Plasma cell leukaemia and other aggressive plasma cell malignancies

Summary

Extramedullary plasma cell cancers, such as plasma cell leukemia (PCL) and multiple extramedullary plasmacytomas (MEP) are very aggressive malignancies. These can be primary (de-novo) or secondary due to progressive prior multiple myeloma (MM). Recent reports suggest an increase in incidence of these disorders. Compared to MM, organ invasion is common in PCL, while soft tissue tumors involving the head, neck or paraspinal area are common sites for MEP. Markers of poor prognosis are frequently observed in these extramedullary forms of plasma cell cancers, and survival is significantly inferior compared to patients with MM. Conventional chemotherapeutic and radiotherapy approaches have been employed with variable results. Even high dose chemotherapy with autologous stem cell rescue has not been able to demonstrate consistent improvement in survival outcome. Although not specifically evaluated, novel anti-plasma cell agents, such as the proteasome inhibitor bortezomib, and immunomodulatory drugs, such as lenalidomide, appear to be active against these aggressive cancers. Clinical and translational research directed at improved understanding of disease biology and development of novel therapeutics is urgently needed.

Introduction

Multiple myeloma (MM) is the most common plasma cell neoplasm (Jemal et al. 2008). Recent advances in myeloma treatment have resulted in significant improvement in patient outcomes. At present, although, MM remains incurable, the median survival of patients has increased from 29.9 months, before 1996, to 44.8 months in last decade (Kumar et al. 2008). With increasingly successful treatment of MM that has resulted in prolonging patient survival, newer challenges, such as difficult-to-treat extramedullary relapse and secondary plasma cell leukemia are an emerging problem. This paper reviews the available literature on two particularly aggressive forms of extramedullary plasma cell cancers; plasma cell leukemia (PCL) and the multiple extramedullary plasmacytomas (MEP) both of which are noted to have a rising incidence.

Plasma cell leukemia

Traditionally, PCL has been diagnosed based on Kyle’s criteria, which requires circulating plasma cells to account for at least 20% of peripheral blood leukocytes and/or an absolute circulating plasma cell count of 2.0 × 10⁹/l, with evidence of monoclonal gammopathy (Jimenez-Zepeda and Dominguez 2006). Some investigators consider this number to be arbitrary and have relied more on the clinical behavior of the disease rather than the absolute number of circulating malignant plasma cells (Avet-Loiseau et al. 2001). Furthermore, it is important to recognize that severe peripheral blood plasmacytosis can be seen in non-malignant conditions, such as severe sepsis (especially staphylococcal), pertusis, dengue fever, infectious mononucleosis, rubella and parvovirus B19 infection; however, in these disorders the oligo or polyclonal plasmacytosis is transient and rapidly resolves with improvement of the antecedent condition. (Gawoski and Ooi 2003; Shtalrid et al. 2003; Bai et al. 2006). Clinically, PCL can be subdivided into primary and secondary types. Primary PCL (pPCL) presents de novo in the leukemic phase without a prior history of a plasma cell cancer, while secondary PCL (sPCL) arises in the context of a pre-existing MM.

Epidemiology

PCL is a rare disease and its accurate incidence and prevalence are not known. Published literature has only reports of case series. Larger series have estimated PCL to account for 0.5 to 3 % of all plasma cell disorders. Han et al (2008) reported on the epidemiological and survival characteristics of 254,702 cases of lymphoid neoplasm diagnosed during 1973 and 2003 at 17 Surveillance, Epidemiology and End Result (SEER) registries. They found 221 cases of PCL among 42,065 cases of plasma cell neoplasm. The median age of PCL patients at diagnosis was 64 years. Non-Hispanic white males and females comprised 38% and 32.6% of cases, respectively. Interestingly, African American males and females accounted for only 7.2 % and 12.2 % of cases, respectively. The African American patients (both males and females) tended to be 10 years younger than the non-Hispanic whites at diagnosis (Han et al. 2008). In this series patients with PCL had the worst prognosis of all lymphoid malignancies. Earlier studies have identified pPCL to be commoner of the two subtypes; however, recent literature suggests that, with the increasing incidence of sPCL, each type now accounts for about 50% of cases (Noel and Kyle 1987; Tiedemann et al. 2008). Consistent with recent observations in the literature (Tiedemann et al. 2008), we have also noted an increase in the incidence of patients with sPCL at our institute over the past few years (unpublished observation). Interestingly, as noted below in the section of multiple extramedullary plasmacytomas (MEP), a similar trend in increasing incidence of secondary MEP has also been noted.

