ORIGINAL PAPERS

 

A comparative study of two histological techniques for the identification of cytomegalovirus infection in colorectal biopsies from patients with chronic inflammatory bowel disease

Estudio comparativo entre dos técnicas histológicas para el diagnóstico de infección por citomegalovirus en biopsias colorrectales de pacientes con enfermedad inflamatoria intestinal crónica idiopática

 

 

M. L. de Castro, A. Tardío¹, V. del Campo², A. Estévez¹, J. R. Pineda, F. Domínguez, J. A. Hermo and J. Clofent

Departments of Gastroenterology, ¹Pathology, and ²Epidemiology. University Hospital, Vigo-CHUVI. Vigo, Spain

Correspondence

 

 

---

ABSTRACT

Background: the role that cytomegalovirus (CMV) plays in inflammatory bowel disease (IBD) is controversial. The diagnosis of CMV infection in IBD depends on viral identification with hematoxylin-eosin (HE) or immunohistochemistry (IHC). Our aim was to compare the sensitivity of HE and IHC for this diagnosis in IBD patients.
Patients and methods: a case-control study. Our database was searched for IBD patients with HE- or IHC-based CMV-positivity from 1997 to 2007. Controls were selected among IBD inpatients matched for age and year of diagnosis with CMV. Their clinical characteristics were analyzed. HE and IHC were performed on biopsies from cases and controls at 6 months before and after inclusion in the study. In the statistical analysis, p values below 0.05 were considered significant.
Results: ten IBD patients with CMV infection were identified. IBD-CMV patients were more steroid-resistant or steroid-dependent (p = 0.03), and underwent a higher number of colonic biopsies (p = 0.03). From 97 biopsies analyzed, 12 were HE-negative and IHC-positive, and 3 showed reversed results. The sensitivity of HE was 58.6%, 95% CI (38.9-78.3), and that of IHC was 89.7%, 95% CI (76.8-100). We did not find a good level of agreement between both techniques: kappa value 0.55, 95% CI (0.36-0.75). CMV positivity with IHC was associated with the use of more than one immunosuppressant drug, OR 13.5, 95% CI (1.2-152.2). Antiviral treatment was useful for CMV patients with steroid-dependent and steroid-refractory IBD.
Conclusions: IHC shows a 30% higher sensitivity than HE for the diagnosis of CMV infection in IBD patients. There is no good level of agreement between both histological techniques.

Key words: Cytomegalovirus infection. Inflammatory bowel disease. Ulcerative colitis. Crohn's disease. Immunohistochemistry. Diagnosis.

---

RESUMEN

Introducción: el papel que el citomegalovirus (CMV) juega en la enfermedad inflamatoria intestinal (EII) es controvertido. El diagnóstico de infección por CMV en estos pacientes reside en su identificación en el tejido colónico mediante hematoxilina-eosina (HE) o inmunohistoquímica (IHC). Nuestro objetivo fue comparar la sensibilidad de ambas técnicas histológicas para establecer este diagnóstico.
Pacientes y métodos: estudio caso-control, identificando en el periodo 1997-2007 aquellos pacientes CMV positivos con HE o IHQ. Sus controles fueron apareados por edad y año de diagnóstico de CMV en los casos. Se realizaron ambas técnicas en las biopsias de casos y controles obtenidas 6 meses antes y tras su inclusión en el estudio. En el análisis estadístico consideramos significativos los valores de p < 0,05.
Resultados: encontramos infección por CMV en 10 pacientes. Estos sujetos presentaban más frecuentemente una EII resistente o dependiente de esteroides (p = 0,03) y un mayor número de biopsias de colon (p = 0,03). De 97 biopsias analizadas, 12 fueron HE-negativas e IHC-positivas y 3 mostraron resultados inversos. La sensibilidad de HE fue 58,6% IC95 (38,9-78,3) y para IHQ 89,7% IC95 (76,8-100). No encontramos una buena concordancia entre ambas técnicas histológicas: kappa 0,55 IC95 (0,36-0,75). La positividad para CMV se asoció al uso de más de un fármaco inmunosupresor, OR 13,5 IC95 (1,2-152,2). El tratamiento antiviral fue útil en la EII refractaria o dependiente de esteroides.
Conclusiones: la IHQ posee una sensibilidad superior a HE (30%) para el diagnóstico de infección por CMV en la EII, no existiendo buena concordancia entre ambas técnicas histológicas.

