Portal de Periódicos da CAPES

Cytomegalovirus in Inflammatory Bowel Disease: Time for Another Look?

2009; Elsevier BV; Volume: 137; Issue: 3 Linguagem: Inglês

10.1053/j.gastro.2009.07.014

1528-0012

Kabir Julka, Christina M. Surawicz,

Mycobacterium research and diagnosis

Domenech E, Vega R, Ojanguren I, et al. (Department of Gastroenterology, Hospital Universitari Germans Trias i Pujol, Universitat Autonoma de Barcelona, Badalona, Catalonia, Spain). Cytomegalovirus infection in ulcerative colitis: a prospective comparative study on prevalence and diagnostic strategy. Inflamm Bowel Dis 2008;14:1373–1379.Cytomegalovirus (CMV) infection is a well-known complication in immunosuppressed patients, including after bone marrow transplants and with HIV infection. CMV has been associated with steroid refractoriness in inflammatory bowel disease (IBD), but its role in inducing inflammation and its relationship with immunosuppressive therapy has not been well elucidated. The authors of this study prospectively studied the prevalence of colonic CMV disease in ulcerative colitis (UC) patients compared with control patients. Additionally, they investigated whether the presence of CMV was related to the severity of UC, which immunomodulator therapies were associated with the presence of CMV, and the efficacy of various diagnostic methods in detecting colonic CMV.This prospective, single-center study of 94 subjects and 25 controls was performed over 4 years in Catalonia, Spain. Group 1 (n = 25) included active UC patients requiring hospitalization and IV steroid therapy. Group 2 (n = 19) comprised active UC patients requiring hospitalization and subsequently found to be steroid refractory after 7–10 days of IV steroids. These patients were then placed on IV cyclosporine (CSA). Group 3 (n = 25) consisted of UC patients with inactive disease on maintenance azathioprine. Group 4 (n = 25) included UC patients with inactive disease on maintenance oral mesalamine. Group 5 (n = 25), the control group, were individuals with no history of UC and a normal colonoscopy. Evaluation included blood tests (CMV immunoglobulin [Ig]G antibodies and CMV-pp65 antigenemia). All also had colonoscopy or flexible sigmoidoscopy and rectal biopsies. A single, blinded pathologist reviewed all the biopsies with hematoxylin and eosin staining as well as immunohistochemical (IHC) staining for CMV. Finally, qualitative analysis of CMV-pp67 mRNA of the tissue was used to detect CMV.Overall, 70% of all subjects had a positive serology for CMV IgG, with no significant differences of seropositive prevalence among the groups. None had CMV disease before steroid therapy. Six cases of colonic CMV were diagnosed, all from group 2, the subset of patients with steroid-refractory UC (prevalence, 6/19 [32%]). All of these patients were CMV seropositive. Of the 11 seropositive patients in this steroid-refractory group, 6 (54%) developed colonic CMV, and seropositivity was significantly associated with an increased risk of colonic CMV (P = .03). Five of the 6 patients with colonic CMV had extensive colitis (at least to the transverse colon). Moreover, 5 out of the 6 patients were diagnosed with CMV after being started on CSA.The diagnosis of colonic CMV was made by biopsy in 4 and in the surgical specimen after urgent colectomy in the other 2. Although standard histology was positive in 5 out of 6 patients, biopsies from all 6 patients had positive IHC staining. Colonic CMV-pp67 mRNA was positive in 5 out of 6 cases. CMV-pp65 antigenemia was present in only 2 patients.Three of the 6 patients with CMV required colectomy compared with 2 out of 13 patients with steroid-refractory disease who did not have CMV. Only 1 of these 3 patients had a diagnosis of CMV before colectomy and had received ganciclovir as therapy. The remaining 3 patients who did not require colectomy all received ganciclovir and ultimately achieved remission of disease.In summary, this study showed a prevalence of colonic CMV of 32% in patients with steroid-refractory UC. Seropositivity strongly correlated with development of colonic CMV; no seronegative patients developed colonic CMV. None of the patients who responded to steroids developed colonic CMV. Chronic use of azathioprine or mesalamine did not seem to incur an increased risk for colonic CMV. The authors conclude that CMV disease is due to reactivation of latent virus triggered by steroid or CSA therapy.The authors propose that CMV serostatus be determined before starting systemic steroids in patients with severe UC. Only those seropositive patients who are steroid refractory would need biopsies to look for CMV if the clinical situation merits. Finally, they propose that larger randomized studies be done to determine the role of early antiviral therapy in seropositive, steroid-refractory patients, as well as to determine if outcomes are significantly worse for steroid-refractory patients with colonic CMV compared with those without colonic CMV.CommentThe role of infections in IBD continues to be of great interest, both in pathogenesis of IBD as well as its role in the course of IBD. There are at least 3 important scenarios: infections that mimic IBD, infections that trigger the onset of IBD, and infections that worsen the course of IBD.Infections that Mimic IBDCampylobacter, Salmonella, and Shigella