Prophylaxis of Pneumocystis carinii pneumonia with atovaquone in children with leukemia




Despite extensive studies of atovaquone in human immunodeficiency virus (HIV)-infected patients, there is little information about its efficacy as a prophylactic agent for Pneumocystis carinii pneumonia (PCP) in pediatric patients with cancer. Therefore, a retrospective analysis was conducted to determine the incidence of PCP in pediatric patients who received prophylactic atovaquone during treatment for acute leukemia.


We reviewed the medical records of all patients treated at our institution for acute lymphoblastic leukemia or acute myeloid leukemia between 1994 and 2004. Only patients who were intolerant of trimethoprim-sulfamethoxazole (TMP-SMZ) and received atovaquone prophylaxis were included in the analysis.


Eighty-six patients were unable to tolerate TMP-SMZ and received daily atovaquone for PCP prophylaxis. PCP was not diagnosed in any patient who received atovaquone prophylaxis: the upper limit of the 95% confidence interval (CI) was 1.74 per 100 person-years.


Atovaquone is an efficacious alternative for PCP prophylaxis in pediatric patients who have leukemia and are intolerant of TMP-SMZ. Cancer 2007. © 2007 American Cancer Society.

Pneumocystis carinii pneumonitis (PCP) is an opportunistic infection that is life-threatening in immunocompromised persons.1 The first reported case of PCP in the US was a pediatric patient in 1956.2 Subsequently, sporadic cases were reported in children whose primary immune function was impaired.3 In 1970, Johnson and Johnson4 described an outbreak of PCP in 19 pediatric patients who were undergoing chemotherapy for cancer at St. Jude Children's Research Hospital. The impaired host immune function in this pediatric cancer cohort was a secondary effect of the immunosuppressive nature of their anticancer therapies.5, 6

Options for the treatment of PCP were limited to pentamidine isethionate until the mid-1970s, when Hughes et al.7 demonstrated the efficacy of trimethoprim-sulfamethoxazole (TMP-SMZ) for treating PCP in pediatric patients. PCP continued to contribute to morbidity and mortality in patients with cancer until chemoprophylactic strategies with TMP-SMZ were demonstrated.8, 9 Unfortunately, the usefulness of TMP-SMZ is limited in some patients by unacceptable toxic effects that include rash, fever, and myelosuppression.7, 10

Atovaquone is an effective PCP preventive agent for human immunodeficiency virus (HIV)-infected patients who cannot tolerate TMP-SMZ, with similar efficacy to aerosolized pentamidine.11 In children with HIV infection atovaquone is safe and well tolerated when taken daily for PCP prophylaxis.12 Despite extensive studies of atovaquone in HIV-infected patients, there is little information about its efficacy as a PCP prophylactic agent in pediatric patients with cancer. We therefore reviewed our institution's experience with the use of atovaquone in pediatric patients who were being treated for leukemia.


Between 1994 and 2004, approximately 1000 patients at St. Jude Children's Research Hospital (St. Jude) were enrolled on institutional and noninstitutional leukemia protocols. During this period patients intolerant of TMP-SMZ received alternative PCP prophylaxis that continued until 6 months after the completion of leukemia therapy or until chronic graft-versus-host disease was resolved. Patients included in this analysis were children enrolled in institutional protocols for the treatment of newly diagnosed acute lymphoblastic leukemia (Total Therapy Studies XIIIA, XIIIB, XIV, and XV) or acute myeloid leukemia (AML91, AML97, AML02) or those enrolled on other protocols for the treatment of these leukemias.13–18 The analysis also included 2 patients treated for myelodysplastic syndrome and juvenile chronic myeloid leukemia. Patient medical records were reviewed to determine demographic data, primary diagnosis, disease status at the time atovaquone therapy was initiated, treatment protocol, length of time on atovaquone therapy, and the occurrence of PCP. TMP-SMZ intolerance included but was not limited to the patient's history of intolerance or hypersensitivity to sulfonamides or trimethoprim, fever, rash, urticaria, allergic reaction, neutropenia, or elevation of transaminases. All leukemia protocols were approved by the St. Jude Institutional Review Board and written informed consent was obtained from patients, parents, or guardians. This retrospective study was conducted after approval by the St. Jude Institutional Review Board.


Table 1 shows patient demographic and clinical characteristics at the time of receipt of atovaquone. Eighty-six patients with hematologic malignancies received atovaquone for prevention of PCP. Their mean age was 10.7 years (SD, 5.8 years). Sixty-five patients were in first complete remission and 21 were being treated for relapsed or refractory disease during receipt of their atovaquone prophylaxis. Thirty-seven of the 86 patients underwent allogeneic hematopoietic stem cell transplantation (HSCT) for leukemia while they received atovaquone.

Table 1. Characteristics of Patients Treated With Atovaquone Prophylaxis
CharacteristicNo. (%)
  1. HSCT indicates hematopoietic stem cell transplantation.

