isaatmIci(

TtsticirtH

ii

etPtM3pe

td

CI

S

PHASE I RESEARCH AND THE MEANING OF DIRECT BENEFIT

LAINIE ROSS, MD, PHD

In this article, I examine whether Phase I pediatric oncology trials offer “the prospect of direct benefit,” a concept foundn Subpart D of the Code of Federal Regulations (CFR), the guidelines that provide additional protections to pediatric researchubjects. In research that offers the prospect of direct benefit, children can be exposed to greater risk than in other researchnd their dissent can be overridden. I argue that Phase I trials do not offer the prospect of direct benefit and classifying thems if they do fails to acknowledge the moral relevance of the researchers’ intent. In Subpart D, research that does not providehe prospect of direct benefit can be approved locally if it does not expose the children to more than a minor increase overinimal risk. If the risks are greater, the research must be approved nationally. To avoid the need for national review for Phaseoncology trials, I propose a new research category that incorporates the concept of “secondary direct benefit.” In this

ategory, the child’s dissent would be dispositive. This new category would improve the protections provided to children byncorporating intentions into Subpart D, the absence of which is a serious flaw in our current regulatory schema.J Pediatr 2006;149:S20-S4)

o review Phase I pediatric research under Subpart D of the Code of Federal Regulations (CFR) 45 part 46, it is necessaryto determine whether the research offers the prospect of direct benefit.1 If the research offers the prospect of directbenefit (also known as “therapeutic research”), then the research is permissible provided that: (1) the risk is justified by

he anticipated benefit to the subjects; and (2) the relation of the anticipated benefit to the risk is at least as favorable to theubjects as that presented by available alternatives (CFR §46.405). If the research does not offer the prospect of direct benefit,hen the research is permissible if: (1) the risks are no more than minimal (CFR §46.404); (2) the risks are no more than a minorncrease over minimal risk; and: (a) the research is likely to yield generalizable knowledge about the subject’s disorder orondition; and (b) the intervention or procedure presents experiences to subjects that are reasonably commensurate with thosenherent in their actual or expected medical, dental, psychological, social, or educational situations (CFR §46.406); or (3) theesearch is not otherwise approvable but presents an opportunity to understand, prevent, or alleviate a serious problem affectinghe health or welfare of children and it is approved by a committee convened by the Secretary of the Department of Health and

uman Services (DHHS) (CFR §46.407).In this article, I examine the issue of whether Phase I pediatric oncology trials offer the prospect of direct benefit, the

mplications of classifying the research as offering or not offering direct benefit, why the current classification schema isnadequate, and an alternative approach to the issue.

PHASE I ONCOLOGY STUDIESFollowing extensive preclinical testing, new drugs must be tested on human beings. Phase I studies are the first time a new

ntity is introduced into a human subject. Most Phase I studies are done on healthy adult volunteers, except in oncology, wherehe toxicities of new entities are potentially significant and Phase I testing is conducted in patients with refractory cancer.2 Inhase I oncology testing, the drugs often have a narrow therapeutic window, meaning that

here is a direct relationship between dose and toxicity, and between dose and efficacy.3

ost Phase I oncology studies are cohort studies, in which patient-subjects (usually n �) are treated at increasing doses to determine the maximal tolerated dose (MTD), theharmacokinetics, and drug toxicities to allow researchers to plan for Phase II studies offficacy.

In traditional adult oncology Phase I trials, there is an attempt to minimize theoxicity of the new agent by beginning at a very low dose. Given the relationship betweenose and efficacy, this means that those who are enrolled earliest in the trial are unlikely

FR Code of Federal Regulations MTD Maximal Tolerated Dose

From the Department of Pediatrics and theMacLean Center for Clinical Medical Ethics,University of Chicago, Illinois.

Reprint requests: Dr Lainie Friedman Ross,University of Chicago, Department of Pe-diatrics, 5841 S. Maryland Ave., MC 6082,Chicago, IL 60637. E-mail: lross@uchicago.edu.

