Medical Oncology

, Volume 26, Issue 1, pp 1–9

Spontaneous remission of “methotrexate-associated lymphoproliferative disorders” after discontinuation of immunosuppressive treatment for autoimmune disease. Review of the literature

Authors

    • Hematology Section, Department of Pathology and HematologyBari University Medical School
  • Paola Curci
    • Hematology Section, Department of Pathology and HematologyBari University Medical School
  • Mario Delia
    • Hematology Section, Department of Pathology and HematologyBari University Medical School
  • Erminia Rinaldi
    • Hematology Section, Department of Pathology and HematologyBari University Medical School
  • Antonia Chiefa
    • Hematology Section, Department of Pathology and HematologyBari University Medical School
  • Giorgina Specchia
    • Hematology Section, Department of Pathology and HematologyBari University Medical School
  • Vincenzo Liso
    • Hematology Section, Department of Pathology and HematologyBari University Medical School
Original Paper

DOI: 10.1007/s12032-008-9069-8

Cite this article as:
Rizzi, R., Curci, P., Delia, M. et al. Med Oncol (2009) 26: 1. doi:10.1007/s12032-008-9069-8

Abstract

There are a number of intriguing reports of lymphoproliferative disorders (LPDs) diagnosed during immunosuppressive treatment for underlying autoimmune disease, and spontaneously abated shortly after treatment discontinuation. Such LPDs, completely or partially regressing, occur in the clinical setting of “Methotrexate (MTX)-associated LPDs”, recognized by the World Health Organization (WHO) among the “Immunodeficiency-associated LPDs”. We identified 26 literature patients achieving spontaneous complete remission (CR) of their LPD, and eight others showing partial remission (PR). Most of them were affected by rheumatoid arthritis, received low-dose and long-term pulsed MTX alone or combined with other immunosuppressants, and developed a lymphoma. By reviewing the patients achieving CR, the following can be drawn: the absence of a unique type of LPD, the occurrence of an increased incidence of diffuse large B cell lymphoma as well as of frequent extranodal involvement, and EBV-infection. Further, CR mostly occurred within 4 weeks after discontinuation of immunosuppressant, and appeared to be persistent overtime. Conversely in the patients experiencing PR, the interval between discontinuation of immunosuppressive treatment and clinical response was mostly reported as longer than 4 weeks; moreover, in many cases the persistence of LPD or its progression induced to start cytotoxic therapy. Increased awareness is needed on the possible occurrence of LPD spontaneous remission following immunosuppressant discontinuation, after that it is therefore advisable to have a careful monitoring of the patient for some weeks, before starting cytotoxic therapy.

Keywords

Autoimmune diseaseImmunosuppressantMethotrexateLymphomaSpontaneous remission

Introduction

There are a number of intriguing case reports and case series of lymphomas and other lymphoproliferative disorders (LPDs), appearing during immunosuppressive treatment for underlying autoimmune disease and spontaneously regressing shortly after treatment discontinuation. Such LPDs occur in the clinical setting of “Methotrexate (MTX)-associated LPDs”, recognized by the World Health Organization (WHO) among the “Immunodeficiency-associated LPDs” [1].

By a computer-assisted search of the published literature (PubMed database to May 2007), we identified 26 patients achieving spontaneous complete remission (CR) of their LPD [224], and eight others showing partial remission (PR) [11, 15, 2528]. Most of them were affected by rheumatoid arthritis (RA), had received low-dose and long-term pulsed MTX alone or in combination with other immunosuppressants, and developed a lymphoma.

Patients

The main characteristics and outcome of the patients achieving CR are shown in Table 1, and those of the patients achieving PR in Table 2; the emerging data are summarized as follows.
Table 1

CR patients

Patient

Autoimmune disease

IS treatment

LPD

EBV

CR interval (between IS stop and LPD CR)

Follow-up (months)

Therapy after IS stop

Reference

N

Age/sex

Type

Duration (years)

Type

Dose (mg)

Duration (months)

Classification

Phenotype

Stage

Mono-clonality

1

83/F

RA

23

MTX

10–15/w

84

NHL, “small cleaved cell”/T

CD3+

CD4+

CD5+

HLA−

DR+

IV

(−)

