Guidelines on the diagnosis, investigation and management of chronic lymphocytic leukaemia

Abstract

The guideline group was selected to be representative of UK based medical experts and patients representatives. Recommendations are based on a review of the literature using Medline/Pubmed searches under the heading, chronic lymphocytic leukaemia, up to August 2011, and data presented at the American Society of Haematology in 2011. The writing group produced the draft guideline, which was subsequently revised by consensus by members of the Haemato-oncology Task Force of the British Committee for Standards in Haematology (BCSH). The guideline was then reviewed by a sounding board of approximately 50 UK haematologists, the British Committee for Standards in Haematology (BCSH) and the British Society for Haematology Committee and comments incorporated where appropriate. The 'GRADE' system was used to quote levels and grades of evidence, details of which are available in the BCSH guideline pack http://www.bcshguidelines.com/BCSH_PROCESS/42_EVIDENCE_LEVELS_AND_GRADES_OF_RECOMMENDATION.html. The objective of this guideline is to provide healthcare professionals with clear guidance on the management of patients with chronic lymphocytic leukaemia. In all cases individual patient circumstances may dictate an alternative approach. The age-adjusted incidence rate of CLL in the UK and USA is 4·2 per 100 000 per year. The incidence increases with age, is higher in men than women and higher in Caucasians than in other racial groups. The median age at presentation is 72 years and 11% of patients are diagnosed under the age of 55 years (Howlader et al, 2012). Epidemiological surveys have shown a sevenfold increase in CLL and a 2·5-fold increase in other lymphoid malignancies, especially lymphoplasmacytoid lymphoma and hairy cell leukaemia, in the relatives of patients with CLL (Goldin et al, 2009). Genome-wide association studies, genotyping single nucleotide polymorphisms in large numbers of patients with CLL and controls provide evidence for a series of loci associated with slightly increased susceptibility to CLL (Di Bernardo et al, 2008; Enjuanes et al, 2008; Crowther-Swanepoel et al, 2010). The diagnosis of CLL is currently based on the combination of lymphocyte morphology, the presence of >5 × 109/l circulating clonal B cells persisting for >3 months and a characteristic immunophenotype (Bene et al, 2011). A recommended scoring system allocates one point each for the expression of weak surface membrane immunoglobulins, CD5, CD23, and absent or low expression of CD79b and FMC7 (Moreau et al, 1997). Using this system, 92% of CLL cases score 4 or 5, 6% score 3 and 2% score 1 or 2. The differential diagnosis of a CD5-positive chronic lymphoproliferative disorder with a low score includes CLL, especially cases with atypical lymphocyte morphology and/or trisomy 12 (Cro et al, 2010), mantle-cell lymphoma and marginal zone lymphomas. Additional investigations, including cytogenetic analysis and histology, may be required to obtain a definitive diagnosis (Dronca et al, 2010). Three disorders, namely CLL, 'clinical' CD5+ve monoclonal B cell lymphocytosis (cMBL), i.e., those cases of MBL with a lymphocytosis detectable on a routine full blood count, and small lymphocytic lymphoma (SLL) share a common immunophenotype, lymphocyte morphology and/or histology and similar biological features (Marti et al, 2005; Hallek et al, 2008; Muller-Hermelink et al, 2008; Rawstron & Hillmen, 2010; Gibson et al, 2011)). Distinguishing features are shown in Table 1. Patients may present with lymphadenopathy, systemic symptoms such as tiredness, night sweats and weight loss or the symptoms of anaemia or infection. However, more than 80% of patients are now diagnosed as an incidental finding on a routine full blood count. Clinical evaluation should elicit a family history of lymphoid malignancy, define the clinical stage (Table 2) and determine whether B symptoms (fever, weight loss, night sweats), profound lethargy and cytopenias are CLL-related, due to marrow infiltration, immune destruction or hypersplenism, or have an alternative cause. The investigation of asymptomatic stage A patients at diagnosis should include: full blood count, reticulocyte count, direct antiglobulin test (DAT), immunophenotype, routine biochemistry and serum immunoglobulins. Additional tests required pre-treatment, include screening for TP53 deletion and for hepatitis B and C infection in patients who are to receive intensive chemotherapy and/or immunotherapy. Human immunodeficiency virus (HIV) testing should be performed according to UK national guidelines [British HIV Association (BHIVA), 2008]. Marrow examination is not essential for the diagnosis of CLL, but is mandatory to define complete response. It is also indicated in determining the cause of cytopenias pre treatment and prolonged cytopenias post-treatment. Lymph node biopsy is indicated when there is diagnostic uncertainty or clinical suspicion of lymphomatous transformation (see below). While computerized