Welcome international healthcare professionals

This site is no longer supported and will not be updated with new content. You are welcome to browse and download all content already included in the site. Please note you will have to register your email address to access the site.

You are here

Pathology, pathogenesis and molecular genetics of follicular NHL

Best Practice & Research Clinical Haematology, 2, 24, pages 95 - 109

Follicular lymphoma (FL) is a germinal centre-derived indolent B-cell lymphoma representing the second most common Non Hodgkin lymphoma in the Western world. This chapter focuses on the pathology of FL and summarizes the current knowledge about genetic and molecular features that are relevant for the pathogenesis of this neoplasm. The translocation t(14;18) is present in approximately 90% of FL leading to the upregulation of the anti-apoptotic protein BCL2, that may constitute a promising molecular target for therapeutic approaches. FL lacking the t(14;18) also exist, and B-cells carrying the t(14;18) can be detected in a subset of healthy individuals. In addition to the t(14;18), secondary genetic alterations are present in most FL and, more recently, deeper insights into the methylation and microRNA expression patterns in the tumour cells have been gained. The tumour microenvironment appears to be particularly important for the biology and the clinical course of FL.

Keywords: follicular lymphoma, pathology, genetics, microenvironment.

Pathology and pathogenesis of follicular lymphoma

Approximately 95% of all malignant Non Hodgkin lymphomas are – at least in the Western hemisphere – of B-cell origin. Among more than 30 different entities and subtypes of B-cell Non Hodgkin lymphomas (B-NHL), follicular lymphoma (FL) constitutes the second most common B-cell malignancy. This chapter attempts to provide an overview of the pathological features of FL, the current knowledge about genetic and molecular genetic events underlying its pathogenesis and prognostic and predictive markers that may help to individualize treatment decisions in FL patients in the future based on defined aberrantly activated or inactivated and potentially ‘druggable’ molecular pathways of the lymphoma.

Clinical appearance of follicular lymphoma

In general terms, FL is considered an indolent lymphoma with a clinical evolution that is characterized by slow progression over many years. However, the clinical course of FL patients can be surprisingly variable and, accordingly, treatment options range from a ‘watch and wait’ approach to aggressive therapy including high dose chemotherapy and stem cell transplantation. In the pre-rituximab treatment era, the median overall survival time of patients with FL ranged from eight to ten years. There is good evidence, however, that modern, more specialized therapeutic approaches including monoclonal antibodies and novel agents have prolonged survival times considerably [1] and [2]. The median age of FL patients at the time of diagnosis is 59 years with a male to female ratio of 1:1.7 [3] . In the paediatric patient population, FL is exceedingly rare and predominantly affects males.

FL involves generally the lymph nodes, but also the spleen, bone marrow, peripheral blood and Waldeyers’s ring, and the majority of patients present with stage III/IV disease at the time of diagnosis [3] . More recently, it has become apparent that FL can also arise at extranodal sites, e.g. in the small intestine, and some of these lymphomas may differ in their molecular properties and also in their clinical behaviour from their nodal and systemic counterparts. Transformation from FL into an aggressive lymphoma, usually diffuse large B-cell lymphoma (DLBCL), takes place in 25–30% of cases. For more detailed information on potentially underlying molecular and genetic events of this scenario the reader is referred to Chapter 5 of this book. In the following paragraphs, we will focus on non-transformed FL of grades 1, 2 and 3A, whereas FL grade 3B is the subject of Chapter 2.

Morphology, immunophenotype and pathogenesis of follicular lymphoma

Generally, B-cell neoplasms represent clonal proliferations of B-cells that frequently mimic “frozen” stages of normal B-cell development in many aspects. Specifically, most B-NHL are of germinal centre (GC) origin, as is FL, or of post-GC origin. This is not surprising given that, physiologically, GCs represent anatomical sites of vigorous proliferation of antigen-triggered B-cells combined with extensive DNA modification steps including ongoing somatic hypermutation of the immunoglobulin heavy chain genes (ongoing SHM) and immunoglobulin class switch recombination (CSR) [3] . Mimicking the situation in normal GC B-cells, FL usually arise in close vicinity to follicular T-cells that express CD3, CD4, and the more specific follicular T-helper cell markers CD57, PD1 and CXCL13, as well as to follicular dendritic cells (FDCs) and varying numbers of histiocytes [4] . The preservation of important morphological features of physiological GCs in concert with the observed preservation of the overall structure of the B-cell receptor in approximately 90% of FL cases and the generation of specific carbohydrate-linking motifs by the hypermutation machinery suggest that stimulation of the B-cell receptor by antigen might play an important role in the pathogenesis of this entity [5] .

Morphologically, FL infiltrates replace the physiological lymph node architecture by neoplastic follicles that harbour atypical germinal centres and frequently lack clear-cut mantle zones. The neoplastic follicles are composed of randomly distributed centrocytes and centroblasts and often display a monomorphic appearance due to the lack of characteristic ‘starry sky’ and zonation patterns of reactive GCs that are divided in a highly proliferating dark zone and a less proliferating light zone ( Fig. 1 ) [3] . FL can be graded into grades 1, 2, 3A and 3B according to the number of blasts and the presence or absence of centrocytes. In grade 1 and 2 FL, the number of centroblasts does not exceed 150 per 10 high power fields. Grade 3 FL present with more than 150 centroblasts per 10 high power fields. While centrocytes are still present in FL 3A, FL 3B are entirely composed of (large) blastic cells (centroblasts).

gr1

Fig. 1 Morphology and immunophenotype of FL. Compared to reactive lymph nodes, the neoplastic follicles are less well defined and frequently lack a mantle zone and the typical ‘starry sky’ pattern (red arrow heads) (A, D). In contrast to normal germinal centres, 90% of FL stain positive for BCL2 due to the presence of the t(14;18) (B, E) and proliferate less as indicated by a reduced Ki67 immunostaining (C, F).

On the molecular level, FL is characterized by ongoing somatic hypermutations of their immunoglobulin heavy chain variable (IgVH) genes. Most of the cases express the GC markers CD10 and BCL6 [3] . More recently, IRF8 was identified by Martinez and colleagues as a novel GC marker that was found to be highly expressed in germinal centre derived lymphomas such as FL, to a lesser extent in pre-germinal centre derived lymphomas (e.g. mantle cell lymphoma) and completely absent in post-germinal centre neoplasms such as multiple myeloma [6] .

Remarkably, in contrast to normal GCs that lack BCL2 expression, the neoplastic GCs in FL stain positively for BCL2, in most cases as a consequence of the translocation t(14;18)(q32.3;q21.3) ( Fig. 1 ). This chromosomal alteration is present in approximately 90% of FL grades 1 and 2 cases and can be viewed as the initial genetic hit in the pathogenesis of FL. FL cases that lack the t(14;18) and thus – at least in part – also lack BCL2 expression may be therefore misdiagnosed as follicular hyperplasia. Notably, a proportion of FL cases in fact do express BCL2, but the protein cannot be detected with some of the commercially available antibodies due to mutations in the antibody binding site in the BCL2 gene [7] . Although FL is generally viewed as a germinal centre-derived B-NHL, the translocation t(14;18) likely arises at an earlier step in B-cell development during VDJ recombination of the IgV genes in the bone marrow [3] . The VDJ recombination process involves recombination signal sequences (RSS)-guided double strand breaks that are initiated by two recombinant activating genes (RAG1 and RAG2) and subsequently resolved by the non-homologous end-joining repair apparatus [8] . Errors in this process may result in translocations such as the t(14;18) in FL that involve the IgH locus [9] . The breakpoints at the IgH locus (14q32.33) predominantly occur within the joining elements of the heavy chain gene locus, and the breakpoints at the BCL2 locus (18q21.3) are located within the major breakpoint region, the minor cluster region or dispersed in the intermediate cluster region [10] . Alternative translocations of the BCL2 gene locus with the immunoglobulin light chain genes (IgL, IgK) resulting in the translocations t(2;18) or t(18; 22) that also lead to an overexpression of BCL2, are exceedingly rare in FL [3] .