Palabras clave: Citomegalovirus. Enfermedad inflamatoria intestinal. Colitis ulcerosa. Enfermedad de Crohn. Inmunodepresión. Diagnóstico.

---

 

Introduction

The role played by cytomegalovirus (CMV) in inflammatory bowel disease (IBD) is controversial (1). This infectious agent has been involved in episodes of IBD exacerbation and in the development of corticoid-refractoriness. Nonetheless, its presence may be merely explained in terms of previous colonization, which reactivates in a context of immunosuppression, thus lacking importance for the evolution and prognosis of IBD.

The diagnosis of CMV infection in these patients relies on the use of hematoxylin-eosin (HE) staining in order to identify the typical cellular changes induced by the infectious agent, or via CMV-specific immunohistochemistry (IHC) in colonic biopsies. Despite an absence of comparative studies between both techniques, a higher sensitivity of IHC compared to HE has been reported in patients suffering from serious IBD outbreaks (2).

Our goal was to compare, in a case-control study, the sensitivity of both histological techniques for the diagnosis of CMV infection in IBD patients.

 

Methods

Clinical study

A review of our Pathology Department archives over the period from 1997 to 2007 allowed an identification of patients suffering from IBD with CMV infection in our health center (400-bed hospital servicing an adult population of 150,000 inhabitants). The diagnosis of IBD was based on clinical, endoscopic and histological findings (3,4). Positivity by histological HE or IHC in colonic biopsies established the diagnosis of CMV infection. As we had no serological data, neither antigenemia nor CMV culture results from every patient included in this study, such data were not analyzed.

Controls were selected among patients admitted to hospital in order to control their IBD in the same year in which the study cases were diagnosed of CMV infection, and with age differences of less than 5 years from their corresponding cases.

Clinical histories of both study cases and controls were reviewed, and type and extension of IBD, time from diagnosis and immunosuppressant medication given over one month prior to admission and in the hospital were all recorded. Categorization as corticoid-refractory IBD was considered in the absence of a favorable clinical response following 10-day intravenous steroid therapy with a dose of 1 mg/kg body weight, whilst corticoid-dependence was established on the reappearance of symptoms after a reduction or suppression of steroidal treatment.

Histology

We carried out a revision of all colonic biopsies from patients with CMV infection and their control group 6 months before and after the date of inclusion in the study. In all these samples both histological techniques (HE and IHC) were reviewed by the same pathologist (AT) in paraffin-embedded sections.

Positivity for CMV infection with HE was established by the presence of characteristic cellular changes: cytomegalia and basophilic intranuclear inclusions. To determine the presence of CMV infection with IHC, a biclonal antibody was employed (CCH2 and DDG9; DAKO 1:25). Tissue slices, previously deparaffined, were rehydrated, and an antigen-retrieval technique was applied using citrate buffer pH6 with heat and proteinase K. Incubation with biclonal mouse antibodies CCH2 and DDG9 was followed by application of the EnVision-DAKO viewing system. We used CMV-infected pulmonary tissue for positive control, and as a negative control we followed all the steps described in the previous protocol without adding the biclonal antibody. Positivity for CMV infection was reviewed in all colonic biopsies.

Statistical analysis

Results are shown using the mean and standard deviation (in brackets) values or percentages where appropriate. In order to analyze continuous variables the Mann-Whitney U test was applied, while categorical variables were evaluated using Fisher's exact test. The Kappa index evaluates the level of agreement between both histological techniques. Statistical significance was considered for values of p < 0.05 and 95% confidence intervals (CI95) of clinically significant variables are shown as well.

 

Results

Over the period from 1997 to 2007, ten patients were diagnosed with IBD with CMV infection (6 men and 4 women), with a mean age of 45.2 (15) years. Their clinical characteristics are showed in table I. Seven of these patients were diagnosed with ulcerative colitis (UC) and 3 with Crohn's disease (CD). Colonic lesions were limited to the left colon in 4 patients, and were pancolonic in 6. Three of these patients were clinically catalogued as corticoid-resistant, 2 were corticoid-dependant, and in 5 cases CMV was identified in colonic biopsies taken during a moderate outbreak of IBD. In 2 patients the diagnosis of CMV was established in the first outbreak of IBD, while in the remaining 8 subjects disease duration was 2.1 years (2). Two of the reported cases proved to be positive for CMV while on an active IBD outbreak in more than one occasion during the study period (cases 2 and 3).

