High-Dose Intravenous Vitamin C Combined with Docetaxel in Men with Metastatic Castration-Resistant Prostate Cancer: A Randomized Placebo-Controlled Phase II Trial

Abstract High-dose intravenous vitamin C (HDIVC) administered to produce pharmacologic concentrations shows promise in preclinical models and small clinical trials, but larger prospective randomized trials are lacking. We evaluated the clinical benefit of combining HDIVC with docetaxel in patients with progressive metastatic castration-resistant prostate cancer (mCRPC). In this double-blind, placebo-controlled phase II trial, 47 patients were randomized 2:1 to receive docetaxel (75 mg/m2 i.v.) with either HDIVC (1 g/kg) or placebo. Coprimary endpoints were PSA50 response and adverse event rates. Secondary endpoints included overall survival, radiographic progression-free survival, and quality of life measured using the Functional Assessment of Cancer Therapy-Prostate instrument. Correlative analyses included pharmacokinetics and oxidative stress markers. Eighty-nine percent of patients previously had three or more lines of therapy. The PSA50 response rate was 41% in the HDIVC group and 33% in the placebo group (P = 0.44), with comparable adverse event rates in both groups. There were no significant differences in Functional Assessment of Cancer Therapy-Prostate scores. The median radiographic progression-free survival was not significantly different between the HDIVC and placebo groups, with durations of 10.1 and 10.0 months (HR, 1.35; 95% confidence interval, 0.66–2.75; P = 0.40), respectively. The median overall survival was 15.2 months in the HDIVC group and 29.5 months in the placebo group (HR, 1.98; 95% confidence interval, 0.85–4.58; P = 0.11). HDIVC did not decrease F2-isoprostanes, indicators of oxidative stress. The study was suspended after prespecified interim analysis indicated futility in achieving primary endpoints. In this patient population, combining HDIVC with docetaxel did not improve PSA response, toxicity, or other clinical outcomes compared with docetaxel alone. Findings do not support the routine use of HDIVC in mCRPC treatment outside of clinical trials. Significance: This is the first randomized, placebo-controlled, double-blind trial to evaluate HDIVC in cancer treatment. The addition of HDIVC to docetaxel in patients with mCRPC does not improve PSA response, toxicity, or other clinical outcomes compared with docetaxel alone. The routine use of HDIVC in mCRPC treatment is not supported outside of clinical trials.


Introduction
Vitamin C (ascorbic acid or ascorbate) is essential for humans.When taken orally, its concentrations in both plasma and tissues are tightly regulated by multiple mechanisms: intestinal absorption, tissue transport, renal reabsorption/excretion, and tissue utilization (1)(2)(3)(4).However, tight physiologic control is bypassed when vitamin C is administered parenterally.In animals and humans, parenteral administration uniquely results in pharmacologic ascorbic acid concentrations, which persists until renal excretion restores homeostasis (4).Only pharmacologic ascorbic acid concentrations, not physiologic concentrations, produce extracellular hydrogen peroxide as a prodrug for reactive oxygen species, which is effective in vitro and in animal models in limiting the growth of a wide variety of cancers (5,6).
Intravenous vitamin C (IVC) has a strong safety record, with only minimal adverse events (AE) reported among the approximately 10,000 individuals who undergo IVC treatment annually, including lethargy, fatigue, nausea, and vomiting in less than 1% (7).In small phase I/II clinical trials, IVC has shown promising efficacy across various cancers, including ovarian cancer, pancreatic cancer, glioblastoma, and multiple myeloma (8)(9)(10)(11)(12)(13)(14)(15).Notably, trials that combined high-dose IVC (HDIVC) with chemotherapy for advanced cancers indicated improved quality of life and reduced toxicities (11,16,17).Specifically, one randomized controlled trial involving newly diagnosed patients with ovarian cancer revealed that the addition of IVC to firstline treatment with paclitaxel and carboplatin chemotherapy prolonged time to disease progression and suggested favorable trends in overall survival (OS).Additionally, ascorbate addition led to a marked reduction in lowgrade toxicities associated with chemotherapy (11).
Metastatic castration-resistant prostate cancer (mCRPC) treatment poses a significant challenge, particularly when patients exhaust standard therapeutic avenues of androgen deprivation therapy and androgen receptor signaling inhibitors.Subsequent-line treatment options include docetaxel or cabazitaxel and are aimed at extending survival (18).The utility of taxane therapies is frequently limited by associated toxicities that encompass a spectrum of AEs ranging from low-grade symptoms like fatigue, nausea/ vomiting, neuropathy, bone pain, and anorexia to more severe grade 3 to 4 AEs like neutropenia, anemia, and thrombocytopenia.Such treatmentrelated complications often necessitate infusion delays, dose adjustments, or even discontinuation of therapy.Approximately 11% to 35% of men with mCRPC receiving docetaxel experience dose interruptions, underscoring the need for alternative or adjunctive treatments (18,19).

