Praluzatamab Ravtansine, a CD166-Targeting Antibody-Drug Conjugate, in Patients with Advanced Solid Tumors: An Open-Label Phase I/II Trial

Abstract

Purpose: Praluzatamab ravtansine (CX-2009) is a conditionally activated Probody drug conjugate (PDC) comprising an anti-CD166 mAb conjugated to DM4, with a protease-cleavable linker and a peptide mask that limits target engagement in normal tissue and circulation. The tumor microenvironment is enriched for proteases capable of cleaving the linker, thereby releasing the mask, allowing for localized binding of CX-2009 to CD166. CX-2009 was evaluated in a phase I/II clinical trial for patients with advanced solid tumors.

Patients and methods: Eligible patients had metastatic cancer receiving ≥2 prior treatments. CX-2009 was administered at escalating doses every 3 weeks (0.25-10 mg/kg) or every 2 weeks (4-6 mg/kg). Primary objective was to determine the safety profile and recommended phase II dose (RP2D).

Results: Of 99 patients enrolled, the most prevalent subtype was breast cancer (n = 45). Median number of prior therapies was 5 (range, 1-19). Dose-limiting toxicities were observed at 8 mg/kg every 3 weeks and 6 mg/kg every 2 weeks. On the basis of tolerability, the RP2D was 7 mg/kg every 3 weeks. Tumor regressions were observed at doses ≥4 mg/kg. In the hormone receptor-positive/HER2-nonamplified breast cancer subset (n = 22), 2 patients (9%) had confirmed partial responses, and 10 patients (45%) had stable disease. Imaging with zirconium-labeled CX-2009 confirmed uptake in tumor lesions and shielding of major organs. Activated, unmasked CX-2009 was measurable in 18 of 22 posttreatment biopsies.

Conclusions: CD166 is a novel, ubiquitously expressed target. CX-2009 is the first conditionally activated antibody-drug conjugate to CD166 to demonstrate both translational and clinical activity in a variety of tumor types.

Figure 1.

[⁸⁹Zr]Zr-CX-2009 evaluation and comparison with mAbs evaluated in previous clinical studies (baseline). A, A total of 10 mg, 37 MBq of [⁸⁹Zr]Zr-CX-2009 was administered to a patient with micropapillary adenocarcinoma of the lung. B, Using Patlak analysis, irreversible uptake (Ki) of [⁸⁹Zr]Zr-CX-2009 in major organs was compared with a validated set of four mAbs. C, Tumor uptake was confirmed in this tumor type with low cellularity (images 162 hours after injection, top = (¹⁸F)FDG, bottom = [⁸⁹Zr]Zr-CX-2009, left images = PET/CT fusion, middle = CT scan in lung setting, right = PET). A tumor lesion in the left lower lung is indicated with a blue circle. Note: additional tumor tissue is present in the right lower lung. However, due to the close proximity of this lesion with the liver, uptake of [⁸⁹Zr]Zr-CX-2009 is not reliably distinguishable.

Figure 2.

A, Preliminary dose 1 intact CX-2009 (solid lines) and total CX-2009 (dashed lines) median plasma concentrations (nmol/L) versus time (days) following administration of up to 10 mg/kg CX-2009 every 3 weeks. B, Preliminary dose 1 Probody-conjugated-DM4 (Pc-DM4), intact CX-2009, total CX-2009, DM4-Me, and free DM4 median plasma concentrations (nmol/L) versus time (days) following administration of 7 mg/kg CX-2009 every 3 weeks. DM4, N-succinimidyl 4-(2-pyridyldithio) butanoate-N2′-deacetyl-N2′-(4-mercapto-4-methyl-1-oxopentyl)-maytansine; DM4-Me, S-methyl DM4; Pc, Probody-conjugated.

Figure 3.

Waterfall plot of change in tumor burden in evaluable patients who received CX-2009 ≥4 mg/kg. A, Breast cancer and CD166 expression. Numbers along the x-axis represent a composite CD166 IHC score (H-score) for each patient's archival/predose biopsy, with the highest potential score being 300 and the lowest being 0. See Supplementary IHC methods for a full description. Two of the 32 response-evaluable patients had new lesions, but no associated measurements to determine the percent change from baseline for the waterfall plot. cPR, confirmed partial response; uPR, unconfirmed partial reponse. B, HNSCC, head and neck squamous cell carcinoma; NSCLC, non–small cell lung cancer; EOC, epithelial ovarian carcinoma. C, Plot of tumor burden reduction by dose in patients with breast cancer treated with CX-2009 ≥ 4 mg/kg every 3 weeks. NE, not evaluable; PD, progressive disease; PR, partial response; SD, stable disease. *Patient started at 6 mg/kg every 2 weeks (Q2W), was dose reduced to 4 mg/kg Q2W due to keratitis. ^#Patient who, at the timepoint, had PD due to worsening nontarget lesion or presence of new lesions.

Figure 4.

Analysis of CX-2009 activation and CD166 levels in on-treatment biopsies. A, CX-2009 is activated in patient biopsies. On-treatment biopsies collected on C1D4 were analyzed by CEI to determine the concentration of activated CX-2009. Values below the lower limit of quantification (LLOQ) are plotted as 0.1. The CX-2009 dose for the on-treatment biopsy patient subset ranged from 4–10 mg/kg. For a given indication, a darker color indicates a higher dose. For biopsies from patients with breast cancer, samples from HR⁺/HER2⁻ patients are represented as circles, and the single sample from a HER2⁺ patient is represented as a square. eOC, epithelial ovarian cancer. B, CD166 target levels correlate with activated CX-2009 in patient biopsies. CD166 in patient biopsy lysates was measured by CEI; the peak area measured for this assay focused on the molecular weight range expected for the glycosylated/membrane-associated form of CD166. Note that the activated CX-2009 values shown in B are a subset of the samples shown in A. CD166 was measured in all 22 evaluable samples, but 6 of the 18 samples with quantifiable activated CX-2009 were below the LLOQ for CD166; therefore, the total number of samples in which both analytes were quantifiable was 12.