Etiologies, diagnostic strategies, and outcomes of diffuse pulmonary infiltrates causing acute respiratory failure in cancer patients: a retrospective observational study

Over the study period, a total of 583 patients with cancer were admitted to the oncology medical ICU. Of these patients, a total of 214 consecutive cancer patients with ARF and diffuse pulmonary infiltrates on chest radiography were included in the study. There was no patient with more than one ICU admission over the study period. ARF was defined as the presence of respiratory distress symptoms such as tachypnea (respiration rate >30/min) with either need for ventilator support or an arterial partial pressure of oxygen (PaO₂)/fraction of inspired oxygen (FiO₂) ratio (PF ratio) <300. Diffuse pulmonary infiltrates were defined as diffuse opacities involving all four quadrants of lung on chest radiography [8].

According to our diagnostic strategy for pulmonary infectious disease in cancer patients with diffuse pulmonary infiltrates, non-invasive tests including blood cultures; examination of the sputum for bacteria, mycobacteria, and fungi; urine tests for antigens of Streptococcus pneumoniae and Legionella pneumophila; and serologic tests for Mycoplasma were performed routinely. Blood tests for cytomegalovirus (CMV) antigen [9] and an Aspergillus galactomannan assay [10] were performed when there was a high suspicion of any of these pathogens. In addition, chest computed tomography (CT) scans were also performed to differentiate etiologies of diffuse pulmonary infiltrates if possible. For etiologic evaluation of acute respiratory failure from causes other than pulmonary disease, echocardiography, and blood tests for cardiac markers were performed in the ICU.

Bronchoaveolar lavage (BAL), transbronchial lung biopsy (TBLB), and surgical lung biopsy (SLB) were performed based on a joint decision of attending physicians and intensivists. BAL was performed using the standard technique as previously described [11]. BAL fluid samples were stained using Gram and Ziehl-Neelsen methods and then cultured for bacteria, mycobacteria, and fungi. Multiplex nested polymerase chain reaction (PCR) assays were used to detect influenza viruses A and B; parainfluenza viruses 1, 2, and 3; respiratory syncytial virus; and adenovirus [12]. Quantification of CMV DNA in BAL fluid was also performed using quantitative real-time PCR [13]. In addition, CMV and adenovirus were cultured using standard techniques. Calcoflour white stain and Grocott-Gomori methenamine silver stain were used to detect Pneumocystis jiroveci in BAL fluid [14]. TBLB and SLB were performed at the segment exhibiting the most severe abnormalities by chest CT scan. Specimens were analyzed by cultures, staining, and histologic testing.

Diagnoses were based on clinical, radiologic, microbiologic, and histopathologic findings [3] and were thoroughly reviewed by two of the authors (HY and KJ). Differences in observed findings were resolved by consensus. All diagnoses were classified as definite, probable, or non-diagnostic as follows. Definite diagnosis of bacterial pneumonia was defined as the pathogen being identified in BAL fluid in an amount >10⁴ CFU/mL or the same pathogen being isolated from respiratory specimens and blood cultures. Probable diagnosis of bacterial pneumonia was defined as improvement of the patient's clinical conditions after antibiotic treatment even though the pathogen was not isolated from a respiratory specimen. Definite viral pneumonia was defined as observation of the cytopathologic effect of a virus in histopathology [15]. Probable viral pneumonia was defined as recovery of a virus in BAL fluid or nasopharyngeal specimens from a patient with clinical features consistent with viral pneumonia. However, a positive CMV PCR test or antigenemia was not sufficient for diagnosis of CMV pneumonia [3]. Pulmonary aspergillosis was diagnosed according to recently developed criteria for invasive fungal disease [16]. The diagnosis of P. jiroveci pneumonia was made by identification of P. jiroveci from a clinically relevant specimen of sputum, BAL fluid or lung tissue [17]. Cardiogenic pulmonary edema was diagnosed when echocardiography and biomarkers showed cardiac dysfunction in patients with a compatible clinical picture [18]. Diffuse alveolar hemorrhage (DAH) associated with an underlying malignancy was diagnosed using previously reported criteria [19]. However, alveolar hemorrhage was considered to be a manifestation of lung disease but not a cause of ARF, except in patients with no BAL evidence of infection and no evidence of congestive heart failure [19, 20]. Drug-induced pneumonitis was defined as combination of the presence of a compatible clinical pattern, drug that is a known or suspected offender, and the exclusion of infection or pulmonary involvement from the underlying malignancy [21]. As this study included only patients with malignancies, other definitions associated with cancer status were defined by previously reported definitions [22‐24]. Patients experiencing a relapse in their malignancies following intensive front-line chemotherapy or who failed to respond to initial chemotherapy were considered to be in a relapsed/refractory status [22]. The extensiveness of malignancy was classified according to the extent of the tumor and major organ involvement, as reported previously [22, 23]. Disease extent was evaluated as extensive disease and/or major organ involvement. Extensive disease was defined as stage III or IV for lymphoma; metastatic or locally extensive disease for solid malignancies; and >80% blasts in bone marrow, >25,000 blasts/μL in peripheral blood, or the need for leukapheresis for hematological malignancies [22]. Major organ involvement was defined as a pathologically confirmed or radiologically suspected invasion of the brain, heart, lung, liver, or kidney [24]. Neutropenia was defined by an absolute neutrophil count (ANC) <0.5×10³/μL [25, 26]. Therapeutic modification was defined as the addition, change of class, or withdrawal of antimicrobial agents, chemotherapy, or steroids according to identification of an etiology or exclusion of a presumptive diagnosis [27].

The following clinical data on ICU admission were extracted from the medical records: age, gender, type of malignancy, disease extent, recent chemotherapy administration, need for mechanical ventilation, need for renal replacement, or need for vasopressor therapy. Severity of illness was assessed by the Simplified Acute Physiology Score 3 (SAPS 3) [28] and Sequential Organ Failure Assessment (SOFA) [29]. Finally, we documented outcomes of cancer patients with ARF with diffuse pulmonary infiltrates including length of stays in ICU and hospital, ICU and hospital mortality, and 90-day mortality.