 Women38 (44.2)
 Men48 (55.8)
 African American12 (14.0)
 Hispanic4 (4.6)
 Other6 (7.0)
 Caucasian64 (74.4)
Primary disease
 Acute lymphoblastic leukemia57 (66.3)
 Acute myeloid leukemia27 (31.4)
 Juvenile chronic myeloid leukemia1 (1.2)
 Refractory anemia1 (1.2)
 AML917 (8.1)
 AML978 (9.3)
 AML023 (3.5)
 Total XIIIA5 (5.8)
 Total XIIIB13 (15.1)
 Total XIV7 (8.1)
 Total XV20 (23.3)
 Other protocols23 (26.7)
Allogeneic HSCT
 Yes37 (43.0)
 No49 (57.0)

The time and efficacy of atovaquone chemoprophylaxis are shown in Table 2. As expected from the duration of treatment of the primary disease, the median duration of atovaquone therapy was longer for the 57 pediatric patients with acute lymphoblastic leukemia (584 days) than for the 29 patients with myeloid malignancy (204 days). The total person-years for atovaquone therapy was 172.1. None of these patients developed PCP. Therefore, the incidence of PCP in this group of patients was zero (95% confidence interval [CI]: 0–1.74) per 100 person-years as determined by Poisson exact statistics.

Table 2. Length of Atovaquone Prophylaxis and Incidence of PCP
  • PCP indicates Pneumocystis carinii pneumonia.

  • *

    Total person-years = 172.1.

No. of patients5729
Atovaquone therapy584 (74–3953)*204 (68–3050)*
Median days (range)
 Incidence of PCP00


In this analysis, we found atovaquone to be effective in preventing PCP in pediatric leukemia patients who were intolerant of TMP-SMZ. This result was equivalent to the success in chemoprophylaxis achieved with TMP-SMZ (given daily for 2 years) in the Hughes et al.8 prospective study in 80 pediatric patients with cancer. In that study, 17 of the 80 patients who did not receive chemoprophylaxis with TMP-SMZ (21%) developed PCP, whereas no cases of PCP were observed in those pediatric patients who received prophylactic TMP-SMZ.8 On the basis of the reported results from this early study of prophylaxis by Hughes et al.,8 we estimate that the PCP incidence was 24.3 per 100 person-years (95% CI: 14.4–38.4) in the placebo group. Additionally, Hughes et al.9 did not observe any cases of PCP in the 74 patients with cancer treated at St. Jude when TMP-SMZ was administered on an intermittent schedule (3 consecutive days per week) over a 2-year period.

The current retrospective pediatric leukemia patient cohort was similar in size to that of Hughes et al.8, 9 and had received even more intensive therapy. Intensive anticancer therapy in the recent era, in the absence of PCP prophylaxis, continues to place patients at risk for PCP. A report by Lyytikainen et al.19 in 1996 provides an indication of the incidence of PCP in acute leukemia in patients receiving no PCP prophylaxis during intensive chemotherapy. Of the 29 cases studied, PCP occurred in 7 patients (24%)19; these results are similar to the 20% incidence reported by Hughes et al.8 in 1977 for acute lymphoblastic leukemia patients before PCP prophylaxis. Furthermore, in 2001 Poulsen et al.20 identified PCP in 13 of 71 Danish children (18%) with ALL in whom PCP prophylaxis was stopped after the induction and consolidation phases of chemotherapy. For the current cohort at St. Jude, atovaquone provided effective PCP chemoprophylaxis for the 65 patients who received intensive contemporary frontline therapy for newly diagnosed leukemia as well as for the 21 patients who received even more intensive therapy for relapsed or refractory leukemia. Moreover, atovaquone chemoprophylaxis was efficacious in the 37 patients who underwent allogeneic HSCT during this 10-year period. As an alternative agent for PCP prophylaxis, we continue to use atovaquone in leukemic or stem cell transplant patients at an oral dose of 30 mg/kg once daily, with the maximum daily dose of 1500 mg/day (prophylactic dose for adults).

In comparison studies of PCP in patients with acquired immunodeficiency syndrome (AIDS), atovaquone was a less effective therapy than TMP-SMZ but had fewer treatment-limiting adverse effects.21 The treatment failure rate for atovaquone was 20%, whereas that of TMP-SMZ was 7%.21 Conversely, treatment-limiting adverse effects occurred in 7% of patients receiving atovaquone and 20% of those on TMP-SMZ, giving overall therapeutic success rates of 62% and 64%, respectively. Therapeutic efficacy was related to the plasma concentration of atovaquone. It is noteworthy that a poorly soluble tablet formulation of atovaquone (no longer in use) was used in the early trials. The improved suspension formulation was introduced in 1995, providing a 2-fold increase in bioavailability.22 Although atovaquone is less effective than TMP-SMZ in treating PCP, its activity and its well-tolerated toxicity profile make it a useful alternative agent for patients intolerant of TMP-SMZ. Atovaquone has been approved by the FDA for the treatment of mild to moderate PCP in patients who cannot tolerate TMP-SMZ.

