Clinical characteristics and prognosis of communityacquired pneumonia in autoimmune diseaseinduced immunocompromised host: A retrospective observational study

Zhong-shu Kuang, Yi-lin Yang, Wei Wei, Jian-li Wang, Xiang-yu Long, Ke-yong Li, Chao-yang Tong, Zhan Sun, Zhen-ju Song

1 Department of Emergency Medicine, Zhongshan Hospital, Fudan University, Shanghai 200032, China

2 Department of Pharmacology, University of Virginia School of Medicine Charlottesville, Virginia, USA

KEY WORDS: Community-acquired pneumonia; Immunocompromised hosts; Autoimmune disease; Prognostic marker

I NTRODUCTION

Autoimmune diseases (AIDs) are a group of chronic diseases that arise from an inappropriate immune response against self-antigens, contributing to the loss of tolerance towards native antigens and subsequently to the development of autoantibodies, which are eventually responsible for tissue inflammation and damage.[1,2]Currently, suppression of the pathologic autoimmune response is the main therapeutic option for autoimmune disorders, including systemic immunosuppression or specific, targeted and local immunosuppression.[2]Treatments with traditional management consisting of glucocorticoids and immunosuppressive drugs, which are commonly known as disease-modifying anti-rheumatological drugs (DMARDs), to biological therapies such as tumor necrosis factor-alpha (TNF-α) inhibitors, may provide symptomatic relief and decelerate the progression of the disease.[3,4]These therapeutic measures prevent appropriate responses to infections and environmental antigens at the same time, thereby decreasing both cellular and humoral immunity and leading to an immunocompromised state, which carries an inherent risk of patient susceptibility to opportunistic infections.[5]

The infection is mild and transient for most AID patients; however, serious infection in some patients leading to hospitalization often results in high mortality, usually due to common bacterial, fungal and viral pathogens.[6]Recently, a multicenter, double-blind, randomized trial in IgA nephropathy patients found that corticosteroid therapy is associated with high rates of serious adverse outcomes, mainly due to an increased risk of serious infection.[7]Moreover, in systemic lupus erythematosus (SLE) patients, infections contributed to a high mortality rate of 20%-55%.[8]In addition, rheumatoid arthritis (RA) patients have been found to carry a 2-fold higher risk for acquiring hospital-related infection than the general population.[6]The increased susceptibility to infection in these immunocompromised hosts (ICHs) is believed to be due to a dysfunctional immune system and continuous exposure to the environment. Moreover, pulmonary infection is one of the most common types of infection in ICH patients, followed by systemic and urinary tract infections.[9,10]

A population-based study of community-acquired pneumonia (CAP) in older adults showed that the incidence rate of CAP among immunocompromised patients is almost 3-fold higher than that in immunocompetent subjects.[11]Furthermore, CAP patients with ICH had higher mortality and morbidity. However, ICH patients were systematically excluded from previous large-sample CAP studies because of high mortality and poor prognosis.[12]Some guidelines, such as those of the Infectious Diseases Society of America and the American Thoracic Society on the management of CAP, did not even include recommendations for immunocompromised patients.[13]Only a few studies elaborated the incidence, causative organisms, clinical characteristics and outcome of CAP in ICH patients.[12,14]Moreover, few specific measures and biomarkers have been used to assess the prognosis of CAP with ICH, especially in AID-induced ICH patients. Therefore, to further describe the condition of CAP in AID-induced ICH, we designed this retrospective study to investigate the clinical characteristics, risk factors and predictors of outcome in CAP patients with AID-induced ICH.

