First-in-Human Phase I Dose Escalation Study of a Second ... · First-in-Human Phase I Dose...

Cancer Therapy: Clinical

First-in-Human Phase I Dose Escalation Study of aSecond-Generation Non-Ansamycin HSP90Inhibitor, AT13387, in Patients with AdvancedSolid TumorsGeoffrey I. Shapiro1,2, Eunice Kwak2,3, Bruce J. Dezube2,4, Murray Yule5, John Ayrton5,John Lyons5, and Daruka Mahadevan6

Abstract

Purpose: AT13387 is a potent second-generation, fragment-derived HSP90 inhibitor. This phase I study investigated themaximum tolerated dose (MTD)/recommended phase II dose(RP2D) and safety, pharmacokinetic, and pharmacodynamicprofiles of two AT13387 regimens in a refractory solid tumorpopulation.

Experimental Design: Standard 3þ3 dose escalation was used.MTD and RP2D determinations were based on the occurrence ofdose-limiting toxicities (DLT) and overall toxicity, respectively.Pharmacokinetic parameters were measured after single andmultiple doses. AT13387-mediated induction of HSP70 wasevaluated in plasma, peripheral blood mononuclear cells, andpaired tumor biopsies.

Results: Sixty-two patients were treated with doses rangingfrom 10 to 120 mg/m2 twice weekly and 150 to 310 mg/m2 onceweekly (both for 3 weeks every 28 days). One DLT of visual

disturbance occurred at 120 mg/m2, which was considered theMTD and RP2D for the twice-weekly regimen. No formal DLTsoccurred in the once-weekly regimen, but multiple moderatelysevere toxicities, including diarrhea, nausea, vomiting, fatigue,and systemic infusion reactions, led to selection of 260 mg/m2 asthe RP2D. Exposures of AT13387 increased proportionally withdose. Target engagement as measured by HSP70 inductionoccurred in plasma and tumor biopsy samples. One patient withgastrointestinal stromal tumor (GIST) who had progressive dis-ease on imatinib had a partial response and remained on treat-ment for 10 months. Twenty-one patients (34%) had stabledisease, which lasted >120 days in 7 patients.

Conclusion:AT13387 administered once or twice weekly hasan acceptable safety profile and demonstrated evidence oftarget engagement and preliminary antitumor activity. ClinCancer Res; 21(1); 87–97. �2014 AACR.

IntroductionHSP90 is an abundant and ubiquitously expressed molecular

chaperone that is crucial for the stability and function of manyoncogenic drivers, including receptor tyrosine kinases and hor-

mone receptors, such as KIT protein tyrosine kinase (c-KIT), EGFR,HER2, and anaplastic lymphoma kinase (ALK), androgen recep-tors, as well as proteins in key signaling pathways, including AKT/protein kinase B (PKB; refs. 1–3). HSP90-dependent clients aretherefore involved in multiple neoplastic processes, includinggrowth factor independence, invasion, and metastasis, sustainedangiogenesis, cell survival, and resistance to antigrowth signals. Inthe absence ofHSP90, these clients are ubiquitinated and targetedfor degradation via the proteasome (4), so that inhibition ofHSP90 is an attractive anticancer therapeutic strategy (5). First-generation geldanamycin-based HSP90 inhibitors such as 17-AAG and IPI-504 have shown evidence of clinical activity, buthave been limited by formulation and toxicity issues (6). Hence,there is a need for improved HSP90 inhibitors, which has led tothe development of second-generation compounds that are cur-rently in clinical development (7–10).

Using a combination of NMR and high-throughput X-raycrystallography, we discovered a number of low-affinity fragmenthits against the N-terminal ATPase domain of the HSP90 protein.Structure-aided drug design allowed for the rapid optimization ofa key fragment hit into a lead compound with inhibitory activityin the low nanomolar range. Subsequent improvement of thephysicochemical properties of the hit series allowed for theidentification of a clinical candidate, AT13387, which was for-mulated for intravenous dosing as a potential cancer therapeutic(11).

1Department of Medical Oncology, Dana-Farber Cancer Institute andDepartment of Medicine, Brigham and Women's Hospital, Boston,Massachusetts. 2Department of Medicine, Harvard Medical School,Boston,Massachusetts. 3MassachusettsGeneralHospital Cancer Cen-ter, Boston, Massachusetts. 4Department of Medicine, Beth IsraelDeaconess Medical School, Boston, Massachusetts. 5Astex Pharma-ceuticals, Inc., Dublin, California. 6Arizona Cancer Center, Tuscon,Arizona.

Note: Supplementary data for this article are available at Clinical CancerResearch Online (http://clincancerres.aacrjournals.org/).

Current address for D. Mahadevan: The West Clinic, Memphis, Tennessee.

Presented in part at American Society of Clinical Oncology Meetings, June 2010and June 2012, Chicago, IL.

Corresponding Authors: Geoffrey I. Shapiro, Early Drug Development Center,Department of Medical Oncology, Dana-Farber Cancer Institute, 450 BrooklineAvenue, Boston, MA 02215. Phone: 617-632-4942; Fax: 617-632-1977; E-mail:[email protected]; or Daruka Mahadevan, TheWest Clinic, PC,100 N. Humphreys Blvd, Memphis, TN 38120. Phone: 901-683-0055; Fax: 901-435-5595; E-mail: [email protected]

doi: 10.1158/1078-0432.CCR-14-0979

�2014 American Association for Cancer Research.

ClinicalCancerResearch

www.aacrjournals.org 87

on August 17, 2020. © 2015 American Association for Cancer Research. clincancerres.aacrjournals.org Downloaded from

Published OnlineFirst October 21, 2014; DOI: 10.1158/1078-0432.CCR-14-0979

http://clincancerres.aacrjournals.org/

As a novel HSP90 inhibitor, AT13387 has the potential forbroad utility in the treatment of advanced solid and hematologicmalignancies that are resistant or refractory to standard therapies.Prolonged drug accumulation in tumors despite rapid systemicclearance is well recognized for several HSP90 inhibitors (12, 13).AT13387 has a half-life of up to 65 hours in tumor xenograftsleading to prolonged knockdown of HSP90 client proteins inthese tumors, suggesting the potential for an extended pharma-codynamic effect following a single dose of the compound (11).In vivo studies have shown potent antitumor activity at doses thatare well tolerated in several different human tumor xenograftmodels inmice, including those derived fromnon–small cell lungcancers (NSCLC), melanomas, and gastrointestinal stromaltumors (GIST; refs. 14, 15). In NSCLC NCI-H1975 xenograftsharboring EGFR L858R/T790M, suppression of mutant EGFRexpression occurred up to 72 hours after a single dose, with tumorgrowth inhibition achieved with twice-weekly dosing (14).Higher doses of AT13387, administered once weekly, also pro-duced significant tumor growth inhibition (14). In addition,once-weekly dosing of AT13387 was effective against imatinib-sensitive and resistant GIST xenografts (15).

Toxicology studies of 1-cycle duration (28 days), with 1-hourintravenous infusion twice weekly for 3 weeks, showed a cleardose–effect relationship in dogs; histopathology revealed changesin thebonemarrow, thymus, and testes,with doses toleratedup to3 mg/kg (60 mg/m2 human equivalent). Overall effects weretransient and reversible, except for testicular lesions, which didnot reverse in 14 days of recovery. In the 1-cycle study in rats usingthe same regimen, doses up to 50 mg/kg (300 mg/m2) weretolerated with minimal changes in hematologic parameters andinfusion-site findings of inflammation. AT13387 is metabolizedmainly by conjugation. In vitro studies in human hepatocytesshowed formation of glucuronide- and sulphate conjugates as themain metabolites formed.

On the basis of preclinical efficacy and pharmacodynamicclient suppression in xenografts, as well as dosing regimens offirst-generation HSP90 inhibitors, this phase I, open-label, dose-

escalation study was designed to determine the maximum toler-ated doses (MTD) and recommended phase II dose (RP2D) ofAT13387 given as a 1-hour intravenous infusion either twiceweekly or once weekly for 3 of every 4 weeks in patients withmetastatic solid tumors that were refractory to standard therapy(16, 17). Secondary objectives included characterization of safetyand tolerability, as well as assessment of the pharmacokinetic andpharmacodynamic profiles of the drug.

