Skip to Content
U.S. National Institutes of Health
Last Updated: 05/29/08

Phase II Window Studies in Pediatric Oncology Meeting Report

Introductory Note

The Cancer Therapy Evaluation Program (CTEP) of the National Cancer Institute sponsored a meeting on July 22,1997 in Bethesda, MD to address issues surrounding the appropriate conduct of pediatric clinical trials using a "phase II window" design and to consider what information needs to be conveyed to patients and their families who are considering participation in this type of clinical trial. The report of the meeting was prepared by a committee that was constituted so as to represent the viewpoints of clinicians, clinical researchers, family/patients, and bioethicists. The committee was chaired by Dr. Thomas Beauchamp (bioethicist, Kennedy Institute of Ethics at Georgetown University). Other members included Dr. John Lukens (pediatric hematologist/oncologist, Vanderbilt University), Dr. Stephen Sallan (pediatric hematologist/oncologist, Dana-Farber Cancer Institute), Dr. Rebecca Pentz (bioethicist, MD Anderson Cancer Center), Dr. Susan Ellenberg (Biostatistician, Food and Drug Administration), and Dr. Susan Weiner (officer in North American Brain Tumor Coalition and Children's Brain Tumor Foundation). Ms. Mary McCabe (NCI) served as the "Executive Secretary" for the committee. At the meeting, presentations were made by representatives from the Children's Cancer Group (CCG), the Intergroup Rhabdomyosarcoma Study Group (IRS), the Pediatric Oncology Group (POG), St. Jude Children's Research Hospital (St. Jude), and CTEP.


The committee appreciates the time and care taken by the investigators in preparing for this meeting, and considers that the data and background material prepared by the Children’s Cancer Group (CCG), particularly its white paper, the Intergroup Rhabdomyosarcoma Study Group (IRS), the Pediatric Oncology Group (POG), and St. Jude Children’s Research Hospital (St. Jude), represent the most current scientific information available.


Standard phase II studies have identified many agents that have improved outcome when incorporated into front line therapy. These agents include the combination of ifosfamide and etoposide for the Ewing’s family of tumors, high-dose cytosine arabinoside for acute myelogenous leukemia, and all-trans retinoic acid for acute promyelocytic leukemia.

Several agents considered inactive in standard phase II studies have shown activity when evaluated in the phase II window setting in previously untreated patients. These include ifosfamide for osteogenic sarcoma, neuroblastoma, and rhabdomyosarcoma; carboplatin for neuroblastoma; topotecan for rhabdomyosarcoma; and melphalan for rhabdomyosarcoma. The extent to which these agents will influence outcome when incorporated into standard therapy remains to be determined. It will be several years before assessment can be made of the overall therapeutic merit of agents that lack promise in conventional phase II studies but are shown to have activity in untreated patients.

As the effectiveness and the intensity of treatment have changed in recent years, so too have the numbers and characteristics of patients eligible for standard phase II studies. At current cure rates, fewer than 30% of children with cancer will have a tumor recurrence and thus be eligible for such studies in the future. These patients will also have received more intensive initial therapy than in the past. Both of these issues were presented by the investigator participants to support the use of phase II window studies.

The investigators representing POG, IRS, and St. Jude were in favor of the phase II window approach and thought that the design had merit in phase II studies of pediatric tumors. Except for reservations expressed by the CCG representative during the discussion, the views of investigators who opposed the phase II window approach were not represented at the meeting. Hence, the committee was at a disadvantage in being able to identify controversial issues associated with this approach.

The development of recommendations regarding this study design was difficult because the data presented were, by necessity, combined from various protocols that studied a range of tumors, drug combinations, and durations of the phase II window.

General Issues

When proposing to conduct these phase II window trials, investigators should carefully distinguish between the therapeutic intent and the experimental intent of the trial. Currently, there appears to be a blurring of this distinction. Therapeutic intent should always be a distinct goal for these studies.

The committee members acknowledge that, in today’s pediatric oncology setting, it is difficult for a clinical investigator to avoid enrolling his/her own patients in his/her own study. However, the committee advises investigators to be cognizant that these dual roles create a potential conflict of interest, and encourages the pediatric community to remain vigilant in its responsibility to act in the best interest of the patient while conducting clinical investigation.

