Survival and prognostic factors after initiation of treatment in Waldenstrom’s macroglobulinemia

M. A. Dimopoulos+, G. Hamilos, K. Zervas, A. Symeonidis, G. Kouvatseas, P. Roussou, D. Gika, T. Karmiris, K. Bourantas, A. Zomas, C. Mitsouli, I. Xilouri, E. Vervessou, K. Matsis, N. Anagnostopoulos and T. Economopoulos

Department of Clinical Therapeutics, University of Athens School of Medicine, Athens, Greece

Received 15 November 2002; revised 7 January 2003; accepted 14 March 2003


    Abstract
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Backgroud

Waldenstrom’s macroglobulinemia (WM) is an unusual lymphoplasmacytoid lymphoma characterized by the presence of a serum monoclonal immunoglobulin M. Although several studies have evaluated possible prognostic factors of this disease, few have focused on the survival and prognosis of symptomatic patients after the initiation of treatment.

Patients and methods:

Our study included 122 previously untreated patients with a median age of 67 years who required systemic treatment. Multiple variables were analyzed for their prognostic value on survival after initiation of treatment using univariate and Cox regression multivariate analysis.

Results:

The median overall survival was 106 months. Pretreatment factors associated with shorter survival were age >=65 years, splenomegaly, B-symptoms (weight loss, fever or night sweats), hemoglobin <10 g/dl, platelets <100 x 106/dl, albumin <3.5 g/dl and bone marrow lymphoplasmacytic infiltrate >=50%. In the multivariate analysis, the two variables with independent prognostic value were age >=65 years and hemoglobin <10 g/dl. Furthermore, we were able to divide our patients into three risk groups based on the presence of two, one or none of these two adverse prognostic factors. The median survival times in the high-, intermediate- and low-risk groups were 46 months, 107 months and 172 months, respectively (P <0.0001).

Discussion

Our findings suggest that advanced age and anemia appear to be the two dominant prognostic factors for survival after initiation of treatment in patients with WM. These two readily available parameters can stratify the patients into three distinct subgroups and may help the selection of appropriate treatment.

Key words: prognosis, survival, treatment, Waldenstrom’s macroglobulinemia


    Introduction
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Waldenstrom’s macroglobulinemia (WM) is a low-grade lymphoplasmacytoid lymphoma characterized by the production of monoclonal immunoglobulin M (IgM). According to the Revised European American Lymphoma Classification of lymphoid neoplasms, the diagnosis of WM is established by the presence of monoclonal IgM in the serum and infiltration of bone marrow by a proliferation of small lymphocytes, plasmacytoid lymphocytes and plasma cells [1, 2].

WM is an infrequent disease that affects ~1500 Americans each year; it is ~10–20% as common as multiple myeloma. Most patients with WM present with symptoms, signs or other complications that require prompt initiation of systemic treatment in order to reduce the proliferation of the malignant clone. However, several patients are diagnosed by chance and do not have symptoms, anemia, significant hepatosplenomegaly or lymphadenopathy, or complications associated with the presence of monoclonal IgM. Such asymptomatic patients are followed without any treatment for several months or years until there is evidence of disease progression [3].

WM is a relatively indolent disease with a median survival of 5–10 years in different series. Several studies have evaluated clinical and laboratory parameters associated with shorter or longer survival of patients with WM [411]. Most of these studies have included a mixed population of patients consisting of symptomatic patients who received treatment at diagnosis and of asymptomatic patients who required treatment several months or years after diagnosis. Thus, in several studies it was not clear which variables were predictive specifically for the symptomatic patients who needed treatment. We thus designed a retrospective study that included previously untreated patients who required systemic treatment, and we focused on their survival and prognosis after the initiation of treatment.


    Patients and methods
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Between January 1985 and June 2001, 122 patients with WM were identified who had lymphoplasmacytic infiltration of their bone marrow along with a serum monoclonal IgM, and who required systemic treatment. Patients with indolent WM were not included.

