Purpose:

TGFβs are overexpressed in many advanced cancers and promote cancer progression through mechanisms that include suppression of immunosurveillance. Multiple strategies to antagonize the TGFβ pathway are in early-phase oncology trials. However, TGFβs also have tumor-suppressive activities early in tumorigenesis, and the extent to which these might be retained in advanced disease has not been fully explored.

Experimental Design:

A panel of 12 immunocompetent mouse allograft models of metastatic breast cancer was tested for the effect of neutralizing anti-TGFβ antibodies on lung metastatic burden. Extensive correlative biology analyses were performed to assess potential predictive biomarkers and probe underlying mechanisms.

Results:

Heterogeneous responses to anti-TGFβ treatment were observed, with 5 of 12 models (42%) showing suppression of metastasis, 4 of 12 (33%) showing no response, and 3 of 12 (25%) showing an undesirable stimulation (up to 9-fold) of metastasis. Inhibition of metastasis was immune-dependent, whereas stimulation of metastasis was immune-independent and targeted the tumor cell compartment, potentially affecting the cancer stem cell. Thus, the integrated outcome of TGFβ antagonism depends on a complex balance between enhancing effective antitumor immunity and disrupting persistent tumor-suppressive effects of TGFβ on the tumor cell. Applying transcriptomic signatures derived from treatment-naïve mouse primary tumors to human breast cancer datasets suggested that patients with breast cancer with high-grade, estrogen receptor–negative disease are most likely to benefit from anti-TGFβ therapy.

Conclusions:

Contrary to dogma, tumor-suppressive responses to TGFβ are retained in some advanced metastatic tumors. Safe deployment of TGFβ antagonists in the clinic will require good predictive biomarkers.

This article is featured in Highlights of This Issue, p. 521

Translational Relevance

Most novel cancer therapies are taken into clinical trials based on preclinical data from a limited number of models that capture little of the variability of the human disease. Antagonists of the TGFβ pathway are currently in early-phase clinical oncology trials. However, this pathway is a particularly complex therapeutic target, as TGFβs can have both stimulatory and suppressive effects on multiple cell types in the tumor ecosystem. Here, we tested a neutralizing anti-TGFβ antibody in a large panel of 12 immunocompetent mouse models of metastatic breast cancer. Although anti-TGFβ therapy suppressed metastasis in many of the models, 25% of models showed an undesirable increase in metastasis following TGFβ blockade. Good predictive biomarkers will be crucial for optimal deployment of TGFβ antagonists in the clinic. Our data suggest that patients with breast cancer with high-grade, hormone receptor–negative disease may be the most likely to benefit.

TGFβs are overexpressed by many advanced human tumors, and high expression frequently correlates with poor prognosis, making the pathway a candidate for therapeutic targeting (1). Multiple biological activities of TGFβs contribute to driving cancer progression. TGFβs have been implicated in promoting invasion and migration of tumor cells, driving tumor cell plasticity and the epithelial-to-mesenchymal transition (EMT), expanding the cancer stem cell (CSC) compartment, and enhancing generation of cancer-associated fibroblasts (1–3). Importantly, TGFβs have strong immunosuppressive activity, and extensive evidence suggests that elevated TGFβ expression by tumor or stromal cells may compromise antitumor immunity and limit the efficacy of immunotherapy (1, 4–7). TGFβs can also mediate resistance to chemotherapy (8) and radiotherapy (9). Compounding the problem, many therapeutic approaches themselves further increase TGFβ production, including radiation (10), chemotherapy (10), and immune checkpoint inhibition (11). Thus, there is a compelling rationale to be made for attempting TGFβ pathway blockade. On the basis of encouraging preclinical data showing therapeutic benefit of targeting the TGFβ signaling axis (1, 11–18), over 40 early-phase clinical oncology trials are now ongoing, using various TGFβ pathway antagonists either as single agents, or in combination with other therapeutics, including immune checkpoint inhibitors (https://clinicaltrials.gov; ref. 1). In general, TGFβ pathway blockade has been well tolerated in the clinic, with some early signs of efficacy (19–21).

The situation is complicated by the highly pleiotropic nature of the biological processes that are regulated by TGFβ, as TGFβ has context-dependent tumor suppressor activity in addition to its prooncogenic properties (1–3). The prevailing dogma is that for most epithelial tumors, TGFβ functions as a tumor suppressor early in the carcinogenic process through homeostatic effects on cell proliferation, survival, genomic integrity, and inflammatory cytokine production (1–3). However, during tumor progression, genetic and epigenetic changes in the tumor cell, coupled with increased local levels of TGFβ and altered TGFβ responsiveness of the tumor cell, frequently lead to selective loss of these tumor-suppressive responses. Proprogression effects of TGFβ on tumor cells and stroma are progressively unmasked, and come to dominate in later stages of the disease. Given this complex dual role, there was initially reluctance to target the TGFβ pathway in cancer, until preclinical studies in the early 2000s suggested that it might be possible to target the prooncogenic TGFβ in advanced disease without disrupting effects on normal homeostasis and tumor suppression (12, 13).

Preclinical studies in mouse models play an important role in supporting the clinical drug development process. However, there has been emerging awareness recently of limitations of the preclinical research endeavor in generating information that translates usefully into clinical practice (22, 23). One significant issue is that much preclinical work is done using a relatively small number of well-studied models that fail to capture the heterogeneity of the human disease. Other limitations have included a heavy reliance on immunodeficient mice, and the failure to use metastatic burden as the clinically relevant endpoint. To overcome some of these issues, we recently assembled and characterized a panel of 12 immunocompetent allograft models of metastatic breast cancer that capture some of the heterogeneity of the human disease (24). Here, we have tested the effect of anti-TGFβ–neutralizing antibodies on the metastatic endpoint across all models. Using this expanded panel, we find that 3 of 12 models respond to anti-TGFβ therapy with an undesirable stimulation of metastasis, suggesting that tumor-suppressive responses to TGFβ may be retained in advanced disease in some cases. We exploit the panel to gain insights into potential predictive biomarkers and underlying mechanisms.

Mouse cell line models of metastatic breast cancer

The 12 mouse metastatic mammary cell lines 4T1, EMT6, TSAE1, MET1, R3T, HRM1, 6DT1, D2A1, E0771, F3II, M6, and MVT1 were obtained from the originating investigators as described in Yang and colleagues (24). Individual characteristics of the cell lines, including culture conditions, are given in Supplementary Table S1, with additional information about genomic characterization in ref. 24.

Animal experiments

All animal experiments were conducted under protocol LC-070 approved by the Animal Care and Use Committee of the NCI (Bethesda, MD). For metastasis therapy studies, tumor cells were either orthotopically implanted into the mammary fat pad (4T1, EMT6, R3T, HRM1, 6DT1, D2A1, E0771, M6, MVT1), or delivered by tail-vein injection (TSAE1, F3II, MET1) into strain-matched mice. Unless otherwise indicated, primary tumors were resected when they reached 0.5 to 0.8 cm diameter. Mice were randomized to treatment groups of 10 to 15 mice/group, and treated with the neutralizing anti-TGFβ mouse mAb 1D11 (Genzyme Corp., or Bio X Cell) or isotype control (13C4, Genzyme Corp. or MOPC21, Bio X Cell) at 5 mg/kg i.p. 3 times per week unless otherwise indicated. At resection or experimental endpoint, tumors were harvested for histology and molecular analyses. Metastatic burden in the lung was determined by counting histologically evident metastases in lung cross-sections. More details are available in Supplementary Methods.

Statistical analysis

Statistical significance in animal-based studies was determined by Mann–Whitney test, or Kruskal–Wallis test with Dunn multiple comparison correction, using GraphPad Prism 7 unless otherwise specified. Statistical significance in human datasets was determined by two-tailed Student t test for independent samples, or one-way ANOVA unless otherwise specified.

Supplementary methods

Additional methodologic details are available in the Supplementary Methods and Supplementary Tables.

Data availability

Gene expression microarray data are available from NCBI Gene Expression Omnibus (GEO, http://www.ncbi.nlm.nih.gov/geo/) under the accession number GSE96006. Copy number variant data and single-nucleotide variant data are in GEO under accession # GSE69902.

