A Single-Center Study of the Utility of Bronchoalveolar Lavage in Critically Ill Patients With Haematological Malignancy or Stem Cell Transplants

Introduction: The aim of this study was to evaluate the yield of bronchoscopy-guided bronchoalveolar lavage (BAL) and decisions on management of antimicrobials in critically ill patients with hematological malignancy and/or hematological stem cell transplant (HSCT). The safety and tolerance of bronchoscopy were also reported. Methods: A retrospective cohort study was conducted by reviewing health charts of all adult patients with a hematological malignancy and/or an HSCT who were admitted to the intensive care unit and underwent bronchoscopy and BAL over four years from April 2016 to April 2020 at King Abdulaziz Medical City, Riyadh. Results: The cohort included 75 critically ill patients. Of these 75 patients, 53 (70.7%) had HSCT (allogenic 66%, autologous 32.1%, haplogenic 3.8%). Computed tomography of the chest was abnormal in all patients. Predominant findings included airspace abnormalities, ground glass opacities, and others. The positive yield was found to be 20% for bacterial, 22% for viral, 21% for fungal, and other organisms were identified in 2%. Although cytology was not performed in 18 patients, malignant cells were identified on BAL in two patients. While the overall mortality of the cohort was high (46.7%), the vast majority (94.7%) tolerated bronchoscopy and BAL without any complications. However, three patients (4%) developed a pneumothorax and one patient bled and developed the acute respiratory distress syndrome post bronchoscopy. Conclusions: BAL can identify and detect microorganisms directly influencing the clinical care of patients who have received non-invasive diagnostic tests that yielded negative culture results. Bronchoscopy and BAL are generally safe and well tolerated by critically ill patients with hematological malignancy or HSCT.


Introduction
Patients with haematological malignancy or stem cell transplants are at high risk of developing pulmonary complications [1][2][3].These can be life-threatening and are difficult to diagnose.The differential diagnosis of lung infiltrates in this population includes infectious, malignant, and non-infectious causes, such as drug toxicity, graft-versus-host disease, and diffuse alveolar haemorrhage [2].The optimal diagnostic approach for these patients remains controversial, as conventional methods such as blood cultures, sputum samples, and serological tests have low sensitivity and specificity [3].
Bronchoalveolar lavage (BAL) involves the instillation of sterile saline into a segment of the lung.Aspiration of the BAL sample then allows the collection of alveolar fluid and cells.This fluid can be analyzed for microbiological, cytological, and immunological markers.The BAL procedure may be performed blindly but is usually performed during bronchoscopy via a flexible bronchoscope.
Bronchoalveolar lavage is often used for the diagnosis of pulmonary complications in patients with hematologic malignancies or stem cell transplants.Bronchoalveolar lavage can provide direct evidence of causative agents or pathological processes, but it has important limitations.These include its invasiveness, potential complications, cost, and variable diagnostic yield.Jorge et al. found that the results of BAL changed management in 57.4% of the cases [1].Therapeutic .The findings of Feinstein et al. [2] were similar to those of Jorge et al. [1].However, other studies reported that survival did not improve despite any change in treatment after a positive result from the analysis of a BAL or transbronchial biopsy [4].Patients treated for hematological malignancies, in particular those undergoing allogeneic hematopoietic stem cell transplantation (HSCT), are immunosuppressed and have an increased risk of serious infections.Neutropenia, decreased cellular immunity, hypogammaglobulinemia, chemotherapyinduced damage to mucosal barriers, and the frequent use of central venous catheters predispose to such infections [5][6].
There is scarce data about the overall utility of BAL as a diagnostic tool to guide decision-making in this vulnerable cohort and the available data is conflicting.Furthermore, given the trend towards empirical antimicrobial therapy, it is uncertain whether the performance of BAL influences management or changes the outcomes of these patients.Moreover, there are no data on the utility or importance of BAL in patients with hematologic malignancies or stem cell transplantation in Saudi Arabia.
The aim of the study was to evaluate the outcomes and management decisions of patients with hematological malignancies, HSCT or both who developed pulmonary complications and required BAL.

