SEARCH

SEARCH BY CITATION

Keywords:

  • Cohort;
  • lymphoma;
  • malignancy;
  • mycophenolate mofetil;
  • renal transplant

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

This large prospectively conducted observational cohort study examined the risk of lymphoma and other malignancies with mycophenolate mofetil (MMF) in de novo renal transplant recipients. A total of 6751 patients receiving MMF, and an equal number of matched controls receiving non-MMF-based immunosuppression, were identified from two large registries (Organ Procurement and Transplantation Network/United Network for Organ Sharing (OPTN/UNOS) and Collaborative Transplant Study (CTS)) and followed for 3 years. The primary endpoint was development of lymphoma. Secondary endpoints included development of any malignancy. There was no evidence of any increased risk of developing lymphoma or malignancy with MMF. The risk of developing lymphoma with MMF compared with the non-MMF cohort was not higher in either the CTS registry (relative risk (95% confidence interval); 0.4 (0.17–0.94)) or the OPTN/UNOS registry (1.04 (0.61–1.78)). In the MMF group, there was a trend toward a lower risk of malignancy in both registries (OPTN/UNOS 0.86 (0.69–1.07); CTS 0.79 (0.61–1.02)) and a significant increase in time to malignancy in the CTS dataset (p < 0.026). This study has demonstrated that MMF is not associated with an increased risk of lymphoma or other malignancies post-renal transplant, and may even be associated with a lower risk in some populations.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

The association between cancer and immunosuppression in renal transplant patients is well recognized (1). The risk of malignancy in a post-transplant population has been reported to be 4–500 times higher than in the general population, depending on study methodology and patient selection, with the most common malignancies being post-transplant lymphoproliferative disease (PTLD) early after transplantation and skin cancer late after transplantation (2). Over a 10-year period, the risk of PTLD in renal transplant recipients was 11.8-fold higher than that in a matched non-transplanted population (3).

The risk of malignancy is known to be related to immunosuppressive dose and total immunosuppression, including duration (4). The immunosuppressive regimen appears to influence the incidence and latency of malignancies (5–8). Both cyclosporine regimens (9,10) and the use of anti-CD3 (OKT3) monoclonal antibodies (7,11) have been reported to increase the incidence of lymphomas and decrease the time to disease occurrence. Spontaneous remission after removal of immunosuppression has been reported, but the majority of patients who develop post-transplant lymphoma require aggressive therapy and the prognosis is generally poor (12,13).

Mycophenolate mofetil (MMF; CellCept®, F. Hoffmann-La Roche, Basel, Switzerland) has been shown to be effective and well tolerated and is now a standard immunosuppressant agent post-transplant (14). The incidence of malignancies in three pivotal trials, involving almost 1500 renal transplant patients followed for at least 1 year, was similar to the incidence reported in the literature for renal allograft recipients. However, there was a slight numerical, but non-statistically significant increase in the incidence of lymphoma/PTLD in the MMF treatment groups compared with the placebo and azathioprine groups. Lymphoma/PLTD was noted in 0.6% of patients receiving MMF 2 g per day, in 1.0% of patients receiving MMF 3 g per day, in 0% of patients receiving placebo and in 0.3% of patients receiving azathioprine (15–17). However, this finding was not investigated further.

This prospective observational cohort study examined the risk of lymphoma and other malignancies when using MMF compared with other treatments in de novo renal transplant recipients.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Patients and study design

This cohort observational study utilized two large registries to identify patients for prospective follow-up and as the means of obtaining outcome data. Data from the United States were contributed from 250 kidney transplant centers participating in the Organ Procurement and Transplantation Network/United Network for Organ Sharing registry (OPTN/UNOS). Data from several European countries and Canada were contributed from 190 kidney transplant centers that participate in the Collaborative Transplant Study (CTS) registry .