Patients and MethodsPatient selection

Patients included in the study were �18 years of age, with anEastern Cooperative Oncology Group (ECOG) performance sta-tus 0–2, and histologic or cytologic evidence of a metastatic solidtumor, including lymphoma that was refractory to standardtherapy. Patients had to have adequate bone marrow function(hemoglobin > 9 g/dL, neutrophils > 1.4 � 109/L and platelets �100 � 109/L), hepatic function [serum bilirubin � 2.5 times theupper limit of reference range (ULRR); alanine aminotransferase(ALT), aspartate aminotransferase (AST), and alkaline phospha-tase < 2.5 times ULRR, unless due to the presence of livermetastases, when ALT and AST could have been up to 5 timesthe ULRR or bone metastases, when an isolated alkaline phos-phatase < 5 times ULRR was acceptable], renal function (serumcreatinine � 1.5 times ULRR, without evidence of more than 1þproteinuria on 2 occasions), and cardiac function (left ventricularejection fraction > 50% on echocardiogram or multigated acqui-sition scan and QTc � 460 msec according to Fridericia correc-tion). Main exclusion criteria were pregnant or lactating females,ongoing central nervous system metastases, treatment with anyanticancer therapies, or with known cytochrome P450 inducers,within 4weeks of the start of AT13387, history of ischemic cardiacevent, myocardial infarction or unstable cardiac disease within 3months of study entry, prior history of infection with HIV orknown activeHBVorHCV infection, and any evidence of severe oruncontrolled systemic conditions.

Study design and treatment administrationThis was an open-label, dose-escalation (3þ3 design) cohort

expansion study of single-agent AT13387. The primary objectiveof the study was to identify the MTD and RP2D for twice-weeklyand once-weekly administration regimens. The protocol definedthe MTD to be the dose one level below which a DLT occurred inmore than one patient (during the study, however, the MTD wasdetermined more conservatively, as described in the Results).Secondary objectives were to characterize the safety and tolera-bility of AT13387 and to identify DLTs; to define the pharmaco-kinetics of AT13387 in plasma and urine, and to demonstrate thepharmacodynamic activity of AT13387 in plasma, in peripheralblood mononuclear cells (PBMC), and in paired tumor biopsiesobtained in consenting patients treated at one of the RP2Ds.

The initial study design assessed a twice-weekly regimen (dos-ing days 1, 4, 8, 11, 15, and 18 of a 28-day cycle) and wassubsequently amended to evaluate a once-weekly regimen (dos-ing days 1, 8, and 15 of a 28-day cycle). Doses for the twice-weeklyregimen were 10, 20, 40, 80, and 120 mg/m2; cohorts of 3 to6 patients each were enrolled, with escalation determined by amodified Fibonacci sequence until the MTD was achieved. ASafety Monitoring Committee determined the dose escalationsaccording to the toxicity observed in the previous cohorts. On the

Translational Relevance

AT13387 is a second-generation, fragment-derived HSP90inhibitor that is active in multiple in vitro and in vivo tumormodels. This single-agent, phase I study identified the recom-mended phase II dose (RP2D) for two administration regi-mens, and characterized pharmacokinetics as well as pharma-codynamic biomarkers in patients with previously treatedrefractory solid tumors. A twice-weekly regimen was initiallystudied on the basis of the predicted half-life and duration ofclient suppression in preclinical models, followed by investi-gation of a once-weekly regimen, justified by the demonstra-tion of efficacy of the regimen and an extended intratumoralhalf-life in xenograft experiments. Target engagement, as mea-sured by HSP70 induction, was demonstrated in both plasmaand tumor biopsy samples at tolerable doses. Preclinical dataled to the enrichment for patientswith gastrointestinal stromaltumor (GIST) who had failed prior imatinib treatment. Nota-bly, one such patient with tumor harboring an acquiredmutation conferring resistance to imatinib achieved a partialresponse.

Shapiro et al.

Clin Cancer Res; 21(1) January 1, 2015 Clinical Cancer Research88




basis of the safety and tolerability data of the 120 mg/m2 twice-weekly cohort, a starting dose of 150 mg/m2 was chosen for theonce-weekly regimen with a maximal escalation of 25% until theMTD was reached. The doses administered on this regimen were150, 180, 220, 260, and 310 mg/m2. On both regimens, up to 12additional patients could be enrolled at the MTD/RP2D in anexpansion phase of the study to further evaluate safety, tolera-bility, and pharmacodynamics.

AT13387 was administered as a 1-hour intravenous infusion,based on body surface area in both regimens, via peripheral orcentral venous access using an appropriate calibrated infusionpump.

The study (ClinicalTrials.gov ID: NCT00878423) was con-ducted according to good clinical practice guidelines and theDeclaration of Helsinki. The Institutional Review Boards of the4 participating U.S. centers approved the study, and patients gavewritten informed consent before enrollment.

DLT definitions and safety assessmentsToxicity was graded according to Common Terminology Cri-

teria for Adverse Events (CTCAE) v3.0. A DLT was defined asneutropenia (neutrophil count <0.5 � 109/L) for >5 days; neu-tropenia (neutrophil counts <1 � 109/L) with fever; thrombocy-topenia (platelet count <25 � 109/L) accompanied by bleeding,or thrombocytopenia (platelet counts <10 � 109/L); any grade 3or 4nonhematologic toxicity thatwasnot a consequence of tumorprogression (other than nausea, vomiting, or diarrhea in theabsence of appropriate prophylaxis); and more than 1 individualdose omission during the first cycle of treatment due to theappearance of toxicity that was considered related to AT13387.

Each patient underwent a full medical history and physicalexamination at study screening. Safety parameterswere assessed atbaseline and at regular intervals during the study. Safety assess-ments included the reporting of all adverse events (AE), changes inclinical laboratory parameters (hematology, serum chemistries,coagulation, urinalysis), vital sign measurements, 12-lead elec-trocardiogram (ECG) results, and physical examination findings.In addition, serial measurements of left ventricular ejection frac-tion were performed, as cardiotoxicity was observed in firstgeneration HSP90 inhibitors (18). Ophthalmologic evaluation,including external examination, fundoscopy, visual acuity, intra-ocular pressure, assessment of visual fields, and measurement ofcolor vision, was tested at baseline and at alternate cycles startingon day 1 of cycle 3.Ophthalmologic assessments were introducedafter the studywas initiated, so that not all patients underwent theprocedures.

Response assessmentsTumor response was assessed by CT or MRI at screening and

every 2 months during the first 6 months, every 3 months for thenext 6 months, and every 6 months thereafter if patients contin-ued to receive treatment with AT13387. Tumor response wasevaluated according to Response Evaluation Criteria in SolidTumors (RECIST) version 1.0 (19).

PharmacokineticsBlood samples for pharmacokinetic analyses were collected on

cycle 1, day 1 and cycle 1, day 18 in the twice-weekly regimen orcycle 1, day 1 and cycle 2, day 15 in the once-weekly regimen[collection times in both regimens were immediately beforeinfusion start and at 0.5, 1 (before infusion end), 1.08 (5minutes

after infusion end), 1.5, 1.75, 2, 2.5, 3, 4, 7, 12, and 24 hours].Urine samples for pharmacokinetic analyses were collected at thefirst void on day 1 predose and at 0 to 7 hours, 7 to 24 hours and24 to 48 hours during and after the start of the first cycle oftreatment. AT13387 concentrations were measured usingAT13387-specific liquid chromatography with mass spectromet-ric detection. The lower limit of quantitation for AT13387 inplasma was 1.0 ng/mL and in urine was 10.0 ng/mL. The linearrange for the validated method was 1.00 to 1,000 ng/mL, withacceptance criteria for precision and accuracy of �15.0%.

Pharmacokinetic parameters of AT13387 were estimated inplasma and urine using standard noncompartmental analysesusing WinNonlin version 5.3 (Pharsight Corp.). Calculated phar-macokinetic parameters included area under the concentration–time curve from 0 to 24 hours (AUC0–24), maximum observedconcentration (Cmax), elimination half-life (t1/2), total bodyclearance (Cl), and volume of distribution (Vz). Urine concen-tration values of AT13387 were used to calculate the followingparameters: cumulative urinary excretion from time zero to time t(Ae0–t), maximum rate of urinary excretion (Rmax), time of Rmax

(Tmax), and percent recovered. Rmaxwas calculated by dividing theamount of drug excreted in each collection interval by the timeover which it was collected.