tomography (CT) scanning is mandatory in patients entered into clinical trials, the role of imaging in routine practice remains controversial. CT scanning has the potential to identify small volume nodal and/or splenic enlargement in patients who would otherwise be diagnosed as having cMBL or stage A, Rai 0 disease, to identify bulky disease in previously untreated or relapsed patients with no other indication for therapy, to provide a more accurate assessment of treatment response and to detect incidental abnormalities which might influence clinical management. Very few studies have addressed the clinical benefits of this additional information (Hallek et al, 2008; Norin et al, 2010; Eichhorst et al, 2011). There is no evidence to support the routine use of imaging in asymptomatic stage A CLL or cMBL (Muntanola et al, 2007; Scarfo et al, 2012). If there is clinical concern regarding the possibility of significant thoracic, abdominal or pelvic nodal disease, or of disease transformation, then a CT scan is indicated using standard indolent lymphoma protocols. Consensus UK expert opinion supports the routine use of pre- and post-treatment CT scans in patients managed with more intensive therapies. There is no evidence to support on-going routine CT surveillance scanning of asymptomatic patients following treatment for CLL. Lymphomas develop in 5–15% of patients with CLL, either pre- or post-therapy. The varying incidence partly reflects the requirement for histological diagnosis and differing policies on the indications for tissue biopsy in CLL (Tsimberidou & Keating, 2005; Rossi et al, 2008, 2009). Histological appearances resemble diffuse large B cell lymphoma (DLBCL) in approximately 80% of cases and Hodgkin lymphoma (HL) in the remainder. Clinical features suggestive of lymphomatous transformation include bulky (>5 cm) lymphadenopathy, rapid nodal enlargement, the appearance of extra nodal disease, the development of B symptoms and marked elevation of lactate dehydrogenase (LDH). As lymphomatous transformation may be localized, biopsy should be directed to the most suspicious clinical site. Positron emission tomography (PET)/CT scanning may help in the choice of the lesion to biopsy (Bruzzi et al, 2006). The prognosis of patients with CLL is dependent on a variety of patient, disease and treatment-related factors (Table 3). Disease-related factors include biomarkers able to predict prognosis and those able to predict response to specific treatments. The Binet and Rai staging systems predict outcome in patients presenting with widespread and/or bulky lymphadenopathy, hepatosplenomegaly or marrow failure but are insensitive to the clinical heterogeneity within early CLL, i.e. those cases with a low tumour burden who have Binet stage A or Rai stage 0/1 disease (Rai et al, 1975; Binet et al, 1977). Adding simple clinical and laboratory parameters, such as age, gender, lymphocyte count, lymphocyte doubling time (LDT) and serum beta 2 microglobulin (B2M) to clinical stage improves the prediction of overall survival (OS) (Wierda et al, 2007; Shanafelt et al, 2009a) and time to first treatment in early stage CLL (Molica et al, 2010; Bulian et al, 2011). These parameters and an increasing number of biomarkers (reviewed in Dal-Bo et al, 2009; Furman, 2010; Stamatopoulos et al, 2010) enable patients to be classified as being at low, intermediate or high risk of disease progression. However, the difficulty of extrapolating population data to individual patients is highlighted by recent studies identifying a small subset of stage A patients with TP53 abnormalities who, nevertheless have stable disease (Best et al, 2009; Tam et al, 2009). Although there is no current evidence that prognostic data should influence the timing of initial therapy in individual patients, we recognize that some patients will still wish to have the clearest possible idea of the likely natural history of their disease. If biomarkers are measured, then a minimum set of investigations should include IGHV gene analysis, serum B2M (interpreted in relation to renal function) CD38 expression and a screen for genomic abnormalities. The results must be interpreted in the clinical context, especially taking account of the patient's age, significant comorbidities and evidence of disease progression since diagnosis (Shanafelt et al, 2010). TP53 loss occurs in 5–10% of patients at the time of initial therapy and in 30% of patients with fludarabine-refractory disease. A further 5% of patients prior to initial therapy and 12% with refractory disease, have a TP53 mutation without loss of the other allele and would not be detected by fluorescence in situ hybridization (FISH) (Stilgenbauer et al, 2009; Oscier et al, 2010; Zenz et al, 2010a; Gonzalez et al, 2011; Pospisilova et al, 2012). Both retrospective and prospective studies of previously untreated patients and those with relapsed CLL show that patients with TP53 loss and/or mutation, have a significantly lower response rate and short progression-free survival (PFS) and OS, when treated with an