Follicular Lymphoma: Current Management and Novel Approaches

This book provides a comprehensive, state-of-the-art overview of follicular lymphoma. The first section of the text explores the current understanding of the biology and pathogenesis of follicular lymphoma, through reviewing recent changes in the WHO classification of low-grade lymphomas, current diagnostic techniques, and emerging research on the importance of the immune microenvironment. The second section focuses on current treatment for localized disease, advanced stage disease, and transformed follicular lymphomas, and details currently FDA approved regimens and evolving radiation techniques for early stage and advanced disease. The last section of the text presents emerging approaches in targeted, immunologic, and vaccine therapy.
Written by experts in the field, Follicular Lymphoma: Current Management and Novel Approaches is a valuable resource for practicing clinicians, practitioners, and researchers who manage and have an interest in follicular lymphoma.

• Language: English
• Publisher: Springer
• Release date: Dec 2, 2019
• ISBN: 9783030262112

Book preview

Part IBiology and Pathogenesis of Follicular Lymphoma

© Springer Nature Switzerland AG 2020

N. H. Fowler (ed.)Follicular Lymphomahttps://doi.org/10.1007/978-3-030-26211-2_1

1. Follicular Lymphoma: Epidemiology, Pathogenesis and Initiating Events

Zi Yun Ng¹   , Connull Leslie², ³    and Chan Yoon Cheah¹, ⁴   

(1)

Department of Haematology, Sir Charles Gairdner Hospital, Nedlands, WA, Australia

(2)

Department of Anatomical Pathology, Pathwest QEII Medical Centre, Nedlands, WA, Australia

(3)

School of Biomedical Sciences, University of Western Australia, Crawley, WA, Australia

(4)

Medical School, University of Western Australia, Crawley, WA, Australia

Zi Yun Ng

Email: [e-mail]

Connull Leslie

Email: [e-mail]

Chan Yoon Cheah (Corresponding author)

Email: [e-mail]

Contributed equally

Keywords

Follicular lymphomaEpidemiologyIncidenceEtiologyPathogenesisLymphomagenesisTumour microenvironmentBCL2 In situ FL

All others contributed equally to this work

Epidemiology

Introduction

Follicular lymphoma (FL) is the second most common lymphoma in the United States (US) and Western Europe and the most common indolent lymphoma [1]. FL is a lymphoproliferative disorder of germinal centre B-cells with a median age of diagnosis of 58 years [1]. It is commonly associated with the inappropriate activation of BCL2, a proto-oncogene which is most commonly activated through the t(14; 18)(q32;q21) chromosomal translocation [2].

General Trend of Incidence

In the United States, Teras et al. analysed data from the Surveillance, Epidemiology, and End Results (SEER) registries to provide estimates of the total numbers of US lymphoid neoplasm cases by subtype as well as a detailed evaluation of incidence and survival statistics. The US age-adjusted incidence rate from 2011 to 2012 for FL was 3.4 per 100,000 population. In this study, while most lymphoid malignancies showed excess risk for males, this was not seen for FL which had an incidence rate ratio (IRR) for gender of 1.18 [3]. Similarly, in the United Kingdom from 2004 to 2014, FL had a higher age-standardised sex rate ratio of 0.93 ([95% CI 0.89–0.98], P = 0.006), meaning there were marginally more females than males diagnosed with FL [4].

From 1992 to 2001, the incidence of FL showed a non-significant rise of 1.8% per year among the elderly [5]. However, the incidence for both genders declined from 2001 to 2012. For males with FL, the annual percentage change in incidence dropped from 4.7% in 2001–2004 to −2.2% in 2004–2012. For females, the annual percentage change declined from 3.4% to −0.8% (2001–2004) and then a further −3.6% from 2007 to 2012. It is hypothesised that the decline in incidence rates is due to declining smoking rates over this period. Gender and race did not significantly influence 2-, 5-, and 10-year survival rates [3].