0022-3476/$ - see front matter

Copyright © 2006 Elsevier Inc. All rightsreserved.

RB Institutional review board

20

10.1016/j.jpeds.2006.04.046

tarudaooea

paohrrp1doctmD0aecq

ntbldbbaictitpobbves

sAdotctCsopIpcfm

tPcenebbs

bspsspddsnr

ptetdsnploOotmbm

P

o get any therapeutic effect. Because of this, some havergued that the major ethical problem with Phase I canceresearch is not the risks of toxicity, but the problem ofnderdosing.4 As the trial advances, dose-limiting toxicitiesevelop, which define the end point of the trial. Meta-nalyses of adult Phase I trials of anticancer drugs show anverall therapeutic response rate of about 5%.5-7 The majorityf these are partial responses with �1% of subjects experi-ncing a complete response. Death from toxic effects is rare,t 0.5%.6

Because of the small number of pediatric oncologyatients and the practice that drugs are tested first on animalsnd adults before children,8 pediatric oncology Phase I trialsften begin at 80% of the adult MTD.9 Starting at such aigh dose means that even the earliest enrolled children mayeceive a dose that has therapeutic benefit. An objectiveesponse in pediatric Phase I oncology trials is defined as aartial or complete tumor remission, and it occurs in 5% to0% of all subjects.10-12 The definition of objective responseoes not include children whose cancer remains stable whilen Phase I experimental protocols nor improvements in painontrol and quality of life. These responses are meaningful tohe children and their families, suggesting that the studiesay offer even greater potential for therapeutic benefit.10,13

eath from toxic effects is rare in pediatric Phase I trials at.5%,6,12 although one pediatric meta-analysis found it to bes high as 2.5%.11 As many as one fifth of the childrenxperience dose-limiting toxicity.12 Other side-effects areommon, but their overall frequency, severity, and impact onuality of life have been poorly documented.13

DO PHASE I STUDIES OFFER THEPROSPECT OF

DIRECT BENEFIT?The federal regulations do not define “direct benefits”

or explain how they differ from indirect benefits or otherypes of benefit. Nancy King proposes that we distinguishetween three types of research benefits: direct benefit, col-ateral benefit, and aspirational benefit.14 Direct benefit isefined as “benefit arising from receiving the interventioneing studied,” that is, a therapeutic benefit.14 Collateralenefit (or indirect benefit) is “benefit that arises from beingsubject, even if one does not receive the experimental

ntervention” (eg, a free medical exam).14 Collateral benefitsan be medical or psychological, and include what is known ashe “inclusion benefit,” the benefit gained from participationtself.15 Aspirational benefit is benefit to society. King noteshat these benefits are often combined and confused.14 Sub-art D focuses on direct or therapeutic benefit, but researchersften combine direct and collateral benefits, and discuss theenefit from participating in this research, which includesoth the possible therapeutic benefit of receiving the inter-ention as well as other therapeutic and nontherapeutic ben-fits, including access to skilled clinicians, improvement inymptom management, and psychological benefits.14

The medical literature is divided as to whether Phase I f

hase I Research And The Meaning Of Direct Benefit

tudies in adults offer a direct or therapeutic benefit. Themerican Society of Clinical Oncology clearly says that theyo: “[p]hase I cancer trials can represent a real therapeuticption for some patients who have failed to respond to otherreatment or for whom no other therapies exist.”2 at 853 Inontrast, ethicists argue that they are nontherapeutic (that is,hey do not offer the prospect of direct benefit). Moreno,aplan, and Wolpe argue that “[a]ll phase I consent forms

hould include the phrase, prominently displayed in bold typen the first page, ‘This medical research project is not ex-ected to benefit you’.”16 at 1954 Brody also argues that Phasetrials, as traditionally designed, must be treated as nonthera-eutic research.17]He argues for newer designs, such as theontinual reassessment method, where subjects are treatedrom the beginning at the dosage anticipated to be recom-ended for study in a Phase II trial.17

