OUP user menu

Myeloid Neoplasms Secondary to Plasma Cell Myeloma: An Intrinsic Predisposition or Therapy-Related Phenomenon?
A Clinicopathologic Study of 41 Cases and Correlation of Cytogenetic Features With Treatment Regimens

Deepti M. Reddi MD, Chuanyi M. Lu MD, George Fedoriw MD, Yen-Chun Liu MD, Frances F. Wang MS, Scott Ely MD, Elizabeth L. Boswell MD, Abner Louissaint Jr MD, PhD, Murat O. Arcasoy MD, Barbara K. Goodman PhD, Endi Wang MD, PhD
DOI: http://dx.doi.org/10.1309/AJCPOP7APGDT9JIU 855-866 First published online: 1 December 2012


We describe 41 cases of myeloid neoplasms (MNs) secondary to plasma cell myeloma (PCM). The types of MN included myelodysplastic syndrome (MDS) in 34 (82.9%), acute myeloid leukemia (AML) in 4 (9.8%), and myeloproliferative neoplasm (MPN) or MDS/MPN in 3 (7.3%) cases. The latency from treatment to diagnosis of MN ranged from 9 to 384 months, with a median of 60 months. Of 37 cases with cytogenetic studies, complex abnormalities were detected in 22 (59.5%), −5(q)/−7(q) in 4 (10.8%), other abnormalities in 8 (21.6%), and normal karyotype in 3 (8.1%) cases. Complex abnormalities and −5(q)/−7(q) correlated directly with multiple chemotherapeutic regimens, particularly with combined melphalan/cyclophosphamide. Moreover, the features of cytogenetic abnormalities in our series were significantly different from those with concomitant PCM/MN who had significantly lower complex abnormalities. The latency, skewed proportion of MDS, and bias toward complex cytogenetic abnormalities/unbalanced aberrations of chromosomes 5/7 suggested an alkylating mutagenic effect on pathogenesis of secondary MN. Kaplan-Meier survival analysis demonstrated a median survival of 19 months, which was better than that for therapy-related (t)–MDS/AML. In contrast to t-MDS, the survival in our patients appeared to depend on subtypes of MDS as seen in de novo diseases.

Key Words:
  • Plasma cell myeloma
  • Myelodysplastic syndrome
  • Acute myeloid leukemia
  • Myeloproliferative neoplasm
  • Myeloid neoplasm
  • Therapy-related

Plasma cell myeloma (PCM) has been the second most common hematologic malignancy in Western countries for the past 15 years.13 The clinical outcome of patients with PCM before the introduction of alkylating agents was poor, with a median survival of 7 months.2,3 During the last 2 decades, high-dose chemotherapy with or without total body irradiation followed by autologous hematopoietic stem cell transplantation (auto-HSCT) and the newer immunomodulatory agents such as thalidomide, lenalidomide, and bortezomib have been introduced and dramatically improved progression-free survival and overall survival of this patient population, with the rate of complete remission increased to 80%.3,4 With overall survival in this patient population improving because of these new treatment modalities, therapy-related neoplasms, as late effects of the treatment, or secondary neoplasms closely associated with PCM become more evident. Since the discovery of alkylating agents and other regimens for treatment of PCM, reports of therapy-related myeloid neoplasms (t-MNs) have increased, which include acute myeloid leukemia (AML), myelodysplastic syndrome (MDS), and myeloproliferative neoplasm (MPN).49 Although some PCM treatment regimens have been known to have mutagenic effects and potentially induce myeloid neoplasms (MNs), their exact contribution to the development of secondary myeloid malignancies is uncertain owing to the possible intrinsic predisposition to MN in patients with PCM. In addition, the quantitative relationships with the dose and duration of treatment, the relative oncogenic potential of different drugs, and drug-drug interactions as well as the pathologic features of secondary MNs have not been well characterized to date. Here, we report a retrospective analysis of 41 cases of MN secondary to PCM from 4 tertiary medical centers.

Materials and Methods

Case Selection

With institutional approval, 41 cases of MN secondary to PCM or related diseases were identified from our bone marrow biopsy database using the search phrase “myelodysplastic,” “leukemia,” or “myeloproliferative,” in combination with “myeloma” or “plasma cell dyscrasia.” These included 17 cases from Duke University Medical Center (Durham, NC), 10 cases from University of California at San Francisco Medical Center, 8 cases from University of North Carolina at Chapel Hill, and 6 cases from Cornell University Medical Center (Ithaca, NY). The diagnoses of secondary leukemia or MDS were confirmed according to the 2008 World Health Organization (WHO) classification.10,11 The primary diagnosis, treatment for PCM or related diseases, treatment for secondary MNs, and other clinical information were collected from corresponding patients’ medical records.