Although many FL show a predominantly follicular growth pattern (i.e. more than 75% of the infiltrate grows in atypical follicular structures), some FL have a follicular and diffuse growth pattern, and a small subset even grows in a predominantly diffuse fashion (defined as an infiltrate with less than <25% follicular growth pattern, see Fig. 2 A) [3] . In the diagnostic setting, diffuse areas in FL may be defined with the help of the follicular dendritic cell markers CD21 and CD23 that are absent in diffuse areas while mostly present in the few remaining neoplastic follicles. A more recent publication described CD23 expression mainly in the diffuse areas of predominantly diffuse FL cases. These predominantly diffuse FL were characterized clinically by their frequent origin in the inguinal region, the formation of large tumours and a low clinical stage ( Fig. 2 B). Genetically, these FL were characterized by a lack of the translocation t(14;18) and a genomic deletion in 1p.36.3 (Katzenberger et al., 2008) [11] . In contrast to the CD23 expression, CD21 staining in this FL subgroup was restricted to the rare neoplastic follicles ( Fig. 2 C).

gr2

Fig. 2 Morphology and immunohistochemical features of predominantly diffuse FL with del(1)(p36.3). Predominantly diffuse FL are defined by less than 25% follicularity and present with a diffuse infiltration pattern, with only few intermingled atypical follicles (A). Diffuse areas of FL are usually defined by the absence of the follicular dendritic cell markers CD21 and CD23. CD23, however, can also be expressed in the tumour cell population in the diffuse areas (B). CD21 staining is restricted to the rare atypical follicles (C).

Cytogenetic alterations in the development of follicular lymphoma

The translocation t(14;18) is currently viewed as the initial genetic hit in FL that juxtaposes the BCL2 gene to the IgH enhancer thus leading to a constitutive expression of the BCL2 protein. Proteins of the BCL2 family govern the permeability of the outer mitochondrial membrane and the release of Cytochrome c to the cytoplasm. They can be either pro-apoptotic or anti-apoptotic [12] . BCL2 itself is an anti-apoptotic molecule that inhibits the release of Cytochrome c and thus the formation of the apoptosome, a protein complex that is composed of cytochrome c, APAF1, ATP and Caspase 9. Moreover, BCL2 was described to be a cell cycle regulator that inhibits G0 to G1 transition via upregulation of p27 which is in line with the lower proliferation index in FL compared to normal GCs [13] . In the physiological setting, GC B-cells of healthy individuals undergo apoptosis unless they are positively selected by a specific antigen to become a plasma cell or a memory B-cell. However, BCL2 expressing FL cells fail to undergo apoptosis during affinity maturation and class switching processes in the germinal centres and, therefore, are more likely to accumulate secondary chromosomal alterations in the presence of activation induced deaminase (AID)-mediated genomic instability [5], [14], and [15].

Secondary chromosomal alterations that are typical for FL include gains in 1q, 2p, 7, 8, 12q, 18q and X as well as deletions in 1p, 6q, 10q, 13q and 17p, combinations of which are present in almost all t(14;18)-positive FL [16] and [17]. Besides the t(14;18), losses and copy number neutral loss of heterozygosity (LOH) in 1p were described to be the most frequent alterations in FL [18] and [19].

In an attempt to define minimally deleted genetic regions in FL to narrow down the search for potential tumour suppressor genes, Cheung and colleagues recently identified a 2Mb overlapping deletion in the chromosomal region 1p36.33-p36.32, including the tumour suppressor gene TP73, using array CGH and 250K SNP arrays from Affymetrix. The role of TP73 alterations, however, in the pathogenesis of FL remains unclear, since in an earlier study sequence analysis of TP73 in 16 FL with 1p deletions did not reveal a mutation in the remaining allele [19] .

Earlier chromosomal alterations in the karyotypic evolution of FL may comprise deletions in 6q, alterations of chromosome 18 (dup der (18)t(14;18)) and gains of chromosome 7 and 8, according to a study of Hoglund and colleagues, whereas losses of the long and/or short arm of chromosome 17, gains of chromosomes 12, 1q, X and deletions in 1p were suggested to be later events [20] .

Methylation patterns in the development of follicular lymphoma

The systematic and global assessment of epigenetic DNA modifications has become a recent research focus in many malignant tumours, since epigenetic changes, such as DNA methylation at CpG-rich sequences, represent reversible DNA modifications that have been associated with transcriptional silencing of pathogenetically relevant genes in neoplastic cells. Accordingly, methylation profiling in FL is currently considered a useful approach for the identification of biomarkers that may contribute to its pathogenesis and that may be promising targets for more specialized treatment strategies using demethylating agents. Specifically, hypermethylated tumour suppressor genes (TSGs) or unmethylated oncogenes may constitute promising targets for new treatment approaches in FL.

Some years ago, Baur and colleagues investigated the methylation status of the cell cycle inhibitors and potential TSGs p14 (CDKN2A), p15 (INK4B or CDKN2B) and p16 (INK4A or CDKN2) in a small cohort of FL samples [21] . Aberrant methylation of the p15 and p16 promoter regions was detected in roughly 50% of studied cases, whereas hypermethylation of the p14 promoter was absent ( Fig. 3 ).

gr3

Fig. 3 Methylation Profiling in FL. FL tumours can be separated according to their methylation profile from non-neoplastic controls by a hierarchical clustering approach. Compared to non-neoplastic B-cells or non germinal centre derived B-cell lymphoma, FL shows a high level of aberrant hypermethylation. Specifically, several tumour suppressor genes such as the cell cycle regulators p16, p15, p57 and p14 and the death-associated protein kinase DAPK were found to be hypermethylated in FL. Abbreviations: M: methylated, U: unmethylated, NTC: non-tumour control, FL: follicular lymphoma

Another cell cycle inhibitor that was found to be frequently methylated in FL is the tumour suppressor gene p57 (KIP2, CDKN1C). In a study of 18 FL samples, 8 cases presented with a hypermethylated p57 promoter region [22] . Moreover, promoter methylation of the detoxifying enzyme GSTP1 and the serine-threonine kinase SNK/PLK2 might also play a role in FL pathogenesis ( Fig. 3 ) [23] . Strikingly, Rossi and colleagues also detected promoter hypermethylation of the death-associated protein kinase (DAPK) in 85% of FL [23] . In line with these findings DAPK1 was also one of the 47 genes that were found to be aberrantly methylated and significantly down-regulated in gene expression in a later study of O’Riain and colleagues and, interestingly, was found to be hypermethylated at three different sites in FL (P10_F, E46_R and P345_R) [24] . DAPK is a calcium-calmodulin-dependent serine/threonine kinase that participates in the extrinsic apoptotic pathway. In FL, DAPK promoter methylation may therefore add to the anti-apoptotic function of BCL2 that deregulates the intrinsic apoptotic pathway.