Etiology and pathogenesis

Like other plasma cell malignancies the causation remains elusive. pPCL has been associated with prior exposure to chemotherapy and/or radiotherapy; however, this association remains difficult to confirm due to low incidence of the disease (Buskard et al. 1977; Candoni et al. 2004). sPCL evolves from pre-existing MM.

Recent attempts at genetic and molecular profiling of PCL have provided important clues to its biology. Using conventional metaphase karyotyping, cytogenetic abnormalities have been identified in over 70% of PCL patients (Fonseca et al. 2004). Hypodiploidy and complex karyotypes with multiple numerical and structural abnormalities involving chromosome 1, 13, and 14 have been identified in a significant number of PCL cases (Colovic et al. 2008) (Avet-Loiseau et al. 2001) (Garcia-Sanz et al. 1999). Interestingly, similar distribution of genetic abnormalities have also been demonstrated in non-Caucasian populations (Xu et al. 2009) (Peijing et al. 2009).

In an important study, using fluorescence in situ hybridization (FISH) probes for del (13q), del (17.1p)-p53, t(11;14), t(4;14), del (1p21) and 1q21 amplifications, Chang et al (2009) compared the incidence of specific genetic abnormalities in 41 PCL (15 with pPCL and 26 with sPCL) cases and 220 MM patients. Interestingly, they found similar genetic abnormalities in pPCL and sPCL. Furthermore, compared to MM, del (17p), del (13q), del (1p21), t(4;14) and 1q21 amplifications (known predictors of poor outcome in MM) were more frequent in PCL (Chang et al. 2009). Rearrangements involving MYC are a relatively uncommon and late genetic event in MM; and have been shown to strongly correlate with disease progression (Skopelitou et al. 1993; Pope et al. 1997). Tiedemann et al (2008) reported MYC rearrangement, by FISH, in 33% of pPCL and sPCL patients; this was associated with a trend toward inferior survival in pPCL. Finally, it is tempting to note that, although both pPCL and sPCL share many genetic abnormalities, the fact that these genetic events are present at the time of initial presentation in pPCL as opposed to their evolution over time in sPCL clearly points to these two forms as biologically distinct entities. Thus, although the underlying mechanism(s) that govern(s) transition of preferentially bone marrow residing malignant plasma cell to the leukemic phase remains unknown, it is clear that the genetic abnormalities known to confer aggressive behavior, drug resistance and early relapse in MM are a common finding in PCL.

Clinical Features

Although both pPCL and sPCL share many clinical features, important differences exist between the two. Patients with pPCL are younger; often have extra osseous organ involvement, with increased frequency of renal failure, fast declining performance status and rapid progression to the terminal stage. The liver and spleen are commonly involved (Colovic et al. 2008). Other extramedullary sites of disease involvement include lymph nodes, skin, central nervous system (CNS), testis, heart and pleura (Colovic et al. 2008) (Turhal et al. 1998) (Iseki et al. 1987; Klug, et al. 1992). Clinical and biological characteristics from six large series of pPCL patients are detailed in Tables I and II. In contrast to pPCL, patients with sPCL have advanced bone disease and often develop some of the above mentioned extramedullary manifestations later in the disease course rather than at the time of initial diagnosis of preceding MM. Nodal or splenic involvement seems be to a rare event in sPCL.

Table I.

Clinical characteristics and outcome of patients with primary plasma cell leukemia, as reported in larger series.

	Garcia-Sanz et al (1999)	Dimopoulos et al (1994)	Colovic et al (2008)	Noel & Kyle (1987)	Tiedemann et al (2008)	Peijing et al (2009)
No. of patients	26	27	30	25	41	22
Median age (years)	65	57	60	53	54.5	49.5
Males (%)	46	NA	73	60	59	64
M-protein (%)
Ig G	54	52	53	8	28	54.5
Ig A	4	15	23	12	13	9.1
Ig D	8		3	4	2
Light chain only	31	26	20	28	41	27.3
Lytic bone disease (%)	48	NA	60	44	35	44.4
Extramedullary sites (%)
Liver	0	32	56	52	32	44.4
Spleen	0	18	53	44	18	33.3
Lymphadenopathy	11	6	3	12	6	NA
β2 microglobulin ≥ 6 mg/l (%)	65	91	64	NA	50	50
High LDH (%)	48	63	37	NA	50	NA
Creatinine ≥ 176.8 μmol/l (%)	44	37	43	50	50	NA

Abbreviations: Ig; immunoglobulin, LDH; lactate dehydrogenase; NA; not available.

