Endogenous Cytokine Responses and Unexpected Survival in a Phase I Trial Using Anti-CD3 x Anti-HER2 Bispecific Antibody Armed Activated T Cells (HER2 BATs) for Treatment of Women with Stage III Breast Cancer

BACKGROUND: Advanced stage breast cancer has unacceptably high recurrence rates. In this study, we report a phase I immunotherapy (IT) trial in women with high risk (>3 positive nodes) breast cancer that seeks to determine safety and maximum tolerated dose (MTD) of anti-CD3 x anti-HER2 bispecific antibody (HER2Bi) armed activated T cells (HER2 BATs), with the secondary objective of determining whether IT induces endogenous cytokine changes. METHODS: The phase I trial consisted of 8 twice weekly infusions of HER2 BATs given in a 3+3 dose escalation with groups of 3 patients receiving 5, 10, or 20 x 109 HER2 BATs per infusion. HER2 BATs were given in combination with daily low dose interleukin 2 (IL-2) and twice weekly low dose granulocyte-macrophage colony-stimulating factor (GM-CSF) to determine safety, the maximum tolerated dose (MTD), technical feasibility, immune cytokine responses, time to progression (TTP), and overall survival (OS). RESULTS: Nine women with stage III breast cancer were enrolled at a single institution. There were no dose limiting toxicities (DLTs) and the MTD was not defined. It was technically feasible to grow 176 x 109 ATC from a single apheresis. Five of 9 women are alive with no evidence of disease with an undefined median OS. Infusions of HER2 BATs induced increases in serum Th1 cytokines, and some of these changes persisted for weeks to months after IT. CONCLUSIONS: Targeting HER2 positive and negative tumors with BATs infusions was safe, induced Th1 cytokines and IL-12 responses in the serum providing evidence for inducing changes in the endogenous immune system of the patients. The prolonged survival that is undefined with a median follow up of 12.7 years suggests that targeting tumors in the adjuvant setting with HER2 BATs may help eliminate microscopic residual disease after surgery, irradiation, and chemotherapy. TRIAL REGISTRATION: The protocol entitled “Combination Chemotherapy Plus Biological Therapy in Treating Patients with Stage II or Stage III Breast Cancer” as a phase I/II was registered


12-15 ]
In this study, long-term follow-up is presented for women with stage III breast cancer who received standard of care adjuvant therapy followed by immunotherapy with HER2 BATs infusions administered with low dose interleukin 2 (IL-2) and granulocyte-macrophage colony-stimulating factor (GM-CSF). We designed the trial to ask three questions: 1) Are HER2 BATs safe and technically feasible? 2) Do HER2 BATs induce changes in serum cytokine patterns before, during, and after infusions? 3) Does this combination of immunotherapy improve time to progression (TTP) or overall survival (OS)? Because this was a phase I pilot study with limited funds, the study was constrained to 9 patients. The impetus to submitting this report was driven by unexpected long-term clinical results.
Infusions of BATs were safe, feasible, induced marked changes in cytokine levels, and may have provided clinical anti-tumor activity in this high-risk adjuvant population of patients. The product was released after bacterial and fungal cultures were negative for 7 days, testing for mycoplasma was negative, and the endotoxin levels were determined.

Methods
Phenotyping and functional testing of the product was documented.
Phase I Clinical Trial Design. The primary endpoint of this standard 3 + 3 dose escalation trial was to determine the safety and maximum tolerated dose (MTD) of HER2 BATs with dose levels of 5, 10, and 20 billion HER2 BATs/infusion. Patients received 2 infusions/week for 4 weeks, for total doses of 40, 80, and 160 x 10 9 BATs. BATs were given in combination with subcutaneous injections of IL-2 (300,000 IU/m 2 /day) and GM-CSF (250 μg/m 2 /twice weekly) beginning 3 days before the 1 st infusion and ending 1 week after the last BATs infusion. Figure 1A shows the treatment schema. Tumor evaluations were performed per standard of care followed by the referring physicians. All patients were followed for OS.  Table 1.
Leukapheresis, T cell Expansion, and Production of BATs. ATC were produced from the leukapheresis product as described.
[ 10 ] Aliquots from cryopreserved BATs were submitted for bacteria and fungus, endotoxin, mycoplasma, phenotype, and cytotoxicity testing. BATs were released for infusions only after passing all release criteria as previously described.[ 7 Dose Modification and Toxicity Scoring (NCI Toxicity Criteria, v2, June 1, 1999). Patients were accrued to each dose level based on the dose-escalation schema. If 1 in the first 3 patients or 2 in the first 6 patients experienced persistent grade 3 non-hematological toxicity or grade 4 toxicity, the dose would not be escalated. Patients with grade 4 nonhematologic toxicity were removed from protocol. Patients who experienced a 10% decrease in the ejection fraction would be removed from protocol and treatment stopped.
Treatment was held for grade 3 toxicity until toxicity decreased to grade 0-1. If grade 3 toxicity recurred, the subsequent dose of BATs was washed to eliminate dimethysulfoxide.
If toxicity persisted, the next dose was reduced by 50%. If the toxicities persisted at the reduced dose of BATs, IL-2 would be stopped but the BATs infusions continued. If grade 3 toxicity occurred again, the BATs infusions would be stopped. Toxicities were assessed for 7 days after each infusion. Toxicities were followed until they resolved.
Immune Function Assays. Cytokine patterns were obtained by multiplex cytokine array using the Bio-Plex system (Bio-Rad Lab., Hercules, CA).
Statistics. The primary endpoint of the study was to determine safety and MTD of BATs.
The secondary endpoint was to assess OS, which was measured from the date of enrollment. Descriptive statistical analyses were used to analyze immune responses using GraphPad Prism (GraphPad, Version 6.0).