Risks and Benefits

At this time, the evidence is inconclusive as to whether phase II window trials harm or benefit participating patients. Since any clinical trial intervention that involves non-trivial risk should also offer potential benefit to the patient, continued assessment of the benefits and risks of this study design is needed.

In considering the potential for increased patient risk in these trials, the committee raised the following areas of concern:

  • Risk of tumor progression
    Tumor progression occurred in phase II window studies of carboplatin for osteosarcoma and of paclitaxel for neuroblastoma. In the absence of a randomized control, it is uncertain whether comparable progression would have resulted with immediate use of multi-agent regimens.
  • Risk of additional toxicities
    Unique toxicities, such as Fanconi’s syndrome induced by ifosfamide, are of concern.
  • Risk of jeopardizing future therapy
    In three phase II window studies, toxicities associated with the trials might have compromised subsequent standard therapy. Examples include:
    1. the use of thiotepa in CNS tumors resulted in unacceptable myelosuppression;
    2. the use of idarubicin in small cell carcinoma of the lung jeopardized subsequent use of CVE; and
    3. the use of melphalan in rhabdomyosarcoma hindered later treatment with VAC due to stem cell toxicity.

Until additional data are developed and analyzed, the level of patient risk in phase II window trials remains a significant concern.

  • Benefits
    Phase II window trials also have the potential to benefit the patient. If a patient responds to the window agent, and that agent is then incorporated into that patient’s treatment regimen, the patient has the opportunity to receive an additional agent known to be active, possibly enhancing the probability of achieving long-term remission. The potential benefit of window treatment for a patient who receives the new agent only during the window trial is less clear, and this limitation of the new agent is therefore discouraged.

The current informed consent process and supporting documents are inadequate. They do not identify the phase II window as a separate study and do not address the risks of and alternatives to the window treatment. The investigator must be explicit and forthright about the investigational nature of the phase II window approach, even at the risk of losing patient participation in the clinical trial.

Therefore, the committee makes the following specific recommendations:

  • Although consent to the window component of the study need not be given in a separate document, the window component should be clearly distinguished from the main treatment or study.
  • A model informed consent process and document should be developed for use in phase II window studies. This model should specify the voluntary nature of study participation, emphasizing that the individual authorized to represent the patient has the final decision about participation.
  • Provisions in the model consent document should include:
    1. a clear presentation of treatment alternatives;
    2. provision for children’s assent or refusal;
    3. a clear explanation as to how the window treatment differs from therapy to be administered following the window;
    4. delineation of the known possible risks of a phase II window study, including the risks associated with delaying standard therapy and the risk of jeopardizing eligibility for future therapies;
    5. the effect on the patient’s quality of life (for example, imposing additional procedures and extending the duration of therapy);
    6. pre-clinical and clinical data relating to the efficacy and toxicities associated with the phase II agents.

Study Design

  • The development of a phase II window study should be based on compelling evidence, such as proven efficacy in adults, a low toxicity profile, novel mechanisms of action, and any relevant positive preclinical data.
  • The committee encourages protocol design that allows incorporation of the phase II window agent into the phase III portion of the study when activity of the investigational agent is observed in an individual patient.
  • Most phase II window studies utilize two courses of the agent under evaluation before instituting phase III treatment. The committee recommends that surrogate measures of tumor response, such as serum concentration of tumor markers or needle biopsies, be used when possible to shorten the duration of the phase II window to one course and thereby minimize the risk of disease progression and of drug toxicity.
  • The committee recommends that phase II window studies include a prospective assessment of the impact of the window treatment on quality of life, when feasible.
  • In some clinical trials presented to the committee, patients were required to participate in the window portion of the protocol because the definitive treatment outcome was determined, in part, by the outcome of the window portion. The committee disagrees with this design and recommends that participation in the window portion of the protocol should be elective, with the phase II window being separable from the remainder of the study.

Appropriate Participants

Currently, 90% of phase II window protocols are for previously untreated patients. Consideration should be given to promoting the incorporation of the phase II window design in the treatment of patients at first relapse, who are clearly at higher risk than previously untreated patients, whatever the tumor type.