Table II.

Biological characteristics of primary plasma cell leukemia as reported in large series.

	Garcia-Sanz et al (1999)X	Dimopoulos et al (1994)†	Colovic et al (2008)*	Noel & Kyle (1987)	Tiedemann et al (2008)¥
Immunopheno type (%)
CD 20+	50	NA	11	NA	NA
CD56−	55		77
Cytogenetics (%)
Hypodiploidy		41	28	NA	60
Complex karyotype		92	47
Ch-13 abnormal	85	33	14		85
Ch-1 abnormal	57	50	28

Abbreviations: Ch; chromosome, NA; not available.

^XImmunophenotype information was available in 21 cases. Cytogenetics (FISH) results are reported for 13 patients.

^†Conventional cytogenetics results from 12 patients.

^*Immunophenotype available on 9 patients. Conventional cytogenetics analyses from 21 patients.

^¥Chromosome 13 abnormalities detected by FISH.

Laboratory Evaluation

Diagnosis of PCL requires a comprehensive history, physical examination, laboratory and radiological investigations. Most laboratory tests are similar to those required for the diagnosis of other plasma cell cancers, such as MM. Imaging studies include skeletal survey, bone mineral density estimation by Dual Energy X-ray Absorptiometry (DEXA) scan as well as computed tomography (CT) scans of chest, abdomen and/or pelvis as clinically indicated.

In pPCL, similar to MM, IgG is the most common monoclonal protein. Increased frequency of Bence-Jones proteinuria as an isolated paraproteinemia has also been reported (Garcia-Sanz et al. 1999). Despite the aggressive nature of PCL there does not seem to be any significant difference in the serum M-protein levels compared to patients with MM. Conversely, PCL patients have lower levels of serum albumin and higher levels of lactate dehydrogenase (LDH), β-2 microglobulin (≥ 6 mg/l in more than 65% of patients), serum calcium, and serum creatinine, all signifying aggressive disease biology. Most patients with PCL are observed to have an advanced clinical stage (Garcia-Sanz et al. 1999; Colovic et al. 2008). Additional differentiating features of pPCL include more extensive bone marrow infiltration, often with plasmablastic cytomorphology (Figure 1). Rare findings, such as erythrophagocytsis (haemophagocytic syndrome) (Butterworth et al. 1953), and hyperammonaemia (Kawano et al. 1991; Nakahashi et al. 1994; Minauchi et al. 2004) have also been reported. Patients with sPCL have laboratory abnormalities that are similar to those of pPCL, except that higher proportions of pPCL patients have renal impairment on presentation.

Figure 1. Histological and flow cytometric features of leukaemic plasma cells.

(A) Leukaemic plasma cells in the peripheral blood with prominent nucleoli (blood smear, Wright-Giemsa stain, x1000). (B) Immunophenotyping of malignant plasma cell from bone marrow of a patient with PCL demonstrates presence of CD138 and absence of CD56 expression by multiparameter four-color flow cytometry (PE; phycoerythrin, APC; allophycocyanin). (C and D) Variability in cytomorphological features of malignant plasma cells is noted in patients with PCL, such as small plasma cells or lymphoplasmacytoid cells (C) and larger cells or plasmablasts (D) (blood smear, Wright-Giemsa stain, x1000).

Flow cytometry is an important diagnostic tool for the evaluation of peripheral blood to confirm circulating clonal plasma cells (Figure 1). Additionally, immunophenotyping has also been used to characterize and compare the malignant plasma cells from patients with monoclonal gammopathy of undetermined significance (MGUS), MM, and PCL. Akin to the malignant plasma cells from MM or MGUS patients, malignant PCL cells also express CD38 and CD138; however, expression of CD 20 is more prevalent in PCL (Perez-Andres et al. 2005). Compared to MM, there is reduced expression of CD117, HLA-DR and CD9 antigens on PCL cells (Garcia-Sanz et al. 1999). Moreover, absence of CD56 expression (a finding associated with aggressive clinical behavior in MM patients) is more frequently seen in PCL (Pellat-Deceunynck et al. 1998; Garcia-Sanz et al. 1999; Rawstron et al. 1999; Sahara et al. 2002; Colovic et al. 2008). Intriguingly, Pérez-Andrés et al (2005) also demonstrated significantly decreased expression of the surface molecules HLA-1, β-2M and CD40 on clonal plasma cells from patients with MM and PCL as compared to those from MGUS patients. These findings suggest a progressive failure of tumour-specific immune response as the disease progresses from relatively indolent to aggressive form, thus providing important insights into mechanisms of immune evasion.

