Bone marrow is the commonest tissue available for histological examination and immunohistochemistry in CLL. Although bone marrow aspirate and biopsy are generally not required to make the diagnosis of CLL, if a marrow is done the aspirate smear must show more than 30 % of all nucleated cells to be lymphoid for the diagnosis of CLL. The pattern of marrow infiltration was previously thought to have prognostic significance, but with the advent of newer prognostic markers it has become less relevant [40]. Even though marrow is not essential for the diagnosis of CLL it is usually recommended before initiating treatment in CLL. The trephine biopsy is a better guide to the overall level of infiltration compared to a bone marrow aspiration. The trephine biopsy core should preferably be a minimum of 1.6 cm, ideally 2 cm, in length taken from the posterior iliac crest [41, 42]. It is valuable particularly in the context of cytopenias both at diagnosis and following treatment as it allows the distinction between marrow failure due to disease, autoimmunity, hypersplenism or treatment-related effects. In around 25 % of patients autoimmune haemolytic anaemia can occur, where the bone marrow will show erythroid hyperplasia [43, 44]. Marrow fibrosis can occur but is rare [45]. Pure red cell aplasia can occur as an autoimmune complication in about 1 % of patients and rarely autoimmune granulocytopenia can also occur [46].

The morphology of lymphocytes in the aspirate is similar to the pattern observed in peripheral blood (Fig. 7.3). Cytology is, however, better appreciated on peripheral blood smears. Medullary spaces can be completely replaced or can be infiltrated in a patchy manner. Bone marrow infiltration in CLL can be interstitial, nodular, diffuse or mixed pattern. In the diffuse growth pattern the marrow is hypercellular and normal haematopoietic cells and fat spaces are replaced by CLL cells (Figs. 7.4 and 7.5). In nodular pattern there is discrete, non-paratrabecular aggregates of small lymphocytes scattered throughout the marrow space. Benign lymphoid aggregates that mimic the nodular pattern can be also observed in systemic autoimmune diseases, chronic myeloproliferative disorders, toxic myelopathy and viral infections [47]. In contrast to benign lymphoid aggregates in the bone marrow, the nodular aggregates of CLL are less compact and have irregular borders with neoplastic lymphoid cells infiltrating into the surrounding space. Bone marrow can also have proliferation centres or pseudofollicles (Fig. 7.6). In the interstitial pattern the lymphocytes are admixed with the normal hematopoietic elements without much effacement of the architecture. In nodular and interstitial pattern of infiltration normal haematopoiesis is retained. A combination of these patterns is probably the commonest especially a combination of interstitial and nodular pattern [40] (Fig. 7.7). The diffuse infiltration was thought to be associated with a poorer prognosis with a median survival time of 28 months compared to 90 months in nodular and 46 months in interstitial pattern [40]. It has, however, become clear that a diffuse pattern of marrow infiltration is not an independent indicator of poor outcome as it appears to be associated with other poor prognostic features such as unmutated IGHV gene and ZAP-70 expression [48].

Patients can present with symptomatic lymphadenopathy or involved lymph nodes may be found during surgery for other pathologies such as mastectomy for breast cancer, bowel resection or hernia repair. Involved lymph nodes in CLL shows effacement of architecture and can be involved with four growth patterns—pseudofollicular, diffuse, tumour forming and paraimmunoblastic. Three types of cells are seen in these infiltrations: small lymphocytes, prolymphocytes and paraimmunoblasts. The morphology of small lymphocytes and prolymphocytes is similar to that in the peripheral blood. Occasionally the nucleus in small lymphocytes can be cleaved mimicking that of a centrocyte which could be ­misinterpreted as mantle cell lymphoma. Paraimmunoblasts are large cells, with a wide rim of cytoplasm. The nucleus is round or oval with fine chromatin, prominent central nucleolus and the mitotic figures ranging from low to moderate. In diffuse pattern of growth small lymphocytes predominate over the other population of cells.