(−)

4 w

24

Steroids

[2, 3]

2

86/F

RA

18

MTX

10–7.5/w

36

NHL, diffuse large/B

CD20+

CD30+

IV

NR

(+)

17 d

16

No therapy

[4]

3

65/F

DM

1

MTX

7.5/w

12

HL, nodular sclerosing

CD15+

CD30+

IV

NR

(+)

4 w

12

Steroids

[4]

4

33/M

CD

13

AZA

100/d

59

NHL, polymorphic/B

NR

IV

(+)

(+)

4 w

12

Mesalamine 4 g/d

[5]

5

57/F

RA

24

MTX

10–15/w

60

NHL, diffuse large/B

CD20+

CD3−

CD45RO−

IIIs

(−)

(+)

4 w

48

NR

[6]

6

56/M

RA

8

MTX

15/w

72

NHL, lymphoplasmacytic/B

CD19+

CD20+

E

(−)

(−)

3 w

16

NR

[7]

7

NR/NR

RA

NR

MTX

15/w

84

NHL, diffuse large/B

NR

NR

NR

(−)

NR

60

NR

[8]

8

65/M

RA

25

CTX

NR

48

NHL, anaplastic large/B

CD20+

CD30+

I

(+)

(+)

6 w

20

Steroids NSAID

[9]

   

AZA

NR

18

         
   

MTX

7.5–15/w

97

         
   

aCD5

0.33/Kg

5(d)

         

9

66/F

RA

21

MTX

7.5–15/w

84

NHL, diffuse large/B

CD20+

CD45+

CD43+

IV

(+)

(+)

4 w

15

Sulfasalazine NSAID

[9]

10

60/F

RA

5

MTX

10/w

9

NHL, immunoblastic with plasmacytic differentiation/B

CD20+

CD45RO+ k+

III

(−)

(+)

12 d

11

Steroids

[10]

11

40/F

LV

16

MTX

50/w

10

NHL, lymphoplasmacytic/B

NR

IV

(+)

(+)

NR

48

NR

[11]

12

58/M

Per. Seroneg A

5

MTX

10/w

48

NHL, lymphocytic/B

CD20+

CD22+

IgM+

E

(+)

(−)

3 w

12

NR

[12]

13

63/F

RA

18

MTX

7.5/w

60

LPD, resembling HL

CD15−

CD30+

CD20−

CD45RB−

III

NR

(−)

4 w

9

No therapy

[13]

14

48/F

RA

NR

MTX

7.5–15/w

96

LPD, resembling HL

CD15−

CD30+

CD20−

CD45RB−

E

NR

(+)

2 w

NR

Steroids

[13]

15

64/F

RA

33

MTX

7.5/w

24

NHL, anaplastic large/T

CD30+

CD3+

III

NR

(+)

4 w

48

Steroids

[14]

16

58/F

RA–SS

3–5

MTX

1.25–7.5/w

7

NHL, angioimmunoblastic/T

CD3+

CD4+

CD8+

III

(−)

(−)

1 w

36

NR

[15]

17

58/F

RA

20

MTX

7.5–10/w

78

NHL, diffuse large/B

CD20+

CD45RB+

CD30+

BCL2−

E

(+)

(−)

8 w

NR

NR

[16]

18

70/M

RA

0.6

MTX

10–15/w

6

NHL, “lymphocytic”/T

CD3+

CD4+

E

(+)

(+)

15 d

8

Steroids

[17]

19

47/F

DM

2.9

MTX

7.5–10/w

60

NHL, polymorphous centroblastic/B

CD3−

CD4−

CD8−

CD56−

E

NR

(+)

15 d

4

NR

[17]

20

46/F

RA

11

MTX

7.5–10/w

11

NHL, diffuse large/B

NR

III

NR

(+)

NR

30

Steroids

[18]

   

CyA

250/d

NR

         

21

69/F

RA

29

MTX

15/w

43

NHL, nodal marginal zone/B

NR

III

NR

NR

4 w

14

NSAID

[19]

22

63/M

RA

10

MTX

20/w

NR

NHL, diffuse large/B

CD20+

CD79+

E

NR

(−)

19 d

15

Sulfasalazine hydroxichloroquine

[20]

   

CyA

250/d

NR

         

23

52/M

Psoriasis vulg.

