Pre-clinical characterization of potent and selective next-generation RET inhibitors

Presented at: AACR Annual Meeting 2021 Date: April 10, 2021 Pre-clinical characterization of potent and selective next-generation RET inhibitors Gabrielle R. Kolakowski, Erin D. Anderson, Joshua A. Ballard, Barbara J. Brandhuber, Kevin R. Condroski, Eliana B. Gomez, Thomas C. Irvin, Manoj Kumar, Nisha A. Patel, Faith D. Watson, Steven W. Andrews # 1464 Loxo Oncology at Lilly, Boulder, CO

Table 2. Next-generation RET inhibitors are very potent Fig 1. A cancer cell line screen demonstrates potency Fig 3. LOX-18228 and LOX-19260 cause complete tumor Background against RET alterations in cell-based assays and selectivity of next-generation RET inhibitors regression in RET-altered PDX models

>10µM GI (µM) CCDC6-RET colorectal cancer PDX tumors

• In May 2020, selpercatinib became the first FDA- LOX-18228 LOX-19260 Less sensitive 50 More sensitive ) GI50 not )

IC (nM) IC (nM) 3 50 50 3 2500 3000

10 1 0.1 0.01 0.001 )

approved selective RET inhibitor, indicated for determined 10 1 0.1 0.01 0.001 ) 3

Founder alterations 3 m

m 2500

2000

m

m (

patients (pts) with RET fusion-positive NSCLC and (

CCDC6-RET 4 4 LOLOX-X1-812822828 e

LOX-18228 e 2000 m

KIF5B-RET 1 1 m 1500 u

thyroid cancer as well as RET-mutant medullary u

)

l )

l 1500

3

3 o

RET M918T 4 6 2500 o 3000 V

1,2 V

m m 1000 Vehicle for LOX-18228 Vehicle for LOX-19260 LOX-19260 r

thyroid cancer. LOX-19260LOX-19260 r 1000

o o

Acquired resistance mutations m

m LOX-18228 10mg/kg LOX-19260 10mg/kg

(

(

m

m

2500 u

2000 u 500 LOX-18228 60mg/kg LOX-19260 60mg/kg e

e 500

KIF5B-RET G810S 8 11 T

Mutations at the solvent front of the ATP pocket T m • m KIF5B-RET V804L 56 21 SeSlepleprecractaintiinbib

Selpercatinib u

u 0 2000 0 l

l 0 6 12 18 24 30 36 o KIF5B-RET V804M 71 34 o 1500 0 4 8 12 16 20 24 28

have been identified as a mechanism of acquired (mm Volume Tumor

Tumor Volume (mm Volume Tumor

V

V

1500

r Days KIF5B-RET G810S+V804M 35 36 TPC-1 (CCDC6r -RET) TreatmentDays Days Treatment Days

VaVnadnedteintiinbib o resistance to RET inhibitors selpercatinib and o

RET M918T+G810S 6 8 Vandetanib 1000 m

LC-2/ad (CCDC6m -RET) 1000 u

pralsetinib, due to a steric clash created at the RET u CCDC6-RET G810S+V804M colorectal cancer PDX tumors

T

T

RET-mutant inducible HEK-293 cell lines were generated and

CCabaobzoaznatnintiinbib MZ-CRC-1 (RET50 M918T)0

n

n )

Cabozantinib ) 500

3 3 a

G810 position and result in a loss of binding treated with LOX-18228 or LOX-19260 for 1 hr. Phospho-RET a 2500 2500

e

e )

TT (RET C634W) ) 3

3,4 3

M

M m

potency. levels were measured using an In Cell Western protocol, and 0 2000 0 m 2000 m

Non-RET-altered m

(

(

0 4 8 12 16 20 24 28 0 6 12 18 24 30 36 e

IC50 values were calculated using a 4-parameter fit. e

)

m m

1500 ) 1500

3

3 u

• Maintaining potency against RET V804 mutations is u l

2500 Days 2500 Dl ays o

LOX-19260 o m

LOLOXX-00-01822818228 LOX-00019260 Vehm icle for LOX-18228 Vehicle for LOX-19260 V

140 140 The potency and selectivity of LOX-18228 and LOX- V

• 1000 1000

r

r m

important to prevent emergence of resistance and LOXm -18228 60mg/kg LOX-19260 30mg/kg

(

( o

120 120 o

LOX -18228 90mg/kg LOX-19260 60mg/kg m 2000 2000 m

19260 are comparable to the selective RET inhibitor,e

e u 100 500 u 500

treat patients with concurrent RET V804 gatekeeper 100 LOX-18228 120mg/kg LOX-19260 90mg/kg