International Variation

Multiple epidemiological studies have shown that FL has higher incidence in Caucasian populations compared to African or Asian [2, 5, 6]. Analysis of 19 case-control studies by the InterLymph Consortium showed that magnitudes of associations with FL according to region (Europe, North America and Australia) were mostly consistent [1]. A study of non-Hodgkin lymphoma (NHL) from 1988 to 1990 showed FL comprised a greater proportion of NHL diagnoses in North America, London and Cape Town (28%–32%) relative to other sites like Hong Kong (8%), Sweden (11%) or France (17%) [7]. Similarly, a study of 4056 cases of NHL at 13 major medical centres in Thailand from 2007 to 2014 found only 5.6% of these cases to be FL [8].

When looking at a migrant population in England from 2001 to 2007, rates of FL were lowest among Chinese and individuals of African descent, intermediate among South Asians and highest among Caucasians. There was little difference between Afro-Caribbeans and Africans, with incidence rates around 60% lower than that of Caucasians. Between the South Asian groups, Pakistanis showed the highest rates, followed by Indians and Bangladeshis (IRRs of 1.11, 0.68 and 0.54, respectively) [9]. FL is less common in India compared to Europe or America; for example, a study from Mumbai showed that FL accounted for only 12.6% of 2773 NHL cases [10].

Interestingly, investigating incidence of FL in Americans of Asian descent, Clarke et al. found the incidence was significantly lower in foreign-born Asian-Americans compared to American-born (IRR 0.57 [95% CI 0.44–0.73]), suggesting a role for environmental factors in the pathogenesis of FL [11]. Supporting this, the risk of FL seems to be lower in first-generation Asian-born Japanese and Chinese migrants compared to their descendants [12].

Genetic Factors

The InterLymph Consortium which comprised of 19 case-control studies (3530 cases and 22,639 controls) in Europe, North America and Australia showed that a family history of non-Hodgkin lymphoma in a first-degree relative confers approximately double the population background risk of FL [1]. The risk was 3.6 times higher in participants with first-degree male relatives with multiple myeloma compared to the general population. Interestingly, this was not evident if there was a first-degree female relative with myeloma. First-degree relatives with leukaemia or Hodgkin lymphoma did not seem to confer an increased risk of FL [1]. Analysis of 4455 individuals in the Swedish Family-Cancer Database found that a parental history of FL was associated with a significantly increased risk of FL (standardised incidence ratio of 6.1), while an affected sibling conferred a 2.3 times risk [13].

There have been an increasing number of genome-wide association studies (GWAS) identifying single-nucleotide polymorphisms (SNPs) associated with risk of developing FL (detailed in Table 1.1).

Table 1.1

Genome-wide association studies with the relevant SNPs identified to be associated with FL

aInversely associated with FL

bSNPs in non-HLA loci

Certain polymorphisms of the DNA repair gene XRCC3 may increase the risk of developing FL, especially in current smokers [20].

Environmental Factors

The aforementioned migrant studies provide some evidence that environmental factors play a role in the pathogenesis of FL [12]. Attempts to study environmental risk factors in epidemiological studies are greatly hampered by unavoidable confounders and bias. As a result, drawing firm conclusions regarding the relative contribution of specific environmental risk factors is challenging, as data from studies are often conflicting.

A number of studies have examined the association between occupation and risk of FL. A reduced risk of FL was found in bakers and millers (OR 0.51 [95% CI 0.28–0.93]) and university or higher education teachers (OR 0.58 [95% CI 0.41–0.83]) [1]. However, a separate meta-analysis showed an increased risk for NHL in teachers at all levels [21]. A small prospective study in Germany which included 92 FL patients showed significant FL risk increases for occupational groups like medical, dental and veterinary workers (OR 3.1 [95% CI 1.4–6.8]); sales workers (OR 2.8 [95% CI 1.3–5.9]); machinery fitters (OR 3.4 [95% CI 1.5–7.8]); and electrical fitters (OR 3.5 [95% CI 1.5–8.4]) [22]. Risk of FL certainly significantly increased with exposure to chemical solvents such as benzene, toluene, xylene and styrene (OR 1.7 [95% CI 1.2–2.5] P = 4 × 10−7) [23]. Spray painters and those working with paint solvents had increased risk of FL (OR 2.66 [95% CI 1.36–5.24]) [1, 24, 25]. Medical doctors who had worked more than 10 years had a significantly elevated risk (OR 2.06 [95% CI 1.08–3.92]) based on 38 cases vs. 13 controls [1]. Employment in other occupations was not associated with risk of FL, including working/living on a farm [1]. The t(14;18) translocation which occurs in up to 70–90% of FL was found to be associated with certain agricultural pesticides in two studies [26, 27]. A different study found that occupational exposure to pesticides would increase BCL2-IGH prevalence together with the frequency of BCL2-IGH-bearing cells especially during periods of high pesticide use [28]. It should be noted that this translocation can be detected in healthy individuals or patients with other cancers. There were also modestly increased risks of FL related to residential proximity to a petroleum refinery (OR 1.3) or a primary metal industry (OR 1.2) [29].

