Attempts to prevent CMV infection and disease in bone marrow transplant recipients have produced mixed results. In CMV-seronegative patients, most CMV infections can be prevented by the use of CMV-seronegative blood products [6]. Prophylactic intravenous immunoglobulin also modifies the severity of CMV infection in CMV-seronegative patients and decreases the risk for acute graft-versus-host disease (GVHD) and interstitial pneumonia [7, 8]. On the other hand, effective prophylaxis for CMV reactivation and pneumonia in patients who are CMV-seropositive at the time of transplantation has not been clearly established. Previous trials of prophylactic vidarabine, human leukocyte interferon, and low-dose acyclovir showed no clinically significant effect [1]. In a nonrandomized controlled trial, high doses of prophylactic acyclovir were associated with a decreased incidence of CMV infection and CMV disease [9]. However, the incidence of CMV infection and CMV-related pneumonia was 59% and 19%, respectively, despite the high doses of acyclovir. The efficacy of CMV immune plasma or immunoglobulin in CMV-seropositive patients is also uncertain [7, 10, 11].

Ganciclovir, an acyclic nucleoside analog of guanosine, has recently become available for treatment of CMV infection in immunocompromised patients [12]. In vitro, ganciclovir is approximately 50 times more active than acyclovir against CMV isolates [13]. Thus, we initiated a placebo-controlled, double-blind, randomized trial of prophylactic ganciclovir in CMV-seropositive allogeneic bone marrow transplants.

Methods

Top Methods Results Discussion Author & Article Info References

From May 1987 to August 1990, patients hospitalized at the UCLA Center for the Health Sciences were enrolled in the study if they met the following criteria: undergoing allogeneic bone marrow transplantation for hematologic malignancy or aplastic anemia; 12 years of age or older; seropositive for CMV antibody; and no evidence of pneumonia or other CMV clinical syndrome. Informed consent approved by the UCLA Human Subject Protection Committee was obtained from each patient. Patients undergoing a second bone marrow transplant were excluded. Only three patients meeting the eligibility criteria and subsequently approached for consent refused to participate in the study.

Transplant Procedure

Details on conditioning therapy before transplantation and clinical management after transplantation have been reported previously [14-16]. Patients were given high-dose chemotherapy alone or with radiation therapy followed by intravenous infusion of bone marrow from a related or unrelated donor. Cyclosporine alone or in combination with methotrexate, corticosteroids, T-cell depletion, or immunotoxin (Xomazyme-CD5; Xoma Corporation, Berkeley, California) was used to prevent GVHD. Patients who developed GVHD were evaluated by standard criteria and treated with either corticosteroids alone or corticosteroids plus immunotoxin [14-16]. Trimethoprim-sulfamethoxazole was administered to all patients between the seventh and second days before transplantation and then for 2 consecutive days of each week between day 40 and day 150 after transplantation to prevent Pneumocystis carinii pneumonia [1]. Neither prophylactic acyclovir nor intravenous immunoglobulin was used. All patients received unscreened blood products that were not tested for CMV antibody.

Study Drugs

Patients were randomly assigned in a double-blind fashion to receive ganciclovir or placebo through a central intravenous catheter. The ganciclovir was given at a dosage of 2.5 mg per kg body weight every 8 hours intravenously, starting on the day that pretransplant conditioning therapy was initiated (usually day 7 before transplant) and continuing until the day before the bone marrow infusion. After transplantation, when the neutrophil count reached 1.0 x 10⁹/L, the ganciclovir was resumed at a dosage of 6 mg/kg once per day, Monday through Friday, and continued until day 120 after transplant. The dosage was adjusted in patients with renal failure. For patients whose neutrophil count fell below 1.0 x 10⁹/L while receiving the study drug, prophylaxis was temporarily discontinued. When the neutrophil count returned to a level greater than 1.0 x 10⁹/L, the ganciclovir was restarted at a dosage of 6 mg/kg once per day on Monday, Wednesday, and Friday. If a patient developed documented interstitial pneumonia, gastrointestinal disease, or other clinical syndromes related to CMV, the primary physician could remove the patient from the study and treat the patient with ganciclovir.