PharmacodynamicsPharmacodynamic assessments of biologic activity of AT13387

in plasma, PBMCs, and tumor tissue were performed. Bloodsamples (15 mL) for pharmacodynamic analyses were collectedbefore infusion and at 3, 7, and 24 hours after the start of infusionduring cycle 1 and immediately before thefirst infusion on cycle 2,day 1. Both plasma and PBMCs were isolated.

Measurement of HSP70 levels in plasma was evaluated byELISA and run as recommended by the manufacturer (ELISA;R&D Systems; Human/Mouse/Rat Total HSP70/HSPA1A Survey-or IC; Cat# SUV1663; ref. 20). In addition, several HSP90 clientproteins weremeasured in PBMC lysates byWestern blot analysis.Optional pretreatment tumor biopsy samples were taken atscreening or before the start of infusion on cycle 1, day 1.Posttreatment tumor biopsy samples were taken on cycle 1, day19 for patients on the twice-weekly regimen and on cycle 1, day 16for patients on the once-weekly regimen. HSP70 levels weremeasured by immunohistochemistry. If feasible, the HSP90 cli-ents CDK4, RAF-1, AKT, and S6, aswell as pAKT, pS6, and caspase-3 were also assessed. pS6 is a downstream target of mTORphosphorylation, which controls for mTOR inhibition; mTOR isa downstream target of PI3K/AKT.

ResultsPatient disposition and characteristics

Sixty-three patients were enrolled (62 patients receivedAT13387; 1 patient withdrew after enrollment but before treat-ment) in the study between May 2008 and August 2013. Table 1shows patient characteristics. Most patients (58; 94%) receivedprior chemotherapy and 45 patients (73%) received 3 or moreprior chemotherapy regimens. The median number of prior regi-mens was 4 (range 0–14). Thirty-four patients (55%) receivedprior radiotherapy.

The median number of cycles administered was 2, with arange of 1 to 12 cycles. Eight patients (13%) completed at least6 cycles of study treatment. The remaining 55 patients (87%)

Phase I Dose Escalation Trial of AT13387 in Metastatic Solid Tumors

www.aacrjournals.org Clin Cancer Res; 21(1) January 1, 2015 89




discontinued earlier: 41 patients (65%) due to disease progres-sion, 3 (5%) due to an AE, 6 (10%) due to withdrawal of consent,3 (5%) due to death, and 2 (3%) for other reasons (stated asclinical progression and investigator decision). All patients whoreceived study treatment were evaluable for safety. Forty-fourpatients (71%) were evaluable for best response to therapy byRECIST based on the investigator assessment [16 patients (26%)had measurable disease at baseline but no posttreatment mea-surements and two patients (3%) were not evaluable for otherreasons]. Sixty-one patients (98%) were evaluable for pharma-cokinetics, and 56 patients (90%) provided samples that wereanalyzed for HSP70 expression in plasma.

Dose escalation and recommended phase II dosePatients were assessed for the presence or absence of DLTs

during the first 4 weeks of treatment only (one cycle). Thenumber of patients treated per dose level was 4 at 10 mg/m2,3 at 20 mg/m2, 3 at 40 mg/m2, 5 at 80 mg/m2, 13 at 120 mg/m2

(all twice-weekly regimen), 4 at 150 mg/m2, 3 at 180 mg/m2,9 at 220 mg/m2, 13 at 260 mg/m2, and 5 at 310 mg/m2

(all once-weekly regimen). The dose escalation is summarizedin Supplementary Table S1.

For the twice-weekly regimen, doses were escalated from 10 to120 mg/m2. One patient of the first 3 enrolled in the 120-mg/m2

treatment group had a protocol-defined DLT involving grade 3visual disturbances with associated electroretinogram changesthat were later reversible. This cohort was expanded to a total of

6 patients, and no further DLTs occurred in the dose-escalationphase of the 120mg/m2 treatment group. However, grade 1 and 2infusion reactions in 2 of 6 patients (including dizziness, hyper-hidrosis, facial flushing, tachycardia, hypotension, bradycardia,and abdominal pain, both during and after infusion) and grade 1prolonged fatigue in 2 of 6 patients (starting 16 and 43 days,respectively, after treatment with AT13387) prompted a decisionnot to escalate further. Thus, the 120 mg/m2 dose level wasconsidered to be both the MTD and the RP2D for the twice-weekly regimen. The tolerability of this dose level was confirmedwith the treatment of an additional 7 patients.

For the once-weekly regimen, dosing started at 150mg/m2 andwas escalated to 310 mg/m2. While no protocol-defined DLTswere observed, 260mg/m2was considered the RP2Dbased on theoccurrence of severalmoderately severe toxicities, including grade2 diarrhea, nausea, vomiting, fatigue, and systemic infusionreactions in patients enrolled in the 310 mg/m2 treatment group.Additional patients were enrolled in the 260 mg/m2 treatmentgroup for a total of 13.

SafetyTables 2 and 3 show the most common AEs considered related

to AT13387 occurring in �10% of patients. Across the study, themajority of patients (73%) experienced diarrhea, 45% experi-enced visual disturbances, 29% experienced injection site events,and 27% experienced systemic infusion reactions. In addition todiarrhea, other gastrointestinal toxicities included nausea, dry

Table 1. Patients' baseline characteristics

Number (%) of patientsAT13387 AT13387 All

twice weekly once weekly patientsn ¼ 28 (%) n ¼ 34 (%) N ¼ 62 (%)

Age, yMean � SD (range) 54 � 11 (28, 71) 59 �10 (42, 76) 56� 10 (28, 76)

Sex, n (%)Females 14 (50) 19 (56) 33 (53)Males 14 (50) 15 (44) 29 (47)

Race, n (%)White 23 (82) 24 (71) 47 (76)Black or African American 2 (7) 5 (15) 7 (11)Asian 1 (4) 2 (6) 3 (5)Hispanic 2 (7) 3 (9) 5 (8)

ECOG PS, n (%)0 13 (46) 19 (56) 32 (52)1 15 (54) 13 (38) 28 (45)�2 0 2 (6) 2 (3)

Number of prior chemotherapy regimens1–2 7 (25) 6 (18) 13 (21)3–8 18 (64) 23 (68) 41 (66)9–14 1 (4) 3 (9) 4 (6)

Type of cancerColorectal cancer 6 (21) 10 (29) 16 (26)NSCLCa 4 (14) 6 (18) 10 (16)Metastatic melanoma 7 (25) 1 (3) 8 (13)GIST 1 (4) 6 (18) 7 (11)Adenocarcinoma of the pancreas 1 (4) 2 (6) 3 (5)Sarcomab 1 (4) 2 (6) 3 (5)Prostate adenocarcinoma 2 (7) 0 2 (3)Bladder cancer 1 (4) 1 (3) 2 (3)Esophageal cancer 1 (4) 1 (3) 2 (3)Otherc 4 (14) 5 (15) 9 (15)

aThere was one patient who had NSCLC with evidence of neuroendocrine differentiation.bThe 3 types of sarcoma were biphasic synovial sarcoma, liposarcoma of retroperitoneum, and osteosarcoma.cThe "other" types of cancerwere breast, pituitary cancer, thyroid cancer, and high-grade gliomas in the twice-weekly regimen and cervical cancer (neuroendocrine),melanoma, mesothelioma, parotid gland cancer, and cholagiocarcinoma in the once-weekly regimen.

Shapiro et al.





