leading to potential bias as lung transplantation outcomes were worse in earlier eras. Post-transplant survival under the lung allocation score
The Lung Allocation Score (LAS) system was implemented in the USA in May 2005, replacing the system in which priority for lung transplant was based on how long an individual had remained on the waiting list. The LAS uses clinical data to predict 1-year survival both with and without lung transplantation, and normalizes the results on a scale of 0–100, with higher scores leading to higher priority for lung transplantation. Clinical factors that increase 1-year mortality without lung transplantation or increase the likelihood of 1-year survival after lung transplantation increase an individual’s LAS [8]. A recent study used a different approach in comparing the risk of death with or without lung transplantation in CF, incorporating LAS into the statistical model as a surrogate for disease severity. Using data from the United Network for Organ Sharing (UNOS), this study included multiple models to assess the impact of lung transplantation on survival. One of the Cox proportional hazard models used lung transplant as the only time-dependent covariate, incorporating other factors as time-independent (using baseline values for those variables), which was a similar approach to prior studies. A separate analysis used a joint model for longitudinal and survival data, a newer approach that incorporates mixed-effects modeling to describe changes of a covariate over time [9], allowing the variable (in this case LAS) to be included as a time-dependent covariate in the Cox model. This model included LAS from 316
the longitudinal (time-variant) model, lung transplant as a time-dependent covariate and an interaction term between the two. Models were stratified by transplant center. Of 704 patients in the dataset, 67% underwent lung transplantation, 14% died on the waiting list and 19% remained on the waiting list. Consistent with registry data cited above, posttransplant survival in CF was greater than post-transplant survival for all diagnoses: 96.5% at 3 months, 88.4% at 1 year and 67.8% at 3 years. Using traditional modeling with lung transplant as the only time-dependent covariate, there was a survival benefit with lung transplantation both for the unadjusted model and on multivariable analysis (suggesting that negative results in prior studies may have been due to the allocation system prior to LAS, rather than the statistical methods used). There was an interaction between lung transplant and LAS at listing such that lung transplantation was found to confer a survival advantage when LAS was greater than 30. The joint model incorporating both lung transplantation and LAS as time-dependent covariates similarly found an interaction between transplant and LAS, with higher LAS conferring a greater survival benefit with lung transplantation. These recent data strongly support lung transplantation as a means to improve survival in the most severely ill CF patients under the LAS system [10]. Microbiology
ISHLT guidelines list overt sepsis as an absolute contraindication to lung transplantation. Colonization with highly resistant or highly virulent bacteria’ is considered a relative contraindication, without further specification regarding particular organisms or antibiotic-resistance pattern as an absolute contraindication in and of itself [11], although the guidelines do note potential increased risk associated with certain pathogens in CF. In this section, we discuss data regarding lung transplant outcomes based on sputum microbiology prior to transplantation. Burkholderia cepacia complex
Burkholderia cepacia complex has been associated with worse outcomes in patients with CF, irrespective of lung transplantation [12]. This is even more significant when patients are colonized with more pathogenic strains like Burkholderia cenocepacia ET12 [13]. Sputum colonization with B. cepacia complex in CF patients prior to lung transplantation has been associated with increased mortality after transplant in several studies. One single-center study in Toronto showed that the most common cause of early post-transplant mortality among B. cepacia-positive patients was widespread pneumonia, including areas of lung necrosis and cavitation in some patients. The genomovar or subspecies was not specified for any patients in this retrospective study [14]. Another single-center retrospective study also found increased post-transplant mortality among patients infected preoperatively with B. cepacia complex, but found excess mortality only among those infected with B. cepacia genomovar III (now known as B. cenocepacia) [15]. Other retrospective studies at individual centers have shown similar results [16,17]. Expert Rev. Respir. Med. 8(3), (2014)
Expert Review of Respiratory Medicine Downloaded from informahealthcare.com by Nyu Medical Center on 05/30/14 For personal use only.
