Leukemia (2002) 16, 632–635  2002 Nature Publishing Group All rights reserved 0887-6924/02 $25.00 www.nature.com/leu Significance of the levels of bone marrow lymphoid infiltrate in chronic lymphocytic leukemia patients with nodular partial remission R Oudat 1 , MJ Keating 2 , S Lerner 2 , S O’Brien 2 and M Albitar 1 1 Department of Hematopathology, University of Texas MD Anderson Cancer Center, Houston, Tx, USA; and 2 Department of Leukemia, University of Texas MD Anderson Cancer Center, Houston, TX, USA Patients with chronic lymphocytic leukemia (CLL) are con- sidered in nodular partial remission (nPR) when they are in remission but bone marrow biopsies show rare nodules. The significance of the level of residual disease in nPR is not known. We studied 91 previously untreated CLL patients who were treated with fludarabine alone, fludarabine with predni- sone, or fludarabine with cyclophosphamide and achieved nPR at the end of six courses. We compared bone marrow lymphoid infiltration before therapy and at the end of three and six courses of therapy as evaluated by a pathologist in retrospec- tive fashion with that of the routine evaluation at the time of performing bone marrow biopsy. We then compared these results with those obtained by computer-aided histomorpho- metry in 28 patients in nPR. There was significant correlation (P  0.05) between pathologists as well as between pathol- ogists and histomorphometry. Upon correlation with clinical characteristics, there was significant correlation (P  0.01) between marrow involvement before therapy and white blood cell counts (WBC), hemoglobin (Hgb), absolute lymphocyte counts, and 2-microglobulin (2-M) but none of these para- meters correlated with the lymphoid infiltrate at the end of three or six courses of therapy. More importantly, lymphoid infil- tration after three and six courses did not correlate with time to progression (TTP) or overall survival (OS). However, patients with 70% marrow involvement before therapy had a signifi- cantly shorter TTP (P = 0.02). All 91 patients showed similar results. However, we found reverse correlation between mar- row lymphoid infiltrate at the end of three courses and OS (P = 0.01). Leukemia (2002) 16, 632–635. DOI: 10.1038/sj/leu/2402439 Keywords: bone marrow; chronic lymphocytic leukemia; nodular partial remisions; histomorphometry; lymphoid infiltrate Introduction Chronic lymphocytic leukemia has a variable course and a wide range of survival time. The survival of some patients may be as short as a few months while others may live more than 25 years after diagnosis and die of unrelated causes. 1–4 The variability of the clinical course and survival prediction is important for the treatment decision and follow-up. Much pro- gress has been made over the past years in our understanding of therapy of CLL, particularly after the introduction of purine analogues. 5–16 Despite the fact that CLL is an incurable hema- tologic malignancy, complete remission can be achieved with conventional chemotherapy. 14–19 In 1996, The National Can- cer Institute-sponsored Working Group (NCI-WG) on CLL defined patients with complete remission (CR) and persistent bone marrow nodules as nPR. 20 The main goal of this study is to investigate the significance of the levels of the lymphoid infiltrate in nPR CLL patients and its correlation with clinical progression and overall survival. To ensure accurate esti- Correspondence: M Albitar, Hematopathology Program, Division of Laboratory Medicine, 1515 Holcombe Boulevard – Box 72, Houston, TX 77030–4095, USA; Fax: 713-794-1800 Received 4 May 2001; accepted 20 December 2001 mation of residual disease in patients with nPR, we first com- pared routine morphologic evaluation with histomorphometry as determined using computer-aided measurement of residual lymphoid tissue in 28 patients with nPR. We correlated the bone marrow infiltrate with TTP and OS in this group of patients. After establishing that the routine estimation of residual disease by pathologists is comparable to that by histo- morphometry, we extended our study to include an additional 64 patients (total of 91 patients) and evaluated the significance of the level of residual disease in overall outcome in this group of patients. Materials and methods Bone marrow biopsies from 91 previously untreated patients, seen at MD Anderson Cancer Center between 1985 and 1999, with CLL who achieved nPR were studied at the start of treat- ment, after three courses of chemotherapy, and after six courses of chemotherapy. Patients were selected because they were previously untreated. They were treated uniformly with minor variations that did not alter CR and OS. 12,21 The patients were treated on three different protocols: fludarabine alone (30 mg/m 2 intravenously daily for 3 days), fludarabine with prednisone (30 mg/m 2 /d for 5 days), and fludarabine with cyclophosphamide (300–500 mg/m 2 daily for 3 days). 12,21 Specimens of the 28 patients were independently evaluated at three time intervals: at the time of obtaining the specimen (P1) as a routine evaluation, in retrospective fashion (P2) where all cases were reviewed again blindly, and by com- puter-aided histomorphometry through mixed (interactive- automatic) image analysis. The histomorphometry mixed image analysis was carried out by digitizing and then quan- tifying the amount of lymphoid infiltrate in each image using the NIH Image program (version 1.62). All histomorphometric evaluations were performed blindly. According to the method suggested by Rywlin, 23 all specimens were classified into four groups: nodular pattern, interstitial pattern, mixed pattern and diffuse pattern. All differential counts were based on counting 500 cells on four different smears. Diagnostic criteria for CLL were based on the NCI-WG guidelines for diagnosis of CLL. 20 All the patients fulfilled these criteria. Pretreatment evalu- ations included medical history, physical examination, com- plete blood cell count and platelet count using automated cell counters; 2-microglobulin, creatinine, bilirubin, immuno- globulin, albumin, magnesium, sodium, potassium, chloride, transaminase, lactic dehydrogenase, glucose, and blood urea nitrogen levels were taken using automated standard bio- chemistry instruments. Bone marrow examination was perfor- med on all patients. Immunoglobulin gene rearrangement analysis using Southern blot analysis was done for each patient. Immunophenotyping was performed using flow cyto- metry and included CD19, CD5, CD11c, CD22, CD3, CD4, CD8, CD23, CD79b, kappa, lambda, CD20, CD103, and