Alimentary Pharmacology and Therapeutics
The impact of cytomegalovirus reactivation and its treatment on the course of inflammatory bowel disease M. Delvincourt*, A. Lopez†, S. Pillet‡,§, A. Bourrier*, P. Seksik*,¶, J. Cosnes*, F. Carrat**, J. Gozlan††, L. Beaugerie*, X. Roblin§,‡‡, L. Peyrin-Biroulet† & H. Sokol*,¶,§§
*Department of Gastroenterology, Saint Antoine Hospital, Assistance Publique H^ opitaux de Paris and Paris VI University, Paris, France. † Department of Hepatogastroenterology, Nancy University Hospital, Vandoeuvre-lesNancy, France. ‡ Department of Bacteriology-VirologyHygiene, Saint Etienne Hospital, Saint Etienne, France. § GIMAP EA 3064, Faculte de medecine de Saint-Etienne, Universite de Lyon, Lyon, France. ¶ Laboratoire INSERM U1057/UMR CNRS 7203, Universite Pierre et Marie Curie 6, Paris, France. **Public Health department, SaintAntoine Hospital, AP-HP Paris; UMRS 707, Universite Pierre et Marie Curie-Paris 6 & INSERM, Paris, France. †† Department of Virology, Saint Antoine Hospital, Assistance Publique H^opitaux de Paris and Paris VI University, Paris, France. ‡‡ Department of Gastroenterology, Saint Etienne Hospital, Saint Etienne, France. §§ Commensal and Probiotics-Host Interactions Laboratory, UMR 1319 Micalis, Jouy-en-Josas, France. Correspondence to: Dr H. Sokol, Service de Gastroenterologie et Nutrition, H^ opital Saint-Antoine 184 rue du faubourg Saint-Antoine, 75571 Paris CEDEX 12, France. E-mail: [email protected] Publication data Submitted 27 November 2013 First decision 3 January 2014 Resubmitted 14 January 2014 Accepted 15 January 2014 EV Pub Online 9 February 2014 This article was accepted for publication after full peer-review.
712
SUMMARY Background Consequences of latent cytomegalovirus (CMV) infection reactivation on inflammatory bowel disease (IBD) flare, as a flare-worsening factor or simple bystander, are debated. Impact of anti-viral treatment on IBD course is poorly known. Aim To assess the impact of CMV reactivation on patients hospitalised for IBD flare and the effect of anti-viral treatment on IBD flare in patients with CMV reactivation. Methods First, a population of UC patients from Saint-Antoine hospital, in flare with positive blood CMV PCR without anti-viral treatment (n = 26), were compared to matched patients with negative blood CMV PCR in a case–control study. Secondly, a total of 110 hospitalisations between October 2003 and May 2012 for IBD flare-up with CMV reactivation (80 diagnosed on blood PCR, 33 on tissue PCR) were identified in three French referral centres. Evolution following CMV reactivation diagnosis was compared between patients receiving anti-viral treatment and those who did not. Results In the case–control study, no differences were observed between the two groups regarding length of hospital stay and colectomy rate. Comparing treated and untreated patients, no differences were observed at inclusion regarding age, gender, IBD type, immunosuppressant, CRP and haemoglobin level. No differences were observed regarding CRP level decrease at 10 days and colectomy rate at 3 months. Anti-viral treatment was associated with lower serum albumin level at inclusion and longer hospitalisation. Conclusions CMV reactivation does not appear to alter the course of IBD flare. CMV treatment does not seem to impact the course of IBD. These results should be confirmed prospectively. Aliment Pharmacol Ther 2014; 39: 712–720
ª 2014 John Wiley & Sons Ltd doi:10.1111/apt.12650
CMV reactivation in IBD INTRODUCTION Cytomegalovirus (CMV) is a member of the Herpesviridae family. Infection, through close contact with different body fluids, is common in all human populations, with seroprevalence rates ranging from 40% to 100%.1 CMV disease manifestations in infected humans are numerous, from the most common asymptomatic form in immunocompetent patients to the severe CMV-induced interstitial pneumonitis in allogeneic bone marrow transplant recipients. After primary infection, CMV persists in a latent state in neutrophils, T lymphocytes, endothelial cells and renal epithelial cells. Reactivation can be triggered by immunosuppression. Thus, at-risk populations include transplanted patients, HIV-positive patients and, more largely, patients receiving immunosuppressive therapy.2 The first case of ulcerative colitis (UC)-associated CMV infection was reported in 1961 in a patient exhibiting viral cytoplasmic inclusions on colonic biopsies.3 Indeed, patients with inflammatory bowel disease (IBD) are often immunosuppressed, through combined action of potentially severe malnutrition, use of immunosuppressants and primary impairment of immune functions. These factors, associated with a CMV tropism for inflammatory tissues, lead to a significant increased risk of developing CMV reactivation.4 Thus, prevalence of CMV reactivation during moderate-to-severe IBD flare-up ranges from 21% to 34% in different studies.5, 6 Among steroid-refractory IBD patients, prevalence increases beyond 30%.1, 7 During IBD flare-up, two distinct studies have identified, on multivariate analysis, the female gender, a pancolic inflammatory disease and active inflammation on histology as independent risk factors for CMV reactivation.1, 8 Various methods are used to detect CMV reactivation. Quantitative whole-blood and plasmatic CMV PCR and histological analysis of colonic biopsies have shown a powerful correlation5, 9 and are often considered the gold standard methods.6 So far, European guidelines (European Crohn’s and Colitis Organization guidelines, 2009) recommended the use of immunohistochemistry and tissue CMV PCR to detect the virus in immunosuppressant-refractory IBD patients.10 Despite the clear association between IBD flare-up and CMV reactivation, the impact of the latter on the former and vice versa is debated.11 Several studies and meta-analysis have suggested a deleterious role of CMV reactivation in IBD flare-ups, increasing the risk for steroid resistance and need for colectomy.12 On the other
Aliment Pharmacol Ther 2014; 39: 712-720 ª 2014 John Wiley & Sons Ltd
hand, other studies have considered CMV as a simple marker of digestive inflammation severity, disappearing under appropriate immunosuppressive treatment.13 In this context, we aimed to evaluate the impact of anti-viral therapy against CMV on IBD course in a group of IBD patients hospitalised for IBD flare-ups complicated with CMV reactivation compared with a group of patients with IBD flare and CMV reactivation who did not receive anti-viral therapy.
MATERIAL AND METHODS A population of UC patients in flare with significant blood CMV viral load without anti-viral treatment (n = 26) entered a retrospective case–control study. In this study, patients were matched 1:1 according to sex, age 5 years and year of hospitalisation in Saint Antoine hospital to patients with UC flare-up with negative blood CMV PCR. Secondly, in the retrospective study, all consecutive patients hospitalised in French tertiary centres of Paris (Saint Antoine Hospital), Nancy and Saint Etienne, between October 2003 and May 2012, for IBD flare-up complicated with CMV reactivation, were included. The diagnosis of IBD was based on clinical, endoscopic, radiological and histological parameters. CMV reactivation was assessed by positive blood CMV PCR and/or positive tissue CMV PCR on colonic biopsies. Viral load significance thresholds were defined in each centre according to the techniques used (500–2500 copies/mL for PCR in blood and semi-quantitative method for PCR in colonic biopsies). These values were determined in accordance with established significant thresholds for CMV viral loads used to diagnose CMV infection in the setting of bone marrow transplants in each centre. When several CMV PCR were positive during a single hospital stay, the first CMV PCR above the significance threshold was considered. In patients for whom anti-viral treatment was initiated, the CMV viral load preceding the anti-viral treatment introduction was considered. Demographic (age, sex, IBD type, extent of the disease), clinical, biological, endoscopic and therapeutic (use of corticosteroids and other ongoing immunosuppressant treatments) profiles of all investigated patients were retrospectively collected. Clinical evaluation at inclusion was based on specific clinical severity indexes, performed on routine clinical practice: Lichtiger score for UC and Harvey–Bradshaw index for Crohn’s disease. Endoscopic evaluation included registering of severe colitis features, such as wide mucosal defects, punched-out ulcerations
713
M. Delvincourt et al. and mucosal detachment. Severe colitis, according to both clinical (Lichtiger score superior or equal to 10; Harvey–Bradshaw index superior to 16) and endoscopic criteria, were recorded. Blood haemoglobin, C-reactive protein (CRP) and albumin levels at inclusion were extracted from patient files for the purpose of the study. All clinical and biological criteria were re-assessed on day 10 4 after inclusion, allowing the calculation of CMV viral load drops, Lichtiger score drops and CRP level drops between inclusion and re-evaluation, in absolute values and percentage. Finally, length of hospital stay, the need for colectomy and, if so, the time from inclusion to colectomy, were assessed. Patients receiving anti-viral treatment and those without anti-viral treatment were compared in terms of disease type, flare-up severity and use of immunosuppressants at inclusion. Disease evolution was assessed and compared between the 2 groups according to evolution of CRP level between inclusion and re-evaluation, average hospital stay and 3-month colectomy rates. For statistical analysis, qualitative variables expressed in absolute values and percentages were compared using Chi-squared test or Fisher’s exact test when appropriate. Quantitative variables, expressed in mean value with standard deviation were compared using Student t-test or Wilcoxon test. The results were considered statistically significant when P values were below 0.05. Statistical analysis of the data was carried out using JMP software version 7.0 (USA).