NR

MTX

NR

6

NHL, polymorphic/B

CD79a+

CD20+

CD30+

IIIs

(+)

(+)

4 w

30

Topical treatment

[21]

   

CyA

3/Kg/d

180

         

24

64/F

RA

8

MTX

15/w

87

NHL, extranodal marginal zone/B

NR

E

NR

NR

12 w

6

NR

[22]

   

MTX

plus

           
   

Inflix

5100 td

26

         

25

48/F

RA

NR

MTX

10/w

120

HL, NR variety

CD15+

CD30+

IV

NR

(−)

4 w

12

NR

[23]

26

67/F

RA–SS–HT

18

MTX

75/W

156

NHL, diffuse large/B

CD20+

CD79+

E

NR

(−)

4 w

12

NR

[24]

LPD, lymphoproliferative disorder; CR, complete remission; IS, immunosuppressant; RA, rheumatoid arthritis; DM, dermatomiositis; CD, Crohn’s disease; LV, leukocytoclastic vasculitis; Per seroneg A, peripheral seronegative arthritis; SS, Sjogren’s syndrome; Psoriasis vulg., Psoriasis vulgaris; HT, Hashimoto thyroiditis; td, total dose; EBV, Epstein-Barr virus; MTX, methotrexate; AZA, azathioprine; CTX, cyclophosphamide; CyA, cyclosporine A; aCD5, anti-CD5 monoclonal antibody; Inflix, Infliximab; NSAID, non-steroid anti-inflammatory drug; y, year; m, month; w, week; d, day; NR, not reported

Table 2

PR patients

Patient

Autoimmune disease

IS treatment

LPD

EBV

PR interval (between IS stop and LPD PR) (weeks)

Autoimmune disease therapy after IS stop

Reference

N

Age/sex

Type

Duration (years)

Type

Dose (mg)

Duration (months)

Classification

Phenotype

Stage

Mono-clonality

27

61/M

RA

3

AZA

100/d

12

HL, mixed cellularity

NR

I

NR

(+)

8

NR

[25]

   

MTX

10/w

8

        
   

MTX

plus

 

3

        
   

CyA

3.5/Kg/d

         

28

66/M

RA

22

MTX

7.5–10/w

144

NHL, follicular mixed/B

NR

III

(+)

(+)

8

MTX restart

[11]

29

70/F

RA

13

MTX

12.5–15/w

7

HL, mixed cellularity

NR

II

NR

(+)

NR

NR

[11]

30

62/F

RA

15

MTX

5/w

60

NHL, diffuse mixed/B

NR

IV

(+)

(+)

9

NR

[11]

31

6/F

JRA

NR

MTX

10–15/w

16

HL, mixed cellularity

CD15+

CD30+

IIIe

NR

(+)

NR

NR

[26]

32

6/F

JRA

NR

MTX

10–7.5/w

23

HL, nodular sclerosing

NR

III

NR

(−)

12

NR

[27]

33

69/F

RA

5

MTX

75/w

36

NHL, diffuse large/B

CD20+

CD45RO−

I

NR

(−)

3

NR

[15]

34

61/F

RA

24

MTX

5–15/w

60

HL, nodular sclerosing

CD15+

CD30+

II

NR

(+)

12

Hydroxichloroquine, steroids

[28]

LPD, lymphoproliferative disorder; PR, partial remission; IS, immunosuppressant; RA, rheumatoid arthritis; JRA, juvenile rheumatoid arthrtis; SScl, systemic sclerosis; EBV, Epstein-Barr virus; NR, not reported; MTX, methotrexate; AZA, azathioprine; CyA, cyclosporine A; y, year; m, month; w, week; d, day