Table

2.Sum

maryofstud

ytrea

tmen

t-relatedad

verseev

ents

occurring

in�1

0%

ofpatients(twice-wee

klysche

dule)

10mg/m

220

mg/m

240mg/m

280mg/m

2120mg/m

2

AT133

87

twice-wee

kly

twice-wee

kly

twice-wee

kly

twice-wee

kly

twice-wee

klyc

AEapreferred

term

/Ove

ralltotal

twice-wee

kly

(n¼

4)

(n¼

3)(n

¼3)

(n¼

5)(n

¼13)

group

term

b(N

¼62)

(n¼

28)

Gr1–2

Gr3–

4Gr1–2

Gr3–

4Gr1–2

Gr3–

4Gr1–2

Gr3–

4Gr1–2

Gr3–

4

Diarrhe

a45(73)

18(64)

1(25)

0(0)

0(0)

0(0)

2(67)

0(0)

4(80)

1(20)

10(77)

0(0)

Visua

ldisturban

cesb

28(45)

10(36)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

9(69)

1(8)

Fatigue

26(42)

11(39)

1(25)

0(0)

1(33)

0(0)

1(33)

0(0)

2(40)

0(0)

6(46)

0(0)

Nau

sea

23(37)

6(21)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

1(20)

0(0)

5(38)

0(0)

Injectionsite

even

tsb

18(29)

1(4)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

1(8)

0(0)

Systemic

infusionreactions

b17

(27)

2(7)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

2(15)

0(0)

Dizzine

ss16

(26)

3(11)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

3(23)

0(0)

Dry

mouth

14(23)

5(18)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

1(20)

0(0)

4(31)

0(0)

Ane

mia

13(21)

7(25)

1(25)

0(0)

1(33)

0(0)

0(0)

0(0)

0(0)

0(0)

5(38)

0(0)

Insomnia

10(16)

1(4)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

1(8)

0(0)

Musclespasms

10(16)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

Abdominal

pain

9(15)

3(11)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

1(20)

0(0)

2(15)

0(0)

Hea

dache

9(15)

1(4)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

1(8)

0(0)

Rash

8(13)

3(11)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

1(20)

0(0)

2(15)

0(0)

Weight

decreased

8(13)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

Decreased

appetite

7(11)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

Dyspho

nia

6(10)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

Hem

oglobin

decreased

6(10)

6(21)

0(0)

0(0)

0(0)

0(0)

1(33)

0(0)

3(60)

0(0)

2(15)

0(0)

Hyp

otension

6(10)

1(4)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

1(8)

0(0)

Pruritis

6(10)

1(4)

1(25)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

aAEsco

nsidered

possibly,p

robab

ly,o

rdefi

nitely

relatedto

thestud

ytrea

tmen

t.bForboth

stud

ytrea

tmen

t-relatedan

dun

relatedAEsin

thisstud

y,thegroup

term

visualdisturban

cesinclud

edtheAEpreferred

term

schloropsia,diplopia,h

alovision,

night

blindne

ss,p

hotopsia,tunn

elvision,

vision

blurred

,visua

limpairm

ent,an

dvitreo

usfloaters.The

group

term

injectionsite

even

tsinclud

edthead

verseev

entpreferred

term

sextravasation,infusionsite

inflam

mation,

infusionsite

pain,injectionsite

reaction,an

dinjectionsite

pain.The

group

term

system

icinfusionreactions

includ

edthead

verseev

entpreferred

term

schills,flushing

,hyp

erhidrosis,hy

persensitivity,infusion-relatedreaction,pruritus,an

dpyrexia,b

utonlyifthese

even

tsweredefi

nitely

orpossibly

relatedto

stud

ytrea

tmen

tan

doccurredwithin24

hoursofadose.

cPatientswho

received

themaxim

umtolerateddose

duringthedose-escalationpha

sean

dtheco

hortexpan

sionpha

se.






Table

3.Sum

maryofstud

ytrea

tmen

t-relatedAEsoccurring

in�1

0%

ofpatients(once-wee

klysche

dule)

150mg/m

2180mg/m

222

0mg/m

226

0mg/m

2310mg/m

2

AT133

87

onc

e-wee

kly

onc

e-wee

kly

onc

e-wee

kly

onc

e-wee

klyc

onc

e-wee

kly

AEapreferred

term

/Ove

ralltotal

onc

e-wee

kly

(n¼

4)

(n¼

3)(n

¼9)

(n¼

13)

(n¼

5)group

term

b(N

¼62)

(n¼

34)

Gr1–2

Gr3–

4Gr1–2

Gr3–

4Gr1–2

Gr3–

4Gr1–2

Gr3–

4Gr1–2

Gr3–

4

Diarrhe

a45(73)

27(79)

3(75)

0(0)

2(67)

0(0)

7(78)

0(0)

11(85)

1(8)

3(60)

0(0)

Visua

ldisturban

cesb

28(45)

18(53)

1(25)

0(0)

1(33)

0(0)

5(56)

0(0)

7(54)

0(0)

4(80)

0(0)

Fatigue

26(42)

15(44)

0(0)

0(0)

1(33)

0(0)

3(33)

0(0)

8(62)

0(0)

3(60)

0(0)

Nau

sea

23(37)

17(50)

4(100)

0(0)

1(33)

0(0)

1(11)

0(0)

8(62)

0(0)

3(60)

0(0)

Injectionsite

even

tsb

18(29)

17(50)

3(75)

0(0)

1(33)

0(0)

6(67)

0(0)

6(46)

0(0)

1(20)

0(0)

Systemic

infusionreactions

b17

(27)

15(44)

0(0)

0(0)

0(0)

0(0)

4(44)

0(0)

6(46)

0(0)

5(100)

0(0)

Dizzine

ss16

(26)

13(38)

0(0)

0(0)

0(0)

0(0)

2(22)

0(0)

8(62)

0(0)

3(60)

0(0)

Dry

mouth

14(23)

9(26)

3(75)

0(0)

1(33)

0(0)

1(11)

0(0)

2(15)

0(0)

2(40)

0(0)

Ane

mia

13(21)

6(18)

1(25)

0(0)

1(33)

0(0)

1(11)

0(0)

3(23)

0(0)

0(0)

0(0)

Insomnia

10(16)

9(26)

0(0)

0(0)

0(0)

0(0)

5(56)

0(0)

2(15)

0(0)

2(40)

0(0)

Musclespasms

10(16)

10(29)

1(25)

0(0)

0(0)

0(0)

3(33)

0(0)

4(31)

0(0)

2(40)

0(0)

Abdominal

pain

9(15)

6(18)

0(0)

0(0)

0(0)

0(0)

1(11)

0(0)

3(23)

0(0)

2(40)

0(0)

Hea

dache

9(15)

8(24)

0(0)

0(0)

1(33)

0(0)

3(33)

0(0)

3(23)

0(0)

1(20)

0(0)

Rash

8(13)

5(15)

0(0)

0(0)

0(0)

0(0)

1(11)

0(0)

1(8)

0(0)

3(60)

0(0)

Weight

decreased

8(13)

8(24)

1(25)

0(0)

0(0)

0(0)

2(22)

0(0)

2(15)

0(0)

3(60)

0(0)

Decreased

appetite

7(11)

7(21)

0(0)

0(0)

0(0)

0(0)

1(11)

0(0)

4(31)

0(0)

2(40)

0(0)

Dyspho

nia

6(10)

6(18)

0(0)

0(0)

0(0)

0(0)

1(11)

0(0)

4(31)

0(0)

1(20)

0(0)

Hem

oglobin

decreased

6(10)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

Hyp

otension

6(10)

5(15)

0(0)

0(0)

0(0)

0(0)

0(0)

0(0)

4(31)

0(0)

0(0)

1(20)

Pruritis

6(10)

5(15)

0(0)

0(0)

0(0)

0(0)

1(11)

0(0)

2(15)

0(0)

2(40)

0(0)

aAEsco

nsidered

possibly,p

robab

ly,o

rdefi

nitely

relatedto

thestud

ytrea

tmen

t.bForboth

stud

ytrea

tmen

t-relatedan

dun

relatedAEsin

thisstud

y,thegroup

term

visualdisturban

cesinclud

edtheAEpreferred

term

schloropsia,diplopia,h

alovision,

night

blindne

ss,p

hotopsia,tunn

elvision,

vision

blurred

,visua

limpairm

ent,an

dvitreo

usfloaters.The

group

term

injectionsite

even

tsinclud

edthead

verseev

entpreferred

term

sextravasation,

infusionsite

inflam

mation,

infusionsite

pain,injectionsite

reaction,

and

injectionsite

pain.The

group

term

system

icinfusionreactions

includ

edthead

verseev

entpreferred

term

schills,flushing

,hyp

erhidrosis,hy

persensitivity,infusion-relatedreaction,pruritus,an

dpyrexia,b

utonlyifthese

even

tsweredefi

nitely

orpossibly

relatedto

stud

ytrea

tmen

tan

doccurredwithin24

hoursofadose.

c Patientswho

received

themaxim

umtolerateddose

duringthedose-escalationpha

sean

dtheco

hortexpan

sionpha

se.

