CRITERIA FOR EVALUATING DISEASE RESPONSE AND PROGRESSION IN PATIENTS WITH MULTIPLE MYELOMA TREATED BY HIGH‐DOSE THERAPY AND HAEMOPOIETIC STEM CELL TRANSPLANTATION

Multiple myeloma (MM) is a malignant plasma cell disorder accounting for about 10% of haematological malignancies. The disease is characterized by the clonal proliferation of plasma cells which produce a monoclonal immunoglobulin heavy and/or light chain (paraprotein, M-protein or M-component). This patient-specific paraprotein is present in the serum and/or urine of all patients except in the 1–2% of patients with non-secretory myeloma. Typical clinical and laboratory features in patients with MM include bone pain (due to lytic lesions or osteoporosis), anaemia, renal insufficiency, hypercalcaemia, increased susceptibility to infection and constitutional symptoms resulting in poor performance status. Less common complications include cord compression due to extramedullary plasmacytomas or vertebral collapse, peripheral neuropathy, amyloidosis and hyperviscosity syndrome ( Malpas, 1998). Prior to the introduction of alkylating agents, the median survival of patients with MM was less than a year ( Korst et al, 1964 ; Holland et al, 1966 ). Approximately 60% of patients respond to initial treatment with conventional chemotherapy, but although survival is prolonged by treatment the median survival remains approximately 3 years ( Bergsagel, 1998). Complete remissions are rare and all patients ultimately relapse, resulting in c 25% survival at 5 years and <10% survival at 10 years. Criteria by which different treatment regimens can be evaluated include the proportion of patients achieving an objective response, the duration of response, and survival. Over the past 10–15 years high-dose therapy followed by haemopoietic stem-cell rescue, either allogeneic or autologous, has been increasingly employed in the treatment of multiple myeloma. For a number of reasons the existing criteria for the assessment of disease response have not proved entirely satisfactory for the analysis of disease outcome after high-dose therapy. In particular, there has been no generally agreed definition of complete response. Agreed definitions of response and progression are essential to ensure consistency of reporting within the transplant registries and to enable comparison of results from different studies and/or different treatment centres. New criteria for response and progression have therefore been developed as a result of discussions between representatives of the Myeloma Subcommittee of the Chronic Leukaemia Working Party of the European Group for Blood and Marrow Transplantation (EBMT) and representatives of the Myeloma Working Committee of the Autologous Blood and Marrow Transplant Registry (ABMTR) and the International Bone Marrow Transplant Registry (IBMTR). These criteria will now form the working definitions of response and progression for the purposes of data collection and registry-based studies. Currently none of the registries include specific diagnostic criteria, although all record the relevant investigations performed at diagnosis. However, we wish to emphasize that all patients undergoing high-dose therapy should have proven myeloma which requires treatment. At present high-dose therapy is not recommended for patients with equivocal myeloma or those with stage I disease. We have not at this stage reviewed the criteria for the diagnosis of myeloma, but there may be a requirement for this in the future. For example, because of the increasing use of high-dose therapy for the treatment of primary amyloidosis, it will be important to establish clear guidelines for the differential diagnosis between this condition and multiple myeloma with amyloid. Changes in the level of the serum paraprotein and/or urinary light chain excretion form the basis of assessing the response to therapy and monitoring the progress of the disease. In a minority of patients disease progression will be manifested by increasing marrow or skeletal involvement, or development of other complications, without a rise in paraprotein. In non-secretory myeloma it is difficult to monitor disease accurately. Serial bone marrow examinations are helpful, although the patchy nature of marrow involvement in myeloma makes it difficult to accurately interpret small changes in the percentage of plasma cells present. The currently used response criteria are shown in Tables I. Response criteria were first developed by the Committee of the Chronic Leukemia and Myeloma Task Force (CLMTF) of the U.S. National Cancer Institute in 1968 and were reviewed by the same group in 1973 ( Chronic Leukemia and Myeloma Task Force, 1968, 1973). The main response parameter is a reduction in the paraprotein of at least 50% ( Table I). In 1972 the Southwest Cancer Chemotherapy Study Group, now the Southwest Oncology Group (SWOG), defined ‘objective response’ as a reduction of at least 75% in the calculated serum paraprotein synthetic rate (rather than paraprotein concentration) and/or a decrease of at least 90% in urinary light-chain excretion, sustained for at least 2 months ( Alexanian et al, 1972 ). Patients with a reduction in serum paraprotein synthetic rate of between 50% and 74% were considered to have improved ( Table II). A review of the literature indicates that the CLMTF or SWOG criteria have been used in most subsequent clinical trials, albeit frequently with some modifications of the original proposals. The relative merits of these two sets of criteria in defining outcomes has never been formally assessed, i.e. there is no evidence to indicate whether a 75% reduction in paraprotein synthetic rate has a better prognostic significance than a 50% reduction in serum paraprotein level. Most groups have used paraprotein concentration to define response because of simplicity. The terms partial response or partial remission are also frequently