Abstract

Background:

An increased incidence of second cancers has been reported in lymphoproliferative disorders.

Patients and methods:

We assessed the frequency, characteristics and predictive factors of second cancers in 230 patients with Waldenström macroglobulinemia (WM) and compared the incidence of second cancers in WM with that of an age- and sex-matched control population.

Results:

Twenty-two patients (10%) developed solid cancers and 10 (4%) second hematologic malignancies. In a competing risk model, the cumulative incidence of solid cancers was 12% at 10 years and 17% at 15 years while the incidence of hematologic malignancies was 6% and 8%, respectively. The overall risk of second cancer in WM was 1.69 times higher than expected (P = 0.002). WM patients were at increased risk for diffuse large B-cell lymphoma [standardized incidence ratio (SIR) 9.24, P < 0.0001], myelodisplastic syndrome/acute myeloid leukemia (SIR 8.4, P < 0.0001), brain cancer (SIR 8.05, P = 0.0004). The risk of a second hematologic malignancy was fourfold higher in patients previously treated, though not reaching statistical significance (P = 0.19).

Conclusions:

WM patients are at higher risk of second cancers as compared with the general population. The sample size does not allow firm conclusions about the effect of therapy on the development of second cancers.

introduction

An increased incidence of second cancers has been reported in non-Hodgkin’s lymphomas (NHLs) [1–5], chronic lymphocytic leukemia (CLL) [6, 7] and hairy cell leukemia (HCL) [8]. Disease-related immunosuppression has been postulated as a possible mechanism predisposing to second cancers [9]. Some studies, however, have raised concerns about the role of therapy in the development of second malignancies, especially when alkylating agents or nucleoside analogs are used [10, 11].

Waldenström macroglobulinemia (WM) is a rare indolent lymphoma characterized by bone marrow infiltration with lymphoplasmacytic cells associated with a serum immunoglobulin M (IgM) monoclonal protein. Some authors have reported a high incidence of transformation to diffuse large B-cell lymphoma (DLBCL) and to myelodysplastic syndrome (MDS)/acute myeloid leukemia (AML) among patients treated with nucleoside analogs [12–16]. There are no data regarding the incidence of second solid tumors in WM. Furthermore, there are no studies comparing the incidence of second cancers in WM with the incidence observed in the general population.

The aims of this study were to assess the frequency, characteristics and predictive factors of second cancers in WM and to evaluate whether WM patients are at higher risk of second cancers as compared with an age- and sex-matched control population.

patients and methods

We retrospectively evaluated 230 consecutive patients with WM seen from January 1980 to December 2009 in two hematologic centers of northern Italy (Division of Hematology, Fondazione IRCCS Policlinico San Matteo, Pavia and Division of Hematology, Niguarda Ca’ Granda Hospital, Milano). The study was approved by the Institutional Review Board and conducted in accordance to the Helsinki Declaration of 1964, as revised in 2000. According to the Consensus Panel Recommendations from the second International Workshop on WM, the diagnosis of WM was based on the histopathological demonstration of lymphoplasmacytic lymphoma on bone marrow biopsy associated with a serum IgM monoclonal component of any size [17]. WM patients were defined symptomatic if they had symptoms attributable either to tumor infiltration or to the monoclonal component.

All second cancers occurring in WM patients were recorded with the exception of nonmelanoma skin cancers and in situ cervical cancer. To evaluate the risk of second cancers in WM, only cancers diagnosed after WM were considered in the statistical analysis. The incidence of second cancers per person per year in the study population was compared with the incidence of malignancies in the northern Italy population applying the sex- and age-specific incidence rates reported by Associazione Italiana Registri Tumori (AIRTUM). AIRTUM (www.registri-tumori.it) is a national database that collects and publishes data on tumors from single accredited Italian tumor registries. The standardized incidence ratio (SIR) was calculated as the ratio between the observed and the expected numbers of cases; 95% confidence intervals (CIs) were based on the assumption that the observed numbers of second cancers were distributed as Poisson variables. The time free of second cancers was calculated from the date of diagnosis of WM to the date of last follow-up or the date of diagnosis of second cancer. The cumulative incidence of second cancers was estimated in a competing risk model, considering solid tumors, hematologic malignancies and death from any cause as competing events. Cox proportional hazards regression was used to evaluate risk factors for second cancers. Therapy was analyzed as a time-dependent covariate. Statistical analyses were carried out using Statistica 8 software and Microsoft Excel 2000. A P value of 0.05 was considered statistically significant.