Treatment and Prognosis of PCL

PCL remains an incurable and highly resistant disease. The overall prognosis remains extremely poor with significantly worse outcome for patients with secondary disease (Garcia-Sanz et al. 1999; Colovic et al. 2008). Although pPCL may respond to treatment initially, these responses are short lived with dismal median overall survival of 8 months (Garcia-Sanz, et al 1999). On the other hand, sPCL is an exceedingly resistant, rapidly progressive and fatal disease with median overall survival of 2 months (Noel & Kyle 1987, Garcia-Sanz, et al 1999, Tiedeman et al 2008). It is noteworthy that similar survival has been observed in the Chinese series of 31 PCL patients (Peijing et al. 2009). Currently there are no standard therapeutic strategies as none of the treatment options have been prospectively evaluated in formal clinical trial investigations due, primarily, to the logistical limitations conducting clinical trials in a rare disease.

The following section reviews the currently employed therapies for PCL utilizing conventional and/or novel agents.

Conventional chemotherapy agents

With the exception of a few case reports, alkylating agents in conventional doses have not shown significant activity in PCL. In their series, Noel & Kyle (1987) reported median survival of 8.7 months for 15 pPCL patients treated with melphalan (Table III). Similarly, Bernasconi et al (1989) reported on 15 PCL patients treated with various combinations of cyclophosphamide, vincristine, melphalan, prednisone and doxorubicin. The median survival for pPCL and sPCL was 10 and 5 months, respectively. Although, higher response rates have been reported with second generation myeloma regimens, such as VAD (vincristine, adriamycin and dexamethasone), overall outcome of PCL has not changed with these strategies (Dimopoulos et al. 1994). In the largest single centre series of 80 patients, Tiedemann et al (2008) reported median overall survival of 11.2 months in pPCL and 1.3 months in sPCL. Treatment regimens incorporating vincristine, dexamethasone, doxorubicin, carmustine and cyclophosphamide in various combinations had better outcome as compared to melphalan and prednisone (median survival 15.4 months vs. 4.1 months). Patients treated with VAD or VBMCP (vincristine, carmustine, melphalan, cyclophosphamide, prednisone) who were consolidated with high dose therapy (HDT) and stem cell rescue had a median survival of 22 months as compared to those who did not undergo HDT. Importantly, these results should be interpreted with caution as patients who underwent HDT were younger and had better performance status as compared to those who did not undergo HDT (Tiedemann et al. 2008).

Table III.

Response rates and survival of primary plasma cell leukemia patients as reported in larger series.

	ORR (%)X	CR (%)	Median OS (months)
Garcia-Sanz et al (1999)	37	0	8
Dimopoulos et al (1994)	59	NA	2† 20*
Colovic et al (2008)	46	23¥	4.5
Noel & Kyle (1987)	NA	NA	6.8
Tiedemann et al (2008)	NA	NA	11.2
Peijing et al (2009)	45**	0	14

Abbreviations: NA; not available, ORR; over all response rate. CR; complete response rate, OS; overall survival.

^XThere was variability in response criteria used in each study.

^¥CR was defined as more than 50% reduction in circulating plasma cells and M protein in serum and urine.

^†With melphalan-prednisone regimen (n=10).

^*With VAD or CE regimen (n= 17).

^**15 patients treated with multi agent chemotherapy (VAD or VBMCP); 5 patients received MP; 2 patients underwent Allo-transplant

Regimens incorporating hyperfractionated cyclophosphamide, such as Hyper-CVAD (cyclophosphamide, vincristine, adriamycin and dexamethasone), have demonstrated limited success in PCL (Hosono et al. 2008; Murthy et al. 2009).

