Background and Aims The prognosis of lymphoma that occurs in patients with inflammatory bowel disease [IBD] is poorly known. Methods A multicentre retrospective cohort analysis was done in seven French tertiary centres from 1999 to 2019. Only lymphoma occurring in patients with previous established diagnosis of IBD were analysed. The primary outcome was progression-free survival at 3 years. Results A total of 52 patients [male 65%, Crohn's disease 79%, median age 48.3 years, median duration of IBD 10.1 years] were included, of whom 37 had been previously exposed to immunosuppressants and/or biologics for at least 3 months and 20 had primary intestinal lymphomas. The lymphoma histological types were: diffuse large B cell lymphomas [ N = 17], Hodgkin lymphomas [ N = 17], indolent B cell lymphomas [ N = 12], and others including T cell lymphomas, mantle cell lymphomas, and unclassifiable B cell lymphoma [ N = 6]. The median follow-up after lymphoma was 5.1 years (interquartile range [IQR] 4–7.8). Progression-free survival at 3 years was 85% in the overall population (95% confidence interval [CI] 75%–96%) with no significant difference between the exposed and unexposed group, 79% for patients exposed to immunosuppressants and/or biologics [95% CI 67%–94%], and 83% for patients diagnosed with primary intestinal lymphoma [95% CI 67%–100%]. No relapse of IBD has been observed during chemotherapy. The IBD relapse rate at the end of the last chemotherapy cycle was 23% at 3 years [95% CI 11%-39%] in the overall population. Conclusions In this large cohort, the prognosis for lymphomas occurring in IBD appears to be good and similar to what is expected, irrespective of the exposure to biologics and/or immunosuppressants.
Keywords: IBD; lymphoma; prognosis
The management of inflammatory bowel diseases [IBD] has been rapidly evolving over the past 20 years, with an increasing use of early combined therapy with tumour necrosis factor alpha [TNF-α] antagonists and immunosuppressants, as well as an expansion in the number of treatment options available.1 Tolerance to immunosuppressive therapies has become a major concern, with an initial focus on serious infections.2,3 However, there are also excess deaths from cancers in the IBD population, which can be attributed to chronic intestinal inflammation or to the carcinogenic effects of the immunosuppressive or biologic drugs.4 In the absence of exposure to immunosuppressants, the absolute risk of lymphoma in IBD patients is relatively low and does not seem to exceed that of the general population.5
Conversely, it has been shown in a recent meta-analysis that IBD patients taking thiopurines have a nearly 6-fold higher incidence of lymphoma when compared with the general population, with a large proportion of post-transplant-like Epstein-Barr virus [EBV]-associated B cell lymphomas. This risk appears to be approximately twice as high in men as in women and related to the duration of exposure, with only a few cases described in patients with less than 1 year of exposure. Young males under 30 years and patients over 50 years are at higher risk. Nevertheless, the absolute risk remains very low and estimated at one in 2000 patients per year in subjects under 50 years of age, and one in 350 patients per year in subjects over 50 years of age.
Additionally, discontinuation of thiopurines does not appear to result in a persistent elevated risk, suggesting that immunosuppression rather than direct DNA damage represents the main risk factor in the development of excess lymphomas.6 It was also recently reported in a nationwide French cohort study that the use of TNF-α antagonists was associated with an increased risk of lymphoma both in monotherapy [adjusted hazard ratio of 2.41] and combination therapy [adjusted hazard ratio of 6.11], with a similar magnitude for non-Hodgkin's lymphoma and Hodgkin lymphoma.7 The risk of lymphoma is greater in men and in patients older than 65 years, except for hepatosplenic T cell lymphomas.8 Moreover in the CESAME cohort, an excess risk in IBD patients of primary intestinal lymphomas arising in inflammatory areas was demonstrated, particularly in patients exposed to thiopurines.9 Although this lymphoma excess risk is well established, there are few data on the prognosis in patients with IBD. Only one small study including a historical cohort of 14 patients was published to date.10 The aim of our study was to assess the prognosis of lymphomas in patients with IBD since the early 2000s, with the addition of the routine use of rituximab, particularly in patients who were given immunosuppressants and/or biologics, and to describe the subsequent evolution of IBD after haematological remission has been achieved.
