ORIGINAL ARTICLE
Prognosis of Patients with Ulcerative Colitis in Sustained Remission After Thiopurines Withdrawal Estefanía Moreno-Rincón, MD,* José Manuel Benítez, MD,* Francisco Javier Serrano-Ruiz, MD,* Juan María Vázquez-Morón, MD,† Héctor Pallarés-Manrique, MD, PhD,† José Manuel Herrera-Justiniano, MD,‡ Eduardo Leo-Carnerero, MD,‡ María Rosario Gómez-García, MD, PhD,§ María José Cabello-Tapia, MD, PhD,§ Manuel Castro-Fernández, MD, PhD,k María Rojas-Feria, MD, PhD,k Luisa Castro-Laria, MD,¶ Federico Argüelles-Arias, MD, PhD,¶ Raquel Camargo-Camero, MD, PhD,** Guillermo Alcaín-Martínez, MD, PhD,** Eva Iglesias-Flores, MD,* and Valle García-Sánchez, MD, PhD*
Background: The ideal length of treatment with thiopurines in patients with ulcerative colitis (UC) in sustained remission remains unknown. It is widely accepted that the drug withdrawal is associated with a worse outcome. The aim of this study was to analyze the outcome after this withdrawal and to identify predictors of relapse.
Methods: A multicenter and retrospective study was designed. A total of 102 patients with UC who discontinued thiopurines in a situation of sustained remission were included. All the patients were followed up until last revision or until relapse (understood as the occurrence of signs and symptoms of UC that required a rescue treatment). Results: After thiopurines withdrawal, overall relapse was recorded in 32.35% of the patients: 18.88% in the first year, 36.48% in the third, and 43.04%
in the fifth year after withdrawal. On multivariate analysis, predictors of relapse were the time from diagnosis of UC until the starting of thiopurines (hazard ratio [HR], 1.01; 95% confidence interval [CI], 1.01–1.02; P ¼ 0.039), the number of relapses before the withdrawal (HR, 1.3; 95% CI, 1.01–1.66; P ¼ 0.029), pancolitis (HR, 5.01; 95% CI, 1.95–26.43; P ¼ 0.028), the duration of treatment with thiopurines (HR, 0.15; 95% CI, 0.03–0.66; P ¼ 0.013) and the situation of biological remission at withdrawal (HR, 0.004; 95% CI, 0.0001–0.14; P ¼ 0.002).
Conclusions: The withdrawal of thiopurines in patients with UC, although in sustained remission, is related to a high relapse rate. Clinical variables such as the extent of the disease, the duration of treatment or time from diagnosis to the start of thiopurines should be considered before stopping these drugs. (Inflamm Bowel Dis 2015;21:1564–1571) Key Words: ulcerative colitis, thiopurines, withdrawal, relapse
T
he analogs of purines (azathioprine [AZA] and its metabolite, the mercaptopurine [MP]) have proven effective in maintaining remission in patients with ulcerative colitis (UC) as shown by
Received for publication January 26, 2015; Accepted February 17, 2015. From the *Department of Gastroenterology, IMIBIC (Maimónides Institute of Biomedical Research of Córdoba), University Hospital Reina Sofía, Córdoba, Spain; † Department of Gastroenterology, University Hospital Juan Ramón Jiménez, Huelva, Spain; ‡Department of Gastroenterology, University Hospital Virgen del Rocío, Sevilla, Spain; §Department of Gastroenterology, University Hospital Virgen de las Nieves, Granada, Spain; kDepartment of Gastroenterology, University Hospital Nuestra Señora de Valme, Sevilla, Spain; ¶Department of Gastroenterology, University Hospital Virgen Macarena, Sevilla, Spain; and **Department of Gastroenterology, University Hospital Virgen de la Victoria, Málaga, Spain. The authors have no conflicts of interest to disclose. All the authors are current members of the “Grupo Andaluz de Trabajo en Enfermedad Inflamatoria Intestinal.” Reprints: Estefanía Moreno-Rincón, MD, Department of Gastroenterology, University Hospital Reina Sofía, Avenida Menéndez Pidal s/n, 14004 Córdoba, Spain (e-mail: [email protected]). Copyright © 2015 Crohn’s & Colitis Foundation of America, Inc. DOI 10.1097/MIB.0000000000000400 Published online 28 April 2015.
1564
| www.ibdjournal.org
several studies, among them, various clinical trials,1–3 2 metaanalyses,4,5 and a Cochrane review.6 In this latest review, the global rate of failure in maintaining remission with and without this drug was of 44% and 65%, respectively, with a relative risk of 0.68 (95% confidence interval [CI], 0.54–0.86). This modest result is explained by poor methodological quality of some of the trials included. After publication of this study, a trial was carried out with this same purpose.3 Seventy-two patients with active UC and steroid dependence were included and randomized to receive AZA or mesalazine for 6 months. The steroid-free clinical and endoscopic remission was obtained in 53% of patients treated with AZA as compared with the 21% of those treated with mesalazine, with an odds ratio of 4.78 (95% CI, 1.57–14.5). The evidence supporting the use of thiopurines (TPs) for UC also comes from observational cohorts of retrospective series.7–11 In a 30-year review of the Oxford Clinic on the use of these drugs, the global remission in 346 patients with UC treated with AZA was 86% after 6 months. After 5 years of treatment, this percentage was of 62%, by applying a strict definition of relapse, and of 81% if a mild relapse was considered using a short course of steroids.12 Inflamm Bowel Dis Volume 21, Number 7, July 2015
Copyright © 2015 Crohn’s & Colitis Foundation of America, Inc. Unauthorized reproduction of this article is prohibited.
Inflamm Bowel Dis Volume 21, Number 7, July 2015
The safety profile of TP is well known. The cumulative incidence of adverse effects was of about 26%, and the annual risk was 7% per patient-year of treatment in the study by Chaparro et al,13 where the existence of adverse effects was analyzed on a cohort of 3931 patients treated with TP for an average period of 44 months. Most of these events happen at the beginning of treatment. After the first few weeks, the main risks of indefinite use of these drugs are the occurrence of myelotoxicity, liver toxicity, infections, and malignancies (lymphoproliferative disorders or nonmelanoma skin cancer). It was determined that the risk of infections and lymphoproliferative diseases14 is increased with age and in situations of associated comorbidity.15 For this reason, the discontinuation of treatment is suggested if the disease remains in remission for a long period.16,17 However, there is no overwhelming evidence for this. The risk of relapse in Crohn’s disease is about 35% to 77% in the first year and 65% to 75% after 5 years of drug discontinuation.18 The most important information obtained about the discontinuation of TPs in patients with UC in remission comes from a randomized controlled clinical trial19 and 4 retrospective studies,12,20–22 where there is a high percentage of significant clinical relapse (SCR) in patients who stopped the treatment with TP. Despite all this, there are no general recommendations yet that can help select by themselves the patients who would benefit from the discontinuation of TP without an increased relapse risk, so currently the choice is to individualize the risk on a patient-bypatient basis. So far, the results obtained in previous studies do not show strong conclusions, since usually; the potential statistically significant predictive factors that can be obtained have low–moderate clinical significance. Therefore, they are not recommended to be used as a basis for decision making. The aim of this study was to clarify whether it is possible to discontinue the TP therapy in patients with UC in sustained remission. To such effect, the outcome of patients after drug withdrawal will be evaluated for identifying relapse rates and predictive factors of a poor prognosis.
MATERIALS AND METHODS Design of the Study A retrospective, observational, and multicenter study was performed, where 7 hospitals of Andalucía (Spain) participated. A total of 102 patients with UC according to the international criteria23 were recruited consecutively according to the database and records of patients available on each center. The following variables were obtained:
Demographics: gender, age at the time of diagnosis, active smoking upon discontinuation (defined as 7 or more cigarettes per week). Disease-related: family history of inflammatory bowel disease, appendectomy, presence of extraintestinal manifestations, disease duration, disease extent (proctitis, leftsided colitis, or pancolitis), biological remission (defined
Patients with UC in Sustained Remission
as normal inflammatory parameters [white cell count level, 4000–10,000 units/mL; C-reactive protein [CRP], 0 to 5 mg/dL; erythrocyte sedimentation rate, 0–15 mm/sec; and platelets, 150,000–300,000 units/mL] upon discontinuation), biological parameters (hemoglobin, white cell count, platelets, CRP, erythrocyte sedimentation rate), endoscopic remission upon discontinuation (defined as 0 in Mayo score), amount of SCRs before the discontinuation of TP. Treatment-related: type of TP used and indication, time since diagnosis until introduction of TP, total time in steroidfree clinical remission, occurrence of adverse effects, need to reduce dosage of TP and reason, concomitant treatments, total duration of TP treatment, reason for drug discontinuation. Main variable-related: occurrence of SCR, time until SCR or until the last review.