used. Some groups have added additional response categories, such as good or very good partial response and minimal response, again based on the degree of paraprotein reduction. An exception is the United Kingdom Medical Research Council (MRC) Myelomatosis trials, which have evaluated the efficacy of treatment not by the degree of paraprotein reduction but by the proportion of patients achieving plateau ( Table III) ( MacLennan et al, 1992 ). Plateau phase consists of a period of stability after chemotherapy in which tumour progression does not occur despite the persistence of measurable disease. The definition of plateau does not require any specific degree of paraprotein reduction. The minimum period of stable observation required to define plateau was 6 months in the early MRC trials but more recently has been reduced to 3 months ( MacLennan et al, 1992 ). Although the concept of plateau phase was introduced almost 20 years ago ( Durie et al, 1980 ), it has not been extensively used for the evaluation of response in multiple myeloma. Neither the CLMTF nor the SWOG response criteria include a definition of complete response/complete remission (CR), since CR was rarely observed with existing treatments. With the introduction of new regimens such as VAD (vincristine, adriamycin and dexamethasone) and high-dose melphalan (140 mg/m2) without stem cell support, measurable paraprotein disappeared in a significant proportion of patients and criteria for complete remission were formulated ( Selby et al, 1987 ; Gore et al, 1989 ; Samson et al, 1989 ). As the use of high-dose therapy has increased there has been a consequent increase in the number of patients entering CR, and other groups have published their own definitions of CR; as shown in Table IV ( Gahrton et al, 1991 ; Anderson et al, 1993 ; Dimopoulos et al, 1993 ; Bjorkstrand et al 1995b ; Attal et al, 1996 ; Vesole et al, 1996 ; Barlogie et al, 1997 ; Ballester et al, 1997 ; Joshua et al, 1997 ; Schiller et al, 1998 ). All groups agreed that there should be no detectable paraprotein in serum or urine together with a normal number of plasma cells in the marrow (i.e. <4–5%), but differed according to whether the absence of paraprotein is based on routine electrophoresis (EP) alone or whether a more sensitive method such as immunoelectrophoresis (IEP) or immunofixation (IF) was required. In the earliest reports either no method was specified or only EP was required. More recently there has been a trend towards a more stringent definition of CR requiring a negative IF. Some groups have also required the plasma cells in the marrow to be of normal morphology whereas others have not included morphological assessment, and some groups have included factors such as transfusion independence and lack of symptoms. It is perhaps surprising that many groups do not exclude transient responses by specifying a minimum duration of time for which the paraprotein must remain undetectable to fulfil the definition of CR. CR has hitherto been defined by the EBMT Myeloma Transplant Registry as absence of detectable paraprotein in serum and urine and < 5% plasma cells in marrow, without specifying the method to be used for excluding the presence of paraprotein, nor the time period for which results must remain negative. The IBMTR and ABMTR have not hitherto used a standard definition of CR. The current North American National Cancer Institute Intergroup (SWOG, INT, CALGB and ECOG) Myeloma Trial, comparing conventional versus high-dose therapy, defines CR as absence of paraprotein in serum and urine by EP and IF on at least two measurements for a minimum of 6 weeks, and <4% plasma cells in the bone marrow. There are also currently no generally accepted criteria for the definition of disease progression or relapse and papers reporting the results of different treatment regimens do not always specify the criteria used to define progression ( MacLennan et al, 1992 ; Ballester et al, 1997 ; Barlogie et al, 1997 ). Bergsagel et al (1979 ) defined progression as a progressive increase in serum paraprotein of at least 10 g/l or a 100% increase in urinary light chain excretion. Belch et al (1988 ) also used a minimum increase of 10 g/l in serum paraprotein but required an increase of 2.0 g/24 h in urinary light chain excretion. In recent reports most groups have defined progression as an increase in serum paraprotein or urinary light chain excretion by 25% ( Oivanen et al, 1997 ) or 50% ( Samson et al, 1989 ; Bjorkstrand et al 1995a ; Attal et al, 1996 ; Joshua et al, 1997 ). Other indicators of progressive disease such as increasing marrow infiltration or an increase in the number of lytic bone lesions are also included in the definition of disease progression by most groups. For patients in CR a reappearance of paraprotein, by whatever method, is generally accepted to constitute relapse. The EBMT has hitherto defined progression as a 50% increase of measurable paraprotein levels ( Bjorkstrand et al, 1995b ; Gahrton et al, 1995 ). The IBMTR and ABMTR have not previously utilized any defined criteria, but a number of groups recently reporting results of high-dose therapy studies have used a 25% increase for defining progression ( Attal et al, 1996 ; Schiller et al, 1998 ; Barlogie et al, 1997 ) and the current North American Intergroup trial adopts the same definition. In the pioneer study dealing with response to treatment in multiple myeloma, the median survival of patients who responded to melphalan was 41 months compared with 9 months in patients who did not respond ( Bergsagel, 1975) and Alexanian et al (1972 ) reported that the survival of patients treated with combination chemotherapy was directly correlated with the extent of reduction of paraprotein synthesis. This has been a frequently quoted reference supporting the relationship between the degree of response and subsequent survival. However, a