Data from 52 patients with an established diagnosis of IBD and diagnosed with a lymphoma [whatever the histological type] at age over 18 years between September 1999 and April 2019 were retrospectively collected in seven French centres. The study protocol has been declared to the CNIL [Commission Nationale de l'Informatique et des Libertés; number 2,212,958 v 0] and has been registered in the public directory managed by the INDS [Institut National des Données de Santé]. Lymphoma cases were identified either from a hospital database known as PMSI [Programme de Médicalisation des Systèmes d'Information] or from databases of the haematology or gastroenterology departments. Histological reports of all lymphomas were reviewed by a referent haematologist [CT] to assess the different subtypes. A lymphoma was defined as EBV-related in case of positive in situ hybridisation by EBER probe [Epstein-Barr virus Encoded Ribonucleic acid] and/or immunohistochemistry targeting LMP1 [Latent Membrane Protein 1]. Patients were considered exposed to immunosuppressants and/or biologics if they had received at least one of these treatments for more than 3 months prior to lymphoma diagnosis, whether the treatment was ongoing or discontinued at that time.
All files except two were reviewed by the same person [TS]. Baseline data included gender, age, smoking status, IBD type and duration, IBD phenotype at diagnosis of lymphoma [according to Montreal classification], history of IBD surgery, previous IBD treatments at diagnosis of lymphoma, clinical activity of IBD at diagnosis of lymphoma, and characteristics of lymphoma. Follow-up data collected included specific medical and surgical treatment of lymphoma, severe infectious complications induced by chemotherapy for lymphoma, evolution of lymphoma, evolution of IBD after haematological diagnosis, and death from all causes.
The primary outcome was progression-free survival at 3 years. Secondary outcomes were: overall survival at 3 and 5 years, progression-free survival at 5 years, the rate of severe chemotherapy-related complications defined by hospitalisation for infection or outpatient antibiotic treatment during chemotherapy and within 3 months after discontinuation of lymphoma treatment, and the cumulative incidence of relapse of IBD after remission had been achieved.
Quantitative variables are expressed by their median and interquartile range [IQR]. Qualitative variables are presented by their numbers and corresponding percentages. Progression-free survival was defined as the time between the date of diagnosis of lymphoma and the date of morphological and/or histological examination documenting the progression of lymphoma. Overall survival was defined as the time between the date of diagnosis of lymphoma and the date of death. The distributions of overall survival and progression-free survival were estimated using the Kaplan-Meier method.
Relapse of IBD after completion of chemotherapy was estimated on the subset of patients who received chemotherapy, in a competing risk framework with death without relapse as the competing event. Cumulative incidence of relapse was estimated using Gray methodology.
All statistical analyses were performed using R version 3.5.2.
We identified 57 cases of lymphoproliferation. Five patients were excluded from analysis after haematological review [two EBV-induced lymphomatoid granulomatosis, one EBV-induced lymphoproliferation, one Castleman disease, one not histologically documented]. All patients had negative HIV serology, and the median age at diagnosis of lymphoma was 48.3 years [IQR 34–63]. The median time between the diagnosis of IBD and lymphoma was 10.1 years, and the median follow-up after lymphoma diagnosis was 5.1 years with a cumulated follow-up of 272.7 person-years. Patients had Crohn's disease in almost three-quarters of cases, and only one-third of the overall population had active disease at diagnosis of lymphoma. The main characteristics of the overall population at diagnosis of lymphoma and of the patients with and without immunosuppressive therapy are presented in Table 1 and Figure 1, respectively.
Table 1. Characteristics of the overall population at lymphoma diagnosis.
Population. Overall Age at lymphoma diagnosis [years]a 48.3 [34–63] Duration of IBD [years]a 10.1 [5–16] Male sex 34 [65%] Smoking status 21 [40%] Current 4 [8%] Previous 17 [33%] Never 22 [42%] Unknown 9 [17%] Type of IBD Crohn's disease 41 [79%] Ulcerative colitis 11 [21%] Extent of ulcerative colitis [Montreal classification] E2 [left-sided] 9 [82%] E3 [extensive] 2 [18%] Location of Crohn's disease [Montreal classification] L1 [ileal] 17 [41%] L2 [colonic] 7 [17%] L3 [ileocolonic] 16 [40%] NA 1 [2%] Previous surgery for IBD 19 [36,5%] Immunosuppressants and/or biologics >3 months 37 [71%] Anti-TNF antagonists 15 [41%] Exposure to a single anti-TNF/two anti-TNFs 10/5 Ongoing [<3 months] 13 Infliximab/adalimumab 5/8 Combination therapy [thiopurines/methotrexate] 3 [1/2] Discontinued 2 Combination therapy [ongoing/discontinued] 13 [3/10] Thiopurinesb 32 [86%] Ongoing [<3 months] 17 Discontinued 15 Combination therapy [with anti-TNF: ongoing/discontinued] 13 [1/12] Methotrexate 8 [22%] Ongoing [<3 months] 3 Discontinued 5 Combination therapy [with anti-TNF: ongoing/discontinued] 7 [2/5] Ustekinumab [ongoing] 2 [5%] Others 3 [8%]