CR patients

The patients achieving CR had a mean age of 57 years (ranging from 33 to 86), with a sex distribution of 18 females and seven males; age and sex were not reported for one patient [8]. With regard to autoimmune disease, 21 patients (80%) had RA [2, 4, 610, 1214, 1620, 22, 23] that was concomitant with Sjögren’s syndrome (SS) in one of them [15], and with SS and chronic autoimmune (Hashimoto’s) thyroiditis in another [24]; among the other patients, two had dermatomyositis (DM) [4, 17], one had Cröhn’s disease (CD) [5], one had leukocytoclastic vasculitis (LV) [11], and one had Psoriasisvulgaris [21]. In most of the patients (20/26), MTX alone was given [2, 4, 617, 19, 23, 24], whereas it was administered in combination with cyclosporine-A (CyA) to two patients [18, 20] or with infliximab (15 infusions) to another [22]. MTX was followed by cyclophosphamide (CTX), AZA, and anti-CD5 monoclonal antibody (Mab) in one patient [9], and by CyA in another [21]. AZA alone was given to one patient [5].

In regard to LPDs, there was a predominance (22/26) of non-Hodgkin lymphoma (NHL) of different subtypes; in particular, a diffuse large B cell lymphoma (DLBCL) was diagnosed in eight patients [4, 6, 8, 9, 16, 18, 20, 24]; in two a marginal zone lymphoma was found, nodal in one case [19] and extranodal in the other [22]. A lymphoplasmacytic/B cell NHL was observed in two patients [7, 11] and a polymorphic/B cell NHL in two others [5, 21]; further, an anaplastic large/B cell NHL [9], an immunoblastic with plasmacytic differentiation/B cell NHL [10], and a polymorphous centroblastic/B cell NHL [17] were described in single patients. Otherwise, the development of a T cell NHL was reported in four patients; among them, one was reported to have a “small cleaved cell” (CD3+, CD4+, CD5+, HLA-DR+) lymphoma with germline immunoglobulin heavy chain genes (JH) and T cell receptor β genes [2], another an anaplastic large cell lymphoma [14], a third one an angioimmunoblastic lymphoma [15], and the last one a “lymphocytic” (CD3+, CD4+) lymphoma restricted to the skin with some large cells (CD30+) and clonal T cells (T cell receptor γ chain locus study) [17]. Finally, there were four cases of Hodgkin’s lymphoma (HL); among these, one was of nodular sclerosis variety [4], two were described as LPD resembling HL [13], and the last one was of unspecified variety and diagnosed on bone marrow biopsy [23]. Regarding the immunophenotypic characteristics, the Reed-Sternberg cells in two patients were classically CD15+CD30+ [4, 23], whereas the large cells in the other two were CD15−CD30+ [13]. The search for clonal disease was performed in 13 patients [2, 57, 912, 1517, 21] with a positive response in seven of them [5, 9, 11, 12, 16, 17, 21].

EBV was demonstrated in lymphoma cells of 14 patients [46, 911, 13, 14, 17, 18, 21]; this search resulted negative in 10 [2, 7, 8, 12, 13, 15, 16, 20, 23, 24] and not reported in two others [19, 22]. CR was detected within 4 weeks after immunosuppressive treatment discontinuation, in all but three patients [9, 16, 22] in whom the time interval was 6, 8, and 12 weeks, respectively; it was instead not reported in three others [8, 11, 18]. The length of follow-up ranged from 4 [17] to 60 months [8], with a mean value of 21 months.

After immunosuppressant withdrawal, the management of autoimmune disease included low-dose steroids in six patients with RA [2, 10, 13, 14, 17, 18], and one patient with DM [4]. Mesalamine was given to one patient with CD [5], non-steroidal anti-inflammatory drugs (NSAIDs) combined with steroids or with sulfasalazine to two patients with RA [9], NSAIDs alone to one patient with RA [19], sulfasalazine and hydroxichloroquine to another patient with RA [20]. Lastly, only topical treatment was instituted in one patient with Psoriasis vulgaris [21]. The duration of immunosuppressive therapy before the diagnosis of LPD ranged from 6 to 180 months (mean 63.3). LPD was extranodal in eight patients [6, 12, 14, 1719, 22, 24].