Laboratory Procedures

The cytomegalovirus serologic status of patients and bone marrow donors was determined by latex agglutination (CMV SCAN; Becton Dickinson, Cockeysville, Maryland). After transplantation, serologic studies for CMV antibody were done every 2 to 4 weeks on all patients by both complement-fixation and enzyme-linked immunosorbent assay (ELISA) (CMV ELISA-IgG; Pharmacia Diagnostics, Fairfield, New Jersey). Viral cultures of throat, urine, and buffy coat were obtained from marrow transplant recipients before entry into the study and then once a week. Whenever appropriate, viral cultures of suspicious lesions, bronchoalveolar lavage, biopsy material, and autopsy tissue were performed. Tissue cultures were initially screened for viral antigen by immunofluorescence using monoclonal antibodies to viral proteins and then observed for 4 weeks to detect characteristic cytopathic effects. Bronchoalveolar lavage and biopsy material were also examined histologically for typical viral inclusions and immunohistochemically by indirect immunofluorescence using murine monoclonal antibodies to early and late CMV proteins. Complete blood counts and tests for serum creatinine, electrolytes, and liver function were done before, during, and after the study period to assess patients for treatment-related side effects.

Diagnosis of Cytomegalovirus Infection and Disease

Cytomegalovirus infection was diagnosed by isolation of CMV from a culture obtained from any site, a fourfold or greater increase in the CMV antibody titer on complement fixation, an increase in the CMV ELISA measurement to 1.1 units or greater, or the presence of typical CMV inclusion bodies in a tissue specimen. Interstitial pneumonia was diagnosed by tachypnea, hypoxemia, fever, and interstitial pulmonary infiltrates on a chest roentgenogram not explainable by other obvious causes. Cultures, histologic examination, and immunochemical staining of bronchoalveolar lavage or lung biopsy were done to determine the cause of interstitial pneumonia. Similarly, cultures, histologic examination, and immunochemical staining of an endoscopic biopsy of the gastrointestinal tract or a biopsy of the liver in a patient with associated symptoms and signs were used to diagnose CMV disease of the gastrointestinal tract and liver. The wasting syndrome related to CMV was defined as fever, anorexia, and weight loss not explainable by other causes in a patient with culture or serologic evidence of CMV infection.

Statistical Analysis

The Fisher exact test was used to compare differences in proportions. The Student t-test was used to compare means. Univariate comparisons of times to specific events were performed by the method of Kaplan and Meier and analyzed by the log-rank test. Confidence intervals (CIs) for 95% of differences are given where appropriate.

Results

Top Methods Results Discussion Author & Article Info References

Patient Characteristics

One hundred thirty patients were enrolled in the study. However, 45 patients (20 placebo patients, 25 ganciclovir patients) were considered nonevaluable and were excluded from the efficacy analysis. Reasons for nonevaluability were as follows: early death within 9 to 34 days after the transplant (24 patients), marrow graft failure preventing administration of the study drug after transplant (9 patients), withdrawal of patient from the study (10 patients), and inadvertent enrollment of a patient undergoing a second transplant (2 patients). Fifteen placebo patients and 18 ganciclovir patients were not evaluable due to either early death or graft failure. Eight of the 45 patients removed from the study subsequently developed CMV infection (asymptomatic CMV excretion in five patients, fever and wasting in one patient, and pneumonia in two patients). The one patient with a CMV wasting syndrome and one of the two patients with pneumonia were initially randomized to receive ganciclovir. The other patient with CMV pneumonia was randomized to receive placebo. None of these patients received study drug after the transplant, and none was taking study drug when CMV disease developed. All determinations of evaluability were done blindly without knowledge of the patient's treatment assignment and before the study code was broken.

