Testosterone Breakthrough Rates during Androgen Deprivation Therapy for Castration Sensitive Prostate Cancer
2020; Lippincott Williams & Wilkins; Volume: 204; Issue: 3 Linguagem: Inglês
10.1097/ju.0000000000000809
ISSN1527-3792
AutoresFred Saad, Neil Fleshner, Tom Pickles, Tamim Niazi, Himu Lukka, Frédéric Pouliot, Ilidio Martins, Laurence Klotz,
Tópico(s)PARP inhibition in cancer therapy
ResumoYou have accessJournal of UrologyReview Article1 Sep 2020Testosterone Breakthrough Rates during Androgen Deprivation Therapy for Castration Sensitive Prostate Cancer Fred Saad, Neil Fleshner, Tom Pickles, Tamim Niazi, Himu Lukka, Frederic Pouliot, Ilidio Martins, and Laurence Klotz Fred SaadFred Saad *Correspondence: Institut du Cancer de Montréal/CRCHUM, CHUM—Pavillon R, 900, rue St-Denis, porte R10-464, Montréal, Québec H2X 0A9 telephone: 514-890-8000; ext 36446; FAX: 514-412-7620; E-mail Address: [email protected] Centre Hospitalier de l’Université de Montréal, University of Montreal, Montreal, Quebec, Canada , Neil FleshnerNeil Fleshner Mount Sinai Hospital, Princess Margaret Cancer Centre (UHN), Toronto General Hospital (UHN), University of Toronto, Toronto, Ontario, Canada , Tom PicklesTom Pickles British Colombia Cancer Agency, University of British Columbia (UBC), Vancouver, British Columbia , Tamim NiaziTamim Niazi McGill University, Montreal, Quebec, Canada , Himu LukkaHimu Lukka McMaster University, Hamilton, Ontario, Canada , Frederic PouliotFrederic Pouliot Laval University, Quebec City, Quebec, Canada , Ilidio MartinsIlidio Martins Kaleidoscope Strategic Inc., Toronto, Ontario, Canada , and Laurence KlotzLaurence Klotz Sunnybrook Health Sciences Centre, University of Toronto, Toronto, Ontario, Canada View All Author Informationhttps://doi.org/10.1097/JU.0000000000000809AboutAbstractPDF ToolsAdd to favoritesDownload CitationsTrack CitationsPermissionsReprints ShareFacebookTwitterLinked InEmail Abstract Purpose: Androgen deprivation therapy is an established therapy for castration sensitive prostate cancer and recent studies have observed that patients whose testosterone levels are suppressed below 0.7 nmol/l have improved outcomes. Testosterone breakthrough, or a rise in testosterone above a target threshold after the first month of androgen deprivation therapy, is generally associated with treatment deficiency. The purpose of this review is to summarize breakthrough rate data and explore the relationship to clinical outcomes in patients with castration sensitive prostate cancer receiving androgen deprivation therapy. Materials and Methods: Our systematic search identified 45 studies with a total of 52 cohorts representing 6,047 total patients reporting testosterone breakthrough rates or derivative measures above the thresholds of 1.7 nmol/l (51 cohorts, 6,015 patients) or 0.7 nmol/l (15 cohorts, 2,495 patients). Results: Significantly higher weighted mean breakthrough rates were seen for the 0.7 nmol/l threshold compared to 1.7 nmol/l (41.3% vs 6.9%, p <0.0001). A significant association between breakthrough rates and worse clinical outcomes overall was not found, although when larger trials (sample size greater than 100) and higher event rates (greater than 50%) were considered for the lowest threshold, significant associations between breakthrough rates and clinical outcomes were observed. Clinical factors such as administration and monitoring frequency, type of testosterone assay and type of androgen deprivation therapy did not significantly affect breakthrough rates, although nonvalidated assays were associated with a large degree of variability. Conclusions: Results from our analysis indicate that testosterone breakthroughs likely result in worse clinical outcomes and should be avoided. Moreover, there is a need to standardize assessment of testosterone levels both clinically and in the research context to better inform treatment decisions and improve the reliability and comparability of results across studies. Abbreviations and Acronyms ADT androgen deprivation therapy CLIA chemiluminescent immunoassay CRPC castration resistant prostate cancer CSPC castration sensitive prostate cancer LC-MS liquid chromatography-mass spectrometry LHRH luteinizing hormone-releasing hormone PRISMA Preferred Reporting Items for Systematic Reviews and Meta-Analyses PSA prostate specific antigen RIA radioimmunoassay RT radiotherapy T testosterone Pathogenesis and progression of prostate cancer is dependent on the activity of the androgen receptor,1 which drives nuclear localization and gene transcription after binding to testosterone and dihydrotestosterone.2 