Testosterone nadir and clinical outcomes in patients with advanced prostate cancer: Post hoc analysis of triptorelin pamoate Phase III studies

Abstract Objective The objective of the study is to evaluate whether low nadir testosterone during treatment with triptorelin pamoate, a luteinising hormone‐releasing hormone (LHRH) agonist, is associated with improved clinical outcomes in patients with advanced prostate cancer using a retrospective analysis of clinical trial data. Patients and methods Data were pooled from three prospective, 9–12‐month Phase III studies of triptorelin monotherapy in patients with advanced prostate cancer (including NCT00751790). The serum testosterone concentration suppression targets evaluated were <0.35 nmol/L (<10 ng/dl), <0.7 nmol/L (<20 ng/dl), <1.7 nmol/L (<50 ng/dl) and ≥1.7 nmol/L. Overall survival (OS) and disease‐specific survival (DSS) by testosterone suppression group were assessed by Kaplan–Meier analysis, with log‐rank test. The time frame for the primary analysis was Days 1–518 (median OS follow‐up 254 days [range, 29–518 days]) and for the sensitivity analyses was Days 1–262. Supplementary analyses combined the ≥0.7‐ to <1.7‐nmol/L and ≥1.7‐nmol/L groups. Results The sample size comprised 592 patients (most received triptorelin monotherapy; four reported concomitant androgen receptor‐axis–targeted therapy). Nadir testosterones of <0.35, ≥0.35 to <0.7, ≥0.7 to <1.7 and ≥1.7 nmol/L were achieved by 96%, 3.2%, 0.34% and 0.17% of patients, respectively. Better OS with decreasing level of nadir testosterone was observed (p < 0.001) and this persisted after sensitivity/supplemental analyses (all p < 0.001). Differences in DSS with decreasing levels of nadir testosterone were not statistically significant in the primary analysis. Sensitivity/supplemental analysis showed better DSS with decreasing level of nadir testosterone (Days 1–262, p = 0.01; combined groups Days 1–518, p = 0.03; combined groups Days 1–262, p = 0.005). Conclusion Low nadir testosterone achieved during treatment with the LHRH agonist triptorelin was associated with improved OS and DSS in patients with advanced prostate cancer.


| INTRODUCTION
Lower levels of testosterone in men on androgen-deprivation therapy have been linked with improved clinical outcomes in advanced prostate cancer. 1 In the setting of testosterone-lowering treatment, castration was originally defined as a serum testosterone level <1.7 nmol/L (<50 ng/dl).This target level was widely adopted by regulatory authorities. 2Clinical experience utilising contemporary assay technology has suggested that a lower serum testosterone level threshold may be more appropriate than the historic definition. 2 In the PR-7 study, nadir testosterone levels ≤0.7 nmol/L (≤20 ng/dl) within the first year of androgen-deprivation therapy correlated with improved cause-specific survival and duration of response in men with recurrent, nonmetastatic prostate cancer. 1 Triptorelin pamoate is a gonadotropin-releasing hormone agonist indicated for the treatment of advanced prostate cancer. 3Pooled data from nine prospective studies showed that 80%-90% of patients with advanced prostate cancer reached testosterone levels <0.7 nmol/L with triptorelin. 4Mean testosterone levels were >0.35 nmol/L (>10 ng/dl) at Month 1, and <0.35 nmol/L for Months 2-12. 4 Testosterone levels continued to decrease between Months 1 and 2. 4 Serum testosterone levels at the end of the study were lowest with the 1-month formulation (median, 0.1 nmol/L [3 ng/dl]; interquartile range, 0.1-0.23 nmol/L [3-7 ng/dl]) compared with the 3-month (0.2; 0.1-0.3nmol/L [6; 3-9 ng/dl]) and 6-month formulations (0.3; 0.2-0.5 nmol/L [9; 6-14 ng/dl]). 4In a Japanese study of androgendeprivation therapy, very low levels of serum testosterone (0.07-0.14 nmol/L [2-4 ng/dl]) were associated with a better prognosis in patients with metastatic prostate cancer and certain SRD5A2 polymorphisms. 5 This retrospective study evaluated data from phase III trials of triptorelin conducted by Debiopharm to assess whether nadir testosterone during treatment is a determinant of clinical endpoints in patients with advanced prostate cancer treated with triptorelin.

| PATIENTS AND METHODS
This was a retrospective pooled analysis of data from three prospective, Phase III studies of triptorelin in patients with advanced prostate cancer: DEB-96-TRI-01 1st phase, DEB-96-TRI-01 2nd phase and DEB-TRI6M-301 (clinicaltrials.govidentifier: NCT00751790) (Table 1). 4,6DEB-96-TRI-01 was not registered in a centralised database because, at the time it was implemented, this was not a requirement.All three studies were conducted in South Africa.The primary objective of the analysis was to explore whether low nadir serum testosterone during androgen-deprivation treatment improves overall survival (OS) and disease-specific survival (DSS) due to prostate cancer or complications of cancer treatment.Secondary objectives included exploring the proportion of time that testosterone was suppressed below the target level, the time to achieve a testosterone level below the suppression target and defining prostate-specific antigen (PSA) kinetics.