are among the bacterial infections that cause acute diarrheal illnesses and colitis that can mimic IBD; unlike IBD, onset of symptoms is usually acute and stool cultures are diagnostic. Both Aeromonas and Plesiomonas have been associated with chronic colitis, but it is unclear whether they actually cause colitis or trigger IBD. When the diagnosis is in question, colorectal biopsy can be very helpful; chronic changes seen in IBD are very uncommon with infectious colitis. Shiga-toxin Escherichia coli (E coli 0157:H7) with a preference for right-sided colitis can mimic ischemic colitis.Yersinia can mimic Crohn's disease when it presents with chronic diarrhea, ileocolitis, and granulomas on colonic biopsies. Differentiation of tuberculosis and Crohn's disease can be very challenging, especially in patients from countries where tuberculosis is endemic, such as Africa. A recent paper by Almadi et al (Am J Gastroenterol 2009;104:1003–1012) provides an excellent comparison of these 2 diseases. In some cases, surgical resection may be the only way to exclude tuberculosis when it cannot be differentiated from Crohn's disease.Parasitic infections can also mimic IBD. Strongyloides stercoralis can cause segmental inflammation in the small bowel, and less commonly in the colon. Entamoeba histolytica, with its predilection for the ileum and right colon can also mimic Crohn's disease. There are case reports of patients with presumed UC who developed fatal amebiasis after therapy with steroids. The diagnosis of amebiasis may require colorectal biopsy or serology.Infections Trigger IBDPostdysenteric UC was reported in 1966, in a series of patients in whom UC developed after an episode of amebic dysentery (Gut 1966;7:438–443). Since then, many series have documented the onset of IBD after acute gastroenteritis both in retrospective series and small prospective series. A recent paper by Gradel et al (Gastroenterology 2009;137:495–501) examined the risk of IBD in individuals with documented Salmonella or Campylobacter gastrointestinal (GI) infection compared with an age- and gender-matched control group from the same population in Denmark. Patients were followed for up to 15 years. Of 13,324 patients with Salmonella or Campylobacter infection 107/13,324 (1.2%) developed IBD compared with 73/26,648 (0.5%) unexposed individuals. The risk was highest during the first year after infection, but remained elevated up to 15 years after the infection. Interestingly, the risk was similar with both infections, and rates of UC and Crohn's disease were also similar.Infections Worsen IBDFinally, GI infections can precipitate flares of IBD. The most common food- or water-borne pathogens, Campylobacter, Salmonella, and Shigella, can worsen or mimic symptoms in patients with IBD. Several series evaluated the yield of stool culture for enteric pathogens in patients with acute IBD. Retrospective analyses suggest that, although small, the yield of positive cultures for enteric pathogens ranges from 10% to 20% and can be associated with an increased rate of hospitalization for treatment of IBD flares. A series of 55 patients with 62 flares of IBD had a prevalence of 20% for positive cultures, nearly all of which were Clostridium difficile (J Clin Gastroenterol 2004;38:772–775). In another series of 197 patients with IBD and 221 relapse events, pathogens were detected in 8%, again most commonly C difficile (Eur J Gastroenterol Hepatol 2004;16:775–778). A more recent prospective study of 79 patients with active IBD (99 episodes) found infections in 6, most commonly Campylobacter (Navarco).GI viral infections have also been recognized to aggravate IBD, including CMV, herpes virus, and Epstein–Barr virus. The seroprevalence of human CMV virus ranges from 30% to 90% in the general population. Seroprevalence increases with increasing age. Most CMV infection in immunocompetent individuals is likely asymptomatic. However, after a CMV infection the virus is latent within the host and can periodically reactivate. Acute CMV infection is most often an infectious mononucleosis type picture, very similar to Epstein-Barr virus with symptoms of fever, adenopathy, and myalgias. The virus is latent, probably in myeloid cells. The T-cell–mediated response is the primary mechanism that controls CMV replication. Thus, severe CMV disease occurs most often in patients who have cellular immunodeficiencies.Clinically significant CMV infection is most common in individuals who are immunosuppressed, with solid organ transplant, stem cell transplant, and patients with HIV/AIDS. In solid organ transplant recipients, CMV causes a febrile illness with a low white blood cell count and can affect the lungs, gut, liver, and eyes, among other organs. With stem cell transplant, lung, gut, eye infection, and brain infection are less common. With HIV/AIDS, retinitis is much more common, as is enterocolitis and other GI infections. There is also evidence that CMV infection can accelerate progression to AIDS and death in those with HIV infection.With solid organ transplantation, CMV is such an important factor that patients are screened for serostatus with a potential donor. The highest risk of infection is in a seronegative donor who receives an organ from a seropositive individual. Acute symptoms in this setting can include fever, low white blood