Most consent forms state that Phase I studies are non-herapeutic.18 In an examination of the “Benefit Section” ofhase I consent forms, Horng et al found that only 1 in 272onsent forms stated that the subjects were expected to ben-fit, 11 forms (4%) stated with certainty that subjects wouldot benefit, 255 (94%) communicated uncertainty about ben-fit, and 5 forms (2%) said nothing about the chance ofenefit. Interestingly, 139 forms (51%) alluded to the possi-ility of benefit in a section other than the designated benefitection.18

Whether Phase I studies offer the prospect of directenefit is not critical in adult oncology because, underubpart A of the federal regulations, research is permissiblerovided that the risks are minimized, and they are rea-onable in relation to the anticipated benefits.19 Phase Itudies are designed to meet both of these criteria. Inediatric oncology, however, whether the research does oroes not offer the prospect of direct benefit is a criticalistinction that influences the level of risk to which theubjects can be exposed, the consent mechanisms that areecessary, and whether the study requires local or nationaleview for approval.

There is division within the medical and bioethicalediatric communities as to whether Phase I studies providehe prospect of direct benefit. Kodish, a pediatric oncologist-thicist, supports viewing Phase I oncology trials as offeringhe prospect of direct benefit, arguing for a “relativistic un-erstanding of prospect of benefit,”3 The Phase I trials offer amall chance of controlling the progression of the disease, ifot reducing the cancer burden, compared with nonpartici-ation and the receipt of palliative care.3 Ackerman, a phi-

osopher-medical ethicist, also argues that Phase I pediatricncology trials should be treated as therapeutic research.20

ne concern with stating that Phase I trials offer the prospectf direct benefit is the possibility that it will promote theherapeutic misconception.21 Many studies have shown thatost subjects enroll in Phase I trials hoping for personal

enefit, despite knowing that the study is designed to deter-ine toxicity and despite what is written on the consent

orm.13,22,23 In pediatrics, this may result in parental delay in

S21

ahtmUesa

oairfn(twf

Ap1odiehaoleot“toNmwi

accatabsbp

dtf

pctia(simthteddtIfImtc

pfste

pbC§

rdep

ecdirbr(

S

ccepting their child’s terminal prognosis, and may make itarder for parents to acknowledge their child’s suffering ifhey decide to pursue cure at all cost,24,25 even though thisay not be in either the child’s24-26 or the family’s interest.27

sing this line of reasoning, Oberman, a lawyer-medicalthicist, and Frader, a pediatric hospice specialist-ethicist,tate that Phase I tests are explicitly nontherapeutic in naturend that it is an error to think they are therapeutic.28

IF PHASE I STUDIES ARE CLASSIFIED ASOFFERING THE PROSPECT OF

DIRECT BENEFITIf Phase I studies are classified as offering the prospect

f direct benefit, then a research protocol can be approved byn institutional review board (IRB) provided that: (1) the risks justified by the anticipated benefit to the subjects; (2) theelation of the anticipated benefit to the risk is at least asavorable to the subjects as that presented by available alter-ative approaches; and (3) consent requirements are metCFR §46.405). Even in studies that are considered to offerhe prospect of direct benefit, the IRB must still examinehether the risk:benefit ratio is favorable and whether it is as

avorable as the available alternatives.Is the risk:benefit ratio favorable to the subjects?

lthough the potential for direct benefit for complete orartial remission is about 10%, possibly an additional 5% to0% of children experience stability of disease for a periodf time. There are also potential collateral benefits. Manyiscuss the positive side benefits of Phase I participation,

ncluding the inclusion benefit, the reduction of symptoms,nhanced attention in the hospital, and the maintenance ofope.3,20 Agrawal and Emanuel examined the literaturend found data to document that participating in Phase Incology studies may actually improve patients’ quality ofife13 or provide some psychological benefit.22,23,29 How-ver, Cox points out that quality-of-life questionnairesften used in oncology interviews may not adequately cap-ure the burdens that trial participation entails, includingan investment of time, emotional and physical energy . . .hat meant other aspects of their lives were effectively putn hold for the duration of their trial involvement.”30