| Testosterone measurements and targets
Testosterone-derived variables were based on testosterone data collected at baseline, during the treatment period and during the 28-day safety follow-up period.The serum testosterone concentration suppression targets assessed were defined as <0.35 nmol/L, <0.7 nmol/L and <1.7 nmol/L, based on the historical suppression target level of 1.7 nmol/L 7 and currently recommended levels of 0.5-0.7 nmol/L (14-20 ng/dl). 7Testosterone was measured by validated radioimmunoassay or liquid chromatography tandem mass spectrometry. 4The testosterone lower limit of quantification (LLOQ) was 0.

| Endpoints
The primary efficacy endpoints were OS and DSS.OS was defined as the elapsed time between the first dose and death from any cause.
Patients who did not die were censored at the last date of contact.
DSS was defined as the elapsed time between first dose and death

| RESULTS
The intention-to-treat (ITT) analysis set included a total of 592 patients (Table 2).Across the studies, most patients (54%) were White (range, 47%-64%) with a mean age of 70.6 years (range, 69.7-71.1 years) [Correction added on 12 February 2024, after first online publication: A percent sign has been added after 47 in the preceding sentence].

| Testosterone concentrations
Mean changes and fold changes from baseline in testosterone and PSA levels by treatment formulation over time are shown in Figure 1.
The proportion of time that testosterone was <0.35 nmol/L ranged from 60% to 87% across the three treatments; for <0.7 nmol/L, it ranged from 92% to 95%; and for <1.7 nmol/L, it was 99% (Table 3).
The time to (first) testosterone level <0.35 nmol/L ranged from 33 to 77 days across the three treatments; for <0.7 nmol/L, it ranged from 32 to 42 days; and for <1.7 nmol/L, it ranged from 30 to 32 days (Table 3).Owing to the small number of patients with nadir testosterone ≥0.7 to <1.7 nmol/L (n = 2) and ≥1.7 nmol/L (n = 1), data are not reported for these groups.

| Overall survival
The probability of survival over time (Days 1-518) was higher in the nadir testosterone <0.35 nmol/L group than in the ≥0.35 to <0.7 nmol/L group (Figure 2A).In the ≥0.7 to <1.7 nmol/L and ≥1.7 nmol/L nadir testosterone groups, the OS trend was consistent but inconclusive due to the small number of patients and short follow-up.This trend for better OS with decreasing levels of nadir testosterone persisted in the sensitivity analysis, which considered T A B L E 2 Summary of demographic and baseline disease characteristics (ITT).'DEB-TRI6M-301 TNM staging according to International Union Against Cancer (UICC) TNM classification of malignant tumours, 6th edition, 2002.9 Abbreviations: ITT, intention to treat; SD, standard deviation; X, cannot be assessed.
durations from Days 1-262 (log-rank for difference between groups, p < 0.001).Results from the landmark analysis of OS (landmark at Day 57), shown in Figure 2B, support these findings.
No linear trend and no statistically significant differences in OS were observed by proportion of time with testosterone levels below the suppression target or time to (first) testosterone below the suppression target (all log-rank p > 0.10).

| Disease-specific survival
A numerically higher probability of DSS over time was observed with decreasing levels of nadir testosterone.Specifically, for groups    These findings are consistent with the pooled analysis of nine prospective studies (including the three from the current analysis), which showed that 80%-90% of patients achieved testosterone levels ≤0.7 nmol/L during 12 months of triptorelin treatment. 4This study shows an additional benefit with further suppression of testosterone ≤0.35 nmol/L compared with those with testosterone between 0.35 and 0.7 nmol/L.This corroborates evidence from prior studies indicating an association between lower testosterone levels during androgen-deprivation therapy and increased OS in patients with metastatic prostate cancer. 5Median testosterone concentrations at the therapy. 11However, the ICELAND study authors noted that the results may have been due to very few patients (n = 3) in the highest median testosterone group.A trend in the same direction as in the PR-7 trial was noted when patients were stratified by minimum and median testosterone levels. 11though an effect of nadir testosterone group on OS was observed, the findings for DSS were less conclusive.This likely reflects, in part, the relatively few prostate cancer deaths in the cohort resulting in lack of power to demonstrate a statistically significant difference, but beyond this, the precise reasons are unclear.
A strength of the current analysis is that the included clinical studies were similar in terms of design, enrolled populations and methodology used for testosterone measurement [Correction added on 12 February 2024, after first online publication: "Limitations of this study:" has been removed from the preceding sentence.]. 4 Given the dearth of large randomised clinical trials for androgen-deprivation therapy, this retrospective analysis helps to address a gap in the literature.Limitations include that this was a noncomparative, retrospective, single-arm analysis.Included studies were conducted before the availability of newer targeted therapies and were too short in duration to allow for a meaningful analysis of the time to castration resistance.
However, although androgen-deprivation therapy is no longer recommended as monotherapy in current guidelines, it was at the time the trials were conducted and is still likely to be used at times in clinical practice.Most patients were treated with luteinising hormonereleasing hormone (LHRH) agonist monotherapy, although four patients reported concomitant androgen receptor-axis-targeted therapy and were not excluded (two patients from the DEB-96-TRI-01 second phase, 3.75-mg group and two from the DEB-TRI6M-3, 22.5-mg group).Inclusion of these four patients is not expected to have a significant impact on the reported outcomes.The significance of testosterone levels in men treated with anti-androgen or androgen receptor-axis-targeted therapies cannot be extrapolated from these data.Testosterone suppression levels were not evaluated by age, although previous studies did not find a correlation between these variables. 5Nadir testosterone suppression did not appear to be associated with extent of disease; however, baseline disease stages were not balanced (most patients were T = 3, N = X and M = 1).

