The PD-1- and LAG-3-targeting bispecific molecule tebotelimab in solid tumors and hematologic cancers: a phase 1 trial

Abstract

Tebotelimab, a bispecific PD-1×LAG-3 DART molecule that blocks both PD-1 and LAG-3, was investigated for clinical safety and activity in a phase 1 dose-escalation and cohort-expansion clinical trial in patients with solid tumors or hematologic malignancies and disease progression on previous treatment. Primary endpoints were safety and maximum tolerated dose of tebotelimab when administered as a single agent (n = 269) or in combination with the anti-HER2 antibody margetuximab (n = 84). Secondary endpoints included anti-tumor activity. In patients with advanced cancer treated with tebotelimab monotherapy, 68% (184/269) experienced treatment-related adverse events (TRAEs; 22% were grade ≥3). No maximum tolerated dose was defined; the recommended phase 2 dose (RP2D) was 600 mg once every 2 weeks. There were tumor decreases in 34% (59/172) of response-evaluable patients in the dose-escalation cohorts, with objective responses in multiple solid tumor types, including PD-1-refractory disease, and in LAG-3⁺ non-Hodgkin lymphomas, including CAR-T refractory disease. To enhance potential anti-tumor responses, we tested margetuximab plus tebotelimab. In patients with HER2⁺ tumors treated with tebotelimab plus margetuximab, 74% (62/84) had TRAEs (17% were grade ≥3). The RP2D was 600 mg once every 3 weeks. The confirmed objective response rate in these patients was 19% (14/72), including responses in patients typically not responsive to anti-HER2/anti-PD-1 combination therapy. ClinicalTrials.gov identifier: NCT03219268 .

Fig. 1. Clinical trial design and patient flow.

Cholangio, cholangiocarcinoma; IV, intravenous; pts, patients. ^aSingle-patient cohort.

Fig. 2. Safety profile of tebotelimab plus margetuximab in patients with HER2⁺ advanced solid tumors (n = 84).

a, Overall summary of safety and AESIs across all patients treated with combination therapy. b, Top 20 TRAEs versus AEs irrespective of causality by severity. ^aGrade 4 TRAEs include thrombocytopenia (n = 1) and ALT increased (n = 1). ^bIncludes MedDRA preferred terms of rash and maculopapular rash. ^cIncludes MedDRA preferred terms of pruritus and generalized pruritus. At each level of patient summarization, a patient was counted once, if the patient reported one or more events. AE, adverse event; ALT, alanine aminotransferase; AST, aspartate aminotransferase; SAE, serious adverse event. Data cutoff: 1 December 2021.

Extended Data Fig. 1. Tebotelimab: a bispecific PD-1xLAG-3 DART checkpoint inhibitor molecule and binding affinities.

a, Tebotelimab is composed of a two-chain protein structure with a molecular weight of 54.4 kDa and 28.9 kDa and yields an anticipated overall molecular weight of 166.7 kDa. b, Size-exclusion chromatography of tebotelimab (n = 1). c, SDS-PAGE of tebotelimab. d, SPR analysis of binding of soluble human PD-1 or LAG-3 to tebotelimab captured on Fab2 goat-anti-human Fc-coated surface. e, Tabular summary of binding kinetics of tebotelimab to soluble PD-1 and soluble LAG-3. f, g, Binding of tebotelimab to murine myeloma NS0-PD-1⁺ (f) and NS0-LAG-3⁺ (g) engineered cells, in the presence of titrating concentrations of tebotelimab or nivolumab^a or relatlimab^a, was assessed by FACS analysis.^aReplicas of nivolumab and relatlimab were generated by MacroGenics based on published sequences. APC, antigen-presenting cell; Fab, fragment antigen-binding; FACS, fluorescence-activated cell sorting; HPLC, high-performance liquid chromatography; kDa, kilodaltons; SDS-PAGE, sodium dodecyl-sulfate polyacrylamide gel electrophoresis; SPR, surface plasmon resonance. Source data

Extended Data Fig. 2. PK/PD and safety profile of tebotelimab monotherapy.