Lung transplantation in patients with CF
A subsequent study coordinated by the B. cepacia Research Laboratory and Repository assessed outcomes associated with different Burkholderia subspecies compared with absence of known Burkholderia among CF patients aged 12 and older listed for lung transplant on the Scientific Registry of Transplant Recipients. There was a significant increase in adjusted post-transplant mortality among patients infected with Burkholderia gladioli and non-epidemic strains of B. cenocepacia. There was no difference in mortality among those infected with Burkholderia multivorans, epidemic strains of B. cenocepacia or other B. cepacia complex species, although caution should be exercised when interpreting these results given the small sample size for each subspecies of Burkholderia [18]. In addition, in this group B. cenocepacia ET12 was very uncommon. A study at another transplant center employed a similar post-transplant antibiotic and immunosuppression protocol in transplant recipients colonized with B. cepacia complex as the regimen described above in the Toronto study. This study did identify the B. cepacia subspecies for each patient. Recipients infected with B. cenocepacia prior to transplant demonstrated excess mortality despite the modified post-transplant treatment regimen, with two of the eight patients dying within 30 days and five dying within 1 year; all five deaths occurred in the context of B. cenocepacia sepsis, often with the presence of empyema. Ten patients were infected with other Burkholderia subspecies (four with Burkholderia vietnamiensis, five with B. multivorans and one with both B. multivorans and Burkholderia stabilis) pre-transplant. Survival among this group was similar to survival among recipients not infected with Burkholderia species, with only one death in the first year [19]. While formal guidelines from the ISHLT do not consider the presence of B. cepacia complex to be an absolute contraindication to lung transplantation [11], most transplant centers do not currently perform lung transplants in CF patients infected with B. cepacia complex [20]. Given the above evidence, we consider B. cenocepacia a strong contraindication to lung transplant. B. gladioli is also a probable contraindication to lung transplantation, although the data are more limited for this subspecies. There is no convincing evidence at this time that other Burkholderia subspecies should be considered contraindications to lung transplantation, although this question warrants continued investigation. Referring centers should consult different transplant centers about transplant eligibility. It is important to note the potential for misidentification of Burkholderia subspecies [21]; given the potential impact of misidentification on prognosis and transplant candidacy, confirmation by a Burkholderia reference lab is of paramount importance. Other bacteria
Two early, single-center studies showed no difference in posttransplant survival among patients whose sputum contained pan-resistant bacteria (primarily Pseudomonas aeruginosa, although other species were also included) compared with sensitive bacteria, although both studies were small [22,23]. informahealthcare.com
Review
A larger retrospective study at two transplant centers assessed post-transplant survival for CF patients based on the presence or absence of pan-resistant bacteria (excluding B. cepacia), defined as bacteria that were resistant to one or more antibiotics from each known class of antibiotics against that bacteria. The vast majority of pan-resistant bacteria were P. aeruginosa. The group with pan-resistant bacteria had higher mortality after lung transplant. There was no statistically significant difference in death from any particular cause, although there was a trend toward increased deaths due to infection. However, mortality in both groups was equivalent to or better than survival among all transplant recipients based on data from UNOS [24]. A single-center retrospective study of CF patients colonized with P. aeruginosa but not B. cepacia compared the use of multiple-combination bactericidal testing (also known as synergy testing) with conventional antimicrobial sensitivity testing in the selection of prophylactic antibiotics immediately after lung transplantation. Of 129 included patients, 2 had pan-resistant P. aeruginosa by conventional testing. The incidence of sepsis was significantly lower among patients treated based on multiple-combination bactericidal testing compared with conventional testing (4 vs 16.5%, p = 0.046), with approximately half of sepsis cases in both groups attributed to P. aeruginosa. There was no difference in mortality at 30 days or 1-year posttransplant [25]. Other organisms identified on pre-transplant cultures, such as Staphylococcus aureus (both methicillin-sensitive and methicillinresistant), Achromobacter xylosoxidans and Stenotrophomonas maltophilia, have not been demonstrated to impact transplant outcomes, although there are less data available for these patients. Mycobacteria
Recent data have shown an increase in Mycobacteria infections among CF patients, primarily consisting of Mycobacterium avium complex (MAC) and Mycobacterium abscessus [26,27], and colonization with Mycobacteria is considered a relative contraindication to lung transplantation [11]. MAC does not appear to worsen outcomes for lung transplant recipients in general [28]. On an international survey of 62 transplant centers regarding experiences with M. abscessus in lung transplant recipients, 50% of centers responded, identifying 17 patients infected with M. abscessus. Only two were infected prior to transplant, both of whom had received more than 12 months of therapy and were not considered actively infected at the time of transplantation. One of these two was transplanted for non-CF bronchiectasis and developed a recurrence of M. abscessus in both breasts, which was successfully treated before she died more than a year later of sepsis associated with pseudomembranous colitis. The other was transplanted for CF and developed a recurrence of pulmonary M. abscessus simultaneous with Pseudomonas and Aspergillus infection; treatment was initiated for all three infections, and his subsequent death was attributed to Aspergillus. Of transplant recipients for whom M. abscessus was first identified post-transplant, only one was not treated due to M. abscessus being considered a colonizer rather than an active infection. 317
Expert Review of Respiratory Medicine Downloaded from informahealthcare.com by Nyu Medical Center on 05/30/14 For personal use only.
