a model for scientific and clinical progress

Multiple myeloma: a model for scientific and clinical progress
Jesus San Miguel1
1Centro de Investigacio´ n Me´ dica Aplicada, Clinica Universidad de Navarra, Pamplona, Spain
Multiple myeloma (MM) is a unique cancer paradigm for investigating the mechanisms involved in the transition from a
premalignant condition (monoclonal gammopathy of undetermined significance) into a malignant disease (MM). In the
pathogenesis of myeloma, the dialogue between plasma cells and their microenvironment is as important as the
genotypic characteristics of the tumor clone. MM is genetically highly complex, with almost all patients displaying
cytogenetic abnormalities and frequent intraclonal heterogeneity that play a critical role in the outcome of the disease.
In fact, it is likely that myeloma will soon no longer be considered as a single entity. This, along with the availability of an
unexpected number of new treatment possibilities, has reinforced the need for better tools for prognosis and for
monitoring treatment efficacy through minimal residual disease techniques. The outcome of MM patients has
significantly improved in the last 2 decades, first through the introduction of high-dose therapy followed by autologous
stem cell transplantation and, more recently, due to the use of proteasome inhibitors (bortezomib and carfilzomib) and
immunomodulatory agents (thalidomide, lenalidomide, and pomalidomide). Moreover, the need to reexamine the
diagnostic criteria of early MM and the possibility of early intervention opens up new therapeutic avenues. New drugs
are also emerging, including second- and third-generation proteasome inhibitors and immunomodulators, monoclonal
antibodies, histone deacetylase inhibitors, and kinesin spindle protein inhibitors, among others. Our goal is to find a
balance among efficacy, toxicity, and cost, with the ultimate aim of achieving a cure for this disease.
Learning Objectives
● To understand that myeloma should no longer be considered
as a single entity
● To understand that better tools for diagnosis and monitoring
treatment efficacy are being implemented
● To understand that the treatment goal is to find the best
possible balance among efficacy, toxicity, and cost
Introduction
Multiple myeloma (MM) is the second most common hematological
malignancy, with an annual incidence of 4 new cases per 100 000
people. It accounts for 1% of all malignant diseases and 15% of all
hematological malignancies. In the pathogenesis of MM, the
mechanisms responsible for the interaction between malignant
plasma cells (PCs) and their microenvironment are as important as
the genetic changes involved in the development of the malignant
clone because these play an important role in bone destruction;
tumor cell growth, survival, and migration; and drug resistance.
Genomic characteristics of myeloma cells
Genome instability is a prominent feature of myeloma cells and, in
fact, almost all patients with MM are cytogenetically abnormal.1
Genomic abnormalities include chromosomal translocations, mainly
involving the IGH locus on chromosome 14q32, copy number
abnormalities, mutations, methylation modifications, and gene and
miRNA dysregulation.2 Unlike other B-cell tumors, MM exhibits a
marked diversity of chromosomal loci involved in IGH transloca￾tions. Approximately 40% of MM tumors have IGH translocations
involving 5 recurrent chromosomal patterns: 11q13 (CCND1), 4p16
(FGFR/MMSET), 16q23 (MAF), 6p21 (CCND3), and 20q11
(MAFB), corresponding to an incidence of 15%-20%, 15%,
5%–10%, and 3% for the latter 2 patterns, respectively.3 Although
IGH translocations induce up-regulation of different oncogenes, it is
possible that all IGH translocations involved in MM converge on a
common pathway that is essential in the pathogenesis of the disease
and cause the inhibition of differentiation and an increase in cell
survival and proliferation. Gene expression profiling (GEP) analysis
has demonstrated that expression of the cyclin proteins (CCND1,
CCND2, and CCND3) is increased in almost all MM patients,
supporting the hypothesis that there is a potential unifying event in
its pathogenesis.4 In addition to these structural changes, numerical
chromosomal abnormalities are frequently observed in MM; in fact,
almost all MM cases are aneuploid. The nonhyperdiploid patients
are characterized by a very high prevalence of IGH translocations,
monosomy/deletion 13, and gains on 1q. In contrast, the hyperdip￾loid group is associated with recurrent trisomies involving odd
chromosomes (3, 5, 7, 9, 11, 15, and 19) and with a low incidence of
structural chromosomal abnormalities.2 Lesions on chromosome 1
are the most common abnormalities in MM; these are mostly 1q
gains that result from tandem and jumping segmental duplications
of the chromosome 1q band, as well as 1p losses. Deletion of
chromosome 13 is present in 40%–50% of MM patients and is
strongly associated with t(4;14) and t(14;16), deletion of 17p, and
gains on 1q. The chromosome 17p deletion, which includes loss of
TP53, occurs at a lower frequency in newly diagnosed MM
(5%–10%), although the proportion is higher in advanced stages of
the disease. Furthermore, 17p deletion is associated with extramed￾ullary MM.
Some genetic changes in MM, such as secondary translocations,
mutations, deletions, and epigenetic abnormalities, are considered
to be late oncogenic events and are associated with disease
progression. Most karyotypic abnormalities involving MYC corre￾spond to complex translocations and insertions that are often
nonreciprocal and frequently involve 3 different chromosomes.