Data were obtained from first time renal transplant patients receiving MMF at the time of discharge from hospital as part of their initial post-transplant treatment (MMF group). Baseline patient data were obtained for patients actually transplanted between January 1, 1996, and December 31, 1998. Patients were entered into the study in the order in which their records arrived at the transplant registries. Patients not receiving MMF as part of their initial post-transplant immunosuppressive treatment were selected at random from the registry databases by registry staff (non-MMF group). Non-MMF patients were then matched 1:1 to MMF patients according to date of transplant (by 6-month windows), age (by 5-year windows) and geographic region. For geographic matching, North American patients were assigned to one of four approximately equal quadrants: north-western, south-western, south-eastern or north-eastern. CTS patients (Europe and Canada) were matched by country. Patient anonymity was maintained according to the United States Privacy Act (HIPAA 1996) and the European Directive on individual privacy.

Post-transplant management and immunosuppressive treatment were determined by the physicians' normal clinical practice. We received endpoint data for follow-up evaluations from the registries on contact dates, 6 and 12 months after the transplant date and, thereafter, at 12-month intervals. No study-specific investigations were required. The study was continued until the last patient enrolled reached the 3-year anniversary of his or her transplant.

The primary endpoint was development of PTLD over a 3 year post-transplant observation period. Secondary endpoints included development of any malignancy (except for squamous skin cancers/carcinomas in situ), graft failure or death. Malignancies were classified as lymphoma, skin cancer, Kaposi's sarcoma, leukemia, myeloma or other. Time to event analysis was performed to examine the time course of each endpoint.

Statistical methodology

The study was designed to have at least 80% power to detect a two-fold increase in the risk of lymphoma associated with MMF therapy at the 95% confidence level. This resulted in a target sample size of 2500 per treatment group, assuming an incidence of lymphoma of 1% in the unexposed treatment group.

The chi-squared test was used to compare the raw incidences of primary and secondary endpoint events in the MMF versus non-MMF cohorts. Time to event was examined using standard survival analysis techniques, with the Kaplan-Meier method used to calculate and plot cumulative survival probabilities. Groups were compared using the log-rank test. Cox regression analysis was also performed on the time to event data. A stepwise backward procedure was applied to determine simultaneously relevant (prognostic) factors among the variables (treatment group, demographic and disease factors). A multivariate model was then evaluated. The model always included the treatment groups as well as any factor that achieved a significance level less than or equal to 0.15. Relative risk was estimated using multivariate Cox regression analysis. Analyses based on data from follow-up visits were performed on the per protocol dataset, which excluded patients with no follow-up, those who died within 30 days of transplant and those with lymphoma diagnosed at the time of transplant.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Patients

A total of 13 502 patients were included from the OPTN/UNOS (n = 8246) and CTS (n = 5256) registries. In the two registries, 70% of patients were followed up for the entire 3-year period, representing more than 22 500 patient-years of follow-up. The proportions of patients lost to follow-up were 1% and 3% for the CTS and UNOS datasets, respectively. Mean follow-up and total person years of follow-up were similar in the MMF and non-MMF groups in both registries (OPTN/UNOS: mean follow-up 3.57 years (MMF) vs. 3.54 years (non-MMF) and person years 14 688 (MMF) vs. 14 583 (non-MMF); CTS: mean follow-up 3.19 years (MMF) vs. 3.54 years (non-MMF) and person years 7889 (MMF) vs. 8278 (non-MMF)).

There were no important differences between MMF and non-MMF groups with respect to demographics and baseline characteristics (Table 1). Overall, there were more males than females in both registries. Some differences in race were seen between the two registries, with a higher percentage of Caucasian patients in the CTS population (Table 1). In both registries, the most common cause of renal failure was glomerulonephritis. In the CTS registry, a greater proportion of donors were cadaveric (87%) compared with the OPTN/UNOS database (66%).

Table 1.  Demographic and baseline data
 OPTN/UNOSCTS
MMF (n = 4123)Non-MMF (n = 4123)MMF (n = 2628)Non-MMF (n = 2628)
  1. Totals may not equal 100% due to rounding.

  2. CMV = cytomegalovirus; CTS = Collaborative Transplant Study; MMF = mycophenolate mofetil; OPTN/UNOS = Organ Procurement and Transplantation Network/United Network for Organ Sharing.