Heterogeneous responses to TGFβ antagonism in a panel of immunocompetent metastatic breast cancer models

We have previously assembled and characterized a panel of 12 immunocompetent mouse allograft models of metastatic breast cancer (24). Here, we assessed the effect of treatment with a pan-TGFβ–neutralizing antibody on the metastatic endpoint across the panel. The antibody used was the mouse monoclonal, 1D11, which has similar properties to the fully human anti-TGFβ antibody fresolimumab (25). Where possible (n = 6 models), the models were run in the clinically relevant format of orthotopic tumor implantation with subsequent surgical resection. In some cases, adequate metastatic efficiency was only achieved if primary tumors were left unresected (n = 3 models), or if tumor cells were injected via the tail vein (n = 3 models). For all models, lung metastatic burden was quantitated in histologic cross-sections (Fig. 1A and B). The results are presented in Table 1, with representative data for each response class in Fig. 1C, and full datasets in Supplementary Fig. S1. The effect of the antibody on primary tumor burden varied between models, but normalization of metastatic burden to primary tumor weight confirmed that effects on metastasis of 4T1, EMT6, and MVT1 models were independent of effects on the primary tumor (Supplementary Fig. S1). Five models (42%) showed inhibition of metastasis in response to anti-TGFβ antibody treatment (“InhibMet” class). Four models (33%) showed no effect of anti-TGFβ antibodies on metastasis (“NoEff” class), while three (25%) showed an undesirable stimulation of metastasis (“StimMet” class). The metastasis-stimulating effects of anti-TGFβ treatment were confirmed in at least one independent experiment for each StimMet model. Although antibody treatment was generally initiated on day+1 after tumor cell implantation for most experiments, similar results were obtained if antibody treatment was delayed until tumors were well established (Supplementary Fig. S2A–S2D).

Figure 1.

Heterogeneous effects of anti-TGFβ antibody therapy on metastasis in breast cancer models. A, Schematic for experimental format used for models in C. B, Representative H&E-stained sections of lungs with metastases (4T1 model; median mouse/group). C, Effects of anti-TGFβ therapy on metastatic burden. One representative model is shown for each the three different response categories (InhibMet, 4T1; NoEff, 6DT1; StimMet, MVT1). Median ± IQ range for n = 11 to 15/group; Mann–Whitney U test. D, Representative virtual Western blot for quantitative CNIA assay assessing effect of anti-TGFβ antibodies on SMAD2 activation in 4T1 primary tumors for four independent tumors (#1 through #4) from each experimental group. E, Effect of anti-TGFβ antibody treatment on activation state of SMAD signaling pathways (top) or non-SMAD signaling pathways (bottom) in primary tumors from one representative model for each response class, assessed by quantitative CNIA. Mean ± SD, n = 4 tumors/model/treatment; unpaired t test for treated versus control comparisons.

Figure 1.

Heterogeneous effects of anti-TGFβ antibody therapy on metastasis in breast cancer models. A, Schematic for experimental format used for models in C. B, Representative H&E-stained sections of lungs with metastases (4T1 model; median mouse/group). C, Effects of anti-TGFβ therapy on metastatic burden. One representative model is shown for each the three different response categories (InhibMet, 4T1; NoEff, 6DT1; StimMet, MVT1). Median ± IQ range for n = 11 to 15/group; Mann–Whitney U test. D, Representative virtual Western blot for quantitative CNIA assay assessing effect of anti-TGFβ antibodies on SMAD2 activation in 4T1 primary tumors for four independent tumors (#1 through #4) from each experimental group. E, Effect of anti-TGFβ antibody treatment on activation state of SMAD signaling pathways (top) or non-SMAD signaling pathways (bottom) in primary tumors from one representative model for each response class, assessed by quantitative CNIA. Mean ± SD, n = 4 tumors/model/treatment; unpaired t test for treated versus control comparisons.

Close modal
Table 1.

Effect of anti-TGFβ therapy on lung metastasis burden in metastatic breast cancer model panel.

Cell lineTumor originMouse strainMetastasis assay formatER statusIntrinsic transcriptomic subtypeTp53 statusRas statusMyc statusPI3K pathway statusEffect of anti-TGFβ on lung metastasisMetastasis effect size (absolute fold change in metastasis number)P value for metastasis numberP value for metastasis index
4T1 Spontaneous BALB/c Orthotopic with resection ER Luminal A Nulla WT WT WT Inhibits 5.8 P = 0.0003 P = 0.02 
EMT6 Spontaneous BALB/c Orthotopic with resection ER+ Luminal A WT WT WT Pten null Inhibits >2 P = 0.0006 P = 0.003 
TSAE1 Spontaneous BALB/c Tail vein injection ER+ Luminal A Mutant mutKras Amp WT Inhibits 2.4 P < 0.0001 n/a 
MET1 MMTV-PVT GEM FVB/N Tail vein injection ER Luminal A Mutant WT WT WTb Inhibits 7.5 P = 0.002 n/a 
R3T DMBA treated OPN−/− GEM 129S1 Orthotopic with resection ER Luminal A Mutant mutKras WT WT Inhibits 3.8 P = 0.05 ns 
HRM1 PIK3CA-Y1047R GEM FVB/N Orthotopic with resection ER+ Luminal A WT WT Amp mutPik3ca No Effect n/a ns  
6DT1 MMTV-Myc GEM FVB/N Orthotopic, no resection ER Luminal A/B WT mutKras Ampc mutPik3ca No Effect n/a ns  
D2A1 Spontaneous BALB/c Orthotopic with resection ER Luminal B WT WT Amp WT No Effect n/a ns  
E0771 Spontaneous C57Bl/6 Orthotopic, no resection ER Luminal B Mutant mutKras Amp WT No Effect n/a ns  
F311 Spontaneous BALB/c Tail vein injection ER+ Luminal A Mutant WT WT WT Stimulates P = 0.0007 n/a 
M6 C3(1)Tag GEM FVB/N Orthotopic with resection ER Basal Nulla WT WT WT Stimulates P = 0.01 ns 
MVT1 MMTV-Myc/VEGF GEM FVB/N Orthotopic, no resection ER+ Luminal A/B WT mutKras Ampc mutPik3ca Stimulates P < 0.0001 P < 0.0001 
Cell lineTumor originMouse strainMetastasis assay formatER statusIntrinsic transcriptomic subtypeTp53 statusRas statusMyc statusPI3K pathway statusEffect of anti-TGFβ on lung metastasisMetastasis effect size (absolute fold change in metastasis number)P value for metastasis numberP value for metastasis index
4T1 Spontaneous BALB/c Orthotopic with resection ER Luminal A Nulla WT WT WT Inhibits 5.8 P = 0.0003 P = 0.02 
EMT6 Spontaneous BALB/c Orthotopic with resection ER+ Luminal A WT WT WT Pten null Inhibits >2 P = 0.0006 P = 0.003 
TSAE1 Spontaneous BALB/c Tail vein injection ER+ Luminal A Mutant mutKras Amp WT Inhibits 2.4 P < 0.0001 n/a 
MET1 MMTV-PVT GEM FVB/N Tail vein injection ER Luminal A Mutant WT WT WTb Inhibits 7.5 P = 0.002 n/a 
R3T DMBA treated OPN−/− GEM 129S1 Orthotopic with resection ER Luminal A Mutant mutKras WT WT Inhibits 3.8 P = 0.05 ns 
HRM1 PIK3CA-Y1047R GEM FVB/N Orthotopic with resection ER+ Luminal A WT WT Amp mutPik3ca No Effect n/a ns  
6DT1 MMTV-Myc GEM FVB/N Orthotopic, no resection ER Luminal A/B WT mutKras Ampc mutPik3ca No Effect n/a ns  
D2A1 Spontaneous BALB/c Orthotopic with resection ER Luminal B WT WT Amp WT No Effect n/a ns  
E0771 Spontaneous C57Bl/6 Orthotopic, no resection ER Luminal B Mutant mutKras Amp WT No Effect n/a ns  
F311 Spontaneous BALB/c Tail vein injection ER+ Luminal A Mutant WT WT WT Stimulates P = 0.0007 n/a 
M6 C3(1)Tag GEM FVB/N Orthotopic with resection ER Basal Nulla WT WT WT Stimulates P = 0.01 ns 
MVT1 MMTV-Myc/VEGF GEM FVB/N Orthotopic, no resection ER+ Luminal A/B WT mutKras Ampc mutPik3ca Stimulates P < 0.0001 P < 0.0001 

Note: Models were treated with anti-TGFβ or control antibodies using the indicated metastasis assay format. Marker, mutation data and transcriptomic classification of the models is from Yang and colleagues' work (24). Effect size is the fold suppression or promotion of the number of metastases/lung following anti-TGFβ treatment. n = 10 to 15 mice/treatment group/model; Mann–Whitney P value; ns, not significant; n/a not applicable. The P value for the metastasis index is the P value for the difference in metastatic burden following correction for effect of therapy on primary tumor size, where applicable. All models were IHC negative for progesterone receptor and genomically negative for HER2 amplification.

Abbreviations: GEM, genetically engineered mouse; Mut, mutant.

aGenetically wild type but functionally null.

bGenetically wild type but functionally activated.

cTransgenically amplified.