PR patients

Among the eight literature patients achieving spontaneous PR of their LPD, six had underlying RA [11, 15, 25, 28] and two had juvenile RA [26, 27]; MTX alone was given to all but one patient, who received different immunosuppressants as well. With regard to LPD type, HL was diagnosed in five patients [11, 2528] and NHL in 3. EBV was detected in lymphoma cells of six patients [11, 25, 26, 28]. Cytotoxic therapy was instituted in all the patients, except for one [11]. In particular, Patient 27 underwent chemotherapy with mechlorethamine, vincristine, procarbazine, prednisone (MOPP) and local radiation therapy after a 50% PR of his EBV-positive mixed cellularity HL. It occurred 8 weeks after stopping combination regimen with MTX and CyA taken for the underlying RA; the previous treatment had been consisting of AZA and MTX taken sequentially [25]. Patient 28 achieved PR of a follicular mixed NHL, occurred within 2 months after MTX withdrawal. For subsequent RA flare, MTX was restarted and, 6 weeks later, the evidence of a liver bulky NHL induced to start chemotherapy with cyclophosphamide, doxorubicin, vincristine, prednisone (CHOP), then resulting in CR [11]. Patient 29, who showed spontaneous PR of a mixed-cellularity HL, was submitted neither to chemotherapy nor to radiation therapy and was reported as “alive with disease” [11]. Patient 30, who developed a diffuse mixed NHL while on MTX for RA, experienced a spontaneous minimal remission 9 weeks after the drug discontinuation. Four months later, she presented a progressive disease treated with chemotherapy (not reported); then, she was described as “alive with disease” [11]. Patient 31 developed an EBV-negative mixed cellularity HL during MTX treatment for juvenile RA; the discontinuation of MTX resulted in rapid reduction of the lymphadenopathies, and the subsequent administration of steroids (2 mg/kg/day) was followed by complete resolution of all the LPD lesions within 2 weeks. Four months later, while the patient was on steroid tapering, a new evidence of HL induced to start chemotherapy with MOPP and adriamycin, vinblastine, bleomycin, dacarbazine (ABVD), in response to which she achieved CR. This was followed, 18 months later, by a relapse treated with salvage chemotherapy and autologous peripheral stem cell transplantation [26]. In Patient 32 with juvenile RA, MTX treatment was at first discontinued for RA improvement, and then restarted for disease flare. Three months later, an EBV-negative nodular sclerosing HL developed, and MTX was withdrawn again. Within 3 weeks, the patient achieved spontaneous PR of her HL and underwent only one cycle of chemotherapy with MOPP-ABVD because of the occurrence of CR, still persistent 1 year later [27]. Patient 33, with RA, achieved spontaneous PR of an EBV-negative DLBCL 3 weeks after MTX discontinuation; 10 months later for a new evidence of LPD, chemotherapy with CHOP was instituted, followed by PR [15]. Patient 34, with RA and subsequent development of EBV-positive nodular sclerosing HL, experienced spontaneous PR of her lymphoma 3 months after stopping MTX; at 10-month follow-up, the documented worsening of HL induced to start chemotherapy with MOPP-ABVD and radiation therapy. After treatment, no findings of lymphoma were reported in the next 2 years [28].

Discussion

The published literature on LPD spontaneous CR or PR, occurring upon reversal of iatrogenic immunosuppression in autoimmune disease, is mainly consisting of anecdotal reporting, implying a large variability of data.

In the present review, we included those patients who received histological evaluation and adequate staging at diagnosis, although in some of them the descriptions of clinical and/or laboratory data appear somewhat incomplete or unsuitable. LPD diagnoses were performed between 1991 and 2006, and the diseases classified according to REAL [29] or WHO [30] classification, except for a few cases reported previously [2, 4]. As shown in Table 1, some discrepancies in the nomenclature and missing immunophenotypic features can be seen. It has to be highlighted that there was the absence of any distinctive feature, particularly of a unique type among the LPDs spontaneously regressed, as well as it is observed among all the “MTX-associated LPDs” [1].

From the present series of patients achieving CR, it can be, however, drawn an increased incidence of DLBCL (24%) as well as of frequent extranodal involvement (32%), and EBV infection (56%) (Table 1). CR occurred within a time interval of 4 weeks after discontinuation of immunosuppressant, except for three patients in whom it occurred later (Table 1) [9, 16, 22]. Moreover, CR appeared to be persistent overtime, as shown by the absence of any detectable recurrence over the follow-up period. The latter was, however, described until the time of the writing of each report, resulting in a mean length of 22 months with the longest period being 60 months [8].