results

patients’ characteristics at diagnosis

A total of 230 patients with WM were included in the analysis. Their main characteristics at presentation are reported in Table 1. Seventeen patients (7%) had a history of one (n = 15) or two (n = 2) solid cancers at the time of the diagnosis of WM. The median time between the diagnosis of cancer and that of WM was 18.7 months (range 3–293). The types/sites of prior cancers were as follows: gastrointestinal (n = 6), breast (n = 4), urinary tract (n = 3), mouth (n = 2), lung (n = 1), prostate (n = 1), brain (n = 1) and thyroid (n = 1). None of the patients had a history of other hematologic malignancies.

Table 1.

Patients’ characteristics at diagnosis

Characteristics Assessable patients Result 
Age (years), median (range) 230 61 (26–86) 
Sex, number of patients (%)   
    Male 230 140 (61) 
    Female  90 (39) 
White blood cells (×109/l) 102 7.2 (2.8–29.6) 
Hemoglobin (g/dl), median (range) 196 12.7 (4.2–16.4) 
Platelets (×109/l), median (range) 195 255 (53–800) 
Serum albumin (g/l), median (range) 168 43 (20–56) 
Erythrocyte sedimentation rate (mm/h), median (range) 156 70 (1–150) 
Lactate dehydrogenase (IU/ml), median (range) 154 304 (86–1039) 
β2-microglobulin (mg/l), median (range) 157 2.3 (0.27–14.5) 
Type of light chain   
    k 207 164 (79) 
    λ  37 (18) 
    k + λ  6 (3) 
Size of serum M protein (g/l), median (range) 196 20 (3–109) 
Uninvolved immunoglobulins reduction, number of patients (%) 
    Immunoglobulin G 179 77 (43) 
    Immunoglobulin A  24 (13) 
Detectable urine M protein, number of patients (%) 147 73 (50) 
Bone marrow lymphoplasmacytic infiltrate (%), median (range) 230 40 (5–100) 
Characteristics Assessable patients Result 
Age (years), median (range) 230 61 (26–86) 
Sex, number of patients (%)   
    Male 230 140 (61) 
    Female  90 (39) 
White blood cells (×109/l) 102 7.2 (2.8–29.6) 
Hemoglobin (g/dl), median (range) 196 12.7 (4.2–16.4) 
Platelets (×109/l), median (range) 195 255 (53–800) 
Serum albumin (g/l), median (range) 168 43 (20–56) 
Erythrocyte sedimentation rate (mm/h), median (range) 156 70 (1–150) 
Lactate dehydrogenase (IU/ml), median (range) 154 304 (86–1039) 
β2-microglobulin (mg/l), median (range) 157 2.3 (0.27–14.5) 
Type of light chain   
    k 207 164 (79) 
    λ  37 (18) 
    k + λ  6 (3) 
Size of serum M protein (g/l), median (range) 196 20 (3–109) 
Uninvolved immunoglobulins reduction, number of patients (%) 
    Immunoglobulin G 179 77 (43) 
    Immunoglobulin A  24 (13) 
Detectable urine M protein, number of patients (%) 147 73 (50) 
Bone marrow lymphoplasmacytic infiltrate (%), median (range) 230 40 (5–100) 

treatment and survival

During the course of the disease, 137 patients (60%) were treated. They received a median of two lines of therapy (range 1–7). Considering all lines of therapies, alkylating agents were given in 134 of 137 patients (98%) and nucleoside analogs in 38 (28%). Rituximab was associated to chemotherapy in 46 patients (34%). The median follow-up was 5.2 years (range 0.5–29.2) after therapy with alkylating agents, 2.9 years (range 0.5–7.1) after nucleoside analogs and 2.4 years (range 0.5–6.3) after rituximab. A watch and wait policy was adopted in 93 patients (40%) with asymptomatic WM (Table 2).