evertheless, even Cox found that most subjects “wouldake the same decision to participate in a similar trial if itas offered to them.”30 This implies that the subjects

nterpret the risk:benefit ratio favorably.Is the risk:benefit ratio to the subjects as favorable as the

lternatives? The only alternative for most children with can-er that is not responsive to curative modalities is palliativeare. Palliative care ensures comfort measures are managedppropriately, but the treatment offers no chance of cure, suchhat the risk:benefit ratio of Phase I trials will often be vieweds “relatively” favorable.3 These modalities, however, need note viewed as diametrically opposed. Providing palliative careervices to children enrolled in Phase I trials would offer theest of both worlds for the child and future children.24 The

alliative services can minimize side-effects of the Phase I t

22 Ross

rug, and thus improve the risk:benefit ratio while enablinghe child to participate in the advancement of medical scienceor future patients.

Research approvable under CFR §46.405 requires one-arent permission and the child’s assent. The child’s dissentan be overridden if “the intervention or procedure involved inhe research holds out a prospect of direct benefit that ismportant to the health or well-being of the child and isvailable only in the context of the research” (CFR §46.408a)). A child’s dissent can be overridden in therapeutic re-earch to promote the child’s medical best interest because its assumed that the parents know what is in their child’s

edical interest and will act to promote this. To allow parentso compel their children to participate in Phase I trials,owever, overstates the potential medical benefit of theserials. Research that may provide unintended (collateral) ben-fits should not qualify as research that offers the prospect ofirect benefit. To classify research as offering the prospect ofirect benefit suggests a certain probability of success, and nothe mere possibility of benefit. Furthermore, to classify Phaseresearch as offering the prospect of direct benefit ignores the

act that the research is not designed to provide direct benefit.gnoring the intent of the research promotes the therapeuticisconception, and thereby inappropriately empowers parents

o compel their children to participate in these trials over theirhildren’s dissent.

Although Phase I pediatric oncology trials may offer aossible direct benefit and may be the best medical alternativeor particular children with cancer, I reject classifying suchtudies under CFR §46.405 because doing so ignores the facthat the researchers’ intent is focused on showing safety notfficacy.

IF PHASE I TRIALS ARE CLASSIFIED AS NOTOFFERING THE PROSPECT OF

DIRECT BENEFITIf Phase I studies are classified as not offering the

rospect of direct benefit, then the research protocol may stille approved under categories CFR §46.404, CFR §46.406, orFR §46.407. In contrast with research approved under CFR46.405, all these categories focus on aspirational benefits.

CFR §46.404 are the regulations that address minimalisk research. By definition, a Phase I trial of an experimentalrug entails more than minimal risk because the new drugntails potentially unknown but significant risks, regardless ofreclinical data.

To qualify under CFR §46.406, the research must: (1)ntail no more than a minor increase over minimal risk; (2) beommensurate with the lived experiences of those with aisorder or condition; and (3) be likely to yield generalizablenformation. Phase I trials are commensurate with the expe-ience of the children to whom it is offered, as many will haveeen through at least one course of chemotherapy (#2). Theequirement that it is likely to yield generalizable knowledge#3) places the onus on the oncology community to design

rials that will provide scientifically useful data. The advances

The Journal of Pediatrics • July 2006

odcmpatarhra

§§wvwbHan

anwacraoilettsphim

pmHrtuprfSsTpi

ptBsatia

tPp4srdsadiewttdotctit

boaepetrwpohP

esrapb

P

f the past two decades in pediatric oncology suggest thatrug testing in pediatric oncology is well-designed.31 Theentral issue for approval, then, is whether the risks are atost a minor increase over minimal risk (#1). Recall that the

rimary end point of these trials is to define MTD. Thesegents are often known to entail serious and anticipatedoxicities, and the trials are an attempt to quantify the risksnd to uncover additional unanticipated side-effects. Death isare, at 0.5%, although one meta-analysis found it to be asigh as 2.5%.6,12 As such, it is probably correct to say that theesearch entails more than a minor increase over minimal risknd the research cannot be classified under CFR § 46.406.

