Correspondence BONE MARROWASSESSMENT IN B-CELL CHRONIC LYMPHOCYTIC LEUKAEMIA: ASPIRATE OR BIOPSY? For more than a decade the bone marrow histological pattern (BMP) has been considered to be a major prognostic factor in B-cell chronic lymphocytic leukaemia (B-CLL) (Rozman et al, 1984; Geisler et al, 1986). In contrast, little attention has been paid to the prognostic value of bone marrow aspirate lymphocytic infiltration (LI), in spite of being one of the French Cooperative Group’s criteria used to subclassify the stage A into A' and A groups (French Cooperative Group on CLL, 1990). However, Montserrat et al (1996) have recently published the results of a study focused on determining not only the merits and prognostic value of BMP, but also those of LI in the evaluation of CLL patients as well. These authors conclude that bone marrow biopsy continues to be the most suitable method to assess bone marrow status in CLL. Nevertheless, they also report some interesting findings regarding the importance of LI in the prognostic assessment of CLL patients, especially of those in early disease (stage A). They found that the survival probability differed significantly according to BMP, as well as to LI, in the entire series and in stage A patients with CLL. Moreover, BMP and LI predicted disease progression of stage A patients. The multivariate analyses showed that BMP entered into the best predictive model for both overall survival and disease progression, but interestingly LI was the only parameter entering into the best prognostic model for survival in stage A patients. In a similar study performed by our group, and recently reported at the Second Meeting of the EHA ( Jarque et al, 1996), we analysed 82 previously untreated patients with CLL in stage A. There were 24 patients (29%) with diffuse BMP and 50 cases (62%) exhibited LI >60%. In our series, patients with diffuse BMP had a shorter survival (median 5 . 2 years) than those with non-diffuse BMP (median 10 years), although the difference was only of borderline statistical significance (P ¼ 006). In contrast, patients with LI >60% had a significantly shorter survival (median 6 . 5 years) than those with LI 60% (median survival not reached) (P ¼ 0003) (Fig 1). In addition, in the multivariate survival analysis, the best predictive model included LI, Rai’s sub- stages, and age, but not BMP. Our results, as those reported by Monserrat et al (1996), emphasize the independent prognostic value of LI, and its superior predictive capacity compared to BMP, in stage A CLL patients. BMP should remain mandatory for patients with more-advanced CLL stages (B and C) and those early CLL patients included in clinical trials. However, an accurate quantification of LI in bone marrow aspirate adds relevant prognostic information and would reduce the need for bone marrow biopsy in clinical practice in most patients with stage A CLL. Department of Haematology, I SIDRO J ARQUE La Fe University Hospital, L UIS L ARREA Valencia, F EDERICO G OMIS Spain G UILLERMO F. S ANZ G UILLERMO M ARTI ´ N J ESUS A. M ARTI ´ NEZ M IGUEL A. S ANZ REFERENCES French Cooperative Group on Chronic Lymphocytic Leukemia (1990) Natural history of stage A chronic lymphocytic leukaemia untreated patients. British Journal of Haematology , 76, 45–57. Geisler, C.H., Ralfkiaer, E., Hansen, M.M. & Hou-Jensen, K. (1986) The bone marrow histological pattern has independent prognostic value in early-stage chronic lymphocytic leukaemia. British Journal of Haematology , 62, 47–54. Jarque, I., Larrea, L., Gomis, F., Sanz, G.F., Martı ´n, G., Martı ´nez, J. British Journal of Haematology , 1996, 95, 754–756 754  1996 Blackwell Science Ltd Fig 1. Probability of survival in 82 previously untreated patients with stage A CLL.