Table 2.

Treatment characteristics

Characteristics  
Any therapy, number of patients (%) 137 (60) 
Watch and wait, number of patients (%) 93 (40) 
Number of therapies, median (range) 2 (1–7) 
Type of therapy, number of patients (% of patients treated)  
    Alkylating agents 134 (98) 
    Nucleoside analogs 38 (28) 
    Monoclonal antibodies 46 (34) 
Follow-up after therapy (years), median (range)  
    Alkylating agents 5.2 (0.5–29.2) 
    Nucleoside analogs 2.9 (0.5–7.1) 
    Monoclonal antibodies 2.4 (0.5–6.3) 
Characteristics  
Any therapy, number of patients (%) 137 (60) 
Watch and wait, number of patients (%) 93 (40) 
Number of therapies, median (range) 2 (1–7) 
Type of therapy, number of patients (% of patients treated)  
    Alkylating agents 134 (98) 
    Nucleoside analogs 38 (28) 
    Monoclonal antibodies 46 (34) 
Follow-up after therapy (years), median (range)  
    Alkylating agents 5.2 (0.5–29.2) 
    Nucleoside analogs 2.9 (0.5–7.1) 
    Monoclonal antibodies 2.4 (0.5–6.3) 

At a median follow-up of 4.8 years (range 0.5–29.3), 171 of 230 patients (74%) were alive and 59 (26%) had died. The median overall survival (OS) of the whole series was 13.4 years with a 10-year survival rate of 65% (95% CI 56% to 73%). In univariate analysis, prognostic factors for OS were age (P = 0.001), hemoglobin (P ≤ 0.0001), β2-microglobulin (P = 0.001), M-component size (P = 0.0002) and degree of bone marrow infiltration (P = 0.006). In multivariate analysis, only age retained its prognostic value (P = 0.01), while hemoglobin had a borderline significance (P = 0.057).

incidence of second cancers in WM patients

Overall, 32 of 230 patients (14%) developed a second cancer after a median time of 51 months (range 6–157) from diagnosis of WM.

solid cancers.

Twenty-two of 230 patients (10%) developed a solid cancer at the following sites: lung (n = 5), gastrointestinal (n = 4), prostate (n = 4), urinary tract (n = 4), breast (n = 2), brain (n = 2; one glioma and one meningioma) and thyroid (n = 1). None of these second solid tumors represented a recurrence of cancers diagnosed before WM. The median time from the diagnosis of WM to the occurrence of a subsequent solid cancer was 53 months (range 6–157). After correction in a competing risk model, the 10 and 15 years cumulative incidences of solid tumors were 12% and 17%, respectively (Figure 1).

Figure 1.

Cumulative incidence of second cancers divided into solid and hematologic malignancies.

Figure 1.

Cumulative incidence of second cancers divided into solid and hematologic malignancies.

hematologic malignancies.

Ten of 230 patients (4%) developed a second hematologic malignancy, represented by DLBCL (n = 6, 2.6%), MDS/AML (n = 3, 1.3%) and chronic myeloid leukemia (CML) (n = 1, 0.5%). The median time from the diagnosis of WM to a subsequent hematologic malignancy was 49 months (range 24–145). When competing risks were taken into account, the 10 and 15 years cumulative incidences of hematologic malignancies were 6% and 8%, respectively (Figure 1). Five of six patients who developed DLBCL had been previously treated for WM. Four of them had received alkylating agents only and developed DLBCL, respectively, 2, 49, 50 and 143 months after the start of therapy; the other patient had received alkylating agents first and six courses of fludarabine, cyclophosphamide, rituximab (FCR) at relapse and developed DLBCL 8.9 years after initial treatment and 4 years after FCR.