Cytomorphologic and Histologic Evaluation

Peripheral blood smears were stained with Wright stain, bone marrow aspirate smears and biopsy touch imprints were stained with Wright-Giemsa stain, and bone marrow core biopsies and clot sections were stained with H&E. The cases were reviewed independently by 4 hematopathologists (E.W., C.M.L., G.F., and S.E.). Morphologic dysplasia and blast count were evaluated on peripheral blood smears and aspirate smears. The presence of dysplastic changes in the erythroid, granulocytic, and megakaryocytic lineages were determined according to the description in the 2008 WHO Classification.11 Bone marrow cellularity was estimated on core biopsy or on clot section in cases in which the biopsy quality was suboptimal, as previously described.12 Briefly, bone marrow cellularity was defined as hypercellular when it was higher than 1 standard deviation (SD) above, hypocellular when lower than 1 SD, and normocellular when within 1 SD of the age-adjusted mean. Reticulin stain was routinely performed on the core biopsy of each case and evaluated based on the 2008 WHO criteria.

Conventional Cytogenetic Studies

Cytogenetic analysis was performed on 2 to 4 mL of bone marrow aspirate from each case. Two cultures from each specimen were initiated from the fresh, anticoagulated specimen in complete tissue culture medium. The cells were incubated for 24 and 48 hours. Both B-cell–stimulated and unstimulated cultures were initiated and examined. Chromosome preparations, including harvesting and Giemsa banding by using trypsin with Wright stain (GTW banding), were made using standard methods. Cytogenetic abnormalities were classified according to the International System for Human Cytogenetic Nomenclature.

Interphase Fluorescence In Situ Hybridization (FISH)

For Detecting Interstitial Deletion of Long Arm of Chromosome 5

Fluorescence in situ hybridization (FISH) analysis was performed using the dual-color Vysis probe (Abbott Molecular, Abbott Park, IL) specific for chromosome 5 at a locus on the short arm (LSI DD5S23/D5S721) at band 5p15.2 and on the long arm (EGR1) at band 5q31. The EGR1 probe was used to identify interstitial deletions, and the LSI DD5S23/D5S721 probe serves to determine the number of copies of chromosome 5.

For Detecting Monosomy 7 or Interstitial Deletion of Long Arm of Chromosome 7

FISH analysis was performed using the dual-color Vysis probe specific for the chromosome 7 centromere (CEP 7) and a locus on the long arm (LSI D7S486) at band 7q31. The LSI D7S486 probe was used to identify interstitial deletions, and the CEP 7 (centromeric) probe served to determine the number of copies of chromosome 7.

For Detecting Cytogenetic Abnormalities Seen Frequently in PCM

FISH analysis was performed using a panel of probes from Abbott Molecular to detect loss or gain of specific loci in chromosomes. This panel is composed of 2 probe sets with multiple colors. Probe set 1 targets the TP53 locus at 17p13.1 and ATM locus at 11q22.3. Probe set 2 targets the D13S319 locus at 13q14.3, the LAMP1 locus at 13q34 (13qtel), and the centromeric region of chromosome 12 (12p11.1–q11). FISH analysis was also performed using dual-color, dual-fusion probes manufactured by Abbott Molecular to detect CCND1/IGH fusion associated with the 11;14 translocation and FGFR3/IGH fusion associated with the 4;14 translocation. These probe sets target the FGFR3 (4p16.3), CCND1 (11q13), and IGH (14q32) loci.

For each interphase FISH analysis, a total of 200 inter-phase nuclei per probe set were examined and percentage of nuclei with abnormal signals was calculated.

JAK2 Mutation Analysis

Testing for the JAK2 V617F point mutation was performed on genomic DNA samples extracted from the blood granulocytes. This test is based on the amplification refractory mutation system (ARMS) technique using a DNA tetra-primer ARMS assay, a method that uses 2 oligonucleotide primer pairs to specifically amplify the normal and mutant sequences as well as a positive control DNA sample in a single polymerase chain reaction. The amplified products were analyzed on an ethidium bromide–stained agarose gel.

Statistical Analysis

Statistical analyses were performed with SAS version 9 (SAS Institute Inc, Cary, NC). The Student t test, Mann-Whitney-Wilcoxon test, and χ2 test were used to test the statistical significance in differences between the groups. P value less than .05 was used to consider the differences between the compared groups to be statistically significant. Overall survivals in stratified groups were analyzed using Kaplan-Meier survival analysis.


