Intravenous Vitamin C and Radio Sensitization in Pancreatic Cancer

Reference

Alexander MS, Wilkes JG, Schroeder SR, et al. Pharmacologic ascorbate reduces radiation-induced normal tissue toxicity and enhances tumor radiosensitization in pancreatic cancer. Cancer Res. 2018;78(24):6838-6851.

Objectives

This paper reports on results from 3 experiments—1 in vitro (cells) and 2 in vivo (animal and human). The objectives of each experiment were as follows:

• In vitro experiments: To determine the differential toxicity of ascorbate in pancreatic cancer vs healthy jejunal cells undergoing radiation
• Animal study: To assess radiation-treated tumor tissue and intestinal jejunal tissue in mice receiving intraperitoneal (IP) ascorbate compared with controls
• Human clinical trial: To evaluate the safety and potential efficacy of intravenous vitamin C (IVC) combined with radiation and gemcitabine for patients with locally advanced pancreatic cancer and identify maximum tolerated dose (MTD) of IVC

Design

In vitro experiments

Cell culture of human pancreatic healthy and cancer cells along with healthy jejunal cells; clonogenic survival assay during administration of ascorbate and radiation; immunoblot analysis of lipid peroxidation; glutathione assay

Animal study

Tumor-bearing mice receive IP ascorbate or saline on days 1 and 2, abdominal radiation on day 3, and IP ascorbate and days 4 and 5, with evaluation of tissue microstructure, mitochondrial damage, and glutathione content

Human clinical trial

Single-institution, open label, phase 1 study of IVC administered concurrently with gemcitabine (GEM) and radiation therapy. Patients with pancreatic adenocarcinoma eligible for the treatments were invited to participate. There were 15 subjects enrolled in the experimental group and there were 19 subjects enrolled as institutional comparators meeting the same inclusion criteria and receiving concurrent gemcitabine and radiation therapy with curative intent. (Some of these were historical controls.) IVC was administered as 15 g test dose in 250 mL sterile water for tolerability, followed by 50 g IVC in 750 mL daily, in the first cohort. Gemcitabine standard dose of 600 mg/m² over 30 minutes was given weekly for 6 weeks, concurrent with radiation delivered as 50 to 50.4 gray in 25 to 28 fractions. Blood samples were drawn weekly measuring plasma ascorbate before and after infusion and at intervals for F₂-isoprostane.

The clinical trial used a 2-stage design. Stage 1 consisted of dose escalation for each patient, with stepwise increases in dose if there was no toxicity at the initial dose. In stage 2, groups of 3 patients were treated in a dose cohort, and if there was no toxicity at the initial dose, each member of the cohort increased to the next dose; any patient who experienced treatment-related toxicity was de-escalated to the previous dose. The study required that a minimum of 6 subjects must be evaluated at the phase 2 dose with ≤ 1 dose-limiting toxicity occurring in those 6 before enrolling more patients.

Participants were assigned to the IVC cohorts (g ascorbate/mL sterile water) as per 3 open slots for each dose: 50 g/750 mL, 75 g/1,000 mL, and 100 g/1,500 mL. So this means 3 would be enrolled to receive 50 g IVC and then if all safe and well-tolerated, they move up to the 75 g dose and 3 more participants are enrolled at the 50 g dose. If the 6 enrolled patients are safe and not experiencing 1 or more dose-limiting toxicity, enrollment continues. The following were considered dose-limiting toxicities: vomiting resulting in hypokalemia that did not respond to treatment; grade 4 febrile neutropenia; grade 4 intra-abdominal hemorrhage; or grade 4 weight loss. Any serious adverse event related to the IVC was also considered a dose-limiting toxicity.

Radiation therapy as intensity-modulated radiation therapy (IMRT) was delivered as 50.4 Gy in 28 fractions or 50 Gy in 25 fractions as determined by the treating radiation oncologist. Intravenous vitamin C was started 20 minutes prior to radiation and continued throughout and after IMRT. Of note, this is the first study to actively infuse IVC during radiation “beam on” time.

