ARTHRITIS & RHEUMATISM Vol. 48, No. 6, June 2003, pp 1543–1550 DOI 10.1002/art.11144 © 2003, American College of Rheumatology

Lymphoma Subtypes in Patients With Rheumatoid Arthritis Increased Proportion of Diffuse Large B Cell Lymphoma ¨m,1 Anders Ekbom,2 Anca I. Catrina,3 Peter Biberfeld,3 Eva Baecklund,1 Christer Sundstro Nils Feltelius,3 and Lars Klareskog3 (DLBCL) (22 of 33 NHL patients, 67%) compared with that in the general population (30–40%). EBV was detected in 5 of 30 examined lymphomas from patients (17%). Twenty of the 22 DLBCL patients had RA with medium or high inflammatory activity, and 6 had been treated with a disease-modifying antirheumatic drug for >1 year. Conclusion. The findings of this study suggest an increased incidence of one specific lymphoma subtype, DLBCL, in RA patients, as well as a possible association with RA disease activity.

Objective. Patients with rheumatoid arthritis (RA) have an increased risk of developing malignant lymphoma. It is not clear whether the increase is confined to certain subtypes of lymphomas. Immunosuppressive therapy and Epstein-Barr virus (EBV) have been linked to the development of these lymphomas. To gain information about the baseline pattern of lymphoma subtypes in RA before the current widespread use of immunosuppressive drugs, we examined the distribution of lymphoma subtypes and the presence of EBV in a cohort of RA patients with a low frequency and duration of treatment with immunosuppressive drugs. Methods. By linking data from the Swedish Hospital Discharge Register and the Swedish Cancer Register, 42 cases of lymphoma were identified among 11,683 patients with RA in the Uppsala Health Care Region between 1964 and 1984. The medical records and paraffin-embedded lymphoma tissues were collected, and the lymphomas were reclassified using the World Health Organization classification. In situ hybridization was used to detect EBV. Results. Tissues from 35 patients were reviewed. Non-Hodgkin’s lymphoma (NHL) was found in 33 patients and Hodgkin’s lymphoma in 2 patients. There was an increased frequency of diffuse large B cell lymphoma

For several decades, it has been known that patients with rheumatoid arthritis (RA) have an increased risk of developing malignant lymphoma (1–5). The reason for this association is still incompletely understood, but immunosuppressive therapy in RA (6– 10) as well as Epstein-Barr virus (EBV) infection may play a role in the development of these lymphomas (11–15). Additionally, both we (16) and others (17) have reported that high RA disease activity in itself may be associated with an increased lymphoma risk. For further studies of the mechanisms of lymphoma development in RA patients, and to understand whether immunosuppression, immunostimulation, or both contribute to lymphoma development in RA, there is a need for more information about the characteristics of these lymphomas. As a consequence of the present widespread use of disease-modifying antirheumatic drugs (DMARDs) in RA patients, most data concerning lymphoma features in RA patients are derived from studies and case reports in which patients were treated with different immunosuppressive drugs (12,14,18–26). Information about lymphoma characteristics in RA patients not influenced by immunosuppressive drug treatment is limited. It is thus not clear whether RA patients develop

Supported by grants from the Swedish Cancer Society and the Swedish Rheumatism Association. 1 Eva Baecklund, MD, Christer Sundstro ¨m, MD, PhD: University Hospital, Uppsala, Sweden; 2Anders Ekbom, MD, PhD: Karolinska Institutet and Hospital, Stockholm, Sweden, and Harvard School of Public Health, Boston, Massachusetts; 3Anca I. Catrina, MD, Peter Biberfeld, MD, PhD, Nils Feltelius, MD, PhD, Lars Klareskog, MD, PhD: Karolinska Institutet and Hospital, Stockholm, Sweden. Address correspondence and reprint requests to Eva Baecklund, MD, Department of Rheumatology, University Hospital, SE-751 85 Uppsala, Sweden. E-mail: [email protected]. Submitted for publication November 21, 2002; accepted in revised form February 27, 2003. 1543