cell count, malaise, arthralgia, and nonspecific symptoms; diagnosis is made by detecting the virus in the blood and/or in tissues. With stem cell transplantation, viral infections are more frequently caused by reactivation of the virus in the recipient rather than a primary infection. Primary infection occurs in about 30% of seronegative recipients, whereas reactivation occurs in about 80% who were seropositive before transplantation.Diagnostic tests for CMV include detection of the antigen in the blood (pp65 antigenemia assay). Detection of CMV DNA by polymerase chain reaction in blood can quantify viral load. GI disease is diagnosed by detection of CMV in biopsy specimens by histologic detection of inclusion bodies, and/or immunohistochemistry, and/or viral culture. Most experts recommend that at least 1 method be used on biopsies to diagnose CMV in addition to histopathology. Immunohistochemistry can increase the yield in biopsies to 93%. It is important to know that GI disease can occur even if there is no detection of CMV in the blood.The role of CMV in patients with IBD is far less clear than in patients posttransplantation or with HIV infection. Patients with IBD, especially when immunosuppressed, seem to be at increased risk for acute CMV infection or CMV reactivation, but there is debate about its pathogenicity and whether CMV is a significant pathogen. In 1 series of 69 patients with UC, 25 had reactivation of CMV but most did not require antiviral treatment. In 2 patients who were given ganciclovir, the clinical outcome was not affected. Moreover, CMV disappeared without treatment in most of these patients (Am J Gastroenterol 2007;102:331–337).In contrast, other series suggest that CMV is a pathogen. There are many published cases of CMV colitis detected in steroid refractory colitis. In some, symptoms improved with antiviral therapy. In others, colectomy was required. The prevalence of biopsy-proven CMV colitis in the setting of IBD flares is estimated to be from 21% to 36% according to several small prospective case series (Am J Gastroenterol 2006;101:2857–2865). In a recent series, CMV was found in 8/23 patients with "steroid-resistant" IBD (Dig Dis Sci December 2008).The paper by Domenech et al adds a significant experience to our knowledge. A relatively large cohort of patients with UC, both active and inactive disease, was evaluated prospectively. Colonic CMV disease developed only in the steroid refractory group, and only in patients who were seropositive for CMV before therapy. In several, CMV colitis seemed to develop after CSA therapy. In the steroid-refractory group, 5 had extensive disease and 1 required colectomy. This study suggests that the combination of active disease and immunosuppressive therapy predisposes to CMV reactivation or infection. No patient with inactive disease on immunosuppressive or mesalamine maintenance developed CMV.Ultimately, many questions remain. Is the risk of CMV infection dependent on disease type (UC vs Crohn's disease) or disease location (colon vs small intestine)? Reports of CMV in IBD focus on UC suggesting that colonic disease is the major factor. Because CMV is trophic to inflamed tissue, this may be the major determinant. Does the type of immunosuppressive therapy increase the risk (steroids vs azathioprine vs tumor necrosis factor antibodies)? In this and other series, steroid therapy seems to confer the greatest risk in UC patients, although this may be a surrogate for severity of colitis. In this series, CSA was also given before CMV diagnosis in several patients. None of the inactive disease patients developed CMV, suggesting that disease activity was more important than the immunosuppressive therapy. Whether tumor necrosis factor inhibitors play a role remains to be seen. In a small, prospective series, none of the 15 IBD patients given infliximab developed CVM infection (J Clin Virol 2008;43:180–183). What is the best diagnostic method? This paper shows that CMV can be absent in the blood, even when present in the colon, and that IHC on colon biopsies was better than conventional histology. Diagnosis of disease depends on correlation of symptoms and documenting presence of CMV in tissue. What criteria should we use to document disease, as opposed to silent infection, namely, the possible "innocent bystander" role, if this exists? What is the best therapy, when to start and when to stop, and should immune suppression be discontinued? Patients whose disease improves with antiviral therapy strengthens the notion that CMV is a pathogen. However, whether antiviral therapy would be better if steroids or CSA were discontinued is not known. It would be ideal to know more about the role of antiviral therapy.Should we be screening high-risk patients for CMV serostatus, as is done in transplant patients? In this series, only CMV-positive patients developed CMV colitis. More effective therapy and potentially lower costs could be achieved if we could be sure that only this subset could develop CMV colitis and needed screening. Is it possible that preemptive antiviral therapy could help a subset of patients? In summary, we do not know enough about the role of CMV in UC, its possible natural history, and role of antiviral therapy. This paper adds significant experience and raises more important questions. Domenech