Abstract
Lung FDG-PET may help to identify aetiology of pulmonary infiltrates in immunocompromised patients http://ow.ly/S24gC
To the Editor:
Diagnostic strategy in haematology patients with pulmonary infiltrates relies on different approaches [1, 2]. Over the past decade, the use of 18F-fluorodeoxyglucose (FDG) positron emission tomography (PET) has been widely used to detect inflammatory and malignant processes, to evaluate the extent of a malignancy, and to ascertain response to therapy. Hot et al. [3] reported diagnostic contribution of FDG-PET in patient with invasive fungal infection. More recently, Veronesi et al. [4] reported PET findings in the diagnostic work-up of screening-detected lung nodules. PET-computed tomography (CT) was highly sensitive for the differential diagnosis of indeterminate nodules detected at baseline, nodules ≥15 mm and solid nodules [4]. However, FDG-PET has never been evaluated in unselected patients with acute respiratory failure that complicates haematological malignancies.
We report the results from a prospective pilot study during which critically ill haematology patients with pulmonary infiltrates benefited from a FDG-PET in addition to routine diagnostic and therapeutic management. Diagnoses were made mostly by noninvasive tests and bronchoalveolar lavage (BAL) fluid analysis; in addition, six patients had bronchial or pulmonary biopsies. FDG-PET was performed between day 1 and day 5, and completed by two senior radiologists specialised in nuclear medicine. None of the patients was intubated at the time of PET-CT, and all received high flow oxygen or intermittent noninvasive mechanical ventilation. According to the level of maximum standardised uptake value (SUVmax), patients were classified as having normal (0–2), mild (2–4), moderate (4–8) or intense (>8) pulmonary parenchymal metabolic activity. Table 1 summarises individual demographic, clinical and FDG-PET data. No adverse effects were observed after isotope injection. Pulmonary lesion related to lymphoma and fungal or viral infection presented with moderate parenchymal activity, whereas bacterial pneumonia presented with focal and mild lung parenchymal activity. Intense lung parenchymal activity was observed with pulmonary tuberculosis (one patient) and pulmonary Kaposi sarcoma (two patients). Normal FDG-PET was associated with normal lungs but cardiac abnormalities (two patients).
The ability of lung PET-CT with a SUVmax >4 to detect pulmonary infiltration by the malignancy had a sensitivity of 100% and a specificity of 33.33%. Also, with a prevalence of malignant infiltration of 40%, the negative predictive value was 100% and the positive predictive value was 50%. Furthermore, the performances of lung PET-CT with a SUVmax >4 were identical than above to detect pulmonary infiltration by opportunistic infection (which prevalence was also 40%).
This preliminary evaluation of FDG-PET in haematology patients with acute respiratory failure depicts FDG-PET patterns that have not been previously reported. We are not able to provide a specific picture related to a single aetiology. However, since there is a direct relationship between diagnosis and prognosis in these patients [1], we believe that every emerging diagnostic strategy needs to be appraised and evaluated carefully to assess its possible contribution to diagnosis and treatments. Further description of FDG-PET patterns in unselected haematology patients with pulmonary involvement is warranted. This prospective evaluation may contribute to the identification of aetiology of pulmonary involvement, to report on the extent of the inflammatory or infectious process, and also to estimate diagnosis contribution of FDG-PET in patients not identified using conventional diagnostic methods. To do so, future studies will need to assess correlations with conventional CT and morphological and immunological features in cells recovered from BAL fluids. In this study of 20 patients, SUVmax was not correlated with the number and the type of BAL cells. Hence, diagnostic performance of lung PET-CT will need to be adjusted for accompanying inflammatory reaction in the normally aerated lung with respect to their neutrophil activation and on correlation of FDG-PET with conventional CT. Only such an evaluation would help determine whether PET-CT can replace conventional CT and bronchoscopy and BAL fluid cytological analysis. Last, patient follow-up using PET-CT may help correlate lung metabolic activity and respiratory situation, in order to assess whether patients are responsive to anti-inflammatory, anti-tumoural or anti-infectious agents.
In summary, these preliminary descriptive results provide interesting insights on the use of PET-CT in immunocompromised patients with pulmonary infiltrates. However, they raise more questions than they bring answers. Studies are needed to confirm these results on a larger scale. Also, investigations to assess diagnostic performance of PET-CT as compared to a well-accepted gold standard, such as pulmonary biopsy, are warranted.
Footnotes
Conflict of interest: None declared.
- Received November 10, 2014.
- Accepted July 22, 2015.
- Copyright ©ERS 2015