T

T

m

m O

80 O 80 u

selpercatinib. u

l

S S

0 l 0

M M

o 1500

and G810 solvent front mutations. 60 60 o 1500

D D

V 0 4 8 12 16 20 24 28 0 4 8 12 16 20 24 28

V

Tumor Volume (mm Volume Tumor (mm Volume Tumor

%

% 40 40 Multi-kinase inhibitors, vandetanib and cabozantinib, are r

• r o Treatment Dayso Treatment Days • Therefore, a next-generation RET inhibitor should 20 20 1000 Days 1000 Days

examples of nonselective RET inhibitors. m

0 0 m

u

u T

-20 -20 T

maintain potency against both solvent front and • LOX-18228 and LOX -19260 were dosed twice daily. Tumor -2 -1 0 1 2 3 4 -2 -1 0 1 2 3 4 Top dose tested in this assay was 10 µM and all cell5 00 n 500

• n a

Log nM Log nM volume data are meana ± SEM (n ≥ 6/group). e gatekeeper resistance mutations individually and e

lines without GI50 values are assigned to the “>10 µM M when co-expressed. CCDC6-RET RET M918T KIF5B-RET V804L KIF5B-RET G810S+V804M 0 M 0 KIF5B-RET KIF5B-RET G810S KIF5B-RET V804M RET M918T+G810S GI50 not determined” portion of the graph. 0 4 8• 1LOX2 16-2182280 24 2 8and LOX-192600 4 were8 12 well16 20-tolerated24 28 (not shown). Table 1. RET G810 solvent front mutations confer Table 3. LOX-18228 and LOX-19260 are highly selective Fig 2. LOX-18228 is well-tolerated and causes complete Days Days resistance to approved RET inhibitors for RET tumor regression in a CCDC6-RET G810S PDX model Conclusions

CCDC6-RET G810S colorectal cancer PDX tumors Selpercatinib Pralsetinib LOX-18228 LOX-19260 • We have identified potent and selective next- IC50 (nM) IC50 (nM) Fold Fold )

% generation RET inhibitors to address the 2000 ( 30

Founder alterations Cell IC50 selectivity Cell IC50 selectivity

)

)

e

3 3 (nM) to WT KIF5B- (nM) to WT KIF5B- m emerging unmet need of patients who relapse on CCDC6-RET 10 5 g

m 20 n

RET RET (

KIF5B-RET 3 3 1500 a selective RET inhibitors. e FGFR1 1608 1780 x 1032 940 x h

RET M918T 9 1 10

C m

Flt3 >10000 >11000 x 280 250 x

t u

Acquired resistance mutations l KDR 684 760 x >10000 >9000 x h • LOX-18228 and LOX-19260 represent promising

o 1000 0

g i

KIF5B-RET G810S 53 81 KIT 376 410 x 772 700 x V

e next-generation RET inhibitor candidates that KIF5B-RET V804L 5 2 r -10

PDGFRb 468 520 x 1467 1300 x o W

could be used to further extend durable disease y

KIF5B-RET V804M 7 10 TrkC 384 420 x 43 40 x m 500 d

KIF5B-RET G810S+V804M 434 370 u -20 o

T control for patients with RET-altered cancers Tumor Volume (mm Volume Tumor RET M918T+G810S 39 141 B -30 0 (%) Change Weight Body following the development of acquired resistance • LOX-18228 and LOX-19260 were highly selective 0 4 8 12 16 20 24 28 0 4 8 12 16 20 24 28 Treatment Days Treatment Days to current agents. against a panel of 374 WT kinases (Reaction Days Days Biology Corp) and cell-based kinase assays Vehicle Vfoerh LicOleX f-o1r8 L2O28X-18228 LOX-18L2O28X -6108m22g8/k g60mg/kg • An IND is planned for 2021. (performed at Eurofins or ProQinase). LOX-18L2O28X -3108m22g8/k g30mg/kg LOX-18L2O28X -110802m28g /1kg00mg/kg • No concerning findings in Cerep Safety Screen 87 References panel at 10 µM compound concentration. • LOX-18228 was dosed twice daily. Tumor volume data are mean ± SEM. Body weight data are mean % body weight 1. Drilon A, et al. N Engl J Med. 2020;383:813-24. • hERG IC50 >10 µM. Crystal structure of Model of G810S solvent front change ± SD (n=8/group). 2. Wirth LJ, et al. N Engl J Med. 2020;383:825-35.

selpercatinib (gray) bound to mutation (in red) shows Ser810 3. Solomon BJ, et al. J Thorac Oncol. 2020; ) ) The efficacy study to test LOX-19260 in the CCDC6-RET

RET V804M (V804M 3 close to the space where 200•3 02000 15:541-49.

gatekeeper residue shown in selpercatinib binds, consistent G810S PDX model is pending. m m 4. Lin JJ, et al. Ann Oncol. 2020;31:1725-33.

green). Selpercatinib is a with a potency loss in KIF5B- m

m All PDX studies were performed at Crown Biosciences. (

• (

potent inhibitor of RET V804M. RET G810S+V804M. e

AACR Annual Meeting, Virtual, April 10-15, 2021 Presenter: Gabrielle R. Kolakowski,1 [email protected] 1500 Sponsored by Loxo Oncology at Lilly

m

m

u

u

l

l

o

o

V

V

1000 1000

r

r

o

o

m

m

u

u T

500T 500

n

n

a

a

e

e M 0M 0 0 40 841281612201624202824 28 DaysDays