This study, which is the largest methylation study in FL to date, was performed on more than 164 untreated FL samples, 10 paired transformed and non-transformed FL cases and 27 non-tumour samples. Technically, the DNA of these tumour samples was bisulfite modified and analyzed for 1536 specific CpG sites in 371 genes by multiplexed genotyping in an array format. On the basis of their methylation profile, the FL tumours could be easily separated from the non-neoplastic samples, which is in line with findings from other studies [25] and [26]. While transformed FL also showed a widely divergent methylation profile in comparison to non-tumour samples, it was not possible to distinguish transformed from untransformed FL with regard to their methylation profile. Moreover, no association between the methylation pattern of a given FL and clinical parameters, specifically survival time, could be established. As an explanation for this disappointing finding, the authors hypothesize that the admixture of a high number of non-malignant bystander cells in many FL specimens may mask the tumour cell-specific profile and therefore prevent a conclusive correlation with survival parameters. As a general important finding of this study, 199 CpG loci were identified in the FL specimens that showed a significant increase in methylation compared to the non-neoplastic control group, whereas only six CpG loci showed a significant loss of methylation in FL. The finding that hypermethylation of CpG loci appears to be more frequent than hypomethylation in FL is in line with the observation in other lymphoid malignancies, especially in germinal centre (GC)-derived B-NHL. A recent microarray-based methylation study by Martin-Subero and colleagues compared various types of B-cell malignancies with normal B-cell counterparts including CD19-positive B-cells and GC B-cells [26] . More than half of the genes that were found to be significantly hypermethylated in any one of the subgroups of B-cell neoplasms were hypermethylated in FL, in contrast to other B-NHL, e.g. mantle cell lymphoma and chronic lymphocytic leukaemia, in which only a small number of genes showed hypermethylation. Thus, these studies provide evidence that the FL genome appears to be characterized by a pronounced hypermethylation pattern. A consistent finding in several studies is the hypermethylation of polycomb repressor targets in FL [27] . This is of interest, since the polycomb group protein and methyl transferase EZH2 which is the catalytic component of the PRC2 complex was found to be highly expressed in proliferating centroblasts in the GC, whereas it was not expressed in non-proliferating centrocytes or naïve B-cells [26] and [28]. This led to the hypothesis that germinal centre derived lymphomas such as FL show high EZH2 expression and, therefore, are affected by a high level of aberrant DNA hypermethylation. However, a very recent ultra-deep sequencing approach of FL samples conducted by Morin and colleagues revealed a mutation in the EZH2 gene that results in the replacement of a single tyrosine in the SET domain of EZH2 in a subset of diffuse large B-cell lymphomas (DLBCL) and in approximately 7% of investigated FL samples leading to a reduced enzyme activity [29] . It is expected that the more detailed characterization of the global methylation pattern in FL will identify altered genetic and signalling pathways in the future that could be of pathogenetic relevance.

MicroRNA patterns in the development of follicular lymphoma

The characterization of microRNAs (miRs) in malignant tumours has gained a considerable momentum recently. MiRs are small non-coding single stranded RNA molecules that have been found to play a key role in many biological processes [30] . These miRs are processed from precursor RNAs by different protein complexes called DROSHA and DICER to mature miRs that are only 19–24 nucleotides in length which are subsequently loaded to so-called RISC complexes. Depending on their sequence, the mature miRs bind to target sequences thereby repressing translation or transcription of a variety of target genes. Abnormal miR expression has been linked to many disease types such as infectious diseases, genetic disorders, and cancer [30] . It is therefore not surprising that several groups have begun to study miR expression in FL. In one study, more than 150 different miRs were investigated in paraffin embedded tumour specimens from 46 FL [31] . In comparison to reactive lymph nodes and diffuse large B-cell lymphomas, 12 miRs showed differential expression in FL. Specifically, nine miRs were overexpressed and three were down-regulated in FL. There was little overlap, however, with another study, but differential expression of miR-150 and miR-135a in FL could be confirmed by Lawrie and colleagues [32] .

A classification tree, based on four miRs (miR330, miR17-5p, miR106a and miR210) was established by Roehle and colleagues that was able to separate FL from DLBCL and non-neoplastic samples with an overall accuracy of 98% [31] . This is supported by another study describing statistically significant differences in miR expression between transformed and non-transformed FL suggesting that FL has its own characteristic miR expression profile that differs from that of other B-NHL and reactive lesions [32] .

BCL2 in follicular lymphoma and in healthy individuals

Because of its well-known anti-apoptotic function, the deregulated expression of BCL2 appears to play a key role in the development of FL and may thus be a promising target in the treatment of this lymphoma [33], [34], and [35]. Consequently, a number of therapeutic agents that target BCL2 are already subject of clinical trials including Oblimersen sodium, Gossypol, ABT-737 and GX-15-070 [36] . Oblimersen is an 18-mer oligonucleotide that binds to the first six codons of the BCL2 gene thus inhibiting BCL2 expression. In FL patients, a synergistic effect was observed between Rituximab and Oblimersen and, notably, Oblimersen may help to overcome Rituximab resistance [35] . For more detailed information, the reader is referred to other chapters of this book.

Nevertheless, the presence of the translocation t(14;18) or BCL2 protein overexpression alone is most likely insufficient for complete neoplastic transformation. Several studies have shown that the t(14;18) can also be detected at a low frequency in the peripheral blood of healthy individuals, the majority of which never develops overt follicular lymphoma [37] and [38].

It has been elegantly shown by Roulland and colleagues that t(14;18)-positive B-cell clones in healthy individuals predominantly belong to the IgD-/CD27+ or IgD+/CD27+ memory B-cell subset and not to the subset of naïve B-cells, as initially thought [39] .

In contrast to peripheral IgD+/IgM+/CD27+ memory B-cells which do not undergo immunoglobulin class switch recombination (CSR), neither in their productively rearranged nor in their nonproductive alleles, t(14;18)-positive IgD+/CD27+ B-cells do undergo CSR. Most importantly, CSR in these B-cells affects the translocated nonfunctional allele which is atypical among circulating memory B-cells. Interestingly, this allelic paradox is also generally observed in FL. A recent study by Agopian and colleagues demonstrated that the peripheral blood of individuals that ore often exposed to pesticides carried a significantly higher numbers of activated t(14;18)-positive B-cells than non-exposed individuals [40] . In addition, they found that the BCL2/IgH positive cell clones in healthy individuals were all CD10-positive and showed ongoing AID-mediated hypermutation activity which is in line with a germinal centre or post-germinal centre stage of B-cell differentiation. Of note, BCL2-Ig transgenic mouse models, over 25 weeks old, demonstrated splenic hyperplasia, that, however, did not progress to a monoclonal stage and overt lympha [37] and [39]. This may indicate a long latency period or a slow pathogenetic evolution from an initial polyclonal B-cell expansion to monoclonal disease, as observed in another study with BCL2-Ig transgenic mice. [38] In this study, 75% of the transgenic mice developed follicular hyperplasia, without progression to FL, whereas a small subset developed diffuse large B-cell lymphoma. The fnding that half of these aggressive lymphomas carried a MYC rearrangement underlines that additional genetic alterations – besides the translocation t(14;18) – are necessary for full neoplastic transformation [38] .

Follicular lymphoma without translocation t(14;18)

Approximately 90% of FL carry the translocation t(14;18) resulting in constitutive overexpression of BCL2 which most likely contributes to the accumulation of secondary chromosomal alterations during clonal evolution. Some FL cases, however, lack the t(14;18) and the majority of these also lack BCL2 protein expression [41] and [42]. In general, the t(14;18)-status of FL is related to the tumour grade or to the site of primary origin. While only few nodal FL grades 1 and 2 lack the translocation t(14;18), it is absent in 30%–40% of FL grade 3A and in 70–85% of FL grade 3B [43] and [44]. Moreover, FL arising at extranodal sites, such as the skin or testis, are also frequently t(14;18)-negative [45] and [46].

