CHILDHOOD CANCERS are rare, accounting for less than 1% of the one million patients diagnosed with cancer in the United States each year.¹ There are between 12,000 and 15,000 pediatric patients per year for all pediatric cancer combined, compared with the more than 170,000 adult patients who are diagnosed with lung cancer, 175,000 adult patients who are diagnosed with breast cancer, and 179,000 adult patients who are diagnosed with prostate cancer annually. Nevertheless, cancer remains the most common fatal disease of childhood.

More than 100 drugs are approved by the Food and Drug Administration (FDA) for the treatment of malignancies. Bleyer et al² noted that about 30 approved drugs are currently used in pediatric oncology. Fifteen of these drugs, listed in Table 1⇓, have been labeled for use in children. The inclusion criteria for this listing are 1) a specific reference to pediatric patients or a pediatric malignancy in the “Indications and Usage” section, 2) approval in a disease setting exclusive to pediatrics (eg, neuroblastoma or Wilms tumor), and 3) specific information regarding pediatric dosing in the “Dosage and Administration” section. The first two criteria provide the most unequivocal support for the use of a drug in pediatric oncology. The 15 drugs that meet these criteria are l-asparaginase, cyclophosphamide, cytarabine, dactinomycin, daunorubicin, doxorubicin, lomustine, mercaptopurine, methotrexate, procarbazine, teniposide, thioguanine, tretinoin, vinblastine, and vincristine.

The FDA approved these drugs predominantly in the 1950s and 1960s. Since 1980, only one new molecular entity, teniposide (Vumon, formerly known as VM-26; Bristol-Myers Squibb, Princeton, NJ), has been labeled for use in a pediatric malignancy (refractory leukemia) compared with more than 50 new molecular entities approved for adult oncology indications. Possible reasons for the disparity between adult and pediatric FDA approvals include the relative rarity of childhood malignancies, the histopathologic and biologic differences between many adult and pediatric tumors, and the limited number of pediatric patients eligible for pharmaceutical trials.

Table 2⇓ lists 26 commercially available oncology medications that are currently prescribed in pediatric oncology, the conditions for which they are prescribed, and whether a pediatric indication or pediatric dosing is in the product label. As an example, none of the three components of the “ICE” (ifosfamide, carboplatin, and etoposide) regimen, which is widely used for relapsed pediatric malignancies, has ever been approved for pediatric use. In addition, the combination of ifosfamide and etoposide has been incorporated into treatment for all patients in a recent cooperative group Ewing’s sarcoma study (Grier HE, personal communication, 1998). Similarly, cisplatin, originally approved in 1978 and at present prescribed in initial therapy for neuroblastoma,^3,4 hepatoblastoma,⁵ and germ cell tumors,⁶ lacks specific pediatric labeling.

The FDA has the responsibility of approving marketing claims about the use of a drug. To be considered for marketing approval, sponsors must submit to the FDA a New Drug Application (NDA). An NDA must contain detailed information regarding the product’s chemistry and manufacturing, preclinical toxicology, pharmacology, and clinical investigations. In addition to intensive review by FDA scientists, NDAs are often reviewed and discussed by the Oncologic Drug Advisory Committee (ODAC), an external panel comprising oncologists, a statistician, a patient representative, and a consumer representative. In an open public forum, the sponsor and the FDA present findings to the ODAC, which makes a recommendation to the FDA regarding the approvability of the marketing claim for a drug. On the basis of the internal recommendations of FDA review staff and the ODAC, the FDA subsequently determines the approvability of the marketing claim and recommendations for the drug’s use to be included in the package insert.

The approved claim is summarized in the product label (package insert) and constitutes the marketing license. All advertising and use claims are based on the label text. Once an NDA is approved, a pharmaceutical sponsor may subsequently submit a supplemental application (sNDA) to expand or modify the drug’s use, including additional indications, modifications of dosage forms or dosing schedules, and different target populations.

As summarized in Table 3⇓, the Center for Drug Evaluation and Research (CDER) of the FDA has reviewed six pediatric submissions involving five drugs during the period from 1979 to June 2002. Teniposide was approved for the therapy of relapsed acute lymphocytic leukemia (ALL). Daunorubicin (Cerubidine; Wyeth-Ayerst, Philadelphia, PA) was initially not approved for pediatric ALL and acute myelogenous leukemia (AML) but was subsequently approved for adult AML in 1979. An sNDA for its use in pediatric ALL was approved in 1983. An sNDA proposing the addition of osteosarcoma as a labeled indication for methotrexate was approved in 1988. Two asparaginase preparations, native Escherichia coli and polyethylene glycol–formulated E. coli, are considered biologic agents by the FDA and were reviewed and approved for use in pediatric ALL by the Center for Biologics Evaluation and Research (CBER) in 1978 and 1993, respectively.