The characteristics of the 85 evaluable patients are summarized in Table 1. Forty-five patients received placebo, and 40 patients were given ganciclovir. The two groups of patients were similar in terms of age, sex, underlying disease, marrow source, GVHD prophylaxis, and the marrow donor's CMV serologic status. More patients in the ganciclovir group received HLA-mismatched marrow (95% CI, –29% to –2%;P = 0.04) and had acute GVHD (CI, –37% to 4%; P = 0.12). On the other hand, more placebo patients received radiation as part of their pretransplant conditioning therapy (CI, 1% to 31%; P = 0.05). The mean time in the study (time from date of transplant to development of CMV disease, death, or study completion) was 87.4 days (range, 24 to 167 days) for the placebo patients and 97.3 days (range, 26 to 122 days) for the ganciclovir patients (P = 0.20).

Effect of Ganciclovir on Cytomegalovirus, Herpes Simplex Virus, and Varicella-Zoster Virus Excretion

The effect of ganciclovir on excretion of CMV is shown in Figure 1. The probability of CMV excretion within the first 120 days after transplant was 0.54 among the placebo patients but only 0.19 among the ganciclovir patients (P = 0.001). The median time to initial excretion of CMV (time of first culture positive for CMV) was day 53 after transplant (range, day 7 before transplant to day 84 after transplant) for the placebo patients and day 61 after transplant (range, day 26 to day 82 after transplant) for the ganciclovir patients (P = 0.004).

View larger version (12K): [in this window] [in a new window]   Figure 1. Probability of cytomegalovirus excretion during the first 120 days after transplantation. The difference in the probability of cytomegalovirus (CMV) excretion between placebo and ganciclovir patients has a P value of 0.001. Kaplan-Meier product limit estimates.

Twenty-three of the 45 placebo patients and 23 of the 40 ganciclovir patients were seropositive for herpes simplex viral antibody before transplantation. Twenty-six placebo patients and 26 ganciclovir patients were also seropositive for varicella-zoster viral antibody before the transplant. Twenty-four placebo patients (53%) and 18 ganciclovir patients (45%) excreted herpes simplex virus during the study (CI, –13% to 30%; P > 0.2). Excretion of herpes simplex virus usually occurred during the immediate post-transplant period when patients were neutropenic before marrow engraftment and were not yet receiving post-transplant study drug. The median time for initial herpes simplex viral excretion was earlier for the placebo patients compared with the ganciclovir patients who received ganciclovir for approximately 1 week before transplant (day 8 after transplant versus day 12 after transplant, P = 0.03). After day 30 following transplant when patients had usually resumed the study drug, fewer ganciclovir patients (5%) than placebo patients (29%) excreted herpes simplex virus (CI, 9% to 39%; P = 0.004). All cases of herpes simplex viral excretion were either asymptomatic or associated with localized, nonfatal infection.

Only two cases of varicella-zoster viral infection occurred during the study; both occurred in the placebo group on days 47 and 53 after transplant (CI, –2% to 10%, P > 0.2). Other viruses excreted during the study were respiratory syncytial virus (two placebo patients), parainfluenza virus (one ganciclovir patient), adenovirus (one placebo patient, one ganciclovir patient), coxsackie A virus (one placebo patient), and rotavirus (one ganciclovir patient).

Effect of Ganciclovir on Cytomegalovirus Infection and Disease

Table 2 compares the incidences of CMV infection and disease in the two study groups. Cytomegalovirus infection developed in 25 of 45 placebo patients (56%) and in 8 of 40 ganciclovir patients (20%) (CI, 16% to 55%; P < 0.001). Cytomegalovirus disease (pneumonia, gastroenteritis, or the wasting syndrome) also occurred less frequently in the ganciclovir patients (4 of 40 patients [10%]) compared with placebo patients (11 of 45 patients [24%]) (CI, –1% to 30%; P = 0.09). When the time at risk for CMV disease is also considered, the probability of CMV disease within the first 120 days after transplant was 0.29 among placebo patients but only 0.12 among the ganciclovir patients (P = 0.06) (Figure 2). Two of the four ganciclovir patients with CMV disease developed disease (pneumonia) while receiving prophylactic ganciclovir. One ganciclovir patient developed CMV pneumonia and another had a CMV-related wasting syndrome after ganciclovir prophylaxis was interrupted because of neutropenia. The median time of onset for CMV disease was day 54 after transplant in the placebo group (range, day 36 to day 81 after transplant) and day 63 after transplant in the ganciclovir group (range, day 42 to day 92 after transplant) (P = 0.06). When all randomized patients were analyzed together, the incidence of CMV infection was 20% in the ganciclovir group (13 of 65 patients) and 43% in the placebo group (28 of 65 patients) (P = 0.008), whereas CMV disease was found in 9% of the ganciclovir patients (6 of 65 patients) and in 18% of the placebo patients (12 of 65 patients) (P = 0.2). No additional cases of CMV infection or disease occurred after day 120 following transplant.