Results
Patient Characteristics. Table 1 summarizes the clinical characteristics, prior therapy, the planned and infused doses of HER2 BATs, and OS for 9 women with high risk breast cancer. Median age was 49 years (range: 38-69 years). All patients received adjuvant chemotherapy and surgery with or without irradiation. Table 2 summarizes the patient characteristics. Five of 9 women were less than 50 years of age. Eight women had >10 nodes positive for tumor, and 1 woman had 2 nodes positive for tumor. Eight women had ER/PR positive tumors, and 1 had an ER/PR negative tumor. Two women had tumors that were HER2 3+, and 2 women had tumors that were HER2 1-2+. Five women had HER2 negative tumors. No one received trastuzumab prior to therapy.
Phase I Evaluation of Toxicities. The most frequent side effect (SE) was grade 3 chills.
Grade 1 and 2 fever and malaise emerged as the second most common SE. Table 4 shows the frequency of side effects (NCI Immunotherapy Protocol Toxicity Criteria, v2). By episode per infusion, the incidence of chills and fever were 66.6% each. All patients with grade 3 chills responded to meperidine. There were no hospitalizations. There were no DLTs attributed to HER2 BATs.
Phase I Evaluation of MTD. The highest dose level completed was 20 x 10 9 BATs per infusion (160 x 10 9 total dose of BATs); thus, 160 x 10 9 given in 8 divided doses is a technically feasible dose. Table 1 shows the planned doses and actual doses for 9 patients.
Overall Survival. Figure 1B shows the K-M curve for all 9 patients. The median OS for 9 patients remains undetermined (14-164 months) with a median follow-up of 12.7 years.
Serum cytokines. Serum was available for 5 of 9 patients to be tested ( Figure 2) for serum cytokine levels pre-IT baseline serum (BASE), after each BATs infusion, and 1 week and 1 month post completion of all infusions to determine whether IT induced changes in serum cytokine profiles (n=5). One patient (FG60031) did not have detectable levels of any of the cytokines tested. Table 5 shows the detailed data shown in Figure 2

]
When interpreting these results in the context of contemporaneous therapies in patients who received adjuvant treatment with anthracyclines and taxanes for breast cancer with >10 lymph nodes, it is notable that median 10-year OS ranged from 29% to 42%. [ 15 ] The median OS in this group has not been reached, with the proportion of patients surviving 14+ years >55%. In this phase I study, most patients had >10 nodes.
The major side effects were chills, fever, headache, fatigue, and hypotension. Mild cytokine "flurries" were observed, but there were no life-threatening cytokine "storms," and no patients were admitted to the hospital. Increases in serum Th 1 cytokine levels leading to high Th 1 /Th 2 ratios and increases in IL-12 that developed mid IT and persisted for weeks after IT show that the immunologic milieu and tumor microenvironment were shifted towards an anti-tumor environment.
Despite the encouraging safety and immunologic findings, an expansion cohort was not pursued due to lack for funding for such a trial. However, in a phase I metastatic breast cancer trial, we showed that the infused HER2 BATs were able to induce endogenous

]
High dose chemotherapy created "immune space" for ATC to expand and accelerate reconstitution of cellular and humoral immunity against breast cancer antigens within 4 weeks after SCT. Low dose IL-2 and GM-CSF were given to a few patients after SCT, but there was no obvious difference in the immune responses. CTL responses and specific IgG to various tumor peptides were detected to different epitopes of HER2, EGFR, carcinoembryonic antigen. These antibody data directed at multiple epitopes of HER2 and the non-targeted antigens EGFR and CEA showed that there was not only epitope spreading, but also antigen spreading after targeting HER2 antigen on the breast cancer.
The specific CTL responses persisted up to 3 years.

Conclusions
These results provide the rationale for the design of phase II clinical trials using HER2 BATs in breast cancer and other solid tumors in the adjuvant setting where there is minimal residual disease. Because this was a phase I study with limited funds, there were only 9 patients. Because the sample size was small, there was originally no intent on submitting the data; however, the delay in reporting led to the unexpected encouraging clinical long-term survival and no evidence of disease on follow-up. These results provide significant rationale for designing approaches that use different form targeted T cells to clean up residual microscopic disease in the adjuvant or neoadjuvant setting. Ethics approval and consent to participate: The protocol was reviewed and approved by protocol review committee, institutional Human Investigational Committees at RWMC, and the Food and Drug Administration. The trial was monitored by RWMC data safety monitoring committee. All patients signed informed-consent prior to enrollment.

Consent for publication:
All patient were informed that the clinical data could potentially be published.
Availability of data and material: The de-identified research and clinical data are available to the public.    Treatment schema shows leukapheresis to obtain T cells for expansion. HER2Bi armed ATC (BATs) were administered twice weekly for four consecutive weeks. All patients received SQ IL-2 (300,000 IU/m2/day) and GM-CSF (250 µg/m2/twice weekly) beginning 3 days before the first BATs infusion and ending 1 week after the last BATs infusion. Immune testing was performed at indicated time points after BATs infusions. 1B. K-M curve for all 9 patients (2 HER2 3+ and 7 HER2 0-2+ patients).