Review
Dorgan & Hadjiliadis
Ten responded to therapy and were considered cured, two deaths were attributed to M. abscessus and the remainder had stabilized or died of other causes [29]. More recent studies have examined the role of Mycobacteria specifically for lung transplants in patients with CF. One singlecenter retrospective study compared 13 patients who had met American Thoracic Society criteria for non-tuberculous mycobacteria (NTM) disease at any time prior to lung transplant. Eleven had undergone M. abscessus therapy; of the other two, one had cleared M. abscessus without NTM-directed therapy and the other was considered actively infected at the time of transplant. Three of the treated patients had cleared M. abscessus from their sputum at the time of transplant. Although nine patients had positive M. abscessus cultures immediately prior to transplant, only four were culture-positive for M. abscessus anytime after transplant. One of those subsequently had negative cultures without treatment. The others all had M. abscessus mediastinitis and/or sternal osteomyelitis; cultures cleared in all three with NTM-directed therapy, although one had persistent sternal caseating granulomata. There was no difference in mortality 1, 3 or 5 years after transplant compared with patients who did not have M. abscessus prior to transplant [30]. Another case series assessed outcomes of lung transplantation among CF patients who had met American Thoracic Society criteria for NTM infection with MAC or M. abscessus. Nine transplant recipients were included, two of whom had previously been infected with MAC but had negative mycobacterial cultures at the time of transplant. Five of the seven had active M. abscessus infection at the time of transplant. There were two deaths within 2 months of transplant, neither of which was attributed to NTM (one from primary graft dysfunction, one from invasive Aspergillus without identification of NTM on autopsy), although both were considered to have active M. abscessus disease at the time of transplant [31]. Finally, the literature has multiple case reports or series of patients infected with M. abscessus who did poorly after lung transplantation, in some cases despite treatment [32,33]. Based on the above data, the presence of NTM should not be considered an absolute contraindication to lung transplantation. However, given the potential significant clinical impacts of these organisms, aggressive treatment in an attempt to eradicate them prior to lung transplantation is warranted. Given the difficulty of eradicating M. abscessus and the limited reports of transplantation during active M. abscessus infections, we consider the evidence inadequate to make a clear recommendation on the safety of lung transplantation during an active infection with M. abscessus. Discussion with transplant centers prior to referral is of paramount importance, as treatment of this organism requires significant expertise. Fungi
Invasive infections with Aspergillus fumigatus can cause severe morbidity and mortality after lung transplantation [34–37]. Two older studies assessed the impact of pre-transplant airway colonization with A. fumigatus in CF on post-transplant outcomes. 318
A single-center retrospective study compared outcomes among lung transplant recipients among three groups: CF patients colonized with A. fumigatus, CF patients not colonized with A. fumigatus and patients with transplant indications other than CF, none of whom was colonized. Of note, this study was performed in the era before newer triazole antifungals such as voriconazole and posaconazole. The airways of 53% of CF patients were colonized with A. fumigatus prior to receiving lung transplants. Among CF patients colonized pre-transplant, 59% continued to have positive cultures after transplant; 40% of CF patients whose airways were not colonized with A. fumigatus pre-transplant developed positive cultures post-transplant, and 28% of non-CF patients developed positive cultures for A. fumigatus after transplant. Tracheobronchial aspergillosis developed after transplant in 4 of the 17 CF patients who had cultures positive for A. fumigatus prior to transplant, compared with none of the non-colonized CF patients and 4% of the non-CF patients. Aspergillus pneumonia developed in 6% of the non-CF patients, but none of the CF patients regardless of pre-transplant cultures. No CF patients colonized with Aspergillus prior to transplant died in the first year after transplant. There was no increase in anastomotic complication associated with pre-transplant Aspergillus colonization [34]. These findings were similar to the results of an earlier single-center study of the impact of pre-transplant Aspergillus colonization on transplant outcomes in CF [35]. Two more recent studies have assessed the impact of Aspergillus infection in lung transplant recipients [36,37]. In the first study, a large series of CF lung transplant recipients was assessed and 70% (65/93) had pre-transplant colonization. Invasive aspergillosis was present in 22.5% of patients and