henhao
Sign in
Cancer BiologyProceedings of the National Academy of Sciences
Peer-reviewed journal article

Dynamic Monitoring of Antibody Drug Conjugates Targeting TROP2 or HER2 in Breast Cancer using Circulating Tumor Cells

Avanish MishraiD, Rachel O. AbelmaniD, Quinn E. CunneelyiD, Victor R. PutaturoiD, Akansha Deshpande, Remy BelliD, Elizabeth M. Seider, Katherine H. XuiD, Mythreayi Shan, J. Kelly, Shih‐Bo HuangiD, Olivia Rieur, Justine Knape, Kaustav A. GopinathaniD, Kruthika KikkeriiD, Jon F.iD, John Walsh, Charles S. Dai, Leif William ElliseniD, David Tsai TingiD, Linda T. NiemaniD, Mehmet ToneriD, Aditya BardiaiD, Daniel A. HaberiD, Shyamala MaheswaraniD

PDFDOIPreprint
Post to X
Threads on XPost to BlueskyThreads on Bluesky
In-depth full-text digest available · Jump to digest ↓

Abstract

Antibody-drug conjugates (ADCs) target surface proteins on cancer cells, leading to internalization and delivery of a drug payload, thereby enhancing selectivity and minimizing toxicity. ADCs against TROP2 (Sacituzumab govitecan) or HER2 (T-DXd) have demonstrated efficacy in metastatic breast cancer, yet paradoxically, outside of HER2-amplified breast cancers, expression levels of these breast cancer-enriched epitopes in tumor biopsies have not been strongly correlated with clinical response. We undertook serial quantitative imaging of circulating tumor cells (CTCs) in a prospective cohort of 35 patients treated with either of these ADCs. At the single-cell level, expression of TROP2 and HER2 within individual patients is highly heterogeneous in both CTCs and paired tumor biopsies. Measurement of these epitopes on CTCs immediately prior to ADC therapy does not predict depth of clinical response. However, absence of CTCs or >80% reduction in CTC numbers after three weeks of treatment (CTCLow) predicts durable response, compared with CTCHigh cases (TROP2: HR 5.15, P = 0.012; HER2: HR 6.01, P<0.001). Targeted epitopes are not commonly downregulated on CTCs at the time of acquired clinical resistance, and switching between TROP2- and HER2-targeting ADCs sharing similar payloads infrequently leads to second-line response. Thus, while CTC burden is correlated with response to these ADCs, the level of TROP2 or HER2 expression is poorly predictive. These findings point to sensitivity to the drug payload as a potential driver of clinical response to currently approved ADCs in breast cancer. SIGNIFICANCE STATEMENTThe clinical efficacy of ADCs may depend both on differential targeting of cancer cells, using antibodies to tumor-enriched epitopes, and on the cleavage and release of drug payloads to which cancer cells are sensitive. ADCs currently used to treat breast cancer target either of two epitopes, TROP2 or HER2, but they share chemically related payloads. While CTC numbers track with response, we find that epitope expression is not strongly predictive. The limited success of sequentially switching between TROP2- or HER2-targeting ADCs as second-line treatment, following progression on a first-line ADC, highlights the need to incorporate non-cross-resistant drug payloads on ADCs to overcome such acquired resistance.

02 /Specification & Classification

Click any tag to browse matching papers
Citations
9
FWCI
19.23
Field percentile
top 1%
OpenAlex Topics

In-depth summary

✦ Multi-agent verified full-text synthesis

Claude Opus 5.5 · xhigh effort · 8 Nov 2026 · from the full text of version 2 (5 May 2026) · CC BY 4.0

In metastatic breast cancer on TROP2- or HER2-targeting antibody-drug conjugates, early CTC disappearance or steep decline tracked response, while CTC epitope levels neither predicted benefit nor vanished at resistance.