CMV + (n = 26) Sex (%): M/F 14 (54%)/12 (46%) Age (years s.d.) 36.8 12.3 Duration of disease 4.2 5.1 (years s.d.) Ongoing immunosuppressants on inclusion Steroids (%) 15 (60%) Steroid dosage 30 28 Time on steroids (d) 32 56 Azathioprine (%) 10 (40%) Methotrexate (%) 1 (4%) Ciclosporin (%) 2 (8%) Anti-TNFa (%) 5 (20%) Lichtiger index s.d. 9.5 4.2 Lichtiger ≥10 (%) 16 (61.5%) Endoscopic Mayo index 2.3 0.7
CMV
RESULTS Case–control study A population of 26 UC patients with positive blood CMV PCR was matched for sex, age 5 years and year of hospitalisation to 26 UC patients with negative CMV PCR. The two groups were similar in terms of age at IBD diagnosis, duration of IBD, ongoing steroid therapy at inclusion, clinical and endoscopic severity as defined by Lichtiger and Mayo score respectively (Table 1). No significant difference was observed between the two groups regarding length of hospital stay (8.7 days vs. 9.7 days for CMV-positive and CMV-negative patients respectively; P 0.42) and 3-month colectomy rate (15.4% and 23.1% for CMV-positive and CMV-negative patients respectively; P 0.48). Retrospective study A total of 110 hospitalisations (in 105 patients) between October 2003 and May 2012 for IBD flare-up with CMV reactivation were included. Patient characteristics are detailed in Table 2. Mean age at inclusion was 40.4 14.4 years. Eighty-seven patients (79%) had UC. As defined by Lichtiger score (≥10), 43 UC patients (52% of all UC patients) had severe colitis at the time of inclusion. Sixty-five per cent of UC patients presented with pancolitis. Treatment at inclusion included steroids, azathioprine and anti-TNF alpha in 70%, 44% and 26% of cases respectively.
(n = 26)
P
14 (54%)/12 (46%) 38.6 12.1 7.6 7.9
1 0.53 0.08
14 (56%) 30 32 10 17 2 (8%) 1 (4%) 0 1 (4%) 10.4 3.2 17 (65.4%) 2.1 0.9
1 0.85 0.22 0.02 1 0.49 0.19 0.61 0.77 0.56
Table 1 | Baseline and demographic characteristics of ulcerative colitis patients in flare, with or without positive blood CMV PCR
CMV, cytomegalovirus; PCR, polymerase chain reaction; M, male; F, female. Categorical variables reported as absolute values and %. Continuous variables reported as mean standard derivation. 714
Aliment Pharmacol Ther 2014; 39: 712-720 ª 2014 John Wiley & Sons Ltd
CMV reactivation in IBD Table 2 | Baseline and demographic characteristics of IBD patients in flare with CMV reactivation diagnosed on blood and/or tissue PCR (n = 110) Sex ratio (M/F) (%) Mean age (years) IBD type (%): CD/UC Montreal classification Age at diagnosis (% CD): A1/A2/A3 Location UC (%) E1 E2 E3 CD (%) L1 L2 L3 Behaviour (% CD): B1/B2/B3/P Severity index UC and IBDU: Lichtiger index [range] CD: Harvey–Bradshaw [range] Lichtiger ≥10 (% UC) CRP (mg/L) Hb (g/dL) Albumin (g/L) Endoscopic severe colitis (Mayo index = 3) (%) Treatments at inclusion Corticosteroids (%) Azathioprine (%) Ciclosporin (% of UC) Anti-TNF alpha (%) Anti-viral therapy (%)
57/43 40.4 14.7 23 (21%)/87 (79%) 9%/83%/35%
5% 30% 65% 18% 30% 52% 59%/9%/32%/30.4% 8 [0–18] 8 [3–20] 43 (52%) 33.5 33.8 11.1 2.1 29.1 5.8 61%
73 (70%) 46 (44%) 6 (7%) 27 (26%) 68%
IBD, inflammatory bowel disease; CD, Crohn’s disease; UC, ulcerative colitis; IBDU, inflammatory bowel disease type unclassified; CRP, C-reactive protein; Hb, Haemoglobin. Categorical variables reported as absolute values and %. Continuous variables reported as mean standard derivation.
CMV reactivation was diagnosed on blood PCR for 80 patients and on tissue PCR in 33 patients. In the last group of patients with positive tissue PCR, 3 had known positive blood PCR CMV as well, while no data regarding blood viral load were available for the others. Anti-viral therapy was initiated in 68% of these patients. Regarding patients with positive blood CMV PCR, there were no significant differences between the treated and the nontreated groups in terms of gender, age, duration of disease or IBD type (Table 3). Distribution of IBD treatments at inclusion was similar between the two groups. Among UC patients, Lichtiger score was similar, and a comparable rate of clinically severe acute colitis was observed between the two groups of subjects. Aliment Pharmacol Ther 2014; 39: 712-720 ª 2014 John Wiley & Sons Ltd
All treated patients experienced a decrease in viral load at 10 days, 63% of them with a control blood CMV PCR below established thresholds (data not shown). Regarding biological parameters, no statistically significant difference was observed in terms of CRP and haemoglobin levels at inclusion between treated and untreated patients. Albumin levels differed significantly, with lower levels observed in the treated group (P = 0.03). Ten days (4) after inclusion, CRP level drops did not differ between the two groups (Figure 1). Hospital stay was significantly longer among patients who received anti-viral treatment (16.3 days vs. 8.3 days; P < 0.001). No difference was observed in terms of colectomy rate at 3 months after inclusion (10.6% vs. 13.3%; P = 0.7). We then performed a subpopulation analysis restricted to the 25 patients with severe UC flare-up (defined by having a Lichtiger index ≥10 at inclusion), with positive blood CMV PCR (15 treated, 10 untreated). Demographic data and ongoing immunosuppressive treatment at inclusion were comparable between the two groups (Table 4). The treated group (n = 15) did not differ from the untreated group (n = 10) in terms of clinical severity (Lichtiger score: 11.9 2.0 vs. 12.3 1.8, respectively; P = 0.48), but had higher biological markers of severity bordering significance (CRP: 42.9 47.5 vs. 17.6 19.2; P = 0.05). On re-evaluation 10 days later, CRP level drop and Lichtiger score drop were comparable between treated and untreated patients (Figure 2). Mean hospital stay was significantly longer among treated patients (15.3 days vs. 8.8 days; P = 0.04). No difference was observed in terms of 3-month colectomy rate (15.0% vs. 10.0%; P = 0.9). Among the 33 patients with positive tissue CMV PCR, 22 received anti-viral treatment, while 11 patients did not (Table 5). The two groups of patients did not differ in terms of sex, age or ongoing treatments at inclusion. Assessment of severity showed no significant difference between the two groups regarding clinical and biological parameters (data not shown). Comparison of CRP level drop, length of hospital stay and 3-month colectomy rate showed no benefit for the anti-viral therapy (data not shown).