At the time of diagnosis with CMV infection all patients were under parenteral corticoids; in 5 cases steroids were associated with azathioprine, and in 3 cases also with cyclosporine A. During the month prior to admission, 60% of cases had been given corticoids and azathioprine, and 20% infliximab (Table I).

No significant differences were observed between both groups when considering sex, age, type or IBD extension. Nevertheless, patients infected with CMV showed a higher percentage of refractory or steroid-dependant IBD (p < 0.03). IBD duration in the control group was 5 (8) years, with this being the first outbreak in 40% of these patients (Table II). An oral intake of 5-ASA was confirmed by 70% of patients in each group. We observed no significant differences with respect to oral steroid intake in the month prior to admission, or in the use of immunosuppressant drugs for the treatment of the acute IBD outbreak between the two groups. Nonetheless, the number of patients in whom more than one immunosuppressant drug had been prescribed considering these two periods of time was significantly greater in the CMV-infected group (90%) than in the control group (40%) p = 0.03, OR 13.5 and CI (1.2-152.2).

A total of 97 colonic biopsies were examined with HE and IHC, 64 of them belonged to the case group and 33 to the control group. Practically all of these biopsies involved the left colon, and were taken from areas showing greater endoscopic abnormalities during an active IBD outbreak. The mean number of biopsies taken from patients suffering from CMV infection was 6.3 (3), and this value was significantly higher compared with 3.1 (2) in the control group, p < 0.003.

In the case group, similar results were observed between both histological techniques in 49 biopsies: 14 of them being positive in both determinations and 35 being doubly negative. However, 12 (19%) colonic biopsies from CMV-infected patients were negative for HE but positive for IHC (Table III). Four of these samples were taken in the period of 6 months prior to CMV infection, these being 19% of a total of 20 biopsies obtained during this period. One of these 12 biopsies was later reclassified as positive also with HE, showing atypical inclusions. Three biopsies showed positivity by HE but no CMV findings in IHC. All 33 biopsies from the control group proved negative for CMV with both techniques.

The sensitivity of both histological techniques obtained from the total of 97 biopsies analyzed was 58.6%, CI95 (38.9-78.3) for HE, and 89.7% for IHC, CI95 (76.8-100). Meanwhile, IHC has a 30% higher sensitivity over HE to establish the diagnosis of CMV infection, and this difference was statistically significant (p = 0.01). The level of agreement between the results obtained in both techniques when all samples were reviewed showed a kappa value of 0.55, CI95 (0.36-0.75), with p = 0.0001.

Patients found positive for CMV were given antiviral treatment. Seven subjects received intravenous ganciclovir (5 mg/kg/12 hours) and the other 3 oral valganciclovir (900 mg/12 hours). We reported clinical improvement in 2 corticoid-resistant patients (66.6%), who reached complete recovery, and in 2 corticoid-dependent subjects (100%) as steroids were discontinued. In one corticoid-resistant patient proctocolectomy was performed due to treatment failure. In patients with moderate IBD outbreaks antiviral treatment did not interfere with IBD outcome.

 

Discussion

CMV-infection is highly prevalent, and is present in 40-100% of the world adult population (1). Once in the body, the virus may remain latent and reactivate itself in a situation of cellular immunodepression (5). The relationship between CMV and IBD is not totally clear. The great affinity that this virus has for organic tissues with active inflammation, combined with its high seroprevalence in patients suffering from UC, led to consider it a mere colonizer (6-8). However, CMV has proven to be a pathological agent in 20-30% of serious outbreaks of IBD, which were refractory to steroidal treatment (9-14). In these patients, antiviral treatment avoids the necessity for further surgical treatment and reduces mortality, thus justifying the active search for this infectious agent (2,15).

At present, the diagnosis of CMV infection in patients suffering from IBD relies on histological techniques, since alternative methods such as serology, antigenemia, viremia, viral isolation or polymerase chain reaction (PCR) on blood or intestinal tissue have not proven their usefulness. These methods do not allow differentiation between a situation of latent colonization or active infection in which CMV has a pathogenic role (1,5,16).

The finding in colonic biopsies of the characteristic CMV changes with HE has traditionally been considered a gold standard for the diagnosis of CMV infection in these patients. This technique is highly specific, although its sensitivity varies widely (10-87%) depending on the number of biopsies examined and the level of experience of the pathologist (5,1). Thus, HE may be negative in 20% of patients with corticoid-resistant ulcerative colitis and CMV infection, and in 37% of HIV patients with lesions of the gastrointestinal tract caused by this virus (2,17).