similar survival analysis carried out by Palmer et al (1989 ) failed to show such a correlation. Several other studies have also reported a lack of correlation between response and survival ( Baldini et al, 1991 ; Marmont et al, 1991 ; Joshua et al, 1991 ; Blade et al, 1994 ). Even with regimens such as high-dose melphalan 140 mg/m2 and VAD, which produced CR in up to 25% of newly diagnosed patients, duration and survival were not prolonged in patients reaching CR as compared with those achieving PR ( Selby et al, 1987 ; Samson et al, 1989 ). With conventional chemotherapy, stabilization of tumour load is a more powerful prognostic factor than the degree of tumour reduction in predicting survival ( Durie et al, 1980 ; Joshua et al, 1991 ; MacLennan et al, 1992 , 1994; Blade et al, 1994 ; Oivanen, 1996). Since the survival of patients who achieve a partial or minimal response is similar to that of those fulfilling more stringent response criteria, all patients attaining a stable state should be considered in plateau phase regardless of the level of paraprotein. The MRC has been unique among those carrying out clinical trials in multiple myeloma in using stable plateau phase to define treatment efficacy rather than response criteria based on a given degree of paraprotein reduction. In some patients the paraprotein does not fall with treatment but does not increase and may remain stable for months or years. These patients have non-responding but non-progressive disease and may be considered to be in plateau phase at diagnosis. Although these patients are classified as non-responders according to the CLMTF and SWOG criteria, the disease does not progress and such patients in fact usually have a long survival ( Blade et al, 1986 ; Joshua et al, 1991 ). This situation is similar to that observed in patients with smouldering myeloma ( Kyle & Greipp, 1980). In summary, few patients treated with conventional chemotherapy enter CR and the correlation between the degree of tumour response and survival is In up to 50% of patients enter CR after high-dose therapy CR defined on the basis of negative after high-dose therapy a correlation between the degree of tumour response and survival has been myeloma patients who enter CR have a and survival than those who enter or remain in PR or who to respond ( Gahrton et al, 1991 , Bjorkstrand et al, 1995a ; Attal et al, 1996 ; Barlogie et al, 1997 ). This be entirely by the increasing use of more stringent criteria for CR in more recent studies of high-dose therapy since in some of these reports CR was based on negative EP without negative IF ( Gahrton et al, 1991 , Bjorkstrand et al, 1995a ). It more that there is a in the of CR after conventional chemotherapy and after high-dose in other the level of minimal disease is in patients in CR than in those who are in CR after therapy. The new criteria are shown in Table are based on existing criteria, with As will form the basis for data reporting from a number of the a was the investigations required are therefore those which are to be the minimum to response and to progression or relapse. paraprotein levels and urinary light chain excretion form the basis for the assessment of response, progression and relapse. levels must remain stable for a minimum of 6 to fulfil the criteria for a given of response. The response criteria for serum paraprotein and urinary light chain must be in patients in are present. Bone marrow examinations are essential only to complete response or to response in non-secretory myeloma. It is that there are patients who increasing bone marrow despite a paraprotein level or non-secretory but this is not common to marrow examinations in all patients and will on In patients to have non-secretory myeloma, marrow is essential to response. In these patients it was also to require a to ensure that the response is not transient and because of the patchy nature of myeloma of the marrow is not but is performed the marrow plasma cell percentage must the skeletal are not required for the definition of response, but performed there must be no evidence of progression of bone disease. to response are also not although examinations are examinations are performed as of routine or for other clinical and show evidence of progressive this will constitute relapse or progression in the absence of any other It is recommended therefore that a skeletal be performed to in to ensure that any new lesions were not in fact present at the time of the It is also that the development of a new vertebral compression may result from bone lesions or and does not response nor constitute relapse. data have not been included in the definitions of response and progression because with this is and the significance of different is not CR is defined on the basis of negative IF on serum and for a minimum of 6 Patients who have no detectable paraprotein on EP without a negative IF result either or not will no be classified as CR. A bone marrow plasma cells is also required for the of CR. Although it is that in patients with myeloma it be very to have of the paraprotein with marrow it was important to exclude this morphology of the plasma cells is not specified because morphological assessment was to be It is not essential to a but a is performed this must also plasma In non-secretory myeloma the marrow must be after a to CR. The main requirement in the definition is the absence of detectable paraprotein by IF as as by Since CR is a for in myeloma, it is that CR should require absence of paraprotein by the most sensitive method in routine of minimal disease at the or level may in patients without detectable paraprotein ( et al, 1993 ; Bjorkstrand et al, 1995b ) but the results of such studies are not and are not in and data at present be included in the criteria for CR. There are in the use of IF to remission status. The requirement for monitoring by IF additional laboratory