Results and perspectives from Phase 1 studies assessing the safety and clinical activity of multiple doses of a NKG2D-based CAR-T therapy, CYAD-01, in metastatic solid tumors

CYAD-01 BACKGROUND

Alain Hendlisz 1, Sylvie Rottey 2, Mateusz Opyrchal 3, Kunle Odunsi 3, Jean-Pascal H. Machiels 4, Solmaz Sahebjam 5, Leila Shaza 1, Sandrine Aspeslagh 1, Ahmad Awada 1, Marc Van Den Eynde 4,Jean-Luc Canon 6, Javier Carrasco 6, Eytan Breman 7, Panagiota A. Sotiropoulou 7, Sarah Snykers 7, Nathalie Braun 7, Caroline Lonez 7, Anne Flament 7, Bikash Verma 8, and Frederic F. Lehmann 7.

1. Institut Jules Bordet, Université Libre de Bruxelles, Brussels, Belgium; 2. Ghent University Hospital, Ghent, Belgium; 3. Roswell Park Comprehensive Cancer Center, Buffalo, NY; 4. Cliniques Universitaires Saint-Luc, Université Catholique de Louvain, Brussels, Belgium; 5. Moffitt Cancer Center, Tampa, FL; 6. Service d’Oncologie-Hématologie, Site Notre-Dame, Grand Hôpital de Charleroi (GHdC), Charleroi, Belgium;

7. Celyad, Mont-Saint-Guibert, Belgium; 8. Celyad, New York, NY

● Chimeric antigen receptor (CAR) T-cell therapieshave yet to demonstrate positive results in thecontext of solid tumors likely because of the inabilityof classical CAR-Ts to infiltrate into the tumor(dense tumor bed and bulky stroma, hypoxic, lowpH, and low nutrient conditions) and overcome thehostile immune suppressive tumor micro-environment (TME).

● CYAD-01 consists of engineered T cells expressinga CAR based on the natural killer group 2, member Dreceptor (NKG2D), a transmembrane receptorexpressed by natural killer cells and some T-cellsubsets.

NKG2D binds to 8 ligands frequently expressed onvarious tumor types [1,2]: MHC class I chain–relatedproteins A (MICA) and B (MICB) and Unique long 16binding proteins (ULBP) 1–6 ligands.

Preclinically, CYAD-01 have anti-tumor effectsbeyond direct cancer cell killing [3] by:o Targeting neovasculature expressing NKG2D

ligands,o Targeting immunosuppressive cells such as

regulatory T cells and myeloid-derivedsuppressor cells expressing NKG2D ligands,

o Recruiting and activating macrophages andmyeloid cells within the tumor stroma, shiftingto an immunostimulatory TME, and

o Inducing a long-term memory immune responsespecific towards tumor antigens.

CYAD-01 DEVELOPMENT

● CYAD-01 is currently evaluated in hematologic andsolid cancer indications in a comprehensive clinicaldevelopment plan (Figure 1). CYAD-01 showedpromising early clinical activity inrelapsed/refractory AML [4] and is also evaluated insolid tumor indications in different settings.

● The THINK (NCT03018405) study – solid tumor arm- aims at evaluating the safety and clinical activity ofCYAD-01 therapy administered as a standalonetherapy in patients with different relapsed/refractory solid tumor indications.

● The THINK CyFlu cohort (NCT03018405) aims atevaluating the safety and clinical activity of CYAD-01 therapy administered after a preconditioningregimen (cyclophosphamide and fludarabine) inpatients with relapsed/refractory mCRC. Thiscorresponds to the schedule used with classicalCAR T-cell therapies and aims at providing a betterCYAD-01 engraftment due to the lymphodepletionof host cells.

● The SHRINK (NCT03310008) study has beendesigned to address the challenges specific to solidtumors and related to the immunosuppressive TMEby evaluating CYAD-01 therapy concurrentlyadministered with the FOLFOX standard of care(SoC) chemotherapy regimen in patients withmCRC. Concurrent administration of SoC aims atproviding a better CYAD-01 engraftment due to thelymphodepletion induced by the FOLFOX andimproving disease control prior to CYAD-01infusion. The study is conducted in mCRC patientseither with potentially resectable liver metastasis(neoadjuvant setting) or with recurrent/progressingdisease after at least one line of systemic therapyfor metastatic disease.

