Methods

Patients

In total, 111 consecutive patients transplanted between 10/5/00 and 12/28/01 who did not receive steroid prophylaxis for ES (group A), and 83 consecutive patients transplanted between 2/1/02 and 3/28/03 who did (group B), were reviewed. Data were prospectively collected, and retrospectively verified. All patients were treated on transplant protocols that were reviewed and approved by the hospital's institutional review board, with written informed consent obtained from all patients.

Transplant protocol

All patients had indwelling central venous catheters (CVC) placed prior to admission to the BMT unit. Mobilization for peripheral blood progenitor cells (PBPC) included either etoposide 2 g/m², given as a continuous i.v. infusion over 4 h, followed by subcutaneous G-CSF 10 /kg daily, or G-CSF alone. Leukopheresis was performed for 5 days, or until collection of 7 10⁶ CD34⁺ cells/kg, whichever came first. A minimum of 2 10⁶ CD34⁺ cells/kg was required to proceed with AHSCT.

The chemotherapeutic preparative regimen in patients with non-Hodgkin's lymphoma (NHL) and HD consisted of oral busulfan 14 mg/kg given as 1 mg/kg every 6 h 14 doses, starting on day -9 and ending on day -5, etoposide 60 mg/kg administered by continuous i.v. infusion over 24–36 h, beginning 4 h after the last dose of busulfan on day –5 and ending on day –4, and i.v. cyclophosphamide 60 mg/kg, beginning at least 6 h after the completion of etoposide, infused over 2 h on day –3 and day –2. Patients with multiple myeloma (MM) received busulfan 16 mg/kg, and cyclophosphamide only. Infusion of PBPC occurred on day 0. Daily i.v. G-CSF 480 g was administered starting day +5, until neutrophil engraftment. Irradiated, leukocyte-depleted blood products were infused to maintain hemoglobin and platelet levels above 90 g/l and 0.015 10⁹/l, respectively. Clinical examination was performed daily, and body temperature was measured three to four times a day. No routine surveillance cultures were performed. If the body temperature was 38°C, blood cultures were collected from the CVC and peripheral blood, and repeated daily for as long as fever and/or bacteremia persisted. Other diagnostic studies for evaluation of infections were guided by clinical findings. For patients to be discharged from the hospital after neutrophil engraftment, they had to be afebrile for at least 24 h, able to drink 1 l of fluid in a day, and tolerating oral medications.

Antimicrobial prophylaxis

Antimicrobial prophylaxis with i.v. vancomycin 1 g once a day, oral clarithromycin 250 mg twice a day, and oral ciprofloxacin 500 mg twice a day was started after the dose of etoposide for PBPC mobilization was administered, and continued until the patient was no longer neutropenic and leukopheresis had been completed.¹⁵ After admission for AHSCT, patients were housed in positive pressure high-efficiency particulate air-filtered rooms on the BMT unit. Mouth care included chlorhexidine 0.12% mouth wash thrice a day, and either clotrimazole troche 10 mg, or nystatin mouth wash (100 000 U/ml) 5 ml four times daily. Upon admission to the BMT unit, oral ciprofloxacin 500 mg twice a day was started, and on day +1, i.v. amphotericin B 0.2 mg/kg once a day, and oral acyclovir 400 mg twice a day, were administered until neutrophil engraftment. Antimicrobial prophylaxis was the same in both groups. Most patients also received oral acyclovir 400 mg twice daily as prophylaxis against herpes simplex virus (HSV) and varicella zoster virus (VZV) for up to 1-y post transplant. Intravenous immune globulin was not administered, except in patients with documented hypogammaglobulinemia (IgG <600 mg/dl). Empiric parenteral antibiotics were administered in neutropenic patients after completion of the chemotherapeutic preparative regimen, at the time of the first recorded fever 38°C, and usually included piperacillin-tazobactam, gentamicin and vancomycin. If fever persisted for more than 3–5 days, the dose of amphotericin B was increased to 0.5 mg/kg daily. Treatment for microbiologically confirmed infections varied according to the individual physician preference.

