Virus infection facilitates the development of severe pneumonia in transplant patients with hematologic malignancies

Allogeneic hematopoietic stem cell transplantation (HSCT) is an effective therapy for patients with hematologic malignancies. Severe pneumonia is associated with high mortality rate in HSCT recipients. Viral co-infection indicates a poor prognosis of HSCT recipients. In this study, a total of 68 allogeneic HSCT recipients were included. Cytomegalovirus (CMV) and Respiratory syncytial virus (RSV) infection was assessed by testing peripheral blood and oropharynx swabs, respectively, collected in the first 180 days after transplantation. We analysed the correlation of CMV and RSV co-infection with severe pneumonia and mortality. The incidence of CMV and RSV co-infection was 26.5% (18/68). Severe pneumonia was diagnosed in 61% (11/18) cases with co-infection compared to only 10% (5/50) cases with mono-infection or no infection. The analysis of potential risk factors for severe pneumonia showed that CMV and RSV co-infection was significantly associated with severe pneumonia (p < 0.001). The 5 patients who died of severe pneumonia were all co-infected with CMV and RSV. In conclusion, CMV and RSV co-infection appears to be an important factor and facilitates the development of severe pneumonia in allogeneic HSCT patients with hematologic malignancies.


INTRODUCTION
Hematopoietic stem cell transplantation (HSCT) is primarily used in the treatment of hematological malignancies [1]. Despite the improved outcomes of HSCT, patients remain at high risk for potentially fatal infections due to immunosuppression after engraftment. Pneumonia is a frequent infectious complication, and development of severe pneumonia is a common cause of mortality after allogeneic HSCT [2,3]. Although many microorganisms are associated with pneumonia, viruses are recognised as the predominant pathogens leading to pneumonia after allogeneic HSCT [4,5].
Cytomegalovirus (CMV) is a ubiquitous herpes virus and the prevalence of CMV infection (anti-CMV IgG positive) in Chinese blood donors is 98.5% [6]. CMV seropositivity and/or documented reactivation have been associated with morbidity and mortality after HSCT [7]. The majority of patients are at risk for CMV reactivation after HSCT, which results in multi-organ inflammation and an increased rate of graft-versus-host disease (GVHD) [8,9]. Additionally, recent studies have identified other viruses, such as respiratory syncytial virus (RSV), Parainfluenza virus (PIV), Influenza virus, and human Metapneumovirus (MPV), as important factors contributing to pneumonia in HSCT recipients [10]. Respiratory syncytial virus (RSV) is a common cause of seasonal respiratory viral infections in patients who have undergone HSCT [11]. In susceptible young children and immunosuppressed elderly individuals, RSV can exploit host immunity and cause a strong Clinical Research Paper www.impactjournals.com/oncotarget inflammatory response that leads to lung damage and viral dissemination [12][13][14]. In immunocompromised hosts, RSV infection usually manifests as an upper respiratory illness (URI) that frequently progresses to a lower respiratory illness [15,16].
Viral co-infections have been noted to increase risks of mortality in HSCT recipients [17,18]. Previous studies have investigated the risk factors for either RSV or CMV infection [7,[19][20][21][22], but none of them examined the risk factors and outcomes of CMV and RSV co-infection in HSCT patients. This prospective study was conducted to investigate the presence of two viruses associated with the development of severe pneumonia among 68 allogeneic HSCT patients. The results indicate that CMV and RSV co-infection facilitated the development of severe pneumonia in allogeneic hematopoietic stem cell transplant patients with underlying hematologic malignancies.

Patients' preliminary characteristics
The patients' preliminary characteristics are shown in Table 1

Incidence of CMV and RSV mono-infection and co-infection
The CMV serology status of the recipients and donors was detected before HSCT. Positive for CMV-IgG and negative for CMV-IgM were confirmed for all of recipients and donors before HSCT. A total of 437 peripheral blood samples were collected and tested for CMV, a mean of 6.4 (4-14) samples per patient. A total of 68 patients had 131 episodes of CMV reactivation during the study period, with a mean of 1.9 (0-8) episodes per patient. A total of 437 oropharynx swabs were collected and tested for RSV, with a mean of 6.4 (3-15) per patient. The 68 patients had 88 episodes of RSV infection during the study period (mean 1.3 episodes, ranging from 0-5 episodes per patient).
The incidence of CMV and RSV co-infection was 26.5% (18/68) in patients with allogeneic HSCT. The 18 co-infected patients had only one episode of CMV and RSV co-infection during the course of the study. The incidences of either CMV or RSV mono-infection were 48.5% (33/68) in allogeneic HSCT patients. CMV mono-infection was identified in 29.4% (20/68) of allogeneic HSCT patients. RSV mono-infection was identified in 39.7% (27/68) of allogeneic HSCT patients. Neither CMV nor RSV infection was tested in the rest of the 17 allogeneic HSCT patients. Patients who had co-infection, mono-infection, and no infection shared common characteristics including age, gender, underlying diseases, donor type, and stem cell source (p = 0.095, p = 0.258, p = 0.95, p = 0.283, p = 0.496). However, the three groups differed significantly in the presence of acute GVHD grade 3-4 or chronic GVHD (p < 0.001, Table 2 ).

