Progression To Lower Respiratory Tract Disease In Parainfluenza Virus Infection After Hematopoietic Cell Transplantation Based On a New Classification

Parainfluenza virus (PIV) is a respiratory virus that infects approximately 10% of hematopoietic cell transplantation (HCT) recipients. A substantial proportion of patients with lower respiratory tract disease (LRD) present without prior upper respiratory tract infection (URI), but reasons for the lack of early detection are poorly understood. Since the definition of LRD differs among studies, we previously proposed a classification of LRD (possible, probable, and proven) that correlates with clinical outcomes (54^thASH meeting; abstract #458). The classification demonstrated that patients with PIV detection in lungs (probable and proven LRD) had significantly worse survival than those with PIV detection in a nasopharyngeal (NP) sample alone (possible LRD and URI), with day 90 overall survival 50% vs. 90%, respectively. The purpose of the study was (1) to evaluate risk factors for progression to proven and probable LRD and (2) to determine factors associated with LRD without preceding URI.

This retrospective cohort study included 391 HCT recipients with PIV infection transplanted between 1990 and 2011 at FHCRC. PIV detection was performed by conventional culture, direct fluorescent antibody testing, and/or polymerase chain reaction. Categories of PIV infection were defined according to site of PIV detection and radiography: URI: PIV detection in NP sample without new pulmonary infiltrates; possible LRD: detection in NP sample with new pulmonary infiltrates; probable LRD: detection in lung (BAL or biopsy) sample without new pulmonary infiltrates; and proven LRD: detection in lung sample with new pulmonary infiltrates. Patients diagnosed with LRD ≥ 2 days after URI diagnosis were considered to have progression to LRD. Risk factors for progression to proven/probable LRD were analyzed using Cox proportional hazards models, and comparisons between LRD cases with or without prior URI were performed by logistic regression models.

Among 391 patients, 344 had a positive NP sample at diagnosis of PIV infection and 47 were diagnosed with a positive lung sample alone. Among 344 patients, 61 progressed to LRD and median time to progression was 7 days (range, 2–40). The cumulative incidence of progression by 30-days among patients who presented with URI was 19%; 5%, 4% and 11% of patients progressed to possible, probable and proven LRD, respectively, and 28% of patients with possible LRD further progressed to proven LRD. In a multivariate model among the 344 patients, bone marrow (BM) or cord blood (CB) as cell source (aHR, 3.1; 95% CI, 1.6-5.9, p < 0.001), lymphocyte count less than 100/µL at time of PIV infection (aHR, 2.7; 95% CI, 1.5-4.8, p = 0.001), steroid use greater than 1mg/kg/day before PIV infection (aHR, 2.1; 95% CI, 1.1-4.1, p = 0.035), and PIV type 3 (aHR, 10.0; 95% CI, 1.4-100, p = 0.021) were significantly associated with progression to proven/probable LRD. New pulmonary infiltrates in patients with upper respiratory viral detection (possible LRD) was also associated with progression to proven/probable LRD (aHR, 5.2; 95% CI, 2.8-9.5, p < 0.001). Among 56 patients with possible LRD, all patients who had BM or CB as cell source and high-dose steroids progressed to proven/probable LRD while patients receiving peripheral blood stem cells (PBSCs) with low or no steroid exposure rarely progressed. Among 94 proven/probable LRD cases, 47 presented with LRD without prior URI. PBSCs as cell source (aOR, 4.20; 95% CI, 1.7-10, p = 0.002) and inpatient status at diagnosis (aOR, 3.86; 95% CI, 1.5-9.7, p = 0.004) were significantly associated with progression to LRD without prior URI. Overall, 9 / 47 patients (19%) without virologically proven prior URI had NP tests that were negative or non-interpretable.

Progression from URI to proven/probable LRD occurs in 21% of PIV-infected patients. Risk factors for progression include PIV-3 subtype, high-dose steroids, BM or CB as cell source and lymphopenia. Progression rates are also high in patients with possible LRD. These data provide important information for clinical trial design as new antivirals are being developed. Strategies are needed to improve detection of PIV prior to progression.

Seo:Ansun BioPharma: Research Funding. Boeckh:Ansun BioPharma: Research Funding.