Unlike other lymphomas, studies suggest that autoimmune diseases are not generally associated with an increased risk of FL with the exception of Sjögren’s syndrome (OR 3.37 [95% CI 1.23–9.19], P = 0.024) [1]. Rather, atopic diseases (with the possible exception of eczema) were associated with a lower risk of FL [1, 30]. Females with allergic rhinitis (OR 0.70 [95% CI 0.56–0.88], P = 0.002) and food allergy (OR 0.74 [95% CI 0.63–0.86], P < 0.001) had lower risk of FL, but this was not apparent in males. Risk for combined and individual atopic/allergic disorders showed greater reduction in Australia compared to Europe or North America [1]. A 22% lower risk of FL was noted if there was a history of a blood transfusion – with reductions in risk most notable if the transfusion was received after 55 years of age and within 40 years of FL diagnosis [1]. Smaller studies examining the impact of prior blood transfusion have suggested either no association [31] or increased risk [32, 33]. Interestingly, although acquired immunosuppression from human immunodeficiency virus (HIV) or organ transplants confer increased risk of lymphoid malignancies such as plasmablastic lymphoma, Epstein-Barr virus (EBV)-driven lymphomas and primary central nervous system (CNS) lymphoma, no increase in FL incidence has been described, suggesting a different mechanism of lymphomagenesis [34, 35].

A population-based case-control study of in-person interviews of 1593 NHL individuals from 1988 to 1995 showed that non-steroidal anti-inflammatory drug use, treatment of type 2 diabetes mellitus with oral hypoglycaemics, a history of hepatitis and three or more lifetime bee stings were inversely associated with FL. On the other hand, a history of heart disease and beta-blocker use were positively associated with FL risk. It is suggested that these conditions exert an immunomodulatory effect that influences the development of FL [36]. In the InterLymph study, positive hepatitis C virus serology was not linked with FL risk (OR 1.28 [95% CI 0.64–2.57]) [1]. Polio vaccination was associated with decreased risk, while influenza vaccination was the opposite; however, the knowledge between vaccinations and FL risk is incomplete [37].

Earlier studies indicated an increased risk of FL for current smokers compared to non-smokers [38, 39], particularly in those with more than a 36-pack-year history [40]. This effect was found in females but not males, for reasons that are unclear [1, 41, 42]. A modest risk of FL among women who ever smoked cigarettes was limited to current smokers, along with a significant positive trend for total duration of smoking. Additionally, duration, rather than frequency of cigarette smoking, appeared more important in the trend in pack-years of smoking in women [1]. The association between smoking and FL is biologically plausible given the increased risk of t(14;18) in heavy smokers [43]. However, two prospective studies showed contrary results, suggesting a lower risk of FL with current/former smokers with one showing a hazard ratio of 0.62 [95% CI 0.45–0.85] [44] and another observing a relative risk of 0.67 [95% CI 0.52–0.86] [45].