Medical androgen deprivation therapy achieves low levels of testosterone by decreasing androgen receptor activity through reversible means using a luteinizing hormone-releasing hormone analog. This approach is often preferred to orchiectomy due to the potential for lifelong negative physical effects of permanent castration and the aversion of men to surgical castration.3–7 LHRH agonists were the first type of LHRH analogs developed, with regulatory approvals starting in the 1980s, followed by antagonists in the 2000s.8–10 For patients with nonmetastatic castration sensitive prostate cancer, guidelines recommend observation or consideration of ADT for patients with a substantial rise in PSA or a rapid shortening of PSA doubling time following radical local therapy. Recent recommendations for newly diagnosed metastatic CSPC include the initiation of ADT11,12 and further therapy based on the extent of disease.11–13 The historic testosterone threshold for castration was 1.7 nmol/l, which largely reflected the inaccuracies of the testosterone assay used at that time for measuring levels below 1.7 nmol/l. Recently, many studies have observed that patients whose testosterone is suppressed below 0.7 nmol/l have an improved time to castration resistant prostate cancer and prostate cancer survival compared to those whose testosterone is not consistently suppressed below this level.11,14–17 There are 4 major categories of testosterone fluctuation that can occur during ADT administration. The first is an initial surge or flare, which is associated with a rise in testosterone due to LHRH receptor stimulation following the initiation of ADT with LHRH agonists.18–24 The second is a micro-surge or acute-on-chronic response, which reflects a transient testosterone increase also caused by LHRH receptor stimulation associated with subsequent agonist injections.19,20,22–27 Testosterone breakthrough (hereafter referred to as breakthrough) is a rise in testosterone above a predetermined threshold occurring any time after the first month of treatment.9,19,25–27 Despite difficulties in establishing their causation based on commonly reported data, breakthroughs are generally indicative of treatment deficiency.19 For example, they could be due to administration or device failures or the tapering off of drug levels towards the end of an administration cycle, or could indicate resistance of the pituitary-gonadal axis to ADT. Finally, treatment failure refers to a prolonged breakthrough defined as 2 or more testosterone measurements above threshold.28 While nadir and median testosterone values have historically been investigated as prognostic or predictive factors, clinical decision making is commonly based on maximum or breakthrough values. Although it has been hypothesized that breakthroughs may be associated with worse clinical outcomes,15,16,28 few published reviews have comprehensively assessed breakthrough rates or their clinical impact. Therefore, a review of current data on breakthrough during ADT for CSPC is warranted. This review will close this gap by quantifying breakthrough rates among patients with CSPC receiving ADT and exploring the relationship between breakthrough rates and clinical outcomes. Methods PubMed® (to January 14, 2019), the proceedings of the American Society of Clinical Oncology (ASCO) 2017-2018, the ASCO Genito-Urinary symposium 2017-2018 and the Annual Congress of the European Society for Medical Oncology 2017-2018 meetings were searched for prospective prostate cancer studies reporting breakthrough rates during ADT (PRISMA diagram, fig. 1) using the key search terms “prostate cancer” AND “androgen deprivation therapy” AND “testosterone” AND “breakthrough/escape” OR respective aliases. A supplemental bibliographic search of review articles and pooled/meta-analyses was also conducted.9,24–26,29–33 Figure 1. PRISMA diagram English language records were vetted at abstract level and confirmed at full text as needed. Only prospective studies (including case series with prospective testosterone assessment) with published or presented clinical results evaluating ADT for predominately castration sensitive populations that reported testosterone levels during ADT and beyond the first month of treatment were eligible. Studies which performed assessments at fixed time points and reported breakthrough rates at well-defined time points or time frames were included (PRISMA diagram, fig. 1). Prior or duplicate reports of the same study were excluded, as were studies in which type of ADT agent was unclear. Testosterone levels during ADT are