Anti-tubulin drugs conjugated to anti-ErbB antibodies selectively radiosensitize

Authors

Stephen R. Adams

Howard C. Yang

Elamprakash N. Savariar

Joe Aguilera

Jessica L. Crisp

Karra A. Jones

Michael A. Whitney

Scott M. Lippman

Ezra E. W. Cohen

Roger Y. Tsien

Sunil J. Advani

Doi

10.1038/ncomms13019

PMID: 27698471 · DOI: 10.1038/ncomms13019 · Journal: Nature Communications (2016)

TL;DR

Adams et al. show that conjugating potent anti-tubulin warheads (auristatins such as MMAE; maytansinoids such as mertansine/DM1) to anti-ErbB antibodies (cetuximab targeting EGFR; trastuzumab targeting ERBB2) restricts radiosensitization to receptor-expressing tumours. Free MMAE and free mertansine radiosensitize cells indiscriminately, but the ADCs C-MMAE, T-MMAE, and the clinically approved ado-trastuzumab emtansine (T-DM1, Kadcyla) only radiosensitize cells that express the cognate surface receptor. In HER2+ xenografts (OE19 esophageal; NCI-N87 gastric), a single 0.25 nmol (~1.9 mg/kg) dose of T-DM1 combined with 2.5 Gy × 3 daily fractions produced long-term tumour control; HER2− HCT116 tumours showed no benefit. The authors propose ADC-directed delivery of highly potent radiosensitizers as an alternative to ErbB signal inhibition for chemo-radiotherapy.

Cohort & data

  • Study type: Preclinical mechanistic and efficacy study in human cancer cell lines and mouse tumour xenografts — no patient cohort.
  • Cell lines: CAL-27 and SCC-25 (HNSC, EGFR+); SCC-35, SCC-61, SQ-9G (HNSC, EGFR+, provided by Weichselbaum, U. Chicago); A549 and CALU3 (NSCLC); OE19 (ESCA, HER2+); BT474 (BRCA, HER2+); NCI-N87 (STAD, HER2+); HCT-116 (COAD, HER2−); LN229 (GB, EGFR−).
  • Xenograft models: Athymic nude mice bearing CAL-27 (EGFR+), OE19 (HER2+), NCI-N87 (HER2+), HCT-116 (HER2−), SCC-61 and LN229 flank / thigh tumours.
  • Agents synthesized: Cy5-labelled C-MMAE (cetuximab–MC-VC-MMAE, drug-to-antibody ratio ~3.7) and T-MMAE (trastuzumab–MC-VC-MMAE, DAR ~3.2); clinical-grade T-DM1, cetuximab, trastuzumab, erlotinib, lapatinib, cisplatin, and paclitaxel obtained from UCSD Moores Cancer Center pharmacy; free MMAE (Concortis); mertansine (Abcam).
  • Readouts: clonogenic survival (SF2 at 2 Gy), neutral comet assay (DNA DSBs at 6 Gy or 2 Gy), propidium-iodide flow cytometry (G2/M fraction), confocal fluorescence, in vivo Cy5 fluorescence imaging, IHC for pS10 Histone H3 and cleaved caspase-3, and tumour volume / survival curves.

Key findings

  • Receptor-restricted binding. Cy5-labelled C-MMAE bound EGFR+ CAL-27, A549, and CALU3 cells in a dose-dependent manner but not HER2+ OE19; T-MMAE bound HER2+ CALU3, OE19, and BT474 but not EGFR+ CAL-27 (PMID:27698471 Fig. 1b,c).
  • Receptor-restricted G2/M arrest. At 2–5 nM overnight, C-MMAE arrested CAL-27 (EGFR+) in G2/M while T-MMAE arrested OE19 (HER2+) in G2/M; antibody-only controls and the non-cognate ADC had no cell-cycle effect (Fig. 1e).
  • Free anti-tubulins radiosensitize indiscriminately; ADCs do not. Free MMAE (2 nM) increased comet-tail length after 6 Gy in both CAL-27 (EGFR+) and LN229 (EGFR−); cetuximab and C-MMAE radiosensitized CAL-27 but not LN229 (Fig. 2d). In CALU3 (EGFR+, HER2+) both C-MMAE and T-MMAE radiosensitized; in OE19 (HER2+) only T-MMAE and free MMAE — not C-MMAE — radiosensitized (Fig. 2f).
  • Mertansine and T-DM1. IC50 of free mertansine was ~10 nM across lines regardless of HER2 status, while T-DM1 IC50 was <1 nM in HER2+ OE19 and NCI-N87 but >100 nM in HER2− HCT116 and CAL-27 (Fig. 3b). In HER2+ OE19 cells, T-DM1 was more potent than paclitaxel, cisplatin, trastuzumab, lapatinib, or erlotinib (Fig. 3c). T-DM1 doses as low as 2 nM radiosensitized OE19 and produced G2/M accumulation (Fig. 3d,f).
  • In vivo tumour targeting. After i.v. 0.5 nmol Cy5-labelled ADC, T-MMAE accumulated in HER2+ OE19 xenografts and C-MMAE in EGFR+ CAL-27, SCC-61, SCC-35, SQ-9G, A549, and HCT-116 xenografts; LN229 (EGFR-low) tumours did not accumulate C-MMAE. C-MMAE persisted in EGFR+ tumours for ≥72 h (Fig. 4a–c).
  • CAL-27 (HNSC) efficacy. In CAL-27 xenografts treated on day 0 with 0.5 nmol cetuximab, free MMAE, or C-MMAE followed by 3 Gy on days 1 and 2, C-MMAE + IR produced significantly greater tumour-growth delay than all other arms (P<0.0001 by day 35; Fig. 4f). Co-administered free MMAE + cetuximab + IR was inferior to C-MMAE + IR (P<0.05 by day 28; Fig. 4g).
  • HER2+ xenograft efficacy. At 1 nmol single dose, T-MMAE and T-DM1 were equally efficacious at slowing OE19 growth vs control/trastuzumab through day 14 (Fig. 5b). At reduced dose (0.25 nmol T-DM1, ~1.9 mg/kg, day 0) plus 2.5 Gy on days 1–3, tumour doubling time increased from 7 (control) → 66 days (T-DM1 + IR) in OE19 and from 9 → 113 days in NCI-N87, while HCT116 (HER2−) showed no benefit from adding T-DM1 to IR (Table 1, Fig. 5d). Long-term tumour control was observed only in HER2+ tumours receiving T-DM1 + IR (Fig. 5e, log-rank P-values in Supplementary Tables).
  • Mechanism. Consistent across readouts, ADC radiosensitization tracks G2/M accumulation (pS10 Histone H3) and DNA double-strand breaks (comet tail length), and is abolished when the antibody–receptor pairing is mismatched — implicating receptor-mediated endocytosis and intracellular release of the anti-tubulin warhead as the gating step.

