Current issues of ACP Journal Club are published in Annals of Internal Medicine


Ganciclovir compared with placebo for cytomegalovirus colitis in patients with AIDS improved colitis scores and reduced new cases of cytomegalovirus retinitis

ACP J Club. 1993 July-Aug;119:13. doi:10.7326/ACPJC-1993-119-1-013

Source Citation

Dieterich DT, Kotler DP, Busch DF, et al. Ganciclovir treatment of cytomegalovirus colitis in AIDS: a randomized, double-blind, placebo-controlled multicenter study. J Infect Dis. 1993 Feb:167:278-82.



To determine the efficacy and safety of ganciclovir for treating cytomegalovirus (CMV) colitis in patients with the acquired immunodeficiency syndrome (AIDS).


Randomized, double-blind, placebo-controlled trial with 14-day follow-up.


4 university medical centers.


62 men with AIDS and biopsy-proven CMV colitis. Patients had diarrhea (> 6 stools/d), malabsorption, or unintentional weight loss of > 4.5 kg. Exclusion criteria were use of interferon or other treatment for cancer within 1 week of the study, use of acyclovir within 2 days of the study, history of chronic liver disease or mental illness, life expectancy of < 1 month, abnormal renal function, extracolonic CMV infection, or colitis severe enough that placebo treatment was considered unethical. Follow-up was complete.


32 patients were allocated to receive ganciclovir, 5 mg/kg body weight in 100 mL of 5% dextrose over 1 hour every 12 hours for 14 days. Medication was adjusted to 3 mg/kg if neutropenia (< 1000 cells/µL) developed. If the neutrophil count did not increase to > 1000/µL in 4 days or if the platelet count decreased to ≤ 25 000, the patient was removed from the study. 30 patients were allocated to receive matching placebo.

Main outcome measures

Colonoscopy was done once in the 10 days before therapy and on day 14 to determine colitis scores (0, normal colon, to 3, pancolitis or large single ulcer and severe inflammation). Treatment success was defined as study completion and a reduction in colitis scores. Symptoms, CMV cultures, and development of extracolonic CMV disease were recorded.

Main results

Patients in the ganciclovir group had fewer premature terminations than did the placebo group (P = 0.047 [Cochran-Mantel-Haenszel row means scores χ2 test]), more treatment successes (P = 0.042), fewer cases of new CMV infections (P = 0.026) (Table), and fewer CMV-positive colon (3% vs 20%, P = 0.034) and urine (6% vs 43%, P < 0.001) cultures. Patients who received ganciclovir also had more improvement in colitis scores than did patients who received placebo (P = 0.028). The groups did not differ for clinical end points (temperature, body weight, diarrhea, abdominal pain, fatigue, serum albumin, cholesterol, or magnesium) or for adverse effects.


Ganciclovir used to treat cytomegalovirus colitis in patients with the acquired immunodeficiency syndrome compared with similar patients taking placebo led to an improvement in colitis scores and fewer new cases of cytomegalovirus retinitis. The groups did not differ for diarrhea and other clinical end points or side effects.

Source of funding: Syntex Research Division.

For article reprint: Dr. D.T. Dieterich, 345 E. 37th Street, Suite 306, New York, NY 10016, USA. FAX 212-682-4712.

Table. Ganciclovir vs placebo at 14 days in patients with AIDS and cytomegalovirus

Outcome Ganciclovir Placebo RBI (95% CI) NNT (CI)
Treatment success 63% 37% 70% (2 to 200) 4 (3 to 109)
Treatment terminations 9% 27% 65% (-10 to 89) Not significant

*Abbreviations defined in Glossary; RRR, RBI, NNT, and CI calculated from data in article.


Although potent antiretroviral therapy has lowered the incidence of HIV-associated CMV disease, CMV in advanced AIDS can be a debilitating and difficult illness to manage. Those treating this condition will note the highly variable response to treatment. In part, this variability may be caused by substantial but unrecognized non-CMV gastrointestinal disease in patients who have positive CMV cultures.

The effects of ganciclovir treatment on colitis in this trial were less than hoped. From the microbiologic perspective, treatment was associated with a substantial reduction in positive urine and colonic cultures. Histologic findings seemed to improve as was indicated by colitis scores. The reliability of this scoring system has not been established, and the clinical importance of the histologic differences remains unclear. Unfortunately, pathologic improvements were not accompanied by improvements in diarrhea or abdominal pain. Clinical improvement probably requires longer than 14 days to become manifest.

Later trials (1, 2) used longer treatment periods and compared intravenous ganciclovir with intravenous foscarnet. On the basis of these studies, current treatment guidelines (3) for CMV gastrointestinal disease support induction therapy with either agent over a 3- to 6-week period. The choice of agent should be based on its toxicity profile, and the drugs can be combined if either has failed. The role of maintenance therapy is less clear, but is recommended after re-induction following a relapse. All patients should be monitored closely for CMV retinitis.

Andrew L. Pattullo, MD
Martin T. Schechter, MDUniversity of British ColumbiaVancouver, British Columbia, Canada


1. Blanshard C, Benhamou Y, Dohin E, et al. Treatment of AIDS-associated gastrointestinal cytomegalovirus infection with foscarnet and ganciclovir: a randomized comparison. J Infect Dis. 1995;172:622-8.

2. Parente F, Bianchi Porro G. Treatment of cytomegalovirus esophagitis in patients with acquired immune deficiency syndrome: a randomized controlled study of foscarnet versus ganiclovir. The Italian Cytomegalovirus Study Group. Am J Gastroenterol. 1998;93:317-22.

3. Whitley RJ, Jacobson MA, Friedberg DN, et al. Guidelines for the treatment of cytomegalovirus diseases in patients with AIDS in the era of potent antiretroviral therapy: recommendations of an international panel. International AIDS Society-USA. Arch Intern Med. 1998;158:957-69.