Lymph nodes

CHAPTER 13 Lymph nodes



Lambert Skoog, Edneia Tani






Introduction


Enlarged lymph nodes were the first organs to be sampled by fine needle aspiration (FNA); today, they are one of the most frequently sampled tissues. In 1904 Greig and Gray reported that trypanosomes could be demonstrated in smears from lymph node aspirates,1 and for some years afterwards, the technique was used to identify various organisms in infected lymph nodes. The earliest report of a wider application of needle aspiration came from the USA in 1921 when Guthrie described using aspirated material to diagnose a variety of diseases causing lymphadenopathy.2 Over the next 30 years the technique was slowly adopted by clinicians and pathologists, resulting in a number of reports on its usefulness. The first study of FNA to show a convincingly high sensitivity was presented by Morrison and co-workers in 1952.3


We now have a large body of evidence supporting the use of FNA as a primary method of diagnosis in reactive, infective and metastatic lymphadenopathy, but the diagnosis of malignant lymphoma by FNA has been much more controversial. However, several recent studies have shown conclusively that a combined cytological and immunological evaluation of aspirated lymphoid cells results in distinctly improved diagnostic accuracy in cases of lymphoma.426 This had inevitably lead to acceptance of FNA cytology as a method which is comparable to histopathology in diagnostic accuracy.



The role of cytology in lymph node diagnosis


Lymph nodes react to a variety of microorganisms and non-specific stimuli by expansion of the follicle centres and/or interfollicular tissue. This results in enlargement of nodes, which may be considerable. The clinical management of patients with enlarged lymph nodes varies with factors such as age, the presence of known infection and the previous medical history. For example, children can present with massive local lymphadenopathy even after mild infections. Accordingly, medical treatment and a period of observation should precede the request for FNA in a child with persistent lymph nodes after a recent history of infection.


In contrast, adult or elderly patients often react to infections with only slight to modest lymph node enlargement: therefore distinct lymphadenopathy in an elderly patient will arouse suspicion of malignancy and justify immediate needle biopsy. For patients between these two extreme clinical settings it is more difficult to decide which patient is more likely to have a reactive or neoplastic lymphadenopathy.


FNA was introduced in most medical centres with a view to reducing the number of excisional biopsies of lymph nodes. Although a routine procedure, surgical excision is considerably more expensive and time consuming, and is afflicted with a distinctly higher morbidity than FNA. Using cytomorphology alone it is often possible to decide if the lymphadenopathy has resulted from reactive lymphadenitis, metastatic malignancy or lymphoma. Patients with reactive lymph node enlargement or metastasis from a known malignancy can thus be spared lymph node excision. In cases with indeterminate cytology or diagnosis of lymphoma, surgical excision has usually been regarded as mandatory.


Today, this still seems to be the prevailing application of FNA cytology in patients with lymphadenopathy. There has, however, been a trend towards accepting cytomorphology alone as sufficient for diagnosis in patients with abdominal or mediastinal lymphomas. Obviously, this is not because lymphomas at these sites are easier to diagnose than those found in superficial sites, but results rather from clinical considerations. The laparotomy and mediastinotomy or mediastinoscopy otherwise required have a distinct morbidity and can also lead to delay in therapy.


Aspirated cells perform excellently in immunocytochemistry, flow cytometry and gene rearrangement analysis, as has been demonstrated by a number of authors.430 This has increased the accuracy of lymphoma diagnosis on FNA material to the same level as histopathology in some series.2325 However, it should be pointed out that the correct subclassification of lymphomas on cytological material requires experience and optimal material, both of which may be difficult to obtain at centres with relatively few lymphoma patients. In principle this variation in reliability of cytological diagnosis means that FNA cytology can be exercised in the management of patients with lymphadenopathy either at a basic or an advanced level.


At a basic level, aspirated cells are evaluated on routine smears alone. This will allow a conclusive diagnosis in the majority of patients with metastatic tumours and in many cases of reactive lymphadenopathy. Most high-grade lymphomas should also be recognisable, while many of the low-grade lymphomas and some cases of reactive lymphadenopathy will not be identified reliably. From this it is clear that conventional FNA cytology should be used to select patients for open biopsy where tissue is needed for histology and immunological evaluation.