A database was constructed, which included the following parameters: age, gender, presence of B-symptoms, splenomegaly, lymphadenopathy, hyperviscosity, cryoglobulinemia, cold agglutinin anemia, amyloidosis, neuropathy, hypercalcemia, bone lesions, or other extranodal involvement. Other parameters included complete blood counts, liver and renal function tests, serum monoclonal protein, serum albumin, quantitation of serum immunoglobulins, serum ß2-microglobulin, type of light chain, Bence Jones proteinuria, serum LDH, and percentage and pattern of bone marrow infiltration. Moreover, the primary indication of the need to start treatment, the type of primary treatment and the response to treatment were coded.

Complete response was defined as the disappearance of the monoclonal protein by immunofixation, resolution of lymphadenopathy and organomegaly, and <20% normal appearing lymphocytes in the bone marrow. Partial response was defined as >=50% reduction of serum monoclonal protein concentration for at least 2 months along with >=50% reduction of tumor infiltrate at all involved sites. The disease was considered stable when the reduction of serum monoclonal protein ranged from 0% to <50% without an increase in lymphadenopathy and organomegaly, and without evidence of additional disease complications. Patients were rated as having progressive disease when they did not meet the criteria for response or stable disease.

Survival was considered from the first day of treatment to the time of death or last follow-up. Several clinical and laboratory characteristics were considered for analysis concerning their individual and simultaneous effects on survival. These parameters were adapted from previous series that evaluated prognostic factors in Waldenstrom’s macroglobulinemia and included: age, gender, hyperviscosity, B-symptoms (weight loss, fever or night sweats), splenomegaly, lyphadenopathy, levels of monoclonal protein, type of light chain, anemia, thrombocytopenia, leukopenia, hypoalbuminemia and bone marrow infiltration. Survival curves were plotted using the Kaplan–Meier method [12]. Statistical comparisons between actuarial survival curves were based on log-rank and Wilcoxon tests [13, 14]. The cut-off point of each variable was selected by comparison of the alternative cut-off points based on the minimum P-value method [15]. Variables considered for possible inclusion in the Cox analysis were those for which there was a significant association in univariate analysis (P <0.05) or for which previous studies had suggested a possible association [16]. The stepwise regression procedure was stopped when the P value for entering an additional factor was >0.01. The model was tested both by expressing values in a continuous way (continuous variables) and by grouping them into categories (dichotomous variables). Since ß2-microglobulin was recorded in 62 patients only, it was not included in the Cox model.


    Results
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Clinical and laboratory characteristics
The median age of the 122 patients was 67 years and the male to female ratio was 1.44 (Table 1). At least one-third of patients had clinical or radiological evidence of splenomegaly and of lymphadenopathy. Approximately one-fifth of patients had B-symptoms before the initiation of treatment. Hyperviscosity was prominent in 24 patients and peripheral polyneuropathy was evident in 10 patients. Symptomatic cryoglobulinemia and cold agglutinin syndrome were seen in a few patients only (Table 1). Amyloidosis was detected in four patients. Involvement of other organs was rare and included infiltration of the parotid gland, the stomach, and the subcutaneous tissue of the face (one patient each).


View this table:
[in this window]
[in a new window]
 
Table 1. Patients and disease features before initiation of treatment
 
Serum monoclonal protein levels ranged from 0.3 to 11.7 g/dl (Table 1). Light-chain type was kappa in two-thirds of patients and light-chain proteinuria was present in 28%. The most frequent hematological abnormality was anemia, which was significant in ~50% of patients. Reduced platelet count to <150 x 106 dl was seen in one-third of patients but thrombocytopenia (platelet count <100 x 109 dl) was found in 13% of patients. Leukopenia [white blood cell count (WBC) <4 x 106/dl] was detected in 16% of patients. Leukocytosis (WBC >10 x 106/dl) was found in 15% and blood lymphocyte count >=5 x 106/dl in 10%. Based on the inclusion criteria of our study, all patients had bone marrow involvement. Extensive infiltration (>=50%) of the marrow by lymphocytes and lymphoplasmacytes was detected in the majority of patients. Pretreatment data on serum ß2-microglobulin were available only in 62 recently diagnosed patients; serum ß2-microglobulin >=3 mg/dl was present in 32 patients (52%).