Clinicians lack biomarkers to choose among TROP2- and HER2-directed antibody-drug conjugates (ADCs) or explain resistance, and a single metastatic biopsy may misrepresent heterogeneous epitope expression across lesions and time. Using circulating tumor cell (CTC) enrichment unbiased by cell size or epithelial markers plus multispectral imaging, the study serially tests whether epitope levels or CTC counts relate to ADC benefit and acquired resistance.

Key results

  1. SuggestedEarly on-treatment CTC clearance (no detectable CTCs, or a fall of more than 80% at three weeks) separated longer from shorter responses to both TROP2- and HER2-targeting ADCs.

    TROP2-ADC (n=13): nine of ten CTC^Low cases had strong responses vs both CTC^High cases poor; median 297 vs 87 days, HR 5.15, p = 0.012. HER2-ADC (n=15): seven (70%) of 10 CTC^Low cases strong vs all four CTC^High poor; median 276 vs 44 days, HR = 6.01, p < 0.001 (log-rank). · Serial CTC enumeration and TROP2 staining in…; Serial monitoring of HER2 expression on CTCs in…; Figure legends

  2. SuggestedThe pretreatment fraction of CTCs expressing TROP2 or HER2 did not distinguish strong from poor responders, and clinical HER2 IHC on biopsies was also uninformative; the only exception was the two HER2-amplified cases (>80% HER2-expressing CTCs), which both responded strongly.

    TROP2 >50% positive CTCs: two strong vs three poor; <50%: two vs two. HER2 >50%: five vs five; <50%: one vs two. A 90% cutoff also failed for both epitopes. Matched biopsies contained more high-expressing cells than CTCs, with TROP2 discordant in five of seven cases and HER2 in one. · Comparison of TROP2 and HER2 protein expression…; Serial CTC enumeration and TROP2 staining in…; Serial monitoring of HER2 expression on CTCs in…

  3. SuggestedSwitching to an ADC against the other epitope after progression mostly gave brief responses, and single cancer cells co-expressed TROP2 and HER2. The authors propose that shared camptothecin-related payloads may explain the poor second-line benefit.

    Nine of ten second-line ADC patients fell below the 6-month cutoff; median time to progression 226 vs 48 days for first- vs second-line (p < 0.001, Welch's t-test); shorter second-line responses followed a mean 12.25 first-line cycles. TROP2-HER2 single-cell correlation r = 0.49 in CTCs, r = 0.68 in biopsy cells (P < 0.0001). · Sequential treatment with TROP2- and…; Figure legends

  4. SuggestedAt progression on either ADC class, CTCs did not consistently lose the targeted epitope, which argues against antigen downregulation as the main resistance route in this cohort.

    TROP2-ADC (six paired patients): CTC TROP2 distribution was unchanged at progression in three of four strong and both poor responders; MGH-009 had a single CTC with higher TROP2. HER2-ADC (five paired patients): one rise (MGH-030), two unchanged, one decrease, one newly HER2-expressing. · Serial CTC enumeration and TROP2 staining in…; Serial monitoring of HER2 expression on CTCs in…; DISCUSSION

  5. ShownFixing blood at collection in Streck tubes extended the CTC-iChip processing window to as long as 72 hours while keeping CTC recovery high and single-cell imaging usable.

    Cultured CTCs spiked into fixed blood stored 24 hours: recovery 96.9 ± 5.1%, Log10 WBC depletion 2.88 ± 0.56, Log10 RBC depletion 4.19 ± 0.47. At baseline, 28 of 35 (80%) patients had ≥3 imageable CTCs per 20mL (median 47, range 4 to 1,481 CTCs/20mL). · Quantitative imaging of TROP2 and HER2 epitope…

Methods, data and reproducibility

Whole blood collected in Streck or EDTA tubes was depleted of RBCs, platelets and magnetically tagged leukocytes (CD45, CD16, CD66b) on the microfluidic CTC-iChip; cytospun cells were imaged at 40X by multispectral microscopy and each CTC scored null, low, medium or high for TROP2 and HER2 against cell-line thresholds at baseline, three weeks and progression.