commonly reported as nadir (minimum) values in a time frame. In our analysis, breakthrough rates are defined as the percentage of patients with any testosterone measurement above a certain threshold (0.7 to 1.7 nmol/l) or within a range (0.7 to 1.7 nmol/l) at a specific time point or within a prespecified time frame. Testosterone measurements were commonly performed on the same day as drug administration. However, studies did not often provide sufficient detail regarding the timing of testosterone sampling relative to drug administration to distinguish between a breakthrough event and an acute-on-chronic response. Similarly, while some studies excluded patients with treatment failure (confirmed breakthrough), most studies did not provide sufficient detail to distinguish between breakthroughs and treatment failure. Breakthrough rates were extracted or derived from full text sources. Only a single value per study cohort or arm (referred to hereafter as cohort) was included in the analysis. In studies reporting values at multiple time frames, the value for the longest time frame was used. In studies reporting multiple time points, the time point with the highest (maximum) rate was used. Achievement of castration levels after initiation of ADT was reported in most studies, although studies were not excluded if this value was not clearly indicated. The total number of testosterone assessable patients per cohort was used to calculate weighted means and medians and respective IQRs. Statistical significance of differences between groups was assessed using the t-test (using GraphPad QuickCalcs) for pairwise comparisons and 1-way analysis of variance (ANOVA) when comparisons involved 3 or more groups, using the conventional threshold for significance of α=0.05. For the plot of breakthrough rates by the number of assessments, the number of “relevant” testosterone assessments was estimated as the number of measurements taken at least 1 week apart that contributed to the reported breakthrough rate. For example, if testosterone was assessed at multiple time points during a study yet the breakthrough rate was reported only at 1 specific point, then the number of relevant assessments was interpreted as 1. Findings and Discussion The literature search identified 45 studies with a total of 52 cohorts reporting breakthrough rates or derivative measures beyond the first month (fig. 1), representing 6,047 total patients. Fifty-one cohorts (6,015 patients) reported breakthrough rates above the 1.7 nmol/l threshold and 15 cohorts (2,495 patients) reported breakthrough rates above the 0.7 nmol/l threshold. Although data reporting across studies was highly variable, the majority of studies reported breakthrough rates as the percent of patients with testosterone levels above a given threshold at a particular time point or within a time frame. When followup periods were considered by assigning testosterone measures collected in a given time frame to the upper time point, the most common followup periods were 6 and 12 months regardless of the threshold used. Few studies extended testosterone followup beyond the first year.34–38 The followup periods reported among the 45 eligible studies showed good alignment with the most commonly recommended 3 to 6-month monitoring schedules and the 12-month followup is in line with guideline recommendations, which confirm the need for followup during the first year.11,28,39,40 Testosterone Breakthrough Rates by Threshold Mean and median breakthrough rates for any time point or time frame were used for our analysis of breakthrough rates at thresholds of 0.7, 0.7 to 1.7 and 1.7 nmol/l. Higher weighted mean breakthrough rates were seen for the lowest threshold (0.7 nmol/l, 15 cohorts, 41.3% [IQR 12.0– 73.0])16,37,41–52 with slightly lower rates seen for the 0.7 to 1.7 nmol/l range (14 cohorts, 31.6% [IQR 11.0–50.0], table 1).16,37,41–50,52 The lowest breakthrough rate was apparent for the 1.7 nmol/l threshold (51 cohorts, 6.9% [IQR 2.0–9.6]; fig. 1, table 1)16,34–37,41–50,52–75 and this rate was highly statistically significantly different from the 0.7 nmol/l threshold (2-tailed t-test p <0.0001). Table 1. Testosterone breakthrough rate weighted means by threshold, administration schedule, type of assay and type of ADT agent No. Cohorts (No. Pts) Mean % (IQR) p Value Threshold (nmol/l): 0.7 15 (2,495) 41.3 (12.0–73.0)* 0.7–1.7 14 (2,463) 31.6 (11.0–50.0)* 1.7 51 (6,015) 6.9 (2.0–9.6)* 1-Way ANOVA <0.00001 t-Test summary (0.7 vs 1.7 nmol/l)† <0.0001 Administration schedule at 1.7 nmol/l threshold: Monthly 20 (2,263) 4.7 (1.7–7.0)* 3-Mo 