E, Vega R, Ojanguren I, et al. (Department of Gastroenterology, Hospital Universitari Germans Trias i Pujol, Universitat Autonoma de Barcelona, Badalona, Catalonia, Spain). Cytomegalovirus infection in ulcerative colitis: a prospective comparative study on prevalence and diagnostic strategy. Inflamm Bowel Dis 2008;14:1373–1379. Cytomegalovirus (CMV) infection is a well-known complication in immunosuppressed patients, including after bone marrow transplants and with HIV infection. CMV has been associated with steroid refractoriness in inflammatory bowel disease (IBD), but its role in inducing inflammation and its relationship with immunosuppressive therapy has not been well elucidated. The authors of this study prospectively studied the prevalence of colonic CMV disease in ulcerative colitis (UC) patients compared with control patients. Additionally, they investigated whether the presence of CMV was related to the severity of UC, which immunomodulator therapies were associated with the presence of CMV, and the efficacy of various diagnostic methods in detecting colonic CMV. This prospective, single-center study of 94 subjects and 25 controls was performed over 4 years in Catalonia, Spain. Group 1 (n = 25) included active UC patients requiring hospitalization and IV steroid therapy. Group 2 (n = 19) comprised active UC patients requiring hospitalization and subsequently found to be steroid refractory after 7–10 days of IV steroids. These patients were then placed on IV cyclosporine (CSA). Group 3 (n = 25) consisted of UC patients with inactive disease on maintenance azathioprine. Group 4 (n = 25) included UC patients with inactive disease on maintenance oral mesalamine. Group 5 (n = 25), the control group, were individuals with no history of UC and a normal colonoscopy. Evaluation included blood tests (CMV immunoglobulin [Ig]G antibodies and CMV-pp65 antigenemia). All also had colonoscopy or flexible sigmoidoscopy and rectal biopsies. A single, blinded pathologist reviewed all the biopsies with hematoxylin and eosin staining as well as immunohistochemical (IHC) staining for CMV. Finally, qualitative analysis of CMV-pp67 mRNA of the tissue was used to detect CMV. Overall, 70% of all subjects had a positive serology for CMV IgG, with no significant differences of seropositive prevalence among the groups. None had CMV disease before steroid therapy. Six cases of colonic CMV were diagnosed, all from group 2, the subset of patients with steroid-refractory UC (prevalence, 6/19 [32%]). All of these patients were CMV seropositive. Of the 11 seropositive patients in this steroid-refractory group, 6 (54%) developed colonic CMV, and seropositivity was significantly associated with an increased risk of colonic CMV (P = .03). Five of the 6 patients with colonic CMV had extensive colitis (at least to the transverse colon). Moreover, 5 out of the 6 patients were diagnosed with CMV after being started on CSA. The diagnosis of colonic CMV was made by biopsy in 4 and in the surgical specimen after urgent colectomy in the other 2. Although standard histology was positive in 5 out of 6 patients, biopsies from all 6 patients had positive IHC staining. Colonic CMV-pp67 mRNA was positive in 5 out of 6 cases. CMV-pp65 antigenemia was present in only 2 patients. Three of the 6 patients with CMV required colectomy compared with 2 out of 13 patients with steroid-refractory disease who did not have CMV. Only 1 of these 3 patients had a diagnosis of CMV before colectomy and had received ganciclovir as therapy. The remaining 3 patients who did not require colectomy all received ganciclovir and ultimately achieved remission of disease. In summary, this study showed a prevalence of colonic CMV of 32% in patients with steroid-refractory UC. Seropositivity strongly correlated with development of colonic CMV; no seronegative patients developed colonic CMV. None of the patients who responded to steroids developed colonic CMV. Chronic use of azathioprine or mesalamine did not seem to incur an increased risk for colonic CMV. The authors conclude that CMV disease is due to reactivation of latent virus triggered by steroid or CSA therapy. The authors propose that CMV serostatus be determined before starting systemic steroids in patients with severe UC. Only those seropositive patients who are steroid refractory would need biopsies to look for CMV if the clinical situation merits. Finally, they propose that larger randomized studies be done to determine the role of early antiviral therapy in seropositive, steroid-refractory patients, as well as to determine if outcomes are significantly worse for steroid-refractory patients with colonic CMV compared with those without colonic CMV. CommentThe role of infections in IBD continues to be of great interest, both in pathogenesis of IBD as well as its role in the course of IBD. There are at least 3 important scenarios: infections that mimic IBD, infections that trigger the onset of IBD, and infections that worsen the course of IBD. The role of infections in IBD continues to be of great interest, both in pathogenesis of IBD as well as its role in the course of IBD. There are at least 3 important scenarios: infections that mimic IBD, infections that trigger the onset of IBD, and infections that worsen the course of IBD. Infections that Mimic IBDCampylobacter, Salmonella, and