Primary reason to start treatment and response to treatment
The most common reason to indicate the need for treatment initiation was anemia (Table 2). Symptomatic hyperviscosity necessitated treatment in one-fifth of patients. Significant or increasing lymphadenopathy and/or splenomegaly or the presence of B-symptoms were the primary cause for treatment in several patients. Ninety-five patients (78%) were treated with alkylating agents. Chlorambucil with or without prednisone was the primary treatment in 60% of patients. Some patients received primary treatment that included combinations of alkylating agents with a vinca alkaloid such as cyclophosphamide, vincristine and prednisone (CVP) (nine patients), or with a nitrosurea such as vincristine, cyclophosphamide, melphalan, carmustine and prednisone (VMCBP) (five patients). Eight patients were treated with cyclophosphamide, doxorubicin, vincristine and prednisone (CHOP). Only nine patients were treated up front with fludarabine and 19 patients received miscellaneous treatments, including rituximab in 12 patients. Overall, an objective response (complete or partial response) was documented in 58% of patients. Complete responses were rare and were seen in only five patients. There was no difference in the response rate and survival among the different regimens.


View this table:
[in this window]
[in a new window]
 
Table 2. Primary reason for initiation of treatment
 
Outcome
The median overall survival of our patients from the date of initiation of treatment was 107 months [95% confidence interval (CI) 84 to 141 months] and 46 patients died. Because the majority of our patients were censored at the time of our analysis, we plotted the survival curves of 59 patients who were diagnosed from January 1985 to December 1995, i.e. for patients who had a minimum follow-up of 7 years. The median survival of this cohort of patients was 106 months (95% CI 75 to 141 months). The cause of death was related to the disease or its treatment in 77% of patients, was considered unrelated in 16% of patients, and could not be clarified in 7% of patients.

Multiple parameters were analysed for their possible prognostic impact on patient survival. The univariate analysis showed that the following factors were associated with an impaired survival: age >=65 years, symptomatic disease, splenomegaly, hemoglobin <10 g/dl, platelets <100 x 106/dl, albumin <3.5 g/dl and bone marrow lymphocytes >=50% (Table 3). Other parameters that were not associated with survival included: gender, type of light chain, WBC <4 x 106/dl, levels of monoclonal protein, hyperviscosity and lymphadenopathy. In the multivariate analysis, the two variables with independent prognostic value were the age >=65 years and hemoglobin <10 g/dl (Table 4). Subsequently we divided our patients into three risk groups based on the presence of two, one or none of the adverse prognostic factors. As shown in Table 5, we were able to separate our patient population into three distinct risk groups with very different survival times. Approximately one-quarter of our patients were assigned to a low-risk group with a median survival of 14 years, whereas one-quarter of them, who had both adverse disease features, survived for a median of 4 years (Figure 1).


View this table:
[in this window]
[in a new window]
 
Table 3. Clinical and laboratory parameters associated with survival on univariate analysis (log-rank test)
 

View this table:
[in this window]
[in a new window]
 
Table 4. Multivariate analysis for survival
 

View this table:
[in this window]
[in a new window]
 
Table 5. Scoring system according to hemoglobin and age
 


View larger version (12K):
[in this window]
[in a new window]
 
Figure 1. Survival of patients with Waldenstrom’s macroglobulinemia with low risk (solid line), intermediate risk (dashed line) and high risk (dotted line), according to Table 5.

 
Furthermore, in our study we attempted to validate Morel’s scoring system [7]. As described previously, one-half of our patients belonged to either the low- or intermediate-risk group, and the other half to the high-risk group (Table 6). There was a statistically significant difference in the survival among the three subgroups (Table 6). However, Morel’s scoring system was of greater value when low- and intermediate-risk groups were grouped together and were compared with high-risk patients (Table 6; Figure 2).


View this table:
[in this window]
[in a new window]
 
Table 6. Morel’s scoring system for all patients
 


View larger version (10K):
[in this window]
[in a new window]
 
Figure 2. Survival of patients with Waldenstrom’s macroglobulinemia according to a modified Morel’s scoring system, as defined in Table 6 (low and intermediate, solid line; high, dotted line).

 
Because pretreatment ß2-microglobulin data were available in one-half of our patients and because the follow-up of those patients was short, ß2-microglobulin was not included in the formal analysis of prognostic factors associated with survival. Nevertheless, we observed that among the 29 patients with ß2-microglobulin <3 mg/dl, only four patients died with a median follow-up of 43 months. Furthermore, we performed a correlation between age and ß2-microglobulin and we found that there was evidence of rising serum ß2-microglobulin levels with increasing age (Spearman’s r = 25.34; P = 0.048) (Figure 3 ).



