Guidelines on the management of Waldenström macroglobulinaemia*

Waldenström macroglobulinaemia (WM) is a distinct disorder characterised by a monoclonal immunoglobulin (Ig)M paraprotein and morphological evidence of lymphoplasmacytic lymphoma; the cells are IgM+, IgD+, CD19+ and CD20+ but usually CD5−, CD10− and CD23−. Therapy should currently be reserved for patients who are symptomatic or in whom there is haematological suppression or clear evidence of disease progression. The aim of treatment should be to improve the quality and duration of life with minimal side-effects in the most cost-effective manner. It is not yet clear that achievement of a complete remission confers clinical benefit, and it is possible that prolonging therapy to maximal response may increase toxicity without extra benefit. Alkylating agents, purine analogues and rituximab are all appropriate choices for the primary treatment of WM. Plasma exchange is indicated for the acute management of patients with severe problems because of a circulating paraprotein. Patients who are primarily refractory or acquire resistance to alkylating agents may be candidates for combination therapy, purine analogues or antibody therapy. Data suggest a higher response rate using fludarabine rather than cyclophosphamide, doxorubicin and prednisolone (CAP). Waldenström macroglobulinaemia (WM) is relatively rare, accounting for approximately 2% of all haematological malignancies. In the United States, the annual incidence is 6·1 per million in white males and 2·5 per million in white females although the incidence appears to be lower in non-Caucasians (Herrinton Phekoo et al, 2005). The cause of WM is unknown. The potential role of viral agents, such as hepatitis C and human herpesvirus-8, remains controversial. Inherited genetic factors may play a role in a minority of patients, as, to date, 12 families containing 31 cases have been described (McMaster, 2003). The PubMed which contains MEDLINE (1951 to date), Cochrane Library and EMBASE (1974 to date) databases in English language were searched using the key words Waldenström macroglobulinaemia/Waldenström macroglobulinemia/macroglobulinaemia/macroglobulinemia and further refined by the subsearches plasmapheresis/plasma exchange. Database searches were augmented by checking the reference lists of useful articles identified. No language or publication restrictions were applied. Where important preliminary data exists only in abstract form this has been included but preference has been given throughout to peer-reviewed publications. The authors are experts in their field. Stakeholder involvement was secured through representation from the UK Waldenström Macroglobulinaemia Support Group. Individual authors were responsible for specific sections of the manuscript namely diagnosis and prognostic factors, plasmapheresis and treatment options. The final recommendations were drawn up by consensus using the methods outlined in the AGREE instrument (http://www.agreecollaboration.org) and were further reviewed by a sounding board of 80 haematologists representing adult practice in both teaching and district hospitals. The levels of evidence used were those of the US Agency for Health Care Policy and Research (summarised in Appendix 1). The presenting clinical features are highly variable in WM. A significant minority of patients are asymptomatic at presentation and have an immunoglobulin (Ig)M paraprotein as a co-incidental finding during unrelated clinical investigations. Patients with symptomatic disease may present with features attributable to tissue infiltration, such as anaemia, systemic symptoms and organomegaly. A proportion of patients present with clinical features directly attributable to physicochemical properties of their paraprotein. These features include hyperviscosity syndrome (HVS), cryoglobulinaemia and autoimmune phenomena, such as peripheral neuropathy and cold agglutinin disease (Dimopoulos et al, 2000). Primary amyloidosis, however, appears to be a rare complication of IgM gammopathies (Gertz et al, 1993). A haematologist should review all patients with an IgM paraprotein who have features suggesting an underlying lymphoproliferative disorder and also those rare Monoclonal Gammopathy of Undetermined Significance (MGUS) patients who have clinical features attributable to the properties of their paraprotein. The diagnostic criteria for WM have recently been refined and are detailed in Table I (Owen et al, 2003a). WM is, by definition characterised by IgM monoclonal gammopathy and bone marrow infiltration in all patients. IgM paraproteins should be demonstrated by agarose gel electrophoresis and quantified by densitometry (Kyle Owen et al, 2001). Bone marrow examination is essential in all patients with clinical and/or laboratory features suggestive of an underlying lymphoproliferative disorder. It is also advisable in those patients with paraprotein-related