During this 23-year period, the National Cancer Institute and pediatric cooperative groups were actively engaged in numerous clinical studies. A careful review of the pediatric oncology drug submissions to CDER over the past 20 years was conducted to examine the apparent discrepancy between approvals and research activity.

DISCUSSION

During the last 20 years, there have been six submissions to the Division of Oncology Drug Products of the CDER of the FDA for pediatric oncology indications. Three of these submissions were approved—two supplemental indications for drugs approved for adults and one new molecular entity, teniposide, for the reinduction of refractory ALL in combination with other active drugs. The last approval occurred in 1990.

To support approval for a pediatric claim, submissions to the FDA are considered favorably if they contain controlled studies of adequate size with disease-appropriate end points, a treatment regimen that can be applied to clinical practice, and prospective data collection with a prospective analysis plan. These characteristics should permit unbiased demonstration of a new drug’s contribution to efficacy.

In the applications that did not lead to approval there were several common features. These included the absence of well-controlled studies (vindesine, teniposide, 5-azacytidine), use of inappropriate controls (teniposide for neuroblastoma), patient heterogeneity (teniposide for neuroblastoma), short response duration (vindesine), and small number of patients (teniposide, 5-azacytidine). Additional reasons relating to nonapproval were the use of an unvalidated surrogate marker (teniposide for neuroblastoma), difficulty in applying the findings to clinical practice (vindesine), and the use of complex regimens that do not indicate the contribution of the new drug to efficacy (teniposide for induction of ALL).

The Code of Federal Regulations has a requirement for adequate and well-controlled studies. The submissions for methotrexate, 5-azacytidine, teniposide, and vindesine all contained uncontrolled or historically controlled studies. Historical controls may not reliably determine therapeutic efficacy. Changes in the standard of care or supportive measures may bias the results in favor of the most recent cohort if a substantial time interval exists between the treatment period for a study cohort and the historical controls. Furthermore, the inclusion of more than one treatment regimen in the historical control group can severely limit and complicate interpretation of the historical comparison, such as in the case of methotrexate for osteosarcoma.

The use of inadequately designed or underpowered studies can be complicated by poor protocol compliance and a high level of patient ineligibility that results in few evaluable patients, as in the cases of teniposide for neuroblastoma or 5-azacytidine for AML. Small numbers of patients or heterogeneity in the patient population do not allow statistical analysis of the results, as in the cases of teniposide for neuroblastoma and ALL, methotrexate for osteosarcoma, 5-azacytidine for acute leukemias, and vindesine for ALL.

The general principles of cancer drug registration are demonstration of clinical benefit in a defined population; demonstration of the effect of the test drug by replication of the findings, usually by demonstration of the clinical benefit in at least two clinical trials; and adequate safety data to assess risk and benefit in a defined population. The conventional approach to obtaining a marketing license for a new drug is to perform two or more studies designed to establish clinical benefit directly. End points that are commonly considered proof of clinical benefit are survival or the diminishing of disease-related symptoms.¹⁶

An alternative strategy is to seek accelerated approval as described under Subpart H of the Code of Federal Regulations, 21 Section 314.¹⁷ Under this provision, improvement of a surrogate marker likely to predict clinical benefit for a serious or life-threatening disease that has no satisfactory available therapy could lead to approval. Examples of diseases that represent unmet medical needs in pediatric oncology include glioblastoma multiforme, brainstem gliomas, infant leukemias, and relapsed metastatic sarcomas. The approval mechanism mandates subsequent studies to directly demonstrate clinical benefit.

The Subpart H approval mechanism allows a planned interim analysis of a randomized controlled study with a clinical benefit end point. The interim analysis could examine a surrogate end point such as response rate and could be used to support accelerated approval with the demonstration of clinical benefit to follow.

In addition, accelerated approval has been granted in single-arm trials examining response rate in unmet medical need conditions. Recent examples of accelerated approval include irinotecan (Camptosar; Pharmacia & Upjohn, Peapack, NJ) for relapsed or refractory colorectal cancer, temozolomide (Temodal; Schering-Plough, Kenilworth, NJ) for refractory anaplastic astrocytoma, and imatinib mesylate (Gleevec; Novartis, East Hanover, NJ) for chronic myelogenous leukemia in blast or accelerated phase.

The FDA continues to be concerned about the paucity of pediatric submissions for oncology. Since 1979, regulations have been issued to address the general issue of the lack of adequate labeling of drug products for children.¹⁸

In 1994, the FDA issued a regulation (the 1994 Final Pediatric Rule) that allowed efficacy data from adult studies to be extrapolated to pediatric populations if the disease or condition under study existed in both the pediatric and adult populations and if the therapeutic response was similar.¹⁹ The intention was to encourage sponsors to develop drugs in pediatrics and to provide pediatric information for drug labeling for both marketed drugs and new molecular entities under regulatory evaluation. Specifically, the provision permits the information from adequate and well-controlled adult studies to support a pediatric indication. Additional pediatric information, such as appropriate pediatric dosage (eg, pharmacokinetic data) and pediatric safety information must also be submitted.