Nowadays, IHC is considered the diagnostic technique of choice, as its sensitivity in patients with serious outbreaks of IBD has been reported to be higher than the one obtained with HE (1,2,18). Nonetheless, these values can decrease whether IHC is carried out in colonic biopsies or in surgical specimens (13). Moreover, its superiority compared to HE has not been established in all patients suffering from active IBD (13,19), and there is no comparative studies between both these histological techniques. Besides, these two histological techniques are performed routinely in different slices of the same biopsy, which makes possible the existence of cytomegalia in one slice and negativity by IHQ in the following one.

Our results show that IHC has a higher sensitivity than HE. Therefore, in the clinical practice we may expect 30% of results to be false negatives if we use HE as the only diagnostic technique for CMV detection in patients suffering from IBD. A similar percentage was observed in the biopsies obtained in the 6 months prior to diagnosis in the case group patients, leading to unnecessary diagnostic delays. On the other hand, the value obtained in this study for the Kappa index demonstrates that there is no good level of agreement when establishing a diagnosis of CMV infection between both histological techniques in colonic biopsies, as this value is lower than 0.8, which discourages the implementation of a single histological technique for this diagnosis.

Nevertheless, a generalized search for CMV infection in active outbreaks of IBD is unjustified due to its low prevalence (17) and the doubtful benefits of antiviral treatment (19-21) in these cases. Our data support this idea, as the benefits of antiviral treatment were only seen in steroid-refractory or steroid-dependeant IBD patients in our group.

Recently, non-pathogenic genotypes of CMV have been identified, which may frequently reactivate in active IBD and spontaneously disappear without specific antiviral treatment (1,20,21). This new perspective allows to understand the initial consideration of CMV as a mere colonizer (7,8) and highlights the multitude of roles that this infectious agent may play in IBD.

CMV infection in patients suffering from IBD is generally considered a reactivation of latent viruses rather than a true primary infection. These patients generally have a long duration of disease and have been under multiple immunosuppressant drugs, which favor the reactivation of previous viral colonization (5,12,15,22). A relationship between the histological detection of CMV and previous steroidal treatment (11,17) or treatment received during an active IBD outbreak (13) has been reported. Nevertheless, data against this fact have been published in some papers as well (10,20), and the heterogeneity of studies carried out do not allow definitive conclusions. In our patients we observed an association between CMV identification in colonic tissue by IHC and the use of more than one immunosuppressant. Therefore, it could be possible that the use of stronger and combined immunosuppressive drugs in IBD may reactivate latent viruses (23), but this topic needs to be confirmed in prospective studies.

In conclusion, IHC has a higher sensitivity than HE for the diagnosis of CMV infection in IBD patients, and there is no good level of agreement between both histological techniques. The use of combined immunosuppressants is related to a higher detection rate of the virus by IHC. CMV still plays a controversial role in patients suffering from IBD, as the significance of this infection is not the same in all clinical settings.

 

References

1. Kandiel A, Lashner B. Cytomegalovirus colitis complicating inflammatory bowel disease. Am J Gastroenterol 2006; 2857-65.        [ Links ]

2. Kambham N, Rohini MS, Cartwright CA, Longacre T. Cytome-galovirus infection in steroid-refractory ulcerative colitis: a case-control study. Am J Surg Pathol 2004; 28: 365-73.        [ Links ]

3. Truelove SC, Witts LJ. Cortisone in ulcerative colitis. Final report on therapeutic trial. Br Med J 1955; 2: 1041-8.        [ Links ]

4. Lennard-Jones JE. Classification of inflammatory bowel disease. Scand J Gastroenterol 1989; 24(Supl. 170): 2-6.        [ Links ]

5. Goodgame RW. Gastrointestinal cytomegalovirus disease. Ann Intern Med 1993; 119: 924-35.        [ Links ]

6. Farmer GW, Vincent MM, Fucillo DA, Horla-Barbosa L, Ritman S, Sever JL. Viral investigations in ulcerative colitis and regional enteritis. Gastroenterology 1973; 65: 8-18.        [ Links ]