Study setting
King Abdulaziz Medical City, Riyadh (KAMCR) is a large regional academic tertiary care center with a total of 1500 beds and a dedicated haemato-oncology service that has a stem cell transplantation program.There are several specialist intensive care units (ICU) with more than 100 beds.When patients with haematological malignancy or stem cell transplantation require organ support, the threshold for admission to a critical care area is low.When indicated, bronchoscopy and BAL may be performed by either an intensivist or a pulmonologist.

Study design
A retrospective cohort study was conducted after approval from the institutional research board (IRB) at King Abdullah International Medical Center, Riyadh, Saudi Arabia (Approval number IRB/2284/22).

Subjects
Patients who underwent bronchoscopy while admitted to an ICU at KAMCR between April 2016 and April 2020 were identified from the records of bronchoscope usage.The electronic healthcare records of these patients were screened to identify those patients with haematological malignancy and/or an HSCT for inclusion in the study.All other patients who underwent bronchoscopy were excluded.
Clinicians who had experience in the field extracted data from the electronic healthcare records of all adult patients with a hematological malignancy and/or an HSCT who were admitted to an ICU at KAMCR and underwent bronchoscopy and BAL between April 2016 and April 2020.

Data collection
Besides standard demographic data, the authors collected data on smoking status, comorbidities (including infection with the human immunodeficiency virus), type of haematological malignancy, type of HSCT, laboratory blood tests, lung function tests, imaging results, microbial culture results, usage of blood products (e.g., platelet transfusions) and the yield from bronchoalveolar lavage.Diagnoses, changes to treatment, length of stay and mortality were also recorded.These data were used to explore the impact of the BAL on the management and outcomes of the present cohort.

Statistical analyses
Statistical analyses of demographics, laboratory blood results, lung function, imaging, treatments, clinical outcomes, and mortality data have been tabulated.The mean, standard deviation, and range were used to describe continuous data.Frequencies and percentages were used to describe categorical data.All statistical analyses were performed using Excel version 2016 (Microsoft, Redmond, WA, USA).

Results
Table  Continuous data are presented as mean and range while categorical data are presented as frequency and percentage.
HIV: human immunodeficiency virus; HSCT: hematological stem cell transplant.Table 3 shows lung function data and the results of laboratory blood tests.Lung function tests were not performed in 27 patients.In the remaining 48 patients (64%) who had lung function tests, the mean forced expiratory volume over 1 second (FEV1) was 77.4% of predicted, the mean forced vital capacity (FVC) was 76.6% of predicted (mean FEV1/FVC ratio was 95.1%).The mean total lung capacity (TLC) was 74.6% of the predicted.The mean C-reactive protein (CRP) was 144 mg/dl, and the mean white cell count (WCC) was 11.4 × 109/L (neutropenia 27/75, 36%).Platelet transfusions were required for 44% of patients (33/75).Abnormalities were identified on the computed tomography (CT) chest scan in all patients.The predominant CT patterns observed were airspace abnormality (consolidation, airspace abnormality, airspace opacities, and air bronchogram), ground glass abnormality (ground glass and ground-glass opacities (GGO)), and others (tree-in-bud opacities, halo signs, micronodulation, nodular abnormality, and crazy paving).
Nearly half of the patients (48%; 36/75) had more than two different patterns of abnormalities.Table 5 shows the results of the microbiological investigations performed on the BAL.As shown in Table 5, the three main groups of pathogens identified in the BAL were bacteria (

