Download Genetic susceptibility to Hodgkin`s lymphoma and to secondary

Symposium article
Annals of Oncology 13: (Supplement 1) 30–33, 2002
DOI: 10.1093/annonc/mdf604
Genetic susceptibility to Hodgkin’s lymphoma and to secondary
cancer: workshop report
A. Staratschek-Jox1*, Y. Y. Shugart2, S. S. Strom3, A. Nagler4 & G. M. Taylor5
1
Department of Internal Medicine I, University of Cologne, Cologne, Germany; 2Department of Epidemiology, Bloomberg School of Public Health, Johns
Hopkins University; 3Department of Epidemiology, University of Texas M.D. Anderson Cancer Center, Houston, TX, USA; 4Bone Marrow Transplantation
Department and Cord Blood Bank, The Chaim Sheba Medical Center, Tel Hashomer, Israel; 5Immunogenetics Laboratory, University of Manchester,
St Mary’s Hospital, Manchester, UK
Although the occurrence of familial Hodgkin’s lymphoma (HL) is a rare event, genetic susceptibility as
a cause of HL and its influence on treatment outcome may not be rare. However, results obtained from
the analysis of HL families will probably have broad implications with regard to understanding
common pathogenic factors leading to the development of the disease. The description of anticipation
among the affected offspring of HL patients further strengthens the view that heritable factors contribute to development of HL. Moreover, the finding that particular human leukocyte antigen (HLA)
alleles are associated with susceptibility to HL may be regarded as a hint to the presence of an as yet
undefined infectious agent, leading to the growth of a malignant lymphoma cell clone in those patients
that are more susceptible to this agent due to their HLA genotype. In addition, since an intrinsic
genomic instability was observed in a proportion of HL patients, it is plausible that these patients are not
only susceptible to the causation of HL, but are also at a higher risk of developing therapy-related (TR)
secondary cancers following treatment. Estimation of sister chromatid exchange was established as a
tool to identify patients at higher risk of TR cancer. In this context the use of therapeutic agents known
to increase genomic instability should be carefully considered prior to determing the best treatment. The
future identification of heritable factors contributing to HL will be of importance both with regard to
diagnosis as well as treatment of HL patients.
Key words: heritability, Hodgkin’s lymphoma, risk, secondary cancer, susceptibility
Introduction
Hodgkin’s lymphoma (HL) is a heterogeneous hemopoietic
maligancy, and little is known about the etiology of this
disease. However, an increased risk of developing HL in the
siblings and monozygotic twins of patients has been reported
[1–3]. In siblings <45 years of age, a seven-fold increased risk
of HL was observed [1], but this was absent for older siblings
diagnosed after the age of 45 years. Moreover, cytogenetic
analysis of Hodgkin Reed–Sternberg (HRS) cells, the malignant cells in the HL tissue, has provided evidence for the
constitutive genomic instability of these cells [4, 5]. Chromosomal instability was also observed in peripheral blood lymphocytes obtained from HL patients prior to treatment [6, 7].
Thus it appears that heritable factors may contribute to the risk
of developing HL in certain patients as well as of developing
secondary cancer following therapy for HL [therapy related
(TR) cancer]. These issues were considered at a workshop
held at the Fifth International Conference on Hodgkin’s Lym-
*Correspondence to: Dr A. Staratschek-Jox, ATABIS GmbH,
Joseph-Stelzmann Str. 9, 50931 Cologne, Germany.
E-mail: [email protected]
© 2002 European Society for Medical Oncology
phoma in Cologne in September 2001. The data presented at
this workshop are summarized below.
Familial HL and anticipation
In 1998, based on published cases obtained from the literature,
it was first suggested that anticipation may play a role in
the clinical presentation of familial HL [8]. More recently,
Shugart et al. [9] estimated the heritabilty of HL and tested the
hypothesis of genetic anticipation using a high quality cancer
database of the Swedish population. Heritability was estimated by employing a threshold-liability model, and this gave a
value of ∼28%.
