Israel Medical Association Journal

Background: Pulmonary infiltrates (PIs) detected in patients with non-Hodgkin lymphoma (NHL) may present a diagnostic challenge due to their wide differential diagnosis, including infection, pulmonary lymphoma and immunochemotherapy-associated pulmonary toxicity.

Objectives: To characterize therapy-associated PIs by positron emission tomography/computed tomography (PET/CT) imaging.

Methods: We conducted a historical analysis of fluorodeoxyglucose-PET/CT (¹⁸F-FDG-PET/CT) PIs in NHL patients treated with combined immunochemotherapy including rituximab. Incidence of PIs, radiological features, patients’ characteristics, underlying NHL type, rituximab/chemotherapy dosing schedules, and symptoms were recorded. Therapy-associated PIs were defined as new or worsening PIs appearing after treatment onset, without evidence of active pulmonary lymphoma or infection.

Results: Among 80 patients who met the pre-specified criteria, therapy-associated PIs were identified in 17 (21%), 6 of whom had accompanying symptoms. Increased FDG uptake was observed in nine, and PI resolution in six. The incidence of PIs was higher in females and in patients with aggressive lymphoma, at advanced stages, and in those who had received treatment consisting of a combination of rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisolone every 14 days (R-CHOP-14).

Conclusions: This characterization of therapy-associated PIs may support the clinician managing NHL patients. Further prospective studies are needed to establish the role of each therapeutic component and the natural history of this phenomenon.

Background: Venous thromboembolism is a well-recognized and relatively frequent complication of malignancy, whereas tumor thrombosis is a rare complication of solid cancers. Correct diagnosis of tumor thrombosis and its differentiation from VTE[1] can alter patient management and prevent unnecessary long-term anticoagulation treatment.

Objectives: To evaluate the contribution of 18F-fluorodeoxyglucose positron emission tomography/computed tomography to the diagnosis of tumor thrombosis and its differentiation from VTE.

Methods: PET/CT[2] scans from 11 patients with suspected tumor thrombosis were retrospectively evaluated. Suspicion arose from positive PET/CT in eight cases, or from findings on contrast-enhanced CT in three patients. Criteria for positivity of PET/CT included increased focal or linear uptake of 18F-FDG[3] in the involved vessel. Findings were categorized as PET/CT positive, or PET/CT negative and compared to contrast-enhanced or ultrasound Doppler, pathology where available, and clinical follow-up.

Results: Eight occult tumor thromboses were identified by PET/CT-positive scans. Underlying pathologies included pancreatic, colorectal, renal cell, and head-neck squamous cell carcinoma, as well as lymphoma (4 patients). Three thrombotic lesions on contrast-enhanced CT were PET/CT negative, due to VTE (2 patients) and leiomyomatosis. Accuracy of PET/CT to differentiate between tumor thrombosis and benign VTE was 100% in this small study.

Conclusions: Contrast-enhanced CT defines the extent of thrombotic lesions, while the functional information from PET/CT characterizes the lesions. It appears that PET/CT may be helpful in the diagnosis of occult tumor thrombosis and its differentiation from VTE.