View larger version (12K): [in this window] [in a new window]   Figure 2. Probability of cytomegalovirus disease during the first 120 days after transplantation. The difference in the probability of cytomegalovirus (CMV) disease between placebo and ganciclovir patients has a P value of 0.06. Kaplan-Meier product limit estimates.

The study groups did not differ in the overall incidence of interstitial pneumonia. Interstitial pneumonia from any cause occurred in 13 of 45 placebo patients (29%) and in 11 of 40 ganciclovir patients (28%) (CI, –18% to 21%; P > 0.2). The types of interstitial pneumonia in the placebo group were CMV (seven cases), idiopathic (four cases), and respiratory syncytial virus (two cases). The types of interstitial pneumonia in the ganciclovir group were CMV (three cases), idiopathic (six cases), parainfluenza (one case), and P. carinii (one case). The ganciclovir patient with P. carinii had previously discontinued his prophylactic trimethoprim-sulfamethoxazole. Eight of 13 patients (62%) in the placebo group and 8 of 11 patients (73%) in the ganciclovir group died of interstitial pneumonia (CI, –48% to 26%; P > 0.2).

Side Effects of Ganciclovir

All 130 patients enrolled into the study before transplantation (65 placebo patients, 65 ganciclovir patients) were evaluated for side effects. Seventeen placebo patients (26%) and 28 ganciclovir patients (43%) had probable or possible side effects related to the study drug (CI, –33% to –1%;P = 0.06). However, the only adverse event that occurred more frequently with ganciclovir was neutropenia. Neutropenia requiring interruption of study drug (usually a fall in the neutrophil count to less than 1.0 x 10⁹/L) occurred in 13 of 47 placebo patients (28%) receiving post-transplant study drug and in 25 of 43 ganciclovir patients (58%) receiving post-transplant study drug (CI, –50% to –11%;P = 0.005). Sixteen of the 25 ganciclovir patients whose prophylaxis was interrupted due to neutropenia subsequently resumed the study drug at a reduced dosing schedule of 6 mg/kg once per day on Monday, Wednesday, and Friday when their neutrophil count returned to greater than 1.0 x 10⁹/L. Four other ganciclovir patients resumed prophylaxis at the original dosing schedule of 6 mg/kg once daily, Monday through Friday. The times for recovery of the neutrophil count after marrow engraftment were similar in the placebo and ganciclovir groups, whereas recovery of the platelet count was delayed in the ganciclovir group (Table 3). Marrow graft failure occurred in seven ganciclovir patients and in six placebo patients. The ganciclovir group and placebo group did not differ with respect to any other adverse events.

Survival

Despite a reduction in CMV infection and disease in the ganciclovir patients, the overall survival before day 120 after transplant was similar in the two groups. Twenty-nine of 45 placebo patients (64%) and 28 of 40 ganciclovir patients (70%) were alive at day 120 after transplant (CI, –25% to 14%; P > 0.2). Causes of death in the placebo group were CMV pneumonia (four cases), other types of interstitial pneumonia (four cases), toxicity (four cases), infection (one case), graft failure (one case), and relapse of leukemia (two cases). Causes of death in the ganciclovir group were CMV pneumonia (three cases), other types of interstitial pneumonia (five cases), toxicity (one case), infection (two cases), and graft failure (one case). Survival in the placebo and ganciclovir patients was also similar when all randomized patients were analyzed together (32 of 65 placebo patients [49%] versus 30 of 65 ganciclovir patients [46%], P > 0.2).