There is some suggestion that a diet high in vitamin D [2, 46] and linoleic acid (a polyunsaturated fatty acid) was associated with a lower risk of FL [2]. Men with a dietary pattern high in fat and meat (highest quartile vs. lowest) had an increased risk of FL (HR 5.16 [95% CI 1.33–20.0]) [47]. A few studies found that a diet high in vegetables and fruit was associated with a decreased risk of FL [47, 48]. An inverse relationship between FL risk and antioxidants like vitamin C, lutein + zeaxanthin, β-cryptoxanthin, isoflavones and flavonols was observed by Frankenfeld et al. [49] Increasing nitrate intake (both plants and animals) was positively associated with FL risk, although geographic and ethnic variability as confounders cannot be excluded [2]. FL risk was modestly reduced in women (OR 0.79), but not men who ever drank alcohol, especially in current drinkers. There was no clear pattern with number of drinks per week, duration or cumulative alcohol consumption due to lack of data collected [1]. Although several studies support a higher risk in non-drinkers [45], other studies yield conflicting results [44, 50]. For example, wine consumption marginally increased the risk (OR 2.19 [95% CI 0.83–5.80]), especially if alcohol consumption started before 20 years of age (OR 4.04 [95% CI 1.19–13.76]) and if the amount exceeded 19 grams of alcohol per day (OR 4.37 [95% CI 1.04–18.45]) [2]. An increasing trend was observed for FL risk and the quantity of coffee assumption – with a doubled risk for an intake of more than four cups per day (OR 2.0 [95% CI 1.2–3.4]) and tripled for a consumption over at least 30 years (OR 3.1 [95% CI 1.7–5.6]). The effect appeared synergistic in current smokers in this Italian population-based case-control study of 161 FL cases [51]. However, a smaller Scandinavian population-based case-control study of 105 FL cases failed to confirm an association between coffee intake and risk of developing FL [48]. An increasing amount of recreational sun exposure was associated with a lower risk of FL (OR 0.7–0.78), but this was attenuated when compared with total sun exposure (OR 0.82–0.88) [1]. This association is dependent on the Ex11 + 32 T > C polymorphism in the vitamin D receptor gene. People homozygous for the C allele with <7 hours per week of sun exposure were six times more likely to develop FL compared to individuals homozygous for the T allele [52]. Examined in women only, there was no association between hair dye use (type, frequency, duration) and FL risk, except a modest increase in those who used hair dyes before 1980 (OR 1.40 [95% CI 1.10–1.78]) [1].

In a population-based control study, increased body mass index (BMI) was positively associated with risk of FL [53]. The InterLymph meta-analysis showed that being overweight or obese as a young adult conferred a higher risk of FL [54] (with 15% increase for each additional 5 kg/m² over BMI of 25 in young adults) [1]. Being overweight or obese as a young adult was associated with ~1.5 times risk of FL [1]. However, a population-based case-control study of 586 FL cases did not find an association between early adult weight and FL risk [55]. Like obesity, lack of physical activity has been associated with decreased immune function. A population-based case-control study suggested that total physical activity of more than 19.1 hours per week may have a protective effect, with the benefits more pronounced for women [56].

Epidemiology: Summary

FL is equally balanced among both males and females and most common in the United States and Europe. The disease arises from a complex interplay of genetic and environmental factors, though most patients do not have clearly identifiable risk factors at presentation. A family history of NHL confers an increased risk, and GWAS have revealed SNPs in both HLA and non-HLA regions that influence this. Exposure to pesticides and chemical solvents (e.g. spray painters), Sjogren’s syndrome, heavy smoking (especially in women), obesity and sedentary lifestyle have all been linked to increased risk of FL in some studies. A diet high in vitamin D, vegetables and fruit and low in fat and meat may be protective. Larger epidemiological studies are needed to answer these questions in further detail.

Follicular Lymphoma Pathogenesis

Follicular lymphoma (FL) cells have dependence on a microenvironment mimicking the normal lymph node germinal centre, as might be expected from a mature B-cell lymphoma showing germinal centre features both morphologically (neoplastic cells appear centrocyte- and centroblast-like) and immunophenotypically. Reflected in the 2016 revision of the WHO classification of lymphoid neoplasms [57], forms of follicular lymphoma not associated with the characteristic BCL2-IGH rearrangement, such as paediatric-type FL [58, 59] and predominantly diffuse FL with 1p36 deletion [60], are increasingly recognised as biologically and clinically distinct neoplasms. These lesions aside from the characteristic t(14;18) BCL2-IGH translocation, present in around 85% of FL cases, are recognised as the likely initial necessary although not sufficient abnormality present early in a multihit pathway which culminates in clinically overt follicular lymphoma.