Definitions

Platelet engraftment was defined as the first day on which the platelet count exceeded 20 10⁹/l unsupported by platelet transfusion, for 3 consecutive days. Neutrophil engraftment was defined as the first day on which the absolute neutrophil count (ANC) exceeded 500 10⁶/l. ES was defined as fever >38°C and rash, developing at the time of engraftment, on day +7 to day +11. Microbiologically confirmed infection was defined as signs or symptoms of infection with an organism isolated on blood culture, histological material or other clinically relevant site of infection. Bacteremia was defined as a fever >38°C, with a positive blood culture for any organism, except for coagulase negative staphylococci, skin coryneforms, Lactobacillus species, and viridans Streptococcus, for which at least two consecutive positive blood cultures were considered significant. i.v. catheter-related bacteremia (CRB) was defined as clinical evidence of infection, accompanied by positive blood culture(s) as defined above, collected through the catheter, with or without the presence of cellulitis or purulence at the exit site or along the tunnel, when a distant primary site of infection could not be clinically identified. Blood stream infection of gastrointestinal origin was defined as the presence of positive blood cultures for Gram negative bacilli in the setting of an erosive bowel process, such as lymphomatous invasion or mucositis, and in the absence of an indwelling CVC. Cytomegalovirus (CMV) viremia was defined as the detection of CMV DNA in blood by Hybrid Capture assay (Digene Corporation, Gaithersburg, MD, USA), with or without the presence of fever. Colitis due to CMV was defined as diarrhea associated with CMV viremia, and pathological evidence of CMV tissue invasion. Colitis due to Clostridium difficile was defined as diarrhea, associated with a positive Clostridium difficile toxin A assay in stool. Urinary tract infection was defined as dysuria, associated with the isolation of a typical urinary pathogen in urine, such as Enterobacteriaceae or Enterococci. Prostatitis was defined as the presence of symptoms of prostatism, associated with the isolation of a typical urinary pathogen in urine. Bronchitis was defined as fever >38°C, respiratory symptoms such as cough or dyspnea, a normal chest radiograph, and a typical lower respiratory pathogen isolated from sputum. Influenza was defined as the presence of symptoms of upper respiratory tract infection, associated with the detection of influenza virus antigen using direct fluorescent antibody (DFA). Shingles was defined as the appearance of typical vesicular skin lesions due to VZV with a dermatomal distribution. Herpes dermatitis and labialis were defined as the presence of the typical vesicular cutaneous rash, or herpetic oral erosions, associated with the detection of HSV antigen using DFA. Early infection was defined as one occurring within 1 month after transplant, and late infection as one occurring 1–6 months later. Length of stay in the hospital was defined as the period between the date of admission for starting the chemotherapeutic preparative regimen and the date of discharge.

Statistical analysis

Categorical variables for demographic data of both groups were compared using the ² test, and continuous variables were compared using either the t-test or Wilcoxon rank-sum test. The incidence and types of microbiologically confirmed early and late infections occurring in groups A and B were compared using the ² test. The Kaplan–Meier method was used to estimate risk of infection, ES, overall survival, and progression-free survival. Outcomes were compared between groups using the log-rank test. Censoring for infection occurred if the patient was either alive at last follow-up without having an infection, or if the patient died without having an infection. The first censored patient for this analysis was censored at day +46. The risk of developing an infection on each of the first 21 days post transplant was compared using the ². All analyses were carried out using SAS® software (Sas Institute, Inc., Cary, NC, USA). All statistical tests were two-sided. P<0.05 was used to indicate statistical significance.

Results

Demographic data

Underlying diagnoses included NHL (55.7%), MM (18%), HD (16.5%), acute myelogenous leukemia (8.2%), and other malignancies (1.5%). There were no significant differences between groups A and B in age, gender, race, prior radiation therapy, number of prior chemotherapy regimens, disease status at transplant, mobilization regimen, days of leukopheresis, CD34+ cell dose, days until platelet engraftment and days until neutrophil engraftment (Tables 1a and 1b). Group B had fewer patients with NHL and MM combined (64 vs 81% (P=0.007)), and shorter median duration from diagnosis to transplant (10.8 vs 14.1 months (P=0.043)). The mean length of hospital stay was shorter in group B compared to group A (21 vs 23 days (P=0.015)). Hospital discharge criteria were the same for both groups. Rates of readmission to our hospital were similar in both groups (Table 2). For the 63 patients who were readmitted, the post transplant day of their first readmission ranged from day 14 to day 659. The mean follow-up among patients who were alive was shorter for patients in group B compared to group A (7.8 months vs 18.6 months (P<0.001)). The last finding is not surprising, as steroid prophylaxis is a more recent protocol.

Table 1a - Demographic data: categorical variables.

Full table

Table 1b - Demographic data: continuous variables.

Full table

Table 2 - Hospital readmission rates.

