26

SUMMARY OF EPIDEMIOLOGY AND IMMUNOBIOLOGY   RA CARTWRIGHT  
    The “Executive Summary” (page 8) and the section on “Workshop Findings, Recommendation, and Agency Responses” (page 10) stand by themselves.  The “Preface” (page 2) has been written as an introduction to the Proceedings Volume. This summary will address the three major sections that comprise the rest of the volume.
 
     1. “Findings from ATSDR Health Studies;”
     2. “Immunobiology and Molecular Genetics of B-Cell Lymphoproliferative Disorders” and,
     3. “Risk Factors and Epidemiological Considerations for B-Cell Lymphoproliferative
          Disorders.”
  FINDINGS FROM ATSDR HEALTH STUDIES  
    These three original reports from the ATSDR, CDC and FDA formed the basis of this workshop. Lymphocyte phenotypes were determined on a total of 5,868 participants in 10 studies: 3,812 from target areas and 2,056 from comparison areas. Of these, 1,499 participants were age 45 and older: 891 from the target areas and 608 from comparison areas. Most of the participants were white.
 
    The CDC identified 11 of these participants with phenotypic characteristics similar to those seen in early B-cell CLL (increased percentage of B lymphocytes, reduced DIM staining of CD20 and/or CD45 and the increased number or pattern of CD5 B-cells). The average age of these 11 persons with the B-CLL-like phenotypes was 63; the youngest was 47 and the two oldest were 72. Six were males and all were white. The overall prevalence of CLL-like phenotypes among participants age 45 or older was 11/1499 (0.7%). The prevalence of B-CLL phenotypes among target area participants over 45 was 8 of 891 (0.9%); the prevalence among comparison area participants was 3 of 608 (0.5%). The prevalence odd ratio associating the CLL-like phenotype with residents in a target area was 1.8 (95% CI = 0.5 to 6.9).
 
    The major finding in the CDC laboratory is the flow cytometric detection of the B-CLL-like lymphocyte phenotypes in eleven participants. The averaged percentiles for these eleven participants for leucocyte count, absolute lymphocyte count, percent lymphocytes, absolute B-cell count and percent B-cells were 80.8, 93.8, 93.6, 97.6 and 96.1 respectively. These findings stress that even in the setting of a normal WBC, a B-cell expansion can and does occur.
 
    The major finding at the FDA was confirmation of the B-cell lymphocytosis and documenting the presence of monoclonal B-cells in three of three individuals tested for light chain restriction, ie, a B-cell monoclonal lymphocytosis (BCML). In addition, it was noted that the morphology of each indi-vidual varied and upon closer detailed inspection, each immunophenotype also varied. An observed increase in G2M cells suggests the presence of near-tetraploid cells, an early step in neoplastic transformation. This observation needs to be confirmed in a larger number of subjects. This report also raised questions concerning the incidence, natural history, differential diagnosis and what to tell patients and their families regarding a BCML.
  IMMUNOBIOLOGY AND MOLECULAR GENETICS  
    It is important to understand B-cell differentiation (Bauer-Chapter 4). All things being equal, for a 70 year old to have a lymphocyte doubling time (LDT) of 500 days, implies that the original neoplastic event may have occurred as early as age 4. We do not know any of the early events of this neoplastic transformation. However, alpha-beta associated Ig proteins are intact.  One possible early event is the acquisition of the common B-CLL antigen (cCLLa) described by Faguet (Chapter 6), but this remains to be confirmed.
 
    We are not the only laboratories to observe low levels of BCML. Maiese and Braylan (Chapter 5) see it in their referral population. And in order to confirm it, they used a polymerase chain reaction (PCR) based assay for Ig rearrangements. Of interest, they find flow cytometry to be just as sensitive or more sensitive in detecting a BCML. Stetler-Stevenson (Chapter 7) has also shown that flow cytometry is very sensitive in detecting clones in the setting of residual disease. In the largest report to date, Jack and colleagues (Chapter 10) also find an increase in BCML using both flow cytometry and PCR in a population believed to be free of hematological malignancy.
 
    The role of Ig genes have been at the heart of B-CLL for over 20 years.  There was the detection of decreased surface IgM expression followed by light chain restriction. With the advent of molecular techniques, Kipps and colleagues (Chapter 9) pioneered the concept of a restricted B-cell Ig repertoire in B-CLL. These investigators have also developed and applied a novel technique (anchored PCR) to further expand their studies. Bertin and colleagues (Chapter 11) have begun an analysis of both heavy and light chain Ig V-region genes.  Although they find no evidence of a restrictive B-cell repertoire in Santiago, sequence analysis may further clarify their findings in this patient population.
 