Cell of Origin: First Hit

The t(14;18)(q32;q21) BCL2-IGH translocation is thought to develop early in B-cell development, during V(D)J recombination of the immunoglobulin heavy-chain locus in B-cell precursors developing within the bone marrow [61]. Low levels of translocation-carrying cells can be detected in the circulating blood of healthy individuals, with an increasing prevalence of up to 66% of individuals aged 50 years or older [62], with the vast majority not developing clinical disease. There is persistence of these BCL2-IGH-carrying clones over multiple years in a given person [63].

The t(14;18) translocation juxtapositions the BCL2 gene with the immunoglobulin heavy-chain gene resulting in overproduction of the BCL2 protein which blocks a final common pathway for programmed cell death, preventing apoptosis [64]. In cases of follicular lymphoma which do not show BCL2 protein expression by immunohistochemistry, a subset may show false-negative staining due to mutations in the BCL2 gene [65].

There is a proven clonal relationship between detected t(14:18)-carrying cells and subsequently developed follicular lymphoma and a link between prevalence of such cells and higher risk of subsequent follicular lymphoma [66]. Evidence that circulating translocation-carrying cells are not naive B-cells but germinal centre-experienced and expanded clones suggests clinical disease requires further mutational events as the germinal centre entry of t(14;18)-carrying B-cells appears an insufficient event to trigger pre-FL to FL progression [63]. As further events in a multihit pathway will not happen at once, the presence of early FL precursors blurs the distinction between healthy individuals and subclinical patients.

Recognising that within the germinal centre environment t(14:18)-mediated anti-apoptotic BCL2 protein expression provokes persistence of such cells by rescue from induced apoptosis, these will be cells with only low-affinity B-cell receptors allowing a larger spectrum of antigen cross-reactivity. Such cells will likely undergo repetitive rounds of expansion within germinal centres during the numerous antigenic challenges faced by the immune system. These cells are at subsequent increased risk of acquiring oncogenic mutations due to repeated exposure to the activation-induced cytidine deaminase (AID) mutator inducing somatic hypermutation in germinal centre B-cells. Although most of the randomly occurring chromosomal alterations will be a selective disadvantage over subsequent iterative cycles (noting that FL prevalence increases with age), there would be further accumulation of chromosomal lesions, some of which provide selective advantage and malignant progression [67]. There is evidence from mouse models that AID is required for germinal centre-derived lymphomagenesis [68], supporting the concept that AID-mediated modification contributes to pathogenesis of follicular lymphoma (Fig. 1.1.)

../images/448586_1_En_1_Chapter/448586_1_En_1_Fig1_HTML.png

Fig. 1.1

A protracted model of multihit FL genesis. FLLC follicular lymphoma-like B-cell clones, FLIS in situ follicular lymphoma, PI follicular lymphoma with partial involvement [67]. (Reprinted from Roulland et al. [67], © 2011, with permission from Elsevier)

Microenvironment

Follicular lymphoma is characterised by numerous closely associated non-malignant immune cells, appreciated in diagnostic samples as nodular morphology representing the expanded follicular dendritic cell (DC) network and the usually numerous small host T-cells (Fig. 1.2). These non-neoplastic immune cells appear to influence disease behaviour, with gene expression profiling studies showing differences between FL which transformed to diffuse large B-cell lymphoma (DLBCL) and FL that did not (within the 7-year follow-up period) being genes involved in T-cell function; and rapidly transforming FL appeared more similar to reactive follicular hyperplasia, while non-transforming FL resembled non-activated lymphoid tissue [69].

../images/448586_1_En_1_Chapter/448586_1_En_1_Fig2_HTML.png

Fig. 1.2

Neoplastic nodules in follicular lymphoma (2A – H+E) include the clonal B-cells (2B – CD20) with admixed host T-cells (2C – CD3) and an expanded distorted follicular dendritic cell network (2D – CD21)

The role of the microenvironment in FL appears to simultaneously support growth and survival of the neoplastic cells and suppress the antitumour immune response (Fig. 1.3). Follicular dendritic cells contribute to B-cell receptor (BCR) signalling and higher levels of sustained signalling eventually supporting survival of FL neoplastic cells [70]. Expression of IL-12 by neoplastic B-cells has also been shown to induce functional intratumoural T-cell exhaustion by promoting TIM-3 expression, similar to changes seen in chronic viral infection [71]. Elevated numbers of infiltrating macrophages associated with increased neovascularisation through angiogenic sprouting have been associated with poor prognosis [72].