DISCUSSION The role of CMV reactivation and thus the usefulness of anti-viral therapy in IBD flare are highly debated. In this case–control study, comparing UC patients with positive blood CMV PCR who did not receive anti-viral 715
M. Delvincourt et al. Table 3 | Baseline and demographic characteristics of patients with positive blood PCR CMV, treated or not with anti-viral treatment Anti-viral treatment (n = 49) Sex (%): M/F Age (years s.d.) IBD type (%): CD/UC Immunosuppressive treatment Steroids (%) Azathioprine (%) Methotrexate (%) Ciclosporin (%) Anti-TNFa (%) CRP (mg/L) Haemoglobin (g/dL) Albumin (g/L) Lichtiger index (UC) Lichtiger index ≥10 (% of UC)
No anti-viral treatment (n = 31)
23 (47%)/26 (53%) 49 37 13 (27%)/36 (73%)
18 (58%)/13 (42%) 31 38.8 10 (32%)/21 (68%)
29 (60%) 18 (38%) 2 (4%) 3 (6%) 15 (31%) 39.2 36.2 10.7 2.1 26.8 4.6 8.1 3.9 11 (52%)
21 (68%) 15 (48%) 2 (6%) 3 (10%) 5 (16%) 28 27 11.5 1.9 30.1 5.9 8.1 4.1 10 (48%)
P 0.33 0.29 0.58 0.51 0.34 0.65 0.58 0.13 0.09 0.09 0.03 1 1
Categorical variables reported as absolute values and %. Continuous variables reported as mean standard derivation.
NS
P = 0.014
P = 0.021
250 CRP (mg/L)
CRP (mg/L)
150
100
50
200 150 100 50 0
0 D0
D 10
No antiviral treatment
D0
D 10
Antiviral treatment
Figure 1 | CRP (mg/L) level evolution between inclusion and re-evaluation (Day 10 4) in IBD patients in flare with positive blood CMV PCR, treated or not with anti-viral treatment.
treatment (n = 26) and a matched control group of UC patients without detected CMV in blood (n = 26), no difference between the two groups was observed in terms of clinical and endoscopic severity, length of hospital stay or colectomy rate. Of note, patients with positive CMV PCR received the same amount of steroids (similar percent of patients treated, similar dosage, similar treatment length) compared with the control group. These results are limited by the retrospective and unrandomised design of the study. To help assess the significance of these first results, a retrospective study evaluating 716
the impact of anti-viral treatment on 110 patients hospitalised for IBD flare-up complicated with blood and/or tissue CMV reactivation was conducted. Anti-viral treatment failed to demonstrate a favourable impact on clinical or biological outcomes, length of hospital stay or colectomy rate at 3 months, thus strengthening the role of CMV as an innocent bystander. On the contrary, longer hospital stays were observed among patients receiving anti-viral therapy as compared with patients without anti-viral treatment, suggesting a higher severity of the treated population at inclusion, thus representing a bias in Aliment Pharmacol Ther 2014; 39: 712-720 ª 2014 John Wiley & Sons Ltd
CMV reactivation in IBD Table 4 | Baseline and demographic characteristics of patients with severe UC flare-up (Lichtiger ≥10) and positive blood PCR CMV, treated or not with anti-viral treatment Anti-viral treatment (n = 15)
No anti-viral treatment (n = 10)
P
7 (47%)/8 (53%) 36.5 11.5
5 (50%)/5 (50%) 35.7 10.8
0.87 0.72
12 (80%) 3 (20%) 1 (7%) 6 (40%) 11.9 2.0 42.9 47.5 10.7 2.2 27.1 5.6
7 (70%) 5 (50%) 0 (0%) 2 (20%) 12.3 1.8 17.6 19.2 11.1 2.3 30.6 5.7
0.57 0.12 0.41 0.29 0.48 0.05 0.64 0.12
Sex (%): M/F Age (years s.d.) Immunosuppressive treatment Steroids (%) Azathioprine (%) Ciclosporin (%) Anti-TNFa (%) Lichtiger index (UC) CRP (mg/L) Haemoglobin (g/dL) Albumin (g/L)
Categorical variables reported as absolute values and %. Continuous variables reported as mean standard derivation.
(a)
NS P = 0.014
60 40 20
150 100 50
0
0 D0
D 10
(b)
D0
D 10
NS P = 0.009
20
P = 0.003
20
15
Lichtiger score
Lichtiger score
P = 0.021
200
CRP (mg/L)
CRP (mg/L)
80
10 5
15 10 5 0
0 D0
D 10
No antiviral treatment
D0
D 10
Antiviral treatment
Figure 2 | CRP level and Lichtiger score evolution between inclusion and re-evaluation (Day 10 4) in patients with severe UC flares and positive blood CMV PCR, treated or not with anti-viral treatment. Aliment Pharmacol Ther 2014; 39: 712-720 ª 2014 John Wiley & Sons Ltd
717
M. Delvincourt et al. Table 5 | Baseline and demographic characteristics of UC patients in flare with positive tissue PCR CMV, treated or not with anti-viral treatment
Sex (%): M/F Age (years s.d.) IBD type (%): CD/UC Immunosuppressive treatment Steroids (%) Azathioprine (%) Ciclosporin (%) Anti-TNFa (%) Lichtiger index (UC) CRP (mg/L) Haemoglobin (g/dL)
Anti-viral treatment (n = 22)
No anti-viral treatment (n = 11)
P
17 (77.3%)/5 (22.7%) 41.2 20.1 0/22 (100%)
7 (63.6%)/4 (36.4%) 52.5 13 0/11 (100%)
0.41 0.29 1
18 (85.7%) 10 (45.5%) 1 (4.6%) 6 (27.3%) 12.7 2.8 53.1 29.8 11.1 1.8
9 (81.8%) 6 (54.6%) 0 1 (9.1%) 13.2 1.3 41.3 21.4 11.4 1.3
0.77 0.34 0.47 0.23 0.97 0.28 0.44
Categorical variables reported as absolute values and %. Continuous variables reported as mean standard derivation.
the study. Indeed, even though clinical and biological status at inclusion did not statistically differ between the two groups, other unassessed parameters, such as hemodynamic status or extra-digestive functions, may have been more severe in the treated population, explaining why these patients were selected to benefit from the anti-viral treatment. However, these results were confirmed when considering only a sub-population of severe UC patients (Lichtiger index ≥10). The large population of this retrospective study is in accordance with available epidemiological data on this topic. We observed a clear predominance of UC patients with mild to severe flare-ups, mostly pancolitis (Table 2).1, 8 A high rate of ongoing steroid therapy at inclusion is in accordance with the previously demonstrated association between this treatment and CMV reactivation. It is not clearly established whether this immunosuppressive treatment facilitates CMV reactivation or if it merely reflects the severity of the underlying inflammatory disease.14 Widmann et al. have suggested potential in vitro dose-dependent effects of steroids on suppression of CMV-specific T-lymphocyte function.15 Other works have confirmed increased viral proteins production in in vitro hydrocortisone-enriched environment.16 CMV reactivation has been witnessed in up to 30% of patients with severe and/or steroid-refractory colitis.14 Current clinical guidelines recommend testing for colonic CMV disease in patients with severe colitis and deliver anti-viral treatment to CMV-positive patients.10 Indeed, several studies have described the potentially pathogenic role of CMV on IBD flare-ups. Thus, Kambham et al. showed, among 40 patients suffering from steroid-refractory ulcerative colitis, a CMV disease 718
assessed by histological features on colonic biopsies in 10 patients (25%), whereas virus activity was demonstrated in only one of 40 patients from a matched population of patients with ulcerative colitis responding to steroids (2,5%).17 Previous works have acknowledged CMV to be an independent risk factor for medical treatment failure and toxic megacolon, both necessitating emergency colectomy.8, 14 Kambham and al. also showed a potential positive impact of the anti-viral treatment, allowing colonic viral clearance among three UC patients treated for CMV reactivation, associated with clinical remission of the ulcerative colitis with steroid suspension. On the contrary, six patients out of 8 UC patients with untreated CMV superinfection necessitated coloprotectomy.17 On the other hand, several recent studies have hypothesised the role of CMV as an innocent bystander of intestinal lesions, with no pathogenic activity by itself. Indeed, in vitro studies have shown spontaneous tropism of CMV towards proliferating cells in inflammatory tissues.18 Thus, Matsuoka and al. compared in 2007 a population of UC patients with mild-to-severe flare-up, complicated with CMV reactivation, assessed by plasmatic PCR and/or pp65 antigenemia (n = 25), with a control population of patients matched for disease activity, without detected CMV (n = 23). Under immunosuppressive treatment alone, without anti-viral treatment, no difference was shown between the two groups in terms of 8-week clinical remission rate and colectomy rate.13 In a French retrospective study, seven IBD patients (three ulcerative colitis, four Crohn’s diseases) with mild-to-severe flare-ups were diagnosed with CMV reactivation assessed on blood PCR. Five of them (71%) showed favourable response to immunosuppressants alone.19 Other studies have demonstrated possible Aliment Pharmacol Ther 2014; 39: 712-720 ª 2014 John Wiley & Sons Ltd
CMV reactivation in IBD clinical improvement in spite of the lack of anti-viral therapy.20, 21 Accordingly, Criscuoli et al. showed no benefit of anti-viral treatment in three patients out of seven in-patients (43%) with severe acute colitis complicated with cytoplasmic CMV inclusions on colonic biopsies.5 In 2004, De Saussure et al. examined prospectively 3 patients with active UC associated with positive CMV viremia. Anti-viral treatment failed to significantly improve the flare in two of these patients.20 In a retrospective study analysing 77 UC patients who required colonic resection, Maconi et al. addressed the issue of a putative association between CMV infection and steroid refractoriness. The study reassessed the strong association between ongoing use of steroids and CMV infection, yet failed to show a clear link between CMV detection on immunohistochemical staining and steroid refractoriness. First, CMV detection rate was not significantly higher in patients with steroid-refractory UC compared with patients with nonrefractory UC. More importantly, unrecognised and untreated CMV infection before surgery did not impact the outcome of UC patients after coloproctectomy.22 On both sides of the debate, studies have suffered from major limitations, preventing definitive answers: retrospective analysis, small populations, lack of uniformity between detection methods, lack of established significance cut-offs for tissue and plasmatic PCR CMV, need for clear definitions for CMV reactivation, CMV superinfection and CMV disease. Thus, we believe, the results observed in our multicentre, large-population, retrospective study ought to be confirmed with a randomised, prospective study testing the benefit of anti-viral treatment in IBD patients in flare. In conclusion, although an association between active IBD on immunosuppressants and CMV reactivation has been clearly established, this study showed no deleterious
role of CMV in patients with IBD flare-ups on disease course. No benefit of viral clearance using anti-viral treatment was demonstrated in terms of inflammation control, length of hospital stay or colectomy rate, supporting the role of CMV as a mere marker of inflammation severity. Our results indicate the need to be confirmed in randomised controlled trials. Pending these results, antiviral therapy, with potential toxicity, could then be limited to severe CMV diseases, involving systemic manifestations.