Shigella are among the bacterial infections that cause acute diarrheal illnesses and colitis that can mimic IBD; unlike IBD, onset of symptoms is usually acute and stool cultures are diagnostic. Both Aeromonas and Plesiomonas have been associated with chronic colitis, but it is unclear whether they actually cause colitis or trigger IBD. When the diagnosis is in question, colorectal biopsy can be very helpful; chronic changes seen in IBD are very uncommon with infectious colitis. Shiga-toxin Escherichia coli (E coli 0157:H7) with a preference for right-sided colitis can mimic ischemic colitis.Yersinia can mimic Crohn's disease when it presents with chronic diarrhea, ileocolitis, and granulomas on colonic biopsies. Differentiation of tuberculosis and Crohn's disease can be very challenging, especially in patients from countries where tuberculosis is endemic, such as Africa. A recent paper by Almadi et al (Am J Gastroenterol 2009;104:1003–1012) provides an excellent comparison of these 2 diseases. In some cases, surgical resection may be the only way to exclude tuberculosis when it cannot be differentiated from Crohn's disease.Parasitic infections can also mimic IBD. Strongyloides stercoralis can cause segmental inflammation in the small bowel, and less commonly in the colon. Entamoeba histolytica, with its predilection for the ileum and right colon can also mimic Crohn's disease. There are case reports of patients with presumed UC who developed fatal amebiasis after therapy with steroids. The diagnosis of amebiasis may require colorectal biopsy or serology. Campylobacter, Salmonella, and Shigella are among the bacterial infections that cause acute diarrheal illnesses and colitis that can mimic IBD; unlike IBD, onset of symptoms is usually acute and stool cultures are diagnostic. Both Aeromonas and Plesiomonas have been associated with chronic colitis, but it is unclear whether they actually cause colitis or trigger IBD. When the diagnosis is in question, colorectal biopsy can be very helpful; chronic changes seen in IBD are very uncommon with infectious colitis. Shiga-toxin Escherichia coli (E coli 0157:H7) with a preference for right-sided colitis can mimic ischemic colitis. Yersinia can mimic Crohn's disease when it presents with chronic diarrhea, ileocolitis, and granulomas on colonic biopsies. Differentiation of tuberculosis and Crohn's disease can be very challenging, especially in patients from countries where tuberculosis is endemic, such as Africa. A recent paper by Almadi et al (Am J Gastroenterol 2009;104:1003–1012) provides an excellent comparison of these 2 diseases. In some cases, surgical resection may be the only way to exclude tuberculosis when it cannot be differentiated from Crohn's disease. Parasitic infections can also mimic IBD. Strongyloides stercoralis can cause segmental inflammation in the small bowel, and less commonly in the colon. Entamoeba histolytica, with its predilection for the ileum and right colon can also mimic Crohn's disease. There are case reports of patients with presumed UC who developed fatal amebiasis after therapy with steroids. The diagnosis of amebiasis may require colorectal biopsy or serology. Infections Trigger IBDPostdysenteric UC was reported in 1966, in a series of patients in whom UC developed after an episode of amebic dysentery (Gut 1966;7:438–443). Since then, many series have documented the onset of IBD after acute gastroenteritis both in retrospective series and small prospective series. A recent paper by Gradel et al (Gastroenterology 2009;137:495–501) examined the risk of IBD in individuals with documented Salmonella or Campylobacter gastrointestinal (GI) infection compared with an age- and gender-matched control group from the same population in Denmark. Patients were followed for up to 15 years. Of 13,324 patients with Salmonella or Campylobacter infection 107/13,324 (1.2%) developed IBD compared with 73/26,648 (0.5%) unexposed individuals. The risk was highest during the first year after infection, but remained elevated up to 15 years after the infection. Interestingly, the risk was similar with both infections, and rates of UC and Crohn's disease were also similar. Postdysenteric UC was reported in 1966, in a series of patients in whom UC developed after an episode of amebic dysentery (Gut 1966;7:438–443). Since then, many series have documented the onset of IBD after acute gastroenteritis both in retrospective series and small prospective series. A recent paper by Gradel et al (Gastroenterology 2009;137:495–501) examined the risk of IBD in individuals with documented Salmonella or Campylobacter gastrointestinal (GI) infection compared with an age- and gender-matched control group from the same population in Denmark. Patients were followed for up to 15 years. Of 13,324 patients with Salmonella or Campylobacter infection 107/13,324 (1.2%) developed IBD compared with 73/26,648 (0.5%) unexposed individuals. The risk was highest during the first year after infection, but remained elevated up to 15 years after the infection. Interestingly, the risk was similar with both infections, and rates of UC and Crohn's disease were also similar. Infections Worsen IBDFinally, GI infections can precipitate flares of IBD. The most common food- or water-borne pathogens, Campylobacter, Salmonella, and Shigella, can worsen or