METHODS

Participants and setting

Heterogeneity between ICHs induced by different disease backgrounds was noted, showing different clinical features, treatments and outcomes. To reduce the clinical heterogeneity, only CAP patients with AIDinduced ICHs were enrolled in this study. From January 2013 to June 2018, a total of 312 patients (＞18 years of age) with a diagnosis of CAP underlying ICH were admitted to the Emergency Department of Zhongshan Hospital, Fudan University (Shanghai, China). Among them, 218 patients were excluded because of HIV infection, organ transplant, or current chemotherapy or radiotherapy due to neoplastic disease. Finally, 94 patients with AID-induced ICH were enrolled in this retrospective observational study. Clinical parameters, laboratory and radiological data and features were collected. The risk stratifi cation of subjects was evaluated according to the Pneumonia Severity Index (PSI) score, including 20 clinical variables to define five severity classes. PSIs of class IV and V were used to define severe pneumonia.[15]The study was approved by the Institutional Research Ethics Committee of Zhongshan Hospital, Fudan University (2006-23). Informed consent was waived because of the retrospective nature of this study.

Def nitions

AID-induced ICH was considered if a patient received corticosteroids in daily doses ＞10 mg prednisoloneequivalents or received immunosuppressive agents, including disease-modifying drugs and biologics for longer than 30 days.[14]The causes of AIDs are shown in Table 1. CAP was defined as a new infiltrate on chest X-ray and one or more of the following symptoms or signs of acute lower respiratory tract infection: cough, chest pain, fever ＞38 °C, temperature ＜35 °C, and dyspnea within the previous 24 hours.[13]Exclusion criteria included: (1) HIV patients; (2) patients undergoing organ transplant; (3) patients with neoplastic disease undergoing chemotherapy or radiotherapy; (4) nursing home residents; (5) patients with nosocomial pneumonia (onset more than 48 hours after hospital admission); and (6) patients with a fi nal diagnosis of AID-induced pneumonia infi ltration.

Data collection

The Electronic Medical Record System (EMRS) and Computerized Physician Order Entry (CPOE) were screened for available data. The following data were obtained for each patient: neutrophil-to-lymphocyte ratio (NLR), pH-value, partial pressure of arterial oxygen (PaO2), partial pressure of arterial carbon dioxide (PaCO2), ratio of PaO2to fraction of inspired oxygen (FiO2) (PaO2/FiO2ratio), C-reactive protein (CRP), prealbumin albumin, globulin, lactate dehydrogenase (LDH), procalcitonin (PCT), prothrombin time ratio (PT ratio), international normalized ratio (INR), D-dimer, serum creatinine (sCr), blood urea nitrogen (BUN), and uric acid (UA). These examinations were fi rst recorded at 24 hours after admission and then dynamically collected on the 3rdday, 7thday, 14thday, 28thday, and fi nally when discharged. Patients were categorized into survival and non-survival groups according to the outcome.

Statistical analysis

The normally distributed continuous variables were expressed as the means±standard error of the mean (SEM). Abnormally distributed continuous variables were expressed as medians (interquartile ranges), and categorical variables were expressed as numbers (percentages). The differences between the survival and non-survival groups were analyzed by the Kolmogorov-Smirnov test. The Mann-Whitney U test or Pearson's chisquare test was used to compare abnormal distributional variables and categorical variables between two groups, while the two-sample t-test of variance was used to compare continuous variables as appropriate.

The repeated measured variables were also collected and reported as descriptive analyses. For the longitudinal regression analysis, generalized estimating equations (GEEs) were the most suitable technique, whereby all measurement cycles of subjects were taken into account to make corrections for correlated measures. With GEEs, the longitudinal effects of existing diagnosed factors on outcomes during the period of the disease were examined.[16]For these analyses, only the dynamic variables were tested with a uni-variate GEE analysis with a P-value ＜ 0.05, and then the Pearson test was used for the bi-variate correlation test to select independent predictors. Subsequently, the independent potential longitudinal predictors were enrolled into a multivariate model to determine the regression to mortality by using multivariate GEEs based on the odds ratio (OR) and 95% confidence interval (CI), with a P-value of 0.05 considered significant. We purposely chose very conservative assumptions and performed sensitivity analyses with the linear mixed model to test whether the results would qualitatively change if a different assumption was used. Single imputation by regression methods was used for all missing values for the determinants. Receiver operating characteristic (ROC) curves were constructed, and the areas under the ROC curves (AUCs) were determined. The optimal cut-off value was determined when the Youden index reached the maximum value. All analyses were programmed by qualified statisticians using SPSS 22.0 (SPSS statistical package, Chicago, IL, USA).