AUTHORSHIP Guarantor of the article: Harry Sokol. Author contributions: MD performed the research, collected the data and wrote the paper. HS designed the study and analysed the data. AL, SP, AB, PS, JC, FC, JG, LB, XR, LPB collected the data and reviewed the manuscript. All authors approved the final version of the manuscript. ACKNOWLEDGEMENTS Declaration of personal interests: Philippe SEKSIK received consulting fees from Biocodex, MSD and Abbott. Jacques Cosnes received consulting fees from Abbvie. Laurent Beaugerie received consulting fees from Abbvie, lecture fees from Merck and Abbvie and unconditional research grants from Biocodex, Ferring Pharmaceuticals and Abbvie. Xavier Roblin received consulting fees from MSD, Abbvie, Norgine, Ferring, Theradiag, Takeda. Laurent Peyrin-Biroulet received consulting and/ or lecture fees from Merck, Abbott, Janssen, Genentech, Mitsubishi, Ferring, Norgine, Tillots, Vifor, Shire, Therakos, Pharmacosmos. Pilege, BMS, UCB-pharma, Hospira, Takeda. Harry Sokol received consulting fees from Danone and Enterome. Declaration of funding interests: None.
REFERENCES 1. Maher MM, Nassar MI. Acute cytomegalovirus infection is a risk factor in refractory and complicated inflammatory bowel disease. Dig Dis Sci 2009; 54: 2456–62. 2. Taylor GH. Cytomegalovirus. Am Fam Physician 2003; 67: 519–24. 3. Powell RD, Warner NE, Levine RS, Kirsner JB. Cytomegalic inclusion disease and ulcerative colitis; report of a case in a young adult. Am J Med 1961; 30: 334–40. Aliment Pharmacol Ther 2014; 39: 712-720 ª 2014 John Wiley & Sons Ltd
4. Papadakis KA, Tung JK, Binder SW, et al. Outcome of cytomegalovirus infections in patients with inflammatory bowel disease. Am J Gastroenterol 2001; 96: 2137–42. 5. Criscuoli V, Casa A, Orlando A, et al. Severe acute colitis associated with CMV: a prevalence study. Dig Liver Dis 2004; 36: 818–20. 6. Kandiel A, Lashner B. Cytomegalovirus colitis complicating inflammatory bowel
disease. Am J Gastroenterol 2006; 101: 2857–65. 7. Cottone M, Pietrosi G, Martorana G, et al. Prevalence of cytomegalovirus infection in severe refractory ulcerative and Crohn’s colitis. Am J Gastroenterol 2001; 96: 773–5. 8. Kishore J, Ghoshal U, Ghoshal UC, et al. Infection with cytomegalovirus in patients with inflammatory bowel disease: prevalence, clinical significance
719
M. Delvincourt et al.
9.
10.
11.
12.
13.
720
and outcome. J Med Microbiol 2004; 53 (Pt 11): 1155–60. Yoshino T, Nakase H, Ueno S, et al. Usefulness of quantitative real-time PCR assay for early detection of cytomegalovirus infection in patients with ulcerative colitis refractory to immunosuppressive therapies. Inflamm Bowel Dis 2007; 13: 1516–21. Rahier JF, Ben-Horin S, Chowers Y, et al. European evidence-based Consensus on the prevention, diagnosis and management of opportunistic infections in inflammatory bowel disease. J Crohns Colitis 2009; 3: 47–91. Pillet S, Pozzetto B, Jarlot C, Paul S, Roblin X. Management of cytomegalovirus infection in inflammatory bowel diseases. Dig Liver Dis 2012; 44: 541–8. Hommes DW, Sterringa G, van Deventer SJ, Tytgat GN, Weel J. The pathogenicity of cytomegalovirus in inflammatory bowel disease: a systematic review and evidence-based recommendations for future research. Inflamm Bowel Dis 2004; 10: 245–50. Matsuoka K, Iwao Y, Mori T, et al. Cytomegalovirus is frequently
14.
15.
16.
17.
18.
reactivated and disappears without antiviral agents in ulcerative colitis patients. Am J Gastroenterol 2007; 102: 331–7. Domenech E, Vega R, Ojanguren I, et al. Cytomegalovirus infection in ulcerative colitis: a prospective, comparative study on prevalence and diagnostic strategy. Inflamm Bowel Dis 2008; 14: 1373–9. Widmann T, Sester U, Gartner BC, et al. Levels of CMV specific CD4 T cells are dynamic and correlate with CMV viremia after allogeneic stem cell transplantation. PLoS ONE 2008; 3: e3634. Forbes BA, Bonville CA, Dock NL. The effects of a promoter of cell differentiation and selected hormones on human cytomegalovirus infection using an in vitro cell system. J Infect Dis 1990; 162: 39–45. Kambham N, Vij R, Cartwright CA, Longacre T. Cytomegalovirus infection in steroid-refractory ulcerative colitis: a case-control study. Am J Surg Pathol 2004; 28: 365–73. Pfau P, Kochman ML, Furth EE, Lichtenstein GR. Cytomegalovirus
19.
20.
21.
22.
colitis complicating ulcerative colitis in the steroid-naive patient. Am J Gastroenterol 2001; 96: 895–9. Leveque N, Brixi-Benmansour H, Reig T, et al. Low frequency of cytomegalovirus infection during exacerbations of inflammatory bowel diseases. J Med Virol 2010; 82: 1694– 700. de Saussure P, Lavergne-Slove A, Mazeron MC, Alain S, Matuchansky C, Bouhnik Y. A prospective assessment of cytomegalovirus infection in active inflammatory bowel disease. Aliment Pharmacol Ther 2004; 20: 1323–7. Kim JJ, Simpson N, Klipfel N, Debose R, Barr N, Laine L. Cytomegalovirus infection in patients with active inflammatory bowel disease. Dig Dis Sci 2010; 55: 1059–65. Maconi G, Colombo E, Zerbi P, et al. Prevalence, detection rate and outcome of cytomegalovirus infection in ulcerative colitis patients requiring colonic resection. Dig Liver Dis 2005; 37: 418–23.
