Results
We recruited 38 patients: 21 without signs of infection disease and 17 with signs of infection,23 (60%) male and 15 (40%) female. Diagnosis was acute leukemia (AL) in 22, brain tumor in 5, lymphoma in 4, sarcoma in 3 (1 renal sarcoma (RS),1 rhabdomyosarcoma (RMS), 1 osteosarcoma (OS)), Wilms tumor (WT) in 2, neuroblastoma (NBL) and desmoid fibromatosis in 1 patient. The mean age was 9.2 years and the median 9 years.
All patients underwent LUS, 16 underwent CXR, 3 chest CT.
Group 1 included 8 cases (21%) of non-infected non-neutropenic patients, mean age 10.3 years, median 9.5 years.
The 8 LUS were negative. CXR was performed in 4 cases, as part of the initial evaluation at onset of cancer, and was negative in all cases. (Table 1)
Group 2 included were 13 (34%) non-infected neutropenic patients, the mean age was 7.1 years, the median age 5 years; in only one case, ultrasonography revealed rare B lines and bilaterally subpleural air bronchograms; according to CXR images these findings were probably related to leukemic infiltrates present at the onset of the disease. In this group there was an agreement between ultrasound and CXR of 83%. Particularly, the finding on x-ray of a nuanced hilar thickening, in a patient with onset of acute lymphoblastic leukemia, afebrile and with normal CRP, was not evidenced on LUS. (Table 1)
Group 3 recluted 7 infected non-neutropenic patients (19%), mean age 11 years, median 9 years. Three patients (42%) presented with respiratory symptoms; two initially asymptomatic patients subsequently presented respiratory symptoms. (Table 1) Of the seven LUS performed, 5 were positive and 2 were negative.
6 CXR were performed: 4 were negative and 2 were positive. In case 25 a second CXR was repeated after the onset of respiratory symptoms and turned out positive.
Between the ultrasound examinations and the radiographic examinations there was agreement in 67% of cases. In this group, 3 pathological chest CT scans were performed with 100% concordance LUS. In one of the three cases (33%) chest CT was discordant with CXR.
LUS images were pathological, in the absence of initial respiratory symptoms, which appeared in the following 24 hours with a rapid decline in lung function in two cases. In these two cases (patient 25 and 26) the negativity of the radiographic examination can be considered as false negative, in fact the patients presented both pathological ultrasound images and afterwards clinical symptoms compatible with lung infection. In both cases, for clinical worsening and the onset of respiratory symptoms, CXR (patient 25), repeated 4 days later, and chest CT (patient 26) performed one day after the CRX execution, were positive for pneumonia. (Fig. 1)
Group 4 included 10 infected neutropenic patients (26%), mean age 8.5 years, median 10 years. We found respiratory symptoms only in 1 case. The results of laboratory and instrumental tests are shown in Table 2.
In 50% of cases the LUS was found to be pathological.
None of the 10 patients underwent a CXR and in no case was it necessary to perform a chest CT for the benign course of the infection.
Between group 3 and 4, 10 patients presented positive LUS and underwent ultrasound follow-up at three and seven days from first LUS showing resolution of the pathological signs in 6 cases. In the 4 cases with persistent pathological images at day seven, a further LUS was performed after 1 month, with resolution of pneumonia in all cases.