The canonical TGFβ signaling pathway involves phosphorylation of the signaling transducers, SMAD2 and SMAD3, which associate with SMAD4 and regulate transcription (26). Noncanonical signaling can also occur through other SMADs (SMAD1,5,9), or non-SMAD pathways such as MAPK, JNK, p38 (MAPK14), and AKT (26). To confirm pharmacodynamic activity of the drug, quantitative capillary nano-immunoassays (CNIA; SimpleWestern) for activation of TGFβ signaling pathways were performed on lysates from primary tumors treated with anti-TGFβ or control antibodies. The results showed that anti-TGFβ treatment could reduce phosphorylation of the canonical TGFβ signaling pathway components, SMAD2 and/or SMAD3, by 30% to 50% in tumors from all three response classes (Fig. 1D and E; Supplementary Fig. S3A and S3B), indicating that lack of response in the NoEff class was not due to the drug failing to engage the target. In contrast, no effect of treatment was seen on noncanonical signaling through the AKT, JNK, p38, or MAPK/ERK pathways in any response class (Fig. 1E). In summary, with this expanded panel of breast cancer models, we find that some models robustly respond to TGFβ antagonism with an undesirable increase in metastatic burden, raising the possibility that this phenomenon might also occur in human patients with breast cancer.

No robust correlation between TGFβ expression or pathway activation and response to anti-TGFβ therapy

Our observation of undesirable stimulatory responses to TGFβ antagonism in 25% of models tested makes the development of good predictive biomarkers critical. We first assessed whether response-to-therapy correlated with any parameters of TGFβ production or signaling in the models. Protein levels for all three TGFβ isoforms were quantitated in treatment-naïve primary tumors, and in plasma from mice bearing large primary tumors (0.5–1 cm diameter). Treatment-naïve primary tumors were used as this material is the most likely to be available in a clinical setting. Total TGFβ in the primary tumors varied over a 4-fold range across the models (Fig. 2A) and was unexpectedly lowest in the InhibMet class, although the classes were not well separated (Fig. 2B). Analyzing by isoform, the NoEff response class tended toward higher TGFβ1 and lower TGFβ3 (Supplementary Fig. S4A and S4B), with a similar, although less pronounced trend at the RNA level (Supplementary Fig. S4C and S4D), but again, differences were not great. With few exceptions, circulating plasma TGFβ1 and TGFβ2 protein levels were in the normal range for all tumor-bearing animals (Fig. 2C and D) and TGFβ3 was undetectable (<0.12 ng/mL). Thus, there is no strong correlation between circulating or local TGFβ ligand levels in the treatment-naïve tumor models and therapeutic response.

Figure 2.

TGFβ expression and signaling in primary tumors from models in the different therapeutic response classes. A, Expression of TGFβ protein isoforms in acid–ethanol extracts of primary mammary tumors compared with normal mammary gland (Mam Gl). Results are mean ± SD for n = 5 to 8 mice/group. ANOVA with Dunnett multiple comparisons test for total TGFβ in tumors from individual models versus normal mammary gland. Dotted line shows mean for normal mammary gland. B, Total TGFβ (all three isoforms) for 5 tumors/model in each therapeutic response class. Results are median ± IQ range; Kruskal–Wallis test. C and D, TGFβ1 (C) or TGFβ2 (D) in plasma from normal or tumor-bearing mice. Dotted lines represent ± 2 SD from mean in normal mice. Data are median ± IQ range for n = 5 to 7 tumor-bearing mice and 21 normal mice. Response class for the tumor models is indicated. E, PhosphoSMAD2 activation status in untreated primary tumors assessed by quantitative CNIA. Mean ±SD, n = 3 to 5 tumors/model. F, Heatmap summarizing relative activation state of SMAD signaling and noncanonical TGFβ signaling pathways in tumors from the model panel. Each data point represents the mean phospho-target/target ratio for the model (n = 3–5 tumors/model), after median centering of the data across the model panel for each target. G, AKT phosphorylation status in untreated individual tumors in the different response classes. Results are median ± IQ range; Kruskal–Wallis test. H, Schematic of canonical and mixed SMAD complexes and their detection by brightfield proximity ligation assay (PLA). I, Representative images for proximity ligation assay staining for mixed SMAD complexes in representative primary tumors showing low (EMT6) and high (6DT1) mixed SMAD formation. Brown dots represent complex formation. Scale bar, 50 μm. J–L, Semiquantitative PLA scores for noncanonical mixed SMAD complexes (J), canonical TGFβ SMAD complexes (K), and canonical BMP SMAD complexes (L) in the three therapeutic response classes. Results are median ± IQ range within the response class, n = 3 tumors/model; Kruskal–Wallis test.

Figure 2.

TGFβ expression and signaling in primary tumors from models in the different therapeutic response classes. A, Expression of TGFβ protein isoforms in acid–ethanol extracts of primary mammary tumors compared with normal mammary gland (Mam Gl). Results are mean ± SD for n = 5 to 8 mice/group. ANOVA with Dunnett multiple comparisons test for total TGFβ in tumors from individual models versus normal mammary gland. Dotted line shows mean for normal mammary gland. B, Total TGFβ (all three isoforms) for 5 tumors/model in each therapeutic response class. Results are median ± IQ range; Kruskal–Wallis test. C and D, TGFβ1 (C) or TGFβ2 (D) in plasma from normal or tumor-bearing mice. Dotted lines represent ± 2 SD from mean in normal mice. Data are median ± IQ range for n = 5 to 7 tumor-bearing mice and 21 normal mice. Response class for the tumor models is indicated. E, PhosphoSMAD2 activation status in untreated primary tumors assessed by quantitative CNIA. Mean ±SD, n = 3 to 5 tumors/model. F, Heatmap summarizing relative activation state of SMAD signaling and noncanonical TGFβ signaling pathways in tumors from the model panel. Each data point represents the mean phospho-target/target ratio for the model (n = 3–5 tumors/model), after median centering of the data across the model panel for each target. G, AKT phosphorylation status in untreated individual tumors in the different response classes. Results are median ± IQ range; Kruskal–Wallis test. H, Schematic of canonical and mixed SMAD complexes and their detection by brightfield proximity ligation assay (PLA). I, Representative images for proximity ligation assay staining for mixed SMAD complexes in representative primary tumors showing low (EMT6) and high (6DT1) mixed SMAD formation. Brown dots represent complex formation. Scale bar, 50 μm. J–L, Semiquantitative PLA scores for noncanonical mixed SMAD complexes (J), canonical TGFβ SMAD complexes (K), and canonical BMP SMAD complexes (L) in the three therapeutic response classes. Results are median ± IQ range within the response class, n = 3 tumors/model; Kruskal–Wallis test.

Close modal

To address whether TGFβ signaling in the tumor cells might differ between the different response classes, we first showed by exome sequencing that the core signaling components Tgfbr1, Tgfbr2, Tgfbr3, Smad2, Smad3, and Smad4 were not mutated or deleted in any of the models (data from ref. 24). Furthermore, TGFβ could induce phosphorylation of SMAD2 and/or SMAD3 in all the tumor cell lines in vitro (Supplementary Fig. S5A), so the canonical TGFβ signaling pathway was intact in all 12 tumor models. We then used quantitative SimpleWestern CNIA to assess the basal activation state of TGFβ signaling through SMAD and non-SMAD pathways in untreated primary tumors from all models. The endogenous activation state of the individual pathways in the tumors varied significantly across the model panel (shown for SMAD2 in Fig. 2E). However, there was no strong correlation with response-to-therapy (Fig. 2F; Supplementary Fig. S5B). The NoEff class showed significantly lower AKT activation, but the groups were not well distinguished (Supplementary Fig. S2B).

Recent work has suggested that certain prooncogenic effects seen at high TGFβ levels may be mediated through activation of noncanonical “mixed SMAD” signaling complexes comprising activated TGFβ SMADs (SMAD2 or SMAD3) together with activated BMP SMADs (SMAD1, SMAD5, or SMAD9; refs. 27, 28). We hypothesized that the presence of these complexes might indicate that the prooncogenic arm of TGFβ signaling was dominant and hence identify the InhibMet class. Using a brightfield proximity ligation assay that we developed (29), we detected mixed SMAD, canonical TGFβ SMAD and BMP SMAD complexes semiquantitatively in treatment-naïve tumor samples (Fig. 2H and I; Supplementary Fig. S6A–S6C). Contrary to our hypothesis, no significant difference was seen for the mixed SMADs (Fig. 2J), while canonical TGFβ SMAD signaling (Fig. 2K) and canonical BMP signaling (Fig. 2L) trended to lower in the StimMet class. Thus overall, we were unable to identify any parameter of TGFβ production or signaling that could robustly distinguish the therapeutic response classes.