1544

BAECKLUND ET AL

specific subtypes of lymphomas, and, in particular, there is a lack of information about possible associations between clinical features of RA, such as disease severity, and different lymphoma subtypes. In this study, we investigated lymphoma biopsy samples obtained from individuals in a previously described cohort of RA patients with lymphoma (16). The lymphoma subtypes were classified, and occurrence of EBV was examined by in situ hybridization (ISH). Using this approach, we could take advantage both of detailed data from the patient records of disease characteristics (such as disease activity and drugs received) and of the fact that the lymphomas were diagnosed during the period 1965–1984, an era when aggressive immunosuppressive therapy in RA was still rare. PATIENTS AND METHODS Selection of patients and collection of clinical data. In the initial case–control study of risk factors for lymphoma in RA patients (16), we used the Swedish Hospital Discharge Register to identify all patients discharged from a hospital with a diagnosis of RA (ICD-7: 722,00; ICD-8: 712,10; 712,38– 712,39), either as a primary or secondary diagnosis, during 1964–1983 in the Uppsala Health Care Region, Sweden. We identified 42 cases of lymphoma (ICD-7, ICD-8: 200–202) among the 11,683 patients with RA, through record linkages between the Swedish Hospital Discharge Register and the Swedish Cancer Register between 1964 and 1984. Excluded first were patients with another malignant disease before the lymphoma and patients with lymphoma diagnosed within the first year after the first discharge for RA. This selection was made to avoid inclusion of patients already having lymphoma not associated with RA or possibly mimicking RA at the first hospitalization. In a previous study (3), the annual incidence of lymphoma in this cohort of RA patients was increased 2-fold compared with that in the general population (standardized incidence ratio 1.98; 95% confidence interval 1.5–2.6). From the data in the Swedish Hospital Discharge Register, the medical records from each patient’s hospital visits were obtained, as, subsequently, were these patients’ medical records from outpatient visits, visits to general practitioners, and the like, to cover the time period from the onset of the RA disease until the death of the patient or the end of the study period (December 31, 1984). The included patients had been treated in 28 different hospitals, including one university hospital and one clinic specializing in rehabilitation of patients with rheumatic diseases. Patients not fulfilling the 1987 American College of Rheumatology (ACR; formerly, the American Rheumatism Association) revised criteria for RA (27) were excluded. Clinical data and treatment information regarding both RA and the lymphoma were obtained. RA disease activity, estimated for the whole RA disease period, was scored as low, medium, or high based on erythrocyte sedimentation rates, numbers of swollen and tender joints, and the treating physician’s global assessments of disease activity. A total score

Table 1.

Rheumatoid arthritis disease activity scoring system* Score at each visit

ESR Number of tender and swollen joints Physician’s global assessment of disease activity

1

2

3

1–30 0–3 Mild

31–60 4–6 Moderate

ⱖ61 ⱖ7 Severe

* At each visit, the scores for erythrocyte sedimentation rate (ESR), number of tender and swollen joints, and physician’s global assessment were summed to yield a total score (maximum 9) for that visit. The final score was the mean of the total scores from all visits. Disease activity was considered low if the final score was 3–4, medium if the final score was ⬎4 to 8, and high if the final score was ⬎8 to 9.

based on these measures was estimated for each reported visit, and the final score was the mean of the total scores from all visits (Table 1). Analysis of lymphoma tissues. The Swedish Cancer Register contains information about the date of diagnosis and the reporting pathology laboratory, as well as a consecutive specimen code number of each diagnosed neoplasm. Using this information, the lymphoma tissues were localized and the original slides and paraffin-embedded lymphoma tissues were collected from the 6 pathology laboratories in the area. One experienced hematopathologist (CS) blinded to all clinical data classified the lymphomas according to the recently described World Health Organization (WHO) classification (28). The original slides were reviewed to confirm the diagnosis of lymphoma, additional sections were cut, and appropriate routine immunostaining was performed. A NexES IHC Staining Module (Ventana Medical Systems, Tucson, AZ) was used for the immunostaining. Antibodies against the following antigens were used: CD20, CD79a, CD5, CD3, CD8, kappa, lambda, Ki-67, epithelial membrane antigen, anaplastic lymphoma kinase, cyklinD1, Bcl-2, myeloperoxidase, CD68, and lysozyme (Dako, Glostrup, Denmark); DBA44 and CD56 (Cell Marque, Hot Springs, AR); CD23 and CD10 (Ventana Medical Systems); CD4 (Novocastra, Newcastle-upon-Tyne, UK); and CD30 (Beckman Coulter, Brea, CA). ISH for the detection of EBV (29) was performed using a fluorescein isothiocyanate (FITC)–labeled EBVencoded RNA (EBER) peptide nucleic acid (PNA) probe (Dako). A PNA ISH detection kit (Dako) was used for detection of the FITC-labeled probe according to the manufacturer’s instructions, and sections were evaluated by light microscopy. Sections from a previously EBER-scored Hodgkin’s lymphoma (HL) were included as positive controls.

RESULTS One of the original 42 patients was excluded for not fulfilling the ACR criteria for RA. This patient had primary Sjo ¨gren’s syndrome. Another 6 patients were excluded, 5 because their lymphoma tissues either could not be located or were of too poor quality for a reliable

LYMPHOMA SUBTYPES IN RA PATIENTS

1545

Table 2. Clinical data of 35 reviewed cases of patients with rheumatoid arthritis (RA) and lymphoma Sex, no. of men/no. of women Age at RA onset, mean (range) years Age at lymphoma diagnosis, mean (range) years RA disease activity, no. low/medium/high*

18/17 54 (30–80) 70 (49–89) 2/14/19

* See Table 1 for RA disease activity scoring system.

