Abstract
Introduction:
Lung infiltrates(LI) are seen in 30-45% of the patients with Febrile neutropenia(FN) in haematological malignancies.In FN patients with LI, accurate/probable microbiological diagnosis is possible in only 30 % with conventional blood culture and serological tests. BAL increases the microbiological diagnostic yield. The battery of microbiological tests done in BAL fluid analysis is not uniform.We evaluated a staged approach while analysing the BAL fluid sample, in the first stage we performed the routine cultures and antigen-based tests and in patients with negative results, we performed Polymerase Chain Reaction(PCR) tests guided by the radiological findings.
Methods:
This was a prospective observational study initiated after Institutional ethics committee approval and conducted at Tata memorial centre,Mumbai between November 2018 and June 2020.BAL testing was done as per a standard protocol(Sampsonas et al.) in hemodynamically stable patients with Spo2 more than 90% and platelet count above 300x10 9/L.Samples were sent for gram stain & bacterial cultures, ziehl neelsen stain and cultures, fungal stain and cultures and Galactomannan (ELISA) and an extra sample was preserved in an EDTA vacutainer at 10-20 degree C. If none of the initial reports were positive, then stored BAL sample was sent for PCR testing guided by the radiology and clinical picture i.e., with nodular infiltrates(Bacteria,Nocardia,Aspergillus,Mucor,P.Jiroveci,Mycobacterium TB,Atypical Mycobacterium TB) and with diffuse micronodular infiltrates Viruses,Legionella,mycoplasma PCR tests were sent. The causal association of the isolated organism was defined as per AGIHO guidelines(G. Maschmeyer et al.)
The Primary objective is proportion of patients with a confirmed microbiological diagnosis using staged BAL analysis. The Secondary Objectives are proportion of patients who had a change in antimicrobial therapy,Feasibility of doing a Bronchoscopy and Proportion of Patients who develop Major or Minor complications during procedure and the 4 and 12weeks Clinical and Radiological Outcomes.
A sample size of 130 patients was required for incidence of 50%(40-60%) positivity with 10% variation at 95% confidence interval.
Results:
A total of 172 patients were eligible of which 50 patients are not enrolled due to physician discretion in 37 patients,9 lost for followup and 2 refused consent and one patient expired and one palliated and 122 patients are enrolled and of these BAL couldn't be done in 20 patients due to hypoxia,low platelets,poor GCS at the time of performing BAL and finally BAL is feasible in 83.6%(n=102/122) patients. Baseline characteristics of patients are mentioned in Table 1.Median age of the patients was 30 (15-65) years with 69.6% males (n=85/122). A confirmed microbiological diagnosis (G. Maschmeyer et al.) was established in 71.3%( 81 /122 ) of cases.Microbiological results are depicted in Table2.
A change of antimicrobial based on BAL (addition and removal of antimicrobial) was done for 78 patients(63.8%) of which 42 had removal of antibacterial and 11 patients had removal of antifungals.
Among 42 patients who had removal of antibiotics, by the end of 4weeks, 36(85.7%) had clinical response and 34 had radiological response,(4 died and 2 lost for followup).
By the end of 12weeks, 31 patients had sustained clinical and radiological response (2 died, 2 lost to follow-up and 1 progressive disease).
Among 11 patients with removal of antifungals 9 had clinical and radiological response by the end of 4weeks (1 died and 1 non responder) which was sustained at week 12.
Complications of BAL
One patient had a major complication (persistent hypoxia), while minor complications were recorded in 27/122 (22%) (Hypoxia-16, hypertension-8,tachycardia-3) during procedure and in 21/122 (17%) (Fever-8, bleeds-6, tachycardia-5,hypertension-2)upto 24 hours post procedure.
Clinical and Radiological responses as per criteria( Figure 1)
Conclusion:
BAL fluid analysis improves the diagnostic yield in febrile neutropenia with lung infiltrates. This leads to a change in antimicrobials in a significant number of patients. It contributes to improved outcomes in this patient population. The test is feasible in a large majority, is safe and the staged approach helps in optimisation of resources.
Figure 1 Figure 1.
Disclosures
No relevant conflicts of interest to declare.