Pseudofollicular pattern is the most common pattern which accounts for 85 % [49, 50]. Pseudofollicles or proliferation centres are characteristic features of CLL. At low power these appears as ill-defined paler staining areas (Fig. 7.8) which at high power are seen to comprise a mixture of prolymphocytes and paraimmunoblasts (Fig. 7.9). The CLL cells in the pseudofollicles have a unique phenotype. Significantly they express the proliferation marker Ki-67 along with other markers such as MUM-1, Oct-1, Bob-1 and CD71 [51] (Fig. 7.10). They are also CD5 positive and stain more intensely for CD20 and CD23 than lymphocytes in the surrounding areas [52]. Expression of cyclin D1 has been noted in tumour cells in proliferation centre by using rabbit monoclonal antibodies [53]. Even though the name pseudofollicle is derived from the fact that they mimic germinal centres of a reactive lymph node, their histological appearance and immunochemistry differ considerably. Pseudofollicles are CD10-, Bcl-6-, Bcl-2+ unlike germinal centres, but both are positive for survivin, an anti-apoptotic protein [54]. Survivin is usually expressed by CLL cells in the proliferation centre but not in the peripheral blood. Their expression in the peripheral blood CLL cells predicts for an aggressive behaviour of the disease [55]. The spleen is usually involved as a part of generalized disorder, but very rare cases of isolated splenic involvement are also reported. Clinically patients can have splenomegaly varying from mild to massive. In one report pathological spleens removed weighed up to 4,130 g [56]. Even with massive splenomegaly the occurrence of splenic infarction is very rare in CLL. On gross examination of the cut surface of the spleen there will be uniformly distributed miliary white nodules ranging from 0.2 to 1.5 cm in diameter. In infiltrated spleen, normal white pulp will be completely replaced and the remnants of B-cell follicles are identified only by special staining for follicular dendritic cells. Unlike in reactive changes in CLL there will also be varying involvement of the red pulp area [57]. The peri-arteriolar tumour will extend along the capillaries to the red pulp. In massive involvement the sinuses will be disrupted. Rarely red pulp involvement can be more prominent than the white pulp involvement which is more common in hairy cell leukaemia. Lymphocytic infiltration within dense fibrous trabeculae and subendothelial regions of trabecular veins is found in SLL/CLL and other malignant infiltrations but not in reactive changes of spleen. Proliferation centres are seen in spleen but are less prominent than in lymph nodes. Iron pigment may be prominent because of haemolysis.

The liver can be involved in CLL but is less commonly observed than the organs described previously. CLL preferentially infiltrate the portal tracts, but sinusoidal and nodular distributions are also described. CLL infiltration could be confused with lymphocytic infiltration in chronic hepatitis or early stages of primary biliary cirrhosis. But CLL cells are cytologically uniform and immunohistochemistry usually distinguishes them.

Currently immunophenotyping is the most useful diagnostic technique available to evaluate various aspects of CLL. It is helpful in various ways including (1) the differential diagnosis of B- and T-cell LPD and reactive conditions; (2) the identification of therapeutic targets and (3) the development of better prognostic markers as well as (4) to facilitate the assessment of minimal residual disease (MRD) after treatment. Phenotyping of CLL and related disorders can be done either by flow cytometry or immunohistochemistry depending on the tissue sample. Immunophenotyping studies, principally by flow cytometry, were first reported in CLL in the early 1980s [65–67]. These studies identified the co-expression of CD5 (Leu1) and pan-B-cell markers as the phenotypic hallmark of CLL. Subsequent studies have further clarified the immunophenotypic features of CLL. Flow cytometry can be used for diagnosis from most tissues. It should be performed in all cases of unexplained persistent lymphocytosis. From solid tissues such as lymph nodes, a cellular suspension can be used for flow cytometry by disaggregating a sample from fresh specimen. A quick diagnosis can be established by this method. There is a considerable overlap of the diagnostic and prognostic antigens used in both methods.

There is convincing evidence that eradication of MRD is associated with an improved outcome in treatment of CLL. MRD can be assessed by multi-colour flow cytometry or real-time quantitative allele-specific oligonucleotide Immunoglobulin heavy chain gene polymerase chain reaction (RQ-ASO IgH-PCR). The initial methods of evaluation of MRD by flow cytometry were not as sensitive as RQ-ASO IgH-PCR which can detect CLL cells down to 0.001 %. But RQ-ASO IgH-PCR as it is expensive, labour-intensive, and cannot be performed unless pre-treatment material is available. In 2007, the European Research Initiative on CLL (ERIC) proposed an international standardized approach after analysing various combinations of antibodies and comparing them against the ASO-PCR technique [105]. After analysing 728 paired blood and bone marrow samples, they derived several conclusions: (1) Blood analysis was equally or more sensitive than marrow in 92 % of samples, but marrow analysis was necessary to detect MRD within 3 months of antibody therapy; (2) The κ/λ/CD19/CD5 combination can be used to screen samples and avoid extended analysis in cases with clear evidence of residual disease where all B cells are CD5+ with light-chain restriction; (3) A CD45/CD14/CD19/CD3 combination or an equivalent can be used to provide a control for CLL cell enumeration and to define the limit of detection; (4) The combination of CD5/CD19 with CD20/CD38, CD81/CD22 and CD79b/CD43 is the best panel to detect MRD with low inter-laboratory variation, low false detection rates, and an accuracy of 95.7 % (Fig. 7.16).

Differential diagnosis for CLL varies depending on the clinical scenarios. For presentation as lymphocytosis and cytopenias the main diseases to be differentiated are mantle cell lymphoma, marginal zone lymphoma and prolymphocytic leukaemia (PLL) (Fig. 7.17). Very rarely splenic diffuse red pulp small B-cell lymphoma, leukaemic presentation of follicular lymphoma and hairy cell leukaemia can give diagnostic confusions. Immunophenotyping using flow cytometry would be the main method for differentiating these disorders. When CLL presents as predominant lymphadenopathy, follicular lymphoma and diffuse large cell lymphoma need to be clearly differentiated. In most cases morphological features and immunohistochemistry are ­adequate to differentiate these disorders. Lymphocytic infiltration at extranodal sites can be a feature of any of the previously mentioned conditions and again morphology and immunohistochemistry serve as a differentiating tool in most cases.