Tumor transcriptomics identify gene signatures associated with the different response classes

Because the candidate approach was unfruitful, we performed transcriptomic analysis on treatment-naïve primary tumors from all models (n = 4/model), as a discovery approach to biomarker identification, and to gain mechanistic insights. The tumor transcriptomes did not clearly segregate by therapeutic response class (Fig. 3A). Supervised hierarchical clustering showed that the two responder classes (InhibMet and StimMet) were transcriptomically more similar to each other than to the nonresponder class (NoEff; Fig. 3B). We started by testing whether existing TGFβ-related transcriptomic markers and/or signatures could discriminate the response classes. A number of prognostic TGFβ response signatures have been generated in preclinical model systems, of which the most widely used is the “TBRS” from the Massague laboratory (30). The TBRS, which, like most TGFβ response signatures, reflects a mix of tumor-promoting and tumor-suppressive effects, showed similar scores in all three response classes (Fig. 3C). In contrast, we previously generated a signature [TGFβ/SMAD3 tumor suppressor signature score (TSTSS)] that specifically reflects just the tumor cell–autonomous suppressive effects of TGFβ, and associates with good outcome in human breast cancer datasets (31). Here, we found that tumors from the StimMet response class show significantly higher expression of the TSTSS signature (Fig. 3D), suggesting that tumor cells in the StimMet models may retain tumor-suppressive responses to TGFβ despite being high grade and aggressively metastatic. However, the TSTSS still only weakly discriminated the StimMet class. We also tested several genes (Pspc1, Klf5, Ywhaz (14-3-3ζ), Six1, Peak1, Rassf1, Dab2) that have been previously proposed to function as molecular “switch” determinants of whether TGFβ acts predominantly as a tumor suppressor or a tumor promoter (32). Of these, only Six1 and Ywhaz showed a significant association with response-to-therapy (Supplementary Fig. S7A–S7C), with the direction of the Ywhaz association being opposite to that predicted from the literature (33). Similarly mutant p53, which has been proposed to cause TGFβ to switch toward oncogenic signaling (34), was not associated with response-to-therapy (Table 1).

Figure 3.

Transcriptomes of untreated primary tumors distinguish the three therapeutic response classes. A, Principal component analysis of primary tumor transcriptomes (n = 4/model). B, Heatmap of top 257 most differentially expressed genes between the three response classes (ANOVA P < 0.0001). Yellow, upregulated; blue, repressed. C, TGFβ response signature score (TBRS) in each response class (median ± IQ range, n = 4 tumors/model; Kruskal–Wallis test with Dunn multiple comparison correction). D, TSTSS in primary tumors (statistics as for C). E, Schematic for generation of response class-specific gene signatures. F and G, InhibMet (F) or StimMet (G) signature score in the different response classes. Mann–Whitney test. H and I, Scores for InhibMet (H) and StimMet (I) signatures in patients with breast cancer from the METABRIC dataset (n = 2,000 patients), stratified by hormone receptor status, intrinsic subtype, or histopathologic grade. ER status, t test; others, ANOVA with Tukey multiple comparison test. ER, n = 439; ER+, n = 1,498; basal, n = 209; Claudin-low, n = 218, HER2, n-224; luminal A, n = 700; luminal B, n = 475; normal-like, n = 148; grade 1, n = 169; grade 2, n = 771; grade 3, n = 952. J, Gene set enrichment analysis (GSEA) showing Hallmark gene sets that are similarly enriched in both responder groups. NES, normalized enrichment score. Results shown for gene sets with absolute NES > 2 in at least one group and difference in NES between groups <1.5. K, GSEA for Hallmark gene sets that are differentially enriched (difference in NES ≥ 3) in the two responder groups. L and M, Representative enrichment plots for the top most affected Hallmark gene set in each responder group. N, Summary of predicted biological differences between response classes.

Figure 3.

Transcriptomes of untreated primary tumors distinguish the three therapeutic response classes. A, Principal component analysis of primary tumor transcriptomes (n = 4/model). B, Heatmap of top 257 most differentially expressed genes between the three response classes (ANOVA P < 0.0001). Yellow, upregulated; blue, repressed. C, TGFβ response signature score (TBRS) in each response class (median ± IQ range, n = 4 tumors/model; Kruskal–Wallis test with Dunn multiple comparison correction). D, TSTSS in primary tumors (statistics as for C). E, Schematic for generation of response class-specific gene signatures. F and G, InhibMet (F) or StimMet (G) signature score in the different response classes. Mann–Whitney test. H and I, Scores for InhibMet (H) and StimMet (I) signatures in patients with breast cancer from the METABRIC dataset (n = 2,000 patients), stratified by hormone receptor status, intrinsic subtype, or histopathologic grade. ER status, t test; others, ANOVA with Tukey multiple comparison test. ER, n = 439; ER+, n = 1,498; basal, n = 209; Claudin-low, n = 218, HER2, n-224; luminal A, n = 700; luminal B, n = 475; normal-like, n = 148; grade 1, n = 169; grade 2, n = 771; grade 3, n = 952. J, Gene set enrichment analysis (GSEA) showing Hallmark gene sets that are similarly enriched in both responder groups. NES, normalized enrichment score. Results shown for gene sets with absolute NES > 2 in at least one group and difference in NES between groups <1.5. K, GSEA for Hallmark gene sets that are differentially enriched (difference in NES ≥ 3) in the two responder groups. L and M, Representative enrichment plots for the top most affected Hallmark gene set in each responder group. N, Summary of predicted biological differences between response classes.

Close modal

In an alternative approach to identify strategies for patient stratification, we generated weighted gene signatures from the most differentially expressed genes (FDR = 0.01 cutoff) in the StimMet class (StimMet signature) or InhibMet class (InhibMet signature) when each responder class was individually compared against the other two classes combined (Fig. 3E–G; Supplementary Table S2), and we used these signatures to query the METABRIC human breast cancer transcriptomic datasets (35). The InhibMet signature was more highly expressed in estrogen receptor–negative (ER) breast cancer, and in the basal and claudin-low transcriptomic subtypes, with expression increasing with increasing tumor grade (Fig. 3H). Conversely, the StimMet signature was more highly expressed in ER+ breast cancer, and in luminal A and luminal B tumors, with highest expression in grade 2 tumors (Fig. 3I). HER2 tumors showed relatively high expression of both signatures suggesting this class may be heterogeneous. Thus our data suggest that anti-TGFβ therapy in breast cancer might be most safely and effectively applied to patients with high grade, ER disease of the claudin-low and basal subtypes.

Transcriptomic analyses identify biological features associated with the different response classes

To gain insight into biological properties of tumors in the different response classes, we performed Hallmark gene set enrichment analysis. Here, we wished to highlight commonalities and differences between the two responder classes, so we compared the StimMet and InhibMet classes individually with the NoEff class. For commonalities, both responder classes exhibited transcriptomic evidence of increased TGFβ signaling, EMT, apoptosis, p53 pathway, and estrogen response, while having decreased MYC and E2F signaling (Fig. 3J). The enrichment for TGFβ signaling in both responder classes is as expected, given that TGFβ antagonism affects both classes, albeit in opposite ways. Transcriptomic approaches are clearly a more sensitive readout of TGFβ signaling than the single target biochemical approaches that we employed earlier. The low relative expression of MYC target genes in both responder classes may reflect the enrichment (P < 0.05, χ2 test) for Myc amplification in the NoEff class (Table 1). KRAS signaling was elevated in both responder classes, although the InhibMet class was enriched for genes upregulated by KRAS and the StimMet class for genes downregulated by KRAS. However, these transcriptomic differences did not correlate with the prevalence of mutant Kras, which was similar across all three response classes (Table 1), suggesting that Kras mutation status alone will not be predictive of response.

Despite these commonalities, the two responder classes also showed important differences from each other (Fig. 3K–M; summary in Fig. 3N). The InhibMet class was strongly enriched for gene sets associated with immunity and inflammation, suggesting that tumors in this class are immunologically active prior to treatment. In contrast, the StimMet class was distinguished by a gene set enrichment profile that reflected strongly suppressed mTORC1 signaling and a relatively inactive metabolic state. The predicted biological differences between the two responder groups give insights into possible differences in mechanism of action of the TGFβ antagonists in these groups (see below), which may be further exploitable for future biomarker development.

Metastasis-inhibiting effects of anti-TGFβ antibodies are immune-dependent while stimulating effects are immune-independent

Transcriptomes of primary tumors from InhibMet models were characterized by a high degree of immune activation and inflammation, and we confirmed that the inhibitory effect of the anti-TGFβ antibodies in the 4T1 InhibMet model was lost in fully immunodeficient NSG mice (Fig. 4A), consistent with therapeutic efficacy being dependent on unmasking effective antitumor immunity. We and others have previously shown that the efficacy of TGFβ antagonists in responsive breast cancer models is dependent on both T cells and natural killer cells (14, 36). In contrast, the stimulatory effect of TGFβ antagonism was immune-independent in the two StimMet models tested (Fig. 4B). The therapeutic response was not correlated with tumor cell neoantigen load, which was similar across response classes (Fig. 4C). To assess whether we could find other markers reflecting the different immune dependence of the response classes, we scored the models for a presence of a transcriptomic cytotoxic T-cell signature (37), and a pan-tumor, T-cell inflamed, IFNγ-driven gene expression signature that predicted clinical response to PD1 blockade (38). The InhibMet models showed highest scores for both signatures in their untreated primary tumors (Fig. 4D and E), and IHC staining also showed a trend toward increased density of CD8+ T cells in tumors from this response class (Fig. 4F and G). The results are consistent with the enrichment in InhibMet models of a T-cell–inflamed microenvironment, similar to that shown to be necessary but not sufficient for clinical response to immune checkpoint blockade (38). The density of other immune cell markers was similar across all model classes (Supplementary Fig. S8), with the exception of reduced F4/80+ cells in the InhibMet models (Fig. 4H), leading to a significantly higher CD8+/F480+ ratio in InhibMet tumors (Fig. 4I). Recent studies have suggested that anti-TGFβ therapy might be particularly effective in tumors with an “immune-excluded” phenotype, because stromally derived TGFβ can reduce immune cell infiltration into the tumor (4, 5). On the basis of CD8+ cell distribution in the primary tumors, we classified the models as immune-excluded (EMT6, TSAE1, D2A1, F3II), strongly infiltrated (MET1, R3T, E0771, MVT1), and weakly infiltrated/immune desert (4T1, 6DT1, HRM1, M6). There was a very weak trend toward InhibMet models being more infiltrated with CD8+ T cells (Fig. 4J). In summary, InhibMet models show evidence of preexisting immune activation, and the therapeutic efficacy of the anti-TGFβ antibodies is dependent on further unmasking of effective antitumor immunity. In the StimMet models, although there is transcriptomic evidence of some level of preexisting immune activation when compared with the NoEff models, this appears not to be limited by TGFβ, and the stimulatory effect of anti-TGFβ therapy in the StimMet models is immune-independent.