review, and 1 because the diagnosis of lymphoma could not be confirmed. Thus, the study finally included 35 reviewed cases of patients with RA and lymphoma. The clinical data are summarized in Table 2. One of the patients had secondary Sjo ¨gren’s syndrome. Tissue samples from 5 of the 35 patients could not be examined for EBV due to a shortage of lymphoma tissue. There were no notable differences in sex, age at onset of RA or at diagnosis of lymphoma, RA disease activity, and treatment for RA between patients whose cases were reviewed and those whose cases were not reviewed. Of the 35 lymphomas, 2 were HLs and 33 were non-HLs (NHLs). Among patients with NHL, there was a predominance of diffuse large B cell lymphoma (DLBCL), diagnosed in 22 of the 33 patients (67%). Five of the remaining 11 patients with NHL were diagnosed as having hepatosplenic T cell lymphoma

(n ⫽ 1), follicular lymphoma (n ⫽ 2), lymphoplasmacytic lymphoma (n ⫽ 1), or Burkitt’s lymphoma (n ⫽ 1). In 6 cases of NHL, it was not possible to make an exact diagnosis according to the WHO classification due to insufficient quality of the material. However, 3 of these were diagnosed as high-grade NHLs, either B cell lymphoma (n ⫽ 1) or unspecified NHL (n ⫽ 2), whereas the other 3 included 1 unspecified B cell lymphoma, 1 unspecified low-grade NHL, and 1 unspecified NHL (Table 3). Compared with the initial diagnosis of lymphoma (according to the Willis classification from 1948 [30], the Rappaport classification from 1966 [31], and the Kiel classification from 1974 [32]), 21 of the 35 cases (60%) had a diagnosis, after review, that was in accordance with, or not excluding, the initial one. The most frequent discrepancy was an initial diagnosis of Hodgkin’s disease (HD) changing to a diagnosis of DLBCL after review in 9 of 14 changed diagnoses. This overdiagnosis of HD in earlier periods has been described previously (33). EBV, searched for by ISH in 30 of the lymphomas, was detected in 4 of the 22 patients with DLBCL and in the patient with a lymphoplasmacytic lymphoma (Table 3). Of the 22 patients with DLBCL, 20 had RA characterized by high (n ⫽ 12) or medium (n ⫽ 8) inflammatory activity. Four of the 5 EBV-positive cases

Table 3. Distribution of lymphoma subtypes in 35 reviewed cases of patients with rheumatoid arthritis and lymphoma and results of in situ hybridization for Epstein-Barr virus (EBV) in 30 lymphomas from these patients

Lymphoma subtype, WHO classification* Mature B cell neoplasms Diffuse large B cell lymphoma Follicular lymphoma Lymphoplasmacytic lymphoma Burkitt’s lymphoma Unspecified high-grade B cell lymphoma† Unspecified B cell lymphoma† Mature T cell and NK cell neoplasms Hepatosplenic T cell lymphoma Hodgkin’s lymphoma Classic Hodgkin’s lymphoma, nodular sclerosis Unspecified lymphomas Unspecified high-grade NHL† Unspecified low-grade NHL† Unspecified NHL†

All lymphomas (n ⫽ 35)

Lymphomas examined for EBV (n ⫽ 30)

EBV-positive lymphomas (n ⫽ 5)

22 2 1 1 1

22 2 1 1 1

4 0 1 0 0

1

0

–

1

0

–

2

1

0

2 1 1

0 1 1

– 0 0

* WHO ⫽ World Health Organization; NK ⫽ natural killer; NHL ⫽ non-Hodgkin’s lymphoma. † Insufficient lymphoma material for diagnosis according to WHO classification.

1546

BAECKLUND ET AL

Table 4. Rheumatoid arthritis (RA) disease activity, lymphoma subtype, and EBV in 35 RA patients with lymphoma* RA disease activity/ no. (%) of patients High/19 (54)

Medium/14 (40)

Low/2 (6)

Lymphomas, no./subtype 12/diffuse large B cell lymphoma 1/follicular lymphoma 1/unspecified high-grade B cell lymphoma 1/unspecified B cell lymphoma 1/unspecified high-grade NHL 1/unspecified low-grade NHL 1/unspecified NHL 1/Hodgkin’s lymphoma 8/diffuse large B cell lymphoma 1/follicular lymphoma 1/lymphoplasmacytic lymphoma 1/Burkitt’s lymphoma 1/hepatosplenic T cell lymphoma 1/unspecified high-grade NHL 1/Hodgkin’s lymphoma 2/diffuse large B cell lymphoma

No. of EBV-positive lymphomas† 4

1

* See Table 3 for other definitions. † Thirty lymphomas were examined for EBV.

