Establishment of baseline toxicity expectations with standard frontline chemotherapy in acute myelogenous leukemia

In evaluating the toxicities of new agents or regimens, the regulatory agencies often consider similar sets of criteria for solid tumors and leukemias. Patients with acute leukemia invariably present with a compromised bone marrow status, severe myelosuppression, and cytopenias; receive intensive chemotherapy; and frequently have infections requiring hospitalizations and affecting organ functions (particularly pulmonary, hepatic, and renal). These preexisting or expected complications are often required to be submitted as adverse or serious adverse events. This requirement results in several problems: (1) overburdening the investigators and regulatory agencies with unnecessary documentation, thus taxing the physical and economic infrastructures; (2) diverting time and attention from potentially important drug- or treatment-associated significant issues; and (3) attributing toxicities to a drug or treatment when that toxicity is not treatment related.

To establish baseline expectations of toxicities before and during therapy in acute leukemia, we reviewed our experience in newly diagnosed patients with acute myelogenous leukemia (AML) receiving induction therapy. Since older patients are expected to have perhaps worse baseline organ functions and to tolerate intensive chemotherapy less well (therefore to have more frequent organ complications), we analyzed the data also in the younger versus the older patient subgroups.

Adults with a diagnosis of AML or high-risk myelodysplastic syndrome (MDS) referred to the Leukemia Department at M. D. Anderson Cancer Center from 1990 to 2006, and who received frontline intensive chemotherapy on current protocols, were reviewed.^1-5  Patients with acute promyelocytic leukemia were excluded. Informed consent was obtained according to institutional guidelines and in accordance with the Declaration of Helsinki, and Institutional Review Board approval was granted by the University of Texas M. D. Anderson Cancer Center (MDACC).

We analyzed the incidence of grade 3 or 4 toxicities both at baseline and during induction therapy within each age group and regimen.

For the purpose of the study, 3 end points for mortality were considered: (1) 6-week mortality, defined as any death occurring in the first 6 weeks regardless of disease status or cause of death; (2) early mortality, defined as death in 14 or fewer days; and (3) total induction death, defined as death occurring without a documented complete remission at any time during induction. The latter is more subjective than the other 2 definitions, since the decision and timing of a second induction course are variable. Variables were compared by the chi-square test.

The patient characteristics are shown in Table 1; 1543 patients were identified; their median age was 60 years (range, 16 to 87 years); 644 (42%) were females. The treatment programs were summarized in previous reports.^1-5  The numbers of patients on each treatment are summarized in Table 2. We then analyzed the incidence of baseline events considered to be grade 3 or 4 by the National Cancer Institute, Cancer Therapy Evaluation Program (NCI-CTEP) criteria version 3.0. Grade 3 or 4 anemia was noted in 50%; neutropenia, in 48%; and thrombocytopenia, in 52%. Febrile episodes with or without a documented infection were noted in 36%. Baseline elevations of creatinine (> 176.8 μM [2 mg/dL]) and bilirubin (> 34.2 μM [2 mg/dL]) were observed in 3% and 3%, respectively. Admission to the intensive care unit was needed in 9%. Those baseline incidences tended to be higher in older patients (Table 1). The distribution of patients in different treatment groups based on age and karyotype was different due to the study inclusion criteria. The more standard idarubicin and cytarabine studies involved mostly better prognosis patients (younger; diploid karyotype) over the temporal course of our research programs. The more investigational studies were designed mostly for worse prognosis patients (older; unfavorable karyotype). This is demonstrated in Table 3. These differences could explain the differences in complete remission (CR) rates and survival, as previously reported.^4,5 

It is often difficult to distinguish mortality during induction by whether it is (1) treatment or drug related, (2) related to the baseline patient condition, or (3) related to ineffective therapy and poor leukemia control. Realistically, it is a combination of the 3 factors and should be evaluated as such. To establish the expected baseline mortality (for comparison with future investigational trials), we analyzed early (2-week), 6-week, and induction mortality by treatment regimen and within age groups. Table 4 shows the 3 mortality end point rates overall and by regimen. The overall induction mortality rate was 20%, and ranged from 13% to 33% by regimen. This range was mostly age dependent. Table 5 shows the mortality rates in younger versus older patients within each regimen. As expected, mortality rates were significantly higher in older patients. Survival also differed among the various regimens compared with the more standard idarubicin and cytarabine (IA) regimen. In previously published studies, using multivariate analyses, the CR and survival rates were similar with different regimens.^4,5 

