Indolent mantle cell lymphoma

Mantle cell lymphoma (MCL) accounts for approximately 6% of all non-Hodgkin’s lymphomas¹ and is characterized by the over-expression of cyclin D1 with cell cycle dysregulation, secondary to the genetic abnormality t(11;14)(q13;q32).² It often has an aggressive clinical course and is generally associated with a poor prognosis. Although there are a number of treatment options for these patients and initial response rates are improving, the remissions achieved are generally not durable and, with the possible exception of allogeneic transplantation, no treatment can be considered curative. The multiply-relapsing pattern of behavior in this disease contributes to a widely quoted median survival of 3–5 years.¹ Given its poor prognosis, MCL is usually treated at diagnosis with aggressive chemotherapy but with a median age at presentation in the mid 60s, most patients are not eligible for intensive therapies.

In the last few years there has been recognition that a subset of patients with MCL have a significantly longer survival (often more than 7–10 years) and a more indolent disease course. Two groups have reported on separate cohorts of patients who did not receive up-front chemotherapy at the time of diagnosis but were instead managed with a ‘watch and wait’ approach.^3,4 Both groups found that this approach did not have adverse effects on survival outcomes, suggesting that if such patients can be reliably identified, chemotherapy for this group, with its attendant morbidity, could reasonably be deferred. Although diagnostic criteria for the identification of these patients are not currently available, there is evolving recognition of clinicopathological differences identifying this group from the group of patients with classical MCL.

It is becoming clear that many, although by no means all, patients with indolent disease present with a leukemic picture rather than nodal disease. Fernandez et al.⁵ found that while both indolent and classical forms of MCL share a common gene expression profile, which differs from that of other leukemic variants of lymphoid neoplasms, there are significant differences between the two. In particular 13 genes were found to be differentially expressed: all were under-expressed in patients with indolent disease and over-expressed in those with classical disease. SOX11 is one of these and potentially has an important role in both the pathogenesis of MCL and in the identification of indolent disease.

SOX11 is a single exon gene and codes for one of the three group C Sry-related HMG box proteins (along with SOX4 and SOX12). These proteins are thought to play a significant role in the control of cell differentiation during many developmental processes, key among these being neuronal differentiation.⁶ Their developmental necessity is evident from single gene knockout studies showing that Sox4-null and Sox11-null mice cannot survive because of heart outflow tract malformations (they die at embryonic day 14 and just after birth, respectively).^7,8 SOX4 is also required for B lymphocyte differentiation. The role of SOX11 in hematologic development (both normal and malignant) is unclear, although there is emerging evidence for SOX11-responsive genes⁹ and a potential tumor suppressor role of SOX11 in MCL.¹⁰

Cytoplasmic SOX11 may be found to a variable extent in B-cell neoplasms and certain normal adult tissues.¹¹ Strong nuclear expression of SOX11 (detected by immunohistochemistry) is confined to MCL, lymphoblastic lymphomas (B and T), and some Burkitt’s lymphomas with weaker nuclear expression present in hairy cell leukemia.¹²

As well as being of diagnostic value, the distribution of SOX11 staining in MCL may be prognostic, since Wang et al. showed that non-nuclear staining is associated with impaired overall survival.¹³ This potentially contradicts the findings from the study by Fernandez et al. in which non-nuclear staining was found in patients with indolent MCL, who had a better overall survival.⁵

The work by Fernandez et al.⁵ also showed that 70% of cases of indolent MCL have highly mutated (>5%) IGVH genes. Although this mutation rate is much higher than that seen in classical MCL (~20%) the significance of this remains controversial, with one group reporting no clinical difference¹⁴ but another group reporting an overall survival advantage.¹⁵ Orchard et al. identified a higher incidence of mutated IGVH genes in leukemic rather than nodal MCL, alongside a more indolent course in some of these leukemic patients.¹⁶