Alkylating agents in higher doses and in combination with steroids have also been used for treatment of PCL. In a study of 24 patients with pPCL Vela-Ojeda et al (2002) compared efficacy and safety of intermediate dose melphalan (80 mg/m² PO day 1) and dexamethasone with that of VMCPA (vincristine, melphalan conventional dose, cyclophosphamide, prednisone, and adriamycin) and VAD regimens. Two of the eight patients, who received higher dose of melphalan, achieved a complete response and four patients achieved a partial response. Only one patient treated with VMCPA responded while no response was seen in patients receiving VAD (n=12) (Vela-Ojeda et al. 2002). These results suggest that dose intensification of alkylating agent in combination with steroid may be of benefit; however, these findings need to be confirmed in a larger cohort of patients before a definitive conclusion can be derived.

High-dose chemotherapy and stem cell transplant

High dose therapy (HDT) with stem cell rescue has not been prospectively evaluated. In a recent series of 75 Italian patients with pPCL, HDT was associated with significantly better overall and progression-free survival compared to conventional chemotherapy (Pagano et al. 2009). Although reports of patients treated with HDT with long remission periods exist (Yamagata et al. 1994; Yeh et al. 1995), available data suggests that HDT is not as effective in PCL as it is for MM. The lack of effective induction regimen and the rapid decline in patient performance status has precluded allogeneic hematopoietic stem cell transplant as a viable option in most patients. Only 3 out of 17 PCL patients (two with allogeneic and one with autologous transplant) reported to International Bone Marrow Transplant registry survived for more than 3 years (Saccaro et al. 2005).

Immunotherapy

Malignant plasma cells in pPCL frequently express surface CD20 molecule, making it an attractive therapeutic target. However, limited experience with rituximab (a chimeric monoclonal antibody targeting CD20) has yielded inconsistent results and at present the role of CD20 directed therapy in PCL is not fully established (Oka et al. 2006) (Korte et al. 1999) (Gemmel et al. 2002).

Novel agents

Increased understanding of malignant plasma cell biology has resulted in the development of novel agents targeting not only the tumor microenvironment (thalidomide, lenalidomide) but specific intracellular targets, such as the proteasome (bortezomib), as well. These agents have transformed the treatment of MM in recent years. However, there is lack of clear understanding of their activity and role in the treatment of PCL. The following section summarizes current available data on the role of these new agents in PCL.

Thalidomide: is the first novel agent to make its way to MM patients. Since its re-birth (in 1999), thalidomide has been extensively used in the treatment of relapsed/refractory and newly diagnosed MM patients. There are conflicting reports about efficacy of thalidomide in PCL. Petrucci et al (2007) described four PCL patients (2 with primary and 3 with secondary disease) treated with thalidomide. None of the patients responded to treatment, and all patients died within 3 months of starting thalidomide (Petrucci et al. 2007). On the contrary, Johnston & Abdalla (2002) described four PCL patients treated with low dose thalidomide (100 –150 mg/day). Three pPCL patients, all progressing on VAD chemotherapy, responded with more than 80% reduction in serum paraprotein, and complete clearance of circulating plasma cells; two patients continued thalidomide until their last follow up 6 and 14 months after starting therapy. One patient with sPCL did not respond (Johnston & Abdalla 2002). Based on above evidence it appears that thalidomide alone is not sufficient for the treatment of PCL, however its role in combination therapy remains to be explored.

Lenalidomide: is a structural analogue of thalidomide with immunomodulatory properties. Its proposed mechanism of action involves altering the tumor microenvironmental components that foster malignant cell growth (Chanan-Khan and Cheson 2008). Antitumor effects of lenalidomide are mediated not only through disruption of the prosurvival cytokine network but also through an immune attack mediated via activation of T and natural killer (NK) cells. Evidence regarding lenalidomide’s efficacy in PCL is accumulating. Benson & Smith (2007) reported a 63-year-old patient with extensively treated MM evolving into sPCL with 70 % of malignant cells having del (13q) and del (17p). The patient responded rapidly to lenalidomide and low dose dexamethasone, and remained in remission 5 months after starting lenalidomide. It is important to note that the patient had failed multiple lines of therapy including thalidomide, intermediate dose melphalan and high dose cyclophosphamide (Benson & Smith 2007). Musto P et al (2008) reported a patient with PCL who had disease progression after treatment with bortezomib, prednisone and melphalan. The combination of lenalidomide and low dose dexamethasone resulted in rapid shrinkage of extra osseous plasmacytomas and disappearance of circulating plasma cells; however, the disease relapsed after 4 months of therapy and patient died of progressive disease (Musto et al. 2008).