Even though the research is not approvable under CFR46.404 or CFR §46.406, it could be approved under CFR46.407 if adequate preclinical testing is done, the studies areell-designed and offer the opportunity to understand, pre-ent, or alleviate a serious problem affecting the health orelfare of children, and the research is reviewed and approvedy a panel convened by the secretary of the Department ofealth and Human Services. This would mean, however, that

ll Phase I pediatric oncology trials would have to undergoational review in addition to local IRB review.

Although many Phase I oncology trials would easily bepproved by a 407 panel, the question is whether this isecessary or desirable. Those who advocate for greater accessill be frustrated by this additional obstacle. Even those who

re wary of the movement toward increased participation ofhildren in research may not support the need for 407 paneleview because it is not clear what additional protections, ifny, the 407 panels would offer. Phase I oncology drugs arenly studied in children after they pass Phase I and II testingn adults; that is, after safety and efficacy have been estab-ished in adults. This does not mean the drug is safe orffective in children; it only confirms that all pre-pediatricesting has been done successfully; that is, that Phase I and IIesting in animals and adults has been completed and there isome reason to believe that the drug will be effective inediatric cancer. Because safety and efficacy of the new drugsave not been established in children, Phase I and II testing

n children is necessary, and the risks must be classified asore than a minor increase over minimal risk for children.

If the 407 process would not provide any additionalrotection, then the logistical costs in terms of time and effortake 407 review ethically and pragmatically undesirable.owever, to approve the research under the current federal

egulations, without resorting to a 407 panel, would meanhat IRBs have to approve Phase I pediatric oncology trialsnder CFR §46.405 by claiming that there is sufficient pros-ect of direct subject benefit. This is problematic for threeeasons. First, and most importantly, such a claim ignores theact that the research is not intended to provide direct benefit.econd, classifying this research under CFR §46.405 over-tates the potential medical benefit of most of these trials.hird, consent for research classified as CFR §46.405 allowsarental authorization over the children’s dissent as explained

n CFR §46.408(a). b

hase I Research And The Meaning Of Direct Benefit

AN ADDITIONAL CATEGORY FOR REVIEWIN WHICH INTENT MATTERS

Phase I research is done to determine toxicity; theossibility of direct benefit is secondary to the objectives ofhe study. In an analysis of risk and benefit for the Nationalioethics Advisory Commission, Weijer points out that re-

earch administered with a therapeutic intent differs in mor-lly relevant ways from research without a therapeutic in-ent.32 Focusing exclusively on whether Phase I trials could benterpreted as offering the prospect of direct benefit fails tocknowledge the moral relevance of the researchers’ intent.

If intentions matter, then Phase I studies cannot be saido offer the prospect of direct benefit. This would mean thathase I studies could only be permissible if approved by a 407anel review. But as discussed above, it is not clear that the07 review process provides any additional protections in thiscenario. Rather, a new category is needed in subpart D foresearch that offers the potential for what I call “secondaryirect benefit” that would not require national review. Aecondary direct benefit is a benefit that arises from receivingn experimental intervention when the research is not beingone to achieve individual benefit. That is, the experimentalntervention offers a potential therapeutic or direct benefitven though the trial lacks therapeutic intent. This contrastsith research in which the experimental intervention offers

he prospect of direct benefit and whose study design isherapeutic (eg, Phase III trials comparing an experimentalrug against standard therapy). Secondary direct benefits arebtained from receiving the experimental intervention, evenhough the study was not designed to promote this benefit, inontrast to the indirect benefits that may accrue from being inhe experiment, regardless of whether one receives the exper-mental intervention. Secondary direct benefits are therapeu-ic; indirect benefits may or may not be.