Aliment Pharmacol Ther 2014; 39: 712-720 ª 2014 John Wiley & Sons Ltd
The impact of cytomegalovirus reactivation and its treatment on the course of inflammatory bowel disease M. Delvincourt*, A. Lopez†, S. Pillet‡,§, A. Bourrier*, P. Seksik*,¶, J. Cosnes*, F. Carrat**, J. Gozlan††, L. Beaugerie*, X. Roblin§,‡‡, L. Peyrin-Biroulet† & H. Sokol*,¶,§§
*Department of Gastroenterology, Saint Antoine Hospital, Assistance Publique H^ opitaux de Paris and Paris VI University, Paris, France. † Department of Hepatogastroenterology, Nancy University Hospital, Vandoeuvre-lesNancy, France. ‡ Department of Bacteriology-VirologyHygiene, Saint Etienne Hospital, Saint Etienne, France. § GIMAP EA 3064, Faculte de medecine de Saint-Etienne, Universite de Lyon, Lyon, France. ¶ Laboratoire INSERM U1057/UMR CNRS 7203, Universite Pierre et Marie Curie 6, Paris, France. **Public Health department, SaintAntoine Hospital, AP-HP Paris; UMRS 707, Universite Pierre et Marie Curie-Paris 6 & INSERM, Paris, France. †† Department of Virology, Saint Antoine Hospital, Assistance Publique H^opitaux de Paris and Paris VI University, Paris, France. ‡‡ Department of Gastroenterology, Saint Etienne Hospital, Saint Etienne, France. §§ Commensal and Probiotics-Host Interactions Laboratory, UMR 1319 Micalis, Jouy-en-Josas, France. Correspondence to: Dr H. Sokol, Service de Gastroenterologie et Nutrition, H^ opital Saint-Antoine 184 rue du faubourg Saint-Antoine, 75571 Paris CEDEX 12, France. E-mail: [email protected] Publication data Submitted 27 November 2013 First decision 3 January 2014 Resubmitted 14 January 2014 Accepted 15 January 2014 EV Pub Online 9 February 2014 This article was accepted for publication after full peer-review.
712
SUMMARY Background Consequences of latent cytomegalovirus (CMV) infection reactivation on inflammatory bowel disease (IBD) flare, as a flare-worsening factor or simple bystander, are debated. Impact of anti-viral treatment on IBD course is poorly known. Aim To assess the impact of CMV reactivation on patients hospitalised for IBD flare and the effect of anti-viral treatment on IBD flare in patients with CMV reactivation. Methods First, a population of UC patients from Saint-Antoine hospital, in flare with positive blood CMV PCR without anti-viral treatment (n = 26), were compared to matched patients with negative blood CMV PCR in a case–control study. Secondly, a total of 110 hospitalisations between October 2003 and May 2012 for IBD flare-up with CMV reactivation (80 diagnosed on blood PCR, 33 on tissue PCR) were identified in three French referral centres. Evolution following CMV reactivation diagnosis was compared between patients receiving anti-viral treatment and those who did not. Results In the case–control study, no differences were observed between the two groups regarding length of hospital stay and colectomy rate. Comparing treated and untreated patients, no differences were observed at inclusion regarding age, gender, IBD type, immunosuppressant, CRP and haemoglobin level. No differences were observed regarding CRP level decrease at 10 days and colectomy rate at 3 months. Anti-viral treatment was associated with lower serum albumin level at inclusion and longer hospitalisation. Conclusions CMV reactivation does not appear to alter the course of IBD flare. CMV treatment does not seem to impact the course of IBD. These results should be confirmed prospectively. Aliment Pharmacol Ther 2014; 39: 712–720
ª 2014 John Wiley & Sons Ltd doi:10.1111/apt.12650
CMV reactivation in IBD INTRODUCTION Cytomegalovirus (CMV) is a member of the Herpesviridae family. Infection, through close contact with different body fluids, is common in all human populations, with seroprevalence rates ranging from 40% to 100%.1 CMV disease manifestations in infected humans are numerous, from the most common asymptomatic form in immunocompetent patients to the severe CMV-induced interstitial pneumonitis in allogeneic bone marrow transplant recipients. After primary infection, CMV persists in a latent state in neutrophils, T lymphocytes, endothelial cells and renal epithelial cells. Reactivation can be triggered by immunosuppression. Thus, at-risk populations include transplanted patients, HIV-positive patients and, more largely, patients receiving immunosuppressive therapy.2 The first case of ulcerative colitis (UC)-associated CMV infection was reported in 1961 in a patient exhibiting viral cytoplasmic inclusions on colonic biopsies.3 Indeed, patients with inflammatory bowel disease (IBD) are often immunosuppressed, through combined action of potentially severe malnutrition, use of immunosuppressants and primary impairment of immune functions. These factors, associated with a CMV tropism for inflammatory tissues, lead to a significant increased risk of developing CMV reactivation.4 Thus, prevalence of CMV reactivation during moderate-to-severe IBD flare-up ranges from 21% to 34% in different studies.5, 6 Among steroid-refractory IBD patients, prevalence increases beyond 30%.1, 7 During IBD flare-up, two distinct studies have identified, on multivariate analysis, the female gender, a pancolic inflammatory disease and active inflammation on histology as independent risk factors for CMV reactivation.1, 8 Various methods are used to detect CMV reactivation. Quantitative whole-blood and plasmatic CMV PCR and histological analysis of colonic biopsies have shown a powerful correlation5, 9 and are often considered the gold standard methods.6 So far, European guidelines (European Crohn’s and Colitis Organization guidelines, 2009) recommended the use of immunohistochemistry and tissue CMV PCR to detect the virus in immunosuppressant-refractory IBD patients.10 Despite the clear association between IBD flare-up and CMV reactivation, the impact of the latter on the former and vice versa is debated.11 Several studies and meta-analysis have suggested a deleterious role of CMV reactivation in IBD flare-ups, increasing the risk for steroid resistance and need for colectomy.12 On the other
Aliment Pharmacol Ther 2014; 39: 712-720 ª 2014 John Wiley & Sons Ltd
hand, other studies have considered CMV as a simple marker of digestive inflammation severity, disappearing under appropriate immunosuppressive treatment.13 In this context, we aimed to evaluate the impact of anti-viral therapy against CMV on IBD course in a group of IBD patients hospitalised for IBD flare-ups complicated with CMV reactivation compared with a group of patients with IBD flare and CMV reactivation who did not receive anti-viral therapy.
MATERIAL AND METHODS A population of UC patients in flare with significant blood CMV viral load without anti-viral treatment (n = 26) entered a retrospective case–control study. In this study, patients were matched 1:1 according to sex, age 5 years and year of hospitalisation in Saint Antoine hospital to patients with UC flare-up with negative blood CMV PCR. Secondly, in the retrospective study, all consecutive patients hospitalised in French tertiary centres of Paris (Saint Antoine Hospital), Nancy and Saint Etienne, between October 2003 and May 2012, for IBD flare-up complicated with CMV reactivation, were included. The diagnosis of IBD was based on clinical, endoscopic, radiological and histological parameters. CMV reactivation was assessed by positive blood CMV PCR and/or positive tissue CMV PCR on colonic biopsies. Viral load significance thresholds were defined in each centre according to the techniques used (500–2500 copies/mL for PCR in blood and semi-quantitative method for PCR in colonic biopsies). These values were determined in accordance with established significant thresholds for CMV viral loads used to diagnose CMV infection in the setting of bone marrow transplants in each centre. When several CMV PCR were positive during a single hospital stay, the first CMV PCR above the significance threshold was considered. In patients for whom anti-viral treatment was initiated, the CMV viral load preceding the anti-viral treatment introduction was considered. Demographic (age, sex, IBD type, extent of the disease), clinical, biological, endoscopic and therapeutic (use of corticosteroids and other ongoing immunosuppressant treatments) profiles of all investigated patients were retrospectively collected. Clinical evaluation at inclusion was based on specific clinical severity indexes, performed on routine clinical practice: Lichtiger score for UC and Harvey–Bradshaw index for Crohn’s disease. Endoscopic evaluation included registering of severe colitis features, such as wide mucosal defects, punched-out ulcerations
713
M. Delvincourt et al. and mucosal detachment. Severe colitis, according to both clinical (Lichtiger score superior or equal to 10; Harvey–Bradshaw index superior to 16) and endoscopic criteria, were recorded. Blood haemoglobin, C-reactive protein (CRP) and albumin levels at inclusion were extracted from patient files for the purpose of the study. All clinical and biological criteria were re-assessed on day 10 4 after inclusion, allowing the calculation of CMV viral load drops, Lichtiger score drops and CRP level drops between inclusion and re-evaluation, in absolute values and percentage. Finally, length of hospital stay, the need for colectomy and, if so, the time from inclusion to colectomy, were assessed. Patients receiving anti-viral treatment and those without anti-viral treatment were compared in terms of disease type, flare-up severity and use of immunosuppressants at inclusion. Disease evolution was assessed and compared between the 2 groups according to evolution of CRP level between inclusion and re-evaluation, average hospital stay and 3-month colectomy rates. For statistical analysis, qualitative variables expressed in absolute values and percentages were compared using Chi-squared test or Fisher’s exact test when appropriate. Quantitative variables, expressed in mean value with standard deviation were compared using Student t-test or Wilcoxon test. The results were considered statistically significant when P values were below 0.05. Statistical analysis of the data was carried out using JMP software version 7.0 (USA).