Clinical Information

Forty-one cases of MN secondary to PCM or related diseases were collected for this study Table 1. These included 17 cases from Duke University Medical Center between 1999 and 2009, during which 811 cases of PCM and 76 cases of plasmacytoma were diagnosed and registered for treatment. Of 41 cases, 29 (70.7%) were male and 12 (29.3%) were female. Patient age at the diagnosis of PCM or related diseases ranged from 33 to 77 years, with a median of 56 (mean ± SD, 56.7 ± 9.2) years. Serum electrophoresis data were available in 38 cases, including immunoglobulin (Ig) G in 24 (63.2%), IgA in 10 (26.3%), IgD in 1 (2.6%), and free light chain in the remaining 3 (7.9%) cases. Thirty-four cases had light chains detected on either serum free light chain quantitation or immunofixation electrophoresis, including κ light chain in 23 (67.6%) cases and λ light chain in 11 (32.4%) cases, respectively. Myeloma treatment information was available for all cases. These included chemotherapy and/or radiotherapy in 38 (92.7%) cases, immunomodulatory therapy alone in 2 (4.9%) cases, and no myeloma-pertinent treatment in the remaining 1 (2.4%) case. Of 38 cases with chemotherapy/radiotherapy, 28 (73.7%) received combination chemotherapy with or without radiotherapy, and 10 (26.3%) cases had melphalan with or without immunomodulatory agents. When sorted by specific chemotherapeutic regimens, 26 (92.9%) of 28 cases with combination chemotherapy received alkylating agents melphalan and/or cyclophosphamide. Three patients (10.7%) received other alkylating agents. Of these, 18 (43.9% of total 41) cases eventually underwent auto-HSCT, and 4 (9.8%) cases received allogeneic HSCT (allo-HSCT). In the patients with auto-HSCT the preparative regimens typically included high-dose melphalan and/or cyclophosphamide and granulocyte colony-stimulating factor, which was used to mobilize hematopoietic stem cells. In the patients with combination chemotherapy, the other common regimens used were vincristine, doxorubicin, and dexamethasone (VAD) with or without immunomodulators. In summary, melphalan was used in 31 (75.6%), cyclophosphamide in 22 (53.7%), doxorubicin in 14 (34.2%), and other topoisomerase II inhibitors such as etoposide in 4 (9.8%) patients. Of note, approximately 60% of patients given initial melphalan repeated the treatment or received another protocol containing melphalan when a disease relapse was clinically evident. All the cases were diagnosed as secondary neoplasms of myeloid lineage, including MDS in 34 (82.9%) cases, AML in 4 (9.8%) cases, MPN in 2 (4.9%) cases (1 case of polycythemia vera [PV] and 1 case of primary myelofibrosis]), and chronic myelomonocytic leukemia, type 1 (CMML-1) in the remaining 1 (2.4%) case (Table 1). When evaluating for secondary MN, 21 (51.2%) patients presented with bicytopenia or pancytopenia, 19 (46.3%) patients with borderline cytopenia or unicytopenia, and the remaining 1 (2.4%) patient presented with relatively normal CBCs but showed absolute monocytosis on differential of WBCs (case 32 in Table 2). The age at the diagnosis of secondary MN ranged from 37.6 to 87 years, with a median of 65 (mean ± SD, 64.2 ± 10.0) years. The median interval between diagnosis of PCM and secondary MN was 66 months, with a range of 12 to 384 months, and the median interval between treatment for PCM and diagnosis of secondary MN was 60 months, with a range of 9 to 384 months.

View this table:
Table 1

Morphologic Evaluation

Bone marrow biopsy was performed on all patients for either persistent cytopenia or restaging of PCM when the diagnosis of secondary MN was made (Table 2). Of 40 cases evaluated for morphologic dysplasia, 23 (57.5%) showed mild/moderate to severe myelodysplasia, 12 (30%) showed minimal to mild morphologic deviations, and the remaining 5 (12.5%) cases did not display changes suggestive of myelodysplasia. Of 34 MDS cases, 27 (79.4%) demonstrated mild/moderate to severe myelodysplasia and/or increased blasts, while 7 (20.6%) cases showed no significant morphologic dysplasia or increase in blasts. Twenty-three (57.5%) of 40 cases demonstrated hypercellularity on bone marrow core biopsies, whereas normocellularity or hypocellularity was noted in the remaining 17 (42.5%) cases. Significantly increased blasts were identified in 16 (39.0%) cases including 4 cases of AML with blasts greater than 20% blasts in bone marrows. Based on morphologic evaluation, the 34 MDS cases were further categorized as refractory cytopenia with unilineage dysplasia (RCUD) in 2 (5.8%) cases; refractory anemia with ring sideroblasts (RARS) in 1 (2.9%) case; refractory cytopenia with multilineage dysplasia (RCMD) in 15 (44.1%) cases; refractory anemia with excess blasts, type 1 (RAEB-1) in 5 (14.7%) cases; refractory anemia with excess of blasts, type 2 (RAEB-2) in 6 (17.7%) cases; and MDS, unclassifiable in the remaining 5 (14.7%) cases. Residual neoplastic plasma cells were seen in 28 (68.3%) cases, ranging from 2% to more than 90%, of which 23 cases (56.1% of 41) had 5% or more abnormal plasma cells in the bone marrows. Representative bone marrow aspirate smears of RCMD Image 1A, RAEB-2 Image 1B, and AML Image 1C are shown, in which coexisting residual/persistent plasma cell neoplasm is demonstrated in each case.