The pathogenesis of t(14;18)-negative FL remains largely unclear. Hypothetically, an anti-apoptotic protein other than BCL2 might be deregulated in these cases, but no such protein(s) has been convincingly demonstrated as yet. Several studies investigated potential differences between t(14;18)-positive and t(14;18)-negative FL, in general, however, without specific exclusion of FL grade 3B or diffuse large B-cell lymphomas with an additional component of FL grade 3B. Not surprisingly, the majority of t(14;18)-negative cases was identified among “high grade FL” and a lack of the t(14;18) was found to be frequently associated with a 3q27/BCL6 rearrangement and a CD10-negative and IRF4/MUM1-positive phenotype ( Table 1 ) [41], [47], [48], and [49].

Table 1 Molecular features of t(14;18)-negative FL in comparison to FL with the t(14;18)

Molecular features of t(14;18)-negative FL Literature
More frequent 3q27/BCL6 rearrangements Horsman et al., 2003 [41]

Guo et al., 2007 [47]
More frequently CD10-negative Guo et al., 2007 [47]

Jardin et al., 2002 [48]

Leich et al., 2009 [42]
More frequently IRF4/MUM1-positive Tagawa et al., 2007 [49]
Increased BCL-XL expression Zha et al., 2004 [50]
No difference in ongoing/aberrant somatic hypermutation and AID expression to t(14;18)-positive FL Gagyi et al., 2008 [51]
Molecular phenotype of a late germinal centre B-cell stage Leich et al., 2009 [42]

In contrast to BCL2 rearranged FL, little information exists about the process of neoplastic transformation in nodal FL grades 1–3A lacking the t(14;18). Specifically, reliable molecular markers and, therefore, attractive future therapeutic targets are lacking. An interesting approach to identify such markers in t(14;18)-negative FL was carried out in a study that applied reverse-phase protein microarrays [50] . In this study, BCL2 protein-negative FL was compared to BCL2 protein-positive FL resulting in an increased BCL-XL expression in BCL2 protein-negative cases, although this difference was only of borderline significance ( Table 1 ).

It has been suggested that the presence of the t(14;18) keeps neoplastic B-cells in the germinal centre stage of B-cell differentiation. Given that somatic hypermutation (SHM) of the IgVH genes is a hallmark feature of GC B-cells and that aberrant SHM has been described in other B-NHL, such as DLBCL, it was of interest to study the occurrence of SHM, potential aberrant SHM of the Myc, RhoH and PAX-5 genes and the level of AID expression in BCL2-negative FL [51] . No differences in these parameters, however, were observed by Gagyi and coworkers between t(14;18)-negative and t(14;18)-positive FL ( Table 1 ) suggesting that t(14;18)-negative FL resemble their t(14;18)-positive counterparts in these molecular features [51] . Evidence for genetic and molecular differences between t(14;18)-positive and t(14;18)-negative FL was provided in a large study of more than 150 FL samples that used gene expression profiling and accompanying genomic profiling using both conventional comparative genomic hybridization and high-resolution 250K SNP profiling [42] . Gene set enrichment analysis of the expression data led to the suggestion that t(14;18)-positive FL carry a “classical” GC B-cell phenotype, whereas the molecular phenotype of t(14;18)-negative FL may be more related to that of a late GC B-cell or post-GC B-cell ( Table 1 ). This finding could be partially validated on the immunohistochemical level in an independent series of 40 t(14;18)-negative and 40 t(14;18)-positive FL cases, in which t(14;18)-negative FL were negative for the germinal centre-associated marker CD10 in approximately one third of cases.

Prognostic and predictive factors in follicular lymphoma

The strongest clinical predictor of outcome to date in FL patients is the International Prognostic Index for FL (FLIPI) which includes information on age, Ann Arbor stage, haemoglobin level, number of extranodal sites affected and serum lactate dehydogenase level [3] . Histological grading and the proliferation index of the tumour cells may also be useful markers for outcome prediction, but – given the highly variable clinical course and the complex biological nature of FL – these ‘simple’ markers may not be sufficient. It is likely that additional molecular and genetic markers are needed to tailor future therapeutic approaches to the underlying biological features of the FL in a given patient. Accordingly, a large number of studies are available in the literature that has investigated the correlation between genetic or molecular features of FL and corresponding clinical parameters, most importantly survival times. In the following paragraphs, these findings will be summarized and some hallmark studies will be presented in more detail.

Genetic alterations and outcome prediction

As a general theme, the existence of overall more genetic alterations in a given FL tumour genome has been associated with more aggressive clinical behaviour. For example, in a study of more than 100 FL cases, the presence of more than six secondary genetic alterations identified by classical chromosome banding and M-FISH analysis, was associated with an inferior overall survival [52] . This finding is in line with a study of O’Shea and colleagues who investigated 185 FL by genomic profiling in which the occurrence of more than three genetic alterations was associated with an inferior outcome [53] . Specifically, recurrent losses in chromosomes 1p, 6q, 10q and 17p, as well as trisomy 21, gains on 1q and acquired uniparental disomy in 16p were found more frequently in FL patients with an adverse clinical course or clinical progression [16], [17], [20], [52], [54], and [55]. In addition, rearrangements of MYC and BCL6 as well as inactivation of the tumour suppressor TP53 appears to be of pivotal importance in the progression and transformation of FL to more aggressive lymphoma [5], [53], [56], and [57].

Interestingly, in a very recent study by Brodtkorb and colleagues that included 27 FL that underwent subsequent transformation and 17 FL that showed no signs of transformation, the inverse prognostic role of losses in 1p or 6q described in other studies could not be confirmed. Instead, gains involving chromosome 2, 3q or 5 were found to be exclusively present in patients with transformation and were associated with poor survival [58] . To add more confusion, another large study with 210 FL found no association between the number or type of cytogenetic alterations and clinical outcome at all [59] . It is difficult to provide an explanation for the varying results in the literature – even between larger studies, but more recent results suggest that the type of treatment during the course of the disease may significantly impact on the relevance of a particular genetic or molecular feature, as it relates to clinical outcome (see below).

Gene expression profiling for outcome prediction in follicular lymphoma

As mentioned above, a particular feature of FL tumours is that they arise in close proximity to non-neoplastic bystander cells, also known as the microenvironment that includes, among other cells, T-cells, histiocytes and macrophages, and dendritic cells. Using gene expression profiling, it became evident in a landmark study that features of the microenvironment are likely to play an important role for the clinical course and biological aggressiveness of FL [60] . In this study, Dave and colleagues investigated tumour specimens from 191 untreated FL patients with whole genome expression arrays and identified two prominent survival-associated signatures, termed immune response 1 (IR1) and immune response 2 (IR2). Most of the genes that are members of these signatures are expressed by the non-neoplastic bystander cells in the tumour infiltrates. Specifically, the IR1 consists of genes that are expressed by T-cells and macrophages (e.g. ACTN1, CD7, IL7R, ITK) and the expression of this signature was associated with a more favourable prognosis. In contrast, IR2 comprises genes that are predominantly expressed by macrophages and dendritic cells (e.g. CEB1, DUSP3, SEPT10, TLR5) and this signature was found to confer a worse prognosis ( Table 2 ). Strikingly, a statistical model composed of the IR1 and IR2 signatures allowed the stratification of FL patients into quartiles with widely differing survival times (between 3.9 years, and 13.6 years).

Table 2 Prognostic markers in follicular lymphoma.