Shapiro et al.





mouth, and abdominal pain. Most of the gastrointestinal-relatedAEs were of grade 1 or 2 severity, responsive to supportivemedications and reversible. Minimal effects on liver functionwere observed, with only mild elevations of transaminases[maximal increase of �2 times upper limit of normal (ULN)],which were transient, without clinical symptoms and not doselimiting.

Forty-seven percent of patients had at least one episode ofvisual disturbance. These events occurred only in patients receiv-ing doses of 80 mg/m2 or above and included transient visualflashes and changes in light/dark accommodation. Patientdescriptions also included "squiggles," "moving objects," and"blobs of light." Except in the one case that resulted in a DLT,these events were all of grade 1 or 2 severity and did not affectpatients' daily lives. The duration varied from 1 day (lastinggenerally a few seconds to a few minutes) to ongoing at studydiscontinuation. Aside from one patient with grade 1 visualimpairment still ongoing at the last follow up (3 months fromthe start of the event), these events were fully reversible. All of thetreatment-related systemic infusion reactions or injection siteevents were mild to moderate (grade 1 or 2) and resolved within48 hours of infusion.

Central ECG recordings (n ¼ 60) revealed no change in heartrate, atrioventricular conduction (PR interval), or depolarization(QRS duration). There were no clear signals of any change inmorphology or new rhythms. No effect of AT13387 was observedon cardiac repolarization (QTcF duration) up to the 120 mg/m2

twice-weekly regimen. On the once-weekly regimen, doses of 220to 310mg/m2 showed a slight trend ofQTcF increase in the 10msto 15 ms range at the time of Cmax.

Eleven patients had 1 or more drug-related grade 3 or 4 AEs,including diarrhea and syncope (in 2 patients each), and dyspnea,urticaria, prolonged QTc on ECG, hypotension, hyponatremia,hypertension, visual disturbance, and asthenia (in 1patient each).Except for visual disturbance, none were DLTs (diarrhea wasexcluded if not maximally treated, and the other events wereeither not considered to be related to treatment by the safetymonitoring committee or occurred in cycle 2 or later). Four of the11 patients discontinued treatment due to drug-related grade 3 or4 AEs, and the remaining 6 (excluding the DLT, described above)received medication. All drug-related grade 3 or 4 AEs resolvedbefore study completion, except for hypotension and asthenia.Seven patients (11%) died within 30 days of last study treatment.Six of these were due to disease progression and one was frompneumonia/sepsis (not considered related to treatment).

Across the study, AT13387 dose was reduced due to AEs in 9patients (in the 80, 120, 260, and 310 mg/m2 dose levels). The

dose reductions were all for grade 1 or 2 AEs, except for 1 patientwith grade 3 hyponatremia, an event that occurred during cycle 2,and so was not considered dose limiting.

Pharmacokinetic analysesThepharmacokinetics of AT13387 showed adose-proportional

increase inAUC0–t andCmax from10 to120mg/m2 (twiceweekly)and from to 150 to 310 mg/m2 (once weekly) with relatively lowinterindividual variability (Table 4, Fig. 1). The t1/2 was dose-independent and ranged from 6.6 to 11.5 hours. There was nonotable accumulation or reduction in exposures between cycle 1,day 1 and day 18 in the twice-weekly regimen or cycle 1, day 1 andcycle 2, day 15 in the once-weekly regimen. Plasma clearance ofAT13387 was relatively high, independent of dose, and rangedbetween 40.9 and 69.1 L/h/m2, likely related to the knownconversion to circulating metabolites, including two isomericO-glucuronide conjugates. The volume of distribution was high(427–872 L/m2).

Urinary excretion of AT13387 was low. The maximum rate ofexcretion was observed within 3.50 to 11.9 hours (mean Tmax)following intravenous infusion for the 10 to 120 mg/m2 doselevels. The percent of the dose recovered in urine ranged from0.92% to 2.98%. The two isomeric O-glucuronide conjugatesrepresenting the major metabolites were detectable in urine. Thelow recovery of the drug dose in urine suggests the existence ofalternative pathway(s) of excretion. ADME studies in humanshave not yet been conducted; however, results of studies innonclinical toxicology models suggest that excretion of AT13387and its conjugated metabolites is mainly through the biliarypathway.

Pharmacodynamic analysesIn addition to client protein degradation, induction of HSP70

typically follows pharmacologic HSP90 inhibition; therefore,HSP70 upregulation is commonly used as biomarker evidenceof HSP90 inhibition and target engagement in clinical trials (6).HSP70 expression was assessed in plasma and tumor biopsysamples when possible. Increased levels of HSP70 in plasmaweredetected across all dose levels and were dose-dependent up to thefifth dose level (120 mg/m2; �2- to 6-fold increase) in the twice-weekly regimen. In the once-weekly regimen, HSP70 inductionappeared to be dose-dependent until the eighth dose level(220 mg/m2; Fig. 2A). Pre- and posttreatment tumor biopsysamples were taken and successfully analyzed from 4 patientstreated at the MTD (120 mg/m2) for the twice-weekly regimen.A consistent increase in HSP70 staining was detected in 3 of the4 samples (Fig. 2B). No relationship was observed between

Table 4. Summary of pharmacokinetic parameters mean (�SD) for AT13387 (cycle 1, day 1)

Dose Dose AUC0–24 (h�ng/mL) t1/2 (h) Cmax (ng/mL) Cl (L/h/m2) Vz (L/m2)Dose level (mg/m2) regimen Mean (SD) Mean (SD) Mean (SD) Mean (SD) Mean (SD)

1 (N ¼ 4) 10 TW 165.1 (53.9) 11.2 (2.0) 69.4 (40.7) 52.0 (14.7) 872 (386)2 (N ¼ 3) 20 TW 309.6 (8.4) 9.1 (2.1) 108 (53.5) 56.5 (2.28) 733 (140)3 (N ¼ 3) 40 TW 754.5 (606) 10.7 (2.1) 463 (509) 45.1 (17.4) 708 (321)4 (N ¼ 5) 80 TW 1,499 (307) 9.4 (2.6) 526 (175) 48.1 (10.4) 639 (174)5 (N ¼ 13) 120 TW 1,830 (558) 11.5 (2.7) 714 (405) 57.0 (21.1) 807 (180)6 (N ¼ 4) 150 OW 2,836 (660) 8.4 (1.1) 1,570 (744) 51.4 (16.5) 617 (165)7 (N ¼ 3) 180 OW 2,414 (216) 8.1 (0.8) 970 (182) 69.1 (7.88) 802 (16.8)8 (N ¼ 9) 220 OW 5,192 (1201) 7.7 (1.5) 2,340 (865) 40.9 (8.45) 448 (93.3)9 (N ¼ 12) 260 OW 4,905 (719) 6.6 (1.6) 1,790 (422) 50.6 (10.7) 491 (79.4)10 (N ¼ 5) 310 OW 7,293 (1986) 7.0 (0.6) 3,790 (2820) 42.0 (8.69) 427 (118)

Abbreviations: TW, twice weekly; OW, once weekly.






clinical outcome and HSP70 induction. Best response was eitherPD (2 patients) or not evaluable (1 patient).

Depletion of HSP90 client proteins in PBMCs showed largevariability and differences in effect. Depletion of HSP90 clientproteins (Raf-1,CDK4, phospho-S6, and levels of cleaved caspase-3) in pre- and posttreatment tumor biopsies taken from the4 patients treated at the MTD for the twice-weekly regimen(120mg/m2) showed large variability across the samples studied.Increased plasma HSP70 was observed in all cohorts; therefore,HSP70 induction in plasma was not sufficient to predict responseof HSP90 inhibitors by evidence of target engagement.