Cytogenetic and Molecular Studies

At the diagnosis of PCM or related diseases, only 11 cases had reported cytogenetic findings. Of these, 7 showed normal karyotype, and the other 4 cases demonstrated clonal cytogenetic abnormalities consistent with PCM, including rearrangement of chromosome 14 (IGH locus) in 2 and deletion of 13q in the other 2 cases. In 37 cases, chromosomal analysis was performed at the diagnosis of secondary MN. These included clonal cytogenetic abnormalities consistent with MN in 33 (89.2%) cases, normal karyotypes in 3 (8.1%) cases (cases 17, 19, and 32), and complex cytogenetic abnormalities with balanced translocation consistent with PCM in the remaining 1 (2.7%) case (case 7) (Table 2). Of note, case 32 received allo-HSCT from a sex-mismatched donor and showed normal karyotype of donor cell origin when a bone marrow biopsy performed 160 months after treatment for PCM showed morphologic evidence of CMML-1. In the other 3 cases with normal karyotypes (including case 7), the diagnosis of MN was based on moderate myelodysplasia (RCMD in case 19), increased blasts (AML in case 17), and abnormal megakaryocytic hyperplasia with a positive JAK2 V617F mutation analysis (PV in case 7), respectively. Of 33 cases with clonal cytogenetic abnormalities, complex cytogenetic abnormalities were identified in 22 (66.7%) cases, whole or partial loss of chromosome 5 and/or 7 was seen in 22 (66.7%) cases (including 18 cases with complex abnormalities), balanced translocations excluding those occurring in complex abnormalities were detected in 2 (6.1%) cases, and other chromosomal aberrations were noted in 6 (18.2%) cases. When stratified by treatment regimens given to the patients during the disease course, complex cytogenetic abnormalities constituted 82.6% (19/23) of the cases treated with combination chemotherapy, including melphalan and/or cyclophosphamide with or without auto-HSCT and/or allo-HSCT, 22.2% (2/9) of the cases treated with melphalan with or without immunomodulators, and 33.3% (1/3) of the cases treated with immunomodulators or without myeloma-related therapy Table 3. Two cases with VAD protocol containing neither melphalan nor cyclophosphamide (cases 7 and 18) showed no complex abnormalities in myeloid clone. The odds ratio (OR) was 16.6 for the group with melphalan- and/or cyclophosphamide-based combination chemotherapy, with a chance of acquiring complex cytogenetic abnormalities in reference to the group with melphalan alone (P < .01). When the group with combination chemotherapy was further stratified, the combination of melphalan and cyclophosphamide demonstrated 91.7% (11/12) complex abnormalities, higher than did melphalan/other cytotoxic regimens (71.4%) and cyclophosphamide/other cytotoxic regimens (75%). The ORs were 38.5 for the group with melphalan and cyclophosphamide, 8.8 for that with melphalan/cytotoxic regimens other than cyclophosphamide, and 10.5 for that with cyclophosphamide/cytotoxic regimens other than melphalan, with a chance of acquiring complex cytogenetic abnormalities in reference to the group with melphalan alone. Similar proportions were noted in each of these groups when a loss of chromosome 5 or 7 was included, but their differences from the reference were smaller than those of complex abnormalities alone. Of 32 MDS cases with cytogenetic analysis, 31 (96.9%) had clonal cytogenetic abnormalities detected with chromosomal analysis, including 20 (62.5% of 32) cases with complex abnormalities. Among these, 7 (21.9%) cases (cases 9, 13, 14, 16, 18, 23, and 26) showed no morphologic evidence of MN, such as significant myelodysplasia or increase in blasts, but the neoplastic myeloid clones in all these cases were identified by incidental finding of myeloid-related karyotypic abnormalities, with 2 cases (cases 13 and 14) demonstrating complex abnormalities. Four cases (cases 1, 5, 20, and 31) in our series did not undergo conventional cytogenetic analysis at the diagnosis of MN. In these cases, the diagnosis was made based on increased blasts in 2 cases (cases 1 and 31; RAEB-2 and AML, respectively), prominent myelodysplasia and increased ringed sideroblasts in 1 case (case 5), and abnormal megakaryocytic hyperplasia with severe myelofibrosis in the remaining 1 case (case 20).

View this table:
Table 2
Image 1

Myeloid neoplasm secondary to plasma cell myeloma. A, Myelodysplasia and coexisting residual plasma cell neoplasm in a case diagnosed as refractory cytopenia with multilineage dysplasia (case 4 in tables). Note prominent dysplastic changes (1 erythroid precursor, 1 nucleated RBC with karyorrhexis, and 2 atypical myelomonocytic precursors) (arrows) and 3 atypical plasma cells (arrowheads) (Wright-Giemsa, ×1,000). B, Increased blasts (arrows) in a case diagnosed as refractory anemia with excess blasts, type 2 (case 27). Note marked increase in atypical plasma cells in addition (Wright-Giemsa, ×600). C, Marked increase in blasts in a case diagnosed as acute myeloid leukemia (case 12). A plasma cell (arrowhead) is present among blast population (arrows) (Wright-Giemsa, ×1,000).