Mean age (±SD)43.4 (±0.23)43.5 (±0.24)45.0 (±0.29)45.0 (±0.29)
Number (%) men2461 (60%)2524 (61%)1640 (62%)1633 (62%)
Race
 Caucasian2450 (60%)2698 (65%)2543 (97%)2490 (95%)
 Black991 (24%)820 (20%)12 (<1%)10 (<1%)
 Unknown/other682 (16%)605 (15%)73 (3%)128 (5%)
Primary cause of renal failure
 Diabetes mellitus917 (22%)955 (23%)313 (12%)267 (10%)
 Glomerulonephritis1291 (31%)1266 (31%)688 (26%)608 (23%)
 Hypertension /renal vascular811 (20%)717 (17%)46 (2%)41 (2%)
 Polycystic disease368 (9%)411 (10%)311 (12%)353 (13%)
 Pyelo-/interstitial nephritis233 (6%)253 (6%)196 (8%)210 (8%)
 Unknown330 (8%)328 (8%)65 (3%)90 (3%)
 Other173 (4%)193 (5%)1009 (38%)1059 (40%)
Cadaveric donor2756 (67%)2669 (65%)2291 (87%)2296 (87%)
Living related/unrelated donor1367 (33%)1454 (35%)336 (13%)332 (13%)
CMV status
 Recipient negative, donor negative734 (18%)792 (19%)422 (16%)460 (18%)
 Recipient positive2229 (54%)2224 (54%)1344 (51%)1393 (53%)
 Recipient negative, donor positive784 (19%)742 (18%)473 (18%)455 (17%)
 Unknown376 (9%)365 (9%)389 (15%)320 (12%)

Post-transplant lymphoproliferative disease

There was no evidence of any increased risk of developing PTLD with MMF from either registry (Table 2). The incidence of lymphoma was similar in the OPT/UNOS registry (0.7% for both MMF and non-MMF groups); in the CTS registry the incidence was 0.3% in the MMF group and 0.9% in the non-MMF group (p = 0.002, Table 2). Cox regression analysis of time to event showed that the overall risk of developing lymphoma was also no higher with MMF in either the OPT/UNOS or the CTS datasets (relative risks 1.0 and 0.4, respectively, Figure 1). In the OPTN/UNOS population, there was no difference between MMF and non-MMF groups regarding timing of occurrence of PTLD (Figure 2A), whereas in the CTS registry, this was significantly delayed in the MMF group relative to the non-MMF group (Figure 2B, p = 0.002).

Table 2.  Incidence of lymphoma, any malignancy, graft failure or death in mycophenolate mofetil versus nonmycophenolate mofetil cohorts (per protocol population)
 OPTN/UNOSCTS
MMF (n = 4118)Non-MMF (n = 4119)p (chi-square)MMF (n = 2628)Non-MMF (n = 2628)P (chi-square)
  1. Abbreviations are same as in Table 1.

Lymphoma, n (%)27 (0.7%)27 (0.7%)0.9997 (0.3%)24 (0.9%)0.002
Malignancy, n (%)146 (3.6%)176 (4.3%)0.088104 (4.0%)146 (5.6%)0.006
Graft failure, n (%)538 (13.1%)544 (13.2%)0.848291 (11.1%)302 (11.5%)0.631
Deaths, n (%)302 (7.3%)316 (7.7%)0.560201 (7.7%)239 (9.1%)0.058
Any event, n (%)882 (22.4%)908 (22.0%)0.490519 (19.8%)591 (24.5%)0.015
image

Figure 1. Relative risk (and 95% CI) of time to event developing lymphoma, any malignancy, graft failure or death in MMF versus non-MMF cohort (Cox regression). CTS = Collaborative Transplant Study; MMF = mycophenolate mofetil; UNOS = United Network for Organ Sharing.

Download figure to PowerPoint

image

Figure 2. Time to event curve for development of lymphoma. (A) log-rank = 0.982 (B) log-rank = 0.002. Abbreviations are same as in Figure 1.