Figure 4.

Metastasis inhibition by anti-TGFβ antibodies is immune-dependent while stimulation is immune-independent. A and B, Effects of anti-TGFβ antibodies on metastasis endpoint in the 4T1 InhibMet model (A) or the MVT1 and M6 StimMet models (B) in immunocompetent or immunodeficient mice (NSG mice for 4T1 and M6; NOD/SCID for MVT1). Median ± IQ range for 9 to 15 mice/group. Mann–Whitney test. C, Predicted neoantigen burden in each model (median ± IQ range). D, CTL signature score (median ± IQ range, n = 4 tumors/model within response class, Dunn multiple comparison test). E, IFNγ signature score (statistics as for D). F, Representative images for CD8a and F4/80 immunostaining in a MET1 primary tumor. Scale bar, 200 μm. G and H, CD8a+ T-cell density (G) or F4/80+ macrophage density (H) in primary tumors determined by IHC (median ± IQ range, n = 3 tumors/model, Dunn multiple comparison test). I, Ratio of CD8+ to F480+ cells in primary tumors assessed by IHC (statistics as for G). J, Spatial distribution of CD8+ T cells in primary tumors assessed by IHC. A representative tumor for each distribution pattern is shown, with models showing that pattern listed below. Scale bar, 200 μm. * M6 was heterogeneous with 2 of 3 tumors having the immune desert phenotype and 1 of 3 being strongly infiltrated.

Figure 4.

Metastasis inhibition by anti-TGFβ antibodies is immune-dependent while stimulation is immune-independent. A and B, Effects of anti-TGFβ antibodies on metastasis endpoint in the 4T1 InhibMet model (A) or the MVT1 and M6 StimMet models (B) in immunocompetent or immunodeficient mice (NSG mice for 4T1 and M6; NOD/SCID for MVT1). Median ± IQ range for 9 to 15 mice/group. Mann–Whitney test. C, Predicted neoantigen burden in each model (median ± IQ range). D, CTL signature score (median ± IQ range, n = 4 tumors/model within response class, Dunn multiple comparison test). E, IFNγ signature score (statistics as for D). F, Representative images for CD8a and F4/80 immunostaining in a MET1 primary tumor. Scale bar, 200 μm. G and H, CD8a+ T-cell density (G) or F4/80+ macrophage density (H) in primary tumors determined by IHC (median ± IQ range, n = 3 tumors/model, Dunn multiple comparison test). I, Ratio of CD8+ to F480+ cells in primary tumors assessed by IHC (statistics as for G). J, Spatial distribution of CD8+ T cells in primary tumors assessed by IHC. A representative tumor for each distribution pattern is shown, with models showing that pattern listed below. Scale bar, 200 μm. * M6 was heterogeneous with 2 of 3 tumors having the immune desert phenotype and 1 of 3 being strongly infiltrated.

Close modal

Metastasis-stimulating effects of anti-TGFβ antibodies target the tumor cell

Because the metastasis-stimulating effects of TGFβ antagonism are immune-independent, we sought other cellular targets. TGFβs are known to have direct tumor suppressive effects on the tumor cell in the early stages of tumorigenesis (1). To address whether the StimMet models retain these tumor cell–targeted tumor-suppressive responses despite their advanced stage of progression, we knocked down TGFβ signaling in the MVT1 StimMet model by overexpression of a dominant negative type II TGFβ receptor (dnTGFBR2), or by silencing of SMAD2 or SMAD3 with shRNA. Blockade of the TGFβ receptor or knockdown of SMAD3, but not SMAD2, increased metastasis, suggesting that TGFβ signaling via the TGFBR2/SMAD3 axis has direct antimetastatic effects on the tumor cell in the MVT1 StimMet model (Fig. 5A–D). Thus, in the StimMet models, anti-TGFβ antibodies are likely promoting metastasis by interfering with metastasis-suppressive effects of TGFβ on the tumor parenchyma.

Figure 5.

StimMet models retain tumor cell-autonomous tumor-suppressive responses to TGFβ. A, Western blot showing expression of FLAG-tagged dominant negative TGFβ receptor (dnTGFBR2) and suppression of SMAD signaling in MVT1 cells. B, Effect on lung metastasis of TGFβ pathway blockade in MVT1 cells with a dominant negative TGFβ receptor (dnTGFBR2; median ± IQ range, n = 15 mice/group, Mann–Whitney test). C, Western blot showing shRNA knockdown of SMAD2 and SMAD3 in MVT1 cells. D, Effect on lung metastasis of TGFβ pathway blockade in MVT1 cells with shSMAD2 or shSMAD3. Median ± IQ range, n = 15 mice/group, Dunn multiple comparison test. E, Effect of TGFβ on cell proliferation in vitro in 3 representative models in the InhibMet and StimMet classes. Mean ± SD, n = 4, t test. F, Effect of TGFβ on clonogenicity in vitro. Mean ± SD, n = 3, t test. G, Schematic for tumorsphere formation assay. H, Effect of TGFβ on tumorsphere formation in vitro. Mean ± SD, n = 6.

Figure 5.

StimMet models retain tumor cell-autonomous tumor-suppressive responses to TGFβ. A, Western blot showing expression of FLAG-tagged dominant negative TGFβ receptor (dnTGFBR2) and suppression of SMAD signaling in MVT1 cells. B, Effect on lung metastasis of TGFβ pathway blockade in MVT1 cells with a dominant negative TGFβ receptor (dnTGFBR2; median ± IQ range, n = 15 mice/group, Mann–Whitney test). C, Western blot showing shRNA knockdown of SMAD2 and SMAD3 in MVT1 cells. D, Effect on lung metastasis of TGFβ pathway blockade in MVT1 cells with shSMAD2 or shSMAD3. Median ± IQ range, n = 15 mice/group, Dunn multiple comparison test. E, Effect of TGFβ on cell proliferation in vitro in 3 representative models in the InhibMet and StimMet classes. Mean ± SD, n = 4, t test. F, Effect of TGFβ on clonogenicity in vitro. Mean ± SD, n = 3, t test. G, Schematic for tumorsphere formation assay. H, Effect of TGFβ on tumorsphere formation in vitro. Mean ± SD, n = 6.

Close modal

To address mechanisms underlying the metastasis suppressive effects of TGFβ on the tumor cell, we took three models in each of the InhibMet and StimMet classes and assessed their biological responses to TGFβ in vitro. There was no correlation between the response to anti-TGFβ antibodies in vivo and effects of TGFβ on tumor cells in vitro with respect to growth inhibition (Fig. 5E), cell survival, migration, or invasion (Supplementary Fig. S9A–S9C). In contrast, TGFβ consistently inhibited clonogenicity and tumorsphere formation by the StimMet models in vitro while having little or no effect on the InhibMet models (Fig. 5F–H). Note that for the tumorsphere assay, tumor cells were pretreated with TGFβ but tumorsphere formation was assessed in the absence of TGFβ, so as to disambiguate effects of TGFβ specifically on the cancer stem cell from more generalized effects on proliferation of all tumor cells. Because tumorsphere formation is a surrogate assay for cancer stem cell activity, the data suggested that TGFβ may have inhibitory effects on the CSC population in the StimMet models, and that TGFβ antibodies release the brakes on CSC expansion. This lead will be pursued in future studies.