Inclusion and Exclusion Criteria All patients had a previous diagnosis of UC and received maintenance treatment with TP (indicated due to steroid dependence or to maintain the remission induced by other drugs) for a period of at least 6 months. The drug was discontinued after at least 6 months of steroid-free clinical remission (defined as a normal/usual presence of stools with no blood or pus and no need of steroid). The cause for discontinuation was the occurrence of adverse events or the decision of the physician or the patient. A follow-up was carried out of all patients until the occurrence of SCR (defined as the occurrence of UC typical signs or symptoms requiring a rescue therapy such as oral or intravenous corticosteroids, biological therapy, immunosuppressant drugs, recapture with TP or surgery) or until the date of the last review in case of patients with no SCR. The following patients were excluded: patients who did not meet the above criteria, patients having an interruption of TP treatment with no clinical remission, and patients receiving a treatment combined with biological drugs and TP where the latter were discontinued in order to continue with the biological drug as monotherapy.
Statistical Analysis The statistical analysis of the obtained data was performed with GStat 2.0, with the statistics software SPSS 15.0 for Windows, and with Microsoft Excel 2007. First, a descriptive analysis of the variables was made. The continuous quantitative variables are represented using medians with interquartile ranges (IQRs) and categorical data with percentages. For comparison of hypothesis according to the main variable, the t tests of Student were used for quantitative variables and x2 test (or accurate Fisher’s test as necessary) for categorical variables. A univariant analysis was performed to determine the factors that predict relapse, and subsequently, a survival analysis with Kaplan–Meier curves, comparison of survival functions (log-rank test), and www.ibdjournal.org |
1565
Copyright © 2015 Crohn’s & Colitis Foundation of America, Inc. Unauthorized reproduction of this article is prohibited.
Inflamm Bowel Dis Volume 21, Number 7, July 2015
Moreno-Rincón et al
multivariant analysis of survival (COX regression) to determine the predictive factors on time to relapse.
Ethical Considerations The study obtained the approval of the Ethical Committee of Clinical Investigation (Comité Ético de Investigación Clínica [CEIC]) of Córdoba and the Spanish Agency of Drugs and Sanitary Products (Agencia Española de Medicamentos y Productos Sanitarios [AEMPS]).
RESULTS A total of 102 patients with diagnosis of UC and TP discontinuation according to the above criteria were recruited throughout the different participating centers. Table 1 includes the main clinical and demographic characteristics of the descriptive analysis and variables related to the TP therapy. The median time since diagnosis of the disease until the start of TP treatment was 32.5 months (IQR, 10–65), and the total median time of treatment with TP was 51 months (IQR, 21–90). In 34.31% of patients, there were adverse effects recorded, being the myelotoxicity the most frequently recorded (up to 15 patients) followed by infections (6 patients), digestive intolerance (5 patients), and liver toxicity (4 patients). No presence of lymphoma was observed in any patient. Twenty-six patients required a decrease of the TP dose. During TP treatment, 43.14% of patients received topical aminosalicylates (5-ASA), whereas 95.1% of patients received oral 5-ASA. After TP withdrawal, and as a maintenance treatment, 97% of patients were treated with oral or topical 5-ASA. The main variables defining the situation of the patient upon discontinuation of TP are shown in Table 1. An average SCR of 0.49 6 0.13 was observed before interruption of TP treatment. The median time of steroid-free remission before discontinuation was of 33 months (IQR, 14–70), and in 76.47% of patients, there was no endoscopic evaluation upon interruption of treatment. After TP discontinuation, a global SCR was observed in 32.35% of patients. The median time until relapse was of 12 months (IQR, 7–24), and the median follow-up time (until last review) in patients with no relapse was of 27 months (IQR, 9–75). The cumulative percentage of SCR was of 18.88% at 12 months, 36.48% at 36 months, and 43.04% after 5 years (Fig. 1). In terms of treatment of relapse, 24 patients (72.73%) required oral or intravenous corticosteroids. Fourteen patients (42.42%) with relapse restarted TP, with a median time since TP withdrawal until reintroduction of 15 months (IQR, 9–24). In 5 patients (15.15%), biological agents were used (1 patient adalimumab, 4 patients infliximab), with a median time until the use of this treatment of 9 months (IQR, 6–10). None of the patients required colectomy for treatment of relapse. In the univariant analysis (Table 2), different variables were compared in terms of the occurrence or not of SCR individually. It was found that the mean of time in steroid-free remission before
1566
TABLE 1. Clinical and Demographic Characteristics of the Population Variable Male gender, n (%) Current smoking, n (%) Appendectomy, n (%) Family history, n (%) Extraintestinal manifestations, n (%) Age at diagnosis, median (IQR), yr Disease duration, median (IQR), yr Disease extent, n (%) Proctitis Left-sided colitis Pancolitis
47 13 2 15 13 32 12
(46.08) (12.75) (1.96) (14.71) (12.75) (24–42) (8–16)
9 (8.82) 46 (45.1) 47 (46.08)
Main Characteristics of Thiopurine Treatment Variable Time from UC diagnosis until start of TP treatment, median (IQR), mo TP indication, n (%) Steroid dependence Maintenance of remission TP treatment duration, median (IQR), mo Drug-related adverse events, n (%) TP dose reduction, n (%) Reason for TP withdrawal, n (%) Adverse events Physician decision Patient decision
32.5 (10–65)
82 20 51 35 26
(80.39) (19.61) (21–90) (34.31) (25.29)
28 (27.25) 45 (44.12) 29 (28.43)
Disease Characteristics at the Time of Thiopurines Withdrawal Variable Time of steroid-free sustained remission 33 (14–70) before TP withdrawal, median (IQR), mo Biological remission, n (%) 98 (96.08) Blood parameters at the time of TP withdrawal, median (IQR) White cell count, units/mL 5250 (4400–6810) Hemoglobin, g/dL 13.1 (12.1–14.5) Platelets, units/mL 255,000 (199,000–294,000) CRP, mg/dL 1.45 (0.3–3.22) ESR, mm/sec 9 (5–18) Endoscopic remission at the time of TP withdrawal, n (%) Yes 22 (21.57) No 2 (1.96) Not evaluated 78 (76.47) ESR, erythrocyte sedimentation rate; TP, thiopurine.
| www.ibdjournal.org
Copyright © 2015 Crohn’s & Colitis Foundation of America, Inc. Unauthorized reproduction of this article is prohibited.
Inflamm Bowel Dis Volume 21, Number 7, July 2015
Patients with UC in Sustained Remission
FIGURE 1. Cumulative relapse rate after thiopurine withdrawal.
TABLE 2. Univariate Analysis: Factors Associated with Relapse After Thiopurines Withdrawal Variable Gender, n (%) Female Male (Ref.) Current smoking, n (%) Family history, n (%) Extraintestinal manifestations, n (%) Age at diagnosis, mean (6SD), yr Disease extent, n (%) Proctitis (Ref.) Left-sided colitis Pancolitis Disease duration, mean (6SD), yr Time of free-steroid sustained remission before TP withdrawal, mean (6SD), mo Biological remission, n (%) White cell count, mean (6SD), units/mL Hemoglobin, mean (6SD), g/dL CRP, mean (6SD), mg/dL ESR, mean (6SD), mm/sec Time from UC diagnosis until start of TP treatment, mean (6SD), mo TP treatment duration, mean (6SD), mo TP treatment duration recoded, n (%) #12 mo (Ref.) 13–48 mo .48 mo TP indication, n (%) Steroid dependence (Ref.) Maintenance of remission Number of relapses before TP withdrawal, mean (SD), mo
Relapse
21 12 3 5 4 30.12 2 15 16 12.41 34.09 31 5485.48 12.67 2.8 14.8 44.27 47.97
(38.18) (25.53) (23.08) (33.33) (30.77) (611.52) (22.22) (32.61) (34.04) (67.11) (635.81) (31.63) (61709.19) (61.91) (64.32) (616.8) (656.69) (651.73)
No Relapse
34 35 10 10 9 34.68 7 31 31 13.32 54.04 67 5710.92 13.54 3.56 13.89 49.99 66.86
P
(61.82) (74.47) (76.92) (66.67) (69.23) (615.65)
0.176 0.448 0.93 0.896 0.141
(77.78) (67.39) (65.96) (65.74) (642.36) (68.37) (61967.19) (61.59) (611.18) (612.01) (55.45) (647.05)
0.541 0.491 0.487 0.027 0.451 0.583 0.024 0.716 0.763 0.627 0.077
8 (47.06) 15 (45.45) 10 (19.23)
9 (52.94) 18 (54.55) 42 (80.77)
0.098 0.133
28 (34.15) 5 (25) 0.49 (61.06)
54 (65.85) 15 (75) 0.51 (61.45)
0.435 0.936
ESR, erythrocyte sedimentation rate; TP, thiopurine. P ,0.05 is considered statistically significant.
www.ibdjournal.org |
1567
Copyright © 2015 Crohn’s & Colitis Foundation of America, Inc. Unauthorized reproduction of this article is prohibited.