Stromal cells within the germinal centre provide signals for malignant cells in two general ways: recruitment to the germinal centre and mediation of growth and survival [73]. While there is evidence that stromal cells in germinal centres (fibroblastic reticular cells) interact with FL B-cells using cross-talk mechanisms similar to those used by reactive B-cells [74], the specific migratory drivers of neoplastic cells compared to the normal counterpart, which lead to re-entry of FL clones to germinal centres with maturation arrest and subsequent amplified gene instability, are an issue requiring further investigation.

../images/448586_1_En_1_Chapter/448586_1_En_1_Fig3_HTML.png

Fig. 1.3

Schematic illustration of interactions between B-cells and their microenvironment in the context of the normal germinal centre (GC) reaction and the follicular lymphoma niche. (a) To make high-affinity, class-switched antibody, B-cells must receive cognate help from T follicular helper (TFH) cells during the GC reaction leading to maturation of activated B-cells along with production of memory B-cells and plasma cells. In the absence of T-cell help during B-cell priming by dendritic cells (DCs) followed by follicular dendritic cells (FDCs), B-cells are driving to apoptosis. A range of cytokines including CD40L, IL-21 and IL-4 produced by TFH cells can direct antibody class switching. Moreover, TFH cells produce high levels of chemokine CXCL13 along with FDCs allowing the B-cell migration within an appropriate GC area where rescued B-cells undergo final maturation. In the opposite, B-cells produce inducible T-cell co-stimulator ICOS-L which engages ICOS-driving production of cytokines in TFH cells. The next critical cell in the development of the GC reaction is the FDC that produces under specific and coordinated signalling from immune accessory cells and B-cells themselves a wide range of factors which support recruitment and survival of B-cells. FDCs also concentrate antigen as immune complexes on their surface bridging B-cell receptor (BCR) on B-cells leading to a specific B-cell signalling involved in the cell activation and maturation. Several other hematopoietic cells are present during the GC reaction holding specific functions such as antigen presentation for DCs and macrophages; innate immune response for macrophages, natural killer (NK) cells and γδ-T cells; and adaptive immune response for CD8+ and T regulatory (Treg) cells. (b) Early in FL emergence, specific changes take place in the microenvironment induced either directly by the BCL2-translocated B-cells (represented by nuclear green-red bar code) or indirectly by emerging cell subsets including Treg cells which attenuate CD8+ T-cell function. TFH cells are highly represented in the FL tumour, and they up-regulate IL-4 production sustaining B-cell survival. FDCs modify released factors in response to cross-talk modifications along with FL B-cells but also through other cell subsets such as macrophages which show significant perturbation. BCL2-translocated FL B-cells present specific modifications including the BCR membrane complex and its secondary signalling. (c) Progressed FL disease shows large modification of the tumour landscape. B-cells present genetic instability (represented by several nuclear bar codes) driving several cell function modifications including a constitutive BCR signal (red star). During progression, cells seen in the normal GC reaction are vanishing (TFH cells, FDCs, CD8+ T-cells, and others), while follicular reticular cell-like cells (pink stromal cells) along with tumour-associated macrophages (TAMs) appear in response to stress signals building a microenvironment specific of tumour aggressiveness including angiogenesis promotion [75]. (Reprinted from de Jong and Fest [75], © 2011, with permission from Elsevier)

Numerous studies have examined the relationship between non-neoplastic immune cells and outcome in follicular lymphoma with contradictory results [75, 76]. Given the evidence that classes of non-malignant cells may have therapeutic implications, there has been interest in objective measurement of these populations by computer-assisted scoring [77] although neither this approach nor other quantifications of the tumour microenvironment are currently in general diagnostic use.

