Transformed Lymphoma

Stephane Doucet¹ and Jane N. Winter²

¹Centre Hospitalier de l’Université de Montréal (CHUM Hospital), Montreal, QC, Canada

²Robert H. Lurie Comprehensive Cancer Center of Northwestern University, Chicago, IL, USA

Multiple Choice Question

1. Which of the following does not constitute transformation?

New diagnosis with areas of follicular lymphoma (FL) grade 1–2 and 5% diffuse large B-cell lymphoma (DLBCL)
Long-standing FL, new mass with biopsy showing DLBCL
Long-standing FL grade 1–2 with progression in one of the lymph nodes to FL grade 3a
Long-standing chronic lymphocytic leukemia (CLL) previously untreated with a rapidly increasing cervical mass with a biopsy that shows DLBCL

There is a lack of consensus regarding the definition of histologic transformation (HT). Most experts agree that grade 1–2 follicular lymphoma that progresses to DLBCL or Burkitt lymphoma represents HT (Figure 46.1). The WHO Classification of Tumours of Haematopoietic and Lymphoid Tissues defines HT as “transformation or progression to a high-grade lymphoma usually DLBCL, but occasionally resembling Burkitt lymphoma or with features intermediate between DLBCL and Burkitt lymphoma.” Progression of FL grade 1–2 to FL grade 3 (situation C) is not considered histologic transformation but rather progression. Situation A is perhaps the most controversial. Some authors have considered the presence of both FL and DLBCL in the same lymph node to represent a “composite” lymphoma, and some have referred to it as “transformation at diagnosis.” Others require a defined interval between the diagnosis of FL and the more aggressive histology. The presence of both an indolent and aggressive lymphoma in the same lymph node implies but does not confirm early transformation. The WHO suggests reporting the two disease entities with their respective percentages of involvement. Composite lymphomas were rarely included in the major series of transformed lymphomas (TLs). Rarely, FL may transform to other histologies such as acute B-cell lymphoblastic leukemia or, as recently described elsewhere, to histiocytic or dendritic cell sarcoma.

web_c46-fig-0001 — Figure 46.1 (A) Follicular lymphoma at diagnosis of an untreated patient. (B) At transformation: top area with follicles but diffuse architecture in the lower area composed of sheets of large cells. (C) High-power field of the diffuse area of Figure 1B. Large cells, prominent nucleoli, and irregular shape are noted. (Color plate 46.1)

Transformation to DLBCL has been described for all the major subtypes of indolent lymphomas: small lymphocytic lymphoma and chronic lymphocytic leukemia (SLL–CLL), lymphoplasmacytic lymphoma (LPL), and marginal zone lymphoma. The first description of HT in SLL–CLL was made in 1928 by Richter and constitutes situation D. SLL–CLL can rarely transform to Hodgkin lymphoma and uncommonly to B-cell prolymphocytic leukemia.

A more scientifically robust definition of “transformation” requires demonstration of a clonal relationship between the original indolent lymphoma and the subsequent aggressive counterpart. This may be best demonstrated by molecular techniques characterizing the immunoglobulin gene. In the daily clinical setting, proving that the two lymphomas have the same light-chain restriction is usually sufficient to suggest a clonal relationship. The immunophenotype of the TL may differ from the original indolent lymphoma. Loss of CD10 expression occurs in 10% of cases, and a gain of MUM1 or CD30 occurs, each in 25% of cases. A change in phenotype does not preclude a clonal relationship between the two lymphomas. Transformation was diagnosed on clinical grounds in two series because of the association between aggressive clinical behavior and transformation. This association is not absolute and, whenever possible, the diagnosis should be made pathologically.

Case study 46.1

A 45-year-old male was recently diagnosed with stage IVA FL.

1. What is his chance of histological transformation to DLBCL?

1% per year
2–3% per year
10% per year
12% per year

The rate of transformation has been estimated at approximately 3% per year. Most of the information on the incidence of transformation is derived from large series. The methodology, follow-up period, and definition of transformation vary among series, which explains the slightly different results. The largest series is from the database of the British Columbia Cancer Agency, which included 600 patients with 170 transformations. The annual risk of transformation was a continuous 3%. This series included patients with clinical transformation when biopsy was not possible. Horning also reported a continuous risk of transformation with long-term follow-up at Stanford. Investigators from the University of Iowa and Mayo Clinic recently reported a series of over 600 patients with follicular lymphoma with an overall transformation rate of 10.7% at 5 years, and an estimated rate of 2% per year. In a single institution series with a median follow-up of 15 years, Montoto et al. (2007) reported a risk of transformation of 28% at 10 years and, unlike the British Columbia series, this group observed a plateau after 15 years. Another single-institution retrospective study reported a probability of transformation of 31% at 10 years with a tendency to plateau after 6 years. Cases of transformation can certainly occur many years after diagnosis of indolent lymphoma. Thirty years ago, the US National Cancer Institute reported a prevalence of transformation of 70% at autopsy. We know that rituximab (R) improves the overall survival (OS) of previously untreated patients with FL when combined with chemotherapy, but we will probably not know its real impact on the rate of transformation for another decade or more.