Inflamm Bowel Dis Volume 21, Number 7, July 2015
Moreno-Rincón et al
TP discontinuation was higher in the group with no relapse as compared with the group with relapse (54.04 versus 34.09 mo, respectively, P ¼ 0.027). There were differences found regarding the mean levels of hemoglobin at withdrawal, as it was significantly higher in the group with no relapse (13.54 versus 12.67 g/ dL, P ¼ 0.024). There was also a trend to statistic significance regarding duration of TP treatment because the mean of duration of treatment was higher in the group with no relapse as compared with the group with relapse (66.86 versus 47.97 mo, respectively, P ¼ 0.077). A recoding with the “total time of treatment with TP” variable was performed, so it was divided into fractions of #12 months, 13 to 48 months, and .48 months. Regarding this variable, there was no difference in the univariant analysis, but when comparing the hypothesis by stratifying by this new variable, there were statistically significant differences (Fig. 2), and the percentage of patients with relapse (19.23%) in the group of the patients treated .48 months with TP was significantly lower as compared with the rest of the other 2 groups (47.06% in the group of #12 mo and 45.45% in the group of 13–48 mo, P ¼ 0.009). A multivariant analysis of survival was performed (Cox regression) (Table 3) to identify the factors related to the presence of SCR in terms of progress. After ordered exclusion of nonsignificant variables in the model, the final variables with statistical significance were the extent of UC, the time since diagnosis of UC until beginning of TP therapy, the total duration of treatment with TP (recoded as indicated in Fig. 2), the number of clinical significant relapses before TP discontinuation, and the presence of biological remission at withdrawal. There were survival curves made by stratifying by each one of the significant variables of the multivariant model, obtaining statistical significance in the survival curves of total duration of treatment with TP (Fig. 3) (.48 versus #12 mo, P ¼ 0.008).
DISCUSSION After discontinuing the TP, a global SCR was recorded in 32.35% of patients. The cumulative percentage was 18.88% in the first year, 36.48% in the third year, and 43.04% after 5 years of
TABLE 3. Multivariate Analysis (Cox Regression) Variable Disease extent Proctitis (Ref.) Pancolitis Time from UC diagnosis until start of TP treatment, mo Number of relapses before TP withdrawal TP treatment duration recoded #12 mo (Ref.) .48 mo Biological remission
HR (95% CI)
P
5.01 (1.95–26.43) 1.01 (1.01–1.02)
0.028 0.039
1.3 (1.01–1.66)
0.029
0.15 (0.03–0.66) 0.013 0.004 (0.0001–0.14) 0.002
TP, thiopurine. P ,0.05 is considered statistically significant.
drug withdrawal. In the univariant analysis, factors such as steroid-free remission time before discontinuation, duration of treatment with TP, and hemoglobin level upon interruption acquired statistical significance. In the multivariant analysis, the independent factors related to the presence of SCR were a greater extent of the disease, a longer time from UC diagnosis to the start of TP treatment, and a higher number of relapses before discontinuation. The factors related to a lower incidence of SCR were a higher duration of treatment with TP and a biological remission at withdrawal. If we analyze the percentage of relapse annually, most of the previous studies show a higher percentage of around 35% to 77% as compared with 18.88% obtained in this study. Only the study by Kennedy et al22 shows similar rates, around 11% after 12 months. These differences probably are due to a longer follow-up period as compared with the first studies and longer previous duration of treatment with TP, which in our trial is of 60 months while in the rest of the studies is of 24 to 47 months. The differences regarding cumulative relapse percentages also appear when comparing rates after 5 years of discontinuation, which are of
FIGURE 2. Rate of relapse stratified by duration of thiopurine treatment.
1568
| www.ibdjournal.org
Copyright © 2015 Crohn’s & Colitis Foundation of America, Inc. Unauthorized reproduction of this article is prohibited.
Inflamm Bowel Dis Volume 21, Number 7, July 2015
Patients with UC in Sustained Remission
FIGURE 3. Survival analysis (log-rank) of relapse stratified by duration of thiopurine treatment.
65% to 75% in the previous studies and 43% in our cohort, and they might be explained also by the above factors. In 1992, Hawthorne et al19 carried out one of the first clinical trials in this field when, on a cohort of 79 patients with UC (67 in remission for at least 2 mo and 12 with steroid dependence), they performed randomization where a group of each arm received AZA and the other placebo. After 1 year of follow-up, the subgroup of patients in remission receiving AZA showed a relapse rate of 36%, as compared with the 59% receiving placebo. These differences were statistically significant, and the conclusion was that the long-term maintenance of AZA treatment in patients with UC in remission prevents the relapse of the disease. Fraser et al12 carried out a retrospective review in 222 patients (79 with Crohn’s disease and 143 with UC), to whom the TP treatment was discontinued after a remission period of more than 6 months. The relapse percentage was of 37% after a year, 66% after 3 years, and 75% after 5 years of follow-up, with no differences between both groups. In another retrospective study, Lobel et al20 analyzed the outcome of 61 patients with UC in steroid-free remission of more than 3 months in treatment with MP with a year of follow-up. All patients had been treated for at least 6 months. Thirty-nine patients continued with MP and 22 discontinued it. The median time until relapse in the first group was 58 weeks and in the second group 24 weeks (P , 0.05). The percentage of relapse after a year was 77% in the group that interrupted the treatment as compared with 43% of the group that continued it. In another retrospective multicenter study performed in 2009, Cassinotti et al21 recruited 127 patients with UC on treatment with TP, to whom this treatment was discontinued in a clinical remission situation, and they recorded the outcome of those patients over time. In their study, one-third of the patients showed clinical relapse after a year of follow-up, half of them after 2 years, and 75% of them after 5 years, while in our study, this information is 18.88%, 34.77% and 43.04%, respectively. It is difficult to
explain these differences from a purely objective point of view. An influencing factor may be the fact that in the study by Cassinotti et al, patients included had to be in a period of steroid-free remission for at least 3 months before the discontinuation. In our study, the period of sustained clinical remission was extended to a minimum of 6 months because it was considered that this would be the maximum period the TPs would need to show efficacy. In the study by Cassinotti et al, there is also a mention to the global colectomy rate, which was 10%. None of the patients of our cohort required colectomy as a rescue therapy for relapse. In the study performed by Kennedy et al,22 108 patients with UC were recruited retrospectively, to whom treatment with TPs was discontinued in a sustained remission situation, with an annual rate of relapse of 12%, which is slightly lower than the rate obtained in our study. This difference may probably be due to the strict inclusion criteria of the study by Kennedy et al, such as the need for a minimum period of 3 years of treatment with TP, which obviously could impact the results. Four studies have assessed the risk factors of relapse after TP discontinuation in patients with UC.12,19,21,22 Only the studies by Kennedy et al and Cassinotti et al identified a population at risk. In the multivariant analysis of the latter, the relapse risk was associated with the lack of sustained remission during the maintenance treatment with AZA (hazard ratio [HR], 2.35; 95% CI, 1.43–3.85; P ¼ 0.001), extensive colitis (HR, 1.79; 95% CI, 1.06– 3.02; P ¼ 0.028), and duration of treatment, so the shorter treatments (3–6 mo) involved a worse prognosis than the treatments that lasted more than 48 months (HR, 2.78; 95% CI, 1.27–6.11; P ¼ 0.008). Information reported in this study aims at the fact that it could be reasonable to maintain TP treatment for a minimum period of 4 years, a statement consistent with our results. The study by Kennedy et al provides relevant information regarding the biological parameters upon interruption, identifying factors such as white cell count level higher than 9100 units/mL www.ibdjournal.org |
1569
Copyright © 2015 Crohn’s & Colitis Foundation of America, Inc. Unauthorized reproduction of this article is prohibited.
