Risk adapted approach: How to treat splenic marginal zone lymphoma in resource-poor settings? - The real-life experience of a Brazilian cancer treatment center

Splenic marginal zone lymphoma (SMZL) is a malignant neoplasm with primary involvement of the spleen, characterized by the proliferation of small B-cell lymphocytes, predominantly in the spleen white pulp, with nodular histological arrangement and indolent behavior [1, 2]. It represents 2% of all non-Hodgkin's lymphomas (NHL), but is the most common histological subtype of splenic primary lymphoma [2]. Usually, patients have massive splenomegaly, peripheral blood cytopenias, and absence of peripheral lymphadenopathy, except for splenic hilum [3]. There is no extranodal involvement, except for the liver and bone marrow. About 50%-60% of patients have bone marrow infiltration and up to 25% may have circulating phase, represented by mild or moderate lymphocytosis. Presence of B symptoms and high levels of lactic dehydrogenase (LDH) are uncommon [4]. Presence of paraproteinemia, usually IgM kappa and autoimmune phenomena occur in up to one third of patients [5].

SMZL arises from marginal zone memory B-cells and probably has a post-germinal origin [6]. Chronic antigenic stimulation of B-cell receptor initially induces polyclonal and later monoclonal B lymphocyte expansion, leading to malignant transformation [6]. In up to 36% of cases, this stimulus is represented by chronic hepatitis C virus infection [7]. Molecularly, SMZL is characterized by dysregulation of nuclear factor kappa-B and mutations affecting the NOTCH, KLF2 (Kruppel like factor 2) and PTPRD (receptor-type protein tyrosine phosphatase delta) pathways [8].

Considered as an indolent and incurable disease, its treatment is only recommended in symptomatic cases, represented by massive visceromegaly, severe cytopenias, presence of constitutional symptoms or autoimmune phenomena unresponsive to steroids [9]. Therapeutic options for SMZL include splenectomy, monotherapy with anti-CD20 monoclonal antibody or combination therapy, including rituximab and classic chemotherapeutic agents [10, 11]. In the last ten years, several studies have shown superiority of rituximab therapy, either alone or in combination [12‐14], but, in resource-poor countries rituximab is not a drug of universal access, making splenectomy still widely used in these particular settings.

The majority of SMZL patients have a disease with favorable prognosis, with a median overall survival (OS) exceeding 10 years, despite the use of specific treatment [15, 16]. However, SMZL prognosis is heterogeneous, and about 20%-30% of the cases show a more aggressive clinical course with a median OS of only 4 years. Histological transformation to high grade B-cell lymphoma may occur in 10%-15% of cases as part of the natural history of this tumor [15, 17]. Thus, the identification of patients with unfavorable outcome becomes necessary for better risk stratification and selection of appropriate therapy.

Clinical and laboratory factors with prognostic impact to guide therapy have been described by different American and European groups using different prognostic scores [15, 18]. Arcaini et al. created a score capable of predicting SMZL prognosis, using a combination of hemoglobin < 120 g/L, elevated LDH and albumin < 3.5 g/dL [18]. A study conducted by the Splenic Marginal Zone Lymphoma Working Group (SMZL-WG), involving 593 patients from different centers, identified Hb < 95 g/L, platelets < 80 x 10⁹/L, elevated LDH and the presence of extra-hilar lymphadenopathy as variables associated with adverse prognosis, representing another widely used prognostic score in clinical practice [15].

In Latin America, particularly in Brazil, there is scarce information regarding SMZL epidemiological data, as well as evolution and prognostic factors that influence survival of these population. The Brazilian national public health system offers limited options to treat these patients. For the most part, anti-CD20 antibodies such as rituximab are not available in public services for SMZL patients. In this context, often the first therapeutic option is splenectomy. Although still a good option for patients with low surgical risk, in elderly patients or in patients with severe comorbidities, splenectomy may be associated with a high risk of morbidity and mortality. In SMZL patients with more aggressive disease, the new treatment progression-free interval can be quite short when using splenectomy. In these scenarios, rituximab therapy may be applied with greater clinical benefit.