Activating RAS mutations are considered to be molecular markers

of disease progression.5 Therefore, the prevalence of activating
KRAS and NRAS mutations is 70% in MM cases at relapse. TP53
inactivation via deletion or mutation also seems to be more
frequently associated with disease progression.
New insights into MM genetics
GEP analysis has confirmed the huge genetic diversity of MM
cases, and several genomic classification models have been pro-
posed by the Arkansas, French, and Dutch groups. The most widely
accepted is the Arkansas TC model, which connects genetic
abnormalities, cell transcriptome, and clinical features of patients
and classifies MM patients into 7 different groups. Each group
displays a specific genetic signature, some of which are associated
with a particular IGH translocation or ploidy status and with a
characteristic clinical behavior.6 However, so far, the reproducibil-
ity of these GEP models has not been optimal and they have not
been implemented in the clinical milieu except in selected centers.6
In a recent large study including 203 MM patients, whole-genome
sequencing strategies have shown that 65% had evidence of
mutations in 1 or more of the 11 recurrently mutated genes: K and
N-RAS, BRAF, FAM46C, TP53, DIS3, TRAF3, CYLD, RB1,
PRDM1, and ACTG1. Interestingly, mutations were often present
in subclonal populations and multiple mutations within the same
pathway (eg, RAS and BRAF) were observed in the same patient.1
This pattern is consistent with other hematological malignancies
such as acute myeloid leukemia, but is in contrast to hairy cell
leukemia and Waldenstrom’s macroglobulinemia, which feature the
single unifying mutations BRAF and MYD88, respectively.
Although the role of epigenetics in cancer has been demonstrated,
there is limited evidence of its role in the pathogenesis of MM.
Silencing of certain tumor-suppressor genes (GPX3, RBP1, SPARC,
CDKN2A, SOCS, and TGFBR2), overexpression of the histone
methyltransferase MMSET, and the presence of mutations of UTX
(histone demethylase) have been described. Furthermore, genome-
wide methylation studies have shown both global DNA hypometh-
ylation and gene-specific DNA hypermethylation in MM, with
certain epigenetic signatures being associated with prognostic
cytogenetic groups.7 Another area of emerging interest in cancer
pathogenesis concerns miRNAs, small, noncoding RNAs that
regulate gene expression at the posttranscriptional level and are
involved in critical biological processes including cellular growth
and differentiation. Various studies have shown that miRNA
expression is deregulated in myeloma cells compared with normal
plasma cells and that their GEP profile is associated with genetic
abnormalities.8 Moreover, several miRNAs are known to be in-
volved in MM pathogenesis. Indeed, a mechanism has been
identified by which miRNAs act on MDM2 expression to regulate
p53; therefore, miR-192, miR-194, and miR-215 reexpression in
myeloma cell lines induces degradation of MDM2, with subsequent
up-regulation of p53 and inhibition of cell growth.9
Multistep pathogenesis and drug resistance
MM is a unique cancer paradigm for investigating the mechanisms
involved in the emergence of a premalignant condition and its
transition to a malignant disease; in other words, from an “early/
benign phase” known as monoclonal gammopathy of undetermined
significance (MGUS), to an “intermediate/indolent phase” [smolder-
ing MM (SMM)] and a final “advanced stage” (symptomatic and
ultimately resistant/refractory MM). Unfortunately, the key ques-
tions in this process are yet to be answered: why does a quiescent
clone become aggressive in some patients but remains stable in
others? and what are the mechanisms responsible for primary and
acquired chemoresistance? Is this dictated only by the genotypic
characteristics of the tumor clone or is the dialogue between
myeloma PCs and their microenvironment also significant in this
process? Until recently, the pathogenic models assumed that MM
develops through a multistep transformation from normal PCs to
MGUS (implying PC immortalization) and subsequent transforma-
tion into active MM, in which clonal PCs are responsible for
end-organ damage. However, studies based on FISH, single-
nucleotide polymorphism arrays, and whole-genome sequencing
have demonstrated that most genetic lesions typically observed in
MM are already present in MGUS patients and that the progression
from MGUS to SMM, and eventually to MM, would involve a
clonal expansion of genetically abnormal PCs, implying a complex
evolutionary process with intraclonal heterogeneity.10 Three distinct
patterns of genomic evolution have been proposed based on data
generated by new genomic approaches: (1) stable genomes, without
differences between diagnosis and relapse clones; (2) linear evolu-
tion, in which the relapse clone apparently derives from the major
subclone at diagnosis, but continues to diversify through addition-
ally acquired lesions; (3) and branching (nonlinear) models, in
which the relapse clone clearly derives from a minor subclone that is
barely present at diagnosis.2,11 Patients with high-risk cytogenetics
usually follow the last 2 evolutionary models. These findings are
also relevant for the treatment of MM, because the presence of
intraclonal heterogeneity with clonal tides supposes a significant
obstacle for targeted therapy. For example, even though patients
harboring the BRAF mutation might respond to BRAF inhibitors,
this effect would be suboptimal if the mutation were not present in a
major PC subclone; in fact, BRAF-negative clones could even
become stimulated. Therefore, mutations are often present in
subclonal populations1,11 and drug combinations targeting coexist-
ing subclones will probably be a more efficient approach.