All three cases of MDS/AML occurred in patients treated for WM: two patients developed AML 2.8 and 5.9 years after treatment with chlorambucil and both died within 2 months from diagnosis. The third patient, who had received chlorambucil as first-line therapy and six courses of FCR at progression, developed MDS 4.5 years after initial treatment and 11 months after FCR. The cytogenetic pattern using conventional karyotyping showed t(2;5), 6q and 5p deletion. He died 14 months after diagnosis of MDS.

risk of second cancer in WM patients with respect to the general population

As compared with an age- and sex-matched population, the overall risk of second cancer in WM was 1.69 times higher than expected (95% CI 1.19–2.38, P = 0.002). The relative risk of a second cancer was not statistically different between males (SIR 1.77, 95% CI 1.18–2.66) and females (SIR 1.50, 95% CI 0.78–2.89) (P = 0.67). The analysis of SIRs showed that WM patients were at increased risk for DLBCL (SIR 9.24, 95% CI 4.15–20.5, P < 0.0001), MDS/AML (SIR 8.4, 95% CI 2.7–26, P < 0.0001) and brain cancer (SIR 8.05, 95% CI 2.01–32.19, P = 0.0004). There was also a borderline significance for thyroid cancer (SIR 5.52, 95% CI 0.78–39.21, P = 0.05). The risk of urinary, lung and prostate cancers was also higher as compared with the general population, albeit not significantly (Table 3).

Table 3.

Types of second cancers and cancer-specific SIRs

Type of cancer Cases observed SIR 95% CI P value 
All cancers 32 1.69 1.19–2.38 0.002 
DLBCL 9.24 4.15–20.5 <0.0001 
Lung 1.74 0.7–4.18 0.21 
Gastrointestinal 0.76 0.28–2.02 0.58 
Urinary tract 2.30 0.86–6.1 0.08 
Prostate 1.36 0.51–3.61 0.54 
MDS/AML 8.4 2.7–26 <0.0001 
Breast 1.08 0.27–4.3 0.91 
Brain 8.05 2.01–32.19 0.0004 
Thyroid 5.52 0.78–39.21 0.05 
CML NA NA NA 
Type of cancer Cases observed SIR 95% CI P value 
All cancers 32 1.69 1.19–2.38 0.002 
DLBCL 9.24 4.15–20.5 <0.0001 
Lung 1.74 0.7–4.18 0.21 
Gastrointestinal 0.76 0.28–2.02 0.58 
Urinary tract 2.30 0.86–6.1 0.08 
Prostate 1.36 0.51–3.61 0.54 
MDS/AML 8.4 2.7–26 <0.0001 
Breast 1.08 0.27–4.3 0.91 
Brain 8.05 2.01–32.19 0.0004 
Thyroid 5.52 0.78–39.21 0.05 
CML NA NA NA 

SIR, standardized incidence ratio; CI, confidence interval; DLBCL, diffuse large B-cell lymphoma; MDS/AML, myelodysplastic syndrome/acute myeloid leukemia; CML, chronic myeloid leukemia; NA, not available.

risk factors for second cancers

In univariate analysis, the risk of second cancer was not influenced by age (P = 0.91), sex (P = 0.45), hemoglobin levels (P = 0.6), erythrocyte sedimentation rate (P = 0.62), lactate dehydrogenase (P = 0.91), β2-microglobulin (P = 0.25), size of serum monoclonal component (P = 0.06), serum albumin (P = 0.65), reduction of immunoglobulin G (IgG) (P = 0.91) or immunoglobulin A (IgA) (P = 0.58), degree of bone marrow infiltration (P = 0.98) and presence or absence of symptoms (P = 0.23).