In this study, we aimed to report the clinical-laboratory and epidemiological findings of 39 Brazilian patients with SMZL followed at a single referral center for cancer treatment, as well as to determine factors related to adverse prognosis and to propose a rational risk-adjusted therapeutic strategy, considering scenarios where access to monoclonal antibody therapy is limited.

Splenic marginal zone lymphoma (SMZL) is a rare B-cell lymphoid malignancy, representing 2% of all non-Hodgkin's lymphomas and 80% of all primary spleen neoplasms [2]. The natural history and clinical-biological characteristics of this disorder are known from case series reports from some European and North American centers [15, 16, 18]. Here, we had the opportunity to describe the characteristics of this tumor in a population derived from a high degree of miscegenation, from a different geographical area, and that may have distinct biological findings. To our knowledge, this is the larger cohort of Latin American SMZL patients reported in the medical literature.

In our cohort, we found a predominance of SMZL in females, with a 3:1 distribution ratio [15, 18], contrarily to what has been observed in series of patients reported from developed countries. Although we believe that women are more careful about their own health and seek medical help earlier than men, particularly in unfavorable socio-economic conditions, we cannot ignore these findings. In agreement with our study, Xing et al. reported 107 cases of SMZL treated at the British Columbia Cancer Agency (BCCA), also observing a predominance of this cancer in women, with a distribution ratio of 1.5: 1.0 [16].

B-symptoms and higher rates of involvement of extralymphonodal sites, particularly bone marrow and peripheral blood, were more frequent in our cohort. We believe that these findings may be explained by the increased difficulty in accessing tertiary health services and referral cancer centers in Brazil, as well as in other developing countries. This may justify the larger tumor mass and incidence of B symptoms attributed to disease diagnosed in later stages.

In our series, we found a rate of 10.2% (04/39) for high-grade B-cell NHL histological transformation, similar to that observed in the BCCA (15/107 - 14%) and Sylvester Comprehensive Cancer Center of Miami cohorts (34/453 - 7.5%) [16, 17]. In the Brazilian cohort all cases of transformation were documented by biopsy. Due to our limited sample (N = 39) we could not access predictors of transformation, but we noted the unfavorable outcome associated with this phenomenon, as all patients with transformed disease died in our cohort. Alderuccio et al., analyzing 453 patients with marginal zone lymphoma (MALT, nodal and splenic), identified failure to achieve CR after initial treatment, elevation of lactic dehydrogenase and more than 4 nodal sites involved at the time of diagnosis as predictors of increased risk of MZL transformation [17]. Overall, prognosis after transformation is quite narrow, with 5-year OS of 0% in our Brazilian patients, 33% in BCCA and 43% in the North-American group [16, 17].

The etiological association between SMZL and chronic infection for hepatitis C virus is commonly related in Southern Europe, and the development of lymphoma under these conditions appears to be associated with activation of CD81 in B-cells by chronic stimulation of the virus E2 glycoprotein [21, 22]. According to some reports, up to 30% of SMZL patients may have chronic hepatitis C [7, 21, 22]. These cases may initially be treated with HCV-guided therapy, such as interferon alpha and ribavirin or a combination of new antiviral agents, and many cases may have remission of lymphoma with such therapies. In our series, 5.1% (2/39) of SMZL patients had positive serology for hepatitis C, but surprisingly 10.2% (4/39) had chronic hepatitis B infection. Recently, some studies have shown coexistence between chronic hepatitis B virus infection and splenic lymphomas, particularly SMZL, suggesting a new etiological association for this neoplasm. Similar to that observed in HCV positive cases, treatment for HBV with antiviral agents is capable of inducing neoplasm remission [23‐26].