Research that has the potential for secondary directenefits, like Phase I pediatric oncology trials, should have itswn classification. Research that is likely to yield generaliz-ble knowledge about the subject’s disorder or conditionntails more than a minor increase over minimal risk, androvides the potential for secondary direct benefit. This cat-gory requires that: (1) the risks are justified by the likelihoodhat the research will yield generalizable knowledge; (2) theesearch is commensurate with the lived experience of thoseith the disease or condition; (3) the research offers theotential for a secondary direct benefit, a benefit that is nottherwise available; and (4) consent requirements are met. Iave already shown that the first three conditions are met inhase I pediatric oncology trials.

The consent mechanisms for research in this new cat-gory should be similar to the mechanisms required for re-earch classified as CFR §46.406. Currently, nontherapeuticesearch beyond minimal risk requires two-parent consentnd the child’s assent, and the child’s dissent is always dis-ositive (CFR §46.408(a)). In contrast, the child’s dissent cane overridden in research that offers the prospect of direct

enefit. For this new category, I believe that the child’s assent

S23

mtmcct

eowPbdbnv“cas

1DM2t3c24ia5TT6I7r18aG9C11P11c1

tKG1i1t111s1t1Dt1Ahr212hH2RI2E2p12C2R2TN2c2t3a23it23sNiA

S

ust be dispositive. Although parents ethically may be au-horized to override a child’s objection for research involvingore than minimal risk that is intended to promote their

hild’s medical best interest, they should not be authorized toompel their child to participate in Phase I research becausehe research is not intended to provide direct benefit.

Creating this new category may be overstating a differ-nce and opening a loophole to allow research that does notffer the prospect of direct benefit to be approved locallyhen it should be reviewed by 407 panels. However, mosthase I pediatric oncology trials are presently being approvedy local IRBs on the grounds that they offer the prospect ofirect benefit (CFR §46.405),3 despite the low likelihood ofenefit and the lack of therapeutic intent. I believe that theew category improves the human subjects protections pro-ided to children, while ensuring that children are not left astherapeutic orphans.” It also incorporates intentions into thelassification of pediatric research subject protections, thebsence of which is a serious flaw in our current regulatorychema.

REFERENCES. US Department of Health and Human Services. (45 CFR 46, Subpart). Protections for children involved as subjects in research. Federal Register.arch 8, 1983;48:9814–20; revised Federal Register. June 18, 1991;56:28032

. American Society of Clinical Oncology. Critical role of Phase I clinicalrials in cancer treatment. J Clin Oncol 1997;15:853–9.. Kodish E. Pediatric ethics and early-phase childhood cancer research:onflicted goals and the prospect of benefit. Accountability Res003;10:17-25.. Ratain MJ, Mick R, Schilsky RL, Siegler M. Statistical and ethicalssues in the design and conduct of Phase I and Phase II clinical trials of newnticancer agents. JNCI 1993;85:1637-43.. Estey E, Hoth D, Simon R, Marsoni S, Leyland-Jones B, Wittes R.herapeutic response in Phase I trials of antineoplastic agents. Cancerreatment Reports 1986;70:1105-15.

. Decoster G, Stein G, Holdner E. Responses and toxic deaths in Phaseclinical trials. Ann Oncol 1990;1:175-81.

. Von Hoff DD, Turner J. Response rates, duration of response, and doseesponse effects in Phase I studies of antineoplastics. Invest New Drugs991;9:115-22.. National Commission for the Protection of Human Subjects. Reportnd Recommendations: Research Involving Children. Washington, DC: USovernment Printing Office; 1977. DHEW Publication No. (OS) 77-0004.

. Smith M, Bernstein M, Bleyer W, Borsi J, Ho P, Lewis IJ, et al.onduct of Phase I trials in children with cancer. J Clin Oncol998;16:966-78.0. Shah S, Weitman S, Langevin AM, Bernstein M, Furman W, Pratt C.hase I therapy trials in children with cancer. J Pediatr Hematol/Oncol998;20:431-8.1. Furman WL, Pratt CB, Rivera GK. Mortality in pediatric Phase Ilinical trials. JNCI 1989;81:1193-4.