Download figure to PowerPoint

Any malignancy

There was no evidence of an increased risk of developing any malignancy (including lymphoma) with MMF. On the contrary, the incidence of any malignancy was numerically lower in the MMF versus non-MMF group in both registries, with the difference being statistically significant in the CTS registry (Chi-square test, p = 0.006, Table 2). Cox regression analysis of time to event showed a trend in both registries toward a lower risk of developing any malignancy in the MMF versus non-MMF cohort (relative risk (95% CI); OPTN/UNOS 0.86 (0.69, 1.07); CTS 0.79 (0.61, 1.02), Figure 1).

Time to development of any malignancy was longer for the MMF versus non-MMF group in both the OPTN/UNOS and CTS registries. In the OPTN/UNOS registry, this reached borderline significance (p = 0.051, Figure 3A) but was significant in the CTS registry (p = 0.026, Figure 3B).

image

Figure 3. Time to event curve for development of any malignancy. (A) log-rank = 0.051 and (B) log-rank = 0.026. Abbreviations are same as in Figure 1.

Download figure to PowerPoint

Graft failure

The incidence of graft failure was similar for the two cohorts in both registries (ranging from 11.1%–13.2%, Table 2), with no significant difference in relative risk by Cox regression (Figure 1) or time to graft failure (log-rank: OPTN/UNOS 0.795; CTS 0.798).

Survival

There was no evidence of any increased risk of death with MMF versus non-MMF in either registry (Table 2). The incidence of death was slightly lower in the MMF versus non-MMF group in the OPTN/UNOS registry (7.3% vs. 7.7%) and 18% lower in the CTS registry (7.7% vs. 9.1%). There were no significant differences in the relative risk for death (Figure 1) or time to death (log-rank: OPTN/UNOS 0.544; CTS 0.252) for MMF and non-MMF groups in either registry.

Post-hoc sub-analysis of MMF-exposed versus MMF-unexposed patients

In order to take into account patients who started in the non-MMF cohort because they were not receiving MMF at baseline but were then switched to an MMF-containing regimen at some time during the observation period, a post-hoc sub-analysis was performed on the OPTN/UNOS dataset, from which sufficient treatment data were available. In this analysis, patients were first classified according to whether they had been exposed to MMF at all during the observation period (n = 5811) or had never been exposed (n = 2426) and the incidence of PTLD in the two groups then compared.

Results from this analysis further support those obtained in the primary analysis (although introduction of some bias cannot be fully excluded as in such post-hoc analyses). The incidence of lymphoma was lower in the MMF-exposed versus MMF-unexposed group (0.53% vs. 0.95%). Cox regression analysis of time to event showed a significantly lower risk of developing lymphoma for the MMF-exposed group. The same results were shown for any malignancy, with a numerically lower incidence in the MMF-exposed versus the MMF-unexposed group (3.4% vs. 4.9%) and a significantly lower relative risk as shown by Cox regression analysis (relative risk (95% CI); 0.73 (0.58–0.91)). Significantly fewer deaths occurred in the ‘ever’-exposed treatment group (6.7% vs. 9.4%). This effect was confirmed in the Cox regression analysis of time to event (relative risk (95% CI); 0.69 (0.58–0.81)).

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

This large cohort study provides evidence that de novo transplant patients treated with MMF from time of transplantation are not at increased risk of developing lymphoma or other malignancies compared with other immunosuppressive treatments.

In the OPTN/UNOS database, there was no significant difference in the risk of lymphoma, malignancy, graft failure or death for the MMF versus non-MMF cohort. However, in the post-hoc sub-analysis of patients, exposure to MMF was associated with a significantly lower risk of lymphoma, any malignancy and death. In the CTS registry, there was a significantly lower risk of developing lymphoma with MMF versus non-MMF, with a trend to a lower risk of malignancy and no significant difference in the risk of graft failure or death.