occurred in patients with high inflammatory activity (Table 4). In addition, these 4 patients had been treated with DMARDs (patient 1, sulfasalazine for 9 months; patient 2, intramuscular [IM] gold for 3 months and antimalarials for 2 years; patient 3, antimalarials for 10 months; and patient 4, antimalarials for 2 years, penicillamine for 1 year, and Prorecid [a drug containing podophyllotoxin which was used particularly in the Nordic countries during this time period] for 1 year until lymphoma diagnosis; in patients 1, 2, and 3, DMARD therapy preceded lymphoma diagnosis by 3, 10, and 15 years, respectively). At the time the lymphoma was diagnosed, a majority of the 35 patients with lymphoma (54%) had a disseminated disease with both nodal and extranodal involvement. According to the Ann Arbor classification system (34), 23 of the patients (66%) had generalized stage IV disease (Table 5). In the 7 patients with extranodal involvement only, the lymphoma was localized in the colon/rectum (2 patients), palate/jawbone (2 patients), and bone marrow, skin, and stomach (1 patient each). The extranodal sites in the patients with both nodal and extranodal involvement were the bone marrow and liver (6 patients each), the lung (5 patients),

the skin (4 patients), the spleen (3 patients), the kidney and bone (2 patients each), and the pancreas and pharynx/tongue (1 patient each). Ten of the patients had involvement of more than one extranodal site. The patient with a verified diagnosis of secondary Sjo ¨gren’s syndrome had a generalized DLBCL that did not affect the salivary glands or mucosal tissue. The median survival time of the 22 patients with DLBCL was short (6 months from diagnosis [range 2 weeks–13 years]) compared with the median survival time of the remaining 13 patients (27 months [range 3 weeks–13 years]). Of all 35 lymphoma patients, 17 had never received DMARD therapy. The DMARDs used in the treated patients included antimalarials (n ⫽ 13), D-penicillamine (n ⫽ 2), IM gold (n ⫽ 9), sulfasalazine (n ⫽ 1), and Prorecid (n ⫽ 2). (Note that one patient could have been treated with more than one drug; 5 patients had received 2 DMARDs, 2 patients had received 3 DMARDs, and 11 patients had received 1 DMARD each.) The duration of treatment with DMARDs was often short, and only 8 patients had been treated with a DMARD for ⱖ1 year during the entire RA disease period before the lymphoma diagnosis was made (median time of treatment 5 years, range 27 months–10 years). These 8 patients also had RA with high disease activity, and all continued to have active disease in spite of therapy. When we divided the patients into groups according to treatment with DMARDs (never treated [n ⫽ 17], treated ⬍1 year [n ⫽ 10], and treated ⱖ1 year [n ⫽ 8]), the different lymphoma subtypes were distributed in a similar way into the different categories (Table 6). The 2 patients with follicular lymphoma had never received DMARD therapy. Of the 2 patients with HL, 1 was never treated with DMARDs and 1 had been treated with antimalarials for 10 years. The only patient with a T Table 5.

Characteristics of the 35 reviewed lymphomas Patients, no. (%)

Nodal involvement only Extranodal involvement only Both nodal and extranodal involvement Ann Arbor stage* I IE II IIE III IIIE IV

9 (26) 7 (20) 19 (54) 1 (3) 1 (3) 4 (11) 2 (6) 4 (11) 0 (0) 23 (66)

* Disease stage according to Ann Arbor classification system (34).

LYMPHOMA SUBTYPES IN RA PATIENTS

1547

Table 6. Disease-modifying antirheumatic drug (DMARD) treatment, lymphoma subtypes, rheumatoid arthritis (RA) disease activity, and EBV in 35 RA patients with lymphoma* DMARD therapy/ no. (%) of patients Never/17 (49)

⬍1 year/10 (29)‡

ⱖ1 year/8 (23)§

Lymphomas, no./subtype

RA disease activity

8/diffuse large B cell lymphoma

Low (n ⫽ 2), medium (n ⫽ 5), high (n ⫽ 1) Medium (n ⫽ 1), high (n ⫽ 1) Medium (n ⫽ 1), high (n ⫽ 1) High Medium High Medium Medium Medium (n ⫽ 3), high (n ⫽ 5) Medium

2/follicular lymphoma 2/unspecified high-grade NHL 1/unspecified B cell lymphoma 1/Burkitt’s lymphoma 1/unspecified low-grade NHL 1/lymphoplasmacytic lymphoma 1/Hodgkin’s lymphoma 8/diffuse large B cell lymphoma 1/hepatosplenic T cell lymphoma 1/unspecified NHL 6/diffuse large B cell lymphoma 1/high-grade B cell lymphoma 1/Hodgkin’s lymphoma

High High (n ⫽ 6) High High

No. of EBVpositive lymphomas†

1 2

2

* See Table 3 for other definitions. † Thirty lymphomas were examined for EBV. ‡ Includes antimalarials 3–11 months (4 patients), sulfasalazine 11 months (1 patient), intramuscular gold 1 month and antimalarials 10 months (1 patient), and intramuscular gold 1–4 months (4 patients). § Includes antimalarials 5–10 years (2 patients), antimalarials 2–5 years and intramuscular gold 3–5 months (4 patients), antimalarials 2 years, D-penicillamine 1 year, and Prorecid 1 year (1 patient), and antimalarials 1 year, D-penicillamine 2 years, and Prorecid 4 years (1 patient). The last 2 patients were taking Prorecid at lymphoma diagnosis.