The incidence of transformation is lower in the nonfollicular lymphoproliferative disorders. For example, in patients with SLL–CLL, Tsimberidou et al. (2006) reported the MD Anderson Cancer Center experience from 1975 to 2005 with a cumulative risk of proven transformation to large-cell lymphoma (Richter’s syndrome) of 3.7%, a risk of transformation to Hodgkin lymphoma of 0.4%, and a risk of transformation to prolymphocytic leukemia of 0.1%. There are less data on the annual incidence for transformation among the non-FL lymphoproliferative disorders, but the incidence is certainly less than that observed in FL.

Case study 46.2

A 56-year-old female with a 10-year history of FL grade 1–2, initially treated with rituximab, cyclophosphamide, vincristine, and prednisone (R-CVP) and 2 years ago with R-bendamustine for symptomatic recurrence, now presents with rapidly increasing left cervical adenopathy, B-symptoms, and high lactate dehydrogenase (LDH). You suspect a histological transformation.

1. Do you need a biopsy to prove your suspicion?

The obvious answer to this question is absolutely yes, or at least whenever possible. What if you suspect transformation in a retroperitoneal lymph node that is difficult to access in a patient with risks or contraindications to biopsy? In the large series of transformed FL, biopsies were generally required, but a high clinical suspicion was considered acceptable evidence of transformation in three series. For example, in the British Columbia series, 36% of the 170 patients with transformation were diagnosed based on clinical criteria because biopsy was not feasible. The presence of at least one of the following was considered indicative of transformation: sudden rise of LDH to more than twice the upper limit of normal, rapid discordant localized nodal growth, new involvement of unusual extranodal sites (liver, bone, muscle, or brain), new B-symptoms, or new hypercalcemia. The median OS was not different in the group of patients diagnosed by biopsy compared to those diagnosed based on these clinical or laboratory findings (20 months vs. 16 months; P = 0.2). It should not be assumed, however, that this more aggressive behavior is the equivalent of transformation and these patients need a biopsy to confirm the diagnosis whenever possible without prohibitive risk. Indolent FL, which remains incurable, can recur after successful treatment of TL, stressing the importance of biopsy after each progression.

2. What are the molecular or genetic events that underlie the process of transformation?

C-myc
P53
Bcl-6 mutations or translocation
C-rel
All of the above

The process of transformation is complex and involves genomic, transcriptional, and epigenetic mechanisms. Interaction with the non-neoplastic immune cells in the tumor microenvironment likely also plays an important role. Many alternative pathways have been implicated. For an in-depth overview of this subject, the reviews by Lossos and Gascoyne (2011) and Montoto and Fitzgibbon (2011) are excellent resources. When paired samples obtained from patients both at diagnosis and at the time of transformation have been studied, a variety of mutations, including those involving proto-oncogenes, have been identified, but none consistently. Genes typically expressed in embryonic stem cells (ESCs) have been implicated in TL, and the expression of an ESC signature may predict for future transformation. It has been suggested that TL derives from a common primitive germinal center B-cell population and does not necessarily represent clonal evolution. This is currently an area of active research.

Rearrangements of MYC have been linked to transformation in a proportion of patients. When arising in patients with FL with t(14;18), these cases are called “double-hit lymphomas.” Morphologically, they may be diagnosed as B-cell lymphoma unclassifiable with features intermediate between Burkitt lymphoma and DLBCL (67% of the cases), or as DLBCL (31%). Most cases arise de novo and cannot strictly be considered transformations, but approximately 35% of cases of double-hit lymphomas arise from a previous FL. This subgroup of TL has a particularly bad prognosis with a median OS of 6 months.

Mutation of the tumor suppressor gene TP53 (on chromosome 17p) has been implicated in the progression of FL to TL in approximately 20–30% of cases. A small subset of patients also has amplification of the proto-oncogene c-REL (10%) and deletion of the tumor suppressor gene CDKN2A (5%) that encodes p16. In some cases, BCL6 mRNA markedly increases upon transformation, but this is not required. Karyotypic abnormalities have also been implicated in the transformation process del 6q, trisomy 7, and trisomy 12. TNFRSF14, a newly recognized tumor suppressor gene, is the likely candidate implicated in transformation when the frequent (35%) deletion 1p36.3 occurs.

3. What is the expected survival for this patient?

Less than one year
1–3 years
3–5 years
More than 5 years

The most commonly reported median OS for transformed FL is between 1 and 2 years. The British Columbia series reported a posttransformation median survival of 1.7 years, while survival was 1.2 years in the series from St Bartholomew’s Hospital and only 7 months in the Lyon series. The majority of patients in these series were treated before 2000 and did not have access to rituximab and the newer therapies. Patients diagnosed between 2002 and 2009 who were included in the recently reported Iowa/Mayo series had a remarkable median overall survival of 50 months from the time of transformation. Patients who were initially observed had the worst prognosis and those treated with rituximab monotherapy, the best in this series. A series from a Swiss database that covers the period from 1979 to 2007 and included patients treated in the rituximab era reported a median survival after a transformation of 2.7 years. In this series, FLIPI score at initial FL diagnosis, age more than 60 years, and Eastern Cooperative Oncology Group (ECOG) performance status >1 at the time of transformation were associated with a worse prognosis. Other series reported elevated LDH level, extensive disease at transformation, and not achieving CR after salvage therapy for TL as poor prognostic factors. In contrast to the Iowa/Mayo experience, patients who did not receive chemotherapy prior to transformation had a better prognosis in the Stanford cohort.

< div class='tao-gold-member'>

Only gold members can continue reading. Log In or Register to continue