Inferior prognosis/Histologic progression
 del1p Horseman et al., 2001 [16] ; Cheung et al., 2009 [54] ; O’Shea et al., 2008 [53]
 del6q Cheung et al., 2009 [54] ; Viardot et al., 2003 [17] ; Horseman et al., 2001 [16]
 del10q Viardot et al., 2003 [17] ;

Horseman et al., 2001 [41]
 +1q Höglund et al., 2004 [20]
 del17p Viardot et al., 2003 [17]
 Trisomy 21 Vangstein Aamot et al., 2007 [52]
 Acquired UPD in 16p O’Shea et al., 2009 [55]
 Rearrangements of MYC Viardot et al., 2003 [17]
 Rearrangement of BCL6 Akasaka, 2003 [56]
 +2 Brodtkorb et al., 2010 [58]
 +3q Brodtkorb et al., 2010 [58]
 +5 Brodtkorb et al., 2010 [58]
 Inactivation of TP53 O’Shea et al., 2008 [53]
 Immune response 2 Dave et al., 2004 [60]
 Dense infiltrate of FOXP3+

T-cells, treatment: fludarabine
de Jong et al., 2009 [77]
 CD69 expression de Jong et al., 2009 [77]
 High number of CD4+ T-cells Wahlin et al., 2010 [62]
 High amount of CD68+ macrophages Farinha et al., 2005 [65] ;

Wahlin et al., 2010 [62]
 CD68+ macrophages, treatment: fludarabine de Jong et al., 2009 [77]
 High density of CD31+ microvessels Clear et al., 2010 [69] ;

Taskinen et al., 2010 [70]
 High number of MCL1-positive centroblasts Michels et al., 2006 [71]
 BCL-XL expression Zhao et al., 2004 [72]
 Increased BCL2/BAK and BCL2/BAX ratios Gulmann et al., 2005 [73]
 YY1 expression Sakhinia et al., 2007 [76]
Good prognosis
 Immune response 1 Dave et al., 2004 [60]
 High number of CD4+ T-cells Lee et al., 2006 [27] ;

de Jong et al., 2009 [77]
 High number of CD8+ T-cells Lee et al., 2006 [27] ;

Wahlin et al., 2007 [66]
 PU1 expression Torlakovic et al., 2006 [74]
 High CCNB1 expression Björk et al., 2005 [75]
 Dense infiltrate of FOXP3+

T-cells, treatment: CVP
de Jong et al., 2009 [77]
 CD68+ macrophages, treatment: CVP de Jong et al., 2009 [77]

Subsequent studies attempted to translate the findings from gene expression profiling into more applicable, immunohistochemistry-based tests for routine diagnostic purposes ( Table 2 ). The results among the different studies, however, are highly inconsistent which might be due to small study cohorts or biases introduced by different treatment regimens. For example, while some immunohistochemical studies describe a significant correlation between the number of FOXP3-expressing T-regulatory cells and improved survival in FL [61] and [62], other authors failed to observe an influence of the number of FOXP3-positive T-cells on outcome [63] . Moreover, it was suggested that the localization of FOXP3-expressing T-regulatory cells rather than their absolute number might be correlated with favourable survival times [64] . In a similar manner, confusing results were obtained for the T-cell markers CD4 and CD8. Some IHC studies observed a correlation between the number of CD4 and CD8 expressing T-cells and improved outcome and others did not [63], [64], [65], and [66]. Of note, in one study the presence of increased numbers of follicular CD4-positive T-cells was associated with poor outcome [62] . Equally frustrating results were obtained when the macrophage content in the tumour infiltrates was investigated, usually by IHC for CD68. In two major studies the number of macrophages correlated with inferior survival, whereas this was not evident in two other series [63], [64], [65], and [66]. Along these lines, upregulation of the chemokine receptor CCR1 that plays a role in the recruitment of monocytes and histiocytes was found to be correlated with inferior outcome in one study [67] , but this could not be confirmed at the immunohistochemistry level in a recent series of 187 tumour specimens from untreated FL patients [68] .

Vascular density, a morphological correlate for potential angiogenic events in the tumour microenvironment, has been studied in FL tumour samples as well recently, and an association between poor outcome and increased angiogenic sprouting was proposed. Interestingly, increased vascular density was associated with an increased number of infiltrating CD163+ macrophages in this study [69] . Technically, the authors quantified CD31/PECAM1 positive vessels using a computerized image analysis system in tumour specimens from 59 FL patients that were divided in a short survival group (<5 years) and a long survival group (>15 years) [69] . In support of this finding, Taskinen and colleagues also observed a significant association between high microvessel density counts/high PECAM1 levels and inferior outcome [70] .

Resistance to apoptosis, e.g. conferred by upregulation of BCL2 as a result of the translocation t(14;18), appears to be a crucial early event in the pathogenesis of FL. Accordingly, other anti-apoptotic genes and proteins besides BCL2 were also subject of several studies in FL ( Table 2 ). MCL1 is a prominent member of the group of anti-apoptotic proteins and increased MCL1 expression in centroblasts demonstrated by IHC correlated with poor overall survival in 85 FL patients [71] . BCL-XL expression was also found to be associated with shorter overall survival times in FL [72] , although this finding could not be confirmed by IHC in another study [65] .

Gulmann and colleagues used reverse-phase protein microarrays and IHC to study apoptotic pathways and found that the BCL2/BAK ratio was able to discriminate FL from follicular hyperplasia, as increased BCL2/BAK ratios were present in FL specimens [73] . Moreover, these authors showed that increased BCL2/BAK and BCL2/BAX ratios were associated with earlier death from FL [73] .

Another IHC study by Torlakovic and coworkers suggested that outcome of FL patients was associated with the level of GC differentiation of the tumour cells, since a strong prognostic significance of the germinal centre-associated marker PU.1 could be demonstrated ( Table 2 ) [74] .

Cell cycle dysregulation may also be a pathogenetic feature of FL cells with prognostic implications. For example, the cell cycle regulator CCNB1 was described to be an independent prognostic factor in a microarray study with 51 FL patients [75] . Specifically, a high mRNA expression of CCNB1 correlated with longer survival, but – again – no significant association between CCNB1 expression and longer survival was observed in another study [76] . Instead, high expression levels of the anti-apoptotic gene YY1 were found to be associated with shorter survival ( Table 2 ) [76] .

The previous paragraphs have empasized the highly conflicting results of potential prognostic markers in FL on the gene expression and protein levels across different studies published to date. As touched upon earlier, the varying treatment modalities of the FL patients in the published series may account, at least in part, for these differences. This has been elegantly demonstrated by de Jong and colleagues who provided evidence that the prognostic role of the composition of the microenvironment might depend on the choice of treatment [77] . More specifically, the authors studied various markers of accessory cells (T-cells, macrophages, etc.) by immunohistochemistry in pre-treatment biopsy samples of FL patients that were either treated with fludarabine alone or with cyclophosphamide, vincristine and prednisone (CVP) in a phase III clinical trial of the European Organization for Research and Treatment of Cancer Lymphoma Group (EORTC). Mechanistically, the CVP protocol may more directly target the tumour cell population, whereas fludarabine exerts its effects both to the tumour cells directly, but also to the microenvironment ( Fig. 4 ). Especially, FOXP3-positive T-regulatory cells may be preferably targeted by fludarabine, as demonstrated in B-cell chronic lymphocytic leukaemias. In both treatment arms of the EORTC study, a dense infiltrate of CD4-expressing T-cells was associated with a good prognosis, whereas expression of CD69 on tumour cells correlated with an inferior outcome. Conversely, the number of tumour-infiltrating CD68-positive macrophages and FOXP3-positive T-regulatory cells was found to have a directly opposite impact on survival depending on the respective treatment regimen ( Table 2 , Fig. 4 ). For example, a dense infiltrate of FOXP3-positive T-regulatory cells was associated with a better survival in patients treated with CVP, while the same number of FOXP3-positive T-regulatory cells in the microenvironment was associated with a shorter time to progression in FL patients treated with fludarabine. Likewise, a dense infiltrate of CD68-positive macrophages was associated with a good prognosis in CVP treated patients and with an inferior prognosis in fludarabine treated patients [77] . The important lesson from this study is that any future prognostic marker in FL needs to be evaluated in the context of specific therapeutic approaches. Such studies should therefore make use of collected tumour specimens from FL patients that have been or will be included in major prospective clinical trials.