positive intraoperative cultures for Aspergillus (odds ratio [OR]: 4.36) and treatment for acute rejection in the first 90 days (OR: 3.53) were risk factors for its development after multivariable analysis; however, it was not associated with increased mortality [36]. The second study assessed patients with invasive aspergillosis (not limited to CF) and it found that patients with nodular lesions were more likely to have a successful outcome [37]. Aside from the possible need to be more vigilant for tracheobronchial aspergillosis after transplant, the presence of Aspergillus in sputum cultures prior to transplant does not appear to be a predictor of transplant outcomes in CF. In contrast to the impact of airway colonization with Aspergillus, the presence of active fungal infection in the form of mycetomas prior to lung transplant has been associated with increased post-transplant mortality [38]. None of the patients in that study had CF. Nutrition
Among all lung transplant recipients, obesity has been associated with decreased survival after transplantation, while studies have shown conflicting results regarding an association between underweight status (BMI
Review
Lung transplantation in patients with cystic fibrosis: special focus to infection and comorbidities Expert Rev. Respir. Med. 8(3), 315–326 (2014)
Daniel J Dorgan and Denis Hadjiliadis* Division of Pulmonary, Allergy and Critical Care, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA 19104, USA *Author for correspondence: Tel.: +1 215 662 4202 Fax: +1 215 614 0869 [email protected]
Despite advances in medical care, patients with cystic fibrosis still face limited life expectancy. The most common cause of death remains respiratory failure. End-stage cystic fibrosis can be treated with lung transplantation and is the third most common reason for which the procedure is performed. Outcomes for cystic fibrosis are better than most other lung diseases, but remain limited (5-year survival 60%). For patients with advanced disease lung transplantation appears to improve survival. Outcomes for patients with Burkholderia cepacia remain poor, although they are better for patients with certain genomovars. Controversy exists about Mycobacterium abscessus infection and appropriateness for transplant. More information is also becoming available for comorbidities, including diabetes and pulmonary hypertension among others. Extra-corporeal membrane oxygenation is used more frequently for end-stage disease as a bridge to lung transplantation and will likely be used more in the future. KEYWORDS: Burkholderia cepacia • cystic fibrosis • cystic fibrosis-related diabetes • extracorporeal membrane oxygenation • lung transplantation • Mycobacterium abscessus
Cystic fibrosis (CF) is the most common lifelimiting genetic disease among Caucasians [1]. The primary cause of death is progressive lung disease due to recurrent infections and destruction of pulmonary tissue. Lung transplantation is often performed in CF patients with endstage lung disease. CF is the third most common indication for lung transplantation, accounting for 16.7% of all lung transplants and 26.3% of bilateral lung transplants performed from January 1995 to June 2011 [2]. Based on registry data from the International Society for Heart and Lung Transplantation (ISHLT), survival after lung transplantation is greater for CF than for other diagnoses, although this comparison is not adjusted for recipient age, bilateral versus single lung transplant or other clinical factors. Overall, median survival after lung transplantation is 7.5 years for CF, compared with 5.5 years for all lung transplants and 6.7 years for all bilateral lung transplants. Combining all diagnoses, posttransplant survival was 88% at 3 months, 79% at 1 year, 64% at 3 years, 53% at 5 years and 30% at 10 years as of the 2012 ISHLT registry informahealthcare.com
10.1586/17476348.2014.899906
report. For CF, post-transplant survival was 90% at 3 months, 83% at 1 year, 69% at 3 years, 60% at 5 years and 43% at 10 years [2]. This article reviews factors that have been studied as potential predictors of outcomes in lung transplantation for CF. However, given prior controversy regarding survival benefit with lung transplantation in CF, we begin with a discussion of studies that have addressed the impact of transplant on survival in patients with advanced lung disease due to CF. Survival benefit of lung transplantation in CF Early studies
Several studies have cast doubt on whether lung transplantation confers a survival benefit for patients with CF, with results suggesting that lung transplantation does not improve survival among CF patients with end-stage lung disease [3–5]. One criticism of these studies is that they model survival using Cox proportional hazards with lung transplantation as the only time-dependent variable [6]. By including pre-transplant clinical data from
Ó 2014 Informa UK Ltd
ISSN 1747-6348
315
Expert Review of Respiratory Medicine Downloaded from informahealthcare.com by Nyu Medical Center on 05/30/14 For personal use only.