RESULTS

The baseline characteristics of CAP patients in AID-induced ICH

The underlying conditions are described in Table 1. The baseline characteristics of the patients are shown in Table 2. In terms of the basic immunosuppressive therapy of autoimmune diseases, 66 (70.21%) individuals underwent combined treatment with glucocorticoids and other immunosuppressive agents, 25 (26.60%) with glucocorticoids and 3 (3.19%) with immunosuppressive agents only.

Clinical characteristics of survival and nonsurvival patients

Of the 94 immunocompromised patients with CAP during the study period, 57 (60.64%) died in the hospital. No signifi cant differences were observed in the frequency of basic immunosuppressive therapy, comorbid illnesses, clinical symptoms or imaging features of the lung. The frequencies of renal insufficiency and dysfunction of coagulation were significantly higher in non-survival patients than in survival patients (P＜0.01). There was a significant difference regarding the condition for mechanical ventilation. Noninvasive ventilation and invasive mechanical ventilation were performed more frequently in non-survival patients (P＜0.05, respectively) (Table 3). High-fl ow nasal cannula or oxygen was primarily used in survival patients (P＜0.05).

Multivariate analysis of risk factors for mortality

Clinical indicators were collected at different time points after admission (the 1st, 3rd, 7th, 14th, 28th, and discharge days) and recorded as the median and IQR. Using uni-variate GEE analysis, the following independent risk factors for prognosis in CAP were included: NLR, PaO2/FiO2, CRP, albumin, LDH, D-dimer, sCr, BUN and UA (Table 4). The factors NLR, CRP, LDH, D-dimer, and sCr were added into the fi nal multivariate GEE model. The variables PaO2/FiO2, albumin, BUN, and UA were excluded according to the clinical value and collinearity analysis. The results showed that NLR, LDH and sCr were statistically significant after adjusting for age and sex (P＜0.01) (Table 5).

As shown in Table 6, the levels of sCr were significantly different on the first day after admission, and the levels of NLR, LDH and sCr were significantly different on the third day after admission. On the third day after admission, the cut-off values of LDH, NLR, and sCr were 695 U/L, 9.5, and 79 μmol/L with AUCs of 0.760 (95% CI 0.626-0.895), 0.747 (95% CI 0.595-0.899), and 0.704 (95% CI 0.555-0.853), respectively. When the three factors were used together, the AUC was 0.856 (95% CI 0.742-0.970, P＜0.001) (Figure 1). All these results suggested that the NLR, LDH and sCr might be independent risk factors for mortality in CAP patients who were AID-induced ICHs.

Table 1. The underlying conditions of autoimmune disease

Figure 1. R eceiver operating characteristic curves to assess the predictive accuracy of sCr, NLR and LDH for clinical outcomes on the 3rd day a fter admission.

DISCUSSION

The current study indicated that CAP patients who were AID-induced ICHs had a higher mortality rate (60.64%) than immunocompetent subjects with CAP.[17]Most patients developed respiratory failure, and noninvasive ventilation was the major oxygenation strategy in both the survival and non-survival patients. Additionally, we found that high levels of NLR, LDH and sCr at different time points were significant for the prediction of poor prognosis. Prior to the current study, data on AID-induced ICHs with CAP were not well described, and the present work was the first to explore the prognostic factors in these patients. In contrast to earlier fi ndings in CAP, our study reinforced the longitudinal variation in the biomarkers, which may be a more comprehensive assessment for the patient's conditions.