View this table:
Table 3

FISH analysis was performed at the diagnosis of secondary MN in 10 cases either to confirm the myeloid clones or to detect a low level of plasma cell neoplasms. Of these, 5 cases demonstrated cytogenetic abnormalities consistent with myeloid neoplasms (data not shown), 2 (cases 4 and 17) cases had the same cytogenetic abnormalities seen in the initial diagnosis of PCM, 1 case (case 13) was positive for both myeloid and PCM-related changes, and the remaining 2 cases were negative for specific myeloid or plasma cell abnormalities (data not shown). Although chromosomal analysis showed isolated myeloid clone with trisomy 8 in case 4, interphase FISH analysis detected 13q deletion in 6% of interphase nuclei consistent with residual plasma cell neoplasm.

Treatment for Secondary Myeloid Neoplasms and Clinical Outcome

The follow-up period after diagnosis of secondary MN ranged from 0 to 99 months, with a median of 7 months in our series. Chemotherapy directed at MN was applied to all cases of AML and the majority of high-risk MDS. Low- to intermediate-risk MDS commonly received supportive care until disease progression, when myeloid-related chemotherapy was typically started. The patient with PV was initially treated with phlebotomy before clinical and laboratory findings showed evidence of multiorgan failure. In 38 cases with at least 1 month of follow-up, 22 patients (57.9%) died of disease progression (13 cases) or complications of the disease or treatment (9 cases), with a median follow-up of 8 months. Kaplan-Meier survival analysis showed estimated median overall survival of 19 months, with 95% confidence interval (CI) of 8 to 70 (mean ± standard error [SE], 36.5 ± 7.8) in our analyzed patient population Figure 1A. The death rates closely correlated with types of secondary MNs, including 3 (100%) of 3 AML cases with a median survival of 2 months, 2 (100%) of 2 MPN cases with a median survival of 10 months, 4 (66.7%) of 6 morphologically high-risk MDS cases (RAEB-2) with a median survival of 4 months, 13 (65%) of 20 intermediate-risk MDS cases (RAEB-1 and RCMD) with a median survival of 14 months, and 0 (0%) of 7 low-risk MDS cases (RCUD and RARS) (P < .05 among 3 risk groups of MDS cases). Of 34 cases with MDS, 7 were categorized as low risk, 12 as intermediate risk 1, 12 as intermediate risk 2, and 3 as high risk, according to the International Prognostic Scoring System (IPSS). Excluding 1 case without follow-up, the death rate was 66.7% (2/3) in the IPSS high-risk group with median overall survival of 4 months, 58.3% (14/24) in the intermediate-risk group with median overall survival of 14 months, and 16.7% (1/6) in the low-risk group with median overall survival of 70 months (P < .01) Figure 1B. When stratified by cytogenetic results, the patients with complex abnormalities or −5/−7 showed median overall survival of 9 months (mean ± SE, 22.3 ± 6.6), with 16 (64%) of 25 patients dead of disease progression or complication compared with a significant longer survival (mean ± SE = 32.6 ± 4.2 months; median, incalculable) in those with other abnormalities or normal karyotype (P < .01) Figure 1C. When stratified by levels of coexisting residual/persistent plasma cell neoplasms, the group with fewer than 5% plasma cells seemed to show a tendency toward a better overall survival (median = 70 months, CI = 2–99) than the group with 5% or more plasma cells (median = 14 months, CI = 6–38), although the difference between these 2 groups was not statistically significant Figure 1D.

Figure 1

Kaplan-Meier survival analysis of the cases with myeloid neoplasm (MN) secondary to plasma cell myeloma or related diseases. A, Overall survival analysis in 38 cases with follow-up period of more than or equal to 1 month after the diagnosis. B, Survival analysis stratified by risk assessment using the International Prognostic Scoring System. P < .05 in a comparison of the intermediate-risk 1 (INT-1) group with the low-risk group or the intermediate-risk 2 (INT-2) or high-risk groups. C, Survival analysis stratified by cytogenetic groups: the group with complex abnormalities, −5 or −7 vs the one with other abnormalities or normal karyotype. P < .05 in a comparison of the 2 groups. D, Survival analysis stratified by coexisting residual/persistent plasma cell myeloma: the group with fewer than 5% plasma cells vs the group with 5% or more plasma cells.