alkylating agent, purine analogue, bendamustine, mitoxantrone and rituximab alone or in combination (Table 4). In contrast, TP53 status has much less effect on the response of patients treated with agents such as alemtuzumab, which kill CLL cells through a TP53-independent mechanism. Unmutated IGHV genes, use of the stereotypic IGHV3-21, deletion of 11q and a raised B2M, independent of clinical stage, also correlated with reduced PFS and OS in a clinical trial of alkylating agent and purine analogue treatment (Oscier et al, 2010). Data from the CLL8 trial comparing FC v FCR indicates that the adverse prognostic significance of del11q may be largely overcome by the addition of rituximab to FC (Hallek et al, 2010). It is currently unclear whether the combination of an anti CD20 antibody with other chemotherapy regimens also improves the outcome of patients with del11q. Patients with active disease, as defined in the International Workshop on Chronic Lymphocytic Leukaemia (IWCLL) guidelines (Hallek et al, 2008), usually benefit from anti leukaemic therapy (Table 5). These criteria apply to both previously untreated and relapsed patients. Neither a high lymphocyte count, in the absence of a rapid LDT or clinical features of hyperviscosity, nor hypogammaglobulinaemia in the absence of serious or recurrent infection is an indication for treatment. The timing of treatment, especially in asymptomatic patients, depends in part upon the rate of disease progression. It is important not to delay treatment to the point at which it may be less effective and/or more difficult to administer (e.g. due to cytopenias) especially in patients with 'high risk' CLL. The clinical heterogeneity of CLL and advanced age of many patients dictate that no single treatment approach is applicable to all patients. Factors influencing the choice of treatment include an assessment of fitness to tolerate chemotherapy and/or immunotherapy, TP53 status, previous or current immune cytopenias and evidence of lymphomatous transformation (Goede & Hallek, 2011). In previously treated patients, the number and nature of prior treatments, their efficacy and toxicity and the availability of a transplant donor should all be taken into consideration. Patients requiring treatment should be assessed for their ability to tolerate myelosuppressive and immunosuppressive therapies. Important factors include age, performance status, significant co-morbid conditions, especially a creatinine clearance of <60 ml/min and susceptibility to infection. The use of scoring systems, such as the Cumulative Illness Rating Scale (CIRS) score may be helpful, but none of the currently used scores is CLL-specific and careful assessment of individual patients remains paramount (Miller et al, 1992; Extermann et al, 1998). Although the outcome of patients over the age of 65–70 years entered into clinical trials of FC and FCR was comparable to that of younger patients (Catovsky et al, 2007; Hallek et al, 2010), this data should not be extrapolated to elderly patients with co-morbidities, as these patients have a higher incidence of myelosuppression, often fail to tolerate a full course of treatment and generally have a poorer outcome with therapy (Tam et al, 2008). Almost all clinical trials have shown that the better a patient responds to therapy the longer their remission, especially in the absence of detectable minimal residual disease (MRD) (Moreton et al, 2005; Bosch et al, 2008). Recent data from the German CLL8 trial showed an improvement in the depth of remission between cycles 3 and 6, and that achieving an MRD negative remission is an independent marker for improved OS as well as PFS (Bottcher et al, 2012). These data provide a clear rationale for the use of the most effective available initial treatment and for completing six cycles of treatment providing toxicity is acceptable. IWCLL criteria for complete response (CR), partial response (PR) and progressive disease are shown in Table 6. The IWCLL define relapse as disease progression at least 6 months after achieving a CR or PR. Refractory disease is currently defined as treatment failure or disease progression within 6 months of anti-leukaemic therapy. However, the duration of response that should influence the choice of second line therapy is an area of continuing debate (Zenz et al, 2012). Patients entered into the German CLL8 trial who had a PFS of <12, 12–24 or >24 months had an OS from the time of second line treatment of 13·1, 20·3 and 44·6 months respectively (Stilgenbauer & Zenz, 2010). In a single centre study, 33 of 112 patients who relapsed after initial treatment with FCR were retreated with FCR. Those patients who relapsed after 3 years had an overall response rate (ORR) and CR of 86% and 23%, respectively, compared to 54% and 0% for those relapsing within 3 years (Keating et al, 2009). Previously untreated or relapsed patients with a TP53 abnormality who require therapy and those who relapse within 2 years of, or are refractory to purine analogue-based therapy regardless of biomarker results, are