    The role of functional studies in this workshop was approached from the standpoint of cytokines. Collins (Chapter 13) suggests using cytokine aberrant receptor expression and flow cytometry as a way of differentiating various lymphoproliferative diseases or disorders (LPD). Fernandes and Raveche (Chapter 15) have provided a comprehensive review of cytokine regulation in B-CLL. TNF-alpha, IL-1b, IL-4, IL-6, IL-7, IL-8, IL-10 and IL-13 are expressed constitutively in varying amounts. This complex network of cytokines is involved in differentiation, disease progression and the progression of apoptosis.  They further suggest that the specific disruption of cytokine genes in B-CLL with antisense oligonucleotides that induce apoptosis, may have therapeutic potential in this disease.
 
    Donnenberg and Donnenberg (Chapter 14) discuss in detail the role of activated large CD4 positive atypical T cells in early B-CLL. They present three hypotheses for the role to T cells: immune surveillance, involving a protective role for T-cells; autoimmune aberrant T cell activation resulting in anemia and thrombocytopenia and the fact that B-CLL might be an autoimmune
disease itself. This latter hypothesis is based on the idea that a self-perpetuating bi-directional stimulatory short circuit between T- and B-lymphocytes gives the B-CLL clone a survival advantage but does not allow it to progress beyond the low sIg stage of development. They conclude that if any of these hypotheses are correct, it may be possible to disrupt the disease process by targeting T cells or the signals that they provide.
 
    Shapiro (Chapter 8) succinctly discusses the etiologic role of infection in a specific low grade NHL (MALT lymphoma) and raises the question whether or not this might be applicable to other closely related lymphoid neoplasm.  The role of chronic antigen stimulation remains to be demonstrated in B-CLL.
  RISK FACTORS AND EPIDEMIOLOGICAL CONSIDERATIONS  
    The lymphoid malignancies generally and CLL in particular provide major challenges for both epidemiologists and biological scientists. The problems are well illustrated by the contributions to this volume. The hitherto lack of understanding of CLL disease progression and the relative paucity of substantive epidemiological studies specifically directed to uncover risk factors for CLL has been a problem. We also need to understand how, if at all, CLL is related to the other lymphoid malignancies, especially NHL.
 
    As Blair et al. (Chapter 16) point out the environmental risk factors for CLL and NHL are numerous but most contribute only weakly to the overall risk. They have concentrated their efforts on pesticide exposures in rural farming areas in the north central US. They also make the important point that if CLL and NHL had similar aetiological risk factors, the most substantial tend to be those associated with immune suppression from artificial sources, or as a result of natural disease progression. With regards to magnetic field exposure, Wartenberg (Chapter 17) concludes that the relative risk is small and may be a result of the exposure assessment methods used.
 
    Marti et al. (Chapter 18) prepared a brief descriptive epidemiology of CLL for this proceedings volume. They note the marked variation in the incidence of CLL both within and between countries compared to the other leukaemia.  There is a gradient from the east to the west hemispherically and from the south to the north in the US. Therefore the highest incidence rates are reported in the north central US and the contiguous central provinces of Canada.  There is a marked male predominance and a steep age-specific rise after the age 30 in both the US and the UK. They also note that a number of studies have suggested that there is an increased incidence of CLL among Jews, particularly those of Russian or Eastern European origin compared to those of Asian or north African origin.
 
    The marked geographical heterogeneity is also seen in NHL and appears to occur in all the populations that have been carefully investigated so far.  However, in the UK, at least some of the variability in CLL distribution may be ascribed to variable investigation rates and indeed chance discoveries of what might be BCML. This makes the descriptive epidemiology of CLL particularly hard to understand at the present time.
 
    The striking familial occurrences and their history are very well described in this volume and form an important contribution for future research. The high incidence of familial B-CLL makes this a potential confounder in epidemiologic studies. Caporaso et al. (Chapter 19) review their experience concerning the NCI Familial B-CLL Registry. The NCI B-CLL Family Registry currently contains 26 kindreds with two or more affected members. The inheritance patterns are varied and complex. Affected sibships; combined first, second generation and affected members in the third generation; and identical male twins. Six of the kindred are of Eastern European origin and of Jewish descent (EEJD). They observed the coexistence of CLL with Hodgkin's disease, myeloma, and hairy cell leukemia in different familial B-CLL kindreds.  The presence of other lymphoid neoplasms was not unexpected. In fact, familial lymphoproliferative disease might be a better term for these kindreds.
 