Review
Dorgan & Hadjiliadis
only a single point in time, the models could underestimate the increasing mortality risk over time due to progression of disease without lung transplantation, thereby creating bias when comparing the risk of mortality without transplant with the risk of mortality after lung transplantation. Perhaps more important, these studies were performed when priority for allocation of lungs for transplantation was based on time on the waiting list rather than on clinical characteristics, which could impact the potential for benefit or harm from lung transplantation. That is, an allocation system based on time on the waiting list preferentially provides transplants to those who are able to remain stable long enough to receive a transplant (and therefore may have a lower instantaneous risk of death without a transplant), while rapidly declining patients who potentially could derive the most benefit from transplantation may not have been able to accrue adequate time on the waiting list to receive a transplant before dying. Older studies that have been done using a priority system for lung allocation have found a survival advantage with lung transplantation in CF [7]. On the other hand, such studies are also biased because survivors to transplant have inherent differences from patients who die; these differences likely cannot all be controlled for, thereby favoring transplant. The controversy was even greater for adolescents with CF when a well publicized study suggested that lung transplant might harm the majority of that group [4]. Aside from the factors noted above, one criticism of this paper was the inclusion of data from all eras, leading to potential bias as lung transplantation outcomes were worse in earlier eras. Post-transplant survival under the lung allocation score
The Lung Allocation Score (LAS) system was implemented in the USA in May 2005, replacing the system in which priority for lung transplant was based on how long an individual had remained on the waiting list. The LAS uses clinical data to predict 1-year survival both with and without lung transplantation, and normalizes the results on a scale of 0–100, with higher scores leading to higher priority for lung transplantation. Clinical factors that increase 1-year mortality without lung transplantation or increase the likelihood of 1-year survival after lung transplantation increase an individual’s LAS [8]. A recent study used a different approach in comparing the risk of death with or without lung transplantation in CF, incorporating LAS into the statistical model as a surrogate for disease severity. Using data from the United Network for Organ Sharing (UNOS), this study included multiple models to assess the impact of lung transplantation on survival. One of the Cox proportional hazard models used lung transplant as the only time-dependent covariate, incorporating other factors as time-independent (using baseline values for those variables), which was a similar approach to prior studies. A separate analysis used a joint model for longitudinal and survival data, a newer approach that incorporates mixed-effects modeling to describe changes of a covariate over time [9], allowing the variable (in this case LAS) to be included as a time-dependent covariate in the Cox model. This model included LAS from 316
the longitudinal (time-variant) model, lung transplant as a time-dependent covariate and an interaction term between the two. Models were stratified by transplant center. Of 704 patients in the dataset, 67% underwent lung transplantation, 14% died on the waiting list and 19% remained on the waiting list. Consistent with registry data cited above, posttransplant survival in CF was greater than post-transplant survival for all diagnoses: 96.5% at 3 months, 88.4% at 1 year and 67.8% at 3 years. Using traditional modeling with lung transplant as the only time-dependent covariate, there was a survival benefit with lung transplantation both for the unadjusted model and on multivariable analysis (suggesting that negative results in prior studies may have been due to the allocation system prior to LAS, rather than the statistical methods used). There was an interaction between lung transplant and LAS at listing such that lung transplantation was found to confer a survival advantage when LAS was greater than 30. The joint model incorporating both lung transplantation and LAS as time-dependent covariates similarly found an interaction between transplant and LAS, with higher LAS conferring a greater survival benefit with lung transplantation. These recent data strongly support lung transplantation as a means to improve survival in the most severely ill CF patients under the LAS system [10]. Microbiology