For PCP prophylaxis in TMP-SMZ-intolerant individuals, the use of atovaquone has been reported and compared with alternative prophylactic agents such as pentamidine and dapsone.11, 23, 24 However, many of the prospective clinical comparative studies have been limited to HIV-infected adults. Chan et al.11 observed daily atovaquone to have similar efficacy when compared with monthly aerosolized pentamidine for PCP prophylaxis in HIV-infected individuals intolerant of TMP-SMZ. In another comparison study of atovaquone and pentamidine in adults with AIDS and PCP, atovaquone showed therapeutic activity and fewer adverse effects.25 In these studies atovaquone was well tolerated and rash was often cited as an adverse event. El-Sadr et al.24 compared atovaquone to dapsone and noted both agents to be similarly effective for PCP prophylaxis in HIV-infected patients (aged 13 years and older). Although it is convenient to administer a PCP prophylactic agent monthly, the effective administration of aerosolized pentamidine may not always be possible in younger pediatric patients. Moreover, methemoglobinemia is an adverse event associated with dapsone administration and has been recently reported by Mandrell et al.26 in 4 pediatric patients from St. Jude and by Williams et al.27 in 3 of 15 children with leukemia—all patients were intolerant of TMP-SMX and were receiving dapsone for PCP prophylaxis. Clinical pediatric studies in patients with HIV using atovaquone for PCP prophylaxis or for the prevention of serious bacterial infections have shown this agent to be well tolerated in children as young as 1 to 3 months of age.12, 28

Adverse events, including rash and diarrhea, have been reported in patients who were receiving atovaquone chemoprophylaxis for PCP.12, 21, 25, 29 However, prospective studies in patients with HIV, and recipients of autologous hematopoietic cell transplants, have confirmed its use as an alternative to TMP-SMZ.12, 29 The frequency of intolerance to prophylactic TMP-SMZ ranges from <6% to 20% in patients with cancer; fever, rash, abdominal pain, neutropenia, and thrombocytopenia are among the most commonly reported adverse effects.9, 10, 30 Neutropenia is of particular concern in oncology, not only because of the risk of infection but also because chemotherapy may be delayed by prolonged neutropenia. Colby et al.29 conducted one of the few prospective, randomized studies that compared atovaquone and TMP-SMZ in transplant patients. PCP chemoprophylaxis was given to 39 patients prior to their autologous peripheral blood stem cell transplants for solid tumors or hematologic malignancies and was restarted after engraftment.29 Engraftment was similar in both groups. No cases of PCP were observed in either group. None of the patients who received atovaquone had any adverse reaction, whereas 8 of the 19 patients (40%) who received TMP-SMZ were removed from the study because of drug intolerance. Similar efficacy using atovaquone prophylaxis for PCP was observed in 28 orthotopic liver transplant patients intolerant of TMP-SMZ.31 Although 50% of the patients were noted to have side effects attributed to atovaquone (diarrhea, bloating, or abdominal pain most frequently reported), Meyers et al.31 observed no cases of PCP in these solid organ transplant patients treated for 1 year with daily atovaquone.

Although the same agents used for PCP prophylaxis in patients with HIV are efficacious in pediatric patients with cancer, it is less well known whether these alternative agents for PCP prophylaxis alter the metabolism of anticancer drugs or whether they have other adverse effects on this patient population. In this regard, the effect of atovaquone on etoposide in pediatric patients has been studied by Van de Poll et al.32 at St. Jude. They compared the intraindividual effects of atovaquone and TMP-SMZ on the pharmacokinetics of the anticancer agent etoposide by using a crossover study design. Nine patients were randomly assigned to receive either TMP-SMZ or atovaquone during continuation chemotherapy. Etoposide pharmacokinetics were then analyzed for each patient after a single dose of intravenous etoposide. Approximately 8 weeks later, each patient was crossed over to the other PCP chemoprophylactic agent and the etoposide pharmacokinetic studies were repeated. The results showed a slight increase in the area-under-the-concentration curve of etoposide and its metabolite (median increase, 8.6% and 28.4%, respectively) with atovaquone. The long-term significance of this effect of atovaquone on etoposide pharmacokinetics is unknown. Such findings emphasize the need for prospective pharmacokinetic studies in patients receiving anticancer therapies when new agents (such as atovaquone) are introduced on a long-term basis, particularly when drugs have similarities in their metabolic pathways.

Although the data presented here are the result of a single-center retrospective analysis and are subject to the limitations inherent in such a study, no episodes of PCP were observed in the 86 pediatric patients who had leukemia and received atovaquone chemoprophylaxis. This finding suggests that atovaquone is an effective alternative agent for PCP chemoprophylaxis in pediatric patients with leukemia who are intolerant of TMP-SMZ.


We thank Dequing Pei and Shelly Lensing for statistical assistance.