    Adler (Chapter 20) provides an editorial viewpoint from another perspective, regarding aging as a cause of BCML. However, there must be a very fine line between what is a 'normal' change in cell physiology and hence an aging phenomena and what is an ‘abnormal’ change in cellular activity and hence a disease progression. This forms the basis of an extremely interesting philosophical approach in which CLL and its possible precursors could play a significant role. Much depends on our knowledge of disease pathogenesis here.
 
    With regards to autoimmunity and B-CLL, Groves and Vogt (Chapter 21) conclude that epidemiologic studies to date have been contradictory, with recent negative studies casting doubt on the early reports of a causal association.  In an early study (Conley, CL et al. Medicine (Baltimore) 1980 Spt; 59 (5) 323-34), an increase in autoimmune disease in relatives of patients with CLL was noted. The NZB mouse is a well known model combining autoimmunity (autoimmune hemolytic anmeia and membranous glomerulonephritis) with lymphoproliferative disease.
 
    The papers by Middleton et al. (Chapter 22) and Slade et al. (Chapter 23) represent responses to recommendations made at the workshop. In the Middleton protocol, a surrogate measure of exposure to hazardous waste sites (case distance) will be determined after mapping residential addresses and hazardous waste sites using the GIS system. The GIS converts addresses into geocoded data (latitude and longitude) for which site distance can be generated.  The health-related data have already been collected by the NCI during the 1980's in three separate studies. These three studies (Iowa/Minnesota, Kansas, and Nebraska) collected data on cases of certain hematologic malignancies (including B-NHL, B-CLL and multiple myeloma) and frequency-matched controls. The studies were all population-based. The combination of these two data bases will permit another approach to testing how important the difference between target cases and controls might be for established B-NHL and B-CLL.
 
    The basic ecological approach of Middleton is similar to a series of studies current in the UK and elsewhere based on 'point source' pollution hypotheses.  Various statistical tests, most notably the isotonic regression model of Stone, argue that higher rates of disease occur nearer to sources of possible pollution and fall away with distance from those points. When all sites of similar pollution types are amalgamated, this can become a powerful tool in terms of hypotheses testing.
 
    The Slade protocol is to re-test the eleven cases using a rigorous case definition for BCML. The laboratory findings common to the “original 11” cases were reviewed to establish sensitive screening criteria. In addition to the original 11 participants, individuals (age 40 or above) will be selected retrospectively for the follow-up study based on results of the immune tests from previous ATSDR health studies. The selection criteria will include: (1) 60% or more of B-lymphocytes (CD19+) that are also CD5+;or,(2) an absolute B-cell lymphocytosis (>/=830 cells/ml). These individuals will then be retested for co-expression of CD19 and CD5, kappa-lambda analysis for monoclonality, polymerase chain reaction (“PCR”) analysis of B-cell DNA using primers specific for the variable and J H regions, review of blood film for lymphocyte morphology and cell cycle analysis.
 
    In this workshop, Steinberg (Chapter 25) led a discussion on the ethical considerations that have arisen with the advent of genetic screening for disease.  She points out that consent is needed prior to genetic testing and that, depending on the seriousness of the test, pre and post-test counseling would be required. Marti et al. (Chapter 3) in their discussion raised the issue of what to tell patients. This is of particular concern in the clinical setting of familial B-CLL. We have demonstrated amply that we have the tools both from flow cytometry and molecular biology to define the presence of a BCML.  However, it is like any other laboratory test in clinical medicine ie, it needs clinical correlation. In addition, B-CLL is thought to be a multi-staged process and BCML would be just one of many steps. Given the frequency of BCML, it is unlikely all will progress to overt disease. At this point we can only recommend that such individuals be followed. This testing remains experimental, requires clinical correlation and is best carried out in an investigative protocol with the availability of genetic counseling unless the testing is done anonymously.
  REVISED DEFINITION OF B-CLL  
    One of the major ideas to emerge from the workshop was a reconsideration of the criteria for the diagnosis of B-CLL (Faguet-Chapter 6). With the combination of automated blood cell counters and flow cytometry, it was inevitable that if BCML existed, it would be detected. We would like to propose the following definition for early or pre-clinical B-CLL (Pre Rai Stage 0 or Pre Binet Stage A). (1) the percentage of B-cells be 50% or greater; (2) the absolute B-cell count (absolute CD19 or CD20 count) be 1,000 cell/ml or greater; (3) that greater than 50% of the B-cells are CD5 positive; (4) kappa lambda light chain restriction; and (5) PCR based evidence of Ig gene rearrangement. And the clinical situation should be appropriate, ie, an unexplained lymphocytosis in an asymtomatic individual in whom a hematologic malignancy is not expected. Such conditions as splenectomy, autoimmunity and HIV infections are excluded. With regards to the differential diagnosis, a review of the blood film and certain other CD markers such as CD23 could be added. Ideally such individuals would be followed in an investigative protocol setting (see below).  The further application of molecular techniques will undoubtedly facilitate a more precise molecular diagnosis. It should also further our understanding of early events in this most common leukemia of aged individuals.
  FUTURE STUDIES  
    The challenges emerging from this workshop take the form of future studies, partly for confirmation of the findings but also to understand more of the pathogenic steps leading to CLL and what, if any, are in common with other lymphoid malignancies.
 