In FL cells with impaired checkpoint selection, the BCR has only loose affinity to any specific antigen. In place of antigen affinity as a driver of cell survival, in some FL cases, a sequence motif introduced during somatic hypermutation characteristic of an N-glycosylation site is present, a finding not seen in normal B-cells or other lymphomas characterised by mutated B-cell subsets [78]. The glycan added to these sites shows unusual termination with high mannose, with evidence that macrophages in FL tissue have up-regulated mannose-binding lectins, which in co-location with surface immunoglobulin has an anti-apoptotic effect [79].

Both the constitutional up-regulation of BCL2 and the acquisition of highly mannosylated BCR appear to be critical steps in lymphomagenesis and substitute for antigen affinity in maintaining FL cells in the germinal centre environment [80]. Disrupting such interactions within the microenvironment may be therapeutic opportunities.

Early Lesions

The term in situ follicular neoplasia (ISFN) should be applied to lymph nodes in which abnormal bright BCL2 expression (associated with the characteristic BCL2-IGH translocation) is seen in follicle centre B-cells where there is preservation of normal lymph node architecture and associated non-neoplastic reactive germinal centres [81] (Fig. 1.4). In the updated (2016) WHO classification, these changes have been renamed ISFN (previously follicular lymphoma in situ) to recognise the low rate of progression to clinically overt disease [57].

../images/448586_1_En_1_Chapter/448586_1_En_1_Fig4_HTML.png

Fig. 1.4

In situ follicular neoplasia (ISFN) is usually an incidental finding noted in lymph nodes with a reactive architecture (4A – H+E ×2) in which follicles are mildly expanded (4B – H+E ×20) and there is an associated follicular dendritic cell network (4C – CD21). The neoplastic cells show strong expression of CD10 (4D) in keeping with germinal centre type, with aberrant expression of anti-apoptotic protein BCL2 (4E) and kappa light-chain restriction (4F)

Such alterations may be seen in patients with synchronous or subsequent clinically evident follicular lymphoma and if seen require clinical assessment, although these are also seen in patients who do not subsequently developed clinically evident follicular lymphoma. In the former cases, the ISFN likely reflects spreading and homing of neoplastic cells to reactive germinal centres of adjacent or distant lymph nodes; however, in the latter, it appears to represent a pre-malignant finding. In several case series, ISFN has been seen in association with a second B-cell lymphoma of other types [82, 83] suggesting increased risk for B-cell neoplasms, although such patients would also have had reason for lymph node excision and hence increased likelihood of incidental detection of ISFN.

Partial involvement of a node by follicular lymphoma identifies patients at greater risk of subsequent clinical follicular lymphoma than ISFN and is identified by altered architecture, expanded follicle size, blurred edge to germinal centre, variable and weaker expression of BCL2 and CD10 and neoplastic cells outside the expanded germinal centre [82] (Fig. 1.5). In contrast to this architecturally abnormal node, ISFN likely represents tissue counterpart of FL-like cells in the peripheral blood of healthy people which have seeded reactive hyperplastic germinal centres and expanded in an antigen-dependent manner. The low risk of progression suggests these cells lack additional mutations required for malignant transformation.

../images/448586_1_En_1_Chapter/448586_1_En_1_Fig5_HTML.png

Fig. 1.5

The node is only partially replaced by follicular lymphoma (bottom right) where there is expansion of neoplastic follicles and spillover of neoplastic cells into interfollicular zones, with adjacent quiescent or benign reactive follicles in the remainder of the node (top left)

Disease Evolution and Clonal Variation

As a germinal centre lymphoma with high expression of AID, follicular lymphoma cells will show ongoing somatic hypermutation of immunoglobulin loci following entry into a germinal centre, and this may occur well before clinically malignant transformation. This enables detailed tracking of neoplastic subclonal evolution, as in a pool of tumour cells each clone can be detected by a unique somatic hypermutation fingerprint. In the small number of examined cases, modelling indicates that FL clones may expand within lymph nodes, then migrate to the bone marrow and stay quiescent for long periods before again expanding lymph nodes with less mutated founder FL cells [84]. Under such a model, these clonally related bone marrow-resident in situ follicular neoplasia cells cause the