Page 364 - Withrow and MacEwen's Small Animal Clinical Oncology, 6th Edition

342 PART III Therapeutic Modalities for the Cancer Patient

events from being DLTs, again if transient and clinically silent in the response rate is higher, a second stage of enrollment continues
nature and are prospectively defined in the study protocol. The to establish an approximate response rate. For example, at least
9 to 14 patients with the same histology or molecular target are
MTD is defined as the highest dose level in which no more than
VetBooks.ir one of six dogs develops a DLT. Traditionally, a fixed-dose modi- treated to test the null hypothesis of insufficient efficacy. If no
responses are observed in this group, the study ends. If a response
fied Fibonacci method of dose escalation is used wherein the dose
is escalated 100%, 67%, 50%, 40%, and then 33% of the previ- is noted in one of the cases, the accrual is increased to 31 patients
9
ous dose as the dosing cohorts increase. If the escalation incre- to establish a more accurate response rate. If a less robust response
ments are too conservative, more patients receive a suboptimal rate is expected (e.g., 20%), then the initial accrual number must
dose; conversely, if the escalations are too rapid, more patients be increased. Sample-size effect on study power calculations will
are at risk for significant toxicity and the accuracy of the MTD is be outlined in a subsequent section. Some have opined that the
compromised. Interdosing cohorts can be added during the study leading cause of drug failure in later phase development is our
period if more refined escalation or deescalation is found to be overdependence on these unpredictable single-arm, uncontrolled
necessary. phase II trials and that, as such, they should be avoided to ensure
Alternative, “accelerated titration” dose-escalation strategies phase III trial resources are not wasted because of the results of
27
have been suggested. 3–5,9,11,12 These include (1) two-stage designs poorly designed phase II trials. It used to be considered that
in which initially single-patient cohorts are used and the dose is the consequence of type I error (false positive) was less deleteri-
increased by a factor of two until a grade II toxicity occurs; then ous than that of type II error (false negative) because false-positive
the second stage involves more traditional three-patient cohorts trials are likely to be repeated, whereas false-negative trials would
29
and acceleration strategies; (2) within-patient escalation in which result in the abandonment of a potentially active treatment.
the same patient receives a higher dose on subsequent treatments The goals of comparative oncology modeling can help minimize
until a DLT is observed; however, this may mask cumulative type II error by mechanistically defining activity of novel agents
toxicity; (3) escalations based on PK parameters, e.g., to achieve through more detailed PK–PD studies. In today’s environment,
a target level of drug exposure; (4) escalations based on target however, with an abundance of novel drugs to be evaluated, false-
modulation (if known); and (5) continual reassessment strategies positive results are just as serious because they tie up patient and
using Bayesian methods. 11–14 In the end, it is always a trade-off of financial resources. With this in mind, the ideal phase II design
risk versus benefit; however, accelerated titration designs are gen- would be randomized across several potential investigational
erally associated with both a reduction in total patient number drugs, blinded, and controlled; modifications of this type applied
and a reduction in the number of patients receiving a suboptimal to standard phase II design are discussed subsequently (controlled
dose. phase II or phase IIB trials). 30
Although the phase I MTD approach works well for traditional
cytotoxic drugs, phase I trials designed to determine the BED may Endpoints of Activity/Efficacy
be more relevant for MTAs. 15–19 Trials evaluating the BED require As the primary goal of phase II trials is assessment of activity/effi-
validated assays that measure an effect on the target in serial tumor cacy, endpoints used to evaluate response are critical to the design.
samples and/or a surrogate tissue or fluid that documents activ- Median survival time is the ultimate “gold standard” primary
ity at the molecular level. Examples of early incorporation of PD endpoint that may be objectively measured to determine whether
markers in tumor and surrogate tissue (peripheral blood) include a drug is beneficial; however, this endpoint requires too great a
canine phase I trials of the KIT kinase inhibitor toceranib, the length of time for early phase trials to be meaningful and can be
20
Btk inhibitor ibrutinib, the histone deacetylase inhibitor sodium affected in veterinary patients by decisions to euthanize or seek
21
22
valproate, and the putative autophagy modulator hydroxychlo- alternative or rescue therapies. Therefore most early phase trials
23
roquine. Such PD modulation studies are increasingly impor- utilize interim or secondary endpoints that are reasonably likely
tant endpoints of phase I and II designs and are now commonly to predict clinical benefit. These surrogate endpoints may allow
required as proof of mechanism for drug approval. for shorter, less expensive, and technically simpler trials, which
allow for early termination if an interim outcome reaches a pre-
Phase II Trials (Activity/Efficacy Trials) specified positive or negative level. It is important that the validity
of the surrogate outcome actually correlates with lower mortality
Several good reviews have outlined phase II trial designs. 4,24–27 (or improved QOL)—this is often validated in larger later phase
The primary goal of phase II trials is, using the MTD or BED trials. With traditional cytotoxic chemotherapeutics, response is
established in phase I, to identify the clinical or biologic activity assessed based on criteria that describe changes in size or volume
in defined patient populations (e.g., tumors with a particular his- according to several published methodologies (e.g., Response
tology or particular molecular target) and inform the decision to Evaluation Criteria in Solid Tumors [RECIST], World Health
embark on a larger pivotal phase III trial. The traditional phase II Organization [WHO]). 31–33 Veterinary consensus modifications
34
design (phase IIA) is a single-arm, open-label activity assessment of RECIST have been published. It is readily evident that such
of a novel drug or therapeutic modality that lacks a control group criteria may not be appropriate for MTAs that are more likely to
or uses historical controls, which are prone to bias (selection, pop- be cytostatic rather than cytotoxic, thus resulting in disease stabi-
ulation drift, and stage migration bias). The number of patients to lization rather than measurable regression. 35,36 In such cases, it is
be enrolled is variable depending on the “minimal useful response critical that a designation of “stable disease” persist for a clinically
rate” and the rate of spontaneous regression (usually <5%). Mul- relevant period of time, generally not less than 6 to 8 weeks, to
tiple references discuss the statistical underpinnings of these power avoid reflection of slow progression rather than a true treatment
calculations, and there are several online resources for performing effect. Alternatively, temporal measures such as progression-free
them. Often, “two-stage min-max” designs are employed where survival (PFS) or time to progression (TTP) may be appropriate
enrollment is halted if the response rate falls below a predeter- endpoints; however, these often take too long to mature for timely
mined minimum level in an initial, small group of patients. If phase II trials. Alternatively, an adequate compromise could be
28
37