gr4

Fig. 4 Impact of different treatment approaches in FL (fludarabine vs. cyclophosphamide, vincristine and prednisone, CVP) on the prognostic role of the microenvironment (according to de Jong et al., 2009). CVP may more directly target the neoplastic cells, whereas fludarabine may affect both the tumour cells and the microenvironment, and specifically FOXP3-positive T-regulatory cells. A dense infiltrate of CD4-expressing T-cells was associated with a good prognosis and expression of CD69 on tumour cells was associated with a poor prognosis in both treatment arms. The prognostic impact of the number of tumour-infiltrating CD68-positive macrophages and FOXP3-positive T-regulatory was dependent on the respective treatment regimen (for details see text).

Summary

Follicular lymphoma arises in close contact to T-cells and follicular dendritic cells, and its morphological and immunophenotypic features resemble those of normal germinal centres. Accordingly, FL expresses BCL6 and CD10 and shows ongoing somatic hypermutation of the immunoglobulin genes in most instances. In contrast to normal GC B-cells, most FL constitutively express BCL2 as a consequence of the translocation t(14;18) which is present in approximately 90% of cases. Aberrant BCL2 expression prevents apoptosis and promotes the accumulation of secondary genetic alterations during clonal evolution. The t(14;18) is also detected in memory B-cells in a subset of healthy individuals, with an increased incidence in people exposed to pesticides arguing that the presence of the t(14;18) alone is insufficient for full neoplastic transformation. Hypermethylation of several tumour suppressor genes including DAPK has been described in FL and hypermethylation of DAPK may add to the anti-apoptotic function of BCL2. 10% of nodal FL lack the t(14;18), but show similar morphological and molecular features including ongoing somatic hypermutation of immunoglobulin genes compared to their t(14;18)-positive counterparts. The gene expression profiles of t(14;18)-negative FL suggest a molecular phenotype corresponding to a late germinal centre B-cell stage. In all FL, the microenvironment appears to play a pivotal role in their pathogenesis as well as for the clinical behaviour. Several genetic, molecular and microenvironmental prognostic markers have been identified, but recent data suggest that at least some of these (e.g. the number of accompanying T-cell subsets) may be highly dependent on the choice of therapy.

Conflict of interest statement

No conflict of interest to declare.