a, Top 20 TRAEs vs AEs irrespective of causality by severity (N = 269)^a. Includes MedDRA preferred terms of rash and maculopapular rash.^b Includes MedDRA preferred terms of pruritus and generalized pruritus. At each level of patient summarization, a patient is counted once if the patient reported one or more events. b, Serum concentration of tebotelimab over time after doses ranging from 1 to 1,200 mg Q2W (N = 46). The published serum C_trough for pembrolizumab at the dose of 2 mg/kg Q3W is 23.6 μg/mL. c, Peripheral blood flow cytometry analyses of receptor occupancy (PD-1 and/or LAG-3) in CD4 and CD8 T-cells before (pre-dose) and after tebotelimab monotherapy (end of infusion). Error bars represents SEM (n = 1 at 1 and 3 mg; n = 4 at 10, 30, and 120 mg; n = 8 at 400 mg; n = 13 at 800 mg; n = 5 at 1,200 mg). AE, adverse event; ALT, alanine aminotransferase; AST, aspartate aminotransferase; IRR, infusion-related reaction; PD, pharmacodynamics; SAE, serious adverse event; SEM, standard error of the mean; TRAEs, treatment-related adverse events. Data cutoff: 1 December 2021.

Extended Data Fig. 3. Best percentage reduction of target lesions with tebotelimab monotherapy in dose escalation (N = 43).

Evaluable patients (n = 44) include those who received at least one dose and had at least one post-baseline tumor evaluation. ^aOne patient had investigator’s overall assessment of PD at the end of treatment, but target lesions were not assessed; therefore, this patient, included in the table, is not shown in the waterfall plot. Data cutoff: 1 December 2021. ACC, adenoid cystic carcinoma; Cholangio, cholangiocarcinoma; CPI, immune checkpoint inhibitor; cPR, confirmed partial response; CRC, colorectal carcinoma; EOC, epithelial ovarian cancer; GC, gastric cancer; GEJ, gastroesophageal junction cancer; HCC, hepatocellular carcinoma; MEC, mucoepidermoid carcinoma; Meso, mesothelioma; NE, not evaluable; NET, neuroendocrine tumor; NSCLC, non-small cell lung cancer; ORR, objective response rate; PD, progressive disease; RCC, renal cell carcinoma; SCC, squamous cell carcinoma; SCCHN, squamous cell carcinoma of the head and neck; SD, stable disease; STS, soft-tissue sarcoma; TNBC, triple-negative breast cancer.

Extended Data Fig. 4. Antitumor activity of tebotelimab monotherapy in the cohort expansion of EOC, TNBC, and NSCLC (N = 90).

a, Best percentage change from baseline in the size of target lesion in patients with EOC (n = 34), TNBC (n = 29), and NSCLC (n = 27). ^aFor 3 of the 36 evaluable EOC patients, no percentage change value was available because target lesions could not be assessed (n = 2; 1 PD, 1 NE) or because no investigator’s RECIST assessment was entered at the time of data cutoff (n = 1; ‘missing’); therefore, 3 EOC patients are not included in the waterfall plot and the spider plot in panel B, but are included in the table and bar charts in Extended Data Fig. 5a. ^bFor 2 of the 31 evaluable TNBC patients, no percentage change value was available because target lesions could not be assessed (n = 1; NE) or because target lesions assessment was not done (n = 1; SD); therefore, 2 TNBC patients are not included in the waterfall plot and the spider plot in panel B, but are included in the table and bar charts in Extended Data Fig. 5a. ^cFor 2 of the 29 evaluable NSCLC patients, no percentage change value was available because target lesions could not be assessed (both PD); therefore, 2 NSCLC patients are not included in the waterfall plot and the spider plot in panel B, but are included in the table and bar charts in Extended Data Fig. 5a. b, Spider plot depicting percentage change from baseline in the size of target lesion over time in patients with EOC, TNBC, and NSCLC (as shown in panel A). cCR, confirmed complete response; CPI, immune checkpoint inhibitor; cPR, confirmed partial response; EOC, epithelial ovarian cancer; NE, not evaluable; NSCLC, non-small cell lung cancer; PD, progressive disease; RECIST, Response Evaluation Criteria in Solid Tumors; SD, stable disease; TNBC, triple-negative breast cancer; uCR, unconfirmed complete response; uPR, unconfirmed partial response.