cell lymphoma had been treated with antimalarials for 3 months. At the time of diagnosis of the lymphoma, only 2 patients were receiving DMARD therapy, and both were taking Prorecid. These patients had a diagnosis of DLBCL and RA with longstanding high disease activity. One of them was EBV positive. Treatment with Prorecid was stopped at diagnosis of lymphoma, and chemotherapy for the lymphoma was started immediately in both cases; it cannot be evaluated whether withdrawal of Prorecid led to any regression of the lymphoma. Oral steroids had been used by 10 patients (by 4 of these patients for ⬎1 year), and there was no obvious relationship between steroid treatment and lymphoma subtype. Most patients had been treated with nonsteroidal antiinflammatory drugs (n ⫽ 32) or aspirin (n ⫽ 32), often in combination and for periods exceeding 10 years. No correlation with a specific lymphoma subtype was evident. DISCUSSION The main finding of this study of lymphomas in RA patients who had not been treated with methotrexate and who had had no treatment, or only short-term

treatment, with other DMARDs was the striking predominance of one specific lymphoma subtype, DLBCL. In the general population, ⬃30–40% of NHLs are DLBCLs (35), compared with 67% in this study. When considering the increasing incidence of DLBCL with time (28,36), and therefore the lower background risk for DLBCL at the time of our study (1964–1984) (33), we estimate that our data may correspond to a 2–3-fold increase in DLBCL incidence in these patients. To our knowledge, a predominance of one lymphoma subtype, not associated with a specific DMARD treatment, has not previously been described in RA patients. Interestingly, in patients with primary Sjo ¨gren’s syndrome, the other rheumatic disease with a welldescribed increased risk of lymphoma (37), another lymphoma subtype predominates. In patients with Sjo ¨gren’s syndrome, extranodal marginal-zone B cell lymphoma of the mucosa-associated lymphoid tissue (MALT lymphoma) is the most common lymphoma, often located in salivary glands and associated with a good prognosis (38). None of the RA patients in our study, including the patient with secondary Sjo ¨gren’s syndrome, had a MALT lymphoma, and although 11 of the patients had mucosal involvement, none had lym-

1548

phoma localized in the salivary glands. This indicates different mechanisms underlying lymphoma development in RA and Sjo ¨gren’s syndrome patients. The main known risk factor for DLBCL is an underlying immunodeficiency, and in patients with primary immune disorders as well as in patients infected by human immunodeficiency virus and in posttransplant patients, the incidence of DLBCL is increased (39). It is therefore of particular interest that the RA patients in this study had not had strong exposure to immunosuppressive therapy. Instead, many of them had an untreated, longstanding, active and destructive RA disease associated with a persistent stimulation of the immune system rather than with a state of immunosuppression. An interesting possible explanation, and one that supports the notion of a role for inflammatory activity in the development of lymphomas, was found in a study by Wahl et al (40) and subsequently confirmed by many groups, including our own (41). In these studies, it has been shown that apart from being associated with a stimulation of certain immune functions, RA may also cause suppression of other functions. An impaired T cell function, as reported in patients with active RA in the studies by Wahl et al and others, could be relevant to an impairment of immune responses toward EBV as well as against emerging malignant B cells. The presence of EBV in the lymphomas is also one of the features of lymphomas in immunosuppressed patients (39). In our patients, EBV was detected by ISH in the lymphomas of 5 of 30 patients (17%). EBV was slightly more frequent in our patients than in the general population, but still not as frequent as in immunosuppressed patients; for example, in posttransplant patients, the majority (50–80%) of the lymphomas are EBV positive (28,39). In RA patients treated with methotrexate, studies have shown varying results. EBV has been detected in 16–44% of lymphomas (14,15,26), and, as in the present study, a role for EBV in lymphoma development cannot be ruled out. Four of the 5 EBV-positive cases occurred in patients with high inflammatory activity, which raises the question of whether there is an association between inflammation and EBV. No conclusions can be drawn from this low number of EBV-positive patients, and, as pointed out, these 4 patients had also been treated with DMARDs, the importance of which is impossible to estimate. Nonetheless, we believe that the immunosuppressive impact of the treatment is likely to have been small in the 3 patients treated with antimalarials, IM gold, or sulfasalazine for relatively short periods and years apart from the diagnosis of the lymphoma. For the

