Burden of severe RSV disease among immunocompromised children and adults: a 10 year retrospective study

From January 1, 2005 to December 31, 2014, we conducted a multi-center observational study of children and adults presenting with an immunodeficiency state who presented with an RSV-URTI or LRTI in an ambulatory setting or upon or during their admission to the University hospitals of Lausanne and Geneva, Switzerland. RSV-positive patients were identified from Laboratory records. Clinical and laboratory data on RSV-positive immunocompromised adults and children were collected retrospectively with the assistance of hospital medical records. All children less than 18 years of age and adults presenting with at least one of the following conditions were included: allogeneic or autologous HSCT recipients, solid organ transplant (SOT) recipients, patients on cancer chemotherapy or long-term immunosuppression for any chronic disease, which included mostly patients with an underlying rheumatologic condition and those with Connective Tissue Disease (CTD) (Additional file 1: Table S1); and primary immunodeficiency disease (PID). Immunocompromised children with other comorbidities qualifying for RSV prophylaxis based on current Swiss guidelines [8] and patients with human immunodeficiency virus (HIV) were excluded. We defined URTI as detection of RSV in upper respiratory secretions together with symptoms only involving the upper respiratory tract and LRTI as any RSV-positive patient associated with cough, tachypnea and any respiratory distress or wheezing Pneumonia was defined as any of the above LRTI symptoms with pulmonary infiltrates reported by a radiologist on chest radiography and the exclusion of other causes as defined elsewhere [9, 10]. Acute-respiratory-tract infection (ARTI)-attributable hospital admission included patient admitted for respiratory symptoms with concomitant documentation of RSV in a respiratory samples. Severe neutropenia was defined as an absolute neutrophil count (ANC) < 500 cells/μl and severe lymphopenia as an absolute lymphocyte count (ALC) < 100 cells/μl [11]. Approval was obtained from the Research Ethics Boards at University hospitals of Lausanne and Geneva.

Information on patient demographic variables such as age, gender, type of immunosuppression, bacterial and viral co-infection, median ANC or ALC counts, oral ribavirin or palivizumab treatments were all abstracted from health records. Primary outcome referred to acute respiratory tract infection (ARTI)-attributable hospital admission. Secondary outcomes consisted of the documentation of LRTI or pneumonia, intensive care unit (ICU) admission and RSV-contributed mortality within 30 days of the diagnosis of RSV infection. RSV-attributable admission to the intensive care unit (ICU) was defined as a documented RSV disease requiring admission to the ICU. Overall mortality referred to death from any cause whereas RSV-contributed mortality was defined as a persistent or progressive RSV infection with respiratory failure documented at the time of death [12]. Each case was adjudicated by 2 members of the study team.

From 2005 to 2013, nasopharyngeal specimens, broncho-alveolar lavage (BAL) fluids and sputum samples were examined by a two-step reverse transcriptase polymerase chain reaction (RT-PCR) assay, targeting RSV A and B. This assay is extensively validated at the Laboratory of Virology, University Hospitals Geneva. In addition, specimens collected from children, which required immediate identification for cohorting purposes, were tested by a rapid antigen assay for RSV (RSV Respi-Strip, Coris Bioconcept) [13, 14] A commercial one-step PCR FTD Respiratory pathogens 21 (ref.FTD.-53–96/12) from Fast-Track Diagnostics (Luxembourg) (used at University hospitals Geneva since 2013) and the Cepheid Xpert Flu/RSV (used at University hospital Lausanne since 2015) further replaced molecular and rapid antigen assays [15, 16]. A patient was defined as being RSV-positive if any specimen was positive by any diagnostic method (antigen assay or any of the above listed molecular assay). Multiple simultaneous virus infections (herein referred to as viral co-infection) were those in which two or more virus pathogens were detected from the same respiratory sample. Bacterial co-infection was defined as the presence of any bacterial pathogen, identified by culture from blood or respiratory samples upon initial consultation with respiratory symptoms or within 30 days of their initial consultation in association with a documented viral infection.We excluded children if blood culture isolates were considered to be contaminants based on international guidelines [17].

Standard descriptive and comparative statistics (the Fisher’s exact test for categorical variables and the Mann-Whitney U test for continuous variables) were used on data categorized by age groups (children under 18 years of age and adults) and in and outpatients. For all statistical testing, only the first RSV-positive nasopharyngeal swab was included in cases with multiple RSV-positive swabs within the same patient because these samples cannot be considered independent. Univariable and multivariable logistic regression were used to assess clinical and laboratory correlates of ARTI-attributable hospital admission, LRTI and pneumonia with stratified analyses for children and adults. Predictor variables included different immunosuppressive categories, age, bacterial infection, and ALC for all patients and the adult subgroup whereas only bacterial co-infection and ALC were included in the children subgroup given the small number of observations. Statistically significant predictors in univariable analyses were included in multivariable analyses [18]. All analyses were conducted by complete case analyses, excluding all patients with missing values for any dependent or independent variables. We evaluated next the robustness of the estimates related to the potential differences in disease severity described for i) HSCT recipients infected with RSV within 2 years of transplantation; ii) patients who presented with viral co-infections and iii) patients detected RSV-positive by molecular assays solely. Therefore, we performed three post hoc sensitivity analyses for primary and secondary outcomes: one in which HSCT recipients infected with RSV after two years from their transplantation were excluded, another one excluding all 52 patients with viral co-infections and one excluding 12 children detected RSV-positive by rapid antigen assays only. Two sided tests were performed, and a P value < 0.05 was considered to be statistically significant. Data were analyzed using SPSS statistical software (version 20.0, SPSS Inc., Chicago, IL, USA), SAS and R (version 3.3.3, R Foundation for Statistical Computing, Vienna, Austria).