To test the hypothesis of anticipation, the t-test procedure
[9] was used to test whether there is a difference in age of onset
of cancer between parents and offspring who were affected
with HL. A randomization test was carried out to test the
validity of the P values. Overall, the mean age of onset in the
12 pairs was 38.41 years [standard deviation (SD) 14.61] for
parents and 25.58 years (SD 10.91) for children. Based on the
birthdate of the parents, data were divided into two groups:
pre-1930 and post-1930. In the pre-1930 group, the mean age
of onset was 27.2 years for the parents (SD 5.3) and 19 years
31
(SD 5.2) for the affected children. The mean anticipation for
the combined sample is 12.8 years (SD 11.2). The corresponding paired t-test statistic was 3.98 (P = 0.0011) and the valid
P value obtained for the systematic randomization procedure
was 0.0007. Separate estimates for anticipation were 17.5
years (SD 13.2) for the group in which parents were born prior
to 1930, and 8.2 years (SD 7) for the group in which the parent
was born after 1930 (P = 0.011). The associated paired t-test
statistics were 3.25 (P = 0.011) and 2.86 (P = 0.018), respectively. The strength of this study is that Shugart and coworkers.
used a population-based cancer registry with 100% coverage
and a registry of the whole population. However, several
weaknesses need to be pointed out: (i) oversampling of
parents with late disease onset, because the HL parents with
early age of onset may have limited reproductive capability; and
(ii) younger offspring may not be old enough to reproduce at
the time this investigation was carried out. These facts may
have biased the results. However, to some extent, their choice
of study design and statistical approaches have allowed them
to reduce some ascertainment bias in the reported estimate for
anticipation in familial HL. More recently, a second study was
performed which included 56 non-HL (NHL) parent–child
pairs [10]. Results show that the anticipation level was even
more pronounced among the NHL–NHL pairs than in HL–HL
pairs (difference = 1.26 years; P = 0.0003). In the near future,
this study will be expanded to test for anticipation in individuals affected with different types of hematological malignancy. The familial aggregation of various cancers will also be
explored further.
Association of HL with particular HLA
class II genes
Epidemiological evidence has long suggested that HL may
arise as a result of the rare outcome of late exposure to a
specific infectious agent [11]. The possibility that this agent
may be Epstein–Barr virus (EBV) is supported by evidence of
EBV in the involved lymph nodes in ∼50% of patients with
HL, depending on age, gender and subtype [12]. However, the
widespread distribution of asymptomatic EBV infection in the
normal adult population suggests that this virus may not be
the only causative factor, implying that gene-environment
interactions may also be involved. The key role played by the
human leukocyte antigen (HLA) genes in the control of
immune responses to viruses suggests that HLA genes might
be important in the etiology of HL. Evidence showing that the
polymorphic peptide-binding pockets of the HLA class II
antigen-binding groove exhibit specificity for different
immunogenic peptides suggests that the causation of HL
could be defined in terms of HLA class II alleles, or by peptide
pocket profiles. This ‘reverse immunogenetic’ footprint of
sequences binding to HL-associated HLA class II alleles may
enable predictions to be made about the role of EBV in HL.
Some years ago Bodmer et al. [13] reported on the observation of an association of DPB1*0201 with resistance to HL,
and DPB1*0301 with susceptibility to HL. Later, an overall
association between the nodular sclerosis subtype of HL and
HLA class II loci was reported [14], whereby susceptibility
seemed to be influenced by more than one haplotype. More
recently, Taylor et al. [15] performed a detailed molecular
analysis of a hospital-based series of 147 adult HL patients
and 183 controls from Manchester, UK. Their results confirmed the previously documented associations with
DPB1*0301 [odds ratio (OR) 1.43, 95% confidence interval
(CI) 0.86–2.36] and 0201 (OR 0.48, 95% CI 0.27–0.90). However, further analysis revealed both gender and subtype differences in associations with these alleles. Thus, the association
of DPB1*0201 with resistance was only significant in females
with nodular sclerosing (NS) HL (OR 0.15, CI 0.03–0.67),
there being no significant association in males with NS or
non-NS, or in each gender with non-NS. In contrast, the
association with *0301 was significant in females with nonNS (OR 4.67, CI 1.11–19.57), but absent in females with NS
and in males with either subtype. Comparison of the polymorphic amino acids lining the peptide binding groove of *0201
and *0301 revealed that they differ at all positions (11, 57, 69,
76 and 87) except position 36, suggesting that valine at this
position is not a key determinant of susceptibility to HL. Support for susceptibility due to lysine at position 69 derives from
results showing increased susceptibility in patients heterozygous for *0301/0401 (OR 13.2, CI 1.60–10.93); these
alleles share only a lysine at position 69, suggesting that this
may exert a role in the combined susceptibility of these heterozygotes.