Although 20 of 32 patients (62%) who developed a second cancer had been previously treated for WM, the risk of second cancer was similar in previously treated and untreated patients [hazard ratio (HR) 1.05, 95% CI 0.5–2.1, P = 0.88]. When only hematologic malignancies were considered, the risk was 3.9 times higher in previously treated than in untreated patients, even though the difference did not reach statistically significance (HR 3.9, 95% CI 0.5–31, P = 0.19). As far as the type of therapy is concerned, the risk of developing a second cancer was not influenced by prior exposure to alkylating agents (HR 1.06, 95% CI 0.5–2.2, P = 0.87) or to purine analogs (HR 1.95, 95% CI 0.7–5.6, P = 0.21).

discussion

Several studies demonstrated an increased incidence of second cancers in lymphoproliferative diseases such as NHL [1–5], CLL [6, 7] and HCL [8]. In WM, information on the occurrence of second cancers is sparse and limited to the risk of therapy-related myelodisplastic syndrome/acute myeloid leukemia and transformation to DLBCL. To the best of our knowledge, this is the first study focusing on the overall incidence of second cancers—either hematologic or solid—in WM and evaluating whether WM patients are at higher risk of second cancers as compared with the general population.

We retrospectively studied 230 WM patients diagnosed and followed in two hematologic centers of northern Italy. We are aware that this patient population, mainly treated with alkylating agents, does not reflect the current therapeutic strategies for WM, where rituximab-based treatments and nucleoside analogs are more frequently used. This fact should to be taken into account in the interpretation of our findings.

The overall incidence of second malignancies in WM patients was significantly increased as compared with an age- and sex-matched population living in the same geographical area. In particular, WM patients were found to be at higher risk of developing second hematologic malignancies such as DLBCL and MDS/AML. The analysis of cancer-specific SIR showed also a significant higher incidence of brain malignancies and a trend for a higher incidence of thyroid cancer. The risk of lung, urinary and prostate cancers was also increased but not significantly. However, these findings should be interpreted cautiously as the number of patients included in this study is probably too small to allow the assessment of small risks, which is the main limitation of cohort studies as compared with population-based studies including thousands of patients.

After correction in a competing risk model, the 15-year cumulative incidence of second was 12% for solid cancers and 8% for hematologic malignancies, higher than reported in prior studies in other NHLs [1, 4, 5] (Figure 1). In our series, the incidence of MDS/AML was 1.3%, in keeping with two prior studies including patients treated exclusively with alkylating agents where MDS/AML was reported to occur in <2% of cases [18, 19]. All three cases of MDS/AML were observed in patients previously treated only with alkylators or with alkylators and nucleoside analogs sequentially. In recent years, some studies have raised concerns about a possible role of nucleoside analogs in the pathogenesis of MDS/AML. In a retrospective study on 439 patients, the incidence of MDS/AML was 1.6% in patients treated with nucleoside analogs, while no cases where observed in patients treated with other regimens and in untreated patients [16].

We observed five cases of DLBCL, four after therapy with alkylating agents and one after therapy with alkylators plus nucleoside analogs. Transformation to DLBCL (Richter’s syndrome) that represents the natural evolution of CLL [20] has been reported also in WM. The incidence of DLBCL in WM ranges from <1% to 9% in different studies [12–14, 16]. Some authors have reported a high incidence of Richter’s transformation in WM patients exposed to nucleoside analogs [12–16, 21]. Actually, in the largest study published so far, the incidence of transformation was significantly higher in WM patients treated with nucleoside analogs as compared with those treated with other regimens (4.6% versus 0.4%, P < 0.001) and no cases were observed in untreated patients. On the opposite, in a prospective randomized trial, the incidence of DLBCL was not statistically different in patients treated with nucleoside analogs or with alternative regimens. In our series, the crude incidence of DLBCL of 2.7% is consistent with previous reports, taking into account that the majority of patients were treated with alkylating agents and only 28% received purine analogs.