View larger version (14K):
[in this window]
[in a new window]
 
Figure 3. Serum ß2-microglobulin levels rise with increasing age.

 

    Discussion
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
We report the clinical features, response to treatment and outcome of 122 patients with WM who required initiation of treatment. Our inclusion criteria required morphological evidence of lymphoplasmacytic infiltration of the bone marrow, but we did not include a cut-off level for serum monoclonal protein, since several studies have indicated that such a restriction may not be of value [5, 11, 17]. Our analysis of 122 patients with WM requiring treatment confirmed the heterogeneity of clinical syndromes associated with this disease. While most patients required treatment due to tumor infiltration of the marrow, the spleen or the lymph nodes, in approximately one-third of patients the unique properties of monoclonal IgM were responsible for complications such as hyperviscosity, neuropathy, cryoglobulinemia and cold-agglutinin disease.

We restricted our analysis to those patients requiring treatment in order to obtain a more homogeneous patient population. Indeed, most studies reported so far have also included asymptomatic patients who were followed for several months or years without treatment, and have analyzed and compared the outcome of asymptomatic patients with that of patients requiring treatment soon after diagnosis. Among those asymptomatic patients, some individuals with IgM monoclonal gammopathy of undetermined significance might also have been included.

All published series with evaluated prognostic factors for survival in WM are listed in Table 7. In some series, up to 30% of those patients included were asymptomatic. In all but two studies a Cox regression multivariate analysis was performed. A review of these studies indicates that in most series, advanced age and anemia are the dominant adverse prognostic factors. Our data confirmed those observations. Our univariate analysis indicated that age >=65 years, B-symptoms, splenomegaly, hemoglobin <10 g/dl, platelet count <100 x 106/dl, albumin <3.5 g/dl and bone marrow lymphocytosis >=50% were adverse prognostic factors. However, the stepwise regression analysis showed that the two independent variables were advanced age and anemia. By combining these two adverse prognostic factors we were able to stratify our patients into three risk groups (high, intermediate and low risk), with significantly different median survival times of 4, 9 and 14 years, respectively.


View this table:
[in this window]
[in a new window]
 
Table 7. Prognostic factors and survival in WM
 
Furthermore, in our patient population we demonstrated that Morel’s scoring system may be used for prognostic stratification of patients with WM; however, we found that Morel’s scoring system appeared to be more powerful when low and intermediate risk patients were grouped together and compared with those at high risk. Using this modified Morel’s scoring system the patients were separated in two subgroups, with median survival times of 7 and 10 years, respectively.

Two recently published studies of prognostic factors in WM that applied a multivariate analysis have indicated that serum ß2-microglobulin is a significant adverse prognostic factor. Dhodapkar et al. [11] found that elevated serum ß2-microglobulin levels, anemia and low levels of serum monoclonal protein were significant adverse prognostic factors for survival. Garcia-Sanz et al. [8] indicated that high ß2-microglobulin and hyperviscosity were the most significant variables. Baseline levels of serum ß2-microglobulin were available in a fraction of our patient population who were recently diagnosed and in whom few events have occurred. Thus, we were not able to evaluate fully the role of this variable in our patient population. Nevertheless it appears that the 3-year survival of patients with lower levels of ß2-microglobulin is better than that of patients with higher levels. The dominant prognostic value of serum ß2-microglobulin in WM is in accordance with its role in multiple myeloma and chronic lymphocytic leukemia [18, 19]. Although serum ß2-microglobulin is considered a reflection of tumor mass, we observed that there was a significantly positive correlation between age and serum ß2-microglobulin levels. Thus, the prognostic significance of serum ß2-microglobulin may be partially explained by this finding.

We conclude that advanced age and anemia appear to be the two dominant prognostic factors for survival after the initiation of treatment in patients with WM. When baseline serum ß2-microglobulin levels are unavailable, these two readily available variables can stratify the patients into three distinct subgroups, and may help the selection of appropriate treatment and the evaluation of the effect of treatment.