17 (1,850) 6.0 (2.8–10.7)* 6-Mo 6 (656) 4.9 (2.0–6.6)* 1-Way ANOVA 0.82 t-Test summary Not applicable Assay type at 1.7 nmol/l threshold: CLIA/unspecified 25 (3,461) 9.1 (2.3–10.7)* RIA 10 (660) 1.6 (0.0–3.0)* LC-MS 7 (1,056) 4.3 (2.8–6.6)* Unspecified (low) 9 (838) 5.1 (2.8–7.0)* 1-Way ANOVA 0.26 t-Test summary (CLIA/unspecified vs RIA)† 0.084 t-Test summary (CLIA/unspecified vs LC-MS)† 0.61 Assay type at 0.7 nmol/l threshold: CLIA/unspecified 11 (2,156) 46.5 (25.5–73.0)* RIA 3 (310) 8.2 (2.5–12.0)* LC-MS 1 (29) 10.0 (not applicable)* 1-Way ANOVA Not applicable t-Test summary (CLIA/unspecified vs RIA)† 0.14 Agent type at 1.7 nmol/l threshold: Agonist 35 (3,589) 5.1 (1.0–7.8) Antagonist 16 (1,844) 4.8 (2.0–8.0) 1-Way ANOVA Not applicable t-Test summary 0.82 Agent type at 0.7 nmol/l threshold: Agonist 14 (1,868) 30.6 (10.0–52.0) Antagonist 0(0) Not applicable 1-Way ANOVA Not applicable t-Test summary Not applicable Weighted mean. 2-Tailed p value. The range of breakthrough rates reported in our analysis at 1.7 nmol/l (0% to 24.7%) was comparable to those reported in previous reviews (0% to 24%),9,19,24–26,33 despite differences in search methodologies and eligibility criteria. Our analysis also showed a wider range of breakthrough rates for 0.7 nmol/l (2.5% to 75%) compared with 1.7 nmol/l (0% to 24.7%) with no studies reporting zero breakthrough rates at the lowest threshold. Given that testosterone levels maintained below 0.7 nmol/l have been associated with improved outcomes,11,14–17,28 our findings underscore the heterogeneity of both individuals and cohorts that have achieved testosterone levels below 0.7 using LHRH agonists. Correlation between Testosterone Breakthrough Rates and Clinical Outcomes The clinical impact of breakthrough rates is relevant to the appropriate management of patients on ADT. To address this issue, prospective studies reporting any measure of correlation between breakthrough rates and clinical outcomes were identified. Seven trials were identified, 3 secondary analyses of phase III trials and 4 prospective cohort trials (table 2). Table 2. Qualitative analysis of correlation between breakthrough rates and outcomes in clinical trials evaluating patients with CSPC treated with ADT Study No. Pts Treatment T Monitoring Period; No. T Measures Median Mos Followup Threshold Comparison (nmol/l) Reported Correlation with Outcomes Survival (castrate specific survival or overall survival) (95% CI) Recurrence or Progression (95% CI) Klotz 201516Phase III PR-7 626 Orchiectomy or leuprorelin 12 Mos; 6 (max) 83 Greater than 1.7 vs less than 0.7 Not reported HR 1.59 (1.12–2.21) p=0.03 (progression to CRPC) 0.7–1.7 vs Less than 0.7 Not reported HR 1.13 (0.83–1.55) p=not significant (progression to CRPC) Nabid 201742Phase III PCS III High risk localized 456 Goserelin + antiandrogen with RT 6 Mos; 1 113 Greater than 1.7 vs 0.7–1.7 vs less than 0.7 p=0.727 p=0.195 (PSA recurrence) p=0.917 (progression*) Intermediate risk localized 347 Goserelin + antiandrogen with RT or RT alone p=0.930 p=0.767 (PSA recurrence) p=0.201 (progression*) Not reported Tombal 201752Phase IIIb ICELAND 345 Leuprorelin + antiandrogen 12 (6–18) Mos; 4 ∼56 (max) Greater than 1.7 vs less than 0.7 HR 3.59 (0.9–10.0) HR 4.57 (0.2–26.8) (PSA progression*) 0.7–1.7 vs Less than 0.7 HR 1.13 (0.6–1.9) HR 2.79 (0.8–9.3) (PSA progression*) Wang 201745 Prospective cohort 206 LHRH analog 6 Mos; 1 14 0.7 Not reported HR 1.99 p=0.001 (progression to CRPC) Bertaglia 201378 Prospective cohort 153 LHRH analog 6–7 Mos; 1 65 1.7 HR 1.35†p=0.32 HR 1.19†p=0.51 (progression*) 1.1 HR 2.22†p=0.034 HR 1.32†p=0.30 (progression*) 0.7 HR 5.26†p=0.020‡ HR 1.72†p=0.12 (progression*) Morote 200737 Prospective cohort 73 LHRH agonist 6 Mos; 3 54.1 (mean) 1.7 Not reported HR 2.8 p=0.008 (PSA progression*) 1.1 Not reported p 0.05 (progression to CRPC) 0.7 Not reported p >0.05 (progression to CRPC) Likely refers to progression “to CRPC” but there were insufficient data at source to confirm. Reported HRs were inverted (1/HR) in order to obtain values consistent with direction of comparisons in other studies. However, on multivariate analysis serum testosterone levels failed to be prognostic in terms of time to progression or overall survival. This correlation was deemed negative based on the finding that “The lowest serum testosterone threshold that was able to significantly distinguish groups related with the survival free of AIP was 32 ng/dl.” The evaluable cohorts of the phase III trials ranged in size between 345 and 626 patients and reported overall results from testosterone measurements