The committee had difficulty determining an appropriate definition for a "high-risk" population, since this designation is variable and has been applied to children with expected cure rates with standard therapy as high as 50%. Some individual committee members were uncomfortable with designating patient groups with these cure rates as "high-risk" and with the inclusion of such patient populations in window studies. These members were more comfortable with cure rates of 20% or less.

  • After much deliberation about whether to include "lower risk" patients in window studies, the committee does not feel it can establish a fixed "cutoff" based on an expected cure rate for entry into window studies, but recommends that priority should be given to patients with expected probabilities of cure well under 50%. This recommendation is made with the acknowledgment that parents will not have uniform views about what cure rate defines "poor prognosis."
  • After the completion of a more extensive evaluation of the risks and benefits of the phase II window study design, as recommended in this report, the risk/benefit ratio might be favorable enough for phase II window trials to be incorporated into the treatment of lower risk patients. Parents of lower risk patients would then have the choice of having their child participate in the phase II window study. However, all patients should have access to the phase III study regardless of their participation in the phase II component.
  • The committee suggests that any trial using lower risk patients should assess the parents’ satisfaction with the informed consent process, with specific attention to parent and patient preferences about being offered such a choice.
  • The committee also suggests that additional designs of phase II protocols be considered and developed.

General Recommendations

  • A more extensive evaluation of the benefits and risks of the phase II window approach is required. The data that were presented were not definitive and did not provide evidence that the phase II window agents improve outcome when added to standard therapy. Limited evaluation of data from existing protocols that have used the window approach suggests that this protocol design may not produce unacceptable harm to high-risk patients. However, the existing evidence is not conclusive. The committee recommends that currently existing data be analyzed extensively, but questions whether collection of further observational data will be of value.
  • The most informative way to evaluate risk/benefit considerations would be to assess the window strategy systematically, by conducting randomized trials in which patients would be randomized to a window study preceding administration of standard treatment, or to standard treatment alone. Given the small number of patients treated within each tumor type, a meta-analysis of a series of such trials might be the most appropriate way to evaluate the window strategy. This strategy would eliminate the arbitrary nature of assigning each patient a risk status and would place the decision to participate in the randomization with the family rather than the investigator. These studies should also include assessment of parent satisfaction with the process. The feasibility of such trials should be considered.
  • A coordinated national strategy to develop new treatments for pediatric patients should be developed in order to assure the most efficient development of cancer therapies. This strategy would include a systematic evaluation of the pediatric cancer population to identify the prevalence of various diseases as well as a review of available trials. Coordinated efforts by different groups and institutions will be necessary to establish a national strategy that is successful in implementing nationwide protocols. The Cancer Therapy Evaluation Program may consider collaborating with the extramural community to elicit changes in protocol design and, as part of the national effort to enhance drug development, utilize the intergroup structure already in place to coordinate the prioritization of trials for this patient population.


The committee is supportive of window studies with the following provisions:

  • There must be a clear presentation of the phase II component in informed consent documents.
  • There should be freedom not to participate in the phase II component without jeopardizing the opportunity to participate in the phase III component of the study.
  • The duration of the window component of the study must be kept as short as possible.
  • The highest risk patients should be given first priority for phase II window participation, with lower risk patients participating only after further evaluation reveals a favorable risk/benefit ratio.
  • The contribution of the window studies to pediatric cancer treatment must continue to be evaluated.

The committee thanks the staff of the Cancer Therapy Evaluation Program for organizing the meeting on this important research question.

NOTE: The committee preparing this report was constituted so as to represent the viewpoints of clinicians, clinical researchers, family/patients, and bioethicists. The committee was chaired by Dr. Thomas Beauchamp (bioethicist, Kennedy Institute of Ethics at Georgetown University). Other members included Dr. John Lukens (pediatric hematologist/oncologist, Vanderbilt University), Dr. Stephen Sallan (pediatric hematologist/oncologist, Dana-Farber Cancer Institute), Dr. Rebecca Pentz (bioethicist, MD Anderson Cancer Center), Dr. Susan Ellenberg (Biostatistician, Food and Drug Administration), and Dr. Susan Weiner (officer in North American Brain Tumor Coalition and Children’s Brain Tumor Foundation). Ms. Mary McCabe (NCI) served as the "Executive Secretary" for the committee. The meeting occurred in Bethesda, MD on July 22, 1997.