Pulmonary Infiltration by CLL/SLL in the Presence of Inflammation: Significant or Incidental Finding?.

Abstract 2352

Poster Board II-329

Radiographically identified pulmonary infiltrates in patients with CLL are usually secondary to infectious causes. However, a subset of patients develop pathologic leukemic infiltrates. Lung biopsies containing both inflammation and CLL infiltrates are particularly problematic. It is unclear in such cases whether the CLL infiltrate represents a nonspecific “passenger effect” secondary to ongoing inflammation versus a pathologic leukemic infiltrate. Furthermore, it is unknown whether there are clinical factors that would help predict who is at higher risk for pulmonary infiltration by CLL. We retrospectively analyzed 45 transbronchial lung biopsies taken from 37 patients at the Cleveland Clinic between 1999 and 2008. We reviewed the biopsies for evidence of acute or chronic inflammation and involvment by CLL. Acute inflammation in our biopsies consisted of acute pneumonias. Cases displaying a lymphocytic infiltrate suspicious for CLL on routine H&E sections were further investigated by immunohistochemistry. The patients studied were 67.5% male with a median age of 67 (45-83) and a median absolute lymphocyte count of 2.75 (0.12 – 86.08). All patients at the time of diagnosis with CLL had flow cytometery of blood and/or bone marrow demonstrating CD5, CD19 positive B cells consistent with CLL. 72% of all patients were previously treated and 52% of patients had a RAI stage > 2 at the time of their diagnosis with CLL. CLL was identified in 20% of the biopsies. Of the 31.1% of biopsies showing acute inflammation, none had leukemia. CLL infiltration was found in association with and without chronic inflammation in 25% and 30.4% of the biopsies respectively. The median peripheral blood absolute lymphocyte count was 3.64 (1.33 – 86.08) in those with leukemic biopsies and 8.31 (0.12 – 79.12) in those without leukemic biopsies. Furthermore, only 33% of the patients with CLL positive biopsies had a RAI stage ≥ 3. At the time of biopsy, 7 previously treated patients were in remission; none of these cases displayed evidence of CLL in lung tissue. We found that the incidence of leukemic infiltration of lung biopsies did not increase in the presence of inflammation. In fact, the incidence of infiltration of biopsies with inflammation was lower than those without inflammation. These results demonstrate that leukemic infiltration on biopsies with inflammation is uncommon, and suggests that this phenomenon represents true, pathologic infiltration of the tissue by CLL and not a “passenger effect”. To the authors knowledge this is the largest study evaluating the association of inflammation and pathologic infiltration of pulmonary tissue by CLL.