Statistics from Altmetric.com
We read with interest the article by Ranoet al in the May edition ofThorax.1 This group showed that, by using non-invasive methods (blood cultures, sputum cultures, nasopharyngeal washes and tracheobronchial aspirates), they were able to make a positive diagnosis in 41% of 200 non-HIV immunocompromised patients with pulmonary infiltrates.
We have looked at a similar group of patients treated in the Leukaemia and Lymphoma Units at the Royal Marsden Hospital who needed investigation for pulmonary infiltrates2 and found a low diagnostic yield from bronchoscopy with neither a positive result (36%) nor a change in management (28%) making any impact on survival. Like Rano et al, we are therefore keen to explore the use of non-invasive investigations.
From our literature review we found many groups advocating the use of high resolution CT scanning (HRCT) of the chest. It has found particular use in those patients with a normal chest radiograph but respiratory symptoms and fever,3 having a negative predictive value after bone marrow transplantation of 97%. The HRCT scan has characteristic appearances in aspergillosis,4 Pneumocystis carinii pneumonia (PCP),5 and cytomegalovirus (CMV) infection.6One group has compared HRCT and bronchoalveolar lavage (BAL) in the diagnosis of fungal pneumonia and found the former to be superior.7 From these data we proposed that such characteristic appearances on CT scanning could obviate the need for bronchoscopy.
In the paper by Rano et al 30 isolates (15%) were of A fumigatus and four (2%) were of PCP. Do the authors feel that these patients might have been diagnosed by HRCT scanning and thus have avoided bronchoscopy altogether?
Antifungal treatment was only instigated after 4–5 days if there was no improvement or positive microbiology. With such a policy the mortality rate in those with Aspergillusinfections was 67%. Other groups recommend early introduction of antifungal treatment due to its high prevalence, particularly in haematological malignancies.7 In the light of their results and the potential of early HRCT scanning, have the group changed their practice to include early use of antifungal treatment?
authors' reply We appreciate the valuable comments by Drs Murray and O'Brien regarding our study on immunocompromised patients. In our study the diagnostic yield of non-invasive methods was 41%.1-1 The good results obtained with these techniques, as well as their availability and relatively low cost, make them very useful in the clinical management of immunocompromised patients. In this sense, HRCT scanning, as Drs Murray and O'Brien point out, has a demonstrated role in the diagnosis of some pulmonary complications, particularlyAspergillus pneumonia, and can be included in the diagnostic strategies of non-invasive diagnostic techniques. Whether or not an HRCT scan might facilitate the early start of antifungal treatment and improve the mortality fromAspergillus pneumonia will need to be evaluated in properly designed studies.
Although non-invasive techniques must be considered as a first step in the clinical management of immunocompromised patients, their use cannot justify any delay in performing a bronchoscopic exploration. In this sense, and contrary to the experience reported by Drs Murray and O'Brien, the systematic use of endoscopic techniques allowed us to increase the diagnostic yield up to 81%.1-1 We believe that obtaining an early diagnosis in immunocompromised patients is of paramount importance. In fact, the low diagnostic yield of the 25 bronchoscopies performed in the series by Murray et al and the lack of impact on survival, also reported by other groups,1-2-1-4 may well be due to the fact that the endoscopic techniques are performed too late in the evolution of the disease when the possibilities of influencing the outcome are very few.
Replacing bronchoscopic techniques by HRCT scanning carries the risk of misdiagnosing polymicrobial lung infectious and non-infectious pulmonary complications in which bronchoalveolar lavage and bronchial aspirates have a known diagnostic role. Remarkably, in 11 of the 15 polymicrobial infections reported in our series,1-1 bothAspergillus and/or cytomegalovirus (CMV) were involved. In these cases the use of HRCT scanning instead of bronchoscopy would probably have missed one of the other co-existent pathogens. We therefore do not believe that HRCT scanning of the chest can replace bronchoscopy in the clinical management of immunocompromised patients with pulmonary infiltrates.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.