One potential explanation for the association of DPB1
alleles with susceptibility and resistance to adult HL is that
they are in linkage disequilibrium (LD) with more strongly
associated DRB1 alleles. To address this question, Taylor and
coworkers DRB1-typed their adult HL series. They found that
DRB1*1501 was associated with susceptibility (OR 1.93, CI
1.21–3.05) and 0101 with resistance (OR 0.38, CI 0.21–0.72)
to HL. However, unlike DPB1 alleles, there was no major
difference in DRB1 susceptibility/resistance with respect to
NS and non-NS subtypes. Next, the two DPB1 alleles (*0201,
0301) were tested for LD with the two DRB1 alleles, and no
significant evidence of preferential associations was found
between the pairs of alleles at the two loci (P >0.05). Although
Taylor and colleagues have not yet been able to examine the
relationship between the presence of EBV and HLA-DPB1
genotype in detail, their preliminary results in a different case
series obtained in collaboration with R. Jarrett, F. Alexander
and colleagues [16] indicate that patients typing DPB1*0101*/
*0301– are more likely to be EBV–, whilst patients typing
DPB1*0101–/*0301+ are more likely to be EBV+. The main
difference in these two alleles is in the amino acids at positions
11, 36 and 57, but further work is needed to clarify this in a
larger series of patients.
32
The tentative conclusion from this work is that different
HLA-DPB1 isotypes bind peptides derived from an infectious
agent that is capable of causing HL in the context of the gender
of the individual, and that this leads to a T cell response to the
agent, which influences the HL subtype that subsequently
develops. Although there is preliminary evidence of differential association with EBV, it is not yet clear whether this is a
causative association.
Sister chromatic exchanges as a predictor for
the development of secondary malignancies
in HL patients
HL survivors face an increased risk of developing second
cancers. Strom et al. [17, 18] evaluated sister chromatid exchange (SCE) in peripheral blood lymphocytes as predictors
of second cancer risk in a cohort of 105 newly diagnosed adult
patients seen at M. D. Anderson Cancer Center, Houston, TX,
USA. They analyzed whether having a high frequency of
pre-treatment SCEs was associated with an increased risk
of developing a second cancer. During the follow-up time
(average 7 years, range 1–13 years), 14 new cancers occurred.
Multivariate Cox regression analysis revealed that high levels
of SCEs [relative risk (RR) 3.9, P = 0.03, 95% CI 1.1–13.7)
and age as a continuous variable (RR 1.1, P = 0.03, CI 1–1.1)
predicted second cancer risk. Histology, stage and treatment
(with alkylating agents versus others) were not associated
with elevated risk. These data suggest that baseline SCE may
be a useful marker to identify patients at increased risk of
developing second cancers. Likewise, as described by Kelly
and Perentesis [19], the detection and characterization of
polymorphisms of drug metabolizing enzymes may be a good
tool for identifying HL patients at risk of treatment-related
complications. HL patients at particular risk of developing
secondary leukemia can be identified using fluorescence in
situ hybridization (FISH) methods in order to detect leukemiaspecific chromosomal aberrations in the developing leukemia
cell clone prior to clinical manifestation. This method is
described in detail by Lillington et al. [20].
Granulocyte colony stimulating factor
influences allelic replication in HL patients
Relapse is the major obstacle for successful autologous stem
cell transplantation (autoSCT) in HL. Event-free survival of
HL patients transplanted during the second achieved complete
remission status or later is 30%. Moreover, survival and
disease-free survival of HL patients with refractory disease or
with resistant relapse is extremely low. Allogeneic SCT (alloSCT) might offer a graft-versus-tumor effect, which may be of
therapeutic potential [21, 22]. However, alloSCT applicability
has been usually limited by conditioning-related toxicity.