Moreno-Rincón et al
(HR, 6.7; 95% CI, 1.86–24.2; P ¼ 0.004) and CRP level above 14 mg/dL (HR, 3.2; 95% CI, 1.48–7.05; P ¼ 0.003) as possible factors that predict relapse after TP discontinuation. Also, in this study, the response to the reintroduction of TP for treatment of relapse was analyzed; 71% of patients had a TP reintroduction, reaching high rates of response of up to 92%. In our cohort, the biological remission situation also reached statistical significance in the multivariant analysis, understanding this as normal biological parameters of inflammation upon discontinuation. The study by Louis et al24 is worth mentioning, which treated a cohort of 115 patients with Crohn’s disease in a sustained remission situation and with previous combined treatment (TP and infliximab), where the factors that predict relapse after infliximab discontinuation were analyzed. In that study, biological parameters such as high levels of white cell count, hemoglobin levels lower than 14.5 g/dL, CRP above 5 mg/ dL, and faecal calprotectin lower than 300 mg/g were factors that predicted a higher risk of relapse. An important factor to be considered is the evaluation of endoscopic remission at withdrawal, defined in our study as a 0 score of the endoscopic Mayo subscore. It is known that reaching the mucosal healing is one of the main therapeutic goals in these patients, having proven in various studies25,26 how its presence is associated with a better prognosis of the disease by reducing the occurrence of complications, specifically the risk of colectomy and colorectal cancer. Its evaluation, therefore, should be a factor to consider before determining the interruption of immunosuppressant medication. It is worth noting that new laboratory parameters have appeared throughout the last few years, such as measurement of faecal calprotectin27–29 and its use as a marker of endoscopic activity, offering valid alternatives to the endoscopic assessment, considering the inherent risks of this procedure. Therefore, it is evident that the use of TP in UC is widely accepted in terms of efficacy, being one of the basic therapeutic strategies in the management of this disease. The different studies performed throughout the years have tried to answer the repeated question of whether the decision to stop these therapies is suitable in patients with long periods of remission, given the noticeable rate of adverse effects associated. The results obtained in these studies have been varied but, as we conclude, the interruption of these therapies seems to be associated with a considerable risk of relapse we should not forget. In this therapeutic strategy, there are doubts before discontinuing the TP, such as minimum duration of treatment that is necessary to maintain, time of remission that should be kept before considering discontinuation, and tools we should use to assess this remission. The most suitable strategy is probably an evaluation that is not only clinical and biological but also endoscopic. The main limitation of our study lies in its retrospective nature. This affects the low percentage of endoscopic evaluation before the discontinuation of TP because a mucosal healing situation, as set forth in the above-mentioned studies, would likely involve a factor that predicts low risk of relapse after interruption
1570
Inflamm Bowel Dis Volume 21, Number 7, July 2015
of treatment. Another important limitation is the absence of measurement of faecal calprotectin, which was determined in the study by Kennedy et al, although not reaching the statistical significance, probably due to the low amount of recruited patients. As conclusion, we should indicate that TP discontinuation in patients with UC in sustained remission is associated with a relapse risk of 18.88% after the first year. Therefore, it is not recommended to interrupt the treatment systematically but to individualize according to the specific characteristics of each patient. Factors such as pancolitis, a longer period of time since diagnosis of UC until TP first administration, or a high number of clinical relapses before withdrawal should be considered before deciding on the discontinuation of these drugs. Further controlled, randomized, and prospective studies may provide more information on this topic.
REFERENCES
1. Jewell DP, Truelove SC. Azathioprine in ulcerative colitis: final report on controlled therapeutic trial. Br Med J. 1974;4:627–630. 2. Sood A, Midha V, Sood N, et al. Role of azathioprine in severe ulcerative colitis: one-year, placebo-controlled, randomized trial. Indian J Gastroenterol. 2000;19:14–16. 3. Ardizzone S, Maconi G, Russo A, et al. Randomised controlled trial of azathioprine and 5-aminosalicylic acid for treatment of steroid dependent ulcerative colitis. Gut. 2006;55:47–53. 4. Gisbert JP, Linares PM, McNicholl AG, et al. Meta-analysis: the efficacy of azathioprine and mercaptopurine in ulcerative colitis. Aliment Pharmacol Ther. 2009;30:126–137. 5. Khan KJ, Dubinsky MC, Ford AC, et al. Efficacy of immunosuppressive therapy for inflammatory bowel disease: a systematic review and metaanalysis. Am J Gastroenterol. 2011;106:630–642. 6. Timmer A, McDonald JWD, Tsoulis DJ, et al. Azathioprine and 6-mercaptopurine for maintenance of remission in ulcerative colitis. Cochrane Database Syst Rev. 2012;9:CD000478. 7. Sood A, Kaushal V, Midha V, et al. The beneficial effect of azathioprine on maintenance of remission in severe ulcerative colitis. J Gastroenterol. 2002;37:270–274. 8. Holtmann MH, Krummenauer F, Claas C, et al. Long-term effectiveness of azathioprine in IBD beyond 4 years: a European multicenter study in 1176 patients. Dig Dis Sci. 2006;51:1516–1524. 9. Sood A, Midha V, Sood N, et al. Long term results of use of azathioprine in patients with ulcerative colitis in India. World J Gastroenterol. 2006; 12:7332–7336. 10. Jharap B, Seinen ML, de Boer NK, et al. Thiopurine therapy in inflammatory bowel disease patients: analyses of two 8-year intercept cohorts. Inflamm Bowel Dis. 2010;16:1541–1549. 11. Chebli LA, Chaves LD, Pimentel FF, et al. Azathioprine maintains longterm steroid-free remission through 3 years in patients with steroiddependent ulcerative colitis. Inflamm Bowel Dis. 2010;16:613–619. 12. Fraser AG, Orchard TR, Jewell DP. The efficacy of azathioprine for the treatment of inflammatory bowel disease: a 30 year review. Gut. 2002;50: 485–489. 13. Chaparro M, Ordás I, Cabré E, et al. Safety of thiopurine therapy in inflammatory bowel disease: long-term follow-up study of 3931 patients. Inflamm Bowel Dis. 2013;19:1404–1410. 14. Lewis JD. Risk of lymphoma in patients with inflammatory bowel disease. Gastroenterol Hepatol. 2012;8:45–47. 15. Gisbert JP, Chaparro M. Systematic review with meta-analysis: inflammatory bowel disease in the elderly. Aliment Pharmacol Ther. 2014;39: 459–477. 16. Kabac¸am G, Törüner M. Can azathioprine/6-mercaptopurine treatment be withdrawn in patients with response in ulcerative colitis: what is the most appropriate time for this? Turk J Gastroenterol. 2012;23:17–20.
| www.ibdjournal.org
Copyright © 2015 Crohn’s & Colitis Foundation of America, Inc. Unauthorized reproduction of this article is prohibited.
Inflamm Bowel Dis Volume 21, Number 7, July 2015
17. Clarke K, Regueiro M. Stopping immunomodulators and biologics in inflammatory bowel disease patients in remission. Inflamm Bowel Dis. 2012;18:174–179. 18. Treton X, Bouhnik Y, Mary JY, et al. Azathioprine withdrawal in patients with Crohn’s disease maintained on prolonged remission: a high risk of relapse. Clin Gastroenterol Hepatol. 2009;7:80–85. 19. Hawthorne AB, Logan RFA, Hawkey CJ, et al. Randomised controlled trial of azathioprine withdrawal in ulcerative colitis. BMJ. 1992;305: 20–22. 20. Lobel EZ, Korelitz BI, Xuereb MA, et al. A search for the optimal duration of treatment with 6-mercaptopurine for ulcerative colitis. Am J Gastroenterol. 2004;99:462–465. 21. Cassinotti A, Actis GC, Duca P, et al. Maintenance treatment with azathioprine in ulcerative colitis: outcome and predictive factors after drug withdrawal. Am J Gastroenterol. 2009;104:2760–2767. 22. Kennedy NA, Kalla R, Warner B, et al. Thiopurine withdrawal during sustained clinical remission in inflammatory bowel disease: relapse and recapture rates, with predictive factors in 237 patients. Aliment Pharmacol Ther. 2014;40:1313–1323.
Patients with UC in Sustained Remission
23. Stange EF, Travis SP. The European consensus on ulcerative colitis: new horizons? Gut. 2008;57:1029–1031. 24. Louis JY, Vernier-Massouille G, Grimaud JC, et al. Maintenance of remission among patients with Crohn’s disease on antimetabolite therapy after infliximab therapy is stopped. Gastroenterology. 2012;142:63–70. 25. Rutgeerts P, Vermeire S, Van Assche G. Mucosal healing in inflammatory bowel disease: impossible ideal or therapeutic target? Gut. 2007;56: 453–455. 26. Dave M, Loftus E. Mucosal healing in inflammatory bowel disease—a true paradigm of success? Gastroenterol Hepatol. 2012;8:29–38. 27. Smith LA, Gaya DR. Utility of faecal calprotectin analysis in adult inflammatory bowel disease. World J Gastroenterol. 2012;18:6782–6789. 28. Moniuszko A, Wisniewska A, Rydzewska G. Biomarkers in management of inflammatory bowel disease. Prz Gastroenterol. 2013;8: 275–283. 29. Burri E, Beglinger C, von Felten S, et al. Fecal calprotectin and the clinical activity index are both useful to monitor medical treatment in patients with ulcerative colitis. Dig Dis Sci. 2015;60: 485–491.
www.ibdjournal.org |
1571
Copyright © 2015 Crohn’s & Colitis Foundation of America, Inc. Unauthorized reproduction of this article is prohibited.