Bortezomib: the first in class inhibitor of ubiquitin-proteosme pathway is a synthetic, boronic acid dipeptide with multiple effects on MM plasma cells, including the ability to overcome tumor resistance to steroids and other conventional agents (Mitsiades et al. 2003). Like other novel agents bortezomib is a relatively new addition to the armamentarium against plasma cell malignancies, therefore literature is emerging about its efficacy in PCL. Based on the available data it appears that bortezomib has consistently shown significant activity against both pPCL and sPCL. In the largest series, Pellegrino Musto et al (2007) reported 12 PCL (8 pPCL; 4 sPCL) patients treated with bortezomib, 3 previously untreated and 9 patients with relapsed/refractory disease after failure of 1–4 regimens, including HDT and autologous stem cell rescue. The overall response rate was 92% and the median overall survival for the whole group was 12 months, while it was not reached at the 21^st month of follow up for the pPCL patients (Pellegrino Musto et al 2007).

In another series, Finnegan et al (2006) reported 4 PCL patients who progressed on VAD chemotherapy; one patient had prior HDT. All four patients responded rapidly to bortezomib-based combinations; however, these responses were temporary and all patients ultimately relapsed and died of disease progression (Finnegan et al. 2006). Ali et al (2007) reported s heavily pretreated patient with sPCL who, after failing multiple salvage regimens, achieved complete remission (CR) with bortezomib, dexamethasone and doxorubicin combination. The patient remained in CR for one year, when the disease relapsed. It is noteworthy that the patient achieved a second CR with retreatment with the same bortezomib-based regimen (Ali et al. 2007). We treated a pPCL patient with a combination of bortezomib, cyclophosphamide, liposomal doxorubicin, thalidomide and dexamethasone as induction regimen for 6 cycles, resulting in partial remission. Maintenance therapy with lenalidomide and dexamethasone was given for an additional 6 months, which resulted in continued disease control. Eventually, the patient developed progressive disease and died of multiple organ involvement at 20 months after the initial diagnosis.

Bortezomib has been shown to overcome the poor prognostic significance of high risk cytogenetic abnormalities in MM (San Miguel et al. 2008). A recent report suggests that the ability of bortezomib to overcome drug resistance, and thus to deliver durable response, may also be seen in PCL patients with adverse cytogenetic features (Katodritou et al. 2008). These reports are encouraging and provide evidence that bortezomib is active against PCL. Furthermore, it appears that bortezomib-based regimens may overcome the adverse prognostic impact of high-risk cytogenetic features.

Multiple Extramedullary Plasmacytomas

Multiple extramedullary plasmacytomas (MEP) are tumor masses consisting of malignant plasma cells outside of the medullary cavity of bone. They can be present at the time of initial diagnosis of MM, as primary MEP, or, more commonly, develop during the course of disease progression in a known case of MM, as secondary MEP.

Epidemiology

There are a few data on accurate incidence of MEP. Two studies have reported the incidence of primary and secondary EMP to be 15% and 20%, respectively (Blade et al. 1994; Blade et al. 1996). In a large study of 1003 consecutive MM patients, Varettoni et al (2009) found MEP to account for 13% of cases (7% had primary and 6% had secondary MEP). Importantly, the authors noted a significant increase in the incidence of MEP overtime. Consistent with these findings, we have also observed an increase in the incidence of secondary MEP at our institution (Manochakian et al. 2006). Intriguingly, there are several recent reports of increasing incidence of extramedullary relapse (EMR) in MM patients treated with novel agents and aggressive treatment strategies, such as HDT and allogenic stem cell transplant (Minnema et al. 2008; Katodritou et al. 2009). The increasing incidence of MEP and sPCL in the era of novel treatments is concerning. The reasons for the changing epidemiology are not entirely clear. It seems plausible that, at least in part, this change is the result of prolongation of the life span of MM patients allowing the natural history of the malignant plasma cell cancers to be more noticeable. On the other hand, use of therapies that interrupt the support of the microenvironmental factors of the bone marrow can select for clones that can escape this dependence and thus result in extramedullary disease.

