We identified 19 adult patients (14 males and 5 females; median age, 38 years) who were diagnosed with adenovirus pneumonia between March 2003 and April 2015 in a single tertiary-referral hospital. All patients underwent Gram stains, bacterial and fungal cultures of blood and urine. Sputum was stained and cultured for bacteria, mycobacteria, and fungi. Specimens from patients who underwent bronchoscopy with the collection of bronchoalveolar lavage (BAL) fluid or lung biopsy, as well as those who underwent surgical biopsy, were stained and cultured for bacteria, mycobacteria, and fungi. In addition, polymerase chain reaction (PCR) analysis for BAL specimens and nasal or throat swabs were performed to detect viral organisms. Patients in whom other infectious organisms were identified in the sputum, blood, urine, nasal and throat swabs, BAL fluid, or lung specimen within four weeks of adenovirus identification were regarded to have a concurrent infection and were excluded from this study. Finally, 19 patients confirmed to have only adenovirus pneumonia were included in this study; these patients included the five patients whose imaging features were published in our previous study (10).

All data including age, sex, premorbid conditions, symptoms, laboratory and pathological findings, clinical course, treatment, and survival outcome were drawn from patient electronic medical records. Patients were regarded as having ARDS when they satisfied the American-European Consensus Conference definition for the condition: acute onset of respiratory failure, bilateral infiltrates on chest radiograph, hypoxemia as defined by a ratio of arterial oxygen partial pressure to fractional inspired oxygen (PaO2/FiO2 ratio) ≤ 200 mm Hg, and no evidence of left atrial hypertension or pulmonary capillary pressure < 18 mm Hg (if measured) to rule out cardiogenic edema (11).

Diagnosis was confirmed in five patients based on positive viral culture for adenovirus in BAL fluid. Diagnosis via lung biopsy specimens obtained by video-assisted thoracoscopic surgery (n = 4) or transbronchial lung biopsy (n = 1) was made in another five patients. The specimens were evaluated using hematoxylin and eosin immunohistochemical staining (1012). In nine patients, BAL fluid was analyzed using the one-step multiplex realtime AdvanSure RV real-time PCR kit (LG Life Sciences, Seoul, Korea) or conventional multiplex reverse transcription PCR (RT-PCR) assay (Seeplex RV12 Detection Kit; Seegene, Seoul, Korea) (1314).

Initial chest radiographs and computed tomography (CT) scans were acquired after an average of 4.4 days (range, 0 to 15 days) or 6.1 days (range, 2 to 15 days) from the commencement of symptoms or signs, respectively. The mean time interval between radiologic examination and pathological or microbiological diagnosis was 9.6 days (range, 1–72 days). In one patient whose interval from initial imaging study to pathological diagnosis was 72 days, confirmative diagnosis of adenovirus pneumonia was delayed due to antibiotics treatment showing gradual deterioration of the disease.

Posterior-anterior chest radiographs were obtained with a computed radiography system (FCR 9501; Fuji, Tokyo, Japan) or digital radiography system (Revolution XQi ADS_28.4; GE Medical Systems, Milwaukee, WI, USA) with operating parameters of 70–120 kVp and 2–3 mAs. Serial chest radiographs were acquired for all patients during hospitalization in order to assess the progression of disease. CT studies were performed using various helical CT scanners (Hi Speed Advantage, Light Speed QXi, Light Speed Ultra, Light Speed VCT and Discovery CT750 HD, GE Healthcare, Chalfont St Giles, England; Brilliance 40, Philips, Best, the Netherlands; Aquilion, Toshiba, Tokyo, Japan). CT scans (114–210 mA, 120 kVp, beam width of 10–20 mm, beam pitch of 1.375–1.5) were performed from the lung apices to the level of the middle portions of both kidneys. Intravenous contrast medium injection was used in approximately half of the patients (10 patients), with 1.5 mL/kg (body weight) of Iomeron 300 (iomeprol, 300 mg iodine/mL; Bracco, Milan, Italy) injected at an infusion rate of 3 mL/s using a power injector (MCT Plus; Medrad, Pittsburgh, PA, USA). The imaging data were reconstructed with section thicknesses of 2.5–5 mm using a bone algorithm. The reconstructed images were then interfaced directly to a picture archiving and communication system (Centricity 3.0; GE Healthcare, Mt. Prospect, IL, USA), which displayed all image data on two monitors (1536 × 2048 matrix, 8-bit viewable gray scale, and 60-ft-lambert luminescence). Both mediastinal (width, 400 HU; level, 20 HU) and lung (width, 1500 HU; level, -700 HU) window images were viewed on these monitors.

Two chest radiologists (MJC and MJJ, with five and 22 years of experience in chest imaging interpretation, respectively) assessed chest radiographs and CT images; decisions were reached by consensus. The abnormalities on initial chest radiographs were assessed as focal opacity or multifocal or diffuse opacity (15). Parenchymal opacity was defined as an area that appears denser than the surrounding lung tissue without a discrete border. The pattern of the disease course was also identified from sequential chest radiographs as described by Wong et al. (16) and Das et al. (17), where type 1 was radiographic improvement, type 2 was radiographic deterioration by one peak level followed by improvement, type 3 was fluctuating radiographic changes with at least two peaks, and type 4 was progressive radiographic deterioration. Peak level was defined as overall lung parenchymal involvement greater than 25% of the initial extent on radiography.

