Establishment of a non-small-cell lung cancer-liver metastasis patient-derived tumor xenograft model for the evaluation of patient-tailored chemotherapy

In order to optimize patient-tailored chemotherapy, a non-small-cell lung cancer (NSCLC)-liver metastasis patient-derived tumor xenograft (PDTX) model is developed. Computed tomography (CT)-guided NSCLC percutaneous biopsy was subcutaneously inoculated into the flank of non-obese diabetic/severe combined immunodeficiency (NOD/SCID) female mice (PDTX F1) and allowed to reach 500 mm3 volume. Then, the tumors were re-transplanted into Balb/c nude mice and liver metastasis was confirmed (PDTX F2), which were further assigned into doxorubicin (DOX), docetaxel (DTX), and non-treatment control group. H&E staining and Keratin 20 (CK20) staining were applied to determine the consistency of PDTX models and primary tumors. Tumor growth curve, body weight, and the expression of p65 nuclear factor (NF)-κB and the secretion of interferon (IFN)-γ were investigated. The successive transplant procedure can induce the NSCLC-liver metastasis PDTX model, and morphological and structural characteristics of PDTX models (F2) were in accordance with primary tumors. DOX and DTX could delay tumor growth, activate the NF-κB pathway, and promote IFN-γ secretion in the PDTX models. The NSCLC-liver metastasis PDTX model is established and provides a powerful mean to assess chemotherapeutic efficacy.


Introduction
As one of the most common cancers worldwide, approximately 30-40% non-small-cell lung cancer (NSCLC) patients following the first diagnosis will present with distant metastases, such as liver, brain, lymph node, bone, and adrenal gland [1,2]. NSCLC metastatic to the liver is classified as stage 4, and the median survival time is only approximately 8 months [3][4][5]. Clinically, treatments of liver metastases from NSCLC consist only of systemic therapy using cytotoxic (doxorubicin (DOX) and docetaxel (DTX)), molecularly targeted agents (EGFR mutations, ALK rearrangements, and ROS1 rearrangements target), palliative radiotherapy, and immunotherapy. In some cases, these medications may result in long-term control of liver metastases from NSCLC [6,7]. It is generally held that the life expectancy of liver metastases from NSCLC depends on the response to chemotherapy [8,9]. Tumor-bearing mice are commonly used to evaluate the potent toxicity and to predict the preclinical efficacy of the chemical. Patient-derived tumor xenograft (PDTX) models will keep the histological and biological characteristics of primary tumors, will be used as ideal models to develop and monitor the drugs. [10][11][12]. Over the past 10 years, some primary cancers targeted PDTX models have been developed [13,14]. More importantly, the passaged PDTX models are biologically stable in terms of tumor architecture, mutational status, global gene-expression patterns, drug responsiveness, and metastatic potential, which can be utilized as an information-rich preclinical resource for predictive biomarker discovery and drug activity screen.
Considering that metastatic disease could lead to more mortality than primary cancer, more efforts should be given to developing xenograft models of metastasis, especially for lung cancer-derived xenograft models [15,16]. The purpose of this investigation is to establish NSCLC-liver metastasis PDTX models and to evaluate patient-specific therapy.

NSCLC lung tissue specimens
From January 2017 to December 2017, three patients were enrolled in the present study with pathologically proven NSCLC-liver metastatic in accordance with the World Health Organization Classification [17,18]. Patients with severe cardiac, bleeding tendency or bleeding, and pulmonary dysfunction were excluded from the cohort. The collection of percutaneous lung cancer biopsy and the design of the whole research were approved by the Cangzhou Central Hospital Ethics Committee. All the participants signed the informed consent.

Mice
Female, non-obese, diabetic (NOD)/severe combined immune deficiency (SCID) mice and female Balb/c nu/nu mice (6-8 weeks old) were ordered from the Vital River Company (Beijing, China). All procedures on mice were performed in accordance with the guideline of Care and Use of Laboratory Animals issued by the China National Institute of Health and the protocols were approved by the Cangzhou Central Hospital.

PDTX
Multidetector-row helical computed tomography (CT)-guided biopsy (Hispeed NX/I; GE Healthcare, U.S.A.) was performed. Four different biopsy locations were chosen and performed based on the enhanced CT measurement to make sure three samples at least could be achieved. One sample was fixed in 4% paraformaldehyde, and the other two samples (F0) cut into fragments (1 mm × 1 mm) were inoculated into the NOD/SCID mice's flank subcutaneously (the PDTX [F1]. When the inoculations reached 500 mm 3 , the tumor tissues were further grafted into the flanks of Balb/c nude mice and liver metastasis was confirmed (the PDTX [F2].

Tumor growth curve and body weight assay
The PDTX [F2] mice were assigned into two treatment groups (DOX and DTX, single dose 10 mg/kg i.v. injection, n=6) and one control group (phosphate-buffered saline, single dose 200 μl i.v. injection, n=6). Both tumor sizes and body weights were recorded every 5 days. Tumor volumes were measured as follows: Tumor volume = (tumor length) × (tumor width) 2 /2.