Epidemiology of CMV or RSV infection
The median duration of viral shedding during RSV infections was 21 days (range: 7-60 days), while the median duration of CMV infection was 60 days (range: 14-160 days) (p < 0.001) ( Figure 1A). In this cohort, RSV infection occurred earlier than CMV infection after HSCT (p = 0.005). Figure 1B shows that CMV infection rate steadily increased throughout the study period, while RSV infection rate was at a medium level at 0-28 days after HSCT, then decreased and occasionally increased for the remainder of the study period. The seasonal characteristics of CMV and RSV infection were further analysed. As shown in Figure 1C, the incidence of RSV infection was higher in winter and spring (p = 0.029). CMV infection rate was evenly distributed throughout the seasons of 2012.

DISCUSSION
Severe pneumonia leads more frequently to organ failure and death than other infections. In this study, the correlation of CMV/RSV co-infection and severe pneumonia was investigated in 68 allogeneic HSCT patients with hematologic malignancies. We observed a high incidence of severe pneumonia in the co-infection group (11/18) and a lower incidence in patients without co-infection (5/50). CMV and RSV co-infection was significantly associated with severe pneumonia and prognosis of allogeneic HSCT patients with hematologic malignancies. To our knowledge, RSV and CMV are very different types of virus. RSV is a single-stranded RNA virus of the family Paramyxoviridae and well known to be seasonal [23]. It is a major cause of lower respiratory tract infections during infancy and childhood. CMV, a DNA virus of the Herpesviridae family [24], is one of the main agents involved in infectious complications after transplantation [25]. The occurrence of RSV infection was earlier than that of CMV, and RSV infections had a shorter duration of viral shedding after HSCT than CMV infection. It's well known that virus infection experiences three processes, including latent infection, active infection and disease [26]. The disease caused by viruses is due to transmission from the transplanted organ, reactivation of latent infection, or after a primary infection in seronegative transplant patients [27]. Simple RSV or CMV infection is mostly in the active period   without signs and symptoms of the disease. However, coinfection will prolong defects in cellular immune function and accelerate the pathological process of infection with increasing the chemokine level and neutropenia [28,29]. Moreover, the signaling pathways associated to innate and acquired immune system will be stimulated by co-infection of CMV and RSV. The innate cellular response to the initial stages of infection is mediated by Toll like receptor (TLR2, TLR4, TLR9) signaling, these receptors have been shown to recognize the surface proteins of CMV and RSV [30][31][32].
Prior studies have focussed on the importance of viral co-infection. Co-infection decreased the patient's response to antiviral treatment and resulted in more severe infection and a greater risk of death. CMV/BKV coinfection (7/8; 87.5%) is the major cause of death among patients with viral co-infection [40]. Furthermore, coinfection with CMV/EBV is associated with poor survival and a greater risk of post-transplantation complications [41]. Virus co-infection induced bronchial epithelial damage, loss of clearance, and pathogen invasion.
Recent advances in diagnostic techniques such as CMV antigenaemia quantification and PCR enable the early detection of CMV reactivation and prevent the subsequent development of CMV diseases [42].
Unfortunately, there is no strategy for addressing RSV associated complications. HSCT patients, especially those with co-infection, should be closely monitored to prevent the development of severe pneumonia. Prophylactic Acyclovir could not prevent co-infection with CMV and RSV because of the dose of Acyclovir, patient sensitivity, and viral variation. Further study is required to establish effective anti-viral prophylactic treatment after allogeneic HSCT.
Pneumonia is caused by more than one pathogens (e.g. bacterial and viral) and co-existence in HSCT patients. Bacterial infections continue to be one of the most frequent complications after HSCT [43]. We also adopted for the investigation of bacterial infection in the patients. In univariate analysis, we found that bacterial infection was associated with severe pneumonia, but was not significant risk factor in multivariable analysis.
This study had several potential limitations. First, the number of patients was small because it was difficult to enrol large number of patients from a single transplant centre. Second, T-cell depletion, antilymphocyte globulin, unrelated or HLA mismatched donors, and GVHD are known to be risk factors for severe viral infections after HSCT [26,44,45]. In this study, T-cell depletion was not assessed and antilymphocyte globulin assays were not performed. Finally, PCR assays were performed on oropharynx swabs not bronchoalveolar lavage fluid. This limited the frequency with which the quantitative PCR tests could be performed. Nonetheless, these data provided new information about the risk factor of CMV/RSV co-infection and the correlation between CMV/RSV co-infection and severe pneumonia in allogeneic HSCT recipients.
In conclusion, this study found that co-infection with CMV and RSV facilitated the development of severe pneumonia in allogeneic HSCT patients with underlying hematologic malignancies. These results indicate that further investigation into viral co-infection following HSCT is necessary and helpful to predict the incidence of severe pneumonia and prognosis.

