The Hematology Journal (2002) 3, 7 ± 9 ã 2002 The European Haematology Association All rights reserved 1466 ± 4680/02 $25.00 www.nature.com/thj
COMMENTARY
Classical and new prognostic factors in chronic lymphocytic leukemia: Where to now? Emili Montserrat*,1 1
Department of Hematology, Hospital Clinic, University of Barcelona, Barcelona, Spain
The Hematology Journal (2002) 3, 7 ± 9. DOI: 10.1038/sj/thj/6200139 Keywords:
chronic lymphocytic leukemia; CLL; prognosis; prognostic factors
Chronic lymphocytic leukemia (CLL) has a variable clinical course. Although the median survival of patients with this form of leukemia is around ten years, in individual patients the prognosis is extremely variable, ranging from a very short to a normal lifespan. The clinical staging systems independently developed by Rai et al.1 and Binet et al.2 in the early 1980s, based on easily obtainable biological and clinical parameters, are extremely useful for assessing prognosis in patients with CLL. These staging systems not only facilitate the treatment of patients according to individual prognosis, but also make it possible to conduct and to compare trials based on the risk of the disease. Although easy to apply and reproducible, staging systems are not devoid of limitations. For example, the mechanisms accounting for cytopenias (ie, bone marrow in®ltration, hypersplenism, autoimmune basis) are not taken into consideration. As a result, patients are classi®ed as being in advanced stage, independently of the origin of the anemia or thrombocytopenia. Yet there is some indication that the mechanism of the cytopenia is also a factor in prognosis.3 Most important, perhaps, is the prognostic heterogeneity within each of the clinical stages. Thus, staging systems do not allow identi®cation of individuals in early stage (40 ± 60% of all patients) who are likely to progress and those in whom the disease will remain stable for many years. This is important since it is possible ± but not yet proven ± that patients in early stage who are likely to progress, could bene®t from being treated immediately after diagnosis, before progression occurs. Conversely, *Correspondence: Prof. Emili Montserrat, Department of Hematology, Hospital ClõÂ nic, University of Barcelona, Barcelona, Spain; Tel/Fax: +34 93 227 54 75; E-mail:
[email protected] Received 31 August 2001; accepted 15 September 2001
some patients presenting in advanced stage run an indolent course and may not require therapy for long periods of time; staging systems do not identify these patients either. Since the introduction of the clinical staging systems there has been a continuous eort to identify new prognostic factors in CLL, as evidenced by the many articles on that issue. Thus, in a MEDLINE search, 119 articles dealing with `Prognostic factors in CLL' can be identi®ed. The reader interested in a comprehensive review of prognostic parameters in CLL is referred to some of the many reviews recently published on this subject.4 With such a plethora of information, how does one separate the wheat (ie, the relevant prognostic factors) from the cha? In this regard, it is important to keep in mind the conditions that prognostic factors should ful®ll, the most important of which are shown in Table 1.5,6 Besides clinical stages, classical parameters with proven independent prognostic value include: the number of lymphocytes in peripheral blood; the degree of bone marrow in®ltration; the proportion of atypical lymphoid cells in peripheral blood; and the lymphocyte doubling time (Table 2). These parameters are simple, reproducible and add prognostic power to clinical stages. Nevertheless, it should be noted that these prognostic factors were identi®ed at a time when CLL treatment was based on alkylating agents. In most of the trials in which the newer and more eective Table 1 Requirements of prognostic marker for acceptance in clinical practice 1. Determination is reproducible and widely available and with quality control 2. Substantial predictive value beyond recognized prognostic systems is demonstrated 3. The predictions have therapeutic implications that are readily interpretable by the clinician and of bene®t to the patient 4. Conclusions are based on independent con®rmed phase III studies
Prognosis in CLL E Montserrat
8 Table 2 Prognostic factors in chronic lymphocytic leukemia Parameter
Favorable
Unfavorable
Classical Clinical stage Binet A B, C Rai 0 I, II, III, IV Bone marrow in®ltration Biopsy Non-diuse pattern Diuse pattern Aspirate 480% lymphocytes 480% lymphocytes WBC count (x 109/l)* 450 450 Prolymphocytes in 410 410 peripheral blood (%)* Lymphocyte doubling 412 months 412 months time New Serum markers** Cytogenetics C38 expression IgVH
normal normal del(13q)isolated 430% mutated
increased del(11q) del(17p) 430% unmutated
*Continous variables. **LDH, b2 microglobulin, thymidine-kinase, CD23 and others.
