The management of gastrointestinal follicular lymphoma: Some observations on a rare disease

The gastrointestinal tract is the commonest presenting site of primary extranodal lymphoma. Patients typically present with localized disease, and either diffuse large B-cell or marginal zone histology [1]. By contrast, gastrointestinal follicular lymphoma (GI-FL) is rare, representing only 1–3% of primary gastrointestinal lymphoma [2]. In an extensive literature review of 249 patients with GI-FL, Yamamoto et al. reported that most patients presented with small bowel involvement, stage I–II disease and grade I histology [3]. Tari et al. reported that GI-FL was more likely than nodal FL to present with early stage disease (68% vs. 17%), grade I histology (81% vs. 61%) and a low Follicular Lymphoma International Prognostic Index (FLIPI) score (61% vs. 33%) [4]. Although usually confined to the GI tract, thorough evaluation of GI-FL utilizing double balloon or capsule endoscopy reveals multiple sites of small intestinal involvement in 70–80% of cases [3,5]. Anatomical location within the GI tract may have prognostic implications, with primary duodenal disease having a significantly higher progression-free survival rate than non-duodenal presentations (p = 0.001) [6]. In this edition of Leukemia and Lymphoma, Kiess and Yahalom contribute to our understanding of GI-FL, providing further evidence that duodenal lymphoma does indeed have a very indolent natural history, and suggesting that outcome may be independent of the initial management strategy [7].

In light of the rarity and potentially distinct biology of GI-FL, how are clinicians best to decide on its management, particularly in patients for whom curative opportunities may exist? It may be helpful to consider the management principles of GI-FL within the context of those applied to the management of nodal FL. Systemic chemotherapy, immunochemotherapy, radio-immunotherapy and high dose chemotherapy with stem cell transplant, all of which can be effectively applied in disseminated nodal FL, can, of course, also be utilized during the course of management of GI-FL. Remissions of reasonable duration in GI-FL have been observed after single-agent Mabthera as well as after combination immuno-chemotherapy, but as is the case with nodal FL, the curative potential of immuno-chemotherapy in GL-FL is limited [7–10]. Indeed, in their analysis, Kiess and Yahalom report that with a median follow-up of 4.8 years, three of seven patients who received chemotherapy have had a treatment failure [7].

Do potentially curative treatment options exist for early stage GI-FL? In principle, surgical resection might be curative for patients who, after thorough evaluation, are considered to have disease confined to one segment of bowel, in a manner analogous to excisional biopsy in cases of nodal FL. However, given the multifocal nature of GI-FL, the role of surgery is generally limited to establishing a diagnosis and treating actual or imminent complications [11].

Radiotherapy provides local tumor control in almost all treated patients with stage I–II nodal FL, and long-term remission/cure in about half of treated cases [12]. With opportunities for more accurate staging through the use of positron emission tomography (PET) scanning and detection of minimal disease in blood and marrow, improved results following radiotherapy may be achieved as the patient population is enriched for those with truly early-stage nodal FL [13]. Can this approach be applied for patients with stage I–II GI-FL? At least three of the series reporting the outcome following radiotherapy for GI-FL have shown results consistent with those achieved in nodal FL [7,9,10]. The series from Schmatz et al. included 19 patients treated with radiotherapy with no treatment failures reported [9]. In the report from Misdraji et al., eight patients were treated with radiotherapy (six duodenal) with only one treatment failure [10]. The analysis from Kiess and Yahalom included 10 patients who received radiotherapy, eight of whom are in ongoing first remission, with no local relapses reported after initial complete remission. This contrasted with six relapses limited to the abdomen in patients treated without radiotherapy [7]. However, in patients with multifocal GI-FL there are two concerns regarding the application of loco-regional radiotherapy. First, as for nodal FL, more extensive loco-regional disease (stage II vs. I, or greater disease bulk) may be associated with a greater likelihood of subclinical dissemination and therefore a lower, but not absent, likelihood of cure by radiotherapy [12]. Second, adequate radiation therapy for patients with multifocal bowel involvement generally requires whole abdominal irradiation (WAR) [12]. Although WAR has a long history in the management of lymphoma and ovarian cancer, there appears to be decreasing familiarity and comfort with this technique in recent years [11]. Several potential toxicities of WAR, including bone marrow suppression, acute and chronic bowel disturbance and renal impairment, are generally avoidable with appropriate attention to radiation dose, technique and delivery. Of greater concern is the risk of second malignancy after wide-field radiation, which must be carefully weighed against the potential for cure of the lymphoma, particularly for younger patients.

Although there is population-based evidence for the value of radiotherapy for early-stage nodal FL, this has not been evaluated in a randomized trial [14]. Some investigators suggest that the results of radiotherapy are at least partly attributable to the indolent natural history of FL, and advocate observation for early-stage nodal FL [15]. Similarly, as reported by Keiss and Yahalom, observation may be appropriate initial management for some patients with asymptomatic GI-FL. Similar observations have been reported in previous series [3,9]. For example, Schmatz et al. reported 24 patients managed by a “watch and wait” approach. Seven demonstrated spontaneous regression while the remainder had stable disease, and no transformation to large cell lymphoma was observed [9]. Observation may be particularly suitable for patients with duodenal lymphoma, and for older patients for whom the natural history of the lymphoma is unlikely to impact survival.

As early-stage FL has a very indolent natural history, it is possible that clinically meaningful differences in outcome might only become apparent 10 or more years after treatment. The median follow-up for studies of GI-FL managed with a “watch and wait” approach is in the range of 5–8 years, and it is unknown whether the indolent tumor behavior observed over this time-frame will continue indefinitely. Patients with nodal FL who progress after observation may require a series of increasingly intensive treatments, including autologous or allogeneic transplant, where the benefits of graft-versus-lymphoma effect must be balanced against the risks of transplant related mortality and morbidity [16,17]. These considerations present a particular challenge in advising younger asymptomatic patients with GI-FL on their initial management choices.

For patients with symptoms, a significant tumor burden or disease that threatens incipient local complications, early initiation of treatment is warranted. This may also be appropriate for patients with non-duodenal GI-FL, which may not demonstrate the indolent behavior of duodenal disease. For example, Damaj et al. reported their experience with 25 patients with predominantly non-duodenal disease. Four of seven patients who received no initial treatment had progressive disease at a median follow-up of only 37 months. Fifteen of 18 treated patients achieved a complete remission, but the median duration of remission was only 31 months, and a third of patients required two or more lines of therapy [18]. The management of rare conditions such as GI-FL is challenging, and analyses of moderate-sized patient cohorts, such as the report from Kiess and Yahalom, may assist clinicians in applying the general management principles of early stage follicular lymphoma to rare disease entities such as GI-FL.

Potential conflict of interest

Disclosure forms provided by the authors are available with the full text of this article at www.informahealthcare.com/lal.