TGFβs have many tumor-promoting activities, including immunosuppressive effects on multiple immune cell targets (1). Recent clinical successes with immune checkpoint inhibitors targeting CTLA4, PD1, or PD-L1 (39) have fueled a renewed surge of interest in targeting the TGFβ pathway, as an alternative or complementary approach to reactivating effective antitumor immunity. TGFβ pathway antagonists of various types, including neutralizing antibodies (fresolimumab, Sanofi-Aventis; NIS793, Novartis), TGFβ traps (AVID200, Forbius), small-molecule receptor kinase inhibitors (galunisertib, Eli Lilly) and a bifunctional anti–PD-L1/TGFβ trap (M7824, GlaxoSmithKline/Merck KGaA) are in early-phase clinical oncology trials. They have been generally well tolerated, with some early signs of efficacy (19–21). In breast cancer, patients with metastatic disease who received the highest dose of fresolimumab in combination with focal radiotherapy had a favorable systemic immune response and longer median overall survival compared with the low-dose arm (40). Five additional trials in breast cancer with TGFβ antagonists are ongoing (clinicaltrials.gov).

However, despite these encouraging results, the TGFβ pathway remains an unusually complex signaling axis to target clinically. Not only is TGFβ known to have tumor suppressive as well as prooncogenic effects, but nearly every cell type in the tumor ecosystem responds to TGFβ in highly context-dependent ways, making the integrated outcome of these effects in vivo is difficult to predict (1). In this study, we have used a large panel of mouse models of metastatic breast cancer to explore whether preclinical therapy studies that capture more disease heterogeneity can give new insights to better support and inform clinical trials that target this highly contextual regulatory network. Our test agent was a pan-TGFβ–neutralizing antibody, but we anticipate that our results may be more broadly applicable to other pathway antagonists.

Previous preclinical work on a smaller scale has shown antitumor efficacy of TGFβ antagonists in multiple tumor types (1, 11–14, 16, 17, 41). Consistent with these results, a plurality (5/12 or 44%) of models in our panel responded to TGFβ antagonism with the desired reduction in metastatic burden. Of these models, four had been independently shown to respond to TGFβ antagonism in other labs, either as allografts (4T1, EMT6, R3T; refs. 13–15) or in the original GEM model (MMTV-PyVT for MET1; ref. 13). Crucially however, use of the expanded panel also revealed adverse metastasis-promoting responses to anti-TGFβ in 25% of the models. Although this effect had not previously been seen with pharmacologic TGFβ pathway inhibition in preclinical studies, tumor cell-specific genetic ablation of the TGFβ pathway enhanced metastasis in the MMTV-PyVT transgenic mouse model of metastatic breast cancer (42). Thus, the mouse model data suggests that, contrary to dogma, tumor-suppressive responses to TGFβ may be retained and dominant in some instances of advanced metastatic breast cancer. We previously showed that high expression of a gene signature specifically reflecting tumor-suppressive effects of TGFβ was associated with improved metastasis-free survival in clinical breast cancer cohorts (31), which supports the possibility that retention of TGFβ tumor-suppressive responses may also be a feature of the human disease. In the context of ongoing clinical trials with TGFβ pathway antagonists, our data suggest that good predictive biomarkers will be critical for safe and effective use of these agents, not just to identify patients who will respond therapeutically, but more importantly to eliminate those patients who are at risk for adverse on-target responses.

To address this need, we used the mouse model panel as a platform for identification of potential predictive biomarkers. A summary of our results across multiple approaches is given in Supplementary Fig. S10, with potential leads highlighted for further exploration. Broadly, we found that at the larger scale of the current study, many individual biomarkers that had looked promising from small studies did not perform well. We were unable to find a robust correlation between response-to-therapy and any individual parameter relating to TGFβ ligand expression or downstream signaling events in the tumor. Similar observations have recently been made in skin cancer models treated with a TGFβ receptor kinase inhibitor (11). Furthermore, published transcriptomic signatures of TGFβ response, and the expression levels or mutation status of genes that were previously proposed to be involved in the switch of TGFβ from tumor suppressor to prometastatic factor (e.g., Dab2, Klf5, Peak1, Pspc1, Rassf1, Six1, mutant Tp53), did not robustly correlate with response to anti-TGFβ therapy.

Moving beyond the candidate approach, we looked at transcriptomes from treatment-naïve primary tumors for biomarker discovery, reasoning that targeted transcriptomic information might be feasible to acquire in a clinical setting. Interestingly, the tumor transcriptomes of models in the two responder classes (InhibMet and StimMet) were much more similar to each other than to the nonresponder (NoEff) class, which poses challenges for discriminating between them. The association of Myc gene amplification and increased MYC pathway activation with lack of therapeutic response in the NoEff class may be worth further assessment. Interestingly, both responder classes were enriched for gene sets relating to increased TGFβ signaling when compared with the nonresponder class, although our biochemical assays of the signaling pathway had shown no clear differences between them. Thus, in general, the transcriptomic approach, which integrates signal over multiple targets, may be more robust to interindividual heterogeneity than assessment of single targets. Where practical, acquisition of pretreatment tumor transcriptomics in the early-phase clinical trials might provide the biggest return on investment for biomarker identification.

Although treatment-naïve primary tumors from the two responder classes were transcriptomically similar, we were able to generate gene signatures that individually distinguished the InhibMet or StimMet response classes. Applying those signatures to the human breast cancer datasets, we showed that the InhibMet signature is significantly enriched in ER breast cancer, particularly in the claudin-low and basal intrinsic subtypes, and in tumors of higher grade. Conversely, the StimMet signature was significantly enriched in ER+ tumors, with highest expression in luminal A, luminal B, and normal-like tumors. The HER2-intrinsic subtype showed high expression of both signatures, possibly reflecting heterogeneity within this subtype. Thus, based on this analysis, we would predict that breast cancer patients most likely to benefit from anti-TGFβ therapy would be those with high grade, ER tumors, particularly of the claudin-low or basal subtypes. As data become available from ongoing clinical trials with TGFβ antagonists in breast cancer, it will be important to validate these predictions.

A number of limitations of this study need to be acknowledged. The panel has no models of HER2+ disease, and in general, the mapping of the mouse models onto human breast cancer subtypes is complex (24). Nevertheless, the mouse model studies allowed us to identify transcriptomic features of the treatment-naïve primary tumors that are associated with stimulatory or inhibitory responses to TGFβ antagonism, and to show that these features are enriched in specific human breast cancer subtypes. The paucity of human ER+ breast cancer cell lines that metastasize efficiently makes confirmation of the StimMet responses using human cell lines challenging currently, but may become possible with the development of metastatic patient-derived xenografts. A second major caveat is that we used anti-TGFβ antibodies as monotherapy, whereas they will likely be used in combination therapies in the clinic. However, we still see the undesirable metastasis-stimulating response to anti-TGFβ treatment in the MVT1 StimMet model when used in combination with otherwise efficacious doses of cyclophosphamide (Supplementary Fig. S11), so our results are likely also relevant in a combination therapy setting. Finally, in three of the models (TSAE1, MET1, and F3II), metastasis was established by tail vein injection rather than orthotopic implantation of tumor cells. We have previously shown in the 4T1 model that TGFβ antagonism has identical activity in both assay formats (14), suggesting that TGFβ antagonism primarily affects metastatic colonization rather than dissemination from the primary tumor, so this limitation may not be a major concern.

One place where mouse models can make a unique contribution is to understanding therapeutic mechanisms. We confirmed earlier findings by ourselves and others (1, 14–16, 43), that the therapeutic effect of TGFβ antagonists in the 4T1 InhibMet model is dependent on an intact immune system. The untreated primary tumors from the InhibMet models were highly enriched for gene sets and signatures relating to inflammation and immune activation, suggesting that the desired inhibitory effect TGFβ antagonism as monotherapy was seen most strongly in the context of some level of pre-existing immune response. In contrast, StimMet tumors showed a weaker enrichment of gene sets for inflammation and immunity, and we showed that the undesirable metastasis-stimulating effect of TGFβ antagonism in StimMet models was fully immune-independent. Notably, the StimMet tumor transcriptomes showed underexpression of pathways relating to metabolism and mTORC1 pathway activation. Because the mTOR pathway is a metabolite and nutrient sensor, and mTORC1 inhibitors such as rapamycin are immunosuppressive, in part by inhibiting expansion of effector T cells (44), it is conceivable that StimMet tumors create a depleted metabolic environment that represses mTORC1 signaling, and hence prevents T-cell expansion and effective antitumor immunity by mechanisms that are independent of TGFβ. Furthermore, direct interactions between TGFβ and mTOR signaling pathways have been observed in other settings and may be worth pursuing (45). In the StimMet models, we found that metastasis-stimulating effects of the anti-TGFβ antibodies were due to interference with tumor-suppressive effects of TGFβ on the tumor cell, with our in vitro data suggesting that the cancer stem cell subpopulation may be a key target of these effects. Although most studies suggest a stimulatory effect of TGFβ on breast cancer stem cells (reviewed in ref. 46), we and others have previously shown that TGFβ can inhibit the cancer stem cell population in select breast cancer models (47, 48), Further study will be necessary to determine the detailed underlying mechanisms.