mimic symptoms in patients with IBD. Several series evaluated the yield of stool culture for enteric pathogens in patients with acute IBD. Retrospective analyses suggest that, although small, the yield of positive cultures for enteric pathogens ranges from 10% to 20% and can be associated with an increased rate of hospitalization for treatment of IBD flares. A series of 55 patients with 62 flares of IBD had a prevalence of 20% for positive cultures, nearly all of which were Clostridium difficile (J Clin Gastroenterol 2004;38:772–775). In another series of 197 patients with IBD and 221 relapse events, pathogens were detected in 8%, again most commonly C difficile (Eur J Gastroenterol Hepatol 2004;16:775–778). A more recent prospective study of 79 patients with active IBD (99 episodes) found infections in 6, most commonly Campylobacter (Navarco).GI viral infections have also been recognized to aggravate IBD, including CMV, herpes virus, and Epstein–Barr virus. The seroprevalence of human CMV virus ranges from 30% to 90% in the general population. Seroprevalence increases with increasing age. Most CMV infection in immunocompetent individuals is likely asymptomatic. However, after a CMV infection the virus is latent within the host and can periodically reactivate. Acute CMV infection is most often an infectious mononucleosis type picture, very similar to Epstein-Barr virus with symptoms of fever, adenopathy, and myalgias. The virus is latent, probably in myeloid cells. The T-cell–mediated response is the primary mechanism that controls CMV replication. Thus, severe CMV disease occurs most often in patients who have cellular immunodeficiencies.Clinically significant CMV infection is most common in individuals who are immunosuppressed, with solid organ transplant, stem cell transplant, and patients with HIV/AIDS. In solid organ transplant recipients, CMV causes a febrile illness with a low white blood cell count and can affect the lungs, gut, liver, and eyes, among other organs. With stem cell transplant, lung, gut, eye infection, and brain infection are less common. With HIV/AIDS, retinitis is much more common, as is enterocolitis and other GI infections. There is also evidence that CMV infection can accelerate progression to AIDS and death in those with HIV infection.With solid organ transplantation, CMV is such an important factor that patients are screened for serostatus with a potential donor. The highest risk of infection is in a seronegative donor who receives an organ from a seropositive individual. Acute symptoms in this setting can include fever, low white blood cell count, malaise, arthralgia, and nonspecific symptoms; diagnosis is made by detecting the virus in the blood and/or in tissues. With stem cell transplantation, viral infections are more frequently caused by reactivation of the virus in the recipient rather than a primary infection. Primary infection occurs in about 30% of seronegative recipients, whereas reactivation occurs in about 80% who were seropositive before transplantation.Diagnostic tests for CMV include detection of the antigen in the blood (pp65 antigenemia assay). Detection of CMV DNA by polymerase chain reaction in blood can quantify viral load. GI disease is diagnosed by detection of CMV in biopsy specimens by histologic detection of inclusion bodies, and/or immunohistochemistry, and/or viral culture. Most experts recommend that at least 1 method be used on biopsies to diagnose CMV in addition to histopathology. Immunohistochemistry can increase the yield in biopsies to 93%. It is important to know that GI disease can occur even if there is no detection of CMV in the blood.The role of CMV in patients with IBD is far less clear than in patients posttransplantation or with HIV infection. Patients with IBD, especially when immunosuppressed, seem to be at increased risk for acute CMV infection or CMV reactivation, but there is debate about its pathogenicity and whether CMV is a significant pathogen. In 1 series of 69 patients with UC, 25 had reactivation of CMV but most did not require antiviral treatment. In 2 patients who were given ganciclovir, the clinical outcome was not affected. Moreover, CMV disappeared without treatment in most of these patients (Am J Gastroenterol 2007;102:331–337).In contrast, other series suggest that CMV is a pathogen. There are many published cases of CMV colitis detected in steroid refractory colitis. In some, symptoms improved with antiviral therapy. In others, colectomy was required. The prevalence of biopsy-proven CMV colitis in the setting of IBD flares is estimated to be from 21% to 36% according to several small prospective case series (Am J Gastroenterol 2006;101:2857–2865). In a recent series, CMV was found in 8/23 patients with "steroid-resistant" IBD (Dig Dis Sci December 2008).The paper by Domenech et al adds a significant experience to our knowledge. A relatively large cohort of patients with UC, both active and inactive disease, was evaluated prospectively. Colonic CMV disease developed only in the steroid refractory group, and only in patients who were seropositive for CMV before therapy. In several, CMV colitis seemed to develop after CSA therapy. In the steroid-refractory group, 5 had extensive disease and 1 required colectomy. This study suggests that the combination of active disease and