1 IBD, inflammatory bowel disease; NA, not available; TNF, tumour necrosis factor.
- 2 a[Median, interquartile range].
- 3 bDuration <3 months in two patients exposed to anti-TNF antagonists.
Graph: Figure 1. class="chapter-para">Flow chart.
The histological characteristics of lymphomas were as follows: 34 aggressive lymphomas [17 diffuse large B cell and 17 Hodgkin lymphomas], 12 indolent B cell lymphomas, 3 T cell lymphomas, 2 mantle cell lymphomas, and one unclassifiable B lymphoma. The main characteristics of lymphomas at diagnosis in the overall population are presented in Table 2.
Table 2. Characteristics of lymphomas at diagnosis in the overall population.
Overall population. Histological type [WHO, 2016] Diffuse large B cell lymphoma 17 [33%] Not Otherwise Specified [NOS] 13 PTLDa 2 Primary of the central nervous system 1 Germinal centre 1 Hodgkin 17 [33%] Classical 14 PTLDa 2 Nodular lymphocyte predominant 1 Indolent 12 [23%] Follicular 8 Marginal zone 2 Gastric MALT 2 T cell lymphoma 3 [5%] Hepatosplenic 1 Anaplastic 1 T/NK cell 1 Mantle cell 2 [4%] B cell unclassifiable 1 [2%] Exposure to immunosuppressants and/or biologics >3 months 37 [71%] Primary intestinal lymphomasb 20 [38%] EBV-associated lymphomasc [ n = 35 15/35 [43%] International Prognostic Index [IPI] criteria Age >60 years 14 [27%] Ann Arbor staging Stage I 9 [17%] Stage II 9 [17%] Stage III 5 [10%] Stage IV 25 [48%] NA 4 [8%] LDH at diagnosis High 14 [27%] Normal 32 [62%] NA 6 [11%] ECOG performance status ECOG Score 0–1 35 [67%] ECOG Score ≥ 2 15 [29%] NA 2 [4%] Number of extranodal sites on presentation ≥2 13 [25%] 0 or 1 35 [67%] NA 4 [8%] Front line chemotherapy regimen 47 [90%] R-CHOP 26 [55%] R-CHOP like 6 [12%] Others 15 [32%] Use of intrathecal chemotherapyd 13 [28%] Autologous stem cells transplantatione 9 [17%] Upfront 5 Upon relapse 4 Radiation therapy [Hodgkin's disease] 3 [6%] Intestinal resection 12 [23%] Others 5 [10%]
- 4 MALT, mucosa-associated lymphoid tissue; EBV, Epstein-Barr virus;NA, not available; R-CHOP, rituximab, cyclophosphamide, doxorubicine, vincristine and prednisone; LDH, lactate dehydrogenases.
- 5 aPost-transplant lymphoproliferative disorders [PTLD].
- 6 bSeventeen of the gastrointestinal tract, two gastric MALT, one pancreatic lymphoma.
- 7 cIn situ hybridisation [EBER] or immunohistochemistry [LMP1].
- 8 dMetotrexate [lumbar puncture N = 11 patients, and intravenous N = 2 patients].
- 9 eBEAM conditioning in eight out of nine patients.
The main inaugural symptoms at lymphoma diagnosis were: appearance or worsening of digestive symptoms [N = 19], peripheral adenopathy [N = 11], B symptoms [N = 9]. The vast majority of the patients were treated with a front line chemotherapy regimen [47/52; 90%]. The remaining patients were managed as follows: simple monitoring, discontinuation of TNF-α antagonists, surgery alone, eradication of Helicobacter pylori and last, one patient died of lymphoma before any treatment.