Three thousand two hundred and twenty-three RSV-positive patients (4′605 positive RSV samples) of whom 239 (7.4%) were eligible RSV-positive immunocompromised subjects (64 (26.8%) children and 175 adults (73.2%) (Additional file 1: Figure S1). From all patients, HSCT recipients (N = 67; 28.0%) and SOT recipients (N = 63; 26.4%) were the most common immunosuppressive conditions included. Most of the 64 children, 35 (54.7%) males, with a median age of 4.9 years (interquartile range [IQR] 3.2–7.7) presented with leukaemia or lymphoma (N = 29; 45.3%), whereas HSCT (N = 59, 33.7%) and SOT (N = 58, 33.1%) were the most common immunosuppressive conditions documented among the 175 adults (Table 1).

RSV was detected positive from 202 (84.5%) nasopharyngeal (NP) swabs and 35 (15%) BAL, of which 25 (28.6%) were collected from patients with pneumonia. Two hundred and twenty-seven patients (95%) were detected by molecular assays and 12 (5%) by rapid antigen assays. From all patients, 36 (15.1%) presented with a documented bacterial co-infection, of whom 29 (80.6%) with bacteremia and 7 (19.4%) with bacterial pneumonia documented from BAL. From the 11 patients who died, only 3 (27.3%) presented with a bacterial co-infection (S. aureus, Stenotrophomonas maltophilia and Legionella pneumophila) documented from BAL. 52 (21.8%) patients presented with a viral co-infection of whom 10 (19.2%) had more than two viruses detected. The most common viral co- infections were human rhinovirus/enterovirus-RSV (16/42; 38.1%), parainfluenza (PIV)-RSV (6/42; 14.3%) and influenza (FLU)-RSV (6/42; 14.3%) (Table 1).

From all immunocompromised subjects, 89 (37.2%) (N = 58; 34.1% adults; N = 31; 48.4% children; p = 0.090) were admitted to hospital for their RSV infection with a median duration of stay of 5 days (IQR: 0–18 days). From those 89, 19 (21.3%) patients were admitted to intensive care of whom 6 (31.6%) presented with chronic disease requiring underlying immunosuppressive therapy, 4 (21.1%) with solid tumors and 4 (21.1%) with HSCT. Patients admitted to the ICU, requiring mechanical ventilation and those who died were almost all adults. Likewise immunocompromised adults presented with significantly higher rates of pneumonia (36.9% vs. 18.8%, p = 0.008) compared to immunocompromised children. Furthermore, all 11 adults who died within 30 days of their RSV-attributable admission, presented with pneumonia despite ribavarin (N = 6, 54.5%) or palivizumab (N = 2,18.2%) administration. Most of those 11 patients, median age of 64.6 years (IQR 34.1–73.9), were HSCT recipients (N = 4, 36.4%) or presented with lymphoma/leukemia (N = 3, 27.3%) or solid tumors (N = 3; 27.3%) (Table 2).

From multivariable analyses, we identified that patients with leukaemia/lymphoma (odds ratio [OR] 3.9; 95% confidence interval [CI]: 1.7–9.7; P = 0.002), solid tumors (OR 7.4; 95% CI: 2.7–21.9; P < 0.001) or those requiring chronic immunosuppression mainly for vasculitis (OR 9.5; 95% CI: 3.3–30.2 P < 0.001) when compared to HSCT, were significantly more likely to be admitted to hospital for RSV. When stratified by age groups (adults vs. children), increasing age (OR 1.4; 95% CI: 1.0–1.8 P = 0.027), patients with solid tumors (OR 5.2; 95% CI: 1.4–20.9; P = 0.015) and those requiring any chronic immunosuppression mainly for vasculitis (OR 4.1; 95% CI: 1.1–16.0; P = 0.034) when compared to HSCT, remained significant predictors of ARTI-attributable hospital admission among adults. Post-hoc sensitivity analyses did not affect our risk estimates and the above findings (Table 3, Additional file 1: Tables S2, S4 and S6).

From multivariable analyses, patients presenting with a bacterial co-infection were significantly more likely to present with LRTI (OR 3.4; 95% CI: 1.5–8.2; P = 0.005) and pneumonia (OR 3.6; 95% CI: 1.6–8.4; P = 0.002). When stratified by age groups (adults vs. children), bacterial co-infection was also an independent and significant predictor for LRTI (OR 3.4; 95% CI: 1.3–9.8; P = 0.018) only in adults. Increasing age was also a significant predictor for LRTI (OR 1.1; 95% CI: 1.0–1.3; P = 0.048) and pneumonia (OR 1.2; 95% CI: 1.1–1.4 P = 0.003) among all patients. Post-hoc sensitivity analyses did not affect our risk estimates and the above findings (Table 4, Additional file 1: Tables S3, S5 and S7).