immunosuppressive therapy predisposes to CMV reactivation or infection. No patient with inactive disease on immunosuppressive or mesalamine maintenance developed CMV.Ultimately, many questions remain. Is the risk of CMV infection dependent on disease type (UC vs Crohn's disease) or disease location (colon vs small intestine)? Reports of CMV in IBD focus on UC suggesting that colonic disease is the major factor. Because CMV is trophic to inflamed tissue, this may be the major determinant. Does the type of immunosuppressive therapy increase the risk (steroids vs azathioprine vs tumor necrosis factor antibodies)? In this and other series, steroid therapy seems to confer the greatest risk in UC patients, although this may be a surrogate for severity of colitis. In this series, CSA was also given before CMV diagnosis in several patients. None of the inactive disease patients developed CMV, suggesting that disease activity was more important than the immunosuppressive therapy. Whether tumor necrosis factor inhibitors play a role remains to be seen. In a small, prospective series, none of the 15 IBD patients given infliximab developed CVM infection (J Clin Virol 2008;43:180–183). What is the best diagnostic method? This paper shows that CMV can be absent in the blood, even when present in the colon, and that IHC on colon biopsies was better than conventional histology. Diagnosis of disease depends on correlation of symptoms and documenting presence of CMV in tissue. What criteria should we use to document disease, as opposed to silent infection, namely, the possible "innocent bystander" role, if this exists? What is the best therapy, when to start and when to stop, and should immune suppression be discontinued? Patients whose disease improves with antiviral therapy strengthens the notion that CMV is a pathogen. However, whether antiviral therapy would be better if steroids or CSA were discontinued is not known. It would be ideal to know more about the role of antiviral therapy.Should we be screening high-risk patients for CMV serostatus, as is done in transplant patients? In this series, only CMV-positive patients developed CMV colitis. More effective therapy and potentially lower costs could be achieved if we could be sure that only this subset could develop CMV colitis and needed screening. Is it possible that preemptive antiviral therapy could help a subset of patients? In summary, we do not know enough about the role of CMV in UC, its possible natural history, and role of antiviral therapy. This paper adds significant experience and raises more important questions. Finally, GI infections can precipitate flares of IBD. The most common food- or water-borne pathogens, Campylobacter, Salmonella, and Shigella, can worsen or mimic symptoms in patients with IBD. Several series evaluated the yield of stool culture for enteric pathogens in patients with acute IBD. Retrospective analyses suggest that, although small, the yield of positive cultures for enteric pathogens ranges from 10% to 20% and can be associated with an increased rate of hospitalization for treatment of IBD flares. A series of 55 patients with 62 flares of IBD had a prevalence of 20% for positive cultures, nearly all of which were Clostridium difficile (J Clin Gastroenterol 2004;38:772–775). In another series of 197 patients with IBD and 221 relapse events, pathogens were detected in 8%, again most commonly C difficile (Eur J Gastroenterol Hepatol 2004;16:775–778). A more recent prospective study of 79 patients with active IBD (99 episodes) found infections in 6, most commonly Campylobacter (Navarco). GI viral infections have also been recognized to aggravate IBD, including CMV, herpes virus, and Epstein–Barr virus. The seroprevalence of human CMV virus ranges from 30% to 90% in the general population. Seroprevalence increases with increasing age. Most CMV infection in immunocompetent individuals is likely asymptomatic. However, after a CMV infection the virus is latent within the host and can periodically reactivate. Acute CMV infection is most often an infectious mononucleosis type picture, very similar to Epstein-Barr virus with symptoms of fever, adenopathy, and myalgias. The virus is latent, probably in myeloid cells. The T-cell–mediated response is the primary mechanism that controls CMV replication. Thus, severe CMV disease occurs most often in patients who have cellular immunodeficiencies. Clinically significant CMV infection is most common in individuals who are immunosuppressed, with solid organ transplant, stem cell transplant, and patients with HIV/AIDS. In solid organ transplant recipients, CMV causes a febrile illness with a low white blood cell count and can affect the lungs, gut, liver, and eyes, among other organs. With stem cell transplant, lung, gut, eye infection, and brain infection are less common. With HIV/AIDS, retinitis is much more common, as is enterocolitis and other GI infections. There is also evidence that CMV infection can accelerate progression to AIDS and death in those with HIV infection. With solid organ transplantation, CMV is such an important factor that patients are screened for serostatus with a potential donor. The highest risk of infection is in a seronegative donor who receives an organ from a seropositive individual. Acute symptoms in this setting can include fever, low white