ISHLT guidelines list overt sepsis as an absolute contraindication to lung transplantation. Colonization with highly resistant or highly virulent bacteria’ is considered a relative contraindication, without further specification regarding particular organisms or antibiotic-resistance pattern as an absolute contraindication in and of itself [11], although the guidelines do note potential increased risk associated with certain pathogens in CF. In this section, we discuss data regarding lung transplant outcomes based on sputum microbiology prior to transplantation. Burkholderia cepacia complex
Burkholderia cepacia complex has been associated with worse outcomes in patients with CF, irrespective of lung transplantation [12]. This is even more significant when patients are colonized with more pathogenic strains like Burkholderia cenocepacia ET12 [13]. Sputum colonization with B. cepacia complex in CF patients prior to lung transplantation has been associated with increased mortality after transplant in several studies. One single-center study in Toronto showed that the most common cause of early post-transplant mortality among B. cepacia-positive patients was widespread pneumonia, including areas of lung necrosis and cavitation in some patients. The genomovar or subspecies was not specified for any patients in this retrospective study [14]. Another single-center retrospective study also found increased post-transplant mortality among patients infected preoperatively with B. cepacia complex, but found excess mortality only among those infected with B. cepacia genomovar III (now known as B. cenocepacia) [15]. Other retrospective studies at individual centers have shown similar results [16,17]. Expert Rev. Respir. Med. 8(3), (2014)
Expert Review of Respiratory Medicine Downloaded from informahealthcare.com by Nyu Medical Center on 05/30/14 For personal use only.
Lung transplantation in patients with CF
A subsequent study coordinated by the B. cepacia Research Laboratory and Repository assessed outcomes associated with different Burkholderia subspecies compared with absence of known Burkholderia among CF patients aged 12 and older listed for lung transplant on the Scientific Registry of Transplant Recipients. There was a significant increase in adjusted post-transplant mortality among patients infected with Burkholderia gladioli and non-epidemic strains of B. cenocepacia. There was no difference in mortality among those infected with Burkholderia multivorans, epidemic strains of B. cenocepacia or other B. cepacia complex species, although caution should be exercised when interpreting these results given the small sample size for each subspecies of Burkholderia [18]. In addition, in this group B. cenocepacia ET12 was very uncommon. A study at another transplant center employed a similar post-transplant antibiotic and immunosuppression protocol in transplant recipients colonized with B. cepacia complex as the regimen described above in the Toronto study. This study did identify the B. cepacia subspecies for each patient. Recipients infected with B. cenocepacia prior to transplant demonstrated excess mortality despite the modified post-transplant treatment regimen, with two of the eight patients dying within 30 days and five dying within 1 year; all five deaths occurred in the context of B. cenocepacia sepsis, often with the presence of empyema. Ten patients were infected with other Burkholderia subspecies (four with Burkholderia vietnamiensis, five with B. multivorans and one with both B. multivorans and Burkholderia stabilis) pre-transplant. Survival among this group was similar to survival among recipients not infected with Burkholderia species, with only one death in the first year [19]. While formal guidelines from the ISHLT do not consider the presence of B. cepacia complex to be an absolute contraindication to lung transplantation [11], most transplant centers do not currently perform lung transplants in CF patients infected with B. cepacia complex [20]. Given the above evidence, we consider B. cenocepacia a strong contraindication to lung