CMV + (n = 26) Sex (%): M/F 14 (54%)/12 (46%) Age (years s.d.) 36.8 12.3 Duration of disease 4.2 5.1 (years s.d.) Ongoing immunosuppressants on inclusion Steroids (%) 15 (60%) Steroid dosage 30 28 Time on steroids (d) 32 56 Azathioprine (%) 10 (40%) Methotrexate (%) 1 (4%) Ciclosporin (%) 2 (8%) Anti-TNFa (%) 5 (20%) Lichtiger index s.d. 9.5 4.2 Lichtiger ≥10 (%) 16 (61.5%) Endoscopic Mayo index 2.3 0.7
CMV
RESULTS Case–control study A population of 26 UC patients with positive blood CMV PCR was matched for sex, age 5 years and year of hospitalisation to 26 UC patients with negative CMV PCR. The two groups were similar in terms of age at IBD diagnosis, duration of IBD, ongoing steroid therapy at inclusion, clinical and endoscopic severity as defined by Lichtiger and Mayo score respectively (Table 1). No significant difference was observed between the two groups regarding length of hospital stay (8.7 days vs. 9.7 days for CMV-positive and CMV-negative patients respectively; P 0.42) and 3-month colectomy rate (15.4% and 23.1% for CMV-positive and CMV-negative patients respectively; P 0.48). Retrospective study A total of 110 hospitalisations (in 105 patients) between October 2003 and May 2012 for IBD flare-up with CMV reactivation were included. Patient characteristics are detailed in Table 2. Mean age at inclusion was 40.4 14.4 years. Eighty-seven patients (79%) had UC. As defined by Lichtiger score (≥10), 43 UC patients (52% of all UC patients) had severe colitis at the time of inclusion. Sixty-five per cent of UC patients presented with pancolitis. Treatment at inclusion included steroids, azathioprine and anti-TNF alpha in 70%, 44% and 26% of cases respectively.
(n = 26)
P
14 (54%)/12 (46%) 38.6 12.1 7.6 7.9
1 0.53 0.08
14 (56%) 30 32 10 17 2 (8%) 1 (4%) 0 1 (4%) 10.4 3.2 17 (65.4%) 2.1 0.9
1 0.85 0.22 0.02 1 0.49 0.19 0.61 0.77 0.56
Table 1 | Baseline and demographic characteristics of ulcerative colitis patients in flare, with or without positive blood CMV PCR
CMV, cytomegalovirus; PCR, polymerase chain reaction; M, male; F, female. Categorical variables reported as absolute values and %. Continuous variables reported as mean standard derivation. 714
Aliment Pharmacol Ther 2014; 39: 712-720 ª 2014 John Wiley & Sons Ltd
CMV reactivation in IBD Table 2 | Baseline and demographic characteristics of IBD patients in flare with CMV reactivation diagnosed on blood and/or tissue PCR (n = 110) Sex ratio (M/F) (%) Mean age (years) IBD type (%): CD/UC Montreal classification Age at diagnosis (% CD): A1/A2/A3 Location UC (%) E1 E2 E3 CD (%) L1 L2 L3 Behaviour (% CD): B1/B2/B3/P Severity index UC and IBDU: Lichtiger index [range] CD: Harvey–Bradshaw [range] Lichtiger ≥10 (% UC) CRP (mg/L) Hb (g/dL) Albumin (g/L) Endoscopic severe colitis (Mayo index = 3) (%) Treatments at inclusion Corticosteroids (%) Azathioprine (%) Ciclosporin (% of UC) Anti-TNF alpha (%) Anti-viral therapy (%)
57/43 40.4 14.7 23 (21%)/87 (79%) 9%/83%/35%
5% 30% 65% 18% 30% 52% 59%/9%/32%/30.4% 8 [0–18] 8 [3–20] 43 (52%) 33.5 33.8 11.1 2.1 29.1 5.8 61%
73 (70%) 46 (44%) 6 (7%) 27 (26%) 68%
IBD, inflammatory bowel disease; CD, Crohn’s disease; UC, ulcerative colitis; IBDU, inflammatory bowel disease type unclassified; CRP, C-reactive protein; Hb, Haemoglobin. Categorical variables reported as absolute values and %. Continuous variables reported as mean standard derivation.
CMV reactivation was diagnosed on blood PCR for 80 patients and on tissue PCR in 33 patients. In the last group of patients with positive tissue PCR, 3 had known positive blood PCR CMV as well, while no data regarding blood viral load were available for the others. Anti-viral therapy was initiated in 68% of these patients. Regarding patients with positive blood CMV PCR, there were no significant differences between the treated and the nontreated groups in terms of gender, age, duration of disease or IBD type (Table 3). Distribution of IBD treatments at inclusion was similar between the two groups. Among UC patients, Lichtiger score was similar, and a comparable rate of clinically severe acute colitis was observed between the two groups of subjects. Aliment Pharmacol Ther 2014; 39: 712-720 ª 2014 John Wiley & Sons Ltd
All treated patients experienced a decrease in viral load at 10 days, 63% of them with a control blood CMV PCR below established thresholds (data not shown). Regarding biological parameters, no statistically significant difference was observed in terms of CRP and haemoglobin levels at inclusion between treated and untreated patients. Albumin levels differed significantly, with lower levels observed in the treated group (P = 0.03). Ten days (4) after inclusion, CRP level drops did not differ between the two groups (Figure 1). Hospital stay was significantly longer among patients who received anti-viral treatment (16.3 days vs. 8.3 days; P < 0.001). No difference was observed in terms of colectomy rate at 3 months after inclusion (10.6% vs. 13.3%; P = 0.7). We then performed a subpopulation analysis restricted to the 25 patients with severe UC flare-up (defined by having a Lichtiger index ≥10 at inclusion), with positive blood CMV PCR (15 treated, 10 untreated). Demographic data and ongoing immunosuppressive treatment at inclusion were comparable between the two groups (Table 4). The treated group (n = 15) did not differ from the untreated group (n = 10) in terms of clinical severity (Lichtiger score: 11.9 2.0 vs. 12.3 1.8, respectively; P = 0.48), but had higher biological markers of severity bordering significance (CRP: 42.9 47.5 vs. 17.6 19.2; P = 0.05). On re-evaluation 10 days later, CRP level drop and Lichtiger score drop were comparable between treated and untreated patients (Figure 2). Mean hospital stay was significantly longer among treated patients (15.3 days vs. 8.8 days; P = 0.04). No difference was observed in terms of 3-month colectomy rate (15.0% vs. 10.0%; P = 0.9). Among the 33 patients with positive tissue CMV PCR, 22 received anti-viral treatment, while 11 patients did not (Table 5). The two groups of patients did not differ in terms of sex, age or ongoing treatments at inclusion. Assessment of severity showed no significant difference between the two groups regarding clinical and biological parameters (data not shown). Comparison of CRP level drop, length of hospital stay and 3-month colectomy rate showed no benefit for the anti-viral therapy (data not shown).