Prognosis of Patients with Ulcerative Colitis in Sustained Remission After Thiopurines Withdrawal Estefanía Moreno-Rincón, MD,* José Manuel Benítez, MD,* Francisco Javier Serrano-Ruiz, MD,* Juan María Vázquez-Morón, MD,† Héctor Pallarés-Manrique, MD, PhD,† José Manuel Herrera-Justiniano, MD,‡ Eduardo Leo-Carnerero, MD,‡ María Rosario Gómez-García, MD, PhD,§ María José Cabello-Tapia, MD, PhD,§ Manuel Castro-Fernández, MD, PhD,k María Rojas-Feria, MD, PhD,k Luisa Castro-Laria, MD,¶ Federico Argüelles-Arias, MD, PhD,¶ Raquel Camargo-Camero, MD, PhD,** Guillermo Alcaín-Martínez, MD, PhD,** Eva Iglesias-Flores, MD,* and Valle García-Sánchez, MD, PhD*
Background: The ideal length of treatment with thiopurines in patients with ulcerative colitis (UC) in sustained remission remains unknown. It is widely accepted that the drug withdrawal is associated with a worse outcome. The aim of this study was to analyze the outcome after this withdrawal and to identify predictors of relapse.
Methods: A multicenter and retrospective study was designed. A total of 102 patients with UC who discontinued thiopurines in a situation of sustained remission were included. All the patients were followed up until last revision or until relapse (understood as the occurrence of signs and symptoms of UC that required a rescue treatment). Results: After thiopurines withdrawal, overall relapse was recorded in 32.35% of the patients: 18.88% in the first year, 36.48% in the third, and 43.04%
in the fifth year after withdrawal. On multivariate analysis, predictors of relapse were the time from diagnosis of UC until the starting of thiopurines (hazard ratio [HR], 1.01; 95% confidence interval [CI], 1.01–1.02; P ¼ 0.039), the number of relapses before the withdrawal (HR, 1.3; 95% CI, 1.01–1.66; P ¼ 0.029), pancolitis (HR, 5.01; 95% CI, 1.95–26.43; P ¼ 0.028), the duration of treatment with thiopurines (HR, 0.15; 95% CI, 0.03–0.66; P ¼ 0.013) and the situation of biological remission at withdrawal (HR, 0.004; 95% CI, 0.0001–0.14; P ¼ 0.002).
Conclusions: The withdrawal of thiopurines in patients with UC, although in sustained remission, is related to a high relapse rate. Clinical variables such as the extent of the disease, the duration of treatment or time from diagnosis to the start of thiopurines should be considered before stopping these drugs. (Inflamm Bowel Dis 2015;21:1564–1571) Key Words: ulcerative colitis, thiopurines, withdrawal, relapse
T
he analogs of purines (azathioprine [AZA] and its metabolite, the mercaptopurine [MP]) have proven effective in maintaining remission in patients with ulcerative colitis (UC) as shown by
Received for publication January 26, 2015; Accepted February 17, 2015. From the *Department of Gastroenterology, IMIBIC (Maimónides Institute of Biomedical Research of Córdoba), University Hospital Reina Sofía, Córdoba, Spain; † Department of Gastroenterology, University Hospital Juan Ramón Jiménez, Huelva, Spain; ‡Department of Gastroenterology, University Hospital Virgen del Rocío, Sevilla, Spain; §Department of Gastroenterology, University Hospital Virgen de las Nieves, Granada, Spain; kDepartment of Gastroenterology, University Hospital Nuestra Señora de Valme, Sevilla, Spain; ¶Department of Gastroenterology, University Hospital Virgen Macarena, Sevilla, Spain; and **Department of Gastroenterology, University Hospital Virgen de la Victoria, Málaga, Spain. The authors have no conflicts of interest to disclose. All the authors are current members of the “Grupo Andaluz de Trabajo en Enfermedad Inflamatoria Intestinal.” Reprints: Estefanía Moreno-Rincón, MD, Department of Gastroenterology, University Hospital Reina Sofía, Avenida Menéndez Pidal s/n, 14004 Córdoba, Spain (e-mail: [email protected]). Copyright © 2015 Crohn’s & Colitis Foundation of America, Inc. DOI 10.1097/MIB.0000000000000400 Published online 28 April 2015.
1564
| www.ibdjournal.org
several studies, among them, various clinical trials,1–3 2 metaanalyses,4,5 and a Cochrane review.6 In this latest review, the global rate of failure in maintaining remission with and without this drug was of 44% and 65%, respectively, with a relative risk of 0.68 (95% confidence interval [CI], 0.54–0.86). This modest result is explained by poor methodological quality of some of the trials included. After publication of this study, a trial was carried out with this same purpose.3 Seventy-two patients with active UC and steroid dependence were included and randomized to receive AZA or mesalazine for 6 months. The steroid-free clinical and endoscopic remission was obtained in 53% of patients treated with AZA as compared with the 21% of those treated with mesalazine, with an odds ratio of 4.78 (95% CI, 1.57–14.5). The evidence supporting the use of thiopurines (TPs) for UC also comes from observational cohorts of retrospective series.7–11 In a 30-year review of the Oxford Clinic on the use of these drugs, the global remission in 346 patients with UC treated with AZA was 86% after 6 months. After 5 years of treatment, this percentage was of 62%, by applying a strict definition of relapse, and of 81% if a mild relapse was considered using a short course of steroids.12 Inflamm Bowel Dis Volume 21, Number 7, July 2015
Copyright © 2015 Crohn’s & Colitis Foundation of America, Inc. Unauthorized reproduction of this article is prohibited.
Inflamm Bowel Dis Volume 21, Number 7, July 2015
The safety profile of TP is well known. The cumulative incidence of adverse effects was of about 26%, and the annual risk was 7% per patient-year of treatment in the study by Chaparro et al,13 where the existence of adverse effects was analyzed on a cohort of 3931 patients treated with TP for an average period of 44 months. Most of these events happen at the beginning of treatment. After the first few weeks, the main risks of indefinite use of these drugs are the occurrence of myelotoxicity, liver toxicity, infections, and malignancies (lymphoproliferative disorders or nonmelanoma skin cancer). It was determined that the risk of infections and lymphoproliferative diseases14 is increased with age and in situations of associated comorbidity.15 For this reason, the discontinuation of treatment is suggested if the disease remains in remission for a long period.16,17 However, there is no overwhelming evidence for this. The risk of relapse in Crohn’s disease is about 35% to 77% in the first year and 65% to 75% after 5 years of drug discontinuation.18 The most important information obtained about the discontinuation of TPs in patients with UC in remission comes from a randomized controlled clinical trial19 and 4 retrospective studies,12,20–22 where there is a high percentage of significant clinical relapse (SCR) in patients who stopped the treatment with TP. Despite all this, there are no general recommendations yet that can help select by themselves the patients who would benefit from the discontinuation of TP without an increased relapse risk, so currently the choice is to individualize the risk on a patient-bypatient basis. So far, the results obtained in previous studies do not show strong conclusions, since usually; the potential statistically significant predictive factors that can be obtained have low–moderate clinical significance. Therefore, they are not recommended to be used as a basis for decision making. The aim of this study was to clarify whether it is possible to discontinue the TP therapy in patients with UC in sustained remission. To such effect, the outcome of patients after drug withdrawal will be evaluated for identifying relapse rates and predictive factors of a poor prognosis.
MATERIALS AND METHODS Design of the Study A retrospective, observational, and multicenter study was performed, where 7 hospitals of Andalucía (Spain) participated. A total of 102 patients with UC according to the international criteria23 were recruited consecutively according to the database and records of patients available on each center. The following variables were obtained:
Demographics: gender, age at the time of diagnosis, active smoking upon discontinuation (defined as 7 or more cigarettes per week). Disease-related: family history of inflammatory bowel disease, appendectomy, presence of extraintestinal manifestations, disease duration, disease extent (proctitis, leftsided colitis, or pancolitis), biological remission (defined
Patients with UC in Sustained Remission
as normal inflammatory parameters [white cell count level, 4000–10,000 units/mL; C-reactive protein [CRP], 0 to 5 mg/dL; erythrocyte sedimentation rate, 0–15 mm/sec; and platelets, 150,000–300,000 units/mL] upon discontinuation), biological parameters (hemoglobin, white cell count, platelets, CRP, erythrocyte sedimentation rate), endoscopic remission upon discontinuation (defined as 0 in Mayo score), amount of SCRs before the discontinuation of TP. Treatment-related: type of TP used and indication, time since diagnosis until introduction of TP, total time in steroidfree clinical remission, occurrence of adverse effects, need to reduce dosage of TP and reason, concomitant treatments, total duration of TP treatment, reason for drug discontinuation. Main variable-related: occurrence of SCR, time until SCR or until the last review.