transplant. B. gladioli is also a probable contraindication to lung transplantation, although the data are more limited for this subspecies. There is no convincing evidence at this time that other Burkholderia subspecies should be considered contraindications to lung transplantation, although this question warrants continued investigation. Referring centers should consult different transplant centers about transplant eligibility. It is important to note the potential for misidentification of Burkholderia subspecies [21]; given the potential impact of misidentification on prognosis and transplant candidacy, confirmation by a Burkholderia reference lab is of paramount importance. Other bacteria
Two early, single-center studies showed no difference in posttransplant survival among patients whose sputum contained pan-resistant bacteria (primarily Pseudomonas aeruginosa, although other species were also included) compared with sensitive bacteria, although both studies were small [22,23]. informahealthcare.com
Review
A larger retrospective study at two transplant centers assessed post-transplant survival for CF patients based on the presence or absence of pan-resistant bacteria (excluding B. cepacia), defined as bacteria that were resistant to one or more antibiotics from each known class of antibiotics against that bacteria. The vast majority of pan-resistant bacteria were P. aeruginosa. The group with pan-resistant bacteria had higher mortality after lung transplant. There was no statistically significant difference in death from any particular cause, although there was a trend toward increased deaths due to infection. However, mortality in both groups was equivalent to or better than survival among all transplant recipients based on data from UNOS [24]. A single-center retrospective study of CF patients colonized with P. aeruginosa but not B. cepacia compared the use of multiple-combination bactericidal testing (also known as synergy testing) with conventional antimicrobial sensitivity testing in the selection of prophylactic antibiotics immediately after lung transplantation. Of 129 included patients, 2 had pan-resistant P. aeruginosa by conventional testing. The incidence of sepsis was significantly lower among patients treated based on multiple-combination bactericidal testing compared with conventional testing (4 vs 16.5%, p = 0.046), with approximately half of sepsis cases in both groups attributed to P. aeruginosa. There was no difference in mortality at 30 days or 1-year posttransplant [25]. Other organisms identified on pre-transplant cultures, such as Staphylococcus aureus (both methicillin-sensitive and methicillinresistant), Achromobacter xylosoxidans and Stenotrophomonas maltophilia, have not been demonstrated to impact transplant outcomes, although there are less data available for these patients. Mycobacteria
Recent data have shown an increase in Mycobacteria infections among CF patients, primarily consisting of Mycobacterium avium complex (MAC) and Mycobacterium abscessus [26,27], and colonization with Mycobacteria is considered a relative contraindication to lung transplantation [11]. MAC does not appear to worsen outcomes for lung transplant recipients in general [28]. On an international survey of 62 transplant centers regarding experiences with M. abscessus in lung transplant recipients, 50% of centers responded, identifying 17 patients infected with M. abscessus. Only two were infected prior to transplant, both of whom had received more than 12 months of therapy and were not considered actively infected at the time of transplantation. One of these two was transplanted for non-CF bronchiectasis and developed a recurrence of M. abscessus in both breasts, which was successfully treated before she died more than a year later of sepsis associated with pseudomembranous colitis. The other was transplanted for CF and developed a recurrence of pulmonary M. abscessus simultaneous with Pseudomonas and Aspergillus infection; treatment was initiated for all three infections, and his subsequent death was attributed to Aspergillus. Of transplant recipients for whom M. abscessus was first identified post-transplant, only one was not treated due to M. abscessus being considered a colonizer rather than an active infection. 317
Expert Review of Respiratory Medicine Downloaded from informahealthcare.com by Nyu Medical Center on 05/30/14 For personal use only.