over a period of 6 or 12 months (table 2). Studies compared clinical outcomes associated with various ranges of maximum testosterone levels (less than 0.7, 0.7 to 1.7 and greater than 1.7 nmol/l), with 6 cohorts assessing outcomes in recurrent/metastatic CSPC and 2 in localized disease. For recurrent disease 2 trials compared the clinical outcomes of patients with breakthrough rates above 1.7 nmol/l to those with breakthrough rates below 0.7 nmol/l.16,76 The PR-7 trial (626) showed a significantly increased risk of progression to CRPC for patients with breakthroughs above the upper threshold relative to those that maintained testosterone in the lower range (HR 1.59, 95% CI 1.12-2.21, p=0.03).16 However, the smaller ICELAND study (345) showed no associations in a similar comparison for biochemical failure (HR 4.57, 95% CI 0.2─26.8) or survival (HR 3.59, 95% CI 0.9-10.0).76 Two trials compared clinical outcomes of patients with breakthroughs between 0.7 and 1.7 nmol/l to those below 0.7 nmol/l, and neither found significant associations with PSA progression, progression to CRPC or survival.16,76 The large PCS III trial (803) evaluated the association between clinical outcomes and breakthrough rates at all 3 thresholds (less than 0.7 vs 0.7 to 1.7 vs greater than 1.7 nmol/l) for patients with localized prostate cancer receiving treatment with radiotherapy and short-term adjuvant ADT.42,77 No significant association between breakthrough rates and progression or survival were reported, which is not surprising given that many patients are cured in this setting, resulting in a diluted effect. Four prospective cohort studies reported associations between breakthrough rates at various thresholds (0.7, 1.1 and 1.7 nmol/l) and clinical outcomes in recurrent/metastatic disease (table 2).37,45,78,79 At the 0.7 and 1.1 nmol/l thresholds, significant associations between testosterone levels above threshold and increased risk of progression were found by Wang et al (2017, 206 patients; HR 1.99, p=0.001)45 and Morote et al (2007, 73, p <0.03),37 respectively. Bertaglia et al (2013) found no association with progression at any threshold (153; 0.7: HR 1.72, p=0.12; 1.1: HR 1.32, p=0.30),78 although a significant association with improved survival was reported for both thresholds (0.7 nmol/l: HR 5.26, p=0.020; and 1.1 nmol/l: HR 2.22, p=0.034, table 2).78 No significant associations were found by Dason et al. (2013) at either of these thresholds.79 When associations with clinical outcomes were considered at the highest threshold (1.7 nmol/l), a significant association between breakthrough and clinical outcomes was reported in Morote et al (2007, sample size 73, progression HR 2.8, p=0.008)37 but not Bertaglia et al (2013, 153, progression HR 1.19, p=0.51; survival HR 1.35, p=0.32).78 Based on our analysis it is unclear whether breakthrough rates have an impact on clinical outcomes. It is possible that associations between breakthroughs and outcomes may have been mitigated due to the use of additional therapies (such as adding an antiandrogen or switching to a different LHRH analog) although studies reviewed did not report subsequent therapies.15 Another confounding factor could be the dramatic heterogeneity in the methodology of the trials reviewed, which ranged from differences in the primary focus of the trial to differences in dosing and administration protocols. There was also variation in the methods by which the breakthrough analyses were conducted and outcomes reported. Although a significant association was not observed in most analyses, the direction of the correlations was consistent across studies with HRs above 1 (ie higher risk of worse outcome with breakthroughs above a particular threshold). Additionally, it must be noted that for most trials, comparisons were exploratory in nature and studies were not often powered to detect significant differences, a factor that would affect smaller trials and studies with low event rates or shorter followup to a greater extent than larger trials. When outcomes from larger trials (sample size greater than 100) and higher event rates (greater than 50%) were considered for the lowest threshold, significant associations between breakthrough rates and clinical outcomes were found in PR-7,16 Wang et al 201745 and Bertaglia et al 2013.78 This association is also supported by large retrospective cohort studies.15,80 Despite preclinical data indicating that testosterone stimulates prostate cancer growth81 and clinical data showing that androgen targeted therapy improves clinical outcomes,11,12,17,28,40 the current