Inclusion and Exclusion Criteria All patients had a previous diagnosis of UC and received maintenance treatment with TP (indicated due to steroid dependence or to maintain the remission induced by other drugs) for a period of at least 6 months. The drug was discontinued after at least 6 months of steroid-free clinical remission (defined as a normal/usual presence of stools with no blood or pus and no need of steroid). The cause for discontinuation was the occurrence of adverse events or the decision of the physician or the patient. A follow-up was carried out of all patients until the occurrence of SCR (defined as the occurrence of UC typical signs or symptoms requiring a rescue therapy such as oral or intravenous corticosteroids, biological therapy, immunosuppressant drugs, recapture with TP or surgery) or until the date of the last review in case of patients with no SCR. The following patients were excluded: patients who did not meet the above criteria, patients having an interruption of TP treatment with no clinical remission, and patients receiving a treatment combined with biological drugs and TP where the latter were discontinued in order to continue with the biological drug as monotherapy.
Statistical Analysis The statistical analysis of the obtained data was performed with GStat 2.0, with the statistics software SPSS 15.0 for Windows, and with Microsoft Excel 2007. First, a descriptive analysis of the variables was made. The continuous quantitative variables are represented using medians with interquartile ranges (IQRs) and categorical data with percentages. For comparison of hypothesis according to the main variable, the t tests of Student were used for quantitative variables and x2 test (or accurate Fisher’s test as necessary) for categorical variables. A univariant analysis was performed to determine the factors that predict relapse, and subsequently, a survival analysis with Kaplan–Meier curves, comparison of survival functions (log-rank test), and www.ibdjournal.org |
1565
Copyright © 2015 Crohn’s & Colitis Foundation of America, Inc. Unauthorized reproduction of this article is prohibited.
Inflamm Bowel Dis Volume 21, Number 7, July 2015
Moreno-Rincón et al
multivariant analysis of survival (COX regression) to determine the predictive factors on time to relapse.
Ethical Considerations The study obtained the approval of the Ethical Committee of Clinical Investigation (Comité Ético de Investigación Clínica [CEIC]) of Córdoba and the Spanish Agency of Drugs and Sanitary Products (Agencia Española de Medicamentos y Productos Sanitarios [AEMPS]).
RESULTS A total of 102 patients with diagnosis of UC and TP discontinuation according to the above criteria were recruited throughout the different participating centers. Table 1 includes the main clinical and demographic characteristics of the descriptive analysis and variables related to the TP therapy. The median time since diagnosis of the disease until the start of TP treatment was 32.5 months (IQR, 10–65), and the total median time of treatment with TP was 51 months (IQR, 21–90). In 34.31% of patients, there were adverse effects recorded, being the myelotoxicity the most frequently recorded (up to 15 patients) followed by infections (6 patients), digestive intolerance (5 patients), and liver toxicity (4 patients). No presence of lymphoma was observed in any patient. Twenty-six patients required a decrease of the TP dose. During TP treatment, 43.14% of patients received topical aminosalicylates (5-ASA), whereas 95.1% of patients received oral 5-ASA. After TP withdrawal, and as a maintenance treatment, 97% of patients were treated with oral or topical 5-ASA. The main variables defining the situation of the patient upon discontinuation of TP are shown in Table 1. An average SCR of 0.49 6 0.13 was observed before interruption of TP treatment. The median time of steroid-free remission before discontinuation was of 33 months (IQR, 14–70), and in 76.47% of patients, there was no endoscopic evaluation upon interruption of treatment. After TP discontinuation, a global SCR was observed in 32.35% of patients. The median time until relapse was of 12 months (IQR, 7–24), and the median follow-up time (until last review) in patients with no relapse was of 27 months (IQR, 9–75). The cumulative percentage of SCR was of 18.88% at 12 months, 36.48% at 36 months, and 43.04% after 5 years (Fig. 1). In terms of treatment of relapse, 24 patients (72.73%) required oral or intravenous corticosteroids. Fourteen patients (42.42%) with relapse restarted TP, with a median time since TP withdrawal until reintroduction of 15 months (IQR, 9–24). In 5 patients (15.15%), biological agents were used (1 patient adalimumab, 4 patients infliximab), with a median time until the use of this treatment of 9 months (IQR, 6–10). None of the patients required colectomy for treatment of relapse. In the univariant analysis (Table 2), different variables were compared in terms of the occurrence or not of SCR individually. It was found that the mean of time in steroid-free remission before
1566
TABLE 1. Clinical and Demographic Characteristics of the Population Variable Male gender, n (%) Current smoking, n (%) Appendectomy, n (%) Family history, n (%) Extraintestinal manifestations, n (%) Age at diagnosis, median (IQR), yr Disease duration, median (IQR), yr Disease extent, n (%) Proctitis Left-sided colitis Pancolitis
47 13 2 15 13 32 12
(46.08) (12.75) (1.96) (14.71) (12.75) (24–42) (8–16)
9 (8.82) 46 (45.1) 47 (46.08)
Main Characteristics of Thiopurine Treatment Variable Time from UC diagnosis until start of TP treatment, median (IQR), mo TP indication, n (%) Steroid dependence Maintenance of remission TP treatment duration, median (IQR), mo Drug-related adverse events, n (%) TP dose reduction, n (%) Reason for TP withdrawal, n (%) Adverse events Physician decision Patient decision
32.5 (10–65)
82 20 51 35 26
(80.39) (19.61) (21–90) (34.31) (25.29)
28 (27.25) 45 (44.12) 29 (28.43)
Disease Characteristics at the Time of Thiopurines Withdrawal Variable Time of steroid-free sustained remission 33 (14–70) before TP withdrawal, median (IQR), mo Biological remission, n (%) 98 (96.08) Blood parameters at the time of TP withdrawal, median (IQR) White cell count, units/mL 5250 (4400–6810) Hemoglobin, g/dL 13.1 (12.1–14.5) Platelets, units/mL 255,000 (199,000–294,000) CRP, mg/dL 1.45 (0.3–3.22) ESR, mm/sec 9 (5–18) Endoscopic remission at the time of TP withdrawal, n (%) Yes 22 (21.57) No 2 (1.96) Not evaluated 78 (76.47) ESR, erythrocyte sedimentation rate; TP, thiopurine.
| www.ibdjournal.org
Copyright © 2015 Crohn’s & Colitis Foundation of America, Inc. Unauthorized reproduction of this article is prohibited.
Inflamm Bowel Dis Volume 21, Number 7, July 2015
Patients with UC in Sustained Remission
FIGURE 1. Cumulative relapse rate after thiopurine withdrawal.
TABLE 2. Univariate Analysis: Factors Associated with Relapse After Thiopurines Withdrawal Variable Gender, n (%) Female Male (Ref.) Current smoking, n (%) Family history, n (%) Extraintestinal manifestations, n (%) Age at diagnosis, mean (6SD), yr Disease extent, n (%) Proctitis (Ref.) Left-sided colitis Pancolitis Disease duration, mean (6SD), yr Time of free-steroid sustained remission before TP withdrawal, mean (6SD), mo Biological remission, n (%) White cell count, mean (6SD), units/mL Hemoglobin, mean (6SD), g/dL CRP, mean (6SD), mg/dL ESR, mean (6SD), mm/sec Time from UC diagnosis until start of TP treatment, mean (6SD), mo TP treatment duration, mean (6SD), mo TP treatment duration recoded, n (%) #12 mo (Ref.) 13–48 mo .48 mo TP indication, n (%) Steroid dependence (Ref.) Maintenance of remission Number of relapses before TP withdrawal, mean (SD), mo
Relapse
21 12 3 5 4 30.12 2 15 16 12.41 34.09 31 5485.48 12.67 2.8 14.8 44.27 47.97
(38.18) (25.53) (23.08) (33.33) (30.77) (611.52) (22.22) (32.61) (34.04) (67.11) (635.81) (31.63) (61709.19) (61.91) (64.32) (616.8) (656.69) (651.73)
No Relapse
34 35 10 10 9 34.68 7 31 31 13.32 54.04 67 5710.92 13.54 3.56 13.89 49.99 66.86
P
(61.82) (74.47) (76.92) (66.67) (69.23) (615.65)
0.176 0.448 0.93 0.896 0.141
(77.78) (67.39) (65.96) (65.74) (642.36) (68.37) (61967.19) (61.59) (611.18) (612.01) (55.45) (647.05)
0.541 0.491 0.487 0.027 0.451 0.583 0.024 0.716 0.763 0.627 0.077
8 (47.06) 15 (45.45) 10 (19.23)
9 (52.94) 18 (54.55) 42 (80.77)
0.098 0.133
28 (34.15) 5 (25) 0.49 (61.06)
54 (65.85) 15 (75) 0.51 (61.45)
0.435 0.936
ESR, erythrocyte sedimentation rate; TP, thiopurine. P ,0.05 is considered statistically significant.
www.ibdjournal.org |
1567
Copyright © 2015 Crohn’s & Colitis Foundation of America, Inc. Unauthorized reproduction of this article is prohibited.
