Background. Computed tomography pulmonary angiography (CTPA) may improve the diagnostic capabilities of CT imaging for invasive mold disease, but its performance relative to other signs (ie, halo sign, hypodense sign, pleural effusion, reversed halo sign) is unknown. Methods. We prospectively compared the diagnostic performance of CTPA vs other CT imaging findings in 100 patients with hematological malignancies and possible invasive mold disease defined by EORTC/MSG criteria. After undergoing extensive diagnostic work-up, patients were upgraded to probable or proven mold disease based on galactomannan antigen, culture or histology; or remained as possible mold disease if an alternative diagnosis could not be established. Results. In total, 46 /100 patients who underwent CTPA were upgraded to probable or proven mold disease. Excluding 8 CTPA cases that were nonevaluable by the radiologist, a positive occlusion sign identified by CTPA was 100% sensitive for the diagnosis of probable or proven mold disease (41/41). Among patients who could not be upgraded from the possible mold disease category (n = 51), 25 (49%) had evidence of vessel occlusion by CTPA with only one positive patient eventually reaching an alternative diagnosis (Staphylococcus aureus septic thrombosis). Intravenous and/or oral antifungal therapy was stopped earlier in patients with a negative vs positive CTPA results (P ≤ .001). Conclusions. Vessel occlusion detected by CTPA is a more sensitive and possibly more specific radiographic sign vs other common CT findings of invasive mold disease in patients with hematological malignancies.

An extensive analysis of the reliability of positron emission tomography (PET) after induction treatment in patients with Hodgkin's disease (HD) or aggressive non-Hodgkin's lymphoma (NHL). In all, 75 untreated patients with HD (n=41) or aggressive NHL (n=34) were studied with both PET and CT scans following standard chemotherapy induction therapy (ABVD or MACOP-B) with/without radiotherapy. Histopathological analysis was performed when considered necessary. After treatment, four out of five (80%) patients who were PET⁺/CT⁻ relapsed, as compared with zero out of 29 patients in the PET⁻/CT⁻ subset. Among the 41 CT⁺ patients, 10 out of 11 (91%) who were PET⁺ relapsed, as compared with 0 out of 30 who were PET⁻. The actuarial relapse-free survival (RFS) rates were 9 and 100% in the PET⁺ and PET⁻ subsets, respectively (P=0.00001). All five patients who were PET⁺/CT⁻ underwent a lymph node biopsy: in four (80%) cases, persistent lymphoma and was confirmed at histopathological examination. Two HD patients who were PET⁻/CT⁺ (with large residual masses in the mediastinum or lung) were submitted to biopsy, which in both cases revealed only fibrosis. In HD and aggressive NHL patients, PET positivity after induction treatment is highly predictive for the presence of residual disease, with significant differences being observable in terms of RFS. PET negativity at restaging strongly suggests the absence of active disease; histopathological verification is important in patients who show PET positivity.