Extended Data Fig. 5. Antitumor activity of tebotelimab monotherapy in the response evaluable population and correlation analyses of outcome by LAG-3, PD-1/PD-L1, and IFN-γ.

a, Summary of clinical activity in the 96 patients with EOC, TNBC, and NSCLC, as well as in all 167 monotherapy cohort expansion patients evaluable for efficacy by RECIST. Evaluable patients include those who received at least one dose and had baseline and at least one post-baseline tumor evaluation. Data cutoff: 1 December 2021. ^aThe 14 patients with diffuse large B-cell lymphoma are not included because they were evaluated using the Revised International Working Group criteria (that is, the Lugano classification). b, The NanoString PanCancer IO 360™ assay was used to analyze gene expression, including the abundance of 14 immune cell types and 32 immuno-oncology signatures from 77 archival biopsies from EOC (n = 28), NSCLC (n = 20), and TNBC (n = 29) expansion cohort patients. Left panel: Pearson coefficient of 0.43 for association between PD-1 and LAG-3 expression. P < 0.05 for the correlation between LAG-3 expression and objective responses. P = 0.25 for the correlation between PD-1 expression and objective responses. Center panel: P < 0.003 for the correlation between proteasome signature and objective responses. Right panel: pP = 0.05 for the correlation between IFN-γ gene signature and objective responses. P-values (2-sided) per Wilcoxon rank sum between CR/PR vs. PD/SD. Box-plot elements: center line, median; box limits, upper and lower quartiles; whiskers, 1.5× interquartile range; points, outliers. cCR, confirmed complete response; CPI, immune checkpoint inhibitor; cPR, confirmed partial response; EOC, epithelial ovarian cancer; NE, not evaluable; NSCLC, non-small cell lung cancer; PD, progressive disease; RECIST, Response Evaluation Criteria in Solid Tumors; SD, stable disease; TNBC, triple-negative breast cancer; uCR, unconfirmed complete response; uPR, unconfirmed partial response.

Extended Data Fig. 6. Antitumor activity of tebotelimab monotherapy in the cohort expansion of DLBCL (N = 14), association of baseline LAG-3/PD-L1 expression with clinical response, and a case of complete response.

a, Swimmer plot depicting time on study and achievement of objective responses, in 14 DLBCL patients evaluable for efficacy (per the Lugano classification). Evaluable patients include those who received at least one dose and had baseline and at least one post-baseline tumor evaluation. The asterisks indicate two post–CAR-T patients who discontinued tebotelimab treatment to pursue allo-SCT after achieving CR; both patients remain in remission approximately 28 and 12 months post allo-SCT, respectively. Data cutoff: 1 December 2021. b, Best percentage change from baseline in the size of target lesion in 12 patients with DLBCL. ^aFor 2 of the 14 evaluable DLBCL patients, no percentage change value was available because target lesion assessment was not done (PD) or because target lesions could not be assessed (PR); therefore, 2 DLBCL patients, both with ‘other/unknown subtype’, are not included in the waterfall plot but are included in the table (C). Data cutoff: December 1, 2021. c, Summary of clinical activity in 14 DLBCL patients evaluable for efficacy (per the Lugano classification). d, Retrospective IHC analyses on pretreatment biopsies available from 11 DLBCL patients to assess the correlation of LAG-3 protein expression at baseline to objective response. LAG-3 score was determined by calculating mean value of LAG-3⁺ cells per 40× field across 5 LAG-3⁺ hot spot fields (n = 1 per patient). Box-plot elements: center line, median; box limits, upper (75%) and lower (25%) quartiles; upper and lower whiskers: the largest and the smallest data points. e, A complete response observed in a post-CAR T patient after single tebotelimab administration. f, Immunofluorescence of a tumor sample (n = 1) from the patient shown in panel (e), stained for CD3 (with antibody 2GV6), CD79a (with antibody SP18), PD-1 (with antibody NAT105), LAG-3 (with antibody EPR4392(2)), and DAPI. g, Venn diagram representing cell count from the post-CAR T (pre-tebotelimab) tumor sample shown in panel (f), showing the profile of LAG-3 and PD-1 expression on CD3 (a marker of TILs), and CD79a (a marker of malignant B cells) populations. allo-SCT, allogeneic stem cell transplantation; CAR T, chimeric antigen receptor T-cell; CPI, immune checkpoint inhibitor; CR, complete response; DAPI, 4’,6-diamidino-2-phenylindole; DLBCL, diffuse large B-cell lymphoma; GCB, germinal center B-cell; HGBL-DH (MYC/BCL2), high-grade B-cell lymphoma double hit, with MYC and BCL2 rearrangements; IFN-γ, interferon-gamma; PD, progressive disease; PR, partial response; SD, stable disease; TILs, tumor-infiltrating lymphocytes. Source data