  • Prospective metastatic breast cancer cohort sampled serially during ADC therapy (22 HR+, 11 TNBC, 2 HER2-amplified); mean 3.3 prior regimens · human (all female) · n = 35 patients (38 treatment courses)
  • First-line TROP2-targeting ADC subcohort with CTCs at baseline, three weeks on treatment and progression · human · n = 13 patients
  • First-line HER2-targeting ADC (trastuzumab deruxtecan) subcohort, mostly HER2-low tumors (13 of 15) · human · n = 15 patients
  • Sequential ADC subcohort: HER2-ADC then TROP2-ADC, or TROP2-ADC then HER2-ADC · human · n = 10 patients (n=5 per sequence)
  • Patient-matched metastatic tumor biopsies (FFPE) and CTCs stained with the same fluorescent antibody panels · human · n = seven technically adequate matched cases
  • Calibration cell lines for expression thresholds: BRx-142 (breast CTC line), Mel-167 (melanoma), MDA-MB-231, AU565 · human · RRID:CVCL_0062 (MDA-MB-231); RRID:CVCL_1074 (AU565)
  • Healthy donor whole blood (internal donors and Research Blood Components, LLC); prior healthy donor samples guided imaging exposure settings · human
Methods & tools
CTC-iChip (inertial separation array plus magnetic leukocyte depletion), Streck blood collection tubes, Dynabeads MyOne Streptavidin T1, Thermo Shandon Cytospin 4, Akoya PhenoImager (PhenoImagerHT) multispectral imaging, Akoya InForm spectral unmixing, HALO image analysis platform, Leica Bond RX automated stainer, Multiplex immunofluorescence (EpCAM/panCK/CK19, CD45/CD16/CD66b, TROP2, HER2, DAPI), Kaplan-Meier and log-rank analysis, Welch's t-test
Reproducibility gaps
Images and data are available only on request via the Open Science Framework (OSF). No code or HALO segmentation parameters are shared. The CTC-iChip is patent-protected and commercialized by TellBio, which some authors co-founded.
Limitations (6)
  • The authors call this a proof-of-principle analysis of a small, heterogeneous standard-of-care cohort. Few patients (two in the TROP2-ADC group) showed the adverse CTC^High pattern, so the early-response marker needs larger studies. · stated by the authors
  • The cohort was mainly heavily pretreated, while ADCs are moving to earlier lines. HER2-amplified (classical HER2+) tumors, where HER2 dependence is established, were not addressed. · stated by the authors
  • Considerable single-cell TROP2 and HER2 heterogeneity, both within and between patients, prevented a direct comparison of CTCs against matched tumor biopsies. · stated by the authors
  • The lack of association with TROP2 levels cannot rule out a bystander effect, in which a few TROP2-expressing tumor cells allow killing of neighboring antigen-negative cells. · stated by the authors
  • Payload cross-resistance is inferred, not tested (no TOP1 sequencing or payload-sensitivity assays). Comparing first-line with second-line courses also mixes treatment line with disease stage, which confounds the 226 vs 48 day contrast. · noted in this summary
  • Response was split at 6 months on therapy rather than by standardized radiographic criteria. Hazard ratios rest on small groups with no adjustment for subtype or prior lines, and expression classes come from quartiles of only a few cell lines. · noted in this summary

Who should read:Read this if you develop liquid-biopsy biomarkers for antibody-drug conjugates, plan sequencing of TROP2- and HER2-directed ADCs in metastatic breast cancer, or study epitope versus payload drivers of ADC resistance.

Reproducibility ledger

Nobody has reported trying to reproduce this paper yet. Did you replicate an experiment or run the code? Your report saves the next lab a week.

Sign in to file a reproducibility report.

Related papers

Similarity Map
SelectedSame fieldCross-field
THISDogru '26Agarwal '22Yamazaki '21Wu '26Das '25Otvodnikova '26
69% semantic matchPharmacology & Toxicology
Sedat Dogru, Ashwatha Suresh, Edward P. Bowman+3 more·bioRxiv·Sep 25, 2026