Fludarabine-based low-intensity conditioning (LIC) or nonmyeloablative conditioning (NST) is a realistic option, since it
can be administered with very low organ toxicity [23]. Most
LIC transplants are being carried out with granulocyte colony
stimulating factor (G-CSF)-mobilized peripheral blood stem
cells (PBSC). As secondary leukemias are an emerging
problem for HL patients post-autoSCT [24], and as most of the
donors in an alloSCT setting are first-degree relatives of the
patients, Nagler et al. [25] evaluated the possible effect of
G-CSF on genomic stability and changes in the temporal order
of allelic replication in lymphoma patients, as well as in
normal HLA-matched sibling donors of G-CSF-mobilized
PBSC. They used the FISH replication assay, which is based
on in situ hybridization with a fluorescently labeled probe that
identifies either the TP53 or the D21S55 loci, which are highly
associated with hematological malignancies [26]. In this
method, the confirmation of the signal indicates the replication status, thus alleles that replicate synchronously characterize concomitantly expressed genes in the common bi-allelic
mode. In patients with hematological malignancies including
HL and NHL, pre-alloSCT a significantly high level of monoallelic replication of both loci was observed after evaluation of
bone marrow (BM) and peripheral blood (PB) samples. Similarly high levels of mono-allelic replication were observed in
normal donors that received 5 days of subcutaneous injections
of G-CSF (10 µg/kg) for mobilization. Moreover, the effect of
G-CSF could be mimicked in vivo by adding the growth factor
to PB cells in cultures. No increase in mono-allelic replication
could be observed in the control cultures. It was therefore
concluded that the frequency of asynchronous replication of
the TP53 and D21S55 loci is low in PB and BM samples of
normal donors, and high in patients with hematological malignancies including HL. Furthermore, in normal donors treated
with G-CSF in vivo and in vitro, the frequency increases to a
level indicative of mono-allelic replication. This effect is
reversible. Long-term follow-up of the stem-cell donors may,
therefore, be required. The significance of these preliminary
findings has yet to be elucidated.
Concluding remarks
Although it is now well established that HRS cells in the
majority of cases are clonally derived from germinal center B
lymphocytes, the factors contributing to the development of
an HL cell clone remain to be elucidated. Epidemiological
studies suggest that heritable factors are important in terms of
susceptibility. Further investigations will reveal whether a
person can be identified as being at high risk of developing HL
prior to the onset of disease, or at least prior to the clinical
manifestation of the lymphoma clone. Moreover the identification of HL patients that are at high risk for developing
of secondary cancer following chemotherapy may help in
determining the best available treatment as well as the manner
of aftercare following treatment. Further identification of
intrinsic host factors contributing to the development of HL,
as well as TR secondary cancers in HL patients, will thus have
33
broad implications in the future with regard to early diagnosis
or even prevention of the disease, as well as in terms of
management of the disease.
References
1. Grufferman S, Cole P, Smith PG, Lukes RJ. Hodgkin’s disease in
siblings. N Engl J Med 1977; 296: 248–250.
2. Manigand G, Macrez C, Chome J et al. Maladie de Hodgkin
familiale. Presse Med 1964; 72: 1871–1874.
3. Mack TM, Cozen W, Shibata DK et al. Concordance for Hodgkin’s
disease in identical twins suggesting genetic susceptibility to the
young-adult form of the disease. N Engl J Med 1995; 332: 413–418.
4. Weber-Matthiesen K, Deerberg J, Poetsch M et al. Numerical
chromosome aberrations are present within the CD30+ Hodgkin and
Reed–Sternberg cells in 100% of analyzed cases of Hodgkin’s
disease. Blood 1996; 87: 4479–4480.
5. Falzetti D, Crescenzi B, Matteuci C et al. Genomic instability and
recurrent breakpoints are main cytogenetic findings in Hodgkin’s
disease. Haematologica 1999; 84: 298–305.
6. Slavutsky I, de Vinuesa ML, Larripa I et al. Sister chromatid
exchange in malignant lymphomas. Cancer Genet Cytogenet 1984;
13: 153–158.
7. Barrios L, Caballin MR, Miro R et al. Chromosome abnormalities in
peripheral blood lymphocytes from untreated Hodgkin’s patients. A
possible evidence for chromosome instability. Hum Genet 1988; 78:
320–324.
8. Shugart YY. Anticipation in familial Hodgkin lymphoma. Am J Hum
Genet 1998; 63: 270–272.
9. Shugart YY, Hemminki K, Vaittinen P et al. A genetic study of
Hodgkin’s lymphoma: an estimate of heritability and anticipation
based on the familial cancer database in Sweden. Hum Genet 2000;
106: 553–556.