Extended Data Fig. 7. Best percentage change from baseline in the size of target lesion in patients with HER2⁺ advanced solid tumors treated with tebotelimab plus margetuximab evaluable for response (N = 71).

Waterfall plot depicting best percentage change from baseline in the size of target lesion in 71 patients with various tumor types evaluable for efficacy. ^aFor 1 of the 33 evaluable breast cancer patients, no percentage change value was available because target lesions could not be assessed (PD); therefore, 1 breast cancer patient is not included in the waterfall plot. Data cutoff: 1 December 2021. PD-L1 CPS was calculated as follows: number of PD-L1+ cells (tumor cells, lymphocytes, and macrophages)/total number of viable tumor cells × 100. ACC, adenoid cystic carcinoma; cCR, confirmed complete response; CPI, immune checkpoint inhibitor (includes PD-1 inhibitors only); cPR, confirmed partial response; CPS, combined positive score; HER2, human epidermal growth factor receptor 2; NA, not available; PD, progressive disease; SD, stable disease; uCR, unconfirmed complete response; uPR, unconfirmed partial response.

Extended Data Fig. 8. Antitumor activity of the combination of tebotelimab and margetuximab in patients with HER2⁺ advanced solid tumors.

a, Summary of clinical activity in 72 patients evaluable for efficacy, by tumor type. Evaluable patients include those who received at least one dose and had baseline and at least one post-baseline tumor evaluation. Data cutoff: 1 December, 2021. b, Waterfall plots depict best percentage change from baseline in the size of target lesion in 21 patients who did not receive prior anti-HER2 therapy (top) and 50 patients who received prior anti-HER2 therapy (bottom). c, Waterfall plots depict best percentage change from baseline in the size of a target lesion in 59 patients who did not receive prior CPI therapy (top) and 12 patients who received prior CPI therapy (bottom). ^aFor 1 of the 33 evaluable patients with breast cancer, no percentage change value was available because target lesions could not be assessed (PD); therefore, 1 patient with breast cancer is not included in the waterfall plots shown in (b). ^bIncludes colorectal and esophageal cancers. ^cIncludes GEJ and esophageal cancers. ^dIncludes colorectal, GEJ, and esophageal cancers. ^eIncludes GEJ and esophageal cancers. Data cutoff: 1 December 2021. PD-L1 CPS was calculated as follows: number of PD-L1+ cells (tumor cells, lymphocytes, and macrophages)/total number of viable tumor cells × 100. cCR, confirmed complete response; Cholangio, cholangiocarcinoma; cPR, confirmed partial response; CPI, immune checkpoint inhibitor (includes PD-1 inhibitors only); CPS, combined positive score; CRC, colorectal cancer; EOC, epithelial ovarian cancer; GEJ, gastroesophageal junction cancer; HER2, human epidermal growth factor receptor 2; NA, not available; NSCLC, non-small cell lung cancer; ORR, objective response rate; PD, progressive disease; SCCHN, squamous cell carcinoma of the head and neck; SD, stable disease; TNBC, triple-negative breast cancer.

Extended Data Fig. 9. Time on study and treatment outcomes in patients with HER2⁺ advanced solid tumors treated with tebotelimab plus margetuximab evaluable for response (N = 72).

Swimmer plot depicting time on study and achievement of objective responses. ACC, adenoid cystic carcinoma.