At an advanced level, aspirated cells are evaluated on smears and the diagnosis is then substantiated by immunocytochemistry, flow cytometry and/or gene rearrangement analysis (see Algorithm below). This approach allows a conclusive diagnosis in the vast majority of metastatic tumours, reactive processes and lymphomas. Confirmation by histology is then only necessary in a minority of lymphomas, namely follicular lymphoma Grade III in which choice of treatment is currently based on whether growth pattern, is nodular or diffuse. Even at this advanced diagnostic level some cases of lymphadenopathy cannot be diagnosed conclusively. In such cases our experience is that lymph node excision with subsequent histology will rarely be of additional diagnostic value. It is accordingly advisable to perform a repeat FNA biopsy after 2–3 weeks. This time is obviously not fixed but will be determined by various factors including any active infection, the condition of the patient and patient anxiety.


In the opinion of the authors, all laboratories involved in the diagnosis of patients with lymphadenopathy should use FNA cytology in conjunction with immunological characterisation. This diagnostic approach will have a substantial impact on the diagnostic accuracy and consequently the clinical management of such patients.



Technical aspects




Cytospin preparations


After using parts of the aspirates for smear making, the remainder should be suspended in a buffered balanced salt (BBS) solution at pH 7.4 for cytospin preparations. An ordinary aspirate from an enlarged lymph node will yield several millions of cells. The number of suspended cells should therefore be calculated and the concentration adjusted to 1–2 × 106 cells/ml. Cell-rich suspensions can be diluted to optimal concentration by adding BBS solution. Vigorous mixing should be avoided since it can destroy lymphoid cells, particularly the large immature cells seen in high-grade lymphomas. If the cell concentration is low the cells can be concentrated by centrifugation at 700 rpm for 3–5 min. The resulting pellet is then gently resuspended in a reduced volume of BBS solution. To prepare the cytospin slides the cell suspension is spun in a cytocentrifuge at 700 rpm for 3 min. Each cytospin should contain 1–2 × 105 nucleated cells.


One of the cytospins should always be stained with MGG and compared with the smears to monitor recovery of all cell components. If the suspension contains a rich admixture of red blood cells, it is possible to purify the lymphoid cells by density gradient centrifugation. Normally this procedure does not result in any significant cell loss except in some cases of large cell lymphomas which may be fragile and therefore lost in density gradient centrifugation.


Air dried cytospin preparations can be stored at room temperature for up to 1 week without detrimental effect on the immunological staining. Alternatively the cytospin can be stored at −20°C either in a plastic box or wrapped in aluminium foil. Under these conditions lymphoid cells retain their immunological and morphological characteristics for at least 1–2 years. It is important that the slides are kept wrapped until fully thawed when brought out for use otherwise the cells are prone to disintegration. Both immunoalkaline phosphatase and immunoperoxidase methods are suitable for cytospin preparations.



Flow cytometry


Aspirated cells can also be immunologically characterised by flow cytometry (FC).14,15,1820,23,26 As in the case of immunocytochemistry on cytospin preparations one part of the aspirate should be used for smear making. The second part should be suspended in BBS solution at pH 7.4. A cell concentration of approximately 1 million cells per ml buffer will be sufficient for a complete characterisation of reactive lesions as well as most B- and T-cell lymphomas. At the moment four colour FC is standard in immunophenotyping of lymphomas. Evaluation of scattering light allows elimination of dead cells and granulocytes. Several studies have shown a good agreement between FC on FNA material and surgical biopsy specimen.21,25


FC is a rapid and sensitive technique which can detect small abnormal cell populations in a reactive background. Since FC does not allow an evaluation of cytomorphology it is of importance that the results are correlated to cytomorphology on routinely stained smears. Lymphomas with large cells are often fragile and such cells are destroyed during FC analysis. A close cooperation between the FC laboratory and the cytopathologist is therefore strongly recommended.



Molecular biology


Aspirated cells also perform well in PCR rearrangement analysis.27,28,30 After the aspirated cells have been suspended in BBS at pH 7.4 they should be pelleted immediately, snap frozen and stored at −70°C until used for rearrangement analysis. Cytospin preparations can also be used for FISH analysis of specific translocations to aid subtyping of lymphomas.