Enzyme-linked immunosorbent assay
Blood samples were obtained by cardiac puncture and placed at room temperature for 1 h. Then the blood samples were centrifuged for 3000 rpm (30 min) to collect serum. The concentration of interferon (IFN)-γ in the serum was assayed with specific enzyme-linked immunosorbent assay (ELISA) kits (Cat# BMS606) (eBioscience, San Diego,

Statistical analysis
Data were shown as the mean + − standard error and analyzed by Student's t test. P<0.05 was considered to be statistically significant.

Construction of NSCLC-liver metastasis PDTX model
Three NSCLC patients accompanied with liver metastasis were recruited, and only one PDTX model derived from one patient was established. Schematic diagram of PDTX model ( Figure 1A) and the CT scan of such patient with NSCLC ( Figure 1B) were shown. The percutaneous biopsies of NSCLC were grafted into NOD/SCID mice (the PDTX [F1] and re-grafted into Balb/c nude mice (the PDTX [F2]. As shown in the tumor growth curve, the mean latency period for the inoculations to reach 500 mm 3 in PDTX F2 was 33 days ( Figure 1C), while 50 days for PDTX F1 (data not shown). Morphological properties and CK20 (a positive marker of metastasis) staining of PDTX F2 tissues resembled the original NSCLC tissues (Figure 2A,B), while such phenomena were not observed in PDTX F1. The pathological liver damage and cellular infiltration in PDTX F2 mice were severe when compared with PDTX F1 mice (Figure 3), suggesting the occurrence of liver metastasis. All of these results indicated that NSCLC-liver metastasis PDTX model was constructed.

Chemotherapeutic efficacy evaluation
PDTX model was used to test the chemotherapeutic efficacies of DTX and DOX. Single administration with DTX (P<0.05) or DOX (P<0.01) suppressed the growth of tumor in the PDTX model when compared with the control group ( Figure 4A). The tumors growth doubling time can be delayed by DTX and DOX to 5 days when compared  with the control group (15 days). Body weight did not show any significant change in any of the three groups ( Figure  4B).

NF-κB pathway and IFN-γ may mediate the chemotherapeutic efficacy
To assess the association with NF-κB pathway, the expression of p65 in the lung tissue at 30 days was measured with Western blot ( Figure 5A). When compared with the control group, both DTX and DOX could significantly increase the expression of p65, which indicated that the active reaction of the NF-κB pathway. It was worth noting that the secretion of IFN-γ in the serum increased in the DTX and DOX treatment groups (P<0.01, Figure 5B).

Discussion
The purpose of this research is to establish NSCLC-liver metastasis PDTX model and to test the chemical sensitivity of DOX and DTX. Our data suggest that NSCLC-liver metastasis PDTX model is successfully established. Both DOX and DTX can significantly inhibit the growth of the metastatic tumor, and NF-κB and IFN-γ may be involved in the process of chemical therapy. The construction of metastatic PDTX model is constrained due to the impracticable surgery of advanced cancer patients. In such circumstances, CT-guided percutaneous lung biopsy may give another option [19], which is used to classify pathological tumor phenotype and to screen tumor-derived gene mutation available with targeting drugs. In the present study, a fresh biopsy of lung tumor tissues is used to ensure tumorigenesis, and further transplanted into NOD/SCID mice and Balb/c nu/nu mice. Less time is needed to construct the PDTX F2 model when compared with the PDTX F1 model, which indicates increasing tumorigenesis and metastatic property. Successive cancer tissue transplants in immunocompromised mice are performed in the present study, which can maintain at least some aspects of the patient-derived tumors in hereditary characteristic (genetic alterations), growth microenvironment, response to chemotherapy [10,13,19,20]. Moreover, such a model also maintains the original tumor heterogeneity, representing a promising model for investigating specific cellular subpopulations [21]. It is worth noting that genes alteration can also be observed in the PDTX models compared with primary tumors [22], which can be attributed to selection pressures during engraftment into different species. Whether such phenomena also happened in our model needs further investigation. Whatever, metastasis-related CK20 is preserved and the PDTX models we establish could preserve the primary tumor characteristics to optimize patient-tailored chemotherapy. The chemosensitivities of DOX and DTX are compared by measuring the volume of the tumor, which shows similar inhibition rates. To our surprise, the activated NF-κB pathway and promoted IFN-γ production in the PDTX model are observed. Inflammatory stimuli can be converted into tumor promotion signals by NF-κB and consequently up-regulates the growth speed of metastatic tumor [23,24]. It is worth noting that the immune cells targeted by NF-κB, which can prevent or promote tumor development in a context-dependent manner, are immunocompromised in our models [25]. Further studies are indicated to explore the source of increased secretion of IFN-γ and the underlying mechanism of up-regulated NF-κB. All in all, the NSCLC-liver metastasis PDTX models established maintain the histological and genomic properties of primary tumors, which can be further integrated with therapeutic strategies to optimize patient-tailored chemotherapy.

Conclusions
CT-guided percutaneous biopsy samples are used to successfully develop NSCLC-liver metastasis PDTX models to assess chemotherapeutic efficacy.

Author Contribution
Xue Yang and Gaopei Meng performed the experiments, analyzed and interpreted the data. Xue Yang wrote the manuscript. All authors read and approved the final submission.