blood cell count, malaise, arthralgia, and nonspecific symptoms; diagnosis is made by detecting the virus in the blood and/or in tissues. With stem cell transplantation, viral infections are more frequently caused by reactivation of the virus in the recipient rather than a primary infection. Primary infection occurs in about 30% of seronegative recipients, whereas reactivation occurs in about 80% who were seropositive before transplantation. Diagnostic tests for CMV include detection of the antigen in the blood (pp65 antigenemia assay). Detection of CMV DNA by polymerase chain reaction in blood can quantify viral load. GI disease is diagnosed by detection of CMV in biopsy specimens by histologic detection of inclusion bodies, and/or immunohistochemistry, and/or viral culture. Most experts recommend that at least 1 method be used on biopsies to diagnose CMV in addition to histopathology. Immunohistochemistry can increase the yield in biopsies to 93%. It is important to know that GI disease can occur even if there is no detection of CMV in the blood. The role of CMV in patients with IBD is far less clear than in patients posttransplantation or with HIV infection. Patients with IBD, especially when immunosuppressed, seem to be at increased risk for acute CMV infection or CMV reactivation, but there is debate about its pathogenicity and whether CMV is a significant pathogen. In 1 series of 69 patients with UC, 25 had reactivation of CMV but most did not require antiviral treatment. In 2 patients who were given ganciclovir, the clinical outcome was not affected. Moreover, CMV disappeared without treatment in most of these patients (Am J Gastroenterol 2007;102:331–337). In contrast, other series suggest that CMV is a pathogen. There are many published cases of CMV colitis detected in steroid refractory colitis. In some, symptoms improved with antiviral therapy. In others, colectomy was required. The prevalence of biopsy-proven CMV colitis in the setting of IBD flares is estimated to be from 21% to 36% according to several small prospective case series (Am J Gastroenterol 2006;101:2857–2865). In a recent series, CMV was found in 8/23 patients with "steroid-resistant" IBD (Dig Dis Sci December 2008). The paper by Domenech et al adds a significant experience to our knowledge. A relatively large cohort of patients with UC, both active and inactive disease, was evaluated prospectively. Colonic CMV disease developed only in the steroid refractory group, and only in patients who were seropositive for CMV before therapy. In several, CMV colitis seemed to develop after CSA therapy. In the steroid-refractory group, 5 had extensive disease and 1 required colectomy. This study suggests that the combination of active disease and immunosuppressive therapy predisposes to CMV reactivation or infection. No patient with inactive disease on immunosuppressive or mesalamine maintenance developed CMV. Ultimately, many questions remain. Is the risk of CMV infection dependent on disease type (UC vs Crohn's disease) or disease location (colon vs small intestine)? Reports of CMV in IBD focus on UC suggesting that colonic disease is the major factor. Because CMV is trophic to inflamed tissue, this may be the major determinant. Does the type of immunosuppressive therapy increase the risk (steroids vs azathioprine vs tumor necrosis factor antibodies)? In this and other series, steroid therapy seems to confer the greatest risk in UC patients, although this may be a surrogate for severity of colitis. In this series, CSA was also given before CMV diagnosis in several patients. None of the inactive disease patients developed CMV, suggesting that disease activity was more important than the immunosuppressive therapy. Whether tumor necrosis factor inhibitors play a role remains to be seen. In a small, prospective series, none of the 15 IBD patients given infliximab developed CVM infection (J Clin Virol 2008;43:180–183). What is the best diagnostic method? This paper shows that CMV can be absent in the blood, even when present in the colon, and that IHC on colon biopsies was better than conventional histology. Diagnosis of disease depends on correlation of symptoms and documenting presence of CMV in tissue. What criteria should we use to document disease, as opposed to silent infection, namely, the possible "innocent bystander" role, if this exists? What is the best therapy, when to start and when to stop, and should immune suppression be discontinued? Patients whose disease improves with antiviral therapy strengthens the notion that CMV is a pathogen. However, whether antiviral therapy would be better if steroids or CSA were discontinued is not known. It would be ideal to know more about the role of antiviral therapy. Should we be screening high-risk patients for CMV serostatus, as is done in transplant patients? In this series, only CMV-positive patients developed CMV colitis. More effective therapy and potentially lower costs could be achieved if we could be sure that only this subset could develop CMV colitis and needed screening. Is it possible that preemptive antiviral therapy could help a subset of patients? In summary, we do not know enough about the role of CMV in UC, its possible natural history, and role of antiviral therapy. This paper adds significant experience and raises more important questions.