Lymphocytes are v arious cell populations with different functional and phenotypical properties involved in adaptive immunity.[18]It was proposed that in critically ill patients, dysregulated lymphocyte apoptosis might lead to immune suppression, leaving the patient vulnerable to subsequent infections or unable to fight existing sepsis.[18]Moreover, neutrophils act as the first cellular defense against infection and the key cell type of the innate immune system.[19]The hypothesis that the NLR was associated with outcomes was based primarily on the physiological link between neutrophilia and lymphopenia with systemic infl ammation.[20]In this study, the predictive factor of NLR was found to be a relevant index to mortality in CAP patients with AIDinduced ICH, in accordance with a proposal by De Jager et al[21]that NLR is an easily measurable parameter to express injury severity. Previous studies also demonstrated that a high NLR ratio in the emergency department is independently associated with hospital mortality.[20]Septic shock patients presented a low NLR at admission; however, mortality in the late stage was associated with an increased NLR.[22]In addition, the NLR showed prognostic value in predicting 30-day mortality and 3-month re-hospitalization rates with CAP in a prospective clinical trial.[23]Our results also indicated that a high NLR was an independent risk factor for the prediction of poor prognosis.

Table 2. Clinical characteristics of CAP patients in AID-induced ICH

Table 4. The univariate GEEs for risk regression analysis

Table 5. The multivariate generalized estimating equation for risk regression analysis

Table 3. The clinical characteristics in survival and non-survival patients, n

Table 6. The dynamic data patterns of repeated measures

LDH, a host enzyme widely expressed in human tissues, is released into the bronchoalveolar space upon damage to the cytoplasmic cell membrane.[24]Many previous studies emphasized the predictive value of serum LDH as one of the serologic biomarkers for the diagnosis of pneumocystis carinii pneumonia (PCP).[25]In recent years, LDH elevations have been described in fatal cases from multiorgan system failure,[26]which corresponded to our fi nding of higher LDH levels in nonsurvival patients at all time points after admission. Rios and colleagues also concluded that high LDH levels in non-survival patients might be a novel prognostic finding in ARDS caused by influenza A (H1N1) viral pneumonitis.[27]Similarly, our study revealed a signifi cant relationship between LDH and prognosis in AID-induced ICH patients with CAP, which might be a novel and convenient prognostic biomarker.

In the current study, it was interesting that s erum concentrations of creatinine were higher in non-survival patients. Additionally, multivariate GEE analysis identified sCr as an independent prognostic factor for mortality. To our knowledge, previous studies have pointed out that increased sCr levels markedly reduce long-term survival during critical illnesses, such as severe kidney failure and renal replacement therapy.[28]In addition, a high level of sCr is common in sepsis and associated with substantially increased morbidity and mortality.[29]In CAP patients, who might be in the early stage of sepsis, the level of sCr might indicate a poor progression. Therefore, we could use this biomarker as a routine test, and new therapies for this complication need to be explored in CAP patients, even those with an underlying immunocompromised state.

The current study had several limitations. First, though all patient laboratory results were thoroughly described and analyzed, complete biochemical indicators and microbiological samplings could not be fully used as a limitation of the retrospective design. Second, the results cannot be generalized to ICH patients with CAP u nder a non-AID background. Other ICH patients, such as those with underlying HIV, those suffering from organ transplant, and those with neoplastic disease receiving chemotherapy or radiotherapy, were not included. As there was no obvious evidence for evaluating the heterogeneity of ICH with different disease backgrounds, a further study with a larger sample size and more comprehensive types of disease needs to be performed.

C ONCLUSION

A signifi cant relationship was demonstrated between the indices of NLR, LDH, sCr and the mortality risk of CAP occurring in AID-induced ICHs. These markers could potentially be widely used in clinical practice due to their convenience. However, further studies in different ethnicities and with larger populations of AIDinduced ICH patients are needed to confirm the current fi nding.