treatments for CLL have been investigated, most of these prognostic factors have not been regularly tested. This is unfortunate because prognostic factors may change as treatments become more eective. Whether classical prognostic factors apply to patients treated with the newest and more eective treatment regimens should be investigated. Regarding new variables (Table 2), prognostic value has been claimed for many serum markers (eg LDH, thymidine-kinase, b2-microglobulin, CD23, CD44). Most of these markers, however, cannot be employed in routine practice due to the lack of standardized methods for their measurement, which often results in dierent prognostic cut-o levels. This limitation, however, does not seem to apply to serum LDH or b2-microglobulin levels, for which the cut-o level is the upper normal value.7 In addition, recent studies have shown that cytogenetic abnormalities are important prognostic factors. Patients with del(11q) or del(17p) progress rapidly, respond poorly to therapy, and have short survival. In contrast, those with normal cytogenetics or isolated del(13q) have a good prognosis.8 However, to detect cytogenetic abnormalities,
FISH techniques, which are expensive and not widely available, are required. Finally, CLL has long been considered a homogeneous disease of naõÈ ve CD5+ B cells, pre-germinal cells not exposed to antigenic stimulation and which, consequently, do not undergo mutations of the IgVH genes. It is now apparent, however, that neoplastic CLL lymphocytes have either unmutated or mutated IgVH genes. In 1999, two dierent groups independently showed that patients with unmutated IgVH gene (type I, naõÈ ve or pregerminal CLL) have a poorer prognosis than those displaying mutated IgVH gene (type II, memory or post-germinal CLL).9,10 It is of note that the mutational IgVH gene status dierentiates two risk groups, even among patients in early stage. These ®ndings, of outstanding importance, have subsequently been corroborated in other studies. Unfortunately, most laboratories are currently unable to isolate and characterize IgVH gene sequences, and this poses a major limitation to the use of this feature on a routine and broad basis. Identifying a simple surrogate for the IgVH gene mutational status would be, therefore, extremely useful. In this regard, there is some correlation between the expression of the CD38 antigen on neoplastic lymphocytes and IgVH gene mutations, but this is far from being absolute. Interestingly enough, CD38 expression appears to have prognostic value by itself. In the light of these facts, how should the prognosis of CLL patients be assessed? If at all possible, the most relevant parameters correlating with prognosis (Table 2) should be evaluated prospectively in trials. Retrospective studies performed on stored tumor samples would also be useful; these would also provide information more rapidly than prospective studies. Furthermore, eorts to either simplify the analysis of IgVH gene mutations or to identify a simple and reliable surrogate for such alterations should be made. At the same time, new prognostic factors (eg, cytogenetics, serum markers, CD38 expression) should be fully standardized. All this should lead to the identi®cation of the most relevant prognostic factors in CLL. Meanwhile, the classical prognostic factors ± particularly clinical stages ± should continue to be the basis for assessing prognosis, as well as the benchmark with which new prognostic parameters should be compared.
References 1 Rai KR, Sawitsky A, Cronkite EP, Chanana AD, Levy RN, Pasternack BS. Clinical staging of chronic lymphocytic leukemia. Blood 1975; 46: 219. 2 Binet JL, Auquier A, Dighiero G, Chastang C, Piguet H, Goasguen J et al. A new prognostic classi®cation of chronic lymphocytic leukemia derived from a multivariate survival analysis. Cancer 1981; 48: 198. 3 Mauro FR, Foa R, Cerretti R, Giannarelli D, Coluzzi S, Mandelli F, Girelli G. Autoimmune hemolytic anemia in chronic lymphocytic leukemia: clinical, therapeutic, and prognostic features. Blood 2000; 95: 2786. 4 Zwiebel JA, Cheson BD. Chronic lymphocytic leukemia: staging and prognostic factors. Semin Oncol 1998; 25: 42. The Hematology Journal
5 Simon R, Altman DG. Statistical aspects of prognostic factor studies in oncology. Br J Cancer 1994; 69: 979. 6 Justice AC, Covinsky KE, Berlin JA. Assessing the generalizability of prognostic information. Ann Intern Med 1999; 130: 515. 7 Molica S. Prognostic value of biological variables in Bcell chronic lymphocytic leukemia. Can we improve upon clinical parameters? Haematologica 1997; 82: 705. 8 DoÈhner H, Stilgenbauer S, Benner A, Leupolt E, Krober A, Bullinger L et al. Genomic aberrations and survival in chronic lymphocytic leukemia. N Engl J Med 2000; 343: 1910.
Prognosis in CLL E Montserrat
9
9 Damle RN, Wasil T, Fais F, Ghiotto F, Valetto A, Allen SL et al. Ig V gene mutation status and CD38 expression as novel prognostic indicators in chronic lymphocytic leukemia. Blood 1999; 94: 1840.
10 Hamblin TJ, Davis Z, Gardiner A, Oscier DG, Stevenson FK. Unmutated immunoglobulin VH genes are associated with a more aggressive form of chronic lymphocytic leukemia. Blood 1999; 94: 1848.
The Hematology Journal