Discussion

Top Methods Results Discussion Author & Article Info References

Our results show that prophylactic ganciclovir is associated with a decrease in the occurrence and severity of CMV infection in patients who were CMV-seropositive before allogeneic bone marrow transplantation. The incidence of CMV infection was reduced from 56% in the placebo patients (25 of 45 patients) to 20% in the ganciclovir patients (8 of 40 patients) (CI, 16% to 55%; P < 0.001), whereas the probability of developing CMV disease within the first 120 days after transplant was 0.29 in the placebo patients but only 0.12 in the ganciclovir patients (P = 0.06).

Our results are consistent with two controlled trials of ganciclovir for early treatment of CMV infection after allogeneic marrow transplantation. At the City of Hope transplant center, asymptomatic patients with a bronchoalveolar lavage-fluid culture positive for CMV on day 35 after transplant were randomized in an unblinded manner to receive either early therapy with ganciclovir or observation [17]. Ganciclovir decreased the incidence of subsequent CMV pneumonia from 70% to 25%. However, 22% of the patients who had a lavage-fluid culture negative for CMV on day 35 also developed CMV pneumonia. In a second study at the Seattle transplant center, asymptomatic patients excreting CMV from any site after transplant were randomized in a double-blind fashion to receive either early therapy with ganciclovir or placebo [18]. Subsequent CMV disease (pneumonia or gastrointestinal infection) was reduced from 43% in the patients receiving placebo to 3% in patients treated with ganciclovir. Thirty-five patients without a previous culture positive for CMV also developed CMV disease.

Despite a reduction in CMV infection and disease in our study (see Table 2), the overall incidence of interstitial pneumonia in the patients receiving prophylactic ganciclovir or placebo was similar. Ganciclovir prophylaxis had no effect on the occurrence of idiopathic interstitial pneumonia, which accounts for approximately one half of all cases of interstitial pneumonia after allogeneic bone marrow transplantation [1, 19, 20]. Attempts to identify an infectious causative agent for these pneumonias by special cultures, immunochemical stains, or molecular diagnostic techniques have generally been unsuccessful [19]. The failure of ganciclovir prophylaxis to decrease the incidence of idiopathic interstitial pneumonias also suggests that these pneumonias are not caused by cryptic CMV infection.

One concern about the use of any antimicrobial agent for prophylaxis is the emergence of resistant organisms. Indeed, ganciclovir-resistant CMV isolates have emerged in patients with the acquired immunodeficiency syndrome (AIDS) who were receiving chronic maintenance ganciclovir for CMV retinitis [21, 22]. Four patients in this study excreted CMV while receiving prophylactic ganciclovir, and two patients developed CMV pneumonia despite uninterrupted ganciclovir prophylaxis. The in-vitro susceptibilities of these CMV isolates to ganciclovir, however, are currently not available.

Neutropenia was the only clinically significant side effect of prophylactic ganciclovir. This neutropenia was reversible, and 20 of the 25 patients whose ganciclovir prophylaxis was interrupted were able to resume the drug after the neutropenia resolved (16 at a reduced dosage, 4 at the original dosage). Similar treatment-related neutropenia has been observed in 40% to 50% of patients with AIDS treated with ganciclovir for CMV disease [12, 23]. In the trials of early treatment of CMV infection after bone marrow transplantation, neutropenia requiring modification or cessation of ganciclovir therapy occurred in 30% to 35% of the patients [17, 18]. Granulocyte-macrophage colony-stimulating factor (GMCSF) has been used in patients with AIDS and in those with marrow transplants to reverse the neutropenia associated with ganciclovir [18, 24, 25]. Thus, the efficacy of prophylactic ganciclovir might be improved by the concomitant administration of GM-CSF or granulocyte colony-stimulating factor. Similarly, the use of ganciclovir in combination with a prophylactic anti-CMV immunoglobulin preparation may permit a lower dosage of ganciclovir and thereby decrease toxicity [26]. Finally, foscarnet, an antiviral agent recently approved by the Food and Drug Administration for treatment of CMV retinitis, may have a prophylactic role in patients unable to tolerate ganciclovir due to neutropenia. Foscarnet is also active against CMV isolates resistant to ganciclovir [27, 28], and a comparative trial for CMV prophylaxis would be of interest.