2. Estlin EJ, Cotterill S, Pratt CB, Pearson ADJ, Bernstein M. Phase I n

24 Ross

rials in pediatric oncology: perceptions of pediatricians from the Unitedingdom Children’s Cancer Study Group and the Pediatric Oncologyroup. J Clin Oncol 2000;18:1900-5.

3. Agrawal M, Emanuel EJ. Ethics of Phase 1 oncology studies: reexam-ning the arguments and data. JAMA 2003;290:1075-82.4. King NMP. Defining and describing benefit appropriately in clinicalrials. J Law Med Ethics 2000;28:332-343.5. Lantos JD. The “Inclusion Benefit” in clinical trials. J Pediatr999;134:130-1.6. Moreno J, Caplan AL, Wolpe PR. Updating protections for humanubjects involved in Research. JAMA 1998;280:1951-8.7. Brody BA. The ethics of biomedical research: an international perspec-ive. New York: Oxford University Press; 1998.8. Horng S, Emanuel E, Wilfond B, Rackoff J, Martz K, Grady C.escriptions of benefits and risks in consent forms for Phase 1 oncology

rials. N Engl J Med 2002;347:2134-40.9. US Department of Health and Human Services. (45 CFR 46, Subpart). Final regulations amending basic HHS policy for the protection ofuman research subjects. Federal Register. January 26, 1981;46:8366–91;evised Federal Register. June 18, 1991;56:28003-18.0. Ackerman TF. Phase I pediatric oncology trials. J Pediatr Oncol Nurs995;12:143-5.1. Appelbaum PS, Roth LH, Lidz CW, Benson P, Winslade W. Falseopes and best data: consent to research and the therapeutic misconception.astings Center Report 1987;17:20-4.

2. Daugherty C, Ratain MJ, Grochowski E, Stocking C, Kodish E, Mick, et al. Perceptions of cancer patients and their physicians involved in Phasetrials. J Clin Oncol 1995;13:1062-72.

3. Hutchison C. Phase I trials in cancer patients: participants’ perceptions.uropean J Cancer Care 1998;7:15-22.4. Ulrich C, Grady C, Wendler D. Palliative care: a supportive adjunct toediatric Phase I clinical trials for anticancer agents. Pediatrics 2004;14:852-5.5. Levetown M. Ethical aspects of pediatric palliative care. J Palliativeare 1996;12:35-9.

6. Ackerman TF. The ethics of Phase I pediatric oncology trials. IRB: Aeview of Human Subjects Research 1995;17:1-5.7. Kreicbergs U, Valdimarsdottir U, Onelov E, Henter J-I, Steineck G.alking about death with children who have severe malignant disease.Engl J Med 2004;351:1175-86.

8. Oberman M, Frader J. Dying children and medical research: access tolinical trials as benefit and burden. Am J Law Med 2003;29:301-17.9. Moore S. A need to try everything: patient participation in Phase 1rials. J Advanced Nurs 2001;33:738-47.0. Cox K. Assessing the quality of life of patients in Phase I and IInti-cancer drug trials: interviews versus questionnaires. Social Science Med003;56:921-34.1. Bernstein ML, Reaman GH, Hirschfeld S. Developmental therapeuticsn childhood cancer: a perspective from the Children’s Oncology Group andhe US Food and Drug Administration. Hematol/Oncol Clin North Am001;15:631-55.2. Weijer C. The ethical analysis of risks and potential benefits in humanubjects research: history, theory, and implications for US regulation. In:ational Bioethics Advisory Commission (NBAC). Ethical and Policy Issues

n Research Involving Human Participants. Vol. 2. Commissioned Papers.ugust 2001. Available online at: http://www.georgetown.edu/research/

rcbl/nbac/human/overvol2.pdf, P-1-P-29.

The Journal of Pediatrics • July 2006