Full table

Engraftment syndrome

Group B had a much lower risk of ES compared to group A (5/83 {6%} vs 63/111 {57%} (P<0.001)) (Figure 1). Mobilization regimen did not influence the incidence of ES (P=0.14). No patients in group B developed ES by day 4, the day steroid prophylaxis was started, while 12% in group A developed ES by day 4. We attribute this to chance, since there was no difference in treatment up to that point. One death due to ES occurred in group A. There were no significant differences in group B between the five patients who developed ES, and the 78 patients who did not, on any of the parameters included in Tables 1a and 1b (results not shown). In these five patients, ES developed on days 7, 8, 9, 9, and 10, respectively, post transplant. Two of these patients were treated for ES with the same dose of steroids used prophylactically, and the remaining three were treated with increased dose of steroids.

Figure 1.

Risk of engraftment syndrome. Thick line: steroid prophylaxis; thin line: no steroid prophylaxis.

Full figure and legend (15K)

Infections

There was a trend toward more infections occurring in group A, but this was marginally significant only on days +6 and +7 (Table 3). The incidence of early and late microbiologically confirmed infections was not significantly different between groups (Figure 2). Early infections occurred in 21/111 (18.9%) in group A vs 8/83 (10%) in group B (P=0.072)). Late infections occurred in 2/111 (1.8%) in group A vs 6/83 (7.2%) in group B (P=0.075)). Types of infections (Table 4a) and types of organisms identified (Tables 4a and 4b) were similar in both groups. The most common type of infection was i.v. CRB, occurring in 13% of patients in group A and 5% of patients in group B within 1 month of transplant (P=0.06). The most common causative organism for CRB was coagulase negative staphylococcus. Two patients (1.8%) in group A had early CMV viremia, and one patient (1.2%) in group B had late CMV colitis. Only one patient in group A had VZV reactivation. No cases of invasive fungal infections occurred in either group. None of the infections were fatal.

Figure 2.

Incidence of microbiologically confirmed infections.

Full figure and legend (48K)

Table 3 - Risk of infection in the first 21 days post transplant.

Full table

Table 4a - Types of infections identified.

Full table

Table 4b - Types of organisms in patients with microbiologically confirmed infections.

Full table

Survival

Overall survival (Figure 3) and progression-free survival (similar figure, not shown) were similar in both groups (P=0.57, and 0.21, respectively).

Figure 3.

Overall survival. Thick line: steroid prophylaxis; thin line: no steroid prophylaxis.

Full figure and legend (14K)

Discussion

Our study demonstrates that steroid prophylaxis decreases the risk of ES following AHSCT by almost 10-fold, without increasing the risk of infection. In addition, decreasing the risk of ES is associated with a shorter mean length of hospital stay. It has no effect on overall survival or progression-free survival.

Limitations of this study include the fact that it was not a randomized trial. However, most BMT centers now appreciate the dramatic response of ES to corticosteroid therapy. The diagnosis of CRB was not based on quantitative blood cultures, endoluminal catheter brush, or differential time to detection. Since we only included microbiologically confirmed infections in this review, clinically and radiographically diagnosed infections, such as cellulitis, sinusitis and pneumonia, may not have been captured. However, the measures for data collection were identical in both groups, and the incidence of infection was similar to prior studies in AHSCT recipients.² Since our hospital is a tertiary care center, patients may have been readmitted to local hospitals. This study was not powered to detect differences between patients who developed ES, and those who did not, in those receiving steroid prophylaxis.

Similar to the findings of our study, previous series have shown that the most common type of infection occurring within the first month following AHSCT was CRB, either due to streptococci,^16,17 or coagulase negative staphylococci.¹⁸ Risk factors for the development of these infections included duration of neutropenia,^17,18 degree of neutropenia,¹⁸ and mucositis.¹⁸ Unlike the current report, reactivation of VZV occurring 1–6 months following AHSCT was seen in 13–15% of patients in previous series.^16,18 Risk factors for this infection included a diagnosis of myeloma,¹⁶ previous use of fludarabine,¹⁶ administration of total body irradiation,¹⁶ and having a CD4⁺ T cell count <200/l on day +30.¹⁸ The use of prophylactic acyclovir for up to 1 year following AHSCT in our series is the most likely reason for this difference. Prophylaxis against Pneumocystis jiroveci pneumonia was not administered, and no cases of infection with this organism occurred. It is reassuring that there were no invasive fungal infections in either group.

Since steroid prophylaxis for ES has not been standard in any of the previous series, its impact on the risk of infection has not been previously assessed. We continue to administer steroids to prevent ES in AHSCT recipients. The optimum dose and duration for steroid prophylaxis in this setting remains to be determined.

References

References