The finding of a significantly lower risk of lymphoma with MMF in the CTS but not the OPTN/UNOS registry is probably attributable to factors inherent in registry-based studies. In particular, standard clinical practice may well have differed in the two populations, and factors such as a greater use of anti-T-cell therapies in both MMF and non-MMF patients in the OPTN/UNOS registry are known to be associated with an increase in the risk of lymphoma (18). Differences in recorded information between the two registries (partially because the CTS registry is voluntary) may also have affected our results.

Any inequality between the registry populations in the proportion of the non-MMF group subsequently switching to MMF could also have influenced outcomes, by obfuscating potential differences between treatment arms. Indeed, this does appear to at least partially explain the numerically different outcomes, because our post-hoc analysis of the OPTN/UNOS data demonstrated a significantly lower risk for lymphoma in the MMF-exposed group. However, it should be noted that post-hoc assessments are less resistant to statistical bias than prospectively defined analyses.

More recently, OPTN/UNOS data have been analyzed for the association of the type of immunosuppression with PTLD, graft survival and patient survival. MMF discharge maintenance immunosuppression was associated with a significant reduced risk of PTLD and graft loss compared with azathioprine (relative risk PTLD (95% CI); 0.64 (0.46– 0.87); relative risk graft loss (95% CI); 0.85 (0.79–0.91)) (19). A previous analysis of data from the CTS registry showed that immunosuppression with tacrolimus increased the risk of malignant lymphomas approximately two-fold compared with cyclosporine, while MMF did not have any influence (3). The effect of immunosuppressant combinations was not assessed in the present study, thus the decrease in lymphomas observed in the MMF cohort may have been a product of the effects of cyclosporine versus those of tacrolimus.

Studies involving registry data are by their nature subject to certain limitations and constraints. For example, we had no control over the completeness or detail of the data recorded (e.g. clinical history, fine details of dosage). Study groups were matched as described, but patients providing data for analysis were not randomized. However, our methods allowed us to gain data from over 22 500 patient-years of follow-up, which would be difficult to achieve in a prospective randomized study.

That the patients in this study are representative of the renal transplant population is supported by the finding that the overall incidence of PTLD was in line with the 1% reported in renal transplant patients (18) as was the incidence of graft failure (20–24). The validity of our data is further supported by similar findings from the OPTN/UNOS database (19).

This large study based on non-overlapping data from two major transplant registries was novel in that it was designed to be prospective, using matched controls and uniform methods of data collection. Unlike most retrospective registry analyses, the cohorts for this study were selected prior to receipt of the endpoint data. The statistical power of a study with more than 13 000 cases and the choice of well-defined endpoints (malignancy, graft failure and death) clearly increase the reliability of the conclusions (25) despite the shortcomings of registry data.

In summary, this study demonstrates that the use of MMF post-renal transplant is not associated with an increased risk of lymphoma and may even be associated with a lower risk in some populations and a lower risk of malignancy.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

We would like to acknowledge the collaboration and commitment of all the investigators and their staff, without whom the present study would not have been possible.