BAECKLUND ET AL

patient with ongoing DMARD therapy (Prorecid) at the time of the lymphoma diagnosis, a drug-induced immunosuppression is possible. The remaining EBV-positive patient with a lymphoplasmacytic lymphoma and medium inflammatory RA activity had never been treated with DMARDs. For the evaluation of a relationship between EBV infection and inflammatory activity, we suggest inclusion of information about RA disease severity and inflammatory activity, as well as about RA treatment and EBV findings, in future studies of lymphomas in RA patients. We used the Swedish Hospital Discharge Register and the Swedish Cancer Register to identify a population-based cohort of RA patients. The reporting to these nationwide registers is mandatory, and quality evaluations have confirmed high reporting rates of ⱖ95% (42). We therefore estimate that the present study covers virtually all patients discharged from a hospital with a diagnosis of RA and a subsequent lymphoma diagnosis in this region of Sweden. This approach to studying RA patients, who have been hospitalized at some point, may of course bring a selection bias toward more severe cases. To increase the identification of RA patients with variable indications for hospital care and not only those treated for severe RA, patients with RA as both the main and secondary diagnoses were included. Data from another study (43) indicate that this approach incorporates ⬃75% of all RA patients in a particular referral area. From the results of the present study, it is clear that the lymphoma cases represent a subgroup of RA patients with severe disease, however, and that this must be considered when addressing lymphomas among RA patients in general. In spite of the retrospective approach, medical records of all patients were obtained, and it was possible to compile data about treatment and disease characteristics. The fulfillment of the 1987 ACR criteria for RA could be investigated in all cases, and only 1 patient was excluded due to incorrect diagnosis. We faced more difficulties in finding all lymphoma tissues, and age had also influenced the quality of the lymphoma material in some cases. This led to the exclusion of 6 of the cases and to problems with applying the WHO classification in another 6 cases. Still, 35 of 41 cases (85%) could be reclassified. The excluded cases did not differ in clinical characteristics from those that were reviewed. The lymphomas in the excluded patients were highly malignant, and the patients had short durations of survival after diagnosis. We therefore have no reason to believe that inclusion of the missing cases would have altered the main results.

LYMPHOMA SUBTYPES IN RA PATIENTS

When discussing and comparing data about lymphoma subtypes, one must also be aware that the lymphoma classification systems have been modified several times during the past decades, and that the classification of lymphomas has been inconsistent between different countries. The worldwide consensus classification (the WHO classification of tumors of hematopoietic and lymphoid tissues, which was used in this study) was first introduced in 2001. The rather low concordance between the initial diagnosis and the WHO diagnosis after review is not surprising. Many of the lymphoma entities in the WHO classification were not recognized in earlier classification systems. In two smaller studies of case series from hospitals, the reported findings were similar to those in the present study. Thus, an increase in DLBCL was reported in the study by Kamel et al of lymphomas in 15 RA patients (11), and the recent study by Dawson et al in 19 RA patients with lymphoma revealed similar results (15). In both studies, most patients were treated with methotrexate or other immunosuppressive drugs, and the increased frequency of DLBCL as well as the high number of EBV-positive lymphomas were suspected to be related to the immunosuppressive therapy. Unfortunately, reports from both studies lack information about the severity of the RA disease. There is only one reasonably large study of arthritis-associated lymphomas in which lymphoma subtypes have been investigated, and that study, using a case–control approach, yielded somewhat different results. In that study, 42 patients with arthritis and lymphomas were identified from the general population (44), and occurrence of DLBCL was reported in 38% of the patients. No difference in lymphoma subtypes between patients with and those without RA was evident. However, the study had factors of uncertainty. Arthritis was self reported, with RA diagnosis confirmed by the treating physician in only 11 of 42 patients, and no data were available describing RA disease status. In conclusion, the present study provides data indicating that high and longstanding inflammatory activity in RA may by itself contribute to the development of lymphomas which are preferentially of the aggressive, diffuse, large B cell type with a poor prognosis. Since the time period of the study was chosen to encompass an era during which immunosuppressive therapy was rare, we cannot draw any conclusions about the impact of today’s more aggressive treatment in RA. However, we believe that data describing the lymphoma risk and lymphoma characteristics, in a situation with a limited influence of immunosuppressive drugs, should help us in the difficult

1549

task of determining the relative roles of disease activity and drug treatment in the present day, in which RA is treated aggressively with DMARDs as well as with new targeted therapies such as tumor necrosis factor– blocking agents. ACKNOWLEDGMENTS We wish to thank the pathology departments in Falun, ¨ stersund for kind cooperaGa¨vle, Va¨sterås, Sundsvall, and O tion, and Associate Professor R. A. Harris for linguistic advice.