Antitumor activityDisease response was also assessed during the study. One

patient (2%) had a partial response, 21 patients (34%) had stabledisease (SD), and 22 patients (35%) had progressive disease asbest response to therapy based on the investigator's assessment ofresponse to treatment.

Across the study, 7patientshadSD for at least 120days (median184 days, range 155–335 days; Table 5), including 2 patientson the twice-weekly regimen (20 and 80 mg/m2, respectively),and 5 patients on the once-weekly regimen (220–310 mg/m2).

The partial response occurred in a 60-year-old male withgastrointestinal stromal tumors (GIST), who had previously beentreated with imatinib (doses escalated from 300 to 800 mg/day).He received AT13387 at 220mg/m2 on the once-weekly regimen,with successive decreases in tumor size resulting in a partialresponse by RECIST after 6 months. He remained progression-free for an additional 113 days (�3.8 months), so that he was onAT13387 for approximately 10 months. Direct DNA sequencingresults for this patient's tumor at diagnosis (imatinib sensitive)and recurrence (imatinib resistant) showed that this patient had aprimary c-KITmutation (exon 11 deletion 558–572) at diagnosis,and secondary mutations (exon 17 mutations D816H, D820H)associated with resistance to imatinib and other tyrosine kinaseinhibitors (TKI; ref. 21) after recurrence and before treatmentwith

Figure 1.AT13387 mean plasma concentration–time profiles. A, twice-weeklyregimen, cycle 1, day 1; B, once-weeklyregimen, cycle 1, day 1.

Shapiro et al.





AT13387, so that the response to AT13387 occurred despite thepresence of mutations conferring TKI resistance.

Among 6 additional patients with GIST, 3 had a best responseof stable disease. These patients remained progression-free for 43days (�1.4 months), 231 days (�7.7 months), and 335 days

(�11.2 months), respectively. A 59-year-old male was progres-sion-free for 231 days. He had a tumor in the liver and hadreceived prior imatinib. He was treated with AT13387 at 220 mg/m2 once weekly and had interval decrease in the size and metab-olism of the hepatic lesion [standardized uptake volume (SUV)decrease of 48% on 18F-fluorodeoxyglucose positron emissiontomography]. A 64-year-old female was progression-free for 335days. She had widespread abdominal tumor seeding followingsurgery for GIST involving the terminal ileum and colon.Although she was not able to tolerate more than 30 days oftreatment with imatinib due to neutropenia, AT13387 (260mg/m2 once weekly) afforded nearly 1 year of disease controlby study end, withmultiple implants observed during surgery butnot measurable according to RECIST (by CT scan). The mostrecent CT scan (3 years later) showed no progressive disease.

Ten heavily pretreated patients with NSCLC were enrolledrepresenting a variety of genomic subsets, although none withknownALK rearrangement. Four patients achieved abest responseof stable disease, including 3 patientswith EGFRwild-type diseaseand 1 patient with tumor harboring a HER2 exon 12 insertionmutation. However, stabilizations were brief, lasting 47 to 85days. Two patients with tumor harboring an EGFR exon 20insertion and a KRAS G12D mutation, respectively, had progres-sive disease as the best response and 1 patient with tumorharboring an EGFR exon 19 deletion was not evaluable.

DiscussionThis first-in-human, phase I study of AT13387 identified the

MTD and RP2D for 2 regimens in 62 patients with previouslytreated solid tumors. Single-agent AT13387 has an acceptablesafety profile and exhibited linear pharmacokinetic properties.Exposure increased proportionally with increasing dose anddemonstrated no accumulation or reduction on day 15 or day18 versus day 1. Systemic pharmacokinetic exposures in the clinicat MTD dose(s) achieved levels associated with efficacy in pre-clinical xenograft studies when adjusted for differences in proteinbinding between nude mice and humans (11, 14).

AEs commonly observed with other HSP90 inhibitors aregastrointestinal disorders (nausea/vomiting/diarrhea), visual dis-turbances, and hepatic toxicity (7–9, 22–24). The gastrointestinaltoxicity of AT13387 includes grade 1–2 diarrhea that was mostlyself-limited or managed with antidiarrheals. Visual disturbancesalso occurred; these events were typically transient and low grade.Ophthalmologic toxicity may be an inherent property of potentwater-soluble HSP90 inhibitors, as these findings have beenreported with other agents, including 17-DMAG and AUY922

Figure 2.Pharmacodynamic assessment of AT13387 in plasma and tumor. A, HSP70plasma ELISA data were expressed as the greatest fold induction postdose foreach individual patient. Patients with limited pharmacodynamic sampling wereexcluded. Patients represented by dose level (mg/m2) were 10 (n ¼ 3), 20(n¼ 2), 40 (n¼ 3), 80 (n¼ 5), 120 (n¼ 11), 150 (n¼ 4), 180 (n¼ 3), 220 (n¼ 9),280 (n ¼ 13), 310 (n ¼ 5). Median fold increases are represented as horizontalbars (n¼ 58patients represented). B, HSP70 induction in paired tumorbiopsiesfrom 3 patients at 120 mg/m2 (MTD) on the twice-weekly regimen.

Table 5. Patients with stable disease (SD) �120 days across the study

Patient CohortDosingregimen Type of cancer

Number of priorchemotherapyregimens/priorradiotherapy (Y/N)

Overall bestresponse

Duration of bestresponse (days)

A 20 mg/m2 TW Thyroid cancer 4/Y SD 164B 80 mg/m2 TW Metastatic melanoma 0/Y SD 217C 220 mg/m2 OW GIST 1/N SD 231D 260 mg/m2 OW GIST 2/N SD 335E 260 mg/m2 OW Adenoid cystic carcinoma

of the right parotid1/Y SD 184

F 310 mg/m2 OW Colon cancer 3/N SD 161G 310 mg/m2 OW Colon cancer 8/N SD 155

Abbreviations: TW, twice weekly; OW, once weekly; SD, stable disease.






(8, 25, 26). Other commonly occurring AT13387 treatment-related AEs (�30%) included nausea and injection site eventsand systemic infusion reactions, all of whichwere grade 1 or 2 andmanaged prophylactically or treated symptomatically.

The unfavorable hepatotoxicity of first-generation ansamycinclass of HSP90 inhibitors (17-AAG, DMAG, and IPI-504) thathampered clinical development was not observed in this study(22, 27, 28). In the few patients with any changes in hepatictransaminases considered possibly-related to AT13387, thechanges were mild (�2X ULN) and generally transient, which isconsistent with findings with other second-generation HSP90inhibitors (9). In addition, cardiotoxicity was not detected in thisstudy. Independent central cardiac review did not show a signif-icant effect on QTc interval.

HSP90 inhibitors have been evaluated on multiple regimens.First-generation geldanamycins and second-generation non-gel-danamycin agents, including AT13387, as well as AUY922 andganetespib, demonstrate preferential accumulation in tumors,with long intratumoral t1/2 (often detectable for 7–10 days intumor), justifying once-weekly regimens (11–14).Here, AT13387has demonstrated preliminary activity against KIT-driven GIST,including a partial response in a patient with tumor harboringmutations conferring resistance to TKIs. This partial response wasin a patient receiving AT13387 on the once-weekly regimen,implying that regimen is sufficient to produce antitumor activityin sensitive tumors.

AT13387has been shown to be active inGIST xenograftmodels(15) and imatinib-resistant GIST remains largely c-KIT–depen-dent; hence, targeting c-KIT via inhibition of the HSP90 chaper-one, essential for stability of all mutant KIT proteins, has been anattractive clinical setting forHSP90 inhibitors (15). Enrollment ofpatients with GIST who failed previous TKI treatment wasenriched in the once-weekly regimen to explore the hypothesisthat patients with acquired resistance to imatinib due to second-ary resistancemutations within c-KIT might respond to AT13387,with 2 additional patients achieving prolonged stable disease. Onthe basis of these findings, a trial in which AT13387 is combinedwith imatinib in patients with GIST is currently underway(NCT01294202).