phenomena particularly if cytoreductive or monoclonal antibody therapy is being considered. The value of bone marrow examination in asymptomatic patients is however less clear. Differentiating patients with occult WM from those with IgM MGUS is desirable as it provides better diagnostic information to both the patient and clinician. There is, however, no published evidence as yet to suggests that having this information will improve the overall outcome of individual patients. Bone marrow examination is not therefore considered essential in asymptomatic patients if there are no clinical or laboratory features to suggest an underlying lymphoproliferative disorder or any paraprotein-related phenomena. If bone marrow examination is considered necessary it is good practice to obtain a trephine biopsy as the pattern of infiltration is diagnostically important (see below). In addition, it provides a better assessment of the extent of bone marrow infiltration than even the best quality aspirate smears. An adequate trephine biopsy (at least 1 cm in length) also ensures that a diagnosis can be reliably made even when the bone marrow aspirate specimen is of poor quality. Waldenström macroglobulinaemia is characterised by bone marrow infiltration by small lymphocytes showing plasma cell differentiation although the extent of this can vary considerably from case to case. Subclassification of cases into lymphoplasmacytoid, lymphoplasmacytic and polymorphous subtypes on this basis is, however, highly subjective and of dubious prognostic value. The pattern of infiltration is diffuse or interstitial in most cases and a solely paratrabecular pattern should raise the possibility of follicular lymphoma (see below). Immunophenotypic studies are recommended in all cases (Owen et al, 2003a). WM is characterised in most cases by a surface IgM+ sIgD+/− CD5− CD10− CD19+ CD20+ CD22+ CD23− CD25+ CD27+ CD75− CD79+ CD103− CD138− FMC7+ BCL-2+ BCL-6− PAX-5+ immunophenotype (Owen et al, 2001; Owen, 2003; San Miguel et al, 2003). It is not necessary to routinely assess the expression of all these antigens as in practice a sIgM+ CD5− CD10− CD19+ CD20+ CD23− immunophenotype in association with a non-paratrabecular pattern of infiltration is diagnostic of WM (Owen et al, 2001; Owen, 2003). In a proportion of patients clonal B cells are identified by flow cytometry without morphologically detectable disease. Such patients are invariably asymptomatic and are best classified as MGUS (Owen et al, 2003a). The immunophenotypic profile seen in WM is highly suggestive of derivation from postgerminal centre memory B cells. This hypothesis is further supported by IGH sequence analysis, which has demonstrated evidence of somatic hypermutation without intraclonal diversity in most cases (Sahota et al, 2002; Kriangkum et al, 2004; Rollett et al, 2004). Conventional cytogenetic analysis is of little or no value in the routine diagnostic assessment of patients suspected of having WM. Most patients are karyotypically normal, which may reflect the low-proliferative rate of the clonal B cells. A wide range of numerical and structural abnormalities have however, been described but there are as yet no disease-defining or prognostically relevant abnormalities. Deletions of 6q appear to be the commonest structural abnormality, occurring in up to 50% of patients, although their prognostic significance remains unclear (Schop et al, 2002). Using interphase fluorescence in situ hybridisation (FISH), translocations involving the IGH locus at 14q32 are characteristically absent in WM and FISH studies for the t(14;18) and t(11;14) may therefore be useful in diagnostically difficult cases including those rare cases of IgM myeloma (Schop et al, 2002; Avet-Loiseau et al, 2003a,b; Owen, 2003; Ackroyd et al, 2005a,b). The following additional investigations are recommended for most patients with WM at presentation: plasma viscosity; renal and hepatic function; direct antiglobulin test and cold agglutinin titre if positive; cryoglobulins; β-2-microglobulin. Computerised tomography is advised as a baseline for all patients prior to chemotherapy but is not required in asymptomatic patients unless the result would influence the need for chemotherapy. Patients who present with peripheral neuropathy should have nerve conduction studies and anti-myelin-associated glycoprotein (MAG) serology. Additional serological investigations to assess for autoantibody specificity against other neural antigens, such as gangliosides and sulphatides, may be appropriate in those patients with severe neuropathy and negative MAG serology. This is particularly relevant if cytoreductive or monoclonal antibody therapy is contemplated as it is advisable to demonstrate a clear aetiological role for the paraprotein before embarking on such therapies (Owen, 2004). Waldenström macroglobulinaemia, in common with all other indolent lymphoproliferative disorders, has a highly