Normal lymph node histology and cytology


Knowledge of the structural, histological and cytological features of normal lymph nodes is essential in the evaluation of FNA smears from enlarged nodes, whether the pathology is reactive, infective or due to a lymphoproliferative disorder. A brief outline of the structure of a normal lymph node is therefore included, followed by a more detailed description of the normal cell population.




Normal cytology (Fig. 13.1)


As could be predicted from the description of the normal histology, aspirates from normal lymph nodes and from some reactive nodes are dominated by different types of lymphocytes, but plasma cells, macrophages and granulocytes are also found.










Reactive lymphadenopathy


Lymph nodes respond to many different agents by enlarging and becoming more active. Depending on the type of stimulus, a node may react with one of three basic histological and cytological patterns: reactive hyperplasia, suppurative lymphadenitis or granulomatous lymphadenitis. In some cases it is possible to identify the causative agent, either in routine preparations or by special stains such as those for mycobacteria, leishmania, histoplasma and trypanosomes, but the majority of reactive nodes show non-specific changes.



Reactive hyperplasia


Histologically, this response may take the form of enlargement of the lymphoid follicles which develop active germinal centres. These are characterised by numerous centroblasts and centrocytes, a rich admixture of macrophages with a poorly defined pale cytoplasm containing tingible bodies, and a surrounding cuff of small lymphocytes. Alternatively, there may be expansion of the interfollicular tissue by numerous mature lymphocytes, lymphoplasmacytoid cells, plasma cells and varying numbers of immunoblasts. A mixture of both patterns is present in some cases. Nodes draining tumours or other sources of tissue breakdown may be further expanded by the presence of numerous histiocytes in the sinusoids, a picture referred to as sinus histiocytosis.




Cytological findings: reactive hyperplasia (Figs 13.2, 13.3)



Non-specific hyperplasia


Non-specific hyperplasia yields a cytological pattern on FNA which depends on the proportions of follicular and interfollicular tissue in the aspirate, and this in turn usually correlates with the histological findings described above. Thus, smears from a node composed predominantly of large follicles with active germinal centres contain many centroblasts and centrocytes, while the interfollicular tissue is comparatively sparse and represented by mature lymphocytes, plasma cells and immunoblasts (Fig. 13.2).




In extreme cases, the pattern may mimic a mixed lymphoma of centroblastic/centrocytic type. The presence or absence of tingible body macrophages is of little diagnostic value. Immunocytochemical evaluation of the lymphoid population may be the only way to resolve this diagnostic problem.


In contrast, when interfollicular tissue predominates, the smears are rich in lymphocytes, plasma cells, lymphoplasmacytoid cells and some immunoblasts. Such smears are difficult to differentiate from those of a low-grade lymphoma. Analysis of light chain immunoglobulin restriction is usually required to arrive at a conclusive diagnosis.


Some conditions lead to a cytological pattern which clearly deviates from the general types described above. A description of the best recognised of these follows. It is important to remember that definitive diagnosis is dependent on good clinical correlation.





Infectious mononucleosis


Infectious mononucleosis can also cause lymphadenitis, which is mainly confined to the interfollicular tissue (Fig. 13.3). Cytologically, it is characterised by numerous immunoblasts, some of which are atypical with large irregular nuclei.32 In rare cases the pattern may even be suggestive of Hodgkin’s disease.32 Serological tests can be helpful, but phenotyping of the atypical cell population may be the only way to rule out a lymphoma.




Dermatopathic lymphadenopathy


This is a special variant of reactive lymphadenitis which is observed in patients with chronic skin disorders such as psoriasis or dermatitis. The germinal centres are hyperplastic and the interfollicular tissue is expanded by cells of histiocytic appearance.33 Smears from such lymph nodes show numerous small lymphocytes, plasma cells, eosinophils and occasional blast cells. There are numerous histiocyte-like cells, also known as interdigitating reticulum cells, with pale indistinct cytoplasm. Macrophages containing brown melanin pigment from the damaged skin are always present.