The main characteristics of the patients according to whether or not they were exposed to immunosuppressants and/or biologics are presented in Table 3.
Table 3. Characteristics of the patients according to their exposure to treatment.
Population. Exposed to immunosuppressants and/or biologics Not exposed to immunosuppressants and/or biologics Age at lymphoma diagnosis [years]a 44.3 [33–63] 53.2 [48–62] Duration of IBD [years]a 9.2 [6–16] 10.5 [1–32] Male sex 26 [70%] 8 [53%] Type of IBD Crohn's disease 29 [78%] 12 [80%] Ulcerative colitis 8 [22%] 3 [20%] Histological type [WHO, 2016] Diffuse large B cell lymphoma 13 [35%] 4 [27%] Not Otherwise Specified [NOS] 10 3 PTLDb 2 - Primary of the central nervous system - 1 Germinal centre 1 - Hodgkin 12 [32%] 5 [33%] Classical 9 5 PTLDb 2 - Nodular lymphocyte predominant 1 - Indolent 8 [22%] 4 [27%] Follicular 6 2 Marginal zone 2 - Gastric MALT - 2 T cell lymphoma 3 [8%] 0 Hepatosplenic 1 - Anaplastic 1 - T/NK-cell 1 - Mantle cell 1 [3%] 1 [6.5%] B cell unclassifiable 0 1 [6.5%] Primary intestinal lymphomas 16 [43%] 4 [27%] EBV-associated lymphomasc [ n = 29 14/29 [48%] n = 6 1/6 [17%] International Prognostic Index [IPI] criteria Age >60 years 10 [27%] 4 [27%] Ann Arbor Stage IV 20 [54%] 5 [33%] High LDH at diagnosis 12 [32%] 2 [13%] ECOG performance status ≥ 2 12 [32%] 3 [20%] Extranodal sites on presentation ≥2 11 [30%] 2 [13%] Deaths 7 [19%] 3 [20%] Macrophage activation syndrome prior to chemotherapy 1 [3%] - Refractory lymphoma 1 [3%] 1 [6,5%] Relapsing lymphoma 2 [5%] 1 [6,5%] Other cancer 2 [5%] - Infection 1 [3%] 1 [6,5%]
- 10 IBD, inflammatory bowel disease; MALT, mucosa-associated lymphoid tissue; EBV, Epstein-Barr virus; LDH, lactate dehydrogenases.
- 11 a[Median, interquartile range].
- 12 bPost-transplant lymphoproliferative disorders.
- 13 cIn situ hybridisation [EBER] or immunohistochemistry [LMP1].
Within our cohort, 37 patients had a current or previous exposure to immunosuppressants and/or biologics of at least 3 months [71% of the whole cohort]. Median follow-up after lymphoma diagnosis in this subgroup was 4.6 years. A full list of cumulative drug exposures at diagnosis can be found in Figure 2. The majority of these patients were exposed to thiopurines [86%] and nearly half were exposed to TNF-α antagonists [41%]. More than one-third of this population was exposed to combination therapy [N = 13, 35%]. The median cumulative duration of exposure to thiopurines and TNF-α antagonists was 30.8 months and 49.9 months, respectively. Considering only this exposed group, the histological characteristics of lymphomas were as follows: 25 aggressive lymphomas [13 diffuse large B cell and 12 Hodgkin lymphomas], eight indolent B cell lymphomas [six follicular lymphomas and two marginal zone lymphomas], three peripheral T cell lymphomas, and one mantle cell lymphoma. These lymphomas were EBV-associated in about half of the cases [14 cases out of 29 tested].
Graph: Figure 2. class="chapter-para">Cumulative exposure to treatment and occurrence of lymphoma [lollipop graph].
Twenty patients had primary intestinal lymphoma [38% of the whole cohort]. The median follow-up after lymphoma diagnosis in this subgroup was 5.2 years. Patients were mainly suffering from Crohn's disease [N = 17; 85%], with ileal or ileocolic location [88%] and a B2/B3 phenotype [N = 14; 82%]. The lymphoma site arose in IBD lesions in 13 cases [65%]. The majority of these patients were exposed to immunosuppressants and/or biologics [N = 16; 80%] distributed as follows: thiopurines [15; among whom 11 were treated at diagnosis; 94%], TNF-α antagonists [four; among whom three were treated at diagnosis; 25%], methotrexate [two; among whom one was treated at diagnosis; 13%], and others [three; 19%]. The histological characteristics of these lymphomas were as follows: 11 diffuse large B cell lymphomas, four Hodgkin lymphomas, three indolent lymphomas [one follicular lymphoma and two MALT lymphomas], one NK/T cell lymphoma, and one unclassifiable B cell lymphoma. These lymphomas were predominantly EBV-associated [10 cases out of 17 tested].
