Automated Model Generation and Calibration Workflow for Pharmaceutical Processes Using Knowledge Graphs

Breast Cancer Brain Metastases (BCBM) occur in 10-15% of patients with breast cancer and are connected to a very poor prognosis with the shortest median survival of 6 months for patients with the Triple Negative Breast Cancer (TNBC) subtype. At present, there are no FDA-approved drugs to treat BCBM, and only <1% of all breast cancer-related trials focus on BCBM. Current therapeutic options considered the gold standard for treating BCBM are surgery and whole-brain radiation therapy which are specifically used in early and later stages of metastasis, respectively, with whole-brain radiation therapy specifically reported to have long-term consequences of associated neurocognitive decline. Additionally, tumor subtype specific targeted therapies are being developed for breast cancer brain metastases. However, since these therapies rely on receptor expression for targeting, they are not applicable for the TNBC subtype that has a brain metastases frequency of 14-38% and is predominantly treated with cocktails of chemotherapy. This chemotherapy-based treatment is often ineffective due to the challenges posed by the heterogeneous permeability of the Blood Brain-Tumor Barrier (BBTB) limiting the cytotoxic dosage in tumors and the development of chemoresistance in these tumors. Thus, there is an urgent unmet need for better therapeutics to treat BCBM.

We look to evaluate the potential of using RadioPharmaceutical Therapy (RPT) with α-particle emitters (αRPT) combined with chemotherapy for the treatment of BCBM. Targeting α-particles to the tumor, results in the deposition of large amounts of ionizing radiation in short distances (5-10 cell diameters) creating lethal and complex DNA lesions that overwhelm cellular repair mechanisms and are often irreparable in tumors. This highly lethal combination of high energy makes α-particle therapy impervious to currently known resistance mechanisms of cancer. Their short range in tissue, makes α-particles ideal for irradiation of BCBM with minimal effects on the peripheral healthy brain. Since much of the damage caused by α-particles comes from their ability to hit the cell directly, causing DNA damage, α-particles' ability to treat solid tumors is hampered by their inability to penetrate the volume of the tumor by itself - its short-range or by the carriers used to delivery α-particles that are also stymied by diffusion-limited penetration depths. This is the root cause of the current treatment failure of α-particle therapies which is exacerbated by the additional challenge of the traditional carriers' crossing of the Blood Brain-Tumor Barrier. Current targeted α-particle therapies focus on the treatment of early metastases to maximize irradiation of tumors with α-particles, however since BCBM is often diagnosed very late due to inadequate screening, it is imperative to improve penetration and distribution of α-particle therapy via a novel drug delivery strategy. Towards this goal, we propose a two-pronged approach: 1) Utilizing novel carriers to improve penetration of α-particles in the tumor, and 2) Combining α-particle therapy (limited penetration potential) with cisplatin-based chemotherapy (high penetration potential) to achieve synergistic antitumor efficacy.

Here we look at using ultrasmall nanoparticles (<8 nm diameter hydroxyl terminated dendrimers) as the carrier for α-particle emitting radionuclides for the treatment of BCBM for several reasons: 1) These particles being in the sub 10 nm range can easily cross the BBTB upon systemic administration; 2) These particles accumulate in the tumor as a result of being taken up by Tumor-Associated Macrophages (TAM), which occupy about 30-50% of BCBM. This property makes this therapy agnostic to the subtype of breast cancer and also provides an opportunity to achieve increased distribution and retention of dendrimer nanoparticles by the tumor-infiltrating TAM leading to uniform and sustained irradiation of the tumor; 3) The small size of the particle enables it to clear quickly from the body and brain when not associated with TAMs thereby reducing off-target toxicities.

The incidence of BCBM has increased in recent times in part due to the success of current therapeutics in extending the survival of patients with advanced breast cancer. The BBTB, although leaky, keeps out most of the drugs used to treat breast cancer, providing a haven for brain metastases of breast cancer. This increased incidence coupled with a lack of routine screening for brain metastases in asymptomatic patients leads to delayed diagnosis of brain metastases with very few treatment options. In such cases, a multimodal treatment approach is pertinent to achieve better therapeutic efficacy. Chemotherapeutics such as cisplatin (CDDP) can penetrate BBTB and by virtue of its small size can uniformly penetrate and distribute in the tumor compared to dendrimer - nanoparticles described above. Combining both these therapeutics might help achieve synergy due to the increased distribution of drugs compared to either of them administered alone. Such combinations of chemotherapy and external beam radiation therapy have been evaluated in clinical trials previously but no therapeutic strategy involving α-particle therapy and chemotherapy has been evaluated for breast cancer brain metastases to date. To this effect, we propose to evaluate the combination of chemotherapeutic cisplatin commonly used to treat BCBM of the TNBC subtype that has the potential to cross BBTB and ²²⁵Ac radiolabeled dendrimers to achieve a synergistic antitumor effect that will likely reduce tumor burden, avoid development of resistance, and prolong survival.

Materials and Methods: In vitro, the sensitivity of 4T1 murine TNBC cells, that can spontaneously metastasize to the brain, to ²²⁵Ac radiolabeled dendrimer nanoparticles with and without cisplatin were evaluated using radio-sensitivity assays on 4T1 monolayers. The spatial micro distributions of cisplatin surrogate CFDA-SE and Cy5 labeled dendrimer were evaluated at different time points in a 3D tumor model of 4T1 cells and time-integrated radial concentrations were calculated and plotted for the two different molecules. The extent of outgrowth inhibition by radiolabeled dendrimers and cisplatin as single treatments and in combination were evaluated using 3D tumor models of 4T1 cells.

Results and Discussion: Cell killing was increased when Radiolabeled Dendrimer and CDDP were given in combination both on a cell monolayer – in the absence of transport barriers - (Figure 1A) as well as on 3D tumor spheroid models (Figure 1B), compared to either of the two therapies alone. Time-integrated concentration along the radius of the spheroid for the CDDP fluorescent surrogate CFDA-SE (Figure 1C) and Cy5 labeled Dendrimer (Figure 1D) confirmed their complementary distribution profiles in the tumor. Specifically, the dendrimer was associated more near the edge of the spheroid and did not penetrate deeper into the tumor spheroid model compared to CDDP surrogate CFDA-SE that significantly infiltrated the spheroid and reached its core region. Therefore, the improved killing effect observed on cell monolayers and on spheroids is probably due to both due to cell biology factors and due to delivery. We are in the process of better understanding this interplay.

Conclusions: This study aims to understand the synergy – due to cancer biology and delivery - between α-particle therapy and chemotherapy in achieving better inhibition of BCBM in a solid tumor surrogate in vitro. Understanding of these processes may ultimately help in engineering therapeutic interventions that can combat BCBM of TNBC subtype and possibly prolong survival rates.

Figure 1: Combining ²²⁵Ac radiolabeled dendrimers and cisplatin shows better antitumor efficacy compared to single treatments of either of the two therapies (A) on a monolayer cell culture as well as (B) in 3D tumor spheroid models of 4T1 cells in vitro. The time-integrated radial spatiotemporal profiles of (C) the Cisplatin surrogate CFDA-SE and (D) Cy5-labeled dendrimer shows the complementary micro-distribution of carriers in the tumor spheroid model which may in part explain the increased antitumor efficacy obtained by combining the two therapies.