On CT, the patterns of parenchymal abnormalities were subdivided into areas of ground-glass opacity (GGO), consolidation, and micronodules (< 5 mm in diameter). GGO was defined as an area of hazy increased lung opacity, within which margins of pulmonary vessels were indistinct. Consolidation was defined as a homogeneous increase in pulmonary parenchymal opacity that obscured the margins of vessels and airway walls (18). The presence of interlobular septal thickening or pleural effusion was also recorded, and the laterality (unilateral vs. bilateral) of the lesions was evaluated. Septal thickening was defined as abnormal widening of an interlobular septum or septa (18). In terms of anatomic location, the distribution of parenchymal abnormalities was classified as central, peripheral, or mixed in the axial plane and upper, lower, or random in the longitudinal plane. The outer third of the lung was defined as peripheral, and the inner two-thirds of the lung were defined as central. The lesions were regarded as distributed in the upper lung zone when they were located above the hilum, in the lower lung zone when they were located below the hilum, or random when the distribution could not be determined. For the acquisition of CT scores, the lung was also divided into upper (above the carina), middle (below the carina and up to the inferior pulmonary vein), and lower (below the inferior pulmonary vein) zones. The involvement of pneumonia within each lung zone was evaluated by scoring each zone from 0 (normal) to 4, with 4 corresponding to nearly total involvement of the lung parenchyma. Total CT scores ranged from 0 to 24 when combined for all six zones (1719).

Differences in the clinico-radiological characteristics of the patients who eventually developed ARDS (ARDS group) and those who did not (non-ARDS group) were assessed by using chi-square test, Fisher's exact test, or Student's t test. Statistical analyses were performed with SPSS software (version 19.0; SPSS Inc., Chicago, IL, USA). A p value less than 0.05 was considered to indicate a significant difference.

Detailed patient characteristics are shown in Table 1. Among the 19 patients with adenovirus pneumonia, nine were immunocompromised due to underlying comorbid conditions, such as acute myeloid leukemia (n = 4), non-Hodgkin's lymphoma (n = 1), multiple myeloma (n = 1), lung adenocarcinoma (n = 1), history of liver transplantation for hepatocellular carcinoma (n = 1), and diabetes mellitus in conjunction with chronic obstructive pulmonary disease and ischemic cardiomyopathy (n = 1). In contrast, 10 patients were immunocompetent without underlying disease.

Most patients had upper respiratory symptoms such as dyspnea, cough, sputum, fever, and chills, as shown in Table 1. Twelve of 19 patients (63%) subsequently progressed to ARDS, six (32%) of whom eventually died. The average time of progression from commencement of symptoms or signs to ARDS was 9.6 days, indicating relatively rapid deterioration. Fifteen patients received antiviral therapy, including ganciclovir (n = 1), ribavirin (n = 3), or cidofovir (n = 11).

Initial chest radiographs of two patients (11%) were normal, whereas 17 patients had abnormal chest radiographs that showed multifocal or diffuse opacity (n = 8, Fig. 1), or focal opacity (n = 9, Figs. 2, 3). In terms of the disease course identified from serial chest radiographs, two showed gradual improvement (type 1, Fig. 2), nine showed initial radiographic deterioration by one peak level followed by improvement (type 2), two showed radiographic fluctuations with at least two peaks (type 3), and six patients showed progressive radiographic deterioration (type 4, Figs. 1, 3).

Patterns and distribution of CT findings are summarized in Table 2. The most frequent CT appearance was GGO, which was seen in all patients (n = 19), combined with other patterns such as consolidation (n = 14) and micronodules (n = 3). Parenchymal abnormalities were distributed bilaterally in 17 patients (Figs. 1, 2), whereas unilateral involvement was seen in two patients (Fig. 3). Interlobular septal thickening was demonstrated in 12 patients, and pleural effusion was present in 11 patients. The average number of lobes involved was 4.2 ± 1.4 (range, 1 to 5). Pulmonary abnormalities showed upper lung zone predominance in four patients (21%) and lower lung zone predominance in six patients (32%). However, nine patients (47%) showed random parenchymal involvement without zonal predominance. In terms of axial distribution, most of the cases (79%) showed mixed central and peripheral involvement.

Table 3 shows the comparison between the ARDS and non-ARDS groups according to clinical and radiological findings. No significant differences between the groups were identified for the clinical variables of patient age (p = 0.651), sex (p = 0.998), or presence of premorbid conditions (p = 0.650). Regarding the radiological findings, there was no significant difference in initial radiograph findings (p = 0.813), but the radiographic pattern of disease course differed significantly between the ARDS and non-ARDS groups (p = 0.048). In the non-ARDS group, types 1 and 2 accounted for more than 85% of patients, and no one was classified as type 4. In contrast, half of the patients in the ARDS group showed progressive deterioration of chest radiographs and were classified as type 4. Among the CT findings, the presence of pleural effusion and total CT scores differed significantly (p = 0.010 and p = 0.007, respectively) between the two groups. Patients who progressed to ARDS showed higher total CT scores and a higher probability of pleural effusion. Other CT features, such as laterality (p = 0.998), presence of consolidation (p = 0.305) and micronodules (p = 0.997), inter- or intralobular septal thickening (p = 0.656), and number of involved lobes (p = 0.973) did not differ significantly between the ARDS and non-ARDS groups.