SMZL is usually an indolent and chronic disease, with a median overall survival exceeding 10 years [2, 15, 16, 18]. In this study, we showed quite prolonged OS and PFS (5-year OS 76.9% and 5-year PFS 63.7%), estimated over several years, similar to what is reported in the literature. However, 20%-30% of SMZL patients are characterized by greater biological aggressiveness, with shortened OS (median 4 years) and greater propensity for histopathological transformation to high grade B-cell lymphoma [15]. Therefore, identifying this subgroup of ominous outcome is essential to choose the best treatment options, particularly in resource-poor settings where the best therapy is not always universally accessible.

To this end, different international collaborative groups analyzed their own series of SMZL patients, seeking to identify factors associated with unfavorable prognosis, as well as to elaborate prognostic scores for outcome prediction and implementation of risk-adapted therapies. In 2006, Arcaini et al. analyzed 309 SMZL patients and identified 3 laboratory variables that directly impacted OS and PFS, with Hb < 120 g/L, LDH > UNV and albumin < 3.5 g/dl, and through a simple scoring system created a mortality predictor prognostic score [18].

Following the same principles, in 2012 a multicenter study involving 593 SMZL patients from different international cancer centers, known as the Splenic Marginal Zone Lymphoma Working Group (SMZL-WG), determined the unfavorable predictor value of Hb < 95 g/L, platelets < 80 x 10⁹/L, LDH > UNV and presence of extra hilum-splenic lymphadenopathy for symptomatic progression-free survival at 5 years. The SMZL-WG score was thus created, stratifying patients at low-risk, intermediate-risk and high-risk, with 5-year PFS of 95%, 87% and 68%, respectively [15].

Although this is the largest Latin-America single-center cohort of SMZL patients published to date, our study had only 39 cases, which imposes some restrictions on the interpretation of our findings. However, we were able to statistically determine variables associated with worse OS and PFS at 5 years, and we could validate the use of Arcaini and SMZL-Working Group scores in our population. Prognostic factors in the Brazilian series involved Hb < 100 g/L, platelets < 100 x 10⁹/L, LDH > UNV and albumin < 3.5 g/dL. These laboratory prognostic factors were similar to those described in the medical literature, but we could not create a prognostic score using such variables due to the limited N of our study.

The therapeutic approach of SMZL has shown substantial progress in the last 15 years. As with other indolent lymphomas, asymptomatic patients do not benefit from early treatment and should only be observed with a “watchful & waiting” approach [9, 19]. However, treatment should be introduced in the presence of B symptoms, massive or symptomatic splenomegaly, cytopenias secondary to hypersplenism or bone marrow lymphomatous infiltration, or unresponsive-corticosteroid immune manifestations [9, 19].

Until 2005, splenectomy was considered the first-line therapy for SMZL (“gold-standard”). Although it is an effective therapy, with an overall response rate (ORR) of almost 90% and PFS of more than 50% in 5 years, the morbidity and mortality associated with the surgical procedure are not negligible, affecting about 5% of patients [27‐29]. After splenectomy, splenomegaly-related symptoms resolve and cytopenias improve. Furthermore, histological analysis of the spleen will allow the diagnosis of SMZL to be confirmed. However, it should be emphasized that this therapeutic strategy is not curative and that most patients will achieve only partial response, with persistent bone marrow and peripheral blood lymphomatous infiltration, even if a significant reduction in lymphocytosis was observed [28].

Among chemotherapy options, different studies have shown that alkylating agents alone, such as chlorambucil, cyclophosphamide or in combination (CVP, CHOP) are associated with low response rates and high toxicity, considering the indolent behavior of the disease. On the other hand, purine analogs, particularly fludarabine, are associated with higher overall response rates (100%) and complete response (70%) [30‐32].

After 2005, incorporation of the anti-CD20 monoclonal antibody (rituximab) changed the therapeutic paradigm of SMZL patients. The weekly regimen of rituximab monotherapy at a dose of 375 mg/m² for 4 to 6 consecutive weeks is associated with minimal toxicity and OS and PFS ranging from 89%-100% and 73%-98%, respectively [12, 13, 33]. In addition, rituximab is also effective for autoimmune disorders such as AIHA and immune thrombocytopenia, found in up to 15-20% cases of SMZL.