Aetiology and pathogenesis

Early in the disease course, malignant MM cells are critically dependent upon surrounding bone marrow stromal cells for their growth and survival. As the disease advances this stromal dependence diminishes and MM “learns” to survive without critical microenviornmental factors, such as interleukin (IL)-6, IL-15 and vascular endothelial growth factor among others (Hjorth-Hansen et al. 1999; Tricot 2000; Roodman 2002). The precise molecular events that underlie stromal independence are not completely known. However, it appears that lack/loss of surface CD 56 expression (Sahara et al. 2002) and the phenomenon of light chain escape (Dawson et al. 2007) play an important role in MEP pathogenesis. As mentioned above, recent reports suggest a possible aetiological link between the increasing incidence of MEP and the use of drugs, such as thalidomide, and treatment strategies involving allogeneic stem cell transplant. The proposed mechanism(s) include alteration of adhesion molecules and other tumor-microenvironmental interactions by thalidomide (Balleari et al. 2004; Katodritou et al. 2009), and, in case of allogeneic transplant, tumor escape from the graft-versus-myeloma effect in the bone marrow (Perez-Simon et al. 2006). Additional studies are needed to confirm these hypotheses.

Clinical Features

Clinically, primary and secondary MEP share many overlapping features. As compared to intramedullary MM, patients with MEP tend to be younger, with slight male predominance, light chain disease, and advanced stage with extensive bone disease is a common finding (Varettoni et al. 2009). Para skeletal soft tissue is the most common (80%) site of extramedullary disease. Plasmacytomas of skin, airways, liver, kidneys, and breast have also been reported (Figure 2). CNS involvement, manifesting as headache, confusion, visual problems and cranial nerve palsies, has also been reported (Gozzetti et al. 2009). Compared to primary MEP, patients with secondary disease tend to have lower levels of serum M-protein and hemoglobin and higher levels of LDH (Varettoni et al. 2009). Plasma cells from MEP, especially in secondary form, frequently show immature, plasmablastic morphology. Cytogenetic abnormalities identified in MEP mirror those that are seen in MM (Blade et al. 2009).

Figure 2. Evaluation of extramedullary plasmacytoma.

Panel A: Computed tomography scan of base of skull showing paravertebral soft tissue extramedullary plasmacytoma (white arrow). Panel B: Positron Emission Tomography ssan of the same patient showing extensive medullary and extramedullary involvement.

Treatment and Prognosis

Patients with MEP have aggressive disease characterized by resistance to therapy, early relapse and shorter survival compared to MM without extramedullary spread (Blade et al. 2009; Varettoni et al. 2009). MEP tends to respond poorly to conventional dose chemotherapy regimens. Palliative radiotherapy is effective in controlling symptoms related to plasmacytomas. Although role of thalidomide is controversial, other novel agents, such as lenalidomide and bortezomib, have been shown to be effective in extramedullary disease (Blade et al. 2009; Sher et al. 2009).

Summary and recommendation

PCL and MEP are two distinct varieties of aggressive extramedullary plasma cell cancers characterized by rapid progression, drug resistance and early relapse. There is an apparent increase in the incidence of highly resistant sPCL and MEP. The overall prognosis remains poor (especially so for PCL). In PCL the single most important prognostic factor is the type of the disease itself, with pPCL having some initial responsiveness to therapy and relatively longer survival versus sPCL. Conventional treatment with alkylating agents alone is not effective for PCL, the benefit of HDT and stem cell rescue as well as allogeneic hematopoietic stem cell transplant is difficult to assess as the advanced stage and rapid progression makes most patients ineligible for these treatments. In MEP early relapse and drug resistance are important challenges. There is mounting clinical evidence that novel agents may be more effective in controlling these extramedullary diseases.

Based on our review of literature it appears that bortezomib is one of the most active novel agents against PCL and MEP. Encouraging results have also been seen with newer agents, such as lenalidomide. Thus, in our opinion, combination regimens, especially incorporating these new agents, may be the most optimal strategy for these aggressive cancers. Although formal evaluation of such combinations in PCL is lacking, safety data is available through their investigation in patients with MM. Recognizing the limitation of lack of formal, prospective clinical trials in these disorders, we note that, in the absence of available clinical trials, bortezomib-based combination therapy should be the first line of treatment for PCL patients. For patients intolerant of bortezomib, lenalidomide-based initial treatment can be a reasonable alternative. Importantly, as is clear from above review, none of the available novel agents used alone can provide durable disease control in PCL, therefore we recommend maintenance therapy with an agent non-cross resistant to the drugs used in the front line setting. Furthermore, the increasing incidence of EMR warrants systematic evaluation of disease biology and novel therapeutic strategies using newer agents.