Defining the optimal regimen for AT13387 and other HSP90inhibitors has proved complex. Despite the long intratumoral t1/2noted with these agents, the suppression of client protein expres-sion has been noted to bemore transient. For example, in mutantEGFR (L858R/T790M) NCI-H1975 NSCLC xenografts, depletionof mutant EGFR has been documented to last a duration of 72hours or shorter in response to a single dose of AT13387 organetespib (13, 14). In a clinical evaluation of once-weeklyganetespib in GIST (29), tumor biopsies demonstrated onlytransient depletionofKIT and inhibitionofdownstream signalingpathways, suggesting that the once-weekly regimen was subop-timal for producing prolonged client protein depletion. Similarly,in one of the first phase I studies of the geldanamycin 17-AAG,depletion of c-RAF and CDK4 in tumor biopsies could be dem-onstrated at 24 hours, but not at 5 days after treatment (22). Theseresults have prompted the evaluation of more frequent dosingregimens, including the twice-weekly regimen evaluated here, aswell as consecutive day dosing (NCT01246102), the latter pro-ducing the most prolonged client depletion in various preclinicalstudies (13). This study demonstrates that twice-weekly dosing ofAT13387 is tolerable with clear evidence of target engagement intumor, documented by induction of HSP70 in posttreatment

biopsy samples (30). Future studies incorporating biopsies atadditional time points, focusing on longevity of client proteindepletion in conjunction with clinical activity in sensitive popu-lations, will be instructive to justify continued development of themore frequent dosing regimens. In the consecutive day dosingtrial of AT13387, indium-labeled trastuzumab scanning will beused to evaluate the drug effect on HER2 levels in patients withHER2-expressing tumors. Whether twice-weekly or consecutive-day dosing of an HSP90 inhibitor will improve efficacy may bedependent on the specific clinical setting (disease type, specificclient protein targeted for depletion, whether single agent or incombination) and ultimately will require randomization to theonce-weekly regimen for direct comparisonof toxicity and efficacyin homogeneous patient populations.

There has been substantial interest in the activity of HSP90inhibitors in oncogene-driven subsets of NSCLC. In particular,twice-weekly retaspimycin and once-weekly AUY922 and gane-tespib have been evaluated in phase II trials. These studies haveroutinely demonstrated activity in ALK-dependent tumors(7, 31, 32), which along with HER2, may be one of the mostsensitive HSP90 clients. More variable activity has been demon-strated in EGFR-mutant tumors. In this study, only 2 patients withEGFR-mutant NSCLC were enrolled, and the 1 evaluable patienthad progressive disease as the best response. Currently, AT13387is being evaluated in combination with crizotinib in a phase I/IItrial for patients withNSCLCharboringALK rearrangement. Oncedoses of the combination are defined, this studywill evaluate bothAT13387 alone and the combination in patients who have pro-gressed on crizotinib (NCT01712217).

In summary, we have demonstrated that AT13387 is toler-able in patients with advanced solid tumors, with linear phar-macokinetics, evidence of target engagement, and preliminaryantitumor activity. In addition to the current clinical evaluationin GIST and ALK-rearranged NSCLC, AT13387 is also beinginvestigated both alone and in combination with abiraterone incastration-resistant prostate cancer (NCT01685268; the trialhas completed enrollment). In addition, we have demonstratedthe activity of AT13387 in preclinical models of BRAF and MEKinhibitor-resistant BRAF mutant melanoma, so that furtherwork is also planned in this disease. These trials should providefurther insight into the activity of AT13387 as monotherapyand in combination with TKIs or hormonal therapy in cancersdriven by HSP90 clients.

Disclosure of Potential Conflicts of InterestG.I. Shapiro and M. Yule are consultant/advisory board members for Astex.

J. Lyons is an employee of Astex. Nopotential conflicts of interest were disclosedby the other authors.

Authors' ContributionsConception and design: G.I. Shapiro, B.J. Dezube, M. YuleDevelopment of methodology: G.I. Shapiro, B.J. DezubeAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): G.I. Shapiro, E. Kwak, B.J. Dezube, J. LyonsAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis):G.I. Shapiro, E. Kwak, B.J. Dezube, J. Ayrton, J. LyonsWriting, review, and/or revision of the manuscript: G.I. Shapiro, E. Kwak,B.J. Dezube, M. Yule, J. Lyons, D. MahadevanAdministrative, technical, or material support (i.e., reporting or organizingdata, constructing databases): G.I. Shapiro, J. AyrtonStudy supervision: G.I. Shapiro, M. YuleOther (conducted the trial as the principal investigator at the Arizona site andwrote part of the manuscript): D. Mahadevan


Shapiro et al.




AcknowledgmentsThe authors thank the patients and their families and acknowledge the

contributions of Gillian Langford (PhD), Harold Keer (MD, PhD), AramOganesian (PhD, DABT), Amarpal Sahai (BScH, MSc), Nicola Wallis (PhD),Samantha Lewis, Ana Rodriguez-Lopez (PhD), Victoria Lock (BSc), JosephIovino, and Mohammad Azab (MD, MSc, MBA) of Astex Pharmaceuticals, Inc.The authors also thank Amy Burdan (MS) from Sandpiper Communicationsand Kristen Mayo (PhD) for their editorial assistance and Chris Corless (MD,PhD), of Oregon Health and Science University (Portland, OR), for sequencingthe KIT mutations.

Grant SupportThis studywas sponsored, monitored, and funded by Astex Pharmaceuticals,

Inc. (formerly SuperGen, Inc.).The costs of publication of this article were defrayed in part by the

payment of page charges. This article must therefore be hereby markedadvertisement in accordance with 18 U.S.C. Section 1734 solely to indicatethis fact.

Received April 20, 2014; revised August 12, 2014; accepted September 9,2014; published OnlineFirst October 21, 2014.

References1. Shimamura T, Lowell AM, Engelman JA, Shapiro GI. Epidermal growth

factor receptors harboring kinase domainmutations associatewith the heatshock protein 90 chaperone and are destabilized following exposure togeldanamycins. Cancer Res 2005;65:6401–8.

2. Pearl LH, Prodromou C,Workman P. The Hsp90molecular chaperone: anopen and shut case for treatment. Biochem J 2008;410:439–53.

3. Garcia-Carbonero R, Carnero A, Paz-Ares L. Inhibition ofHSP90molecularchaperones: moving into the clinic. Lancet Oncol 2013;14:e358–69.

4. XuW, Marcu M, Yuan X, Mimnaugh E, Patterson C, Neckers L. Chaperone-dependent E3 ubiquitin ligase CHIPmediates a degradative pathway for c-ErbB2/Neu. Proc Natl Acad Sci U S A 2002;99:12847–52.

5. Neckers L, Workman P. Hsp90 molecular chaperone inhibitors: are wethere yet? Clin Cancer Res 2012;18:64–76.

6. Biamonte MA, Van deWater R, Arndt JW, Scannevin RH, Perret D, Lee WC.Heat shock protein 90: inhibitors in clinical trials. J Med Chem 2010;53:3–17.

7. Socinski MA, Goldman J, El-Hariry I, KoczywasM, Vukovic V, Horn L, et al.A multicenter phase II study of ganetespib monotherapy in patients withgenotypically defined advancednon-small cell lung cancer. ClinCancer Res2013;19:3068–77.

8. Sessa C, Shapiro GI, Bhalla KN, Britten C, Jacks KS, Mita M, et al. First-in-human phase I dose-escalation study of the HSP90 inhibitor AUY922 inpatients with advanced solid tumors. Clin Cancer Res 2013;19:3671–80.

9. Goldman JW, Raju RN, Gordon GA, El-Hariry I, Teofilivici F, Vukovic VM,et al. A first in human, safety, pharmacokinetics, and clinical activity phase Istudy of once weekly administration of the Hsp90 inhibitor ganetespib(STA-9090) in patients with solidmalignancies. BMCCancer 2013;13:152.

10. Hong D, Said R, Falchook G, Naing A, Moulder S, Tsimberidou AM, et al.Phase I study of BIIB028, a selective heat shock protein 90 inhibitor, inpatients with refractory metastatic or locally advanced solid tumors. ClinCancer Res 2013;19:4824–31.