It is well documented that the chance of secondary hematopoietic neoplasms after other primary malignancies is increased compared with the risk in the general population.3,1316 This increased risk is either related to chemotherapy/radiotherapy15,16 or attributed to the intrinsic predisposition to secondary malignancies in patients with primary neoplasms.3,13,14 Leukemia, particularly MDS or other MNs, is the most frequent secondary neoplasm following a diagnosis of primary malignancy.3,15,16 PCM is a neoplasm derived from terminally differentiated B cells, which accounts for 21,700 cases and approximately 10,710 expected deaths per year in the United States, as seen in a 2012 survey.17 MN was reported to be associated with PCM,3,18 occurring mostly after but occasionally before or along with PCM. The incidence of MN associated with PCM has increased in recent years, particularly with increased survival of patients because of improved myeloma-related therapy.3,4 However, whether secondary MN is mainly caused by PCM therapy or by increased susceptibility in patients with myeloma remains controversial despite many case reports and case series. In their recent study, Mailankody et al3 demonstrated a risk of MDS/AML in patients with monoclonal gammopathy of undetermined significance that was comparable with the risk in patients with PCM. In the latter population, the risk was the same before or after the introduction of high-dose chemotherapy with melphalan followed by auto-HSCT. Based on these results, the authors were able to suggest a significant role of nontreatment-related factors in the development of MDS/AML in patients with PCM or related diseases; however, the degree of interaction with myeloma-related therapy could not be determined in their epidemiologic study.

In our case series, we demonstrated secondary neoplasia of myeloid lineage in all 41 patients with PCM. Of these, more than 80% of the cases were MDS, and the remaining cases were AML, MPN, or MPN/MDS. This proportion is much higher than that of de novo MDS among entire myeloid malignancies in general.11 Of the cases with conventional cytogenetic analysis in our series, about 90% demonstrated clonal karyotypic abnormalities, the majority of which were complex cytogenetic abnormalities or deletion of chromosome 5 and/or 7 (78.8%). The rate of karyotypic abnormalities and the types of abnormalities are different from those in de novo MDS/AML but are similar to therapy-related MNs after treatment for primary malignancies other than PCM.10,15 The single case (case 8) without PCM therapy in our series developed RCMD after 21 months of PCM diagnosis with interstitial deletion of the long arm of chromosome 5, raising a possibility of natural predisposition to MN in patients with PCM. Two other cases with immunomodulatory therapy (case 16 and 41) showed complex chromosomal abnormalities or interstitial deletion of chromosome 16, respectively, when a diagnosis of MDS was made. Although these 3 cases with immunomodulators or no pertinent treatment had a lower rate of complex abnormalities than combination chemotherapy containing melphalan and/or cyclophosphamide, the difference was not statistically significant. This finding may be explained by a decreased detection sensitivity because of a low number of cases in the former group (OR = 9.5 for combination therapy; P = .09). In a separate study (unpublished data, 2012), we described 16 cases with concomitant PCM and MN in which therapy was excluded as the cause of MN. Of 16 cases, MDS comprised 50% (8/16) of cases, clonal cytogenetic abnormalities were detected in 43.8% (7/16) of cases, and complex cytogenetic abnormalities were seen in 18.8% (3/16) of cases. No cases demonstrated loss of chromosome 5 or 7. All these fractional numbers are similar to those seen in de novo MN but significantly different from the aforementioned proportions of MN secondary to PCM in the present case series (∼83% MDS [P < .05], 89.2% clonal cytogenetic abnormalities [P < .001], 59.5% complex abnormalities [P < .01], and 70.3% complex abnormality/−5 or −7 [P < .005]). These data suggest that therapy for PCM likely plays an important role in the development of MN after PCM or related diseases.