CYAD-01

Hematological malignanciesSolid tumors

Monotherapy THINKNCT03018405

THINK NCT03018405

EPITHINKNCT03612739

SHRINKNCT03310008

Concurrent standardchemotherapy

DEPLETHINKNCT03466320

THINK CyFluNCT03018405

Preconditioningchemotherapy

LINKNCT03370198

Loco-regional administration

FIGURE 1: CYAD-01 clinical development

THINKwithout preconditioning

THINK CYFLUwith prior preconditioning

SHRINKwith concurrent FOLFOX

STUDY DESIGN

• Open-label Phase I study, with distinct dose-escalation segments in hematological and solid cancer indications,• No bridging therapy, no preconditioning,• 3+3 design dose escalation with 3 dose-levels of CYAD-01 (3×108, 1×109, and 3×109 cells/injection),• 3 doses of CYAD-01 every 2 weeks. Potential 3 additional doses of CYAD-01 at 1x109 cells/injection according

the clinical status at first tumor assessment.

• 1 cohort of 3 patients (3+3 design),• Preconditioning regimen (300 mg/m² cyclophosphamide

and fludarabine 30 mg/m² daily) at Day -5, -4 and -3 priorto a single CYAD-01 at 3×108 cells/injection,

• Potential 3 additional doses of CYAD-01 withoutpreconditioning regimen at 3x109 cells/injection.

• 3+3 design dose escalation with 3 dose-levels ofCYAD-01 (1×108, 3×108, and 1×109 cells/injection),

• 6 cycles of FOLFOX every 2 weeks,• 3 doses of CYAD-01 every 2 weeks administered 48 hours

after the end of a chemotherapy cycle,• Potential 3 additional doses of CYAD-01 at the same dose.

DEMOGRAPHICSDL-1 DL-2 DL-3 DL-13x108 1x109 3x109 3x108 1x108

N=4 N=4 N=6 N=2 N=3Age: Mean/Range (years) 56.8 / 37-72 66.0 / 57-74 52.7 / 44-63 58.0 / 51-65 54.8 / 53-57Gender: Male/Female 3 / 1 1 / 3 4 / 2 1 / 1 0 / 3Tumor type

Colorectal cancer 3 2 6 2 3Ovarian cancer 0 2 0 0 0

Pancreatic cancer 1 0 0 0 0Metastatic patients: Relapsed/Refractory 0 / 4 1 / 3 1 / 5 2 / 0 0 / 0 (neoadjuvant setting)No. of prior therapy lines: Mean/Range 4 / 3-7 5.5 / 3-8 3.8 / 2-6 2.5 / 2-5 0 (first line)ECOG performance score: Grade 0/1 3 / 1 2 / 2 0 / 6 0 / 2 3 / 0Best overall response

Complete response 0 0 0

ongoing

1 pCR (HRCM showing TRG1)°Partial response 0 0 0 2 pPR (HRCM showing TRG3)°

Stable disease 1 CRC* 1 OVA* 2 CRC* 0Progressive disease 3 3 4 0

* Confirmed by RECIST 1.1 criteria ° Confirmed by Rubbia-Brandt pathological response score system. HCRM = Hepatic Colorectal Metastases; TRG = Tumor regression grade

Grade 3/4 RELATED AEs3x108 1x109 3x109 3x108 1x108

N=4 N=4 N=6 N=2 N=3Grade 3 Grade 4 Grade 3 Grade 4 Grade 3 Grade 4 Grade 3 Grade 4 Grade 3 Grade 4