In this study, none of the clinical and hematologic characteristics of patients was found to predict the occurrence of second cancers. From a statistical point of view, the incidence of second cancers was not significantly different between asymptomatic and symptomatic WM and between treated and untreated patients. However, about two-thirds of second cancers were diagnosed in previously treated patients. In particular, the risk of developing a second hematologic malignancy was almost fourfold higher in treated than in untreated patients. The lack of statistical significance is probably due to the sample size and therefore, we cannot rule out a treatment effect in the development of second cancers. In particular, the number of patients treated with nucleoside analogs is too low and their follow-up is too short to allow conclusions regarding the role of nucleoside analogs. An alternative explanation for the preponderance of second cancers in WM patients is disease-related immune suppression. The immunologic impairment associated to lymphoproliferative disorders could contribute to the pathogenesis of WM as well as to the development of additional malignancies. Interestingly, IgA and IgG levels in WM are usually subnormal at diagnosis and tend to remain low also in patients achieving complete remission after therapy. Treon et al. [15] demonstrated that IgA and IgG hypogammaglobulinemia is a constitutive feature of patients with WM and may be related to mutations associated with common variable immunodeficiency disorder [21]. In our study, reduction of IgG and IgA at diagnosis was not associated with a higher risk of second cancers.

In conclusion, WM patients are at higher risk of second cancers as compared with the general population. In particular, they are at increased risk of DLBCL, MDS/AML and brain cancers. Multicentric studies with large number of patients and longer follow-up are needed to clarify whether the increased incidence of second cancers is related to treatments or to the immunologic impairment associated with the disease. Regardless of the underlying pathogenetic mechanisms, the awareness of an increased risk of second cancers suggests a careful oncohematological surveillance of WM patients.

disclosure

The authors declare no conflict of interest.

References

1.
Travis
LB
Curtis
RE
Glimelius
B
, et al.  . 
Second cancers among long-term survivors of non-Hodgkin’s lymphoma
J Natl Cancer Inst
 , 
1993
, vol. 
85
 (pg. 
1932
-
1937
)
2.
Brennan
P
Scelo
G
Hemminki
K
, et al.  . 
Second primary cancers among 109000 cases of non-Hodgkin’s lymphoma
Br J Cancer
 , 
2005
, vol. 
93
 (pg. 
159
-
166
)
3.
Tward
JD
Wendland
MM
Shrieve
DC
, et al.  . 
The risk of secondary malignancies over 30 years after the treatment of non-Hodgkin lymphoma
Cancer
 , 
2006
, vol. 
107
 
1
(pg. 
108
-
115
)
4.
Mudie
NY
Swerdlow
AJ
Higgins
CD
, et al.  . 
Risk of second malignancy after non-Hodgkin’s lymphoma: a British cohort study
J Clin Oncol
 , 
2006
, vol. 
24
 
10
(pg. 
1568
-
1574
)
5.
Sacchi
S
Marcheselli
L
Bari
A
, et al.  . 
Secondary malignancies after treatment for indolent non-Hodgkin’s lymphoma: a 16-year follow-up study
Haematologica
 , 
2008
, vol. 
93
 
3
(pg. 
398
-
404
)
6.
Travis
LB
Curtis
RE
Hankey
BF
, et al.  . 
Second cancers in patients with chronic lymphocytic leukemia
J Natl Cancer Inst
 , 
1992
, vol. 
84
 (pg. 
1422
-
1427
)
7.
Tsimberidou
AM
Wen
S
McLaughlin
P
, et al.  . 
Other malignancies in chronic lymphocytic leukemia/small lymphocytic lymphoma
J Clin Oncol
 , 
2009
, vol. 
27
 
