Patient-Derived Xenograft Engraftment and Breast Cancer Outcomes in a Prospective Neoadjuvant Study (BEAUTY)

Patient-derived xenografts (PDX) recapitulate the cell morphology, architecture, microenvironment, and molecular signatures of patient tumors (1, 2), and additionally faithfully reproduce drug response phenotypes seen in patients (2). Thus, PDXs are a standard research tool used across multiple different solid tumors.

While many factors affect PDX engraftment, including the amount of tissue injected, site of injection (subcutaneous vs. orthotopic), and mouse strain, the aggressiveness of the patient malignancy has been associated with higher engraftment. For example, in breast cancer, higher engraftment are seen in tumors with higher grade and estrogen receptor (ER)-negative status (2). These observations have led to the hypothesis that engraftment may be prognostic for disease outcomes. For example, a retrospective study of 24 patients with newly diagnosed breast cancer without prior cancer treatment and followed for a median time of 28 months found engraftment to be associated with relapse (3).

We previously reported results of the prospective multi-center Breast Cancer Genome Guided Therapy Study (BEAUTY) clinical trial (4). Of the 140 patients enrolled, 113 were treated at centers with established PDX programs and had percutaneous tumor biopsies (PTB) obtained for PDX prior to neoadjuvant chemotherapy (NAC; weekly taxane ± trastuzumab followed by anthracycline-based chemotherapy) and 34 had surgical tissue samples obtained from resection following chemotherapy (4, 5). In this prospective study, we obtained 6–8 image-guided percutaneous biopsies from the tumor (>90% were T2 or larger) and the cores were used for pathologic confirmation, tumor sequencing, and PDX generation. Within an hour of sample collection, one to two cores were implanted with Matrigel (BD Biosciences) in the flanks of 6- to 8-week-old female immunodeficient mice which were pretreated and maintained with 17β-estradiol. Samples were implanted subcutaneously, and mice were monitored daily. The process handling was consistent across the entire study period. A histologic evaluation of the tissue was not performed to assess for necrosis prior to implanting the tumor.

We defined PDX engraftment as the percent of patients with at least one stably transplantable xenograft pathologically confirmed as breast cancer and passed at least for four generations. Pre-NAC PDX engraftment was 27.4% (31/113), and differs significantly by clinical molecular subtype [triple-negative breast cancer (TNBC), 51.3% (20/39); HER2⁺, 26.5% (9/34); and ER⁺/HER2⁻ at 5.0% (2/40)] as well as grade and type of mouse strain (NSG vs. NOD-SCID mice), but not by baseline tumor Ki-67 (4, 5). Chemotherapy response [pathological complete response (pCR)] was not found to differ significantly by pre-NAC engraftment (5). Herein, we report on the association between PDX engraftment and breast cancer outcomes.

With a median follow-up of 5.7 years, local, regional, or distant disease events (either disease progression during NAC or recurrence after NAC) were reported in 17 patients. Gray test was used to assess whether the cumulative incidence of a breast cancer event (CI-BCE), differs with respect to either pre-NAC PDX engraftment or post-NAC PDX engraftment. Breast cancer events (outcomes) were defined as progression on chemotherapy prior to surgery or local, regional, or distant recurrence after surgery. Contralateral breast cancer and second primary disease were considered competing events. Time at risk started the day of registration (at most 21 days prior to start of NAC) when assessing pre-NAC PDX engraftment and the day of surgery when assessing post-NAC PDX engraftment. The CI-BCE was not found to differ according to pre-NAC PDX engraftment [5-year rate (95% confidence interval, CI): 13.6% (7.2%–22.1%) no engraftment; 13.4% (4.1%–28.2%) engraftment; P = 0.89; Fig. 1A].

In the 39 patients with TNBC, the subgroup with the highest pre-NAC engraftment, CI-BCE was not found to differ significantly between those without PDX engraftment [5-year rate (CI): 21.4% (6.3%–42.3%)] and those with PDX engraftment [16.2% (3.8%–36.4%), P = 0.73; Fig. 1B].

We previously evaluated the association between mouse strain and engraftment and found that pre-NAC engraftment rates were higher in NSG versus NOD-SCID (65.4% vs. 25.5%; P = 0.001; ref. 5). Therefore, we determined whether CI-BCE differed with respect to engraftment for each mouse strain. CI-BCE was not found to differ significantly between patients with pre-NAC PDX engraftment and patients without pre-NAC PDX engraftment according to type of mouse strain used. In patients from whom tumors were injected into the NOD-SCID mice, the 5-year CI-BCE rate was 7.7% (0.4%–30.3%) in those with pre-NAC engraftment and 13.7% (6.3%–23.8%) in those without pre-NAC engraftment (P = 0.65). In patients from whom tumors were injected into the NSG mice, the 5-year CI-BCE rate was 16.7% (3.9%–37.2%) in those with pre-NAC PDX engraftment and 13.6% (3.3%–31.4%) in those without pre-NAC engraftment (P = 0.99; Fig. 1C and D).