considered to have 'high risk' CLL. These patient groups have a poor outcome when treated conventionally and should be considered for alternative therapies as discussed in section 4.5.4. A diagnosis of stage A CLL is associated with an increased incidence of infections and auto-immune cytopenias. The quality of life of patients and family/partners may also be affected by a variety of factors including use of the term 'leukaemia', uncertainty about the long-term outlook, concerns about transmission of the disease to offspring and practical issues such as difficulties in obtaining insurance. These issues should be explored and addressed at presentation and regularly during the course of the disease (Shanafelt et al, 2007, 2009b; Evans et al, 2012). The issue of patient communication from both the haematologist's and patient's perspective is discussed on the UK CLL Forum website (www.ukcllforum.org). Patients with early CLL should be reviewed at least twice within the first year from diagnosis to assess the rate of disease progression. For those with stable disease, particularly if they have 'good risk' clinical and/or laboratory features, monitoring can be extended to an annual check. This may be performed in primary care (providing there are clear local guidelines for specialist referral), or in hospital clinics (medical, nurse practitioner or via teleclinics) depending on local arrangements. A meta-analysis of 2048 patients in six trials of immediate treatment with chlorambucil plus or minus prednisolone versus deferred treatment showed no significant difference in 10-year-survival (CLL Trialists' Collaborative Group, 1999). The benefits of early versus delayed treatment using FCR, FR or lenalidomide in asymptomatic early stage patients with poor risk prognostic factors are currently being evaluated in randomized trials. Table 7 summarizes the results of recent phase III trials that showed an improved outcome for patients treated with FC compared to F or chlorambucil (Eichhorst et al, 2006; Catovsky et al, 2007; Flinn et al, 2007) and with FCR compared to FC (Hallek et al, 2010). Phase II studies have demonstrated the efficacy of FR (Woyach et al, 2011), BR (Fischer et al, 2009) and FCMR (fludarabine + cyclophosphamide + mitoxantrone + rituximab) (Bosch et al, 2009) (Table 8), and Phase III studies comparing these regimens with FCR are in progress. Chlorambucil remains widely used in the UK for patients considered unfit for intensive therapy. Although there is no international consensus as to the optimal dose or duration of chlorambucil therapy, the highest response rate and longest PFS have been reported in the UK Leukaemia Research Fund trial in which chlorambucil was at a dose of for 7 4 for 6 months to 12 in patients still after 6 months treatment (Catovsky et al, 2011). The results of phase 3 studies comparing single agent chlorambucil with other regimens are in Table and show no benefit for single agent over A recent phase III randomized patients to chlorambucil or et al, 2009) and showed a higher response rate and longer PFS for the The and PFS in the chlorambucil was lower than in the UK trial but between the studies are by the use of chlorambucil dose regimens and differing data from phase II studies are shown in Table A higher was with the combination of chlorambucil and rituximab compared to a from patients single agent chlorambucil in the UK trial et al, 2010). A high of was in patients, of were over the age of treated with and rituximab (Fischer et al, 2009). Phase III trials of chlorambucil or in combination with an antibody are in progress. In of the efficacy of FC and FCR in CLL, small phase II studies have evaluated regimens in patients considered unfit for full dose treatment. et al, 2008; et al, 2010) Although high response with toxicity are randomized studies with prolonged up are to this treatment approach et al, 2010). Patients who relapse and have not a TP53 abnormality can be to to a further course of their initial therapy, the PFS is usually than after initial therapy and often to However, with the previous therapy is not recommended in patients initial treatment was or if a treatment, shown to be to the initial therapy, The results of studies that include patients with relapsed disease are shown in Table for patients relapsing after FC or FCR and for patients relapsing after or refractory to are There are no phase III studies of patients relapsing after FC or FCR. A phase II of FCR in patients with relapsed CLL showed a higher CR rate and longer PFS and OS than in a treated with of and patients prior therapy with regimens that and an alkylating per and were refractory to and chlorambucil The was with CR + and the PFS was The CR + rate for cases was compared to for fludarabine-refractory cases et al, et al, 2011). A phase II of BR has shown response of and for relapsed or fludarabine-refractory patients, respectively, with an survival of patients with a deletion (Fischer et al, 2008; et al, 2011). patients who relapse after chlorambucil will to with The phase II trials of BR