The response to the initiative was less than anticipated, prompting both a mandate (the Final Pediatric Rule of 1998) and an incentive program under the Food and Drug Modernization Act of 1997. The application of the Pediatric Rule of 1998 is based on the principle that an adult indication under review is similar to a pediatric disease or condition.²⁰ The application of the 1998 Rule to pediatric oncology was a subject of ongoing discussions of the Pediatric Oncology Subcommittee of the ODAC.

The incentive program outlined in Section 111 of the FDA Modernization Act of 1997 (FDAMA) was intended to address the situation both when a pediatric disease is the same as in the adult indication and also when it is different, when extrapolation of data is not possible.²¹ The incentive is a 6-month extension to marketing exclusivity based on existing patent or other marketing exclusivity and applies to the entire product line, independent of indication or dosage form, of an active chemical moiety. It can be granted if a study report fairly responds to a FDA-generated Written Request. A Written Request may be initiated by a proposal submitted to the FDA from a drug sponsor or other party with the permission of the drug sponsor, or it may be initiated internally by the FDA. It may be granted regardless of study outcome (ie, a negative study may qualify, because the intent is to provide appropriate pediatric information). Drugs under development as well as currently approved drugs may receive a Written Request. The law did not permit the incentive to apply to biologicals, selected antibiotics, or devices.

The Division of Oncology Drug Products has interpreted the incentive program to allow a pediatric oncology development plan to qualify for determination for an exclusivity extension even when a specific pediatric indication is not targeted.²² If a sponsor submits data from an adequate phase I and specified phase II studies in response to a FDA Written Request, a determination for an exclusivity extension may be requested. In the unusual circumstance that a drug is prohibitively toxic for continuing development beyond phase I trials, a request to the FDA for a pediatric determination exclusivity may be made after the phase I studies. Completion of phase II studies, irrespective of a demonstration of efficacy, may lead to granting of an exclusivity extension. If the results of phase II studies are promising in a disease that is life threatening and for which no effective therapy exists, then a sponsor may apply for accelerated approval using a surrogate marker under Subpart H of the Code of Federal Regulations 21 Section 314.510. Drugs that exhibit evidence of activity would be expected to be developed in phase III studies, although this may not be a condition for granting pediatric exclusivity. A guidance document and sample written request template are posted on the FDA Web site at www.fda.gov/cder/cancer.²³ The incentive program associated with FDAMA expired on December 31, 2001. On January 4, 2002, the Best Pharmaceuticals for Children Act (BPCA) was signed into law.²⁴ The BPCA has a new incentive program that retains most of the features of the FDAMA initiative. The BPCA has several additional features, including a program to study off-patent drugs and an emphasis on dissemination of pediatric information through product labeling and other mechanisms. The BPCA incentive program expires in 2007.

As of December 2002, the FDA received 18 proposals for pediatric oncology studies and had, in addition, issued unsolicited written requests. Not all proposals for pediatric studies resulted in a written request. More than 25 written requests for oncology products were issued, of which 14 were for approved products and the remainder were for products under development. Four products have submitted the requested study reports, and pediatric exclusivity was granted. Overall, the FDA has issued more than 260 Written Requests, with more than 75 pediatric exclusivity determinations made and more than 60 granted.²⁵

The FDA has a desire to be flexible and practical in accepting data intended to support a marketing claim. It should be noted that studies performed outside the United States have been used to support domestic approval and may be acceptable for a pediatric study report. The FDA recommends discussion with a pediatric cooperative group to use their expertise and resources to assist with study design and patient accrual and further recommends meeting with the FDA to review the specifics of any proposal.

The intent of the FDA pediatric initiatives is to stimulate the development of new therapeutics for pediatric indications and to encourage the submission of quality clinical data. These initiatives are only part of a larger matrix to provide new therapies for pediatric oncology patients.²⁶ The FDA has established a Pediatric Subcommittee of the ODAC, incorporated additional pediatric oncologists in the review staff, and established a new Office with responsibility to coordinate and facilitate pediatric drug development. Additional initiatives that may facilitate the initiation of studies in pediatric patients with cancer include the following: the pharmaceutical industry consortium of pediatric oncologists; greater coordination and cooperation between the Children’s Oncology Group and the pharmaceutical industry; involvement of patients, patient families, and patient advocacy groups in all phases of development; and the development of predictive preclinical models.