10. Shugart YY, Hemminki K, Vaittinen P, Kingman A. Apparent
anticipation and heterogeneous transmission patterns in familial
Hodgkin’s and non-Hodgkin’s lymphoma: report from a study based
on Swedish cancer database. Leuk Lymphoma 2001; 42: 407–415.
11. MacMahon B. Epidemiology of Hodgkin’s disease. Cancer Res
1966; 26: 1189–1201.
12. Weiss LM, Strickler JG, Warnke RA et al. Epstein–Barr viral DNA in
tissues of Hodgkin’s disease. Am J Pathol 1987; 129: 86–91.
13. Bodmer JG, Tonks S, Oza AM et al. HLA-DP based resistance to
Hodgkin’s disease. Lancet 1989; 1: 1455–1456.
14. Klitz W, Aldrich CL, Fildes N et al. Localization of predisposition to
Hodgkin’s disease in the HLA class II region. Am J Hum Genet 1994;
54: 497–505.
15. Taylor GM, Gokhale DA, Crowther D et al. Further investigation of
the role of HLA-DPB1 in adult Hodgkin’s disease (HD) suggests an
influence on susceptibility to different HD subtypes. Br J Cancer
1999; 80: 1405–1411.
16. Alexander FE, Jarrett RF, Cartwright RA et al. Epstein–Barr virus
and HLA-DPB1-*0301 in young adult Hodgkin’s disease: evidence
for inherited susceptibility to Epstein–Barr virus in cases that are
EBV(+ve). Cancer Epidemiol Biomarkers Prev 2001; 10: 705–709.
17. Strom SS, Hess KR, Sigurdson AJ et al. Evaluation of sister chromatid exchange and chromosome breaks in a cohort of untreated
Hodgkin’s disease patients. Cancer Epidemiol Biomarkers Prev
1997; 6: 291–293.
18. Strom SS, Gu Y, Sigurdson AJ et al. Chromosome breaks and sister
chromatid exchange as predictors of second cancers in Hodgkin’s
disease. Leuk Lymphoma 1998; 28: 561–566.
19. Kelly KM, Perentesis JP. Polymorphisms of drug metabolizing
enzymes and markers of genotoxicity to identify patients with
Hodgkin’s lymphoma at risk of treatment related complications. Ann
Oncol 2002; 13 (Suppl 1): 107–111.
20. Lillington DM, Micallef INM, Carpenter E et al. Fifth International
Symposium on Hodgkin’s Lymphoma Proceedings. Genetic susceptibility to HD and secondary neoplasias: FISH analysis reveals
patients at high risk of developing secondary neoplasia. Ann Oncol
2002; 13 (Suppl 1): 40–43.
21. Gajewski JL, Phillips GL, Sobocinski KA et al. Bone marrow transplants from HLA-identical siblings in advanced Hodgkin’s disease.
J Clin Oncol 1996; 14: 572–578.
22. Nagler A, Slavin S, Naparstek E et al. Allogeneic peripheral blood
stem cell transplantation using a fludarabine-based low intensity
conditioning regimen for malignant lymphoma. Bone Marrow Transplant 2000; 25: 1021–1028.
23. Morecki S, Gelfand Y, Nagler A et al. Immune reconstitution following allogeneic stem cell transplantation in recipients conditioned by
low intensity vs myeloablative regimen. Bone Marrow Transplant
2001; 28: 243–249.
24. Milligan DW, Ruiz De Elvira MC, Kolb HJ et al. Secondary leukaemia and myelodysplasia after autografting for lymphoma: results
from the EBMT. EBMT Lymphoma and Late Effects Working
Parties. European Group for Blood and Marrow Transplantation. Br
J Haematol 1999; 106: 1020–1026.
25. Nagler A, Toren A, Samuel S et al. Effect of granulocyte colony
stimulating factor (G-CSF) administration for stem cell mobilization
on temporal order of allelic replication. Blood 1999; 94: 605a
(Abstr 2693).
26. Levanon D, Negreanu V, Bernstein Y et al. AML1, AML2, and
AML3, the human members of the runt domain gene-family: cDNA
structure, expression, and chromosomal localization. Genomics
1994; 23: 425–432.