As expected, all patients developed one form or another of grade 3 or 4 myelosuppression (Table 6). Febrile episodes occurred in 94%, documented infections occurred in 64%, and admission to the intensive care unit was required in 28%. Other extramedullary toxicities included renal dysfunction in 3% and liver dysfunction in 22%. Respiratory support and dialysis were required in 16% and 7%, respectively. Of interest is the higher rate of dialysis than grade 3 or 4 renal toxicities. This is due to the definition of grade 3 toxicity that requires a creatinine level 3 to 6 times or more of the upper limit of normal (ULN). The ULN in our institution is 132.6 μM (1.5 mg/dL); a grade 3 toxicity is defined as a creatinine value higher than 397.8 μM (4.5 mg/dL), a level before which dialysis may have already been initiated. The rate of grade 2 renal toxicity varies from 12% to 20% according to regimen. The rates of nonhematologic grade 3 or 4 toxicities were significantly higher, as expected, in older patients (Table 7). This was particularly relevant for organ dysfunctions such as pulmonary, hepatic, cardiac, and renal failure. This is likely due to the worse baseline organ dysfunction in older patients and/or their worse tolerance of chemotherapy.

Reporting grade 3 or 4 toxicities on investigational regimens, which are not treatment related and not higher than baseline expectations, burdens heavily the infrastructures of regulatory organizations (Food and Drug Administration [FDA], NCI-CTEP) and investigators (who need to report within 24 hours or 7 days).^6,7  Such toxicities can still be reported and analyzed in the periodic (6-month or annual) reports and in the final study analysis/publication. Based on our experience, reporting a serious adverse event takes on average 1 to 3 hours of a research nurse/researcher at the site of investigation, as well as significant time and effort at the site of the regulatory organization. Others have reported a mean of 35 minutes and up to 6 hours for clinical trial documentation including serious adverse event reporting.⁸ 

This issue is of significance for tumors in which there is a baseline compromise of organs such as the bone marrow, and primarily concern 2 disciplines: leukemias (and perhaps some lymphomas with marrow involvement) and stem cell transplantation. Following the current FDA/NCI-CTEP guidelines, “serious adverse events” would have to be reported before starting any therapy in more than 50% of patients, and on therapy in more than 90% of patients. Reporting such serious adverse events may at times vary depending on the interpretations of the guidelines published by the FDA.⁹ 

In this study, the first of its kind, we aimed to establish baseline expectations for these serious adverse events in patients with AML, before therapy as well as during induction therapy. Although we analyzed this in a large cohort of patients (1534 patients), we recommend using the anthracycline-cytarabine regimens as the baseline comparator for future reference. We also realize that the incidence of some toxicities (eg, dialysis, ICU admission, or respiratory support) may differ from one institution to the other depending on the institutions' guidelines. However, we hope that this study would initiate a fruitful dialogue between leukemia researchers, regulatory bodies, and pharmaceutical companies interested in developing drugs in leukemia, as to what should really be reported. Based on our analysis, it appears that many of the grade 3 or 4 toxicities that some pharmaceutical companies insist on reporting to the FDA as serious adverse events are not truly so. Consequently, reporting them as serious adverse events overburdens the infrastructure of the investigators' institutional (Office of Protocol Research, Institutional Review Board [IRB]) and the government regulatory bodies. While the ratio of true to overall reported serious adverse events is unknown, we estimate, based on our reporting experience that of approximately 10 “serious adverse events” reported to the FDA in leukemia studies, perhaps only 1 is a true serious adverse event (ie, related to the drug or treatment). In another study, 31% of reported grade 3 to 5 toxicities were unrelated or unlikely to be related to the drug.⁶ 

Because of these important concerns in centers where leukemia research is prominent, several senior faculty representing the leukemia research department met with colleagues who were members of the M. D. Anderson Institutional Review Board to develop formal guidelines for reporting adverse events on leukemia clinical studies. These are outlined in Table 8. These are the events that require individual case report forms and reporting within 24 hours or 7 days. All other grade 3 or 4 toxicities will also be reported in the periodic summaries as well as in the final analyses/reports. These proposals should be considered as a work-in-progress, and perhaps as a starting point for initiating discussions that will hopefully be beneficial to all parties involved including investigators, regulatory organizations, and pharmaceutical companies.

Contribution: E.A. wrote the paper and analyzed the data; J.C. and S.O. reviewed the paper and made corrections; S.P. and M.B.R. collected and analyzed the data; E.E., M.M., M.K., and E.J.F. reviewed the paper and made corrections; and H.K. wrote the paper and analyzed the data.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: Hagop Kantarjian, Department of Leukemia, M. D. Anderson Cancer Center, 1515 Holcombe Blvd, Box 428, Houston, TX 77030; e-mail:hkantarj@mdanderson.org.