11. WoodheadAJ, AngoveH, CarrMG, Chessari G, CongreveM, Coyle JE, et al.Discovery of (2,4-Dihydroxy-5-isopropylphenyl)-[5-(4-methylpiperazin-1-ylmethyl)-1,3-dihydroisoindol-2-yl]methanone (AT13387), a novelinhibitor of the molecular chaperone Hsp90 by fragment based drugdesign. J Med Chem 2010;53:5956–69.

12. Eccles SA, Massey A, Raynaud FI, Sharp SY, Box G, Valenti M, et al. NVP-AUY922: a novel heat shock protein 90 inhibitor active against xenografttumor growth, angiogenesis, andmetastasis. Cancer Res 2008;68:2850–60.

13. Shimamura T, Perera SA, Foley KP, Sang J, Rodig SJ, Inoue T, et al.Ganetespib (STA-9090), a nongeldanamycin HSP90 inhibitor, has potentantitumor activity in in vitro and in vivo models of non-small cell lungcancer. Clin Cancer Res 2012;18:4973–85.

14. GrahamB,Curry J, Smyth T, Fazal L, Feltell R,Harada I, et al. The heat shockprotein 90 inhibitor, AT13387, displays a long duration of action in vitroand in vivo in non-small cell lung cancer. Cancer Sci 2012;103:522–7.

15. Smyth T, Van Looy T, Curry JE, Rodriguez-Lopez AM, Wozniak A, Zhu M,et al. The HSP90 inhibitor, AT13387, is effective against imatinib-sensitiveand -resistant gastrointestinal stromal tumor models. Mol Cancer Ther2012;11:1799–808.

16. Mahadevan D, Rensvold DM, Kurtin SE, Cleary JM, Gandhi L, Lyons JF,et al. First-in-human phase I study: Results of a second-generation non-ansamycin heat shock protein 90 (HSP90) inhibitor AT13387 in refractorysolid tumors. J Clin Oncol 2012;30:3028.

17. Mahadevan D, Shapiro G, Kurtin SE, Cleary JM, Lyons JF, Rodriguez-LopezA, et al. Activity of AT13387, a novel, non-ansamycin inhibitor of heatshock protein 90 is active against gastrointestinal stromal tumors (GIST). JClin Oncol 2012;31:105.

18. Peterson LB, Eskew JD, Vielhauer GA, Blagg BS. The hERG channel isdependent upon the Hsp90a isoform for maturation and trafficking. MolPharm 2012;9:1841–6.

19. Therasse P, Arbuck SG, Eisenhauer EA,Wanders J, Kaplan RS, Rubinstein L,et al. New guidelines to evaluate the response to treatment in solid tumors.European Organization for Research and Treatment of Cancer, NationalCancer Institute of theUnited States, National Cancer Institute of Canada. JNatl Cancer Inst 2000;92:205–16.

20. Ramanathan RK, Egorin MJ, Eiseman JL, Ramalingam S, Friedland D,Agarwala SS. Phase I and pharmacodynamic study of 17-(allylamino)-17-demethoxygeldanamycin in adult patients with refractory advancedcancers. Clin Cancer Res 2007;13:1769–74.

21. Gajiwala KS, Wu JC, Christensen J, Deshmukh GD, Diehl W, DiNitto JP,et al. KIT kinase mutants show unique mechanisms of drug resistance toimatinib and sunitinib in gastrointestinal stromal tumor patients. ProcNatl Acad Sci U S A 2009;106:1542–7.

22. Banerji U, O'Donnell A, Scurr M, Pacey S, Stapleton S, Asad Y, et al. Phase Ipharmacokinetic and pharmacodynamic study of 17-allylamino, 17-demethoxygeldanamycin in patients with advanced malignancies. J ClinOncol 2005;23:4152–61.

23. Lancet JE, Gojo I, Burton M, Quinn M, Tighe SM, Kersey K, et al.Phase I study of the heat shock protein 90 inhibitor alvespimycin(KOS-1022, 17-DMAG) administered intravenously twice weeklyto patients with acute myeloid leukemia. Leukemia 2010;24:699–705.

24. Nowakowski GS, McCollum AK, Ames MM, Mandrekar SJ, Reid JM,Adjei AA, et al. A phase I trial of twice-weekly 17-allylamino-demethoxy-geldanamycin in patients with advanced cancer. Clin Can-cer Res 2006;12:6087–93.

25. Rajan A, Kelly RJ, Trepel JB, Kim YS, Alarcon SV, Kummar S, et al. A phase Istudy of PF-04929113 (SNX-5422), an orally bioavailable heat shockprotein 90 inhibitor, in patients with refractory solid tumor malignanciesand lymphomas. Clin Cancer Res 2011;17:6831–9.

26. Pacey S, Wilson RH, Walton M, Eatock MM, Hardcastle A, Zetterlund A,et al. A phase I study of the heat shock protein 90 inhibitor alvespimycin(17-DMAG) given intravenously to patients with advanced solid tumors.Clin Cancer Res 2011;17:1561–70.

27. Taldone T, Gozman A, Maharaj R, Chiosis G. Targeting Hsp90: small-molecule inhibitors and their clinical development. Curr Opin Pharmacol2008;8:370–4.

28. Jhaveri K, Taldone T, Modi S, Chiosis G. Advances in the clinical devel-opment of heat shock protein 90 (Hsp90) inhibitors in cancers. BiochimBiophys Acta 2011;1823:742–55.

29. Demetri GD, Heinrich MC, Chmielowski B, Morgan JA, George S, BradleyR, et al. An open-label phase II study of the Hsp90 inhibitor ganetespib(STA-9090) in patients (pts) with metastatic and/or unresectable GIST.J Clin Oncol 29: 2011 (suppl; abstr 10011).

30. Zhang H, Chung D, Yang YC, Neely L, Tsurumoto S, Fan J, et al. Identi-fication of new biomarkers for clinical trials of Hsp90 inhibitors. MolCancer Ther 2006;5:1256–64.

31. Sequist LV, Gettinger S, Senzer NN, Martins RG, Janne PA, Lilenbaum R,et al. Activity of IPI-504, a novel heat-shock protein 90 inhibitor, in patientswith molecularly defined non-small cell lung cancer. J Clin Oncol2010;28:4953–60.

32. Garon EB,Moran T, Barlesi F, Gandhi L, Sequist LV, Kim S-W, et al. Phase IIstudy of the HSP90 inhibitor AUY922 in patients with previously treated,advanced non-small cell lung cancer (NSCLC). J Clin Oncol 30, 2012(suppl; abstr 7543).






2015;21:87-97. Published OnlineFirst October 21, 2014.Clin Cancer Res Geoffrey I. Shapiro, Eunice Kwak, Bruce J. Dezube, et al. Patients with Advanced Solid TumorsSecond-Generation Non-Ansamycin HSP90 Inhibitor, AT13387, in First-in-Human Phase I Dose Escalation Study of a

Updated version

10.1158/1078-0432.CCR-14-0979doi:

Access the most recent version of this article at:

Material

Supplementary

http://clincancerres.aacrjournals.org/content/suppl/2014/10/22/1078-0432.CCR-14-0979.DC1

Access the most recent supplemental material at:

Cited articles

http://clincancerres.aacrjournals.org/content/21/1/87.full#ref-list-1

This article cites 30 articles, 19 of which you can access for free at:

Citing articles

http://clincancerres.aacrjournals.org/content/21/1/87.full#related-urls

This article has been cited by 11 HighWire-hosted articles. Access the articles at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected]

To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://clincancerres.aacrjournals.org/content/21/1/87To request permission to re-use all or part of this article, use this link



http://clincancerres.aacrjournals.org/lookup/doi/10.1158/1078-0432.CCR-14-0979

http://clincancerres.aacrjournals.org/content/suppl/2014/10/22/1078-0432.CCR-14-0979.DC1

http://clincancerres.aacrjournals.org/content/21/1/87.full#ref-list-1

http://clincancerres.aacrjournals.org/content/21/1/87.full#related-urls

http://clincancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]

http://clincancerres.aacrjournals.org/content/21/1/87


First-in-Human Phase I Dose Escalation Study of a Second ... · First-in-Human Phase I Dose...

Documents

Transcript of First-in-Human Phase I Dose Escalation Study of a Second ... · First-in-Human Phase I Dose...