PDXs were established from residual tissue from surgery (post-NAC) in 6 of 34 patients (17.6%), specifically, TNBC: 5/9, Her2⁺: 1/8, and luminal: 0/17. The CI-BCE was higher among those with post-NAC PDX engraftment relative to those without post-NAC PDX engraftment [5-year rate (CI): 19.6% (6.9%–37.0%) no engraftment; 50.0% (7.7%–82.9%) engraftment] but this did not reach statistical significance (P = 0.11; Fig. 2).

The process of generating a PDX model results in the selection of tumors that engraft and propagate in mice. As demonstrated in the BEAUTY study as well as in prior reports, more aggressive tumors have a higher engraftment (2, 6, 7). A series of prior studies have reported that tumor engraftment was prognostic for recurrence and survival (8–11). However, these studies are small series of retrospectively collected patients who received a variety of treatments who had sufficient tumor (from breast, nodes, or metastatic site) to attempt implantation. In this prospective study collecting PTB from patients receiving standard-of-care anthracycline and taxane-based chemotherapy, including trastuzumab for HER2⁺ breast cancer, we found that the CI-BCE did not differ according to success of pre-NAC tumor engraftment into an immunocompromised mouse. Because tumor engraftment tends to be highest in TNBC, we further analyzed the association between tumor engraftment and outcome. Again, in this group, the CI-BCE between those with and without pre-NAC PDX engraftment did not significantly differ (P = 0.73).

In contrast, in patients with residual disease following NAC whose tumors exhibited chemotherapy resistance and who are at higher risk of local and distant recurrence (compared with those with pathologic complete response), post-NAC PDX engraftment tended to not only be associated with tumor subtype but also with a higher CI-BCE. These data suggest that for chemotherapy-resistant tumors that remain in the breast at the time of surgical resection, engraftment may identify a subset of patients at greater risk of recurrence.

Taken together, PDXs derived from pre-NAC and post-NAC settings provides an important tool to interrogate cancer biology and drug response. However, PDX engraftment as a biomarker of prognosis has not been definitively validated and remains a research question, in part due to the limited ability to study only in larger academic centers and lack of standardization of the methodologies surrounding PDX engraftment. Given the limitations of this study size, overall take rate and event rate, larger cohorts of PDXs collected in a standardized fashion from patients homogenously treated with appropriate therapies are needed to establish whether PDX engraftment in both treatment-naïve and treatment-resistant settings is prognostic for cancer recurrence.

J.C. Boughey reports grants from benefactor funding, including private foundations as listed in the manuscript, during the conduct of the study, as well as grants from Lilly outside the submitted work. V.J. Suman reports grants from NIH during the conduct of the study, as well as grants from NIH, DOD, American Association for Cancer Research, and ASCO's Conquer Cancer Foundation outside the submitted work; in addition, V.J. Suman has a patent for Methods and Materials for Assessing Chemotherapy Responsiveness and Treating Cancer issued. J. Yu reports other support from Wuxi AppTec Co. outside the submitted work. A. Moreno-Aspitia reports grants from NIH during the conduct of the study, as well as other support from Wuxi licensed PDX mouse models outside the submitted work. R.J. Gray reports a patent for Patient-Derived Xenograft Methodology issued and with royalties paid. J.N. Ingle reports grants from NIH during the conduct of the study, as well as grants from The Breast Cancer Research Foundation outside the submitted work. A. Moyer reports grants from NIH and several foundations during the conduct of the study. R. Weinshilboum reports grants from NIH during the conduct of the study, as well as other support from OneOme LLC outside the submitted work. J.A. Copland III reports other support from Wuxi licensed PDX mouse models outside the submitted work. M.P. Goetz reports other support from Eagle Pharmaceuticals, Lilly, Biovica, Novartis, Sermonix, Context Therapeutics, Pfizer, Biotheranostics, and AstraZeneca, as well as grants from Pfizer, Sermonix, and Lilly outside the submitted work. No disclosures were reported by the other authors.