Inflamm Bowel Dis Volume 21, Number 7, July 2015
Moreno-Rincón et al
TP discontinuation was higher in the group with no relapse as compared with the group with relapse (54.04 versus 34.09 mo, respectively, P ¼ 0.027). There were differences found regarding the mean levels of hemoglobin at withdrawal, as it was significantly higher in the group with no relapse (13.54 versus 12.67 g/ dL, P ¼ 0.024). There was also a trend to statistic significance regarding duration of TP treatment because the mean of duration of treatment was higher in the group with no relapse as compared with the group with relapse (66.86 versus 47.97 mo, respectively, P ¼ 0.077). A recoding with the “total time of treatment with TP” variable was performed, so it was divided into fractions of #12 months, 13 to 48 months, and .48 months. Regarding this variable, there was no difference in the univariant analysis, but when comparing the hypothesis by stratifying by this new variable, there were statistically significant differences (Fig. 2), and the percentage of patients with relapse (19.23%) in the group of the patients treated .48 months with TP was significantly lower as compared with the rest of the other 2 groups (47.06% in the group of #12 mo and 45.45% in the group of 13–48 mo, P ¼ 0.009). A multivariant analysis of survival was performed (Cox regression) (Table 3) to identify the factors related to the presence of SCR in terms of progress. After ordered exclusion of nonsignificant variables in the model, the final variables with statistical significance were the extent of UC, the time since diagnosis of UC until beginning of TP therapy, the total duration of treatment with TP (recoded as indicated in Fig. 2), the number of clinical significant relapses before TP discontinuation, and the presence of biological remission at withdrawal. There were survival curves made by stratifying by each one of the significant variables of the multivariant model, obtaining statistical significance in the survival curves of total duration of treatment with TP (Fig. 3) (.48 versus #12 mo, P ¼ 0.008).
DISCUSSION After discontinuing the TP, a global SCR was recorded in 32.35% of patients. The cumulative percentage was 18.88% in the first year, 36.48% in the third year, and 43.04% after 5 years of
TABLE 3. Multivariate Analysis (Cox Regression) Variable Disease extent Proctitis (Ref.) Pancolitis Time from UC diagnosis until start of TP treatment, mo Number of relapses before TP withdrawal TP treatment duration recoded #12 mo (Ref.) .48 mo Biological remission
HR (95% CI)
P
5.01 (1.95–26.43) 1.01 (1.01–1.02)
0.028 0.039
1.3 (1.01–1.66)
0.029
0.15 (0.03–0.66) 0.013 0.004 (0.0001–0.14) 0.002
TP, thiopurine. P ,0.05 is considered statistically significant.
drug withdrawal. In the univariant analysis, factors such as steroid-free remission time before discontinuation, duration of treatment with TP, and hemoglobin level upon interruption acquired statistical significance. In the multivariant analysis, the independent factors related to the presence of SCR were a greater extent of the disease, a longer time from UC diagnosis to the start of TP treatment, and a higher number of relapses before discontinuation. The factors related to a lower incidence of SCR were a higher duration of treatment with TP and a biological remission at withdrawal. If we analyze the percentage of relapse annually, most of the previous studies show a higher percentage of around 35% to 77% as compared with 18.88% obtained in this study. Only the study by Kennedy et al22 shows similar rates, around 11% after 12 months. These differences probably are due to a longer follow-up period as compared with the first studies and longer previous duration of treatment with TP, which in our trial is of 60 months while in the rest of the studies is of 24 to 47 months. The differences regarding cumulative relapse percentages also appear when comparing rates after 5 years of discontinuation, which are of
FIGURE 2. Rate of relapse stratified by duration of thiopurine treatment.
1568
| www.ibdjournal.org
Copyright © 2015 Crohn’s & Colitis Foundation of America, Inc. Unauthorized reproduction of this article is prohibited.
Inflamm Bowel Dis Volume 21, Number 7, July 2015
Patients with UC in Sustained Remission
FIGURE 3. Survival analysis (log-rank) of relapse stratified by duration of thiopurine treatment.
65% to 75% in the previous studies and 43% in our cohort, and they might be explained also by the above factors. In 1992, Hawthorne et al19 carried out one of the first clinical trials in this field when, on a cohort of 79 patients with UC (67 in remission for at least 2 mo and 12 with steroid dependence), they performed randomization where a group of each arm received AZA and the other placebo. After 1 year of follow-up, the subgroup of patients in remission receiving AZA showed a relapse rate of 36%, as compared with the 59% receiving placebo. These differences were statistically significant, and the conclusion was that the long-term maintenance of AZA treatment in patients with UC in remission prevents the relapse of the disease. Fraser et al12 carried out a retrospective review in 222 patients (79 with Crohn’s disease and 143 with UC), to whom the TP treatment was discontinued after a remission period of more than 6 months. The relapse percentage was of 37% after a year, 66% after 3 years, and 75% after 5 years of follow-up, with no differences between both groups. In another retrospective study, Lobel et al20 analyzed the outcome of 61 patients with UC in steroid-free remission of more than 3 months in treatment with MP with a year of follow-up. All patients had been treated for at least 6 months. Thirty-nine patients continued with MP and 22 discontinued it. The median time until relapse in the first group was 58 weeks and in the second group 24 weeks (P , 0.05). The percentage of relapse after a year was 77% in the group that interrupted the treatment as compared with 43% of the group that continued it. In another retrospective multicenter study performed in 2009, Cassinotti et al21 recruited 127 patients with UC on treatment with TP, to whom this treatment was discontinued in a clinical remission situation, and they recorded the outcome of those patients over time. In their study, one-third of the patients showed clinical relapse after a year of follow-up, half of them after 2 years, and 75% of them after 5 years, while in our study, this information is 18.88%, 34.77% and 43.04%, respectively. It is difficult to
explain these differences from a purely objective point of view. An influencing factor may be the fact that in the study by Cassinotti et al, patients included had to be in a period of steroid-free remission for at least 3 months before the discontinuation. In our study, the period of sustained clinical remission was extended to a minimum of 6 months because it was considered that this would be the maximum period the TPs would need to show efficacy. In the study by Cassinotti et al, there is also a mention to the global colectomy rate, which was 10%. None of the patients of our cohort required colectomy as a rescue therapy for relapse. In the study performed by Kennedy et al,22 108 patients with UC were recruited retrospectively, to whom treatment with TPs was discontinued in a sustained remission situation, with an annual rate of relapse of 12%, which is slightly lower than the rate obtained in our study. This difference may probably be due to the strict inclusion criteria of the study by Kennedy et al, such as the need for a minimum period of 3 years of treatment with TP, which obviously could impact the results. Four studies have assessed the risk factors of relapse after TP discontinuation in patients with UC.12,19,21,22 Only the studies by Kennedy et al and Cassinotti et al identified a population at risk. In the multivariant analysis of the latter, the relapse risk was associated with the lack of sustained remission during the maintenance treatment with AZA (hazard ratio [HR], 2.35; 95% CI, 1.43–3.85; P ¼ 0.001), extensive colitis (HR, 1.79; 95% CI, 1.06– 3.02; P ¼ 0.028), and duration of treatment, so the shorter treatments (3–6 mo) involved a worse prognosis than the treatments that lasted more than 48 months (HR, 2.78; 95% CI, 1.27–6.11; P ¼ 0.008). Information reported in this study aims at the fact that it could be reasonable to maintain TP treatment for a minimum period of 4 years, a statement consistent with our results. The study by Kennedy et al provides relevant information regarding the biological parameters upon interruption, identifying factors such as white cell count level higher than 9100 units/mL www.ibdjournal.org |
1569
Copyright © 2015 Crohn’s & Colitis Foundation of America, Inc. Unauthorized reproduction of this article is prohibited.