References

  • [1] M. Herold, A. Haas, S. Srock, et al. Rituximab added to first-line mitoxantrone, chlorambucil, and prednisolone chemotherapy followed by interferon maintenance prolongs survival in patients with advanced follicular lymphoma: an East German Study Group Hematology and Oncology Study. J Clin Oncol. 2007 May 20;25(15):1986-1992
  • [2] W. Hiddemann, M. Kneba, M. Dreyling, et al. Frontline therapy with rituximab added to the combination of cyclophosphamide, doxorubicin, vincristine, and prednisone (CHOP) significantly improves the outcome for patients with advanced-stage follicular lymphoma compared with therapy with CHOP alone: results of a prospective randomized study of the German Low-Grade Lymphoma Study Group. Blood. 2005 Dec 1;106(12):3725-3732
  • *[3] S. Swerdlow, E. Campo, N.L. Harris (Eds.) et al. WHO classification of tumors of haematopoietic and lymphoid tissues (IARC, Lyon, 2008)
  • [4] M. Eray, V. Postila, J. Eeva, et al. Follicular lymphoma cell lines, an in vitro model for antigenic selection and cytokine-mediated growth regulation of germinal centre B cells. Scandinavian Journal of Immunology. 2003 Jun;57(6):545-555
  • [5] R.J. Bende, L.A. Smit, C.J. van Noesel. Molecular pathways in follicular lymphoma. Leukemia. 2007 Jan;21(1):18-29
  • [6] A. Martinez, S. Pittaluga, M. Rudelius, et al. Expression of the interferon regulatory factor 8/ICSBP-1 in human reactive lymphoid tissues and B-cell lymphomas: a novel germinal center marker. The American Journal of Surgical Pathology. 2008 Jun 18;
  • *[7] M. Schraders, D. de Jong, P. Kluin, et al. Lack of Bcl-2 expression in follicular lymphoma may be caused by mutations in the BCL2 gene or by absence of the t(14;18) translocation. The Journal of Pathology. 2005 Feb;205(3):329-335
  • [8] S.D. Fugmann, A.I. Lee, P.E. Shockett, et al. The RAG proteins and V(D)J recombination: complexes, ends, and transposition. Annual Review of Immunology. 2000;18:495-527
  • [9] A.L. Shaffer, A. Rosenwald, L.M. Staudt. Lymphoid malignancies: the dark side of B-cell differentiation. Nature Reviews. 2002 Dec;2(12):920-932
  • [10] A. Albinger-Hegyi, B. Hochreutener, M.T. Abdou, et al. High frequency of t(14;18)-translocation breakpoints outside of major breakpoint and minor cluster regions in follicular lymphomas: improved polymerase chain reaction protocols for their detection. The American Journal of Pathology. 2002 Mar;160(3):823-832
  • [11] T. Katzenberger, J. Kalla, E. Leich, et al. A distinctive subtype of t(14;18)-negative nodal follicular non-Hodgkin lymphoma characterized by a predominantly diffuse growth pattern and deletions in the chromosomal region 1p36. Blood. 2009 Jan 29;113(5):1053-1061
  • [12] F. Ghiotto, F. Fais, S. Bruno. BH3-only proteins: the death-puppeteer’s wires. Cytometry A. 2010;77(1):11-21
  • [13] S. Zinkel, A. Gross, E. Yang. BCL2 family in DNA damage and cell cycle control. Cell Death and Differentiation. 2006 Aug;13(8):1351-1359
  • [14] G. Lenz, L.M. Staudt. Aggressive lymphomas. The New England Journal of Medicine. Apr 15 2010;362(15):1417-1429
  • [15] L. Pasqualucci, G. Bhagat, M. Jankovic, et al. AID is required for germinal center-derived lymphomagenesis. Nature Genetics. 2008 Jan;40(1):108-112
  • [16] D.E. Horsman, J.M. Connors, T. Pantzar, R.D. Gascoyne. Analysis of secondary chromosomal alterations in 165 cases of follicular lymphoma with t(14;18). Genes, Chromosomes & Cancer. 2001 Apr;30(4):375-382
  • [17] A. Viardot, T.F. Barth, P. Moller, et al. Cytogenetic evolution of follicular lymphoma. Seminars in Cancer Biology. 2003 Jun;13(3):183-190
  • [18] K.J. Cheung, A. Delaney, S. Ben-Neriah, et al. High resolution analysis of follicular lymphoma genomes reveals somatic recurrent sites of copy-neutral loss of heterozygosity and copy number alterations that target single genes. Genes, Chromosomes & Cancer. 2010;49(8):669-681
  • [19] C.W. Ross, P.D. Ouillette, C.M. Saddler, et al. Comprehensive analysis of copy number and allele status identifies multiple chromosome defects underlying follicular lymphoma pathogenesis. Clinical Cancer Research. 2007 Aug 15;13(16):4777-4785
  • *[20] M. Hoglund, L. Sehn, J.M. Connors, et al. Identification of cytogenetic subgroups and karyotypic pathways of clonal evolution in follicular lymphomas. Genes, Chromosomes & Cancer. 2004 Mar;39(3):195-204
  • [21] A.S. Baur, P. Shaw, N. Burri, et al. Frequent methylation silencing of p15(INK4b) (MTS2) and p16(INK4a) (MTS1) in B-cell and T-cell lymphomas. Blood. 1999 Sep 1;94(5):1773-1781
  • [22] Y. Li, H. Nagai, T. Ohno, et al. Aberrant DNA methylation of p57(KIP2) gene in the promoter region in lymphoid malignancies of B-cell phenotype. Blood. 2002 Oct 1;100(7):2572-2577
  • [23] D. Rossi, D. Capello, A. Gloghini, et al. Aberrant promoter methylation of multiple genes throughout the clinico-pathologic spectrum of B-cell neoplasia. Haematologica. 2004 Feb;89(2):154-164
  • *[24] C. O’Riain, D.M. O’Shea, Y. Yang, et al. Array-based DNA methylation profiling in follicular lymphoma. Leukemia. 2009 Oct;23(10):1858-1866
  • [25] J.K. Killian, S. Bilke, S. Davis, et al. Large-scale profiling of archival lymph nodes reveals pervasive remodeling of the follicular lymphoma methylome. Cancer Research. 2009 Feb 1;69(3):758-764
  • [26] J.I. Martin-Subero, O. Ammerpohl, M. Bibikova, et al. A comprehensive microarray-based DNA methylation study of 367 hematological neoplasms. PloS One. 2009;4(9):e6986
  • [27] T.I. Lee, R.G. Jenner, L.A. Boyer, et al. Control of developmental regulators by polycomb in human embryonic stem cells. Cell. 2006 Apr 21;125(2):301-313
  • [28] J.C. van Galen, D.F. Dukers, C. Giroth, et al. Distinct expression patterns of polycomb oncoproteins and their binding partners during the germinal center reaction. European Journal of Immunology. 2004 Jul;34(7):1870-1881
  • [29] R.D. Morin, N.A. Johnson, T.M. Severson, et al. Somatic mutations altering EZH2 (Tyr641) in follicular and diffuse large B-cell lymphomas of germinal-center origin. Nature Genetics. 2010;42(2):181-185
  • [30] C.H. Lawrie. MicroRNA expression in lymphoid malignancies: new hope for diagnosis and therapy?. Journal of Cellular and Molecular Medicine. 2008 Sep–Oct;12(5A):1432-1444
  • *[31] A. Roehle, K.P. Hoefig, D. Repsilber, et al. MicroRNA signatures characterize diffuse large B-cell lymphomas and follicular lymphomas. British Journal of Haematology. 2008 Sep;142(5):732-744
  • [32] C.H. Lawrie, J. Chi, S. Taylor, et al. Expression of microRNAs in diffuse large B cell lymphoma is associated with immunophenotype, survival and transformation from follicular lymphoma. Journal of Cellular and Molecular Medicine. 2009 Jul;13(7):1248-1260
  • [33] D.T. Chao, S.J. Korsmeyer. BCL-2 family: regulators of cell death. Annual Review of Immunology. 1998;16:395-419
  • [34] J.E. Ehlert, M.H. Kubbutat. Apoptosis and its relevance in cancer therapy. Onkologie. 2001 Oct;24(5):433-440
  • [35] B. Pro, B. Leber, M. Smith, et al. Phase II multicenter study of oblimersen sodium, a Bcl-2 antisense oligonucleotide, in combination with rituximab in patients with recurrent B-cell non-Hodgkin lymphoma. British Journal of Haematology. 2008 Nov;143(3):355-360
  • [36] M.H. Kang, C.P. Reynolds. Bcl-2 inhibitors: targeting mitochondrial apoptotic pathways in cancer therapy. Clinical Cancer Research. 2009 Feb 15;15(4):1126-1132
  • [37] T.J. McDonnell, N. Deane, F.M. Platt, et al. bcl-2-immunoglobulin transgenic mice demonstrate extended B cell survival and follicular lymphoproliferation. Cell. 1989 Apr 7;57(1):79-88
  • [38] T.J. McDonnell, S.J. Korsmeyer. Progression from lymphoid hyperplasia to high-grade malignant lymphoma in mice transgenic for the t(14; 18). Nature. 1991 Jan 17;349(6306):254-256
  • *[39] S. Roulland, J.M. Navarro, P. Grenot, et al. Follicular lymphoma-like B cells in healthy individuals: a novel intermediate step in early lymphomagenesis. The Journal of Experimental Medicine. 2006 Oct 30;203(11):2425-2431
  • [40] J. Agopian, J.M. Navarro, A.C. Gac, et al. Agricultural pesticide exposure and the molecular connection to lymphomagenesis. The Journal of Experimental Medicine. 2009 Jul 6;206(7):1473-1483
  • [41] D.E. Horsman, I. Okamoto, O. Ludkovski, et al. Follicular lymphoma lacking the t(14;18)(q32;q21): identification of two disease subtypes. British Journal of Haematology. 2003 Feb;120(3):424-433