Funding:This work was supported by the Shanghai Traditional Medicine Development Project (ZY3-CCCX3-3018, ZHYYZXYJH-201615), the National Natural Science Foundation of China (81471840, 81171837), the Zhongshan Hospital Distinguished Young Scholars and the Shanghai Municipal Planning Commission of science and Research Fund (20134Y023), and Key Project of Shanghai Municipal Health Bureau (2016ZB0202).

Ethical approval:The protocol was approved by the medical ethics committee of Zhongshan Hospital, Fudan University.

Conf icts of interests:They have no confl icts of interest.

Contributors:ZSK and YLY contributed equally to this work. ZJS, ZS, and CYT conceived the project and designed the study, ZSK, YLY, WW, JLW, and XYL analysed and interpreted the data. ZSK and YLY drafted the primary draft of the manuscript. ZJS, ZS, CYT and KYL revised the manuscript critically for important intellectual content. All authors revised and approved the final version of the manuscript.

R EFERENCES

1 Gronlund O, Herweijer E, Sundstrom K, Arnheim-Dahlstrom L. Incidence of new-onset autoimmune disease in girls and women with pre-existing autoimmune disease after quadrivalent human papillomavirus vaccination: a cohort study. J Intern Med. 2016;280(6):618-26.

2 Gharagozloo M, Majewski S, Foldvari M. Therapeutic applications of nanomedicine in autoimmune diseases: from immunosuppression to tolerance induction. Nanomedicine. 2015;11(4):1003-18.

3 Kovarik JM, Burtin P. Immunosuppressants in advanced clinical development for organ transplantation and selected autoimmune diseases. Expert Opin Emerg Drugs. 2003;8(1):47-62.

4 Murdaca G, Orsi A, Spano F, Puppo F, Durando P, Icardi G, et al. Influenza and pneumococcal vaccinations of patients with systemic lupus erythematosus: current views upon safety and immunogenicity. Autoimmun Rev. 2014;13(2):75-84.

5 Luo X, Miller SD, Shea LD. Immune tolerance for autoimmune disease and cell transplantation. Annu Rev Biomed Eng. 2016;18:181-205.

6 Thomas K, Vassilopoulos D. Immunization in patients with inflammatory rheumatic diseases. Best Pract Res Clin Rheumatol. 2016;30(5):946-63.

7 Lv J, Zhang H, Wong MG, Jardine MJ, Hladunewich M, Jha V, et al. Effect of oral methylprednisolone on clinical outcomes in patients with IgA nephropathy: the TESTING randomized clinical trial. JAMA. 2017;318(5):432-42.

8 Xie H, Zhou ZG, Jin W, Yuan CB, Du J, Lu J, et al. Ventilator management for acute respiratory distress syndrome associated with avian influenza A (H7N9) virus infection: A case series. World J Emerg Med. 2018;9(2):118-24.

9 Thomas K, Vassilopoulos D. Infections and vasculitis. Curr Opin Rheumatol. 2017;29(1):17-23.

10 Letourneau AR, Issa NC, Baden LR. Pneumonia in the immunocompromised host. Curr Opin Pulm Med. 2014;20(3):272-9.

11 Vila-Corcoles A, Ochoa-Gondar O, Rodriguez-Blanco T, Raga-Luria X, Gomez-Bertomeu F, Group ES. Epidemiology of community-acquired pneumonia in older adults: a populationbased study. Respir Med. 2009;103(2):309-16.

12 Sousa D, Justo I, Dominguez A, Manzur A, Izquierdo C, Ruiz L, et al. Community-acquired pneumonia in immunocompromised older patients: incidence, causative organisms and outcome. Clin Microbiol Infect. 2013;19(2):187-92.

13 Mandell LA, Wunderink RG, Anzueto A, Bartlett JG, Campbell GD, Dean NC, et al. Infectious Diseases Society of America/American Thoracic Society consensus guidelines on the management of community-acquired pneumonia in adults. Clin Infect Dis. 2007;44 Suppl 2:S27-72.