Because parenteral administration of pharmacologic doses of vitamin C
inhibited growth of CRPC in an animal model (20), a noncomparative phase II clinical study was previously conducted, wherein a cohort of 20 men diagnosed with mCRPC received a regimen of single-agent IVC once weekly with step-up dosing to a target dose of 60 g over 12 weeks (0.74 g/kg; ref. 21).
There were no reductions in PSA levels, oxidative damage markers (such as 8-oxoguanidine excretion), or improvements of bone metastases.However, the lack of response could be attributed to suboptimal dosing and singleagent therapy (21).
At present, there are no appropriately powered, randomized, prospective trials using HDIVC for the treatment of any cancer.We present here results from a randomized, placebo-controlled phase II clinical trial that evaluated the therapeutic potential of combining HDIVC with docetaxel in patients with mCRPC.We hypothesized that HDIVC would increase PSA response and/or mitigate toxicities.

Statistical analysis
The sample size of 63 patients (42 in

rPFS and OS
The median, 12-month, and 24-month rPFS and OS were estimated with the Kaplan-Meier (KM) method, and HRs and 95% confidence intervals (CI) with Cox proportional-hazards regression models.The median follow-up was calculated using the KM method.

PSA response rate
The Fisher exact test was used to compare PSA response rates.An interim analysis of PSA response applied a predictive probability approach; the trial would halt for futility if, after 30 patients completed the required PSA followup, the probability of concluding the trial with significant results was below

FACT-P
The FACT-P comprises two main components: FACT-General, a 27-item self-report questionnaire with four subscale domains (physical, social/family, emotional, and functional well-being) designed to measure general quality of life in patients with cancer, and a 12-item prostate cancer subscale (PCS) tailored to assess prostate cancer-specific quality of life.The FACT-P total score is calculated by summing the scores from the FACT-General subscales and the PCS, with higher total scores indicating better quality of life.Two additional scores from the FACT-P questionnaire were used: the FACT Advanced Prostate Symptom Index score, which includes eight items from the FACT-P, and the FACT-P PCS pain-related score, which comprises four questions from the FACT-P specifically addressing pain.As with the overall FACT-P, higher scores on these indices reflect better health-related quality of life.FACT-P total scores at cycles 4, 6, and 8 were evaluated with analysis of covariance, adjusting for baseline FACT-P total scores.

PK
PK parameters (maximum concentration and AUC) were compared between treatment arms using the Wilcoxon rank-sum test, with data presented as the mean ± SD.A P value <0.05 was defined to be statistically significant.

F2-IsoPs
Concentrations of F 2 -IsoPs were determined at Vanderbilt Eicosanoid Core Laboratory using gas chromatography/negative ion chemical ionization mass spectrometry assays.Whole blood was centrifuged at 4,000 g for 10 minutes to yield plasma, and 0.5 to 1 mL of plasma was used for the quantification of F 2 -IsoPs.The sample was derivatized to the pentafluorobenzyl ester, trimethylsilyl ether derivative for gas chromatography/negative ion chemical ionization mass spectrometry analysis.The lower limit of sensitivity was about 5 pg.The precision of the assay was ±6%, and the accuracy was 96%.
The final results were standardized and expressed as nanograms per milligram of creatinine.Comparisons were made between study arms based on infusions at cycles 4 and 6 using the two-sample t test.