If therapy contributes to the development of MN in patients with PCM or related diseases or it accelerates a latent leukemia, which PCM therapy regimens have a major impact on neoplastic transformation? Because of multiple medications used intermittently during repeated relapses of PCM or related diseases and heterogeneous treatment modalities, it is difficult to sort out which specific regimen(s) is/are responsible for inducing the myeloid transformation. In our case series, the median latency from the treatment for PCM to the diagnosis of MN was 60 months, which is similar to that reported in therapy-related MDS/AML associated with alkylating agents.10,15 In addition, 2 different patterns of cytogenetic abnormalities have been documented in association with therapeutic regimens. Unbalanced chromosomal aberrations, primarily whole or partial loss of chromosomes 5 and/or 7 that is often associated with other abnormalities in complex changes, constitute approximately 70% of therapy-related MNs. These cytogenetic abnormalities correlate with a long latency period, an MDS phase preceding AML, and the use of alkylating agents and/or radiotherapy. The other pattern of cytogenetic changes comprises the remaining 20% to 30% of therapy-related MNs and involves balanced rearrangements of chromosomes. The latter pattern is associated with a short latency, overt AML without a preceding MDS phase, and chemotherapy with topoisomerase II inhibitors.10 The cytogenetic changes in our case series include complex abnormalities in 59.5% (22/37) of the cases, other unbalanced aberrations in 24.3% (9/37) of the cases, balanced chromosomal rearrangements in 5.4% (2/37) of the cases, and normal karyotype in 10.8% (4/37) of the cases. These compositions of cytogenetic abnormalities seem to fall into the category associated with alkylating agents in reports of therapy-related MDS/AML.10,15 For more than 4 decades, the use of melphalan was considered the main cause of the elevated risk for MN in patients with PCM.1820 Although a few studies demonstrated an increased risk for MDS in patients with long-term melphalan treatment before the era of auto-HSCT,21 other studies suggested that high-dose melphalan as a preparative regimen for auto-HSCT was likely a key factor in the neoplastic process.19 In the present case series, more than 90% of the patients received chemotherapy, in which melphalan was the major alkylating agent used. Based on our analysis, complex cytogenetic abnormalities seemed to be associated with melphalan therapy in combination with other chemotherapeutic agents, particularly with cyclophosphamide, whereas melphalan without other cytotoxic agents showed no significant difference in the pattern of cytogenetic abnormalities compared with the cases with concomitant PCM and MN. The results suggested that exposure to multiple cytotoxic agents, particularly a simultaneous or metachronous use of multiple alkylators, might have a major impact on neoplastic transformation in patients treated for myeloma. A skewed proportion of MDS in our case series also suggested the role of alkylating agents in the process and may reflect early detection of the disease because of frequent bone marrow biopsies in this patient population. Interestingly, 15 patients in our cases received doxorubicin and 2 patients were exposed to more than 1 topoisomerase II inhibitor as combination chemotherapy, but no balanced translocations were detected among 14 cases with chromosomal analysis performed at the diagnosis of secondary MN. All these 14 cases showed unbalanced cytogenetic changes, with 12 cases having either complex abnormalities or deletion of chromosomes 5 and/or 7. No balanced translocation involving 11q23 (MLL gene) was identified in any of our 37 cases in which cytogenetic analyses were performed. The aberration involving 11q23 in case 10 appeared to be a tandem duplication of the locus as part of unbalanced complex changes rather than a balanced rearrangement associated with therapy-related AML caused by topoisomerase II inhibitor. Therefore, the absence of balanced chromosomal aberrations suggests that doxorubicin and other topoisomerase inhibitors may not play a major role in leukemogenesis in our patient population. However, whether the use of doxorubicin or other related drugs enhances the effect of alkylating agents and whether a lack of balanced translocation in our series is the result of a relatively short duration of doxorubicin treatment remain to be further investigated. Recent studies have suggested that maintenance therapy with thalidomide or lenalidomide in patients with myeloma might increase the risk for MDS/AML.22,23 Nonetheless, only 1 case (case 39) in our series received lenalidomide without exposure to other cytotoxic agents, making it impossible to assess its role in leukemogenesis.

The incidence of MNs secondary to PCM varies in different studies. Mailankody et al3 demonstrated an 11-fold increased risk of developing AML/MDS in patients with PCM or related diseases, with an annual incidence of 139.0 per 100,000 after the primary diagnosis. In a separate study with more than 3,000 patients with PCM who received high-dose chemotherapy, MDS defined by clonal cytogenetic abnormalities was reported to be 3%.24 Other studies demonstrated different rates, ranging from 1% to 12.2% of patients with myeloma receiving treatment.23 In our case series, 17 cases were from a pool of 887 patients with myeloma/plasmacytoma treated at Duke University Medical Center, giving a prevalence rate of 1.9% in this particular patient population. This lower rate compared with the majority reported in the literature could be explained by a relatively short period of disease follow-up in our patients related to the search criteria, because the risk for secondary MN correlates with the length of disease course. In a series of 648 patients with PCM who received treatment, the prevalence of MDS was reported to be 3% in a 5-year period and 10% in an 8-year period.21 The incidence appeared to depend on the types of treatment, with the highest rate reported in the population with auto-HSCT, which was estimated at 18% 5 years after transplantation.19 However, a lead bias cannot be excluded because most patients with PCM receive other treatment before starting auto-HSCT. Nonetheless, the observations in our case series and the literature could possibly underestimate the prevalence/incidence of this disease complication in patients with myeloma. Based on our experience, peripheral blood cytopenia in patients with PCM is often attributed to either residual/recurrent PCM or bone marrow toxicity caused by multiple medications when significant myelodysplasia and/or increased blasts are absent on bone marrow examination. About half of the 41 cases in our series did not show significant morphologic dysplasia, and only 39% of cases demonstrated an increase in blasts. In 9 cases (22%; 7 MDSs and 2 MPNs), neither significant dysplasia nor increased blasts were identified on bone marrow examination, and factors other than myelodysplasia were considered as the cause of cytopenia in patients before conventional cytogenetic analysis or molecular testing detected clonal abnormalities suggestive of myeloid neoplasm in each case. In addition, mild to moderate dysplasia or mild increase in blasts may be masked or detracted by coexisting plasma cell neoplasm. More than 90% of our cases demonstrated residual PCM, and more than 50% of the cases had more than 5% abnormal plasma cells in the bone marrow when biopsies were performed for peripheral anemia, thrombocytopenia, or pancytopenia. Therefore, we believe that, in a significant number of PCM cases with treatment and sustained cytopenia, the nature of cytopenia would be clinically uncertain without significant morphologic deviations or clonal cytogenetic abnormalities; the diagnosis of MDS may not be rendered until the cases transform with detectable chromosomal abnormalities or evolve to higher-grade MDS or AML.