    We anticipate that we will see a continued integration of bio- and molecular technologies with epidemiological and clinical studies. Some of the questions that need to be addressed are listed below.
 
     1. Disease Natural History: Markers of disease progression are required urgently. This is
         essential for both a BCML and B-CLL.
 
     2. Molecular fingerprints: Spectral karyotyping (SKY) was recently introduced but has not
         been reported in B-CLL yet (Sci 273: 494-497, July 1996). However, it promises to be an
         important technique and we plan to use this method on Giemsa banded samples from
         affected individuals with familial B-CLL. Comparative genomic hybridization (CGH) has
         been reported for B-CLL (Blood 85(10): 2813-2816, 1995). A systematic analysis of
         genomic DNA in B-CLL via the loss of heterozygosity (LOH) has not been reported, and
         representational analysis display (RAD) studies have not been reported in B-CLL. The
         recent finding of BRAC-2 involvement in B-CLL needs to be confirmed.
 
     3. Model of B-CLL leukemogenesis: B-cell activation in terms of the expression of CD23
         occurs early as does expression of the cCCLa.  The decrease in surface Ig probably also
         occurs early while the down regulation of CD20 expression comes later. Bcl2 expression is
         believed to be increased and IL-10 is constitutively expressed. Cluster analysis suggests
         intra clonal diversity and a combination of cytogenetics and southern bloting suggest that
         clonal evolution does occur in B-BLL. Image analysis of lymphocytes seen on the blood
         film show that as the nuclear optical density increases the cell size gets smaller.  And with
         the proper interval of time, ie, the lymphocyte doubling time, one can imagine waves of cells
         being added to the peripheral blood. These findings need to be integrated with known
         molecular markers of leukemogenesis.
 
     4. NZB murine model: The NZB mouse as noted above is seen as a model of LPD. A large
         number of individuals from an F1 backcross [(NZB x DBA/2) F1 x NZB] have been
         phenotyped and a systematic PCR based linkage analysis is underway. This should lead to
         the definition of loci that are linked to LPD (personal communication, Metcalf and Marti).
 
     5. Flow cytometry: Enlarge the flow cytometric patterns and correlate these with
         morphology and clinical course. A consensus meeting to discuss and define the common
         B-CLL immunophenotype for epidemiological studies is being planned.
 
     6. Genetic markers and family studies: New markers would probably complement
         disease progression markers and be useful for family studies. The collection and analysis of
         more families both purely CLL and with other lymphoid malignancies will require a major
         and important international effort. If 70 to 100 sib pairs with other family members were to
         be assembled, molecular analyses of DNA would enable location of relevant parts of the
         genome for intensive study and reveal common patterns between disease types and give
         other studies (see below) the relevant DNA tests for associations.
 
     7. Epidemiological Study:
 
        (a) Further basic descriptive epidemiology is required along the lines of the data collection
             studies in the UK. With good accurate data, many aspects of CLL can be investigated.
             These would include secular trends and geographical distribution.
        (b) Longitudinal or cohort epidemiological studies on BCML, common and familial B-CLL
             will answer urgent questions regarding disease progression and help to answer questions
             about epidemiological studies.
       (c) The incidence and prevalence of B-cell lymphocytosis (BCL), CD5+ B-cell
             lymphocytosis (CD5 BCL) and B-cell monoclonal lymphocytosis (BCML) will need to
             be known. This sequence may be prototypic for the initiation and progression of B-CLL.
 
    In conclusion, future epidemiologic, clinical and laboaratory investigations promise further insight into the aetiology and pathogenesis of BCL, BCML and B-CLL. This bright future will undoubtedly translate into improved care and understanding of these patients and their families.
   Return To Contents