Among these patients a surgical resection was required in 12 cases: eight emergency surgery revealing the lymphoma [four small bowel obstructions, three peritonitis, and one refractory acute severe flare] and four secondary resections [coloprotectomy only without chemotherapy, small bowel resection for incomplete remission after front line chemotherapy, ileocaecal resection for small bowel obstruction during chemotherapy course, ileocaecal resection for small bowel obstruction due to fibrotic stenosis]. Of note, no perforation under chemotherapy was noted in patients with conservative treatment.
The progression-free survival at 3 years for the whole population was 85% [95% CI 75%–96%] [Figure 3]. According to histological subtypes, the progression-free survival at 3 years was 94% for diffuse large B cell lymphomas, 91% for follicular lymphomas, and 70% for Hodgkin lymphomas. The progression-free survival at 5 years for the whole population was 69% [95% CI 56%–86%]. The overall survival at 3 and 5 years for the whole population was 89% [95% CI 80%–99%] and 76% [95% CI 64%–92%], respectively. According to histological subtypes, the overall survival at 3 years was 94% for diffuse large B cell lymphomas, 91% for follicular lymphomas, and 85% for Hodgkin lymphomas. Severe chemotherapy-related complications were observed in 55% [N = 23/42; five missing data]; among them, 18 patients were hospitalised for an infection and five outpatients were managed with only antibiotics. Infections were characterised as follows: no microbiological documentation [N = 10], central line infection [N = 5], infectious colitis [N = 3], pneumonitis [N = 2], urinary tract infection and dental infection [N = 1 each]. In the subgroup of patients exposed to immunosuppressants and/or biologics, the progression-free survival at 3 years was 79% [95% CI 67%–94%]. In the subgroup of patients presenting a primary intestinal lymphoma, the progression-free survival at 3 years was 83% [95% CI 67%–100%].
Graph: Figure 3. class="chapter-para">Progression-free survival of the overall population and according to histological subtypes.
At the time of lymphoma diagnosis, immunosuppressants and/or biologics were stopped in all patients exposed to thiopurines [N = 17], TNF-α antagonists [N = 13], and ustekinumab [N = 2]. A maintenance treatment of IBD was continued concomitantly with lymphoma treatment in 39% of patients in the overall population [N = 22/52]. The distribution was as follows: aminosalicylates [N = 13], budesonide [N = 5], methotrexate [N = 2], tacrolimus [N = 1], and leucocyte apheresis [N = 1]. The cumulative incidence of relapse of IBD after lymphoma remission was 23.1% at 3 years [95% CI 10.5%–38.5%] [Figure 4]. Of note, no relapse of IBD has been observed during chemotherapy course. In the subgroup of patients exposed to an immunosuppressants and/or biologics, the cumulative incidence of relapse of IBD after lymphoma remission was 27.6% at 3 years [95% CI 11.5%–46.4%]. In the primary intestinal lymphoma subgroup, the cumulative incidence of relapse of IBD after lymphoma remission was 14.3% at 3 years [95% CI 2.1–37.5]. Of the 12 patients with a relapse of IBD during follow-up, 11 patients received one or several successive treatments [methotrexate N = 4, vedolizumab N = 4, budesonide N = 3, infliximab N = 3, golimumab N = 1, ustekinumab N = 1, tacrolimus N = 1]. In addition, four patients underwent intestinal resection [ileocolic resection N = 2, sigmoidectomy N = 1, subtotal colectomy N = 1].
Graph: Figure 4. class="chapter-para">Cumulative incidence of inflammatory bowel disease [IBD] relapse after the end of chemotherapy.
During the follow-up, 10 patients died. Among them, death was directly related to lymphoma in six cases. Their characteristics were as follows: one prior to any treatment [EBV-associated Hodgkin lymphoma with inaugural refractory macrophage activation syndrome], two refractory lymphomas [a diffuse large B cell lymphoma and a hepatosplenic T cell lymphoma], and three relapsed lymphoma [a primary diffuse large B cell lymphoma of the central nervous system, a Hodgkin lymphoma, and a mantle cell lymphoma]. Two patients died of infections [one pneumonitis in a patient receiving R-CHOP chemotherapy and one septic shock complicating a pneumococcal pneumonia 1 year after an allogeneic bone marrow transplantation], and two of other cancers [a metastatic small bowel adenocarcinoma and a metastatic prostate cancer].