REFERENCES 1. Prior P, Symmons D, Hawkins CF, Scott DL, Brown R. Cancer morbidity in rheumatoid arthritis. Ann Rheum Dis 1984;43: 128–31. 2. Hakulinen T, Isomaki H, Knekt P. Rheumatoid arthritis and cancer studies based on linking nationwide registries in Finland. Am J Med 1985;78 Suppl 1A:29–32. 3. Gridley G, McLaughlin JK, Ekbom A, Klareskog L, Adami HO, Hacker DG, et al. Incidence of cancer among patients with rheumatoid arthritis. J Natl Cancer Inst 1993;85:307–11. 4. Mellemkjaer L, Linet MS, Gridley G, Frisch M, Moller H, Olsen JH. Rheumatoid arthritis and cancer risk. Eur J Cancer 1996;32A: 1753–7. 5. Thomas E, Brewster DH, Black RJ, Macfarlane GJ. Risk of malignancy among patients with rheumatic conditions. Int J Cancer 2000;88:497–502. 6. Kinlen LJ. Incidence of cancer in rheumatoid arthritis and other disorders after immunosuppressive treatment. Am J Med 1985;78: 44–9. 7. Silman AJ, Petrie J, Hazleman B, Evans SJW. Lymphoproliferative cancer and other malignancy in patients with rheumatoid arthritis treated with azathioprine: a 20 year follow up study. Ann Rheum Dis 1988;47:988–92. 8. Matteson EL, Hickey AR, Maguire L, Tilson HH, Urowitz MB, the Rheumatoid Arthritis Azathioprine Registry Steering Committee. Occurrence of neoplasia in patients with rheumatoid arthritis enrolled in a DMARD registry. J Rheumatol 1991;18: 809–14. 9. Jones M, Symmons D, Finn J, Wolfe F. Does exposure to immunosuppressive therapy increase the 10 year malignancy and mortality risks in rheumatoid arthritis? A matched cohort study. Br J Rheumatol 1996;35:738–45. 10. Asten P, Barrett J, Symmons D. Risk of developing certain malignancies is related to duration of immunosuppressive drug exposure in patients with rheumatic diseases. J Rheumatol 1999; 26:1705–14. 11. Kamel OW, van de Rijn M, LeBrun DP, Weiss LM, Warnke RA, Dorfman RF. Lymphoid neoplasms in patients with rheumatoid arthritis and dermatomyositis: frequency of Epstein-Barr virus and other features associated with immunosuppression. Hum Pathol 1994;25:638–43. 12. Liote F, Pertuiset E, Cochand-Priollet B, D’Agay M-F, Dombret H, Nume´ric P, et al. Methotrexate related B lymphoproliferative disease in a patient with rheumatoid arthritis: role of Epstein-Barr virus infection. J Rheumatol 1995;22:1174–8. 13. Van de Rijn M, Cleary ML, Variakojis D, Warnke RA, Chang PP, Kamel OW. Epstein-Barr virus clonality in lymphomas occurring in patients with rheumatoid arthritis. Arthritis Rheum 1996;39: 638–42. 14. Salloum E, Cooper DL, Howe G, Lacy J, Tallini G, Crouch J, et al. Spontaneous regression of lymphoproliferative disorders in pa-

1550

15.

16. 17.

18. 19.

20.

21. 22. 23. 24.

25. 26.

27.

28. 29.

tients treated with methotrexate for rheumatoid arthritis and other rheumatic diseases. J Clin Oncol 1996;14:1943–9. Dawson TM, Starkebaum G, Wood BL, Willkens RF, Gown AM. Epstein-Barr virus, methotrexate, and lymphoma in patients with rheumatoid arthritis and primary Sjogren’s syndrome: case series. J Rheumatol 2001;28:47–53. Baecklund E, Ekbom A, Spare´n P, Feltelius N, Klareskog L. Disease activity and risk of lymphoma in patients with rheumatoid arthritis: nested case-control study. BMJ 1998;317:180–1. Wolfe F. Inflammatory activity, but not methotrexate or prednisone use predicts non-Hodgkins lymphoma in rheumatoid arthritis: a 25-year study of 1,767 RA patients [abstract]. Arthritis Rheum 1998;41 Suppl 9:S188. Ellman MH, Hurwitz H, Thomas C, Kozloof M. Lymphoma developing in a patient with rheumatoid arthritis taking low dose weekly methotrexate. J Rheumatol 1991;18:1741–3. Kingsmore SF, Hall BD, Allen NB, Rice JR, Caldwell DS. Association of methotrexate, rheumatoid arthritis and lymphoma: report of 2 cases and literature review. J Rheumatol 1992;19: 1462–5. Kamel OW, van de Rijn M, Weiss LM, Del Zoppo GJ, Hench PK, Robbins BA. Reversible lymphomas associated with Epstein-Barr virus occurring during methotrexate therapy for rheumatoid arthritis and dermatomyositis. N Engl J Med 1993;328:1317–21. Bachman TR, Sawitzke AD, Perkins SL, Ward JH, Cannon GW. Methotrexate-associated lymphoma in patients with rheumatoid arthritis: report of two cases. Arthritis Rheum 1996;39:325–9. Moder KG, Tefferi A, Cohen MD, Menke DM, Luthra HS. Hematologic malignancies and the use of methotrexate in rheumatoid arthritis: a retrospective study. Am J Med 1995;99:276–81. Usman AR, Yunus MB. Non-Hodgkin’s lymphoma in patients with rheumatoid arthritis treated with low dose methotrexate. J Rheumatol 1996;23:1095–7. Georgescu L, Quinn GC, Schwartzman S, Paget SA. Lymphoma in patients with rheumatoid arthritis: association with the disease state or methotrexate treatment. Semin Arthritis Rheum 1997;26: 794–804. Fam AG, Perez-Ordonez B, Imrie K. Primary cutaneous B cell lymphoma during methotrexate therapy for rheumatoid arthritis. J Rheumatol 2000;27:1546–9. Mariette X, Cazals-Hatem D, Warszawki J, Liote F, Balandraud N, Sibilia J, et al. Lymphomas in rheumatoid arthritis patients treated with methotrexate: a 3-year prospective study in France. Blood 2001;99:3909–15. Arnett FC, Edworthy SM, Bloch DA, McShane DJ, Fries JF, Cooper NS, et al. The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis. Arthritis Rheum 1988;31:315–24. World Health Organization. World Health Organization classification of tumours: pathology and genetics of tumours of haematopoietic and lymphoid tissues. Geneva: IARC Press; 2001. Chang KL, Chen Y-Y, Shibata D, Weiss LM. Description of an in situ hybridization methodology for detection of Epstein-Barr virus