Conversely, in the patients experiencing PR (Table 2), the time interval between discontinuation of immunosuppressive treatment and clinical response was mostly reported as longer than 4 weeks, often occurring about 2–3 months later. In 7/8 cases, the evidence of LPD or its progression induced to start treatment. Although the overall outcome in these patients is not always described [11, 25], it seems to overlap that of LPDs occurring in general population. The length of the follow-up, reported in four literature patients, is ranging from 12 to 24 months [15, 2628].

Since 1993, literature reports have linked MTX treatment for RA with the development of LPDs, primarily lymphomas, possibly regressing upon MTX discontinuation. The reversible LPDs were considered by various authors similar to those occurring in iatrogenic post-transplant immunosuppression as well as in congenital or acquired (HIV) immunodeficiency. In particular, the spontaneous remissions were often related to the evidence of EBV in malignant cells, and explained with a possible influence of MTX on EBV behavior [3133]. However, LPDs developing in patients with autoimmune disease on MTX treatment really lack a strong association with EBV infection.

The World Health Organization (WHO) classification of 2001 recognized as related to MTX-associated immunosuppression a clinical setting of LPDs, developing in patients on MTX for autoimmune disease. In particular, “MTX-associated LPDs” included changes ranging from lymphoid proliferation to lymphoma, often EBV-related and possibly regressing with MTX discontinuation. These LPDs more commonly included DLBCL, HL, or HL-like lymphoproliferations and, in smaller proportions, follicular lymphoma, Burkitt lymphoma, peripheral T cell lymphomas, polymorphous small lymphocytic, or lymphoplasmacytic proliferations. The immunophenotypic findings were overlapping those of LPDs sporadically occurring. The underlying autoimmune disease was very often RA, but in a minority of cases it was DM or psoriasis [1].

Although scarce with respect to the widespread use of MTX for autoimmune (rheumatic) diseases, the literature reports on spontaneous remission of LPDs upon MTX discontinuation strongly supported a causative relationship between the drug and LPD development. Of concern it is also the use of other therapeutic agents, including immunosuppressants such as CyA, AZA, and CTX [3437], that may lead to LPD by altering immune functions. The question of whether or not there is a causal link between TNF-α antagonists, recently introduced in the therapy of RA, and lymphoma risk has been debated [3840]. A patient included in the present series experienced a spontaneous CR of her LPD developed during treatment with MTX combined with infliximab, following MTX alone [22]. Over time, the scenario of the autoimmune diseases, reported as background in the association between immunosuppressive treatment and LPD spontaneously remitting, has become wider including CD, LV, and SS as well [41, 42]. However, RA disease activity seems to be associated with a major risk of LPD [43, 44].

Conclusion

To address the issue of LPD spontaneous regression following MTX withdrawal in individual cases, a driving mechanism could be hypothesized where chronic iatrogenic immunosuppression would lead to loss of immune surveillance, possibly restored on treatment discontinuation with subsequent control of the pathway of lymphomagenesis. It is, however, questionable to dissect the contribution of the immune system of autoimmune disease patients, by itself “set-up for failure” [44], from that of MTX or other therapeutic agents. The spontaneous CR of “MTX-associated LPDs” might be therefore attributable to an interplay involving autoimmune disease activity, iatrogenic immunosuppression, and lymphoproliferation clonality; in a proportion of cases, this combination of factors can also include the EBV-infection.

Furthermore, although “MTX-associated LPDs” share features with the LPDs observed in general population, their possible resolution upon reversal of iatrogenic immunosuppression provides a clue to their distinction, and induces a particular clinical approach. At the time of LPD diagnosis in a patient receiving immunosuppressive treatment for autoimmune disease, it is advisable to withdraw the treatment and perform a careful monitoring for some weeks. It has to be noted that spontaneous CR of LPD may occur and that this favorable outcome cannot be predictable so far.

Acknowledgment

We are indebted to Dr. Maria Palma for her thoughtful collaboration.

Copyright information

© Humana Press Inc. 2008