To date, there are few data describing the follow-up of lymphomas in patients with diagnosed IBD [cf. Table 4] and, to our knowledge, our work represents the largest series in the literature. We report in the whole study population a progression-free survival and an overall survival at 3 years of 85% and 89%, respectively. These survival data appear to be close to those observed in studies with non-IBD patients, especially in subjects under 60 years of age, which have validated the R-CHOP protocol in first line treatment of diffuse large B cell lymphomas.11,12 Progression-free survival at 3 years for Hodgkin lymphoma ranges from 66% to 93%, depending on the type of protocol used and the population treated.13 The progression free survival at 3 years for Hodgkin lymphoma observed in our cohort was 85%, which therefore seems to be in line with the data in the literature.
Table 4. Series on IBD patients diagnosed with lymphoma.
First author [year]reef.. Study design. Lymphoma cases. Lymphoma types. Outcome. Kumar Cases report 4 Primary gastrointestinal Hodgkin lymphomas Two deaths, two in remission at 17 and 25 months Kandiel Meta-analysis 11 Seven non-Hodgkin's lymphomas, two Hodgkin lymphomas, two not reported Not reported Beaugerie Cohort 23 22 non-Hodgkin's lymphomas, one Hodgkin lymphoma Eight deaths after a mean follow-up of 35 months Siegel Meta-analysis 13 Non-Hodgkin's lymphomas Six deaths at the time of most recent follow-up El Mourabet Retrospective case series 9 Five non-Hodgkin's lymphomas, four Hodgkin lymphomas£ No death within a mean follow-up of 72.8 monthsa Sokol Cohort 14 Primary intestinal lymphoma disorder Not reported Sultan Retrospective case series 14 13 non-Hodgkin's lymphomas, one Hodgkin lymphoma 3 years overall survival: 71% Khan Retrospective cohort 119 Not reported Not reported Nyboe Andersen Cohort 8 Six non-Hodgkin's lymphomas, two Hodgkin lymphomas Not reported Kotlyar Meta-analysis 93 59 non-Hodgkin's lymphomas, eight Hodgkin lymphomas, 26 not recorded Not reported Lemaitre Cohort 336 281 non-Hodgkin's lymphomas, 55 Hodgkin lymphomas Not reported Barzilai Case report and review 22 Primary gastrointestinal Hodgkin lymphomas Not reported
- 14 IBD, inflammatory bowel disease.
- 15 aDescription of long-term survivors only.
Follicular lymphomas represent a heterogeneous entity whose management remains debated. Progression-free survival at 3 years for follicular lymphoma ranges from 51% to 91%, depending on the risk groups.14 In a recent large observational retrospective study, 5-year overall survival was 91% in patients under 50 years of age, which is identical to that in our series.15 In the subgroup of patients exposed to immunosuppressants and/or biologics, progression-free survival and overall survival at 3 years were 79% and 85%, respectively. Although a statistical comparison is not relevant given the small number of patients, this figure is close to that of the overall population of the study, suggesting that the prognosis for lymphoma does not appear to be affected by immunosuppressive treatment exposure. The chemotherapy-induced toxicity observed in our cohort also appears to be close to that usually observed in the literature, in particular for antibiotic use.16 As expected, a majority of patients who developed lymphoma were exposed to immunosuppressive therapy, mostly by thiopurines with a clear association with EBV infection.6,17
In our series, in patients exposed to thiopurines, lymphoma was EBV-associated in 48% of the cases compared with one EBV-associated case among six screened cases of lymphoma without exposure to immunosuppressive agents. The occurrence of lymphoma almost systematically required the discontinuation of immunosuppressants and/or biologics in our study, which is consistent with a recent French series.18 Another striking finding of our study is the high proportion of primary lymphomas of the digestive tract, an exceptional feature in patients who do not have IBD, although they represent 38% of our cohort with a high proportion of EBV-associated lymphomas similar to those of post-transplantation lymphoma disorders. These data are consistent with the case-control study nested in the CESAME cohort.9 Moreover, this is in line with data from other studies.19,20
No relapse of IBD requiring the reintroduction of specific treatment during chemotherapy has been observed. This may be related to the use of high doses corticosteroids for antitumour or antiemetic purposes. However, a direct beneficial effect of chemotherapy cannot be excluded. We report a cumulative incidence of relapse of IBD after chemotherapy of 23.1% at 3 years. These data on such a large cohort have never been reported before. The only published series, in a population of eight patients with lymphoma developed during a known Crohn's disease, reported a relapse in six cases, two of which required immunosuppressant re-treatment within 3 years.21 In our study, this rate is relatively low given the discontinuation of IBD maintenance treatments. Indeed, the expected relapse rate of IBD when immunosuppressants are stopped is approximately 75% at 5 years.22
Nevertheless, IBD was clinically quiescent in nearly three-quarters of patients at the time of lymphoma diagnosis, and 12 patients had early intestinal resection removing the associated inflammatory zone in addition to the lymphoma. Interestingly, only one patient relapsed with IBD out of nine treated with autologous haematopoietic stem cell transplantation, which corroborates the preliminary data of this therapeutic strategy in refractory IBD.23 We have not observed any cases of lymphoma relapse after the reintroduction of treatment for IBD.