This study was supported by F. Hoffmann–La Roche Ltd.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  • 1
    Hanto DW, Shelton MW, Simmons RL. Malignancies after organ transplantation. In: PaulLC, SolezK eds. Organ Transplantation: Long Term Results. New York : Marcel Dekker, 1992: 319351.
  • 2
    Vial T, Descotes J. Immunosuppressive drugs and cancer. Toxicology 2003; 185: 229240.
  • 3
    Opelz G, Dohler B. Lymphomas after solid organ transplantation: a collaborative transplant study report. Am J Transplant 2004; 4: 222230.
  • 4
    McCann J. Lowering immune suppression drugs post-transplant may cut cancer risk. J Natl Cancer Inst 2003; 95: 848849.
  • 5
    Diethelm AG, Deierhoi MH, Hudson SL et al. Progress in renal transplantation. A single center study of 3359 patients over 25 years. Ann Surg 1995; 221: 446457.
  • 6
    Kehinde EO, Petermann A, Morgan JD et al. Triple therapy and incidence of de novo cancer in renal transplant recipients. Br J Surg 1994; 81: 985986.
  • 7
    Opelz G, Henderson R. Incidence of non-Hodgkin lymphoma in kidney and heart transplant recipients. Lancet 1993; 342: 15141516.
  • 8
    Penn I. Cyclosporine and oncogenesis. Mt Sinai J Med 1987; 54: 460464.
  • 9
    Penn I. Neoplastic complications of transplantation. Semin Respir Infect 1993; 8: 233239.
  • 10
    Gruber SA, Gillingham K, Sothern RB, Stephanian E, Matas AJ, Dunn DL. De novo cancer in cyclosporine-treated and non-cyclosporine-treated adult primary renal allograft recipients. Clin Transplant 1994; 8: 388395.
  • 11
    Swinnen LJ, Costanzo-Nordin MR, Fisher SG et al. Increased incidence of lymphoproliferative disorder after immunosuppression with the monoclonal antibody OKT3 in cardiac-transplant receipts. N Engl J Med 1990; 323: 17231728.
  • 12
    Muti G, Cantoni S, Oreste P et al. Cooperative Study Group on PTLDs. Post-transplant lymphoproliferative disorders: improved outcome after clinico-pathologically tailored treatment. Haematologica 2002; 87: 6777.
  • 13
    Weissmann DJ, Ferry JA, Harris NL, Louis DN, Delmonico F, Spiro I. Posttransplantation lymphoproliferative disorders in solid organ recipients are predominantly aggressive tumors of host origin. Am J Clin Pathol 1995; 103: 748755.
  • 14
    Fulton B, Markham A. Mycophenolate mofetil. A review of its pharmacodynamic and pharmacokinetic properties and clinical efficacy in renal transplantation. Drugs 1996; 51: 278298.
  • 15
    European MMF Co-operative Study Group. Placebo-controlled study of MMF combined with cyclosporine and corticosteroids for prevention of acute rejection. Lancet 1995; 345: 13211325.
  • 16
    Sollinger HW for the US Renal Transplant Mycophenolate Mofetil Study Group. Mycophenolate mofetil for the prevention of acute rejection in primary cadaveric renal allograft recipients. Transplantation 1995; 60: 225232.
  • 17
    Tricontinental Mycophenolate Mofetil Study Group. A blinded, randomised clinical trial of mycophenolate mofetil for the prevention of acute rejection in cadaveric renal transplantation. Transplantation 1996; 61: 19291937.
  • 18
    Swinnen LJ. Diagnosis and treatment of transplant-related lymphoma. Ann Oncol 2000; 11: 4548.
  • 19
    Cherikh WS, Kauffman HM, McBride MA, Maghirang J, Swinnen LJ, Hanto DW. Association of the type of induction immunosuppression with posttransplant lymphoproliferative disorder, graft survival, and patient survival after primary kidney transplantation. Transplantation 2003; 76: 12891293.
  • 20
    Frei U, Brunkhorst R, Schindler R et al. Present status of kidney transplantation. Clin Nephrol 1992; 38: S46S52.
  • 21
    Albrechtsen D, Leivestad T, Fauchald P, Flatmark A, Sodal G, Thorsby E. Results of the National Kidney Transplantation Program in Norway. Clin Transpl 1992; 6: 207213.
  • 22
    Bentdal OH, Leivestad T, Fauchald P et al. The national kidney transplant program in Norway still results in unchanged waiting lists. Clin Transpl 1998; 12: 221228.
  • 23
    Candinas D, Schlumpf R, Rothlin MA, Binswanger U, Largiader F. Thirty-two years of renal transplantation in Zurich. Clin Transpl 1996; 10: 241247.
  • 24
    Graeme R. Transplantation. In: DisneyAPS ed. ANZDATA Registry Report 2000. Adelaide , South Australia : Australia and New Zealand Dialysis and Transplant Registry, 2000.
  • 25
    Ward RA, Brier ME. Retrospective analyses of large medical databases. What do they tell us? J Am Soc Nephrol 1999; 10: 429432.