6
(pg. 
904
-
910
)
8.
Hisada
M
Chen
BE
Jaffe
ES
Travis
LB
Second cancer incidence and cause-specific mortality among 3104 patients with hairy cell leukemia: a population-based study
J Natl Cancer Inst
 , 
2007
, vol. 
99
 (pg. 
215
-
222
)
9.
Greene
MH
Hoover
RN
Fraumeni
JF
Jr
Subsequent cancers in patients with chronic lymphocytic leukaemia: a possible immunologic mechanism
J Natl Cancer Inst
 , 
1978
, vol. 
61
 (pg. 
337
-
340
)
10.
Cheson
BD
Vena
DA
Barrett
J
, et al.  . 
Second malignancies as a consequence of nucleoside analogs therapy for chronic lymphocytic leukaemias
J Clin Oncol
 , 
1999
, vol. 
17
 (pg. 
2454
-
2460
)
11.
Thornton
PD
Bellas
C
Santon
A
, et al.  . 
Richter’s transformation of chronic lymphocytic leukaemia: the possible role of fludarabine and the Epstein-Barr virus in its pathogenesis
Leuk Res
 , 
2005
, vol. 
29
 (pg. 
389
-
395
)
12.
LeBlond
V
Ben-Othman
T
Deconinck
E
, et al.  . 
Activity of fludarabine in previously treated Waldenström’s macroglobulinemia: a report of 71 cases. Groupe Cooperatif Macroglobulinemie
J Clin Oncol
 , 
1998
, vol. 
16
 (pg. 
2060
-
2064
)
13.
Leblond
V
Levy
V
Maloisel
F
, et al.  . 
Multicenter, randomized comparative trial of fludarabine and the combination of cyclophosphamide-doxorubicin-prednisone in 92 patients with Waldenström macroglobulinemia in first relapse or with primary refractory disease
Blood
 , 
2001
, vol. 
98
 (pg. 
2640
-
2644
)
14.
Tamburini
J
Levy
V
Chaleteix
C
, et al.  . 
Fludarabine plus cyclophosphamide in Waldenström’s macroglobulinemia: results in 49 patients
Leukemia
 , 
2005
, vol. 
19
 (pg. 
1831
-
1834
)
15.
Treon
SP
Branagan
AR
Ioakimidis
L
, et al.  . 
Long-term outcomes to fludarabine and rituximab in Waldenström macroglobulinemia
Blood
 , 
2009
, vol. 
113
 
16
(pg. 
3673
-
3678
)
16.
Leleu
X
Soumerai
J
Roccaro
A
, et al.  . 
Increased incidence of transformation and myelodysplasia/acute leukemia in patients with Waldenström macroglobulinemia treated with nucleoside analogs
J Clin Oncol
 , 
2009
, vol. 
27
 
2
(pg. 
250
-
255
)
17.
Owen
RG
Treon
SP
Al-Katib
A
, et al.  . 
Clinicopathological definition of Waldenström’s macroglobulinemia: consensus panel recommendations from the Second International Workshop on Waldenström’s Macroglobulinemia
Semin Oncol
 , 
2003
, vol. 
30
 (pg. 
110
-
115
)
18.
Garcia-Sanz Montoto
S
Torrequebrada
A
de Coca
AG
, et al.  . 
Spanish Group for the Study of Waldenström Macroglobulinaemia and PETHEMA (Programme for the Study and Treatment of Haematological Malignancies)
Waldenström macroglobulinaemia: presenting features and outcome in a series with 217 cases
Br J Haematol
 , 
2001
, vol. 
115
 
3
(pg. 
575
-
582
)
19.
Facon
T
Brouillard
M
Duhamel
A
, et al.  . 
Prognostic factors in Waldenström’s macroglobulinemia: a report of 167 cases
J Clin Oncol
 , 
1993
, vol. 
11
 
8
(pg. 
1553
-
1558
)
20.
Tsimberidou
AM
Keating
MJ
Richter syndrome: biology, incidence and therapeutic strategies
Cancer
 , 
2005
, vol. 
103
 (pg. 
216
-
228
)
21.
Hunter
ZR
Manning
RJ
Hanzis
C
, et al.  . 
IgA and IgG hypogammaglobulinemia in Waldenström’s macroglobulinemia
Haematologica
 , 
2010
, vol. 
95
 
3
(pg. 
470
-
474
)