Abstract 3753

Poster Board III-689

Pulmonary infiltrates (PIs) detected in non-Hodgkins lymphoma (NHL) patients may present a diagnostic challenge. One possible etiology among patients receiving rituximab combined with chemotherapy is rituximab-induced interstitial lung disease. We studied a historical prospective cohort of consecutive patients with NHL receiving combined immunochemotherapy. This study aimed to characterize PIs by PET-CT imaging in patients with suspected rituximab associated findings and to describe their clinical correlates, including patient characteristics, underlying disease, rituximab dosing schedule and symptoms. Patients with lymphoma who received rituximab combined with chemotherapy between January 2007 and May 2009 at our institute were identified. Patients were included in the study cohort if the first (baseline) rituximab dose was administered during the study period and if they had a CT or a PET-CT scan prior to baseline and at least one PET-CT scan thereafter. All scans were analyzed by two nuclear medicine physicians and a radiologist blinded to diagnoses and treatment. Cases of therapy associated PIs were defined as patients with new or worsening PIs, with either alveolar, interstitial or ground glass appearance, with or without FDG uptake and without clinical evidence for pulmonary infection or active lymphoma. A total of 199 NHL patients were identified, 80 of whom met the pre-specified eligibility criteria. NHL categories included diffuse large B-cell lymphoma (DLBCL) (61%), follicular lymphoma (20%), mantle cell lymphoma (10%), MALT lymphoma (4%) and other/non-specified lymphoma (5%). Lymphoma involved the lungs at baseline in 36 % of the patients. Chemotherapy regimens included RCHOP-21 (60%), RCHOP-14 (29%), RCOP (9%) and RCHOP/RDHAP (3%). Therapy associated PIs were identified in 17 patients (21%), 6 of whom had accompanying symptoms. NHL categories included DLBCL (n=14) and low grade lymphoma (n=3) among PI cases. PIs were detected after an average of 3.8 rituximab cycles (range 2-8) and were described as ground glass (n=11), interstitial (n=5) or alveolar (n=6). Resolution of infiltrates in follow-up scans was detected in 6 of the 17 patients. Pulmonary manifestations included cough (n=5), dyspnea (n=3), fever (n=2) and decreased diffusion capacity on pulmonary function tests (n=3). In 4 patients symptoms resolved after treatment with corticosteroids and/or antibiotics and cessation of rituximab and one patient died shortly after improvement of symptoms. FDG uptake was observed in 9 of the cases with PIs. No association between FDG uptake and patient characteristics or symptoms was identified. The incidence of PIs was higher in patients with lymphoma involving the lungs (40% vs. 10%, p=.0009), in patients with aggressive vs. indolent lymphoma (28% vs. 10%, p= .06), in females vs. males (30% vs. 16%, p = .09) and in patients who received R-CHOP-14 vs. RCHOP-21 (26% vs. 18%, p =.5). Similarly, the incidence of symptomatic PIs was higher in patients with lymphoma involving the lungs (17.2% vs. 2.0%, p=.02), in patients who received RCHOP-14 vs. RCHOP-21 (17.0% vs 4.0% p=.07) and in females vs. males (10.8% vs. 4.7%, p= .09). In conclusion, therapy associated PIs were not a rare phenomenon in a cohort of NHL patients treated with rituximab/chemotherapy regimens, accompanying symptoms occurred in a third of them and FDG uptake was noted in over a half. PIs were more common in patients with lymphoma involving the lung, possibly due to targeted activity of rituximab in the affected lungs. Patients receiving RCHOP-14 and those with aggressive lymphoma were more prone to develop PIs. Further prospective studies are required to establish the role of rituximab, to compare between symptomatic and asymptomatic cases and to define independent risk factors. Therapy related PIs should be considered in NHL patients receiving combined rituximab and chemotherapy, especially in those developing PIs in association with the relevant predisposing factors.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.