DISCUSSION The role of CMV reactivation and thus the usefulness of anti-viral therapy in IBD flare are highly debated. In this case–control study, comparing UC patients with positive blood CMV PCR who did not receive anti-viral 715
M. Delvincourt et al. Table 3 | Baseline and demographic characteristics of patients with positive blood PCR CMV, treated or not with anti-viral treatment Anti-viral treatment (n = 49) Sex (%): M/F Age (years s.d.) IBD type (%): CD/UC Immunosuppressive treatment Steroids (%) Azathioprine (%) Methotrexate (%) Ciclosporin (%) Anti-TNFa (%) CRP (mg/L) Haemoglobin (g/dL) Albumin (g/L) Lichtiger index (UC) Lichtiger index ≥10 (% of UC)
No anti-viral treatment (n = 31)
23 (47%)/26 (53%) 49 37 13 (27%)/36 (73%)
18 (58%)/13 (42%) 31 38.8 10 (32%)/21 (68%)
29 (60%) 18 (38%) 2 (4%) 3 (6%) 15 (31%) 39.2 36.2 10.7 2.1 26.8 4.6 8.1 3.9 11 (52%)
21 (68%) 15 (48%) 2 (6%) 3 (10%) 5 (16%) 28 27 11.5 1.9 30.1 5.9 8.1 4.1 10 (48%)
P 0.33 0.29 0.58 0.51 0.34 0.65 0.58 0.13 0.09 0.09 0.03 1 1
Categorical variables reported as absolute values and %. Continuous variables reported as mean standard derivation.
NS
P = 0.014
P = 0.021
250 CRP (mg/L)
CRP (mg/L)
150
100
50
200 150 100 50 0
0 D0
D 10
No antiviral treatment
D0
D 10
Antiviral treatment
Figure 1 | CRP (mg/L) level evolution between inclusion and re-evaluation (Day 10 4) in IBD patients in flare with positive blood CMV PCR, treated or not with anti-viral treatment.
treatment (n = 26) and a matched control group of UC patients without detected CMV in blood (n = 26), no difference between the two groups was observed in terms of clinical and endoscopic severity, length of hospital stay or colectomy rate. Of note, patients with positive CMV PCR received the same amount of steroids (similar percent of patients treated, similar dosage, similar treatment length) compared with the control group. These results are limited by the retrospective and unrandomised design of the study. To help assess the significance of these first results, a retrospective study evaluating 716
the impact of anti-viral treatment on 110 patients hospitalised for IBD flare-up complicated with blood and/or tissue CMV reactivation was conducted. Anti-viral treatment failed to demonstrate a favourable impact on clinical or biological outcomes, length of hospital stay or colectomy rate at 3 months, thus strengthening the role of CMV as an innocent bystander. On the contrary, longer hospital stays were observed among patients receiving anti-viral therapy as compared with patients without anti-viral treatment, suggesting a higher severity of the treated population at inclusion, thus representing a bias in Aliment Pharmacol Ther 2014; 39: 712-720 ª 2014 John Wiley & Sons Ltd
CMV reactivation in IBD Table 4 | Baseline and demographic characteristics of patients with severe UC flare-up (Lichtiger ≥10) and positive blood PCR CMV, treated or not with anti-viral treatment Anti-viral treatment (n = 15)
No anti-viral treatment (n = 10)
P
7 (47%)/8 (53%) 36.5 11.5
5 (50%)/5 (50%) 35.7 10.8
0.87 0.72
12 (80%) 3 (20%) 1 (7%) 6 (40%) 11.9 2.0 42.9 47.5 10.7 2.2 27.1 5.6
7 (70%) 5 (50%) 0 (0%) 2 (20%) 12.3 1.8 17.6 19.2 11.1 2.3 30.6 5.7
0.57 0.12 0.41 0.29 0.48 0.05 0.64 0.12
Sex (%): M/F Age (years s.d.) Immunosuppressive treatment Steroids (%) Azathioprine (%) Ciclosporin (%) Anti-TNFa (%) Lichtiger index (UC) CRP (mg/L) Haemoglobin (g/dL) Albumin (g/L)
Categorical variables reported as absolute values and %. Continuous variables reported as mean standard derivation.
(a)
NS P = 0.014
60 40 20
150 100 50
0
0 D0
D 10
(b)
D0
D 10
NS P = 0.009
20
P = 0.003
20
15
Lichtiger score
Lichtiger score
P = 0.021
200
CRP (mg/L)
CRP (mg/L)
80
10 5
15 10 5 0
0 D0
D 10
No antiviral treatment
D0
D 10
Antiviral treatment
Figure 2 | CRP level and Lichtiger score evolution between inclusion and re-evaluation (Day 10 4) in patients with severe UC flares and positive blood CMV PCR, treated or not with anti-viral treatment. Aliment Pharmacol Ther 2014; 39: 712-720 ª 2014 John Wiley & Sons Ltd
717
M. Delvincourt et al. Table 5 | Baseline and demographic characteristics of UC patients in flare with positive tissue PCR CMV, treated or not with anti-viral treatment
Sex (%): M/F Age (years s.d.) IBD type (%): CD/UC Immunosuppressive treatment Steroids (%) Azathioprine (%) Ciclosporin (%) Anti-TNFa (%) Lichtiger index (UC) CRP (mg/L) Haemoglobin (g/dL)
Anti-viral treatment (n = 22)
No anti-viral treatment (n = 11)
P
17 (77.3%)/5 (22.7%) 41.2 20.1 0/22 (100%)
7 (63.6%)/4 (36.4%) 52.5 13 0/11 (100%)
0.41 0.29 1
18 (85.7%) 10 (45.5%) 1 (4.6%) 6 (27.3%) 12.7 2.8 53.1 29.8 11.1 1.8
9 (81.8%) 6 (54.6%) 0 1 (9.1%) 13.2 1.3 41.3 21.4 11.4 1.3
0.77 0.34 0.47 0.23 0.97 0.28 0.44
Categorical variables reported as absolute values and %. Continuous variables reported as mean standard derivation.
the study. Indeed, even though clinical and biological status at inclusion did not statistically differ between the two groups, other unassessed parameters, such as hemodynamic status or extra-digestive functions, may have been more severe in the treated population, explaining why these patients were selected to benefit from the anti-viral treatment. However, these results were confirmed when considering only a sub-population of severe UC patients (Lichtiger index ≥10). The large population of this retrospective study is in accordance with available epidemiological data on this topic. We observed a clear predominance of UC patients with mild to severe flare-ups, mostly pancolitis (Table 2).1, 8 A high rate of ongoing steroid therapy at inclusion is in accordance with the previously demonstrated association between this treatment and CMV reactivation. It is not clearly established whether this immunosuppressive treatment facilitates CMV reactivation or if it merely reflects the severity of the underlying inflammatory disease.14 Widmann et al. have suggested potential in vitro dose-dependent effects of steroids on suppression of CMV-specific T-lymphocyte function.15 Other works have confirmed increased viral proteins production in in vitro hydrocortisone-enriched environment.16 CMV reactivation has been witnessed in up to 30% of patients with severe and/or steroid-refractory colitis.14 Current clinical guidelines recommend testing for colonic CMV disease in patients with severe colitis and deliver anti-viral treatment to CMV-positive patients.10 Indeed, several studies have described the potentially pathogenic role of CMV on IBD flare-ups. Thus, Kambham et al. showed, among 40 patients suffering from steroid-refractory ulcerative colitis, a CMV disease 718
assessed by histological features on colonic biopsies in 10 patients (25%), whereas virus activity was demonstrated in only one of 40 patients from a matched population of patients with ulcerative colitis responding to steroids (2,5%).17 Previous works have acknowledged CMV to be an independent risk factor for medical treatment failure and toxic megacolon, both necessitating emergency colectomy.8, 14 Kambham and al. also showed a potential positive impact of the anti-viral treatment, allowing colonic viral clearance among three UC patients treated for CMV reactivation, associated with clinical remission of the ulcerative colitis with steroid suspension. On the contrary, six patients out of 8 UC patients with untreated CMV superinfection necessitated coloprotectomy.17 On the other hand, several recent studies have hypothesised the role of CMV as an innocent bystander of intestinal lesions, with no pathogenic activity by itself. Indeed, in vitro studies have shown spontaneous tropism of CMV towards proliferating cells in inflammatory tissues.18 Thus, Matsuoka and al. compared in 2007 a population of UC patients with mild-to-severe flare-up, complicated with CMV reactivation, assessed by plasmatic PCR and/or pp65 antigenemia (n = 25), with a control population of patients matched for disease activity, without detected CMV (n = 23). Under immunosuppressive treatment alone, without anti-viral treatment, no difference was shown between the two groups in terms of 8-week clinical remission rate and colectomy rate.13 In a French retrospective study, seven IBD patients (three ulcerative colitis, four Crohn’s diseases) with mild-to-severe flare-ups were diagnosed with CMV reactivation assessed on blood PCR. Five of them (71%) showed favourable response to immunosuppressants alone.19 Other studies have demonstrated possible Aliment Pharmacol Ther 2014; 39: 712-720 ª 2014 John Wiley & Sons Ltd
CMV reactivation in IBD clinical improvement in spite of the lack of anti-viral therapy.20, 21 Accordingly, Criscuoli et al. showed no benefit of anti-viral treatment in three patients out of seven in-patients (43%) with severe acute colitis complicated with cytoplasmic CMV inclusions on colonic biopsies.5 In 2004, De Saussure et al. examined prospectively 3 patients with active UC associated with positive CMV viremia. Anti-viral treatment failed to significantly improve the flare in two of these patients.20 In a retrospective study analysing 77 UC patients who required colonic resection, Maconi et al. addressed the issue of a putative association between CMV infection and steroid refractoriness. The study reassessed the strong association between ongoing use of steroids and CMV infection, yet failed to show a clear link between CMV detection on immunohistochemical staining and steroid refractoriness. First, CMV detection rate was not significantly higher in patients with steroid-refractory UC compared with patients with nonrefractory UC. More importantly, unrecognised and untreated CMV infection before surgery did not impact the outcome of UC patients after coloproctectomy.22 On both sides of the debate, studies have suffered from major limitations, preventing definitive answers: retrospective analysis, small populations, lack of uniformity between detection methods, lack of established significance cut-offs for tissue and plasmatic PCR CMV, need for clear definitions for CMV reactivation, CMV superinfection and CMV disease. Thus, we believe, the results observed in our multicentre, large-population, retrospective study ought to be confirmed with a randomised, prospective study testing the benefit of anti-viral treatment in IBD patients in flare. In conclusion, although an association between active IBD on immunosuppressants and CMV reactivation has been clearly established, this study showed no deleterious
role of CMV in patients with IBD flare-ups on disease course. No benefit of viral clearance using anti-viral treatment was demonstrated in terms of inflammation control, length of hospital stay or colectomy rate, supporting the role of CMV as a mere marker of inflammation severity. Our results indicate the need to be confirmed in randomised controlled trials. Pending these results, antiviral therapy, with potential toxicity, could then be limited to severe CMV diseases, involving systemic manifestations.