Patients
Between June 20, 2016, and September 21, 2021, 62 patients were screened, and 50 were randomized in a 2:1 ratio, with 34 in the HDIVC group and 16 in the control group (Fig. 1).Baseline demographic and clinical characteristics (

Primary outcomes
In the HDIVC group, 41% of patients (13 of 32) achieved a PSA50 response compared with 33% (5 of 15) in the control group (

Secondary outcomes
The median rPFS was 10.1 months (95% CI, 5.85-14.In this study, 18 patients achieved a PSA response, with 5 in the control arm and 13 in the HDIVC arm.A post hoc analysis assessed the duration of PSA response, calculated from the initial day of the cycle wherein the response was first noted to the day of PSA progression defined as the initial PSA increase ≥25% and ≥2 ng/mL above the nadir.The median duration of PSA response (Fig. 2C) was shorter in the HDIVC group than in the control group, 4.3 versus 6.2 months, respectively (HR, 2.5; 95% CI, 0.52-12; P ¼ 0.25).

PK results
PK data for docetaxel were collected from 17 patients, with 7 patients (7/15, 47%) in the cohort receiving docetaxel with placebo and 10 patients (10/32, 31%) receiving docetaxel with HDIVC.The analysis revealed minimal changes in the maximum concentration achieved for docetaxel alone (2,367.1 ± 730.7 ng/mL) compared with when combined with HDIVC (2,762.0± 787.6 ng/mL; P ¼ 0.28).Additionally, there was no statistically significant difference in dose-normalized exposure for docetaxel alone (3,124.7 ± 404.0 ng � h/mL) or in combination with HDIVC (3,623.4± 1,432.9 ng � h/mL; P ¼ 0.33).However, we observed a trend toward higher docetaxel exposure in the plasma of patients receiving the combination treatment.
PK data for vitamin C were voluntary and only available for five subjects who received docetaxel with HDIVC and two subjects who received docetaxel alone (Supplementary Fig. S2).Vitamin C concentrations were as expected in those who received HDIVC (5,24) and in controls (1)(2)(3)(4).
F 2 -IsoPs were analyzed at completion of vitamin C infusions and 60 minutes postinfusion, and they were compared with baseline concentrations at cycles 4 and 6.There were no reductions in F 2 -IsoP concentrations when compared with baseline measurements, although sample sizes were limited (Supplementary Tables S13-S17).

Discussion
This is the first randomized, placebo-controlled, double-blind trial to evaluate HDIVC in cancer treatment.Here, we investigated PSA response rates and selected chemotherapy-related toxicities (fatigue, nausea, bone pain, and anorexia) in patients with mCRPC who received a combination of HDIVC with standard-of-care docetaxel.We also studied the impacts of HDIVC on other AEs, rPFS, OS, and quality-of-life metrics.Comparison between HDIVC and control groups revealed no significant difference in PSA response (PSA50), irrespective of prior docetaxel use in the hormone-sensitive setting.Similar patterns were observed across all endpoints without significant differences between groups.Interim analyses highlighting inadequate PSA responses led to trial suspension due to futility.Although the final sample size was limited by the futility analysis, the trial efficiently evaluated PSA50 response and toxicity outcomes in this patient population.The PSA50 response in the HDIVC arm at the time of the interim analysis, 41%, was below the trial design's hypothesized null, 45%, which was based on the TAX 327 trial (18).Additionally, with the posterior probability of having a successful trial less than 5% and median rPFS and OS both being shorter in the HDIVC arm at the interim, the decision to halt the trial could be made with a savings of 20% of the original planned size (50 vs. 63).Compared with controls, in men who received docetaxel and HDIVC, there were nonsignificant trends of increased AEs, reduced OS, and higher docetaxel exposure in plasma.
Two factors that might explain the lack of efficacy are frequency of administration and dose of HDIVC.Daily administration is common in preclinical models (20).However, twice-weekly administration, used here and in other clinical trials (10,11,15), may have been insufficient for mCRPC and was logistically challenging.Because most HDIVC safety data are for doses at or below 1 g/kg, we were reluctant to increase dosing (6,11,(13)(14)(15)25).
Other factors may have contributed to the lack of efficacy.The trial size was limited, and 89% of those enrolled had previously failed three or more lines of prior therapy.Prespecified coprimary endpoints may have been overly optimistic, such that the trial, in retrospect, perhaps was underpowered.The trend toward higher docetaxel exposure in the combination arm might have contributed to observed toxicity.We also noted that controls in this study lived unexpectedly longer (29.5 (24).Nonresponsiveness in some cases can be attributed to increased production of the enzyme catalase, which dismutates