Gene expression analysis applied to MCL has provided further prognostic information by way of a proliferation gene expression signature.¹⁷ By assigning a value to each of 20 proliferation genes that were differentially expressed among MCL cases a proliferation signature average was calculated and found to correlate inversely with survival. The cohort with low proliferation gene scores had a similar survival to that of the patients with indolent MCL identified in other studies. A more pragmatic way of assessing proliferation is through Ki-67 staining. This provides prognostic information and has been validated within the MCL-specific International Prognostic Index (MIPI) score.¹⁸

In a study published in this issue of the journal, Ondrejka et al. investigated a cohort of patients with leukemic MCL exhibiting an indolent disease course.¹⁹ They retrospectively analyzed archived samples and data from eight patients who presented with leukemic MCL over a 10-year period. All these cases were found incidentally in asymptomatic patients with lymphocyte counts varying between 4.5 and 14.2×10⁹/L and all were t(11:14)-positive. These patients had no progression or very slowly rising lymphocyte counts and only two required any form of treatment as of the time of analysis. All the available bone marrow biopsies from the time of diagnosis showed low level MCL involvement detectable by immunohistochemistry. When examined by flow cytometry a significant proportion had a subset of CD23 positive cells (usually negative in MCL) and the majority of the patients in the study showed kappa light chain restriction, as opposed to the more typical lambda light chain restriction usually found in MCL.

The finding of kappa light chain restriction in the majority of these patients is interesting and may lead to further insights into the development of different subsets of MCL, due to the differences in the timing of kappa and lambda light chain rearrangement during B-cell development.

Given the previous reports regarding the potential prognostic value of SOX11 expression, immunolabeling of SOX11 was performed. The results concur with other studies examining leukemic MCL with an absence of staining in the majority of patients (one patient had some very minor nuclear staining), compared to the strongly positive staining observed in conventional nodal MCL (P=0.00001).

With this study, Ondrejka et al. provide further evidence for both the potential prognostic use of SOX11 and the existence of a subset of MCL patients with very indolent disease. However, by excluding patients with lymphadenopathy, splenomegaly and gastrointestinal symptoms, their study omitted a potentially large group of patients with indolent MCL. From their own data, only 3% of patients with MCL have a leukemic form, while the unselected cohorts studied by Martin et al.³ and Eve et al.⁴ indicated that up to 30% of all MCL may exhibit an indolent behavior.

It is possible that while leukemic MCL was traditionally associated with a poorer prognosis,²⁰ the better overall survival of the patients seen in the more recent studies may be due to changes in entry criteria. As postulated by Ondrejka et al., some of the cases of leukemic MCL may in fact be considered a form of MCL-type monoclonal B-cell lymphocytosis rather than MCL, and that many more of these patients might be detected if all patients with monoclonal B-cell lymphocytosis are screened for t(11:14)/IGH@-CCND1. This may explain the heterogeneity seen within patients in the non-nodal group in the study by Orchard et al. who found increased variability in karyotype complexity and IGVH mutation status in leukemic compared to nodal MCL.¹⁶

Now that a reliable SOX11 antibody is available this may become incorporated into routine MCL diagnosis especially as SOX11 can identify the rare cyclin D1-negative patients.²¹ The differences reported regarding the prognostic value of nuclear SOX11 staining could be reconciled if nodal indolent MCL is considered as an entity separate from, but related to leukemic indolent MCL. The report by Wang et al. showing a shorter overall survival in patients with non-nuclear SOX11 staining¹³ did not describe the population of patients, so it may be that SOX11-negative nodal MCL behaves differently from SOX11-negative leukemic disease.

While there is no doubt that indolent MCL exists, and it is well recognized within patients presenting with non-nodal, leukemic disease, it is likely that this is not the only clinical scenario. SOX11 and other genes are likely to become useful in the identification of these patients at diagnosis and this will ultimately provide clinicians with the confidence to explore less intensive treatment approaches.