variable clinical outcome. A significant minority of patients remain asymptomatic and never require therapy while others have advanced lymphoma-requiring therapy. Several studies have therefore evaluated the effect of standard clinical prognostic factors on overall survival and a number of stratification schema have been developed (Gobbi et al, 1994; Morel et al, 2000; Dhodapkar et al, 2001; Merlini et al, 2003). These are detailed in Table II. Age, haemoglobin concentration, serum albumin and β-2-microglobulin were consistently identified as very strong predictors of outcome in these studies. An international collaboration is underway to develop a WM-specific prognostic index. Interestingly in most studies the paraprotein concentration had no prognostic value, which further confirms the pathological data of Bartl et al (1983) who failed to demonstrate a relationship between the extent of bone marrow infiltration and paraprotein concentration. The International Waldenström Macroglobulinemia Foundation web site at http://www.iwmf.com provides a directory of information for sufferers of WM, their families and carers. Information is also provided for healthcare professionals, scientists and others interested in learning about WM. The Foundation operates a talk list specifically set up for WM patients that enables them to ask questions and give answers to one another; the process is started by sending an e-mail to iwmf-talksubscribe-request@home.ease.lsoft.com (do not sign or put anything in the subject or message area); once approved, comments and questions can be posted by sending an e-mail to iwmf-talk@home.ease.lsoft.com. To contact the UK WM Support Group for an information pack, advice or support, patients should send an e-mail to info@septemberservices.com. Research information on WM and information on clinical trails is available at http://www.waldenstromsresearch.org. The usual criteria for starting patients with WM on active treatment consist of clinical evidence of adverse effects of the paraprotein, e.g. hyperviscosity with neurological or ocular disturbance, peripheral neuropathy, amyloidosis, symptomatic cryoglobulinaemia; haematological suppression (haemoglobin <10 g/dl, or platelet count <100 × 109/l); progression to high-grade lymphoma or development of constitutional symptoms (Kyle et al, 2003). Published studies on the treatment of WM are predominantly small phase which considerably in their criteria and assessment of direct of the response and of response between studies are to be very may include those described in the following or acute is in WM, usually in patients who have been to alkylating Plasma exchange is used in the management of severe by the paraprotein. The effect is and usually in the form of is used to the paraprotein from and of hyperviscosity include neurological symptoms and Patients with have an plasma and may also et al, et al, et al, There is however, a relationship between paraprotein concentration and plasma or An increase in IgM concentration from to 30 g/l in an increase in plasma of but an increase from to g/l the plasma by et al, a plasma exchange in a in IgM concentration but in up to a in plasma et al, 1983; et al, et al, In patients with WM the plasma may that given the data a exchange is therefore appear to the plasma to levels and of the paraprotein to levels for et al, plasma is advised for the treatment of in WM. In patients who are this may be indicated as management of evidence of neuropathy may be with an IgM paraprotein. are predominantly the lower and usually 2001; 2002). Although there are studies that the role of plasma exchange in the treatment of there is a clear for of evidence of The treatment should be and used in the cell to during the with or without prednisolone is used as the therapy in WM are usually and toxicity is minimal the is if The response rate is approximately on the criteria and the median survival approximately 60 months et al, et al, 2000; et al, 2001). A no in response rate or survival when was or for 1 (Kyle et al, 2000). The duration of treatment is unknown. There are no data on the of in WM. alone or in combination is also in WM but there are no data with et al, 2001). There are phase studies using combination chemotherapy as primary treatment in WM The for therapy, patient and response criteria are variable but there is no evidence that combination chemotherapy is to alkylating agents et al, et al, and et al, by et al, 2005). Indeed there has been no of combination therapy. Alkylating therapy is appropriate for the and treatment of WM of evidence of of studies of treatment with fludarabine and are given in and to fludarabine as therapy range from to (Dimopoulos et al, et al, et al, 2000; Dhodapkar et al, 2003). In a phase of fludarabine an overall response rate of with complete was in patients an by a further for patients et al, survival data in this were from not the of treatment the of therapy