In summary, our preclinical study has shown that the assumptions that tumor-suppressive effects of TGFβ are either lost in advanced breast cancer, or are not susceptible to TGFβ antagonism, are not valid in a significant minority of cases. This raises the specter that some patients on anti-TGFβ therapy will have their disease course accelerated. Although we do not know whether this phenomenon will also be observed for cancers other than breast, or for other classes of TGFβ pathway antagonists, we suspect that it will be the case. A complex balance between tumor-suppressive effects of TGFβ on the tumor cell and tumor-promoting effects of TGFβ on the immune stroma determines the outcome of anti-TGFβ therapy. Larger scale preclinical studies, such as the one we performed here for breast cancer, may help inform the safe use of TGFβ antagonists in other cancer histologies. Although the TGFβ field was alert to the possibility of adverse outcomes with TGFβ antagonism, it is important to note that undesired stimulatory effects are also evident with other targeted therapies. Disease hyperprogression in response to treatment with immune checkpoint inhibitors (ICI) has been seen in several clinical studies, with incidence rates ranging from 9% to 29% of patients enrolled (49). Interestingly, gene expression profiling of pretreatment tumors in one clinical study revealed an underexpression of pathways associated with cell metabolism in the hyperprogressors (50). Because we saw underexpression of these same pathways in the StimMet class of tumors, there may be mechanistic commonalities between hyperprogression on ICIs and on anti-TGFβ therapy that could be worth pursuing. Furthermore, given that we saw higher macrophage infiltration in StimMet and NoEff tumors, it is also intriguing that tumor-associated macrophages have been implicated in hyperprogression on, and resistance to, immune checkpoint inhibitors (50). The bottom line is that good predictive biomarkers will be crucial to the safe and effective deployment of TGFβ pathway antagonists, and that preclinical studies designed to capture more of the disease heterogeneity can provide useful information to complement and guide the clinical trials.

No potential conflicts of interest were disclosed.

Conception and design: Y. Yang, D.S. Weinberg, L.M. Wakefield

Development of methodology: Y. Yang, H.H. Yang, B. Tang, N. Moshkovich, J. Chen, R.M. Simpson, L.M. Wakefield

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): Y. Yang, B. Tang, K.C. Flanders, D.S. Weinberg, M.A. Welsh, J. Weng, T.Y. Hu, M.A. Herrmann, J. Chen, R.M. Simpson, L.M. Wakefield

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): Y. Yang, H.H. Yang, K.C. Flanders, J. Weng, T.Y. Hu, M.A. Herrmann, J. Chen, E.F. Edmondson, R.M. Simpson, F. Liu, H. Liu, M.P. Lee, L.M. Wakefield

Writing, review, and/or revision of the manuscript: Y. Yang, H.H. Yang, A.M.L. Wu, N. Moshkovich, D.S. Weinberg, H.J. Ochoa, R.M. Simpson, F. Liu, M.P. Lee, L.M. Wakefield

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): Y. Yang, A.M.L. Wu, H.J. Ochoa, F. Liu, L.M. Wakefield