The treatment-related risk for MN in patients with PCM is difficult to assess precisely because of possible interaction with intrinsic susceptibility to the disease. According to our analysis of cytogenetic results, complex abnormalities occurred in 82.6% of cases given combination chemotherapy containing melphalan and/or cyclophosphamide, in only 22.2% of cases with melphalan alone, and in 33.3% of cases with immunomodulators or no pertinent treatment. The risk for the combination therapy group is estimated to be 16.6 times higher than the group receiving melphalan without other cytotoxic agents (P < .005; Table 3) and 9.5 times higher than the group receiving immunomodulators/no treatment, although the difference in the latter is not statistically significant. In addition, the OR for developing complex cytogenetic abnormalities is 6.4 and complex cytogenetic abnormalities or loss of chromosomes 5 and/or 7 is 10.2 in our patients, compared with the patients with concomitant PCM and MN in whom the treatment factor was excluded. These data could be potentially improved with an increased number of patients, particularly in the group with concomitant diseases or the group with immunomodulators/no treatment.

The prognosis of t-MDS/AML in general is poor, with an overall 5-year survival rate of less than 10%.10,15,25,26 The clinical outcome is strongly associated with karyotypic abnormalities at the diagnosis of t-MDS/AML, with abnormalities involving chromosomes 5 and/or 7 and a complex karyotype carrying a poor prognosis and a median survival of less than 1 year.10,15 Similar prognostic features have been reported in MNs in patients treated for PCM or related diseases, with median overall survival ranging from less than 1 year to about 2 years in the majority of the studies.59,1820 In our case series, the median overall survival after diagnosis of MN was 19 months, which is within the range of reported cases with PCM-related treatment. The estimated survival in our patient population seems to be better than that reported in t-MDS/AML in general.10,15 Previous studies reported that neither blast count nor subtype demonstrated a significant impact on clinical outcome in t-MDS, in contrast to de novo disease, and thus a category of combined t-MDS/AML was recommended.27 Based on our analysis, the clinical outcome of MN secondary to PCM seemed to correlate well with the types of leukemia, with overall survival longest in low-risk MDS (RCUD, RARS) and shortest in high-risk MDS (RAEB-2)/AML. It is unclear whether the better overall survival and subtype dependence of clinical outcome are the result of the inclusion of nontreatment factors in the pathogenesis of our cases. An analysis of previously reported MNs with well-documented myeloma therapy and follow-up information revealed a median overall survival of 6 months, which is much shorter than seen in our series; however, in those series AML comprised nearly 50% of the reported cases59 in contrast to 9.8% in our case series. This raises another possibility that the better survival in our series could reflect an early detection because of the stringent diagnostic workup for PCM or related diseases. In addition, our survival analysis might be significantly limited by an increased censor rate because of the short follow-up in a significant number of our cases. Of note, about 50% of our patients who were alive or lost to follow-up experienced disease progression or sought supportive care in hospice when follow-up was abruptly terminated; these cases had to be censored in the analysis. In addition, comorbidity of underlying PCM might be overlooked in the reported cases. Based on our analysis, the survival could be influenced by coexisting PCM, although the difference between the groups with stratified levels of residual/persistent PCM was not statistically significant in our case series.

In conclusion, complex abnormalities and whole or partial loss of chromosomes 5 and/or 7 constituted 70% of total cases and approximately 80% of cytogenetic abnormalities in our case series. The aberrations correlated directly with therapeutic regimens, particularly with combination chemotherapy containing melphalan and/or cyclophosphamide. Moreover, the features of cytogenetic abnormalities in MN secondary to PCM were significantly different from those of the cases with concomitant PCM/MN, in which a much lower rate of complex abnormalities and/or unbalanced aberrations on chromosomes 5 and/or 7 and higher rate of normal karyotype were observed. The bias toward complex cytogenetic abnormalities and unbalanced aberrations on chromosomes 5 and/or 7, as well as the long latency from treatment to development of MN and skewed proportion of MDS, suggest a therapeutic mutagenic impact of alkylating agents on the pathogenesis of secondary MN in our patient population. The median survival in our patients was 19 months, which is slightly better than that reported in t-MDS/AML cases in general. In contrast to previous studies, the survival of patients with MDS in our series appeared to depend on subtypes of the disease as seen in de novo cases. Whether an intrinsic predisposition to MN in patients with myeloma plays a role in determining this unique clinical outcome remains to be further investigated.


We thank Steven Conlon in the Department of Pathology at Duke University Medical Center for the technical assistance of photo images and figure presentations. We thank Iris A. Katz in the Office of Tumor Registry at Duke University Medical Center for providing the pertinent information regarding myeloma registration at Duke University Hospital.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
View Abstract