J.C. Boughey: Conceptualization, resources, data curation, software, formal analysis, supervision, funding acquisition, validation, investigation, visualization, methodology, writing–original draft, project administration, writing–review and editing. V.J. Suman: Data curation, software, formal analysis, validation, investigation. J. Yu: Data curation, software, formal analysis, validation, investigation. K. Santo: Data curation, software, formal analysis, validation, investigation, visualization, methodology. J.P. Sinnwell: Data curation, software, formal analysis, validation, investigation, visualization, methodology. J.M. Carter: Conceptualization, resources, data curation, software, formal analysis, supervision, funding acquisition, validation, investigation, visualization, methodology, writing–original draft, project administration, writing–review and editing. K.R. Kalari: Data curation, software, formal analysis, validation, investigation, visualization, methodology. X. Tang: Data curation, software, formal analysis, validation, investigation, visualization, methodology. S.A. McLaughlin: Conceptualization, resources, data curation, software, formal analysis, supervision, funding acquisition, validation, investigation, visualization, methodology, writing–original draft, project administration, writing–review and editing. A. Moreno-Aspitia: Conceptualization, resources, data curation, software, formal analysis, supervision, funding acquisition, validation, investigation, visualization, methodology, writing–original draft, project administration, writing–review and editing. D.W. Northfelt: conceptualization, resources, data curation, software, formal analysis, supervision, funding acquisition, validation, investigation, visualization, methodology, writing–original draft, project administration, writing–review and editing. R.J. Gray: Conceptualization, resources, data curation, software, formal analysis, supervision, funding acquisition, validation, investigation, visualization, methodology, writing–original draft, project administration, writing–review and editing. K.N. Hunt: Conceptualization, resources, data curation, software, formal analysis, supervision, funding acquisition, validation, investigation, visualization, methodology, writing–original draft, project administration, writing–review and editing. A.L. Conners: Conceptualization, resources, data curation, software, formal analysis, supervision, funding acquisition, validation, investigation, visualization, methodology, writing–original draft, project administration, writing–review and editing. J.N. Ingle: Conceptualization, resources, data curation, software, formal analysis, supervision, funding acquisition, validation, investigation, visualization, methodology, writing–original draft, project administration, writing–review and editing. A. Moyer: Conceptualization, resources, data curation, software, formal analysis, supervision, funding acquisition, validation, investigation, visualization, methodology, writing–original draft, project administration, writing–review and editing. R. Weinshilboum: Conceptualization, resources, data curation, software, formal analysis, supervision, funding acquisition, validation, investigation, visualization, methodology, writing–original draft, project administration, writing–review and editing. J.A. Copland III: Conceptualization, resources, data curation, software, formal analysis, supervision, funding acquisition, validation, investigation, visualization, methodology, writing–original draft, project administration, writing–review and editing. L. Wang: Conceptualization, resources, data curation, software, formal analysis, supervision, funding acquisition, validation, investigation, visualization, methodology, writing–original draft, project administration, writing–review and editing. M.P. Goetz: Conceptualization, resources, data curation, software, formal analysis, supervision, funding acquisition, validation, investigation, visualization, methodology, writing–original draft, project administration, writing–review and editing.

This work was supported by: Mayo Clinic Center for Individualized Medicine, Nadia's Gift Foundation, John P. Guider, The Eveleigh Family, George M. Eisenberg Foundation for Charities, Pharmacogenomics Research Network (U19 GM61388, to M.P. Goetz, L. Wang, R. Weinshilboum, K.R. Kalari, and J.N. Ingle), NIH (R01 CA196648, to L. Wang), Mayo Clinic Cancer Center (CA15083-40A2), and Mayo Clinic Breast Specialized Program of Research Excellence (SPORE P50CA116201, to M.P. Goetz, V.J. Suman, K.R. Kalari, J.M. Carter, L. Wang, and J.N. Ingle). The BEAUTY study is funded in part by the Mayo Clinic Center for Individualized Medicine, Nadia's Gift Foundation, John P. Guider, the Everleigh Family, George M. Eisenberg Foundation for Charities, generous support from Afaf Al-Bahar, and the Pharmacogenomics Research Network (to M.P. Goetz, J.C. Boughey, L. Wang, R. Weinshilboum, K.R. Kalari, V.J. Suman, and J.M. Carter). J.C. Boughey is the W.H. Odell Professor of Individualized Medicine. R. Weinshilboum is the Mary Lou and John H. Dasburg Professor of Cancer Genomics Research. M.P. Goetz is the Erivan K. Haub Family Professor of Cancer Research Honoring Richard F. Emslander.