14 Arumugam SK, Mudali I, Strandvik G, El-Menyar A, Al-Hassani A, Al-Thani H. Risk factors for ventilator-associated pneumonia in trauma patients: A descriptive analysis. World J Emerg Med. 2018;9(3):203-10.

15 Chen JH, Chang SS, Liu JJ, Chan RC, Wu JY, Wang WC, et al. Comparison of clinical characteristics and performance of pneumonia severity score and CURB-65 among younger adults, elderly and very old subjects. Thorax. 2010;65(11):971-7.

16 Bisschop MI, Kriegsman DM, Deeg DJ, Beekman AT, van Tilburg W. The longitudinal relation between chronic diseases and depression in older persons in the community: the Longitudinal Aging Study Amsterdam. J Clin Epidemiol. 2004;57(2):187-94.

17 Prina E, Ranzani OT, Torres A. Community-acquired pneumonia. Lancet. 2015;386 (9998):1097-108.

18 Wesche DE, Lomas-Neira JL, Perl M, Chung CS, Ayala A. Leukocyte apoptosis and its significance in sepsis and shock. J Leukoc Biol. 2005;78(2):325-37.

19 Kolaczkowska E, Kubes P. Neutrophil recruitment and function in health and infl ammation. Nat Rev Immunol. 2013;13(3):159-75.

20 Salciccioli JD, Marshall DC, Pimentel MA, Santos MD, Pollard T, Celi LA, et al. The association between the neutrophilto-lymphocyte ratio and mortality in critical illness: an observational cohort study. Crit Care. 2015;19:13.

21 de Jager CP, van Wijk PT, Mathoera RB, de Jongh-Leuvenink J, van der Poll T, Wever PC. Lymphocytopenia and neutrophillymphocyte count ratio predict bacteremia better than conventional infection markers in an emergency care unit. Crit Care. 2010;14(5):R192.

22 Riche F, Gayat E, Barthelemy R, Le Dorze M, Mateo J, Payen D. Reversal of neutrophil-to-lymphocyte count ratio in early versus late death from septic shock. Crit Care. 2015;19:439.

23 Cataudella E, Giraffa CM, Di Marca S, Pulvirenti A, Alaimo S, Pisano M, et al. Neutrophil-to-lymphocyte ratio: an emerging marker predicting prognosis in elderly adults with communityacquired pneumonia. J Am Geriatr Soc. 2017;65(8):1796-801.

24 Esteves F, Cale SS, Badura R, de Boer MG, Maltez F, Calderon EJ, et al. Diagnosis of pneumocystis pneumonia: evaluation of four serologic biomarkers. Clin Microbiol Infect. 2015;21(4):379.e1-10.

25 Tasaka S, Hasegawa N, Kobayashi S, Yamada W, Nishimura T, Takeuchi T, et al. Serum indicators for the diagnosis of pneumocystis pneumonia. Chest. 2017;131(4):1173-80.

26 Perez-Padilla R, de la Rosa-Zamboni D, Ponce de Leon S, Hernandez M, Quinones-Falconi F, Bautista E, et al. Pneumonia and respiratory failure from swine-origin infl uenza A (H1N1) in Mexico. N Engl J Med. 2009;361(7):680-9.

27 Rios FG, Estenssoro E, Villarejo F, Valentini R, Aguilar L, Pezzola D, et al. Lung function and organ dysfunctions in 178 patients requiring mechanical ventilation during the 2009 infl uenza A (H1N1) pandemic. Crit Care. 2011;15(4):R201.

28 Linder A, Fjell C, Levin A, Walley KR, Russell JA, Boyd JH. Small acute increases in serum creatinine are associated with decreased long-term survival in the critically ill. Am J Respir Crit Care Med. 2014;189(9):1075-81.

29 Rewa O, Bagshaw SM. Acute kidney injury-epidemiology, outcomes and economics. Nat Rev Nephrol. 2014;10(4):193-207.