MATERIALS AND METHODS Patients
A total of 437 oropharynx swabs and 437 peripheral blood samples were collected from 68 recipients who underwent allogeneic HSCT at the first affiliated hospital, Sun Yat-sen University, Guangzhou, China, from March 2011 to February 2013 (preliminary data for patients are shown in Table 1 ). Samples were tested regardless of the presence of symptoms of respiratory infection. Most of allogeneic HSCT patients were monitored every two weeks for six months (once a week in the first month). The CMV serology status of the donor was tested before HSCT. The donors were all positive for CMV-IgG and negative for CMV-IgM. Clinical and laboratory data were abstracted from the existing electronic medical records. Bacterial infections were detected in blood, tracheal secretions or broncho-alveolar lavage.

CMV-IgG and CMV-IgM detection
The assays were performed using the Abbott Laboratories Microparticle enzyme immunoassay on an Architech C16000. Briefly, the microparticle bottle was mixed to suspend the microparticles. The minimum sample cup volume was calculated and detected. The

Quantification of CMV DNA copy number
The detection of CMV DNA was performed using the Diagnostic Kit for Quantification of Human Cytomegalovirus DNA (Daangene, Guangzhou, China). Peripheral blood leukocyte DNA was extracted and added to the CMV-PCR reaction tube. The negative control and positive quantification controls were installed. Quantitative-PCR was performed in an ABI Prism 7000 machine using the following protocol: 2 min at 93°C, 10 cycles of 45 s at 93°C, 60 s at 55°C, and 30 cycles of 30 s at 93°C and 45 s at 55°C.

Virus prophylaxis
Ganciclovir (5 mg/kg i.v. q12h daily) was given prophylactically to all HSCT patients from the day that conditioning chemotherapy was initiated until day 1 after allogeneic HSCT, followed by Acyclovir (200 mg P.O. 3 times daily) for 6 months. If the CMV viral load exceeded 500 copies/mL or the CMV IgM was ≥1.0, indicating CMV reactivation, Ganciclovir (5 mg/kg i.v. q12h daily) was used instead of Acyclovir until the patient's parameters improved (CMV viral load <500 copies/mL or CMV IgM <1.0). Allogeneic HSCT patients received intravenous immunoglobulin (0.2 g/kg every month) for 6 months.

Criteria for analysis and definitions
Mono-infection was defined by the detection of CMV reactivation or detection of RSV in the period lasting more than two weeks. Co-infection was defined by the detction of both CMV reactivation and RSV infection in two weeks. The definitions for severe pneumonia (severe community-acquired pneumonia) included five "minor" criteria and four "major" criteria. The minor criteria included respiratory rate >30/min, bilateral pneumonia or multilobar pneumonia, systolic BP <90 mm Hg, and diastolic BP <60 mm Hg. The major criteria included a need for mechanical ventilation, an increase in the size of infiltrates by ≥50% within 48 h, either septic shock or the need for pressors for >4 h, and acute renal failure (either urine output <80 ml in 4 h or serum creatinine levels >2 mg/dl in the absence of chronic renal failure) [46,47].

Statistical analysis
To assess associations between continuous variables and the outcome variable, Mann-Whitney U test or ANOVA was used. The Chi-square test or Fisher's exact test was used to analyse categorical data, as appropriate. Logistic regression models were used to estimate the associations of potential risk factors for severe pneumonia. Kaplan-Meier was used for survival analysis, and the differences in survival probabilities between patient subsets were assessed by the log-rank test. Two-sided p values of less than 0.05 were considered to be statistically significant. Statistical analyses were performed using SPSS.20.0 software (SPSS, Inc, Chicago, IL).