Immunocytochemistry (Figs 13.513.7)


In reactive lymphadenitis the small lymphocytes are mostly T cells of which the helper type predominate. The B cells are of various sizes and are polyclonal, expressing both kappa and lambda light chains. Atypical immunoblasts are either of T or B phenotype, some of which may also express CD30 (Ki-1). They do not express CD15 (Leu M1). This phenotype is inconsistent with that of the neoplastic cells in Hodgkin lymphoma, where the mixed cellular infiltrate might be mistaken for a reactive picture.


image

Fig. 13.5 Immunocytochemistry in reactive lymphadenopathy. Cytospin material from the aspirate shown in Figure 13.2. B- and T-cell markers identify a mixed population of lymphoid cells (A). B cells of varying sizes (B). Small mature T cells (alkaline phosphatase).



image

Fig. 13.7 8-colour flow cytometry analysis of fine needle aspirate from a lymph node with reactive follicular hyperplasia. The following antibody panel was used: Lambda FITC/ Kappa PE/CD19 PerCP-Cy5/ CD10 APC/ CD5 PE-Cy7/ CD4 APC-Cy7/ CD8 AM-Cyan/ CD3 Pacific Blue. Data were acquired using FACS-Canto and DIVA software (Becton Dickinson). Data analysis was done using Infinicyt (Cytognos, Salamanca, Spain) software. Upper row: Left: forward scatter/side scatter plot shows that most cells were in the lymphocyte area. Dead cells (yellow) are excluded from analysis. Middle: CD19 vs CD3 plot shows 47% B cells (red) and 43% T cells (blue). Right: Kappa (orange)/ Lambda (green) analysis in CD19/SSC gated B cells shows normal kappa/lambda ratio 1.3. Middle row: Left: CD19 vs CD5 plot within CD19/SSC gate shows a small population of CD5+ B cells that correspond to mantle zone B cells (5% of B cells). Middle: CD19 vs CD10 plot within CD19/SSC gate shows a somewhat larger population of CD10+ B cells that correspond to germinal center cells (20% of B cells). Right: CD3 vs CD5 plot shows that most T cells were positive for both markers (blue). Lower row: Left: Kappa/lambda analysis of CD5+ B cells shows normal kappa/lambda ratio (1.2). Middle: Kappa/lambda analysis of CD10+ B cells shows normal kappa/lambda ratio (1.5). Right: CD4 and CD8 expression in CD3-gated T-cell population. CD4/CD8 ratio was increased to 8.5.


(Courtesy Professor A. Porwit, Hematopathology Division, Dept Pathology and Cytology, Karolinska University Hospital Solna, Stockholm, Sweden)


The flow cytometry pattern of reactive lymphadenitis is seen in Figure. 13.7.






Acute infective lymphadenopathy


A more definitive morphological categorisation of lymph node disease is sometimes possible in certain infections directly involving nodes and in the group of inflammatory or infective disorders associated with granuloma formation. It is of the utmost importance in these conditions, however, that microbiological culture is undertaken for confirmation of the infectious agent.







Tuberculous lymphadenitis


Infection of lymph nodes by Mycobacterium tuberculosis is usually the result of spread from primary lung infection and can present clinically with massive generalised lymphadenopathy, especially of the cervical nodes, even to the extent of simulating lymphoma. The hallmark of tuberculosis histologically is the presence of caseating necrosis associated with epithelioid giant cell granulomata. Early diagnosis is particularly important since the condition is treatable.











Malignant lymphomas



Introduction


Malignant lymphomas are divided into two major categories: Hodgkin lymphoma and non-Hodgkin lymphomas. They can be further divided into several subgroups, which are important to identify because of their different clinical behaviour. Hodgkin lymphoma is most commonly subclassified according to the Rye scheme which was proposed in 1966,53 which is also followed in the recent WHO classification.54


The classification of non-Hodgkin lymphomas has been more controversial. The Kiel classification propounded in 1975 and the 1982 Working Formulation have been the two most commonly used schemes for this group of tumours.55,56 Histological assessment of architectural and cytological features has traditionally formed the basis for all of the classifications. The updated Kiel classification, published in 1988, also incorporated data from immunophenotypic analysis.57 In the REAL (Revised European-American classification of Lymphoid neoplasms) an attempt was made to define clinical relevant subgroups of lymphomas that could be recognised with available morphological, immunological and genetic techniques.58 The ‘WHO classification of tumours of haematopoetic and lymphoid tissues’ is based on the same parameters as the REAL classification and is today generally accepted.54



The role of cytology in lymphoma diagnosis


Much effort has been spent on the diagnosis of malignant lymphomas by FNA, attempts which have until recently been only partially successful. One major reason for this is that most neoplastic lymphoid cells lack the traditional cytological features of malignancy. Such cells are close replicas of their benign counterparts. In many instances the cytological diagnosis therefore rests on evaluation of whether or not the smears show a spectrum of cells in proportions typical of benign conditions.