AUTHORSHIP Guarantor of the article: Harry Sokol. Author contributions: MD performed the research, collected the data and wrote the paper. HS designed the study and analysed the data. AL, SP, AB, PS, JC, FC, JG, LB, XR, LPB collected the data and reviewed the manuscript. All authors approved the final version of the manuscript. ACKNOWLEDGEMENTS Declaration of personal interests: Philippe SEKSIK received consulting fees from Biocodex, MSD and Abbott. Jacques Cosnes received consulting fees from Abbvie. Laurent Beaugerie received consulting fees from Abbvie, lecture fees from Merck and Abbvie and unconditional research grants from Biocodex, Ferring Pharmaceuticals and Abbvie. Xavier Roblin received consulting fees from MSD, Abbvie, Norgine, Ferring, Theradiag, Takeda. Laurent Peyrin-Biroulet received consulting and/ or lecture fees from Merck, Abbott, Janssen, Genentech, Mitsubishi, Ferring, Norgine, Tillots, Vifor, Shire, Therakos, Pharmacosmos. Pilege, BMS, UCB-pharma, Hospira, Takeda. Harry Sokol received consulting fees from Danone and Enterome. Declaration of funding interests: None.
REFERENCES 1. Maher MM, Nassar MI. Acute cytomegalovirus infection is a risk factor in refractory and complicated inflammatory bowel disease. Dig Dis Sci 2009; 54: 2456–62. 2. Taylor GH. Cytomegalovirus. Am Fam Physician 2003; 67: 519–24. 3. Powell RD, Warner NE, Levine RS, Kirsner JB. Cytomegalic inclusion disease and ulcerative colitis; report of a case in a young adult. Am J Med 1961; 30: 334–40. Aliment Pharmacol Ther 2014; 39: 712-720 ª 2014 John Wiley & Sons Ltd
4. Papadakis KA, Tung JK, Binder SW, et al. Outcome of cytomegalovirus infections in patients with inflammatory bowel disease. Am J Gastroenterol 2001; 96: 2137–42. 5. Criscuoli V, Casa A, Orlando A, et al. Severe acute colitis associated with CMV: a prevalence study. Dig Liver Dis 2004; 36: 818–20. 6. Kandiel A, Lashner B. Cytomegalovirus colitis complicating inflammatory bowel
disease. Am J Gastroenterol 2006; 101: 2857–65. 7. Cottone M, Pietrosi G, Martorana G, et al. Prevalence of cytomegalovirus infection in severe refractory ulcerative and Crohn’s colitis. Am J Gastroenterol 2001; 96: 773–5. 8. Kishore J, Ghoshal U, Ghoshal UC, et al. Infection with cytomegalovirus in patients with inflammatory bowel disease: prevalence, clinical significance
719
M. Delvincourt et al.
9.
10.
11.
12.
13.
720
and outcome. J Med Microbiol 2004; 53 (Pt 11): 1155–60. Yoshino T, Nakase H, Ueno S, et al. Usefulness of quantitative real-time PCR assay for early detection of cytomegalovirus infection in patients with ulcerative colitis refractory to immunosuppressive therapies. Inflamm Bowel Dis 2007; 13: 1516–21. Rahier JF, Ben-Horin S, Chowers Y, et al. European evidence-based Consensus on the prevention, diagnosis and management of opportunistic infections in inflammatory bowel disease. J Crohns Colitis 2009; 3: 47–91. Pillet S, Pozzetto B, Jarlot C, Paul S, Roblin X. Management of cytomegalovirus infection in inflammatory bowel diseases. Dig Liver Dis 2012; 44: 541–8. Hommes DW, Sterringa G, van Deventer SJ, Tytgat GN, Weel J. The pathogenicity of cytomegalovirus in inflammatory bowel disease: a systematic review and evidence-based recommendations for future research. Inflamm Bowel Dis 2004; 10: 245–50. Matsuoka K, Iwao Y, Mori T, et al. Cytomegalovirus is frequently
14.
15.
16.
17.
18.
reactivated and disappears without antiviral agents in ulcerative colitis patients. Am J Gastroenterol 2007; 102: 331–7. Domenech E, Vega R, Ojanguren I, et al. Cytomegalovirus infection in ulcerative colitis: a prospective, comparative study on prevalence and diagnostic strategy. Inflamm Bowel Dis 2008; 14: 1373–9. Widmann T, Sester U, Gartner BC, et al. Levels of CMV specific CD4 T cells are dynamic and correlate with CMV viremia after allogeneic stem cell transplantation. PLoS ONE 2008; 3: e3634. Forbes BA, Bonville CA, Dock NL. The effects of a promoter of cell differentiation and selected hormones on human cytomegalovirus infection using an in vitro cell system. J Infect Dis 1990; 162: 39–45. Kambham N, Vij R, Cartwright CA, Longacre T. Cytomegalovirus infection in steroid-refractory ulcerative colitis: a case-control study. Am J Surg Pathol 2004; 28: 365–73. Pfau P, Kochman ML, Furth EE, Lichtenstein GR. Cytomegalovirus
19.
20.
21.
22.
colitis complicating ulcerative colitis in the steroid-naive patient. Am J Gastroenterol 2001; 96: 895–9. Leveque N, Brixi-Benmansour H, Reig T, et al. Low frequency of cytomegalovirus infection during exacerbations of inflammatory bowel diseases. J Med Virol 2010; 82: 1694– 700. de Saussure P, Lavergne-Slove A, Mazeron MC, Alain S, Matuchansky C, Bouhnik Y. A prospective assessment of cytomegalovirus infection in active inflammatory bowel disease. Aliment Pharmacol Ther 2004; 20: 1323–7. Kim JJ, Simpson N, Klipfel N, Debose R, Barr N, Laine L. Cytomegalovirus infection in patients with active inflammatory bowel disease. Dig Dis Sci 2010; 55: 1059–65. Maconi G, Colombo E, Zerbi P, et al. Prevalence, detection rate and outcome of cytomegalovirus infection in ulcerative colitis patients requiring colonic resection. Dig Liver Dis 2005; 37: 418–23.
Aliment Pharmacol Ther 2014; 39: 712-720 ª 2014 John Wiley & Sons Ltd