Adverse Event (AE) Preferred TermTotal pts with at least ≥1 related AE (%) 1 ( 25.0) 0 1 (25.0) 0 2 ( 33.3) 1 ( 16.7) 0 0 1 (33) 0Cytokine release syndrome (CRS) 0 0 1 ( 25.0) 0 0 1 ( 16.7) 0 0 0 0Chills 1 ( 25.0) 0 0 0 0 0 0 0 0 0Dyspnoea 0 0 0 0 1 ( 16.7) 0 0 0 0 0Acute respiratory distress syndrome 0 0 0 0 0 1 ( 16.7) 0 0 0 0Anaemia 0 0 0 0 0 0 0 0 1 (33) 0Alanine aminotransferase increased 0 0 0 0 1 ( 16.7) 0 0 0 0 0Lymphocyte count decreased 0 0 0 0 0 1 ( 16.7) 0 0 0 0

MAIN RESULTS

• 14 patients with solid tumor indications have been enrolled at three different dose-levels.• CYAD-01 as stand alone was overall well tolerated. Only 9 treatment-related grade 3/4 adverse events were

observed in 5 different patients (uncleaned database). The grade 4 CRS in DL-3 was considered as a DLT. The 5other patients recruited at the same dose (DL-3) showed no further evidence of severe toxicity.

• Four patients showed disease stabilization, including 3/11 (27%) mCRC patients and 2/6 (33%) mCRC at DL-3,all for a minimum of 3 months and confirmed according RECIST 1.1. (Figure 5).

• Detection of peripheral CYAD-01 cells by PCR method seem to correlate with dose level (Figure 4) and clinicalbenefit (Figure 3).

• Increase in cytokine release (e.g. IFN-γ and MCP-1) in the peripheral blood post CYAD-01 injections weredetected by luminex assay, mainly at DL-3 (Figure 2).

• Two patients completed their treatment schedule as oftoday.

• CYAD-01 post standard Cy/Flu preconditioning was welltolerated. No DLT nor SAE occurrence (uncleaneddatabase).

• The 2 enrolled patients were not yet evaluable for clinicalresponse.

• Increased CYAD-01 cell expansion after a single injectionwith the Cy/Flu preconditioning (Figure 6).

• The first DL (1x108/injection) is completed.• CYAD-01 post FOLFOX as SOC was well tolerated. No

DLT nor SAE occurrence (uncleaned database).• All patients underwent resection without delays in surgery.• The activity endpoint assessed by pathological response

criteria is promising with 1 major/complete response and 2partial responses.

• At DL-1, the peripheral CYAD-01 cell expansion is similarto the one observed with the standalone therapy but doesnot seem to correlate with the clinical benefit (Figure 7).

FIGURES

FIGURE 4: Peaks of the CYAD-01 versus the clinical response

Clinical benefit (SD)Progressive disease (PD)

FIGURE 3: Kinetics of the CYAD-01

FIGURE 5: Overall clinical responses

FIGURE 2: Cytokine release post CYAD-01 administration

REFERENCES

[1] Nausch N, Cerwenka A. Oncogene 2008;27:5944-58.[2] Lanier L. Cancer Immunol Res 2015;3(6):575-82.

[3] Lonez C, et al. BMJ Open 2017;7:e017075.[4] Sallman DA, et al. Haematologica. 2018;103(9):e424-e426

• CYAD-01 has shown promising early clinical activity in relapsed/refractory AML [4].

• In solid cancer indications, the safety profile of CYAD-01 as stand-alone treatment seemsfavorable. The addition of Cy/Flu or FOLFOX, while still early, does not seem to increase the related-AEs rate.

• In terms of clinical activity, the THINK trial demonstrated encouraging disease stabilization inrelapsed / refractory metastatic CRC and ovarian cancer patients. Some interesting preliminarycorrelation can be done between the clinical benefit and the expansion of peripheral CYAD-01 cells.

• Cell kinetics data in the THINK Cy/Flu cohort are encouraging as we await clinical data to explore the correlation between clinical activity and cell expansion.

CONCLUSIONS

• Early data in the neoadjuvant setting (SHRINK trial) with the 3 patients showing a pathologicalobjective response are encouraging and will be further explored in the next dose-levels.

THINK DL1 THINK DL3

CYAD-01 treatment CYAD-01 treatment

Cy/Flu treatment

CYAD-01 treatment

FOLFOX treatment

FIGURE 6: Kinetics of the CYAD-01 FIGURE 7: Kinetics of the CYAD-01