If the FNA sample is composed of only one cell type a confident diagnosis of non-Hodgkin lymphoma can usually be made. However, some lymphomas are composed of several types of neoplastic cells, while others contain a confusing admixture of benign lymphoid cells with neoplastic elements, which obviously obscures the picture. The complexity of such samples may be an overwhelming task even for the most experienced cytopathologist. In the case of Hodgkin lymphoma the finding of cells with large atypical nuclei and multilobated nuclei has been considered diagnostic.


At present no system of classification has been constructed for FNA cytology material. From published data, it seems clear that a histological diagnosis based on the Kiel classification correlates very well with FNA findings.59 This partly results from the fact that the Kiel classification has only two subgroups in which growth pattern is of importance for diagnosis and choice of therapy. In both the REAL and WHO classification systems there is a much greater emphasis on cytomorphology, immunophenotyping and molecular studies than on architecture growth pattern. Growth pattern, whether nodular or diffuse, contributes to diagnosis in only one subtype. Thus this system will allow a conclusive diagnosis and subtyping of most lymphomas on cytological material if the morphological evaluation is combined with immunophenotypic studies and sometimes cytogenetics.6062


Diagnosis and subclassification of Hodgkin lymphoma have been attempted on FNA material.6365 Again, the classification schemes in current use are based on both architectural and cytological features in excised tissue, making their application to FNA material somewhat difficult.


The reported accuracy of cytological diagnosis and classification of lymphomas on FNA samples varies between 10% –90%.66 Not surprisingly, this degree of variation has impeded the acceptance of FNA cytology as the sole diagnostic modality in patients with suspected lymphoma. However, as previously pointed out, FNA cytology is more readily accepted for evaluation of patients with suspected recurrent lymphoma, or deep-seated primary lymphomas. This attitude is somewhat puzzling since the diagnostic difficulties encountered in these special circumstances are identical irrespective of the fact that the lymphoma is primary or recurrent, superficial or deep-seated.


The use of ancillary techniques including immunocytochemistry, cytogenetics and DNA hybridisation has greatly increased the utility of cytological material for conclusive diagnosis of lymphoma. In fact, cytology specimens seem ideal for immunological evaluation; recent studies show a high diagnostic accuracy if the cytological findings are combined with results from immunophenotyping and clonal restriction analysis.



Hodgkin lymphoma




Hodgkin lymphoma subtypes





These subgroups can in most cases be identified in smears of aspirates by evaluation of the proportion of large atypical cells and reactive cells.636567


The subtyping of Hodgkin lymphoma has clinical relevance with respect to prognosis. The nodular variant of the lymphocyte predominant Hodgkin lymphoma has an excellent prognosis sometimes even when untreated. Of the classical Hodgkin lymphoma subtypes, nodular sclerosis has been reported to have the best and lymphocyte depletion the worst prognosis.


It is important to realise that the histopathological identification of Hodgkin lymphoma can be difficult. Cases of non-Hodgkin lymphoma may be misdiagnosed as Hodgkin lymphoma. This problem seems to occur most often in the diffuse lymphocyte predominant and the lymphocyte depleted subgroups. If cases of non-Hodgkin lymphoma can be completely excluded, the prognostic difference between the subgroups of Hodgkin lymphoma diminishes. In addition, it has been shown that stage of disease, i.e. the extent of spread, rather than subtype, is the most important prognostic factor.68 As a consequence, the current choice of treatment of classical Hodgkin lymphoma is often based on tumour extension, irrespective of histological subtype.


Several other neoplasms both of lymphoid and non-lymphoid origin may present with a morphological picture mimicking that of Hodgkin lymphoma. It is therefore important that the morphological diagnosis is confirmed by immunocytochemistry.


Jun 8, 2017 | Posted by in PATHOLOGY & LABORATORY MEDICINE | Comments Off on Lymph nodes

Full access? Get Clinical Tree

Get Clinical Tree app for offline access