Review
Dorgan & Hadjiliadis
Ten responded to therapy and were considered cured, two deaths were attributed to M. abscessus and the remainder had stabilized or died of other causes [29]. More recent studies have examined the role of Mycobacteria specifically for lung transplants in patients with CF. One singlecenter retrospective study compared 13 patients who had met American Thoracic Society criteria for non-tuberculous mycobacteria (NTM) disease at any time prior to lung transplant. Eleven had undergone M. abscessus therapy; of the other two, one had cleared M. abscessus without NTM-directed therapy and the other was considered actively infected at the time of transplant. Three of the treated patients had cleared M. abscessus from their sputum at the time of transplant. Although nine patients had positive M. abscessus cultures immediately prior to transplant, only four were culture-positive for M. abscessus anytime after transplant. One of those subsequently had negative cultures without treatment. The others all had M. abscessus mediastinitis and/or sternal osteomyelitis; cultures cleared in all three with NTM-directed therapy, although one had persistent sternal caseating granulomata. There was no difference in mortality 1, 3 or 5 years after transplant compared with patients who did not have M. abscessus prior to transplant [30]. Another case series assessed outcomes of lung transplantation among CF patients who had met American Thoracic Society criteria for NTM infection with MAC or M. abscessus. Nine transplant recipients were included, two of whom had previously been infected with MAC but had negative mycobacterial cultures at the time of transplant. Five of the seven had active M. abscessus infection at the time of transplant. There were two deaths within 2 months of transplant, neither of which was attributed to NTM (one from primary graft dysfunction, one from invasive Aspergillus without identification of NTM on autopsy), although both were considered to have active M. abscessus disease at the time of transplant [31]. Finally, the literature has multiple case reports or series of patients infected with M. abscessus who did poorly after lung transplantation, in some cases despite treatment [32,33]. Based on the above data, the presence of NTM should not be considered an absolute contraindication to lung transplantation. However, given the potential significant clinical impacts of these organisms, aggressive treatment in an attempt to eradicate them prior to lung transplantation is warranted. Given the difficulty of eradicating M. abscessus and the limited reports of transplantation during active M. abscessus infections, we consider the evidence inadequate to make a clear recommendation on the safety of lung transplantation during an active infection with M. abscessus. Discussion with transplant centers prior to referral is of paramount importance, as treatment of this organism requires significant expertise. Fungi
Invasive infections with Aspergillus fumigatus can cause severe morbidity and mortality after lung transplantation [34–37]. Two older studies assessed the impact of pre-transplant airway colonization with A. fumigatus in CF on post-transplant outcomes. 318
A single-center retrospective study compared outcomes among lung transplant recipients among three groups: CF patients colonized with A. fumigatus, CF patients not colonized with A. fumigatus and patients with transplant indications other than CF, none of whom was colonized. Of note, this study was performed in the era before newer triazole antifungals such as voriconazole and posaconazole. The airways of 53% of CF patients were colonized with A. fumigatus prior to receiving lung transplants. Among CF patients colonized pre-transplant, 59% continued to have positive cultures after transplant; 40% of CF patients whose airways were not colonized with A. fumigatus pre-transplant developed positive cultures post-transplant, and 28% of non-CF patients developed positive cultures for A. fumigatus after transplant. Tracheobronchial aspergillosis developed after transplant in 4 of the 17 CF patients who had cultures positive for A. fumigatus prior to transplant, compared with none of the non-colonized CF patients and 4% of the non-CF patients. Aspergillus pneumonia developed in 6% of the non-CF patients, but none of the CF patients regardless of pre-transplant cultures. No CF patients colonized with Aspergillus prior to transplant died in the first year after transplant. There was no increase in anastomotic complication associated with pre-transplant Aspergillus colonization [34]. These findings were similar to the results of an earlier single-center study of the impact of pre-transplant Aspergillus colonization on transplant outcomes in CF [35]. Two more recent studies have assessed the impact of Aspergillus infection in lung transplant recipients [36,37]. In the first study, a large series of CF lung transplant recipients was assessed and 70% (65/93) had pre-transplant colonization. Invasive aspergillosis was present in 22.5% of patients and positive intraoperative cultures for Aspergillus (odds ratio [OR]: 4.36) and treatment for acute rejection in the first 90 days (OR: 3.53) were risk factors for its development after multivariable analysis; however, it was not associated with increased mortality [36]. The second study assessed patients with invasive aspergillosis (not limited to CF) and it found that patients with nodular lesions were more likely to have a successful outcome [37]. Aside from the possible need to be more vigilant for tracheobronchial aspergillosis after transplant, the presence of Aspergillus in sputum cultures prior to transplant does not appear to be a predictor of transplant outcomes in CF. In contrast to the impact of airway colonization with Aspergillus, the presence of active fungal infection in the form of mycetomas prior to lung transplant has been associated with increased post-transplant mortality [38]. None of the patients in that study had CF. Nutrition
Among all lung transplant recipients, obesity has been associated with decreased survival after transplantation, while studies have shown conflicting results regarding an association between underweight status (BMI