7. Berk T, Gordon SJ, Choi HY, Cooper HS. Cytomegalovirus infection of the colon: A possible role in exacerbation of inflammatory bowel disease. Am J Gastroenterol 1985; 80: 355-60.        [ Links ]

8. Eyre-Brook A, Dundas S. Incidence and clinical significance of colonic cytomegalovirus infection in idiopathic inflammatory bowel disease requiring colectomy. Gut 1986; 27: 1419-25.        [ Links ]

9. Vega R, Bertrán X, Menacho M, Doménech E, Moreno de Vega V, Hombrados M, et al. Cytomegalovirus infection in patients with inflammatory bowel disease. Am J Gastroenterol 1999; 94: 10536.        [ Links ]

10. Alcalá MJ, Casellas F, Pallarés J, de Torres I, Malagelada JR. Infection by cytomegalovirus in patients with ulcerative colitis requiring colonia resection. Med Clin (Barc) 2000; 114: 201-4.        [ Links ]

11. Cottone M, Pietrosi G, Martorana A, Casá G, Pecoraro L, Oliva A, et al. Prevalence of cytomegalovirus infection in severe refractory ulcerative and Crohn's colitis. Am J Gastroenterol 2001; 96: 773-5.        [ Links ]

12. Papadakis A, Tung JK, Binder SW, Kam LY, Abreu MT, Targan SR, et al. Outcome of cytomegalovirus infections in patients with inflammatory bowel disease. Am J Gastroenterol 2001; 96: 2137-41.        [ Links ]

13. Maconi G, Colombo E, Zerbi P, Sampietro GM, Fociani P, Bosani M, et al. Prevalence, detection rate and outcome of cytomegalovirus infection in ulcerative colitis patients requiring colonia resection. Digestive and Liver Disease 2005; 37: 418-23.        [ Links ]

14. Maher MM, Nassar MI. Acute cytomegalovirus infection is a risk factor in refractory and complicated inflammatory bowel disease. Dig Dis Sci 2008 Dec18 (Epub ahead of print).        [ Links ]

15. Pfau P, Kochman ML, Furth EE, Lichtenstein GR. Cytomegalovirus colitis complicating ulcerative colitis in the steroid-naïve patient. Am J Gastroenterol 2001; 96: 895-9.        [ Links ]

16. Hommes D, Sterringa G, van Deventer S, Tytgat G, Weel J. The pathogenity of cytomegalovirus in inflammatory bowel disease. Inflamm Bowel Dis 2004; 10: 245-50.        [ Links ]

17. Cotte L, Drouet E, Bissuel F, Denoyel GA, Trepo C. Diagnostic value of amplification of human cytomegalovirus DNA from gastrointestinal biopsies from human immunodeficiency virus-infected patients. J Clin Microbiol 1993; 31: 2066-9.        [ Links ]

18. Criscuoli V, Rizzuto MR, Cottone M. Cytomegalovirus and inflammatory bowel disease: is there a link? World J Gastroenterol 2006; (12830): 4813-8.        [ Links ]

19. De Saussure P, Lavergne-Slovet 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.        [ Links ]

20. Dimitroulia E, Spanakis N, Konstantinidou A, Legakis N, Tsakris A, Path MRC. Frequent detection of cytomegalovirus in the intestine of patients with inflammatory bowel disease. Inflamm Bowel Dis 2006; 12: 879-84.        [ Links ]

21. Matsuoka K, Iwao Y, Mori T, Sakuraba A, Yajima T, Hisamatsu T, et al. Cytomegalovirus is frequently reactivated and disappears without antiviral agents in ulcerative colitis patients. Am J Gastro 2007; 102: 331-7.        [ Links ]

22. Villafruela M, Moya E, de Manuel J, Sandoval A, Becerro J. Síndrome mononucleósico de etiología múltiple en paciente con enfermedad de Crohn en tratamiento con azatioprina. Rev Esp Enferm Dig 2009; 101: 225-6.        [ Links ]

23. Martínez-Montiel MP, Muñoz-Yagüe MT. Tratamientos biológicos en la enfermedad inflamatoria crónica intestinal. Rev Esp Enferm Dig 2006; 98: 265-91.        [ Links ]

 

 

Correspondence:
María Luisa de Castro Parga.
C/ República Argentina, 10, 8º D.
36201 Vigo. Pontevedra, Spain.
e-mail: luisadecastroparga@mundo-r.com

Received: 17-01-09.
Accepted: 05-06-09.