Moreno-Rincón et al
(HR, 6.7; 95% CI, 1.86–24.2; P ¼ 0.004) and CRP level above 14 mg/dL (HR, 3.2; 95% CI, 1.48–7.05; P ¼ 0.003) as possible factors that predict relapse after TP discontinuation. Also, in this study, the response to the reintroduction of TP for treatment of relapse was analyzed; 71% of patients had a TP reintroduction, reaching high rates of response of up to 92%. In our cohort, the biological remission situation also reached statistical significance in the multivariant analysis, understanding this as normal biological parameters of inflammation upon discontinuation. The study by Louis et al24 is worth mentioning, which treated a cohort of 115 patients with Crohn’s disease in a sustained remission situation and with previous combined treatment (TP and infliximab), where the factors that predict relapse after infliximab discontinuation were analyzed. In that study, biological parameters such as high levels of white cell count, hemoglobin levels lower than 14.5 g/dL, CRP above 5 mg/ dL, and faecal calprotectin lower than 300 mg/g were factors that predicted a higher risk of relapse. An important factor to be considered is the evaluation of endoscopic remission at withdrawal, defined in our study as a 0 score of the endoscopic Mayo subscore. It is known that reaching the mucosal healing is one of the main therapeutic goals in these patients, having proven in various studies25,26 how its presence is associated with a better prognosis of the disease by reducing the occurrence of complications, specifically the risk of colectomy and colorectal cancer. Its evaluation, therefore, should be a factor to consider before determining the interruption of immunosuppressant medication. It is worth noting that new laboratory parameters have appeared throughout the last few years, such as measurement of faecal calprotectin27–29 and its use as a marker of endoscopic activity, offering valid alternatives to the endoscopic assessment, considering the inherent risks of this procedure. Therefore, it is evident that the use of TP in UC is widely accepted in terms of efficacy, being one of the basic therapeutic strategies in the management of this disease. The different studies performed throughout the years have tried to answer the repeated question of whether the decision to stop these therapies is suitable in patients with long periods of remission, given the noticeable rate of adverse effects associated. The results obtained in these studies have been varied but, as we conclude, the interruption of these therapies seems to be associated with a considerable risk of relapse we should not forget. In this therapeutic strategy, there are doubts before discontinuing the TP, such as minimum duration of treatment that is necessary to maintain, time of remission that should be kept before considering discontinuation, and tools we should use to assess this remission. The most suitable strategy is probably an evaluation that is not only clinical and biological but also endoscopic. The main limitation of our study lies in its retrospective nature. This affects the low percentage of endoscopic evaluation before the discontinuation of TP because a mucosal healing situation, as set forth in the above-mentioned studies, would likely involve a factor that predicts low risk of relapse after interruption
1570
Inflamm Bowel Dis Volume 21, Number 7, July 2015
of treatment. Another important limitation is the absence of measurement of faecal calprotectin, which was determined in the study by Kennedy et al, although not reaching the statistical significance, probably due to the low amount of recruited patients. As conclusion, we should indicate that TP discontinuation in patients with UC in sustained remission is associated with a relapse risk of 18.88% after the first year. Therefore, it is not recommended to interrupt the treatment systematically but to individualize according to the specific characteristics of each patient. Factors such as pancolitis, a longer period of time since diagnosis of UC until TP first administration, or a high number of clinical relapses before withdrawal should be considered before deciding on the discontinuation of these drugs. Further controlled, randomized, and prospective studies may provide more information on this topic.
REFERENCES
1. Jewell DP, Truelove SC. Azathioprine in ulcerative colitis: final report on controlled therapeutic trial. Br Med J. 1974;4:627–630. 2. Sood A, Midha V, Sood N, et al. Role of azathioprine in severe ulcerative colitis: one-year, placebo-controlled, randomized trial. Indian J Gastroenterol. 2000;19:14–16. 3. Ardizzone S, Maconi G, Russo A, et al. Randomised controlled trial of azathioprine and 5-aminosalicylic acid for treatment of steroid dependent ulcerative colitis. Gut. 2006;55:47–53. 4. Gisbert JP, Linares PM, McNicholl AG, et al. Meta-analysis: the efficacy of azathioprine and mercaptopurine in ulcerative colitis. Aliment Pharmacol Ther. 2009;30:126–137. 5. Khan KJ, Dubinsky MC, Ford AC, et al. Efficacy of immunosuppressive therapy for inflammatory bowel disease: a systematic review and metaanalysis. Am J Gastroenterol. 2011;106:630–642. 6. Timmer A, McDonald JWD, Tsoulis DJ, et al. Azathioprine and 6-mercaptopurine for maintenance of remission in ulcerative colitis. Cochrane Database Syst Rev. 2012;9:CD000478. 7. Sood A, Kaushal V, Midha V, et al. The beneficial effect of azathioprine on maintenance of remission in severe ulcerative colitis. J Gastroenterol. 2002;37:270–274. 8. Holtmann MH, Krummenauer F, Claas C, et al. Long-term effectiveness of azathioprine in IBD beyond 4 years: a European multicenter study in 1176 patients. Dig Dis Sci. 2006;51:1516–1524. 9. Sood A, Midha V, Sood N, et al. Long term results of use of azathioprine in patients with ulcerative colitis in India. World J Gastroenterol. 2006; 12:7332–7336. 10. Jharap B, Seinen ML, de Boer NK, et al. Thiopurine therapy in inflammatory bowel disease patients: analyses of two 8-year intercept cohorts. Inflamm Bowel Dis. 2010;16:1541–1549. 11. Chebli LA, Chaves LD, Pimentel FF, et al. Azathioprine maintains longterm steroid-free remission through 3 years in patients with steroiddependent ulcerative colitis. Inflamm Bowel Dis. 2010;16:613–619. 12. Fraser AG, Orchard TR, Jewell DP. The efficacy of azathioprine for the treatment of inflammatory bowel disease: a 30 year review. Gut. 2002;50: 485–489. 13. Chaparro M, Ordás I, Cabré E, et al. Safety of thiopurine therapy in inflammatory bowel disease: long-term follow-up study of 3931 patients. Inflamm Bowel Dis. 2013;19:1404–1410. 14. Lewis JD. Risk of lymphoma in patients with inflammatory bowel disease. Gastroenterol Hepatol. 2012;8:45–47. 15. Gisbert JP, Chaparro M. Systematic review with meta-analysis: inflammatory bowel disease in the elderly. Aliment Pharmacol Ther. 2014;39: 459–477. 16. Kabac¸am G, Törüner M. Can azathioprine/6-mercaptopurine treatment be withdrawn in patients with response in ulcerative colitis: what is the most appropriate time for this? Turk J Gastroenterol. 2012;23:17–20.
| www.ibdjournal.org
Copyright © 2015 Crohn’s & Colitis Foundation of America, Inc. Unauthorized reproduction of this article is prohibited.
Inflamm Bowel Dis Volume 21, Number 7, July 2015
17. Clarke K, Regueiro M. Stopping immunomodulators and biologics in inflammatory bowel disease patients in remission. Inflamm Bowel Dis. 2012;18:174–179. 18. Treton X, Bouhnik Y, Mary JY, et al. Azathioprine withdrawal in patients with Crohn’s disease maintained on prolonged remission: a high risk of relapse. Clin Gastroenterol Hepatol. 2009;7:80–85. 19. Hawthorne AB, Logan RFA, Hawkey CJ, et al. Randomised controlled trial of azathioprine withdrawal in ulcerative colitis. BMJ. 1992;305: 20–22. 20. Lobel EZ, Korelitz BI, Xuereb MA, et al. A search for the optimal duration of treatment with 6-mercaptopurine for ulcerative colitis. Am J Gastroenterol. 2004;99:462–465. 21. Cassinotti A, Actis GC, Duca P, et al. Maintenance treatment with azathioprine in ulcerative colitis: outcome and predictive factors after drug withdrawal. Am J Gastroenterol. 2009;104:2760–2767. 22. Kennedy NA, Kalla R, Warner B, et al. Thiopurine withdrawal during sustained clinical remission in inflammatory bowel disease: relapse and recapture rates, with predictive factors in 237 patients. Aliment Pharmacol Ther. 2014;40:1313–1323.
Patients with UC in Sustained Remission
23. Stange EF, Travis SP. The European consensus on ulcerative colitis: new horizons? Gut. 2008;57:1029–1031. 24. Louis JY, Vernier-Massouille G, Grimaud JC, et al. Maintenance of remission among patients with Crohn’s disease on antimetabolite therapy after infliximab therapy is stopped. Gastroenterology. 2012;142:63–70. 25. Rutgeerts P, Vermeire S, Van Assche G. Mucosal healing in inflammatory bowel disease: impossible ideal or therapeutic target? Gut. 2007;56: 453–455. 26. Dave M, Loftus E. Mucosal healing in inflammatory bowel disease—a true paradigm of success? Gastroenterol Hepatol. 2012;8:29–38. 27. Smith LA, Gaya DR. Utility of faecal calprotectin analysis in adult inflammatory bowel disease. World J Gastroenterol. 2012;18:6782–6789. 28. Moniuszko A, Wisniewska A, Rydzewska G. Biomarkers in management of inflammatory bowel disease. Prz Gastroenterol. 2013;8: 275–283. 29. Burri E, Beglinger C, von Felten S, et al. Fecal calprotectin and the clinical activity index are both useful to monitor medical treatment in patients with ulcerative colitis. Dig Dis Sci. 2015;60: 485–491.
www.ibdjournal.org |
1571
Copyright © 2015 Crohn’s & Colitis Foundation of America, Inc. Unauthorized reproduction of this article is prohibited.