5%.
Post hoc analysis involved the exact Cochran-Mantel-Haenszel test to explore the correlation of HDIVC treatment with PSA response, factoring in prior docetaxel exposure.An assumption of the Cochran-Mantel-Haenszel stratified analysis is that the ORs within each stratum (prior docetaxel, yes or no) are homogeneous.The Breslow-Day test for homogeneity of ORs is used to confirm this assumption for a stratified analysis.For the duration of PSA response, the end dates for several responders are censored, requiring a KM analysis for comparison.Statistical analyses were performed using R software, version 4.1.3(Comprehensive R Archive Network, www.cran.r-project.org), with a P value <0.05 defining statistical significance.

FIGURE 2
FIGURE 2 Key secondary endpoints.Key secondary endpoints include rPFS, OS, duration of PSA response, and quality of life measured using the FACT-P instrument.A, rPFS and (B) OS are presented for docetaxel + HDIVC vs. docetaxel + placebo.The KM method was used to estimate the median, 12-month, and 24-month rates of rPFS and OS.C, Duration of PSA response and (D) FACT-P scores are presented for docetaxel + HDIVC vs. docetaxel + placebo.
the HDIVC treatment group and 21 in Separate queries were conducted to analyze AE data specific to each study drug.The earliest instance of each AE with the highest grade and attribution combination for each patient and AE type was extracted.It is important to note that some patients experienced multiple types of AEs and recurring episodes of the same Common Terminology Criteria for Adverse Events code.In such cases, the earliest instance of that AE with the highest grade and attribution combination was reported.
the control group) would provide 80% power to detect the hypothesized 35% absolute improvement in PSA response with a one-sided 5% Fisher exact test.Efficacy and safety analyses were conducted on a modified intention-totreat basis, incorporating all treated patients.To preserve the coprimary AACRJournals.orgCancerResCommun; 4(8) August 2024 2175 A Randomized Trial of Vitamin C and Docetaxel in mCRPC significance level at 15%, the α-level was set at 5% for PSA response and 10% for toxicity.Detailed description of the statistical analysis methods is provided in the Supplementary Methods.ToxicityGrading of the coprimary toxicity outcome was categorical (no toxicities, grades 1-2, and grades 3-4), and the Cochran-Armitage trend test was used for analysis.

Table 2
).However, this 2176 Cancer Res Commun; 4(8) August 2024 https://doi.org/10.1158/2767-9764.CRC-24-0225 | CANCER RESEARCH COMMUNICATIONS Paller et al.difference failed to reach the anticipated 35% absolute improvement (P ¼ 0.44).Post hoc analysis stratified by prior docetaxel use corroborated these findings (OR, 1.26; 95% CI, 0.29-5.91;SupplementaryTableS2). Due to an insufficient PSA response rate in the interim analysis, accrual to the trial was suspended by the Data and Safety Monitoring Board for futility.Table3presents the worst grade of four AEs of interest (fatigue, nausea, bone pain, and anorexia) encountered by patients during the 24-week treatment.Overall, the groups exhibited comparable AE profiles (Supplementary TablesS3 and S4).Most

TABLE 3
(32)(33)(34)(35)(36)dpoints Randomized Trial of Vitamin C and Docetaxel in mCRPC in tumor growth compared with the groups treated with monotherapy (bioRxiv 2023.03.23.533944).Combination therapy of parenteral ascorbate may also be synergistic with checkpoint inhibitors and chemotherapy in animal and cell models(24,31,32).Certain patient subgroups with specific genetic mutations may be particularly susceptible to combination therapies involving HDIVC.For example, emerging preclinical studies indicate that HDIVC could display increased effectiveness in cancers with mutations in KRAS, BRAF, TET2, IDH1, IDH2, VHL, FH, or SDH and in those with mismatch repair deficiencies or high expressions of GLUT1, and at least some of these gene mutations are associated with prostate cancer(32)(33)(34)(35)(36).
cellular fluid by pharmacologic concentrations of vitamin C.Although nearly 80% of tested cancer cells show responsiveness to such levels of vitamin C, approximately 20% do not