Although large prospective and randomized studies comparing rituximab monotherapy versus splenectomy are still lacking, the Greek group, retrospectively, compared rituximab monotherapy versus splenectomy in a historical cohort and found 5-year OS and PFS of 94% vs 77% (p = 0.09) and 72% vs 58% (p = 0.09) for rituximab [13]. In 2018, the same group published an update of their data (n = 108), confirming the high efficacy of rituximab monotherapy, its role as a splenectomy-sparing therapeutic strategy and the potential benefit of maintenance treatment with immunotherapy. In this study, responses to rituximab were long lasting, with 10-year PFS exceeding 60%, with minimal toxicity, and maintenance therapy improved response quality and was associated with better PFS [14].

Another strategy associated with a high success rate in the treatment of SMZL is a chemo-immunotherapy approach with bendamustine/rituximab (BR) regimen. According to recent studies, this association provides high overall and complete response rates (91% and 73%, respectively), with manageable toxicity and unquestionable benefit, particularly in SMZL patients with intermediate and high-risk scores [34, 35]. Faced with the unquestionable benefits of rituximab monodrug or BR combination chemoimmunotherapy, some authors propose that splenectomy may be abandoned in the up-front management of SMZL [36]. We believe that splenectomy still plays an important therapeutic role in the management of SMZL, particularly in patients not responsive to immunotherapy or chemotherapy and as up-front modality in those with confined spleen disease, in which this is a diagnostic and therapeutic procedure that occurs in about 15%-20% of SMZL patients.

Our cohort comprised SMZL patients treated at our service for about 25 years, therefore, the therapeutic strategies used were quite heterogeneous, including chemotherapy (single or combined), splenectomy and rituximab immunotherapy. In our series, none of these therapeutic approaches showed statistically significant superiority in terms of OS or PFS, but it should be noted that in a context of poor-resource settings, such as in the Brazilian public health system, rituximab has been reserved for cases with known worse prognosis, such as elderly patients or those with high surgical risk and adverse prognostic factors. In addition, our study has restrictions regarding the small casuistic and short segment time of patients receiving immunotherapy (incorporated in our institution for reserved cases of SMZL starting in 2008).

As demonstrated by the data in Table 2, some baseline characteristics differ in statistical significance form (p < 0.05) between the three primary treatment groups. In general, these differences are strongly expressed between splenectomized (N=21) and non-splenectomized (N=17, rituximab or low-intensity cytotoxic chemotherapy) patients.

Patients undergoing splenectomy had lower levels of tumor markers (B2-microglobulin and LDH), thus inferring a smaller tumor mass than patients treated with rituximab or cytotoxic chemotherapy. In addition, surgically treated patients had better general clinical conditions than the non-splenectomized group, reflected by higher levels of serum albumin, lower rate of clinical comorbidities and better performance status according to the Eastern Cooperative Oncology Group (ECOG) scale. Finally, the group undergoing splenectomy had a lower proportion of patients classified as intermediate or high-risk by the prognostic index of the SMZL-Working Group compared to patients treated with medications.

Thus, we noticed that SMZL patients who received immunotherapy had worse general clinical conditions, and laboratory characteristics were consistently associated with a poor prognosis, which led us to infer that this is one of the reasons why we did not find a concrete benefit in terms of OS and PFS of the rituximab group versus splenectomy group in our cohort.

In conclusion, based on the risk factors identified in our real-life study and according to the latest evidence available in the literature for SMZL therapy, we propose that in resource-poor settings, where access to rituximab therapy is not universal, as is the case in the Brazilian public health system, a risk-adapted therapeutic approach should be employed in the management of SMZL patients (Table 4). Thus, rituximab may be reserved for first-line therapy in elderly patients, in those with high operative risk or who have one or more of the following factors: Hb < 100 g/L, < 100 x 10⁹/L platelets, LDH > UNV, albumin < 3.5 g/dL or high-risk Arcaini and SMZL-WG scores. Patients who do not present these factors can be satisfactorily approached with splenectomy, when therapy is indicated.