Genes & alterations

  • EGFR — surface expression level (not mutation status) gates C-MMAE binding, internalization, and radiosensitization. EGFR+ lines (CAL-27, A549, CALU3, SCC-61, SCC-35, SQ-9G, HCT-116) accumulate C-MMAE; LN229 (EGFR-low) does not.
  • ERBB2 (HER2) — HER2 expression gates T-MMAE and T-DM1 activity. HER2+ lines (OE19, NCI-N87, CALU3, BT474) show sub-nanomolar T-DM1 IC50 and G2/M arrest; HER2− lines (HCT-116, CAL-27, LN229) are ~100-fold less sensitive. The paper frames the broader ErbB family (EGFR, HER2, HER3, HER4) as tumour radioresistance drivers but only EGFR and ERBB2 are experimentally targeted.

Clinical implications

  • Biomarker-driven radiosensitization. The paper proposes ErbB-directed ADCs as a selective radiosensitizer class for HER2+ or EGFR+ locally advanced cancers, as an alternative to ErbB signal-transduction blockade (which suffers from bypass-pathway “whack-a-mole” resistance; refs 21, 24, 25).
  • Translational path via T-DM1. Because ado-trastuzumab emtansine is already FDA-approved for HER2+ metastatic breast cancer, the authors argue combining T-DM1 with IR “warrants speedy clinical evaluation” in HER2+ locally advanced disease (esophageal, gastric, breast).
  • Context of clinical failures. The authors note that adding cetuximab to cytotoxic chemo-radiotherapy in NSCLC or HNC failed to improve outcomes (refs 37–38), motivating the ADC-delivery paradigm over receptor signal inhibition.

Limitations & open questions

  • Preclinical only. All efficacy data are in cell lines and athymic-nude mouse xenografts. No human trial data; the proposal to clinically test T-DM1 + IR is hypothesis-generating.
  • Normal-tissue toxicity. The paper argues the therapeutic index should improve because the ADC is tumour-restricted, but no formal normal-tissue (e.g., cardiac, mucosal) toxicity study with combined T-DM1 + IR is reported.
  • Linker instability. The authors acknowledge that the MC-VC-MMAE linker undergoes slow retro-Michael reaction in circulation, potentially releasing free MMAE; improved linkers exist but were not used to preserve clinical fidelity.
  • Single-dose xenograft regimens. Long-term control was demonstrated with a single 0.25 nmol T-DM1 dose + 3 IR fractions; optimal fractionation, repeat dosing, and interaction with conventional concurrent cisplatin are not addressed.
  • No patient tumour data. HER2/EGFR expression cut-offs predictive of ADC radiosensitization in human tumours remain to be defined.
  • ErbB3/ErbB4. Although the introduction names all four ErbB family members, ERBB3 and ERBB4 are not experimentally interrogated.

Citations from this paper used in the wiki

  • “Here we report potent anti-tubulin drugs conjugated to anti-ErbB antibodies selectively radiosensitize to tumours based on surface receptor expression.” (Abstract)
  • “While two classes of potent anti-tubulins, auristatins and maytansinoids, indiscriminately radiosensitize tumour cells, conjugating these potent anti-tubulins to anti-ErbB antibodies restrict their radiosensitizing capacity.” (Abstract)
  • “a clinically used maytansinoid ADC, ado-trastuzumab emtansine (T-DM1), with IR prolongs tumour control in target expressing HER2+ tumours but not target negative tumours.” (Abstract)
  • “In HER2+ cell lines (OE19 and NCI N87 (gastric cancer)) the IC50 of T-DM1 was <1nM but was >100 nM in HER2− cells (HCT116 and CAL-27).” (Results, Fig. 3b)
  • “T-DM1 combined with IR was superior to all other treatment regimens … In stark contrast, HER2-HCT116 tumour xenografts showed no benefit of adding T-DM1 to IR.” (Results, Fig. 5d, Table 1)
  • “we propose an alternative ErbB mediated radiosensitization paradigm based on ErbB directed ADC that is more potent and potentially superior to signal inhibition.” (Discussion)

This page was processed by crosslinker on 2026-05-04.