Study supervision: Y. Yang, L.M. Wakefield

We acknowledge the expert technical assistance of Anthony Vieira, Elena Kuznetsova, Maria Figueroa, and Geneti Gaga (Laboratory Animal Sciences Program, NCI) with the animal experiments, and of Dr. Xiaolin Wu (Laboratory of Molecular Technology, Frederick National Laboratory for Cancer Research) with the microarray analysis. This work was supported by funding from the Intramural Research Program of the National Cancer Institute Center for Cancer Research, NIH project ZIA BC 010881 (to L.M. Wakefield).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Akhurst
RJ
,
Hata
A
. 
Targeting the TGFβ signalling pathway in disease
.
Nat Rev Drug Discov
2012
;
11
:
790
811
.
2.
Massague
J
. 
TGFβ in cancer
.
Cell
2008
;
134
:
215
30
.
3.
Seoane
J
,
Gomis
RR
. 
TGF-β family signaling in tumor suppression and cancer progression
.
Cold Spring Harb Perspect Biol
2017
;
9
:
pii
:
a022277
.
4.
Mariathasan
S
,
Turley
SJ
,
Nickles
D
,
Castiglioni
A
,
Yuen
K
,
Wang
Y
, et al
TGFβ attenuates tumour response to PD-L1 blockade by contributing to exclusion of T cells
.
Nature
2018
;
554
:
544
8
.
5.
Tauriello
DVF
,
Palomo-Ponce
S
,
Stork
D
,
Berenguer-Llergo
A
,
Badia-Ramentol
J
,
Iglesias
M
, et al
TGFβ drives immune evasion in genetically reconstituted colon cancer metastasis
.
Nature
2018
;
554
:
538
43
.
6.
Batlle
E
,
Massague
J
. 
Transforming growth factor-β signaling in immunity and cancer
.
Immunity
2019
;
50
:
924
40
.
7.
Miao
Y
,
Yang
H
,
Levorse
J
,
Yuan
S
,
Polak
L
,
Sribour
M
, et al
Adaptive immune resistance emerges from tumor-initiating stem cells
.
Cell
2019
;
177
:
1172
86
.
8.
Brown
JA
,
Yonekubo
Y
,
Hanson
N
,
Sastre-Perona
A
,
Basin
A
,
Rytlewski
JA
, et al
TGF-β-induced quiescence mediates chemoresistance of tumor-propagating cells in squamous cell carcinoma
.
Cell Stem Cell
2017
;
21
:
650
64
.
9.
Hardee
ME
,
Marciscano
AE
,
Medina-Ramirez
CM
,
Zagzag
D
,
Narayana
A
,
Lonning
SM
, et al
Resistance of glioblastoma-initiating cells to radiation mediated by the tumor microenvironment can be abolished by inhibiting transforming growth factor-β
.
Cancer Res
2012
;
72
:
4119
29
.
10.
Biswas
S
,
Guix
M
,
Rinehart
C
,
Dugger
TC
,
Chytil
A
,
Moses
HL
, et al
Inhibition of TGF-β with neutralizing antibodies prevents radiation-induced acceleration of metastatic cancer progression
.
J Clin Invest
2007
;
117
:
1305
13
.
11.
Dodagatta-Marri
E
,
Meyer
DS
,
Reeves
MQ
,
Paniagua
R
,
To
MD
,
Binnewies
M
, et al
alpha-PD-1 therapy elevates Treg/Th balance and increases tumor cell pSMAD3 that are both targeted by anti-TGFβ antibody to promote durable rejection and immunity in squamous cell carcinomas
.
J Immunother Cancer
2019
;
7
:
62
.
12.
Yang
YA
,
Dukhanina
O
,
Tang
B
,
Mamura
M
,
Letterio
JJ
,
MacGregor
J
, et al
Lifetime exposure to a soluble TGF-β antagonist protects mice against metastasis without adverse side effects
.
J Clin Invest
2002
;
109
:
1607
15
.
13.
Muraoka
RS
,
Dumont
N
,
Ritter
CA
,
Dugger
TC
,
Brantley
DM
,
Chen
J
, et al
Blockade of TGF-β inhibits mammary tumor cell viability, migration, and metastases
.
J Clin Invest
2002
;
109
:
1551
9
.
14.
Nam
JS
,
Terabe
M
,
Mamura
M
,
Kang
MJ
,
Chae
H
,
Stuelten
C
, et al
An anti-transforming growth factor β antibody suppresses metastasis via cooperative effects on multiple cell compartments
.
Cancer Res
2008
;
68
:
3835
43
.
15.
Ge
R
,
Rajeev
V
,
Ray
P
,
Lattime
E
,
Rittling
S
,
Medicherla
S
, et al
Inhibition of growth and metastasis of mouse mammary carcinoma by selective inhibitor of transforming growth factor-β type I receptor kinase in vivo
.
Clin Cancer Res
2006
;
12
(
14 Pt 1
):
4315
30
.
16.
Yingling
JM
,
McMillen
WT
,
Yan
L
,
Huang
H
,
Sawyer
JS
,
Graff
J
, et al
Preclinical assessment of galunisertib (LY2157299 monohydrate), a first-in-class transforming growth factor-β receptor type I inhibitor
.
Oncotarget
2018
;
9
:
6659
77
.
17.
Ostapoff
KT
,
Cenik
BK
,
Wang
M
,
Ye
R
,
Xu
X
,
Nugent
D
, et al
Neutralizing murine TGFβR2 promotes a differentiated tumor cell phenotype and inhibits pancreatic cancer metastasis
.
Cancer Res
2014
;
74
:
4996
5007
.
18.
Vanpouille-Box
C
,
Diamond
JM
,
Pilones
KA
,
Zavadil
J
,
Babb
JS
,
Formenti
SC
, et al
TGFβ is a master regulator of radiation therapy-induced antitumor immunity
.
Cancer Res
2015
;
75
:
2232
42
.
19.
Strauss
J
,
Heery
CR
,
Schlom
J
,
Madan
RA
,
Cao
L
,
Kang
Z
, et al
Phase I trial of M7824 (MSB0011359C), a bifunctional fusion protein targeting PD-L1 and TGFβ, in advanced solid tumors
.
Clin Cancer Res
2018
;
24
:
1287
95
.
20.
Morris
JC
,
Tan
AR
,
Olencki
TE
,
Shapiro
GI
,
Dezube
BJ
,
Reiss
M
, et al
Phase I study of GC1008 (fresolimumab): a human anti-transforming growth factor-β (TGFβ) monoclonal antibody in patients with advanced malignant melanoma or renal cell carcinoma
.
PLoS One
2014
;
9
:
e90353
.
21.
Ikeda
M
,
Takahashi
H
,
Kondo
S
,
Lahn
MMF
,
Ogasawara
K
,
Benhadji
KA
, et al
Phase 1b study of galunisertib in combination with gemcitabine in Japanese patients with metastatic or locally advanced pancreatic cancer
.
Cancer Chemother Pharmacol
2017
;
79
:
1169
77
.
22.
Begley
CG
,
Ellis
LM
. 
Drug development: raise standards for preclinical cancer research
.
Nature
2012
;
483
:
531
3
.
23.
Ioannidis
JPA
,
Kim
BYS
,
Trounson
A
. 
How to design preclinical studies in nanomedicine and cell therapy to maximize the prospects of clinical translation
.
Nat Biomed Eng
2018
;
2
:
797
809
.
24.
Yang
Y
,
Yang
HH
,
Hu
Y
,
Watson
PH
,
Liu
H
,
Geiger
TR
, et al
Immunocompetent mouse allograft models for development of therapies to target breast cancer metastasis
.
Oncotarget
2017
;
8
:
30621
43
.
25.
Grutter
C
,
Wilkinson
T
,
Turner
R
,
Podichetty
S
,
Finch
D
,
McCourt
M
, et al
A cytokine-neutralizing antibody as a structural mimetic of 2 receptor interactions
.
Proc Natl Acad Sci U S A
2008
;
105
:
20251
6
.
26.
Derynck
R
,
Budi
EH
. 
Specificity, versatility, and control of TGF-β family signaling
.
Sci Signal
2019
;
12
:
pii:eaav5183
.
27.
Gronroos
E
,
Kingston
IJ
,
Ramachandran
A
,
Randall
RA
,
Vizan
P
,
Hill
CS
. 
Transforming growth factor β inhibits bone morphogenetic protein-induced transcription through novel phosphorylated SMAD1/5-SMAD3 complexes
.
Mol Cell Biol
2012
;
32
:
2904
16
.
28.
Daly
AC
,
Randall
RA
,
Hill
CS
. 
Transforming growth factor β-induced SMAD1/5 phosphorylation in epithelial cells is mediated by novel receptor complexes and is essential for anchorage-independent growth
.
Mol Cell Biol
2008
;
28
:
6889
902
.
29.
Flanders
KC
,
Heger
CD
,
Conway
C
,
Tang
B
,
Sato
M
,
Dengler
SL
, et al
Brightfield proximity ligation assay reveals both canonical and mixed transforming growth factor-β/bone morphogenetic protein SMAD signaling complexes in tissue sections
.
J Histochem Cytochem
2014
;
62
:
846
63
.
30.
Padua
D
,
Zhang
XH
,
Wang
Q
,
Nadal
C
,
Gerald
WL
,
Gomis
RR
, et al
TGFβ primes breast tumors for lung metastasis seeding through angiopoietin-like 4
.
Cell
2008
;
133
:
66
77
.
31.
Sato
M
,
Kadota
M
,
Tang
B
,
Yang
HH
,
Yang
YA
,
Shan
M
, et al
An integrated genomic approach identifies persistent tumor suppressive effects of transforming growth factor-β in human breast cancer
.
Breast Cancer Res
2014
;
16
:
R57
.
32.
Yeh
HW
,
Lee
SS
,
Chang
CY
,
Lang
YD
,
Jou
YS
. 
A new switch for TGFβ in cancer
.
Cancer Res
2019
;
79
:
3797
805
.
33.
Xu
J
,
Acharya
S
,
Sahin
O
,
Zhang
Q
,
Saito
Y
,
Yao
J
, et al
14-3-3zeta turns TGF-β's function from tumor suppressor to metastasis promoter in breast cancer by contextual changes of SMAD partners from p53 to Gli2
.
Cancer Cell
2015
;
27
:
177
92
.
34.
Adorno
M
,
Cordenonsi
M
,
Montagner
M
,
Dupont
S
,
Wong
C
,
Hann
B
, et al
A Mutant-p53/SMAD complex opposes p63 to empower TGFβ-induced metastasis
.
Cell
2009
;
137
:
87
98
.
35.
Curtis
C
,
Shah
SP
,
Chin
SF
,
Turashvili
G
,
Rueda
OM
,
Dunning
MJ
, et al
The genomic and transcriptomic architecture of 2,000 breast tumours reveals novel subgroups
.
Nature
2012
;
486
:
346
52
.
36.
Zhong
Z
,
Carroll
KD
,
Policarpio
D
,
Osborn
C
,
Gregory
M
,
Bassi
R
, et al
Anti-transforming growth factor β receptor II antibody has therapeutic efficacy against primary tumor growth and metastasis through multieffects on cancer, stroma, and immune cells
.
Clin Cancer Res
2010
;
16
:
1191
205
.
37.
Jiang
P
,
Gu
S
,
Pan
D
,
Fu
J
,
Sahu
A
,
Hu
X
, et al
Signatures of T cell dysfunction and exclusion predict cancer immunotherapy response
.
Nat Med
2018
;
24
:
1550
8
.
38.
Ayers
M
,
Lunceford
J
,
Nebozhyn
M
,
Murphy
E
,
Loboda
A
,
Kaufman
DR
, et al
IFN-gamma-related mRNA profile predicts clinical response to PD-1 blockade
.
J Clin Invest
2017
;
127
:
2930
40
.
39.
Ribas
A
,
Wolchok
JD
. 
Cancer immunotherapy using checkpoint blockade
.
Science
2018
;
359
:
1350
5
.
40.
Formenti
SC
,
Lee
P
,
Adams
S
,
Goldberg
JD
,
Li
X
,
Xie
MW
, et al
Focal irradiation and systemic TGFβ blockade in metastatic breast cancer
.
Clin Cancer Res
2018
;
24
:
2493
504
.
41.
Ganapathy
V
,
Ge
R
,
Grazioli
A
,
Xie
W
,
Banach-Petrosky
W
,
Kang
Y
, et al
Targeting the transforming growth factor-β pathway inhibits human basal-like breast cancer metastasis
.
Mol Cancer
2010
;
9
:
122
.
42.
Forrester
E
,
Chytil
A
,
Bierie
B
,
Aakre
M
,
Gorska
AE
,
Sharif-Afshar
AR
, et al
Effect of conditional knockout of the type II TGF-β receptor gene in mammary epithelia on mammary gland development and polyomavirus middle T antigen induced tumor formation and metastasis
.
Cancer Res
2005
;
65
:
2296
302
.
43.
Akhurst
RJ
. 
Targeting TGF-β signaling for therapeutic gain
.
Cold Spring Harb Perspect Biol
2017
;
9
:
pii
:
a022301
.
44.
Waickman
AT
,
Powell
JD
. 
mTOR, metabolism, and the regulation of T-cell differentiation and function
.
Immunol Rev
2012
;
249
:
43
58
.
45.
Asrani
K
,
Sood
A
,
Torres
A
,
Georgess
D
,
Phatak
P
,
Kaur
H
, et al
mTORC1 loss impairs epidermal adhesion via TGF-β/Rho kinase activation
.
J Clin Invest
2017
;
127
:
4001
17
.
46.
Bellomo
C
,
Caja
L
,
Moustakas
A
. 
Transforming growth factor β as regulator of cancer stemness and metastasis
.
Br J Cancer
2016
;
115
:
761
9
.
47.
Tang
B
,
Yoo
N
,
Vu
M
,
Mamura
M
,
Nam
JS
,
Ooshima
A
, et al
Transforming growth factor-β can suppress tumorigenesis through effects on the putative cancer stem or early progenitor cell and committed progeny in a breast cancer xenograft model
.
Cancer Res
2007
;
67
:
8643
52
.
48.
Bruna
A
,
Greenwood
W
,
Le
QJ
,
Teschendorff
A
,
Miranda-Saavedra
D
,
Rueda
OM
, et al
TGFβ induces the formation of tumour-initiating cells in claudin-low breast cancer
.
Nat Commun
2012
;
3
:
1055
.
49.
Knorr
DA
,
Ravetch
JV
. 
Immunotherapy and hyperprogression: unwanted outcomes, unclear mechanism
.
Clin Cancer Res
2019
;
25
:
904
6
.
50.
Lo Russo
G
,
Moro
M
,
Sommariva
M
,
Cancila
V
,
Boeri
M
,
Centonze
G
, et al
Antibody-Fc/FcR interaction on macrophages as a mechanism for hyperprogressive disease in non-small cell lung cancer subsequent to PD-1/PD-L1 blockade
.
Clin Cancer Res
2019
;
25
:
989
99
.