Participants

In vitro experiments

The following cells were used in the in vitro experiments:

• MIA PaCa-2—adenocarcinoma of the pancreas originating in the body and tail; negative for carcinoembryonic antigen (CEA) or alkaline phosphatase
• PANC-1—adenocarcinoma of the pancreas originating in the head; with duodenal and lymph node invasion; metastatic; negative for CEA
• Patient-derived pancreatic cancer cell line 403
• FHs 74 Int—normal fetal human intestinal epithelial cells
• H6c7—normal pancreatic ductal epithelial cells

Animal study

For the animal experiments, 30-day-old athymic nude mice (Foxn1^nu) were injected with MIA PaCa-2 cancer cells; tumor and jejunal tissue from irradiated mice with and without IVC was studied.

Human clinical trial

Patients (N=16) enrolled in the phase 1 clinical trial had histologically or cytologically confirmed locally advanced pancreatic adenocarcinoma and were candidates for treatment with gemcitabine (GEM), radiation therapy, and IVC. Additional inclusion criteria included the following:

• Glucose-6-phosphate dehydrogenase (G6PD)<7 U/g Hb
• Karnofsky performance status ≥60
• Absolute neutrophil count (ANC) ≥1500/mm³
• Creatinine ≤1.5 mg/dL or creatinine clearance ≥60 mL/min
• Total bilirubin ≤2 x upper limit of normal (ULN)
• Transaminases ≤2.5 x ULN
• Normal prothrombin and international normalized ratio (INR)
• Ability to tolerate a test dose of 15 g intravenous ascorbate in 250 mL sterile water for human use

Patients with active comorbidities such as end-stage congestive heart failure, unstable angina, or myocardial infarction within 6 months of enrollment were excluded from the study.

Two patients dropped out of the study, leaving 14 eligible for final analysis.

Study Parameters Assessed

In vitro experiments

Clonogenic survival of pancreatic cancer and pancreatic healthy cells along with healthy fetal jejunal cells were measured after exposure to ascorbate and radiation.

Animal study

Mouse tumor and jejunal tissue was excised and processed via immunoblot analysis to determine 4-hydroxy-2 nonenal (4HNE)–modified protein levels as a marker of lipid peroxidation/tissue oxidative stress. Jejunal tissue from mice with xenograft pancreatic tumors treated with radiation and IP ascorbate vs saline were obtained and the tissue microstructure was examined for eosin and collagen deposition and given quantitative scores. The total mitochondria in the fixed tissues were visualized using electron microscopy and counted along with the number of damaged mitochondria. Glutathione assay levels in the jejunal tissue were quantified by measuring glutathione disulfide (GSSG) levels.

Human clinical trial

The following were used to assess the patients involved in the clinical trial:

• Karnofsky Performance Score—assessed before and after chemoradiotherapy
• Adverse events—graded according to the National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE) v.4.03. Events rating lower than grade 3 were assumed to be related to tumor burden, side effects of antineoplastic therapy, and activities of daily life
• Plasma ascorbate measurements—taken within 15 minutes before or after IVC
• F-2 isoprostanes plasma samples (formed from nonenzymatic free radical peroxidation)—collected prior to initiating chemoradiotherapy, 3 weeks into chemoradiotherapy, and at the completion of treatment

Key Findings

Safety and tolerability

• Three adverse events were considered attributable to IVC: dry mouth, thirst, and transient blood pressure elevation.
• Grade 3 and 4 hematologic toxicities included anemia (n=1), leukopenia (n=8), decreased lymphocyte count (n=13), decreased ANC (n=7), and decreased platelet count (n=2). These were described to be consistent with GEM and radiation therapy.
• The MTD was found to be 100 g, with 75 g selected as a recommended phase II dose.
• There was no change in hemoglobin (Hb) level in patients during ascorbate combined with chemoradiation treatment (P=0.44), while a statistically significant decrease in Hb was found in 13 out of 19 comparator patients (P<0.01).
• Transient blood pressure elevation was reported in patients (grade 2: n=7; grade 3: n=5; grade 4: n=1). These elevations typically resolved within 30 minutes of completion of the IVC infusion in all patients except 1.
• Oxidative injury assessed via plasma F2-isoprostane levels decreased significantly in the IVC treated patients (P=0.02) but did not change in comparator subjects (P=0.88).
• Treatment adherence was much better in the treatment group compared with controls.