The main strengths of this work are the size of the cohort, with a median follow-up after lymphoma diagnosis of more than 5 years, and a relevant subset of cases exposed to immunosuppressants and/or biologics—given the rarity of this type of pathology in IBD patients. In addition, all haematological data [histology, specific treatments administered] were reviewed centrally by a haematologist highly specialised in the field of lymphomas.
This study clearly has some limitations due to its retrospective design and limited number of patients. However, even if the excess risk of lymphoma in IBD is clearly demonstrated in patients exposed to immunosuppressants, it should be put into perspective with the very low absolute risk of developing lymphoma.28 It therefore seems difficult to obtain more robust data within a multicentre cohort. Recruitment bias cannot be eliminated either, as patients are managed in tertiary centres. This may explain the high number of Hodgkin lymphoma cases observed. Nevertheless, given the complex management of these patients, it seems likely that most of them will be managed in expert centres. Our cohort could therefore constitute a representative sample.
Another limiting factor is the lack of accurate molecular data to better characterise lymphomas, due to the lack of additional analysis on some tumour blocks. This could mask some prognostic factors that may have led to discussions about alternative therapeutic management. Nevertheless, some patients were treated when these molecular tools were not available. In addition, the protocols used seem fully in line with the recommendations. Finally, even if the survival rates observed in our cohort appear to be superimposed on those observed in non IBD patients with sporadic lymphoma, only a case-control study could have confirmed these data.
In conclusion, our study confirms that lymphomas occurring in IBD are frequently associated with exposure to immunosuppressants, with a high rate of primary intestinal lymphomas and EBV-associated lymphomas. We report for the first time that progression-free survival at 3 years in IBD patients presenting with lymphomas is good and appears to be superimposed on that of non-IBD lymphomas. We also provide new data on the IBD course after lymphoma diagnosis, with no relapse during chemotherapy and a low relapse rate at 3 years. We believe that our work provides additional information to the clinician who is usually confronted with this type of potentially life-threatening problem in a context of drug toxicity.
None
The authors have no financial conflict of interest to declare for this study.
TS: writing and data collection. AA: data collection, copyediting of the manuscript. JK, VA, XT, GM, PC, TA, LB, MA: copyediting of the manuscript. JL, AW-P: statistical analysis. CT, LG: study design, copyediting of the manuscript. JMG: manuscript coordination, writing, and data collection.
By T Severyns; J Kirchgesner; J Lambert; C Thieblemont; A Amiot; V Abitbol; X Treton; D Cazals-Hatem; G Malamut; P Coppo; L Galicier; A Walter-Petrich; B Deau-Fischer; C Besson; T Aparicio; L Beaugerie; M Allez and J M Gornet
Reported by Author; Author; Author; Author; Author; Author; Author; Author; Author; Author; Author; Author; Author; Author; Author; Author; Author; Author
Dr J. M. Gornet, MD, Service de Gastroentérologie, Hôpital Saint Louis, 1 Avenue Claude Vellefaux, 75010 Paris, France. Tel.: +33 1 42 49 95 75; Fax: +33 1 42 49 91 68