be has been with the treatment of patients who a response of to fludarabine may at progression et al, 2000). In studies using by or the response rate has between and (Dimopoulos et al, et al, et al, et al, 2000). There is no to by and no in response with as Most patients had between and of The median duration of response to purine analogues has between and is primarily with of patients and autoimmune are as in A small of patients has been in which response or disease was with fludarabine of treatment with et al, but of the of to patients was (Dimopoulos et al, fludarabine and are therapy for patients who are primarily or who alkylating There are phase studies of purine analogues in patients who have prior the response vary from to and are for patients to their primary therapy et al, et al, et al, et al, et al, 2001; et al, et al, et al, 2000; Dhodapkar et al, 2003). but not all studies have higher response in patients with primary refractory disease than in those with refractory The which evaluated the role of fludarabine in both and patients, no significant in outcome between the et al, 2003). occurring as as 1 from the of therapy have been (Dimopoulos et al, however, occurring months were in of patients between and of fludarabine et al, and also in of patients in a et al, 2005). The of a fludarabine with in patients in or with primary refractory disease following treatment with alkylating agents a higher response rate in the fludarabine and a duration of the response months et al, 2001). There was a in of survival et al, although there was no in overall median analogues and alkylating agents are to be of and have recently been with the combination in and patients (Dimopoulos et al, et al, et al, 2005). data also suggest that the combination has and is of further et al, 2005). of response have also been in patients of and with or without rituximab et al, 2004). analogues are appropriate for the and treatment of WM. There is no consensus on the duration of treatment with or or on which purine is of evidence of is active than as therapy of evidence of with other CD20+ the monoclonal antibody rituximab is active in the treatment of WM vary between and 50% of patients have been to however, the median duration of response is for patients and may be by treatment in patients et al, et al, 2000; et al, et al, et al, 2004). There is evidence that rituximab may be in the management of peripheral neuropathy, a complication which may be difficult to even with plasmapheresis et al, 2000; et al, 2003). Although the of rituximab in the management of WM may to chemotherapy and patients in the is important as there appears to be a significant of in the levels of IgM paraprotein, which may result in hyperviscosity and a need for plasmapheresis with the of antibody (Dimopoulos et al, et al, 2004; et al, 2004). Although there has been a case of a therapy et al, this appears to be a with rituximab and was not when the is in with chemotherapy et al, in the treatment of WM. is of potential in the treatment of patients who have alkylating agents, purine analogues and antibody therapy. agents are currently only recommended in the of clinical of evidence of peripheral cell has been in patients with WM. of alkylating alone or in combination with have been in small of WM patients. The number of patients is small and the up to the role of this treatment in WM et al, et al, et al, et al, 2003; et al, 2003). of have also been in this et al, et al, 2003; et al, with complete response in patients. Most WM patients are to the with the but peripheral cell to be a treatment in refractory patients et al, 2001). therapy supported by cell has a role in the management of patients with WM with primary refractory or disease of evidence of The definition of the best therapy for patients with a wide range of lymphoproliferative treatment is unclear because of the of a further response when patients or at a The of and therapies have an on the overall clinical and analysis of the effect of therapy on survival is The of immunophenotypic criteria can a of patients with WM, which will the of including other distinct malignancies. A has to the effect of therapy with fludarabine or The is to require a of patients and currently of in the and it will assess the effect of treatment in of response to therapy, duration of in haematological toxicity of therapy, quality of life and and data are and the site can be at of with for patients at in the particularly those who are refractory to agents, is an important detailed studies of the of such as both alone and in combination with are in this The for therapy and may be by the age of many patients with however, such as may to be of to patients. the advice and information in these is to be and at the of to the the for the any for the of these has to to and for at from and and has on the of and has to and for and from and for from and has to from and The is supported by an from