Other

• Although the trial is small and not prospectively powered to determine differences in survival among the treatment groups, the authors report that the ascorbate-treated group had a median progression-free survival of 13.7 months, lower than the University of Iowa’s institutionally treated median range (4.6 months) and the median range in the ECOG-E4201 trial (11.1 months). Two subjects in the treatment group became resectable following trial completion (both were initially borderline resectable) and were alive at follow-up at 44 and 35 months from diagnosis.
• Plasma vitamin C levels in participants averaged 15 mmol/L for the 15 g dose cohort and 20 mmol/L for both the 75 g and 100 g dose cohorts.
• In vitro, ascorbate potentiates radiation-induced cell death, while simultaneously protecting normal cells.
• In vivo, ascorbate reduces radiation-induced jejunal mucosal damage, intestinal collagen deposition, normal tissue mitochondrial damage, and oxidative stress.

Practice Implications

This study is an intriguing addition to the data supporting IVC in oncology care. It corroborates earlier data showing IVC is safe and well-tolerated in humans with cancer when given in combination with chemotherapy.^1-4 The unique and exciting part of this study is that chemotherapy and radiation were combined with IVC in doses of 50-100 g and found to be safe in all participants with locally advanced and borderline resectable pancreatic cancer. This study differs from others in that the vitamin C was infused during the actual radiation treatment.

Two participants out of 14 became eligible for surgery, underwent resection, and at 44 months and 35 months post-diagnosis were without evidence of recurrence.

For locally advanced and borderline resectable pancreatic cancer patients who are not eligible for surgery as an option initially, it is now thought that chemoradiation can improve survival and the chances of becoming resectable compared with chemotherapy alone or going straight to surgery. So IVC + GEM + IMRT could be a very interesting combination for further study in comparison with chemoradiation regimens such as FOLFIRINOX prior to surgery. Perhaps it could also be considered for those who are not able to tolerate 5-fluorouracil (FU) combination therapy.

In the IVC-treated patients, 2 out of 14 became eligible for and underwent surgery after receiving the IVC combined with chemoradiation. Both had received FOLFIRINOX as initial treatment, which was discontinued due to progression of disease and side effects, respectively, before entering the trial.

As for the MTD of IVC, the study reports information comparable to other IVC trials. Six participants entered the 75 g dose cohort, receiving IVC over 120 minutes, and 5 moved up to the 100 g dose cohort, receiving IVC over 180 minutes. One of the patients at the 100 g dose experienced transient hypertension resolving within 30 minutes after infusion. A second participant at the 100 g dose level who did not have a baseline history of hypertension developed post-infusion elevation of blood pressure that did not resolve within 30 minutes. This may have been attributable to the vitamin C as a dose-limiting toxicity; hence the patient resumed the study in the 75 g dose cohort.

Regarding plasma levels of ascorbate, the authors selected the study doses based on preclinical and human data reporting synergy between gemcitabine and IVC in doses sufficient to generate high millimolar plasma concentrations (10-15 mmol/Lol/L) thought to have antitumor effect. A prior trial, the PACMAN study,⁵ provided IVC with gemcitabine and erlotinib to 9 node-positive pancreatic cancer patients and reported a possible antitumor effect. In the Cullen study, IVC given at the 75 g dose was found to generate average plasma vitamin C concentrations of 20 mmol/L plasma levels (n=3; 95% confidence interval [CI]: 19-21) while the 100 g dose also created average plasma levels of 20 mmol/L (n=32; 95% CI: 19-22). The plasma ascorbate levels for the 50 g dose level averaged 15 mmol/L (n=17; 95% CI: 13-17) and was significantly lower than the other dose cohorts (P<0.05).

A means of dosing has been to use grams of ascorbate per kilogram of weight. As in prior studies, comparison of plasma ascorbate levels to subject metrics allowed for predictable dosing by weight in kilograms, ranging from 1.2 or 1.5 g/kg.

In the Cullen study, 75 g was selected as the recommended phase II dose for future study. In patients with advanced disease and poor performance status where quality of life is the priority, lower doses of IVC may be more appropriate. Although higher doses are reported as safe in studies providing IVC to people with advanced metastatic disease, there are risks to the administration of any high osmolarity fluid, so dilution of concentration and total volume should be considered.

Adverse events in this study are similar to those reported in other studies on IVC and are comparable to what we see in clinical practice: thirst, dry mouth, and transient blood pressure and/or urinary changes that resolve soon after infusion completion. We can also see some dizziness, hypotension, urinary urgency, and cramping pains.