BAECKLUND ET AL

30. 31.

32.

34. 34.

35.

36.

37.

38.

39. 40.

41.

42. 43.

44.

RNA in paraffin-embedded tissues, with a survey of normal and neoplastic tissues. Diagn Mol Pathol 1992;1:246–55. Willis RA. Pathology of tumours. London: Butterworth & Co.; 1948. Rappaport H. Tumors of the hematopoietic system. Atlas of tumor pathology, section III. Washington (DC): Armed Forces Institute of Pathology; 1996. p. 97–206. Ge´rard-Marchant RG, Hamlin I, Lennert K, Rilke F, Stansfeld AG, van Unnik JAM. Classification of non-Hodgkin’s lymphomas. Lancet 1974;2:406–8. Martinsson U, Glimelius B, Sundstrom C. Lymphoma incidence in a Swedish county during 1969-1987. Acta Oncol 1992;31:275–82. Carbone PP, Kaplan HS, Musshoff K, Smithers DW, Tubiana M. Report of the Committee on Hodgkin’s Disease Staging Classification. Cancer Res 1971;31:1860–1. A clinical evaluation of the International Lymphoma Study Group classification of non-Hodgkin’s lymphoma. The Non-Hodgkin’s Lymphoma Classification Project. Blood 1997;89:3909–18. Groves FD, Linet MS, Travis LB, Devesa SS. Cancer surveillance series: non-Hodgkin’s lymphoma incidence by histologic subtype in the United States from 1978 through 1995. J Natl Cancer Inst 2000;92:1240–51. Kassan SS, Thomas TL, Moutsopoulos HM, Hoover R, Kimberly RP, Budman DR, et al. Increased risk of lymphoma in sicca syndrome. Ann Intern Med 1978;89:888–92. Voulgarelis M, Dafni UG, Isenberg DA, Moutsopoulos HM, the Members of the European Concerted Action on Sjo ¨gren’s Syndrome. Malignant lymphoma in primary Sjo ¨gren’s syndrome: a multicenter, retrospective, clinical study by the European Concerted Action on Sjo ¨gren’s Syndrome. Arthritis Rheum 1999;42: 1765–72. Knowles DM. Immunodeficiency-associated lymphoproliferative disorders. Mod Pathol 1999;12:200–17. Wahl SM, Wilder RL, Katona IM, Wahl LM, Allen JB, Scher I, et al. Leukapheresis in rheumatoid arthritis: association of clinical improvement with reversal of anergy. Arthritis Rheum 1983;26: 1076–84. Berg L, Lampa J, Rogberg S, van Vollenhoven R, Klareskog L. Increased peripheral T cell reactivity to microbial antigens and collagen type II in rheumatoid arthritis after treatment with soluble TNF␣ receptors. Ann Rheum Dis 2001;60:133–9. Mattsson B, Wallgren A. Completeness of the Swedish cancer register: non-notified cancer cases recorded on death certificates in 1978. Acta Radiol Oncol 1984;23:305–13. Allebeck P, Ljungstrom K, Allander E. Increased mortality among persons with rheumatoid arthritis, but where RA does not appear on death certificate: eleven-year follow-up of an epidemiological study. Scand J Rheumatol 1981;10:301–6. Kamel OW, Holly EA, van de Rijn M, Lele C, Sah A. A population based, case control study of non-Hodgkin’s lymphoma in patients with rheumatoid arthritis. J Rheumatol 1999;26: 1676–80.