Discussion
The present retrospective study determined that the diagnostic yield of BAL for patients with respiratory diseases was around 60%.This allowed the identification of bacterial, viral, and fungal infections.Pulmonary infiltrates are common in unwell patients with hematological malignancy.This finding is associated with high rates of morbidity and mortality [7][8][9][10][11].The differential diagnosis for the etiology of these infiltrates is broad [12].However, around 70% are infective in nature.
BAL is a well-established method of identifying the cause of pulmonary infiltrates.The routine use of prophylactic antibacterial agents, antiviral agents, and antifungal agents in combination with empirical antimicrobial therapy, before bronchoscopy, may reduce the diagnostic yield [12,13].These observations are similar to the findings in previous studies, showing a diagnostic yield of 31% to 83% depending on the underlying disease [14,15].The BAL cultures were positive for viral and fungal infections in 32 cases (42%).Bacterial cultures were positive in 14 cases (18.6%).Those findings are similar to the study by Ghandili et al. [16].
In contrast, non-invasive diagnostic methods identified only 29 positive microbiological results.These were mainly due to respiratory viral swabs (15 cases) or sputum cultures (nine cases).However, the respiratory viral swabs were not done in 21 cases.Serum galactomannan index was positive in only five cases (6.7%) but was not checked in 14 cases (18.7%) before the BAL.Only one sputum sample was positive for Mycobacteria.
In nearly 80% of cases, the BAL was negative for bacteria, which can explain why empirical antibiotics were not effective in those patients.However, the bacterial cultures may be falsely negative in the setting of ongoing treatment with empirical or pre-emptive broad-spectrum antibiotics that were administered to the majority of the cases in the present cohort.
In almost half of the cases, negative BAL results resulted in the withdrawal of antibiotics and antifungal agents.The positive BAL results also led to a modification in therapeutic regimes in 51 cases (68%), including the addition of antifungal or antiviral agents and the change of antibiotics to one with a narrower spectrum (32%).The negative BAL analysis allowed discontinuation of unnecessary therapies, minimizing the adverse effects and possible development of multi-drug resistant organisms.The findings are consistent with those of previous studies which have reported that bronchoscopy-guided BAL can inform complex antimicrobial stewardship decisions leading to the adjustments of therapies in patients with hematological disease [5,6,[17][18][19].However, the rates of therapeutic interventions reported in previous studies have varied widely, ranging from 5% to 50% [5,6,[17][18][19].
In the present study, malignant cells were identified in the BAL of only two patients.These findings were related to the primary hematological diagnosis.This is similar to previous studies that confirmed pulmonary infections as the primary cause for the pulmonary infiltrates in this cohort of patients [20][21][22].
In the present study, several BAL analyses were non-diagnostic.Therefore, regular re-evaluation of the usefulness and cost-effectiveness of such interventions is important.Bronchoscopy and BAL seemed to be safe in the present cohort.Only three patients had pneumothorax and one developed acute respiratory distress syndrome (ARDS) and bleeding post bronchoscopy.Nearly half of the patients died from their primary pathologies.No deaths were directly attributed to the performance of the bronchoscopic procedure itself.

Limitations
The main limitations of the present study are its retrospective design with the potential for missing data as well as selection bias.The collection of data from a single institution and the relatively small sample size also have the potential for overestimation of the effect.Patient selection may have been affected by the lack of precisely defined criteria for the performance of BAL in critically ill patients with hematological malignancy and or HSCT.The positive yield of BAL might have been closer to the non-invasive diagnostic methods e.g., respiratory viral multiplex PCR on nasopharyngeal swab, had that been performed in all of the cases observed.The lack of a controlled group is also a potential limitation of our study.

Conclusions
Bronchoscopy and BAL are generally safe and well tolerated by critically ill patients with hematological malignancy or HSCT.The bronchoalveolar lavage can detect and identify organisms.It can have a direct impact on the clinical management of patients with negative cultures from non-invasive diagnostic tests.However, the mortality of critically ill patients with hematological malignancy and/or HSCT remains high.
Multicenter randomized controlled trials are required to determine the actual impact of BAL on the management of this complex group of patients.
additions of new antimicrobials to empiric regimens[1]

TABLE 3 : Laboratory blood tests and lung function FEV1
: forced expiratory volume over one second; FVC: forced vital capacity; TLC: total lung capacity.