Another thing we might expect to see in people getting chemotherapy and radiation treatment is anemia, which can impact quality of life and potentially disrupt oncologic treatment. The present study reports that Hb levels in the IVC + GEM + IMRT group remained stable (P=0.44). In contrast, the controls had significant decreases in Hb over the course of treatment (P<0.01).

In terms of borderline resectable patients, another study, presented at the June 2018 American Society of Clinical Oncology (ASCO) meeting, had interesting findings. It also gave gemcitabine and radiation to patients with local or locally advanced disease. The PREOPANC-1 trial enrolled 246 patients and reports that preoperative and postoperative chemoradiotherapy significantly improves outcomes in borderline resectable and resectable pancreatic cancer compared to immediate surgery or chemotherapy alone.⁶ Patients in the neoadjuvant group had an R0 resection rate of 31% vs 65% (P=<0.001), disease-free survival (DFS); median 7.9 vs 11.2 months; HR: 0.67; P=0.010) and no significant difference was observed in grade ≥ 3 adverse events between groups (P=0.17). A subgroup analysis of patients who underwent a resection showed a median overall survival (OS) of 16.8 months in the surgery-alone group and 29.9 months in the chemoradiation group (P<0.001).

Though the current study was not powered to make any conclusive reporting on the efficacy of ascorbate and chemoradiation in pancreatic cancer, the results are encouraging. Two participants out of 14 became eligible for surgery, underwent resection, and at 44 months and 35 months post-diagnosis were without evidence of recurrence.

In terms of surgery, another study presented at ASCO’s June 2018 meeting has some significance for discussion here. The PRODIGE 24 trial enrolled 493 participants and reports that FOLFIRINOX given after pancreatic surgery offers better outcomes than gemcitabine despite having more side effects and even though patients were less likely to finish their regimen.⁷ At 3 years, the disease-free survival rate was 39.7% in the modified-FOLFIRINOX group and 21.4% in the gemcitabine group. The median OS was 54.4 months in the modified-FOLFIRINOX group and 35.0 months in the gemcitabine group (stratified HR for death, 0.64; 95% CI: 0.48-0.86; P=0.003). The OS at 3 years was 63.4% in the modified-FOLFIRINOX group and 48.6% in the gemcitabine group.

Anti-inflammatory effects of IVC have been reported in oncology patients in a prior study.⁸ This anti-inflammatory effect is taking place despite the pro-oxidative effects of high dose IVC.⁹ This present study implies that IVC may be synergizing with chemoradiation without raising oxidative body burden. This was demonstrated by the drop of the oxidative marker, F₂-Isoprostanes, in the participant group. F₂-Isoprostanes were measured before treatment, at week 3, and after completion of all 6 weeks of combination therapy. They showed significant decrease over time. In comparison, F₂-isoprostane levels did not change in comparator subjects.

Lastly, an important aspect of the present study is that more patients in the IVC + GEM + IMRT group actually completed treatment—meaning they received their chemotherapy and radiation vs having dose reductions or stopping treatment. This is something I believe is happening on a larger scale with integrative patients as a whole. With fewer side effects, less damage to peripheral tissues, and improved overall outcomes, patients using integrative therapies may be more adherent to standard cancer treatment regimens and have better outcomes overall. Specifically in this trial, 14 patients received the combined therapy and 57% received GEM as prescribed in 6 cycles and 100% received the prescribed radiation dose. With comparison to well-established data that provides comparative information on GEM alone or combined with radiation therapy for locally advanced pancreatic patients in the ECOG 2011 study,¹⁰ we see that 29% of patients completed all cycles of chemotherapy and 24% received less than 45 Gy of radiation therapy.

With evidence supporting the safety and tolerability of IVC in oncology care, some interesting information suggesting chemotherapy/radiation therapy sensitization and decreased oxidative stress with use of IVC concurrent to chemotherapy, and IVC’s apparent ability to improve adherence to standard treatment, it’s likely we will see phase II and III studies to guide use of this approach.

As a clinician, I currently advocate for use of IVC as a reasonable and rational approach for supportive care in oncology treatment and consider doses between 15 g and 75 g on an individual basis. I find these doses to be well-tolerated for people who have been carefully screened and deemed appropriate candidates to receive IVC by their providers.¹¹