| CONCLUSIONS
A pooled analysis of data from three Phase III studies supports the observation that very low nadir testosterone achieved during monotherapy treatment with the LHRH agonist triptorelin is associated with improved OS and DSS in patients with advanced prostate cancer.
Deep testosterone suppression <0.35 nmol/L was associated with improved survival compared with those whose testosterone was between 0.35 and 0.7 nmol/L.
due to prostate cancer (preferred term of adverse event with fatal outcome: asthenia, malignant neoplasm progression, metastases to liver, prostate cancer and pulmonary mass) or a complication of cancer treatment.Patients who died from other causes or who did not die from prostate cancer or a complication of cancer treatment were censored at the last date of contact.The primary explanatory risk variable was serum nadir testosterone level achieved during androgen-deprivation treatment with triptorelin.Secondary explanatory risk variables included the proportion of time with serum testosterone below the suppression target and time to (first) testosterone below the suppression target.Descriptive variables were serum testosterone and PSA concentrations and the change in these variables from baseline.
DEB-96-TRI-01 first phase DEB-96-TRI-01 second phase DEB-96-TRI-01 combined first + second phase DEB-96-TRI-01 1st phase DEB-All patients were male.All patients had staging established by histology/cytology.DEB-96-TRI-01 1st phase and DEB-96-TRI-01 2nd phase TNM staging according to Schroder et al. 8 with adequate sample size and follow-up, a difference in the DSS survival curves was observed between the group with nadir serum testosterone <0.35 nmol/L and the ≥0.35 to <0.7 nmol/L group (Figure3A).However, this difference was not statistically significant (Days 1-518, log-rank for difference between the four nadir testosterone groups p = 0.08).The sensitivity analysis, which considered durations from Days 1-262, supported the observed difference in DSS (log-rank p < 0.05).Results from the landmark analysis of DSS (landmark at Day 57), shown in Figure3B, support these findings.No statistically significant differences in DSS were observed by proportion of time with testosterone levels below the suppression target or time to (first) testosterone below the suppression target (all log-rank p > 0.10).F I G U R E 1 Geometric mean (left panels) and fold change from baseline (right panels) over time in testosterone and PSA concentrations in patients treated with (A) triptorelin 3.75 mg (1-month formulation), (B) triptorelin 11.25 mg (3-month formulation) and (C) triptorelin 22.5 mg (6-month formulation) (ITT analysis set).Dotted reference lines in left panels correspond to testosterone suppression targets: <0.35 nmol/L (<10 ng/dl), <0.7 nmol/L (<20 ng/dl), <1.7 nmol/L (<50 ng/dl).To convert testosterone concentrations from nmol/L to ng/dL, divide by 0.0347.ITT, intention to treat; PSA, prostate-specific antigen.

1
Overview of included studies.
Testosterone concentration at baseline, nadir, maximum and end of study; testosterone suppression groups; proportion of time testosterone below target suppression and time to testosterone target suppression (ITT analysis set).
4 | DISCUSSIONIn this analysis of 592 patients with advanced prostate cancer, almost all patients (96%) achieved low nadir testosterone levels ≤0.35 nmol/L during treatment with triptorelin, and those low nadir testosterone levels correlated with improved OS.A similar nonsignificant difference was observed for DSS.Rates of OS and DSS were not affected by proportion of time with testosterone T A B L E 3 Kaplan-Meier plots of DSS by testosterone suppression group for (A) primary analysis and (B) landmark analysis, with landmark at Day 57.Number of subjects at risk and DSS probability estimates at Days 169 and 337 (Days 169 and 262 for sensitivity analysis) are shown underneath each graph.Data are shown for primary analysis and landmark analysis (time frame Days 1-518; left panel) and sensitivity analysis (time frame Days 1-262; right panel), with number of subjects at risk (intention-to-treat [ITT] analysis set).CI, confidence interval; DSS, diseasespecific survival; ITT, intention to treat; MT, maximum testosterone; NT, nadir testosterone.
10 testosterone nadir of ≤0.7 nmol/L correlated with improved outcomes in the PR-7 study.10A pc analysis of the ICELAND study found no significant differences in cause-specific survival and time to PSA progression among testosterone level subgroups in men with advanced prostate cancer in Year 1 of androgen-deprivationF I G U R E 3