  • *[42] E. Leich, I. Salaverria, S. Bea, et al. Follicular lymphomas with and without translocation t(14;18) differ in gene expression profiles and genetic alterations. Blood. 2009 May 26;
  • [43] T. Katzenberger, G. Ott, T. Klein, et al. Cytogenetic alterations affecting BCL6 are predominantly found in follicular lymphomas grade 3B with a diffuse large B-cell component. The American Journal of Pathology. 2004 Aug;165(2):481-490
  • [44] G. Ott, T. Katzenberger, A. Lohr, et al. Cytomorphologic, immunohistochemical, and cytogenetic profiles of follicular lymphoma: 2 types of follicular lymphoma grade 3. Blood. 2002 May 15;99(10):3806-3812
  • [45] L.S. Finn, D.S. Viswanatha, J.B. Belasco, et al. Primary follicular lymphoma of the testis in childhood. Cancer. 1999 Apr 1;85(7):1626-1635
  • [46] J.R. Goodlad, P.J. Batstone, D.A. Hamilton, et al. BCL2 gene abnormalities define distinct clinical subsets of follicular lymphoma. Histopathology. 2006 Sep;49(3):229-241
  • [47] Y. Guo, K. Karube, R. Kawano, et al. Bcl2-negative follicular lymphomas frequently have Bcl6 translocation and/or Bcl6 or p53 expression. Pathology International. 2007 Mar;57(3):148-152
  • [48] F. Jardin, P. Gaulard, G. Buchonnet, et al. Follicular lymphoma without t(14;18) and with BCL-6 rearrangement: a lymphoma subtype with distinct pathological, molecular and clinical characteristics. Leukemia. 2002 Nov;16(11):2309-2317
  • [49] H. Tagawa, K. Karube, Y. Guo, et al. Trisomy 3 is a specific genomic aberration of t(14;18) negative follicular lymphoma. Leukemia. 2007 Dec;21(12):2549-2551
  • *[50] H. Zha, M. Raffeld, L. Charboneau, et al. Similarities of prosurvival signals in Bcl-2-positive and Bcl-2-negative follicular lymphomas identified by reverse phase protein microarray. Laboratory investigation. A Journal of Technical Methods and Pathology. 2004 Feb;84(2):235-244
  • [51] E. Gagyi, Z. Balogh, C. Bodor, et al. Somatic hypermutation of IGVH genes and aberrant somatic hypermutation in follicular lymphoma without BCL-2 gene rearrangement and expression. Haematologica. 2008 Dec;93(12):1822-1828
  • [52] H.V. Aamot, E.E. Torlakovic, M.B. Eide, et al. Non-Hodgkin lymphoma with t(14;18): clonal evolution patterns and cytogenetic-pathologic-clinical correlations. Journal of Cancer Research and Clinical Oncology. 2007 Jul;133(7):455-470
  • [53] D. O’Shea, C. O’Riain, C. Taylor, et al. The presence of TP53 mutation at diagnosis of follicular lymphoma identifies a high-risk group of patients with shortened time to disease progression and a poorer overall survival. Blood. 2008 Jul 15;
  • [54] K.J. Cheung, S.P. Shah, C. Steidl, et al. Genome-wide profiling of follicular lymphoma by array comparative genomic hybridization reveals prognostically significant DNA copy number imbalances. Blood. 2009 Jan 1;113(1):137-148
  • [55] D. O’Shea, C. O’Riain, M. Gupta, et al. Regions of acquired uniparental disomy at diagnosis of follicular lymphoma are associated with both overall survival and risk of transformation. Blood. 2009 Mar 5;113(10):2298-2301
  • [56] T. Akasaka, I.S. Lossos, R. Levy. BCL6 gene translocation in follicular lymphoma: a harbinger of eventual transformation to diffuse aggressive lymphoma. Blood. 2003 Aug 15;102(4):1443-1448
  • [57] J.A. Martinez-Climent, A.A. Alizadeh, R. Segraves, et al. Transformation of follicular lymphoma to diffuse large cell lymphoma is associated with a heterogeneous set of DNA copy number and gene expression alterations. Blood. 2003 Apr 15;101(8):3109-3117
  • [58] M.B. Eide, K. Liestol, O.C. Lingjaerde, et al. Genomic alterations reveal potential for higher grade transformation in follicular lymphoma and confirm parallel evolution of tumor cell clones. Blood. May 26 2010;
  • [59] N.A. Johnson, A. Al-Tourah, C.J. Brown, et al. Prognostic significance of secondary cytogenetic alterations in follicular lymphomas. Genes, Chromosomes & Cancer. 2008 Dec;47(12):1038-1048
  • *[60] S.S. Dave, G. Wright, B. Tan, et al. Prediction of survival in follicular lymphoma based on molecular features of tumor-infiltrating immune cells. The New England Journal of Medicine. 2004 Nov 18;351(21):2159-2169
  • [61] J. Carreras, A. Lopez-Guillermo, B.C. Fox, et al. High numbers of tumor-infiltrating FOXP3-positive regulatory T cells are associated with improved overall survival in follicular lymphoma. Blood. 2006 Nov 1;108(9):2957-2964
  • [62] B.E. Wahlin, M. Aggarwal, S. Montes-Moreno, et al. A unifying microenvironment model in follicular lymphoma: outcome is predicted by programmed death-1–positive, regulatory, cytotoxic, and helper T cells and macrophages. Clinical Cancer Research. Jan 15 2010;16(2):637-650
  • [63] A.M. Glas, M.J. Kersten, L.J. Delahaye, et al. Gene expression profiling in follicular lymphoma to assess clinical aggressiveness and to guide the choice of treatment. Blood. 2005 Jan 1;105(1):301-307
  • [64] A.M. Lee, A.J. Clear, M. Calaminici, et al. Number of CD4+ cells and location of forkhead box protein P3-positive cells in diagnostic follicular lymphoma tissue microarrays correlates with outcome. Journal of Clinical Oncology. 2006 Nov 1;24(31):5052-5059
  • [65] P. Farinha, H. Masoudi, B.F. Skinnider, et al. Analysis of multiple biomarkers shows that lymphoma-associated macrophage (LAM) content is an independent predictor of survival in follicular lymphoma (FL). Blood. 2005 Sep 15;106(6):2169-2174
  • [66] B.E. Wahlin, B. Sander, B. Christensson, E. Kimby. CD8+ T-cell content in diagnostic lymph nodes measured by flow cytometry is a predictor of survival in follicular lymphoma. Clinical Cancer Research. 2007 Jan 15;13(2 Pt 1):388-397
  • [67] R.J. Byers, E. Sakhinia, P. Joseph, et al. Clinical quantitation of immune signature in follicular lymphoma by RT-PCR-based gene expression profiling. Blood. 2008 May 1;111(9):4764-4770
  • [68] M.W. Anderson, S. Zhao, W.Z. Ai, et al. C-C chemokine receptor 1 expression in human hematolymphoid neoplasia. American Journal of Clinical Pathology. 2006;133(3):473-483
  • [69] A.J. Clear, A.M. Lee, M. Calaminici, et al. Increased angiogenic sprouting in poor prognosis FL is associated with elevated numbers of CD163+ macrophages within the immediate sprouting microenvironment. Blood. Jun 17 2010;115(24):5053-5056
  • [70] M. Taskinen, E. Jantunen, V.M. Kosma, et al. Prognostic impact of CD31-positive microvessel density in follicular lymphoma patients treated with immunochemotherapy. European Journal of Cancer. Jul 12 2010;
  • [71] J. Michels, V. Foria, B. Mead, et al. Immunohistochemical analysis of the antiapoptotic Mcl-1 and Bcl-2 proteins in follicular lymphoma. British Journal of Haematology. 2006 Mar;132(6):743-746
  • [72] W.L. Zhao, M.E. Daneshpouy, N. Mounier, et al. Prognostic significance of bcl-xL gene expression and apoptotic cell counts in follicular lymphoma. Blood. 2004 Jan 15;103(2):695-697
  • [73] C. Gulmann, V. Espina, E. Petricoin 3rd, et al. Proteomic analysis of apoptotic pathways reveals prognostic factors in follicular lymphoma. Clinical Cancer Research. 2005 Aug 15;11(16):5847-5855
  • [74] E.E. Torlakovic, N. Bilalovic, R. Golouh, et al. Prognostic significance of PU.1 in follicular lymphoma. The Journal of Pathology. 2006 Jul;209(3):352-359
  • [75] E. Bjorck, S. Ek, O. Landgren, et al. High expression of cyclin B1 predicts a favorable outcome in patients with follicular lymphoma. Blood. 2005 Apr 1;105(7):2908-2915
  • [76] E. Sakhinia, C. Glennie, J.A. Hoyland, et al. Clinical quantitation of diagnostic and predictive gene expression levels in follicular and diffuse large B-cell lymphoma by RT-PCR gene expression profiling. Blood. 2007 May 1;109(9):3922-3928
  • *[77] D. de Jong, A. Koster, A. Hagenbeek, et al. Impact of the tumor microenvironment on prognosis in follicular lymphoma is dependent on specific treatment protocols. Haematologica. 2009 Jan;94(1):70-77

Footnotes

a Institute of Pathology, University of Würzburg, Josef-Schneider-Str. 2, 97080 Würzburg, Germany

b Institute of Clinical Pathology, Robert-Bosch-Krankenhaus, Stuttgart, Germany

c Margarete Fischer-Bosch Institute of Clinical Pharmacology, Stuttgart, Germany

Corresponding author. Tel.: +49 931 201 47776; Fax: +49 931 201 47440.

1 Tel.: +49 931 201 47696; Fax: +49 931 201 47440.

2 Tel.: +49 711 8101 3390; Fax: +49 711 8101 3619.