discussed in the previous section elderly patients and the toxicity indicated that this may be for relapsed or refractory patients unfit for FCR. The randomized patients relapsing after single agent alkylating or purine analogue therapy to FC v FCR and showed an improved CR and PFS in the FCR et al, 2010). There have been no randomized studies for patients with high risk CLL and/or combination therapy within 2 The results of phase II and III studies using either FCR, or with or without high dose and previously untreated patients with a TP53 abnormality are shown in Table FCR and are associated with similar response and However, combination therapy with and response and PFS to those with FCR or plus or should be as the of This is associated with a significant risk of infection and should be to and with and and monitoring for is As the duration of remission following regimens is with (see is recommended in patients. The results of studies of fludarabine-refractory CLL are shown in Table The outcome of fludarabine-refractory patients treated with chemotherapy is poor with a median OS of approximately alone results in an of about with results in an improved but the PFS and OS are nevertheless that include and have in patients refractory to either agent alone but are not and the risk of is high et al, 2005; et al, 2011). As with the initial treatment of disease, the duration of remission following regimens is short and therapy, such as (see is recommended in patients. For patients for is not an with should be considered in those patients who relapse more than 12 months after initial treatment et al, 2011). for patients who fail or relapse early after therapy are agents include lenalidomide et al, and with or without rituximab et al, 2011). at dose can provide disease and The choice of therapy depends on patient previous treatment and In the (Wierda et al, 2010), an of in patients refractory to both and and in patients with fludarabine-refractory CLL for was considered The median PFS was approximately 6 months for both groups (Wierda et al, 2010). The of was not by bulky lymphadenopathy, prior rituximab or to FCR. The was lower patients with a deletion in the bulky fludarabine-refractory but in refractory disease. should be considered for patients for has been or is and can provide effective in cases with bulky of × 2 can be effective in this and a higher dose in may be required in patients with disease or those to have a TP53 abnormality et al, 2011). the of achieving long-term survival for patients with CLL, including those with TP53 abnormalities. for and Marrow retrospective of performed between and for CLL showed that about of patients long-term remission et al, 2008). In the German CLL the was and was similar for all et al, 2010), that deletion adverse prognostic significance in this A of data that reduced may be to the in disease using a reduced approach is more than for by the in treatment-related Recent data from the that the following from are to those following from and will increase the donor et al, 2010). of prospective trials of in CLL that not being in remission has adverse prognostic significance than the number of of prior therapy et al, 2009). Data from the group also identify the poorer for both overall and in patients with et al, 2008). The results of recent transplant studies are in Table and current Forum and British Society for Marrow for in CLL, which the consensus indications et al, are shown in Table In of the that MBL is detectable in of and in of of patients with CLL et al, et al, et al, the as to the benefit of screening potential especially family on the outcome of CLL patients donor had MBL is and the risk of progressive CLL from the donor should be the prognosis of the potential transplant particularly if no alternative donor or other of treatment is available et al, 2007; et al, 2010; et al, 2010). There is currently no national or international consensus on the to screen potential for It would to with either early CLL or cMBL in the of B are clonal & Hillmen, 2010). The that an remission is associated with prolonged both in previously untreated (Bosch et al, 2008; Tam et al, and relapsed cases (Moreton et al, has to studies of additional treatment in patients with residual disease post-therapy. The use of following initial therapy with regimens has to an improved CR MRD and prolonged but the potential for careful to the timing of therapy and to and treatment et al, 2009; et al, 2010; et al, 2010; et al, 2011). data that therapy with rituximab may PFS et al, et al, 2008; Bosch et al, 2010). The role of antibody therapy and as therapy are currently being evaluated in clinical trials. Recent phase III studies have evaluated the role of cell in patients who a response to initial therapy. showed of PFS or compared to the with no improvement in OS et al, 2011; et al, 2011; et al, 2012). The of patients had not been to rituximab and it is possible that of a similar might have been with therapy. The outcome of CLL patients with lymphomatous transformation is significantly poorer