and and many do not IF EP is negative. it is the to that IF be performed EP is negative. A is classified as in CR only a negative IF has been on at a minimum of 6 In patients achieving CR, IF must also be performed at all subsequent in to the time of disease relapse. Most IF months in these using IF rather than EP to define CR will relevant will on the evaluation of outcomes CR as a prognostic this the EBMT and will record EP and IF results and will outcomes in patients in CR and those who are but or A 50% decrease in serum paraprotein is required for as in the CLMTF However, a 50% decrease in urinary light chain excretion was not considered to define Most light are by the and the urinary excretion therefore only the that renal a given degree of tumour reduction has a more on urinary light chain excretion than on serum paraprotein level. & Alexanian observed that in a of patients with serum paraprotein and urinary light a 50% decrease in serum paraprotein level was always by a decrease of in urinary light chain excretion. We have therefore used the SWOG of 90% decrease in urinary light chain excretion to define However, in to the SWOG criteria, it is not for urinary light chain excretion to fall h there has been a 90% reduction. urinary light chain excretion may decrease by and for PR it to < since it is difficult to accurately of light chain excretion which be to a reduction of in patients with an initial light chain excretion of 2 g/24 h or a transient response as CR or PR a minimum period of negative results or stable paraprotein level to be although this is than that required to fulfil the criteria for plateau It has been agreed that 6 will be the minimum required this assessment of response to be at which with the initial data collection Some patients will their response after and in this the response will be on the first The paraprotein level must be compared with a reference in to accurately response. A is to use the paraprotein level to transplant as the reference However, this may to the of a as of a remission a the paraprotein level does not decrease by a Patients in CR who remain in CR be classified as transplant has been performed as of a remission the response will be by comparing the and paraprotein levels with those to the chemotherapy a may from PR to CR, or from PR to or from disease to PR or CR. Patients in CR who remain in CR will be as in complete response. Patients who have not responded to initial chemotherapy nor to subsequent transplant will be classified as no response. For patients who have not chemotherapy within the 6 months to transplant the response to the transplant alone will be by comparing paraprotein levels with those to patients will include some patients with primary disease or in relapse, as as those who have stable for months after of The of plateau is an important prognostic for the outcome of patients treated with conventional It may therefore be important to patients in plateau transplant have a better in to transplant However, this may be difficult to since transplant is now performed as the of a treatment and time may have for stable plateau to be the transplant It may also be important to whether reaching plateau is also of prognostic significance in those patients who do not achieve CR, and whether the of plateau is more important than the degree of partial response It has been agreed that plateau phase will be defined on the basis of stable paraprotein levels for a minimum of 3 as in the current MRC Plateau will require to be within 25% of the response is assessed, a rise 25% of the criteria for disease We have used the progression to a increase in disease in patients in partial remission or plateau whereas the relapse to a of disease in patients previously in CR. is usually defined as an increase of in serum paraprotein or urinary light chain excretion, with reference to the levels at the time of response. However, the paraprotein level or urinary light chain excretion is at a very and it not be to a in serum paraprotein from a level of for example, as evidence of We have therefore defined progression as an increase of in paraprotein or urinary light chain excretion marrow plasma cell percentage in the but in we have minimum in these These criteria have been to a increase in disease it is that many patients will be and may not require treatment at this may also be defined on the basis of increasing marrow infiltration or skeletal as it is not essential to these investigations there is a clinical to do is defined as reappearance of detectable paraprotein or other of disease in patients previously in CR. is a more than progression in these patients as there was no evidence of disease were in CR. Since a negative IF is the for the definition of CR, of on IF on at least relapse, whether or not the paraprotein detectable again by This is a very stringent definition of relapse, since of IF is not always followed by an increase in paraprotein and such patients may remain for a prolonged period ( Bjorkstrand et al, 1995b ). the IF result may only be in some patients, a situation to that of patients with in for may be but not ( et al, 1993 ). In other of IF does not at least in the to clinical disease This sensitive definition of relapse in CR patients at least to a remission duration in CR patients than those who do not enter CR. It will therefore be important to record the time treatment was after progression or relapse, in to whether the criteria are of subsequent disease These criteria for response, progression and relapse have been developed with the of the evaluation of new in multiple myeloma, high-dose therapy with haemopoietic stem cell It is that and subsequent of these criteria, again using an may be in the as are in clinical and as new For the these criteria for complete relapse and progression should a for clinical trials and was by a from the and Gahrton by the Cancer of