    Footnotes
 
+ Correspondence to: Dr M. A. Dimopoulos, 227 Kifissias Avenue, Kifissia, Athens 14561, Greece. Tel: +30-210-338-1541; Fax: +30-210-813-1383; E-mail: mdimop{at}med.uoa.gr Back


    References
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
1. Harris NL, Jaffe ES, Stein H et al. A revised European–American classification of lymphoid neoplasms: a proposal from the International Lymphoma Study Group. Blood 1994; 84: 1361–1392.[Free Full Text]

2. Harris NL, Jaffe ES, Diebold J et al. World Health Organization classification of neoplastic diseases of the hematopoietic and lymphoid tissues: report of the Clinical Advisory Committee meeting–Airlie House, Virginia, November 1997. J Clin Oncol 1999; 17: 3835–3849.[Abstract/Free Full Text]

3. Dimopoulos MA, Panayiotidis P, Moulopoulos LA et al. Waldenstrom’s macroglobulinemia: clinical features, complications and management. J Clin Oncol 2000; 18: 214–226.[Abstract/Free Full Text]

4. Facon T, Brouillard M, Duhamel A et al. Prognostic factors in Waldenstrom’s macroglobulinemia: a report on 167 cases. J Clin Oncol 1993; 11: 1553–1558.[Abstract]

5. Dimopoulos MA, Alexanian R. Waldenstrom’s macroglobulinemia. Blood 1994; 83: 1452–1459.[Free Full Text]

6. Gobbi RG, Bettini R, Montecucco G et al. Study of prognosis in Waldenstrom’s macroglobulinemia: a proposal for a simple binary classification with clinical and investigational utility. Blood 1994; 83: 2939–2945.[Abstract/Free Full Text]

7. Morel P, Monconduit M, Jacomy D et al. Prognostic factors in Waldenstrom’s macroglobulinemia: a report on 232 patients with the description of a new scoring system and its validation on 253 other patients. Blood 2000; 96: 852–858.[Abstract/Free Full Text]

8. Garcia-Sanz R, Montoto S, Torrequebrada A et al. Waldenstrom’s macroglobulinemia: presenting features and outcome in a series with 217 cases. Br J Haematol 2001; 115: 575–582.[CrossRef][ISI][Medline]

9. Kyrtsonis MC, Vassilakopoulos TP, Angelopoulou MK et al. Waldenstrom’s macroglobulinemia: clinical course and prognostic factors in 60 patients. Ann Hematol 2001; 80: 722–727.[CrossRef][ISI][Medline]

10. Owen RG, Barrans SL, Richards SJ et al. Waldenstorm’s macroglobulinemia: Development of diagnostic criteria and identification of prognostic factors. Am J Clin Pathol 2001; 116: 420–428.[CrossRef][ISI][Medline]

11. Dhodapkar MV, Jacobson JL, Gertz MA et al. Prognostic factors and response to fludarabine therapy in patients with Waldenstrom’s macroglobulinemia: results of United States intergroup trial (Southwest Oncology Group S 9003). Blood 2001; 98: 41–48.[Abstract/Free Full Text]

12. Kaplan EL, Meier P. Non parametric estimation from incomplete observations. J Am Stat Assoc 1958; 53: 457–481.[ISI]

13. Peto R, Peto J. Asymptotically efficient rank invariant test procedures. J Stat Soc 1972; 135: 185–198.[ISI]

14. Gehan EA. Estimating survival functions for the life table. J Chronic Dis 1969; 21: 629–644.[ISI][Medline]

15. Altman DG, Lausen B, Sanerbrei W et al. Dangers of using ‘optimal’ cut points in the evaluation of prognostic factors. J Natl Cancer Inst 1994; 86: 829–835.[ISI][Medline]

16. Cox DR. Regression models in life tables. J R Stat Soc 1972; 34: 187–220.[ISI]

17. Kyle RA, Garton JP. The spectrum of IgM monoclonal gammopathy in 430 cases. Mayo Clin Proc 1987; 62: 719–731.[ISI][Medline]

18. Greipp PR, Lust JA, O’Fallon WM et al. Plasma cell labeling index and b2-microglobulin predict survival independent of thymidine kinase and C-reactive protein in multiple myeloma. Blood 1993; 81: 3382–3387.[Abstract]

19. Zwiebel JA, Cheson BD. Chronic lymphocytic leukemia: staging and prognostic factors. Sem Oncol 1998; 25: 42–59.[ISI][Medline]