analysis was unable to establish a clear correlation between breakthrough rates and clinical outcomes. However, the larger trials generally did show a significant adverse effect on outcomes. The smaller trials in most cases showed an adverse effect, although not reaching statistical significance, suggesting they were underpowered. Therefore, we recommend that breakthroughs are likely detrimental and should be avoided. Clinical Factors Influencing Testosterone Breakthrough Rates Considerable variability in breakthrough rates was evident among the studies reviewed, especially for the 0.7 nmol/l threshold (IQR 12.0–73.0, fig. 2). There is a broad range of clinical factors that can affect breakthrough rates, including the natural variability of testosterone fluctuations in castrated patients82 and the greater breakthrough rates seen in younger compared to older men.16 Administration schedule,83,84 monitoring frequency,40,85,86 types of assays used for testosterone level detection,28,40,86 and types of agents40,87 may also impact breakthrough rates. To better understand the impact of these later factors, additional exploratory analyses were conducted. Figure 2. Testosterone breakthrough rates by threshold in patients with CSPC treated with ADT. Administration Schedule and Testosterone Monitoring Frequency Some have speculated on the potential influence of administration schedule and/or monitoring frequency on breakthrough rates.40,83–86 To assess this relationship, breakthroughs at the most commonly reported threshold (greater than 1.7 nmol/l) were analyzed for patients on ADT with respect to 1, 3 and 6-month administration schedules (fig. 3, A). The most common schedule was monthly (20 cohorts) followed by the 3 (17 cohorts) and 6-month (6 cohorts) schedules. Weighted means were numerically higher for the 3-month (6.0% [IQR 2.8–10.7]) compared with the monthly (4.7% [IQR 1.7–7.0]) or 6-month (4.9% [IQR 2.0–6.6]) schedules, although differences were not statistically significant (p=0.82, 1-way ANOVA). To assess the impact of monitoring frequency on breakthrough rates, only assessments taken at least 1 week apart were considered. Although the number of testosterone assessments within the first 12 months varied considerably (1 to 25), most studies assessed testosterone between 10 and 12 times with no apparent association between breakthrough rates and monitoring frequency (fig. 3, B). Our analysis did not identify any administration or monitoring factors that significantly affected breakthrough rates for patients receiving ADT. However, other factors such as longer followup periods are known to affect breakthrough rates, as seen in the phase III ICELAND study which reported a breakthrough rate of 2.9% with a total followup of 12 months and 6% at 42 months.38,52 These findings do not account for increases in breakthrough rates related to patient compliance,88,89 administration errors,90–92 failures to adhere to the recommended administration schedule or breakthrough rates while on intermittent ADT.93,94 The impact of rising testosterone while on a break from ADT therapy was not assessed in this analysis and is likely not as clinically relevant, as testosterone increases in this context are not indicative of treatment failure.17 Figure 3. Relationship between testosterone breakthrough rates at 1.7 nmol/l threshold, and administration schedule (A) and monitoring frequency (B) in patients with CSPC treated with ADT. Type of Assay Testosterone assays have varying sensitivity and specificity in detecting testosterone, especially at lower levels.28,95–97 Early hormone assays had limited sensitivity and this had a significant influence on establishing the historic castrate testosterone threshold at 1.7 nmol/l.98–101 As methodologies improved, immunoassays such as chemiluminescent immunoassays102 and radioimmunoassays103 became widely adopted due to their simplicity, ease of use, low cost and high throughput.31,95–97,99,104–109 However, these assays have limited accuracy at low testosterone levels, which limit their reliability28,95,109,110 and may result in overestimation of testosterone levels.111,112 Liquid chromatography-mass spectrometry has recently emerged as the gold standard for assaying low levels of testosterone due to its greater sensitivity and accuracy28,96,99,106,113 and immunoassays validated against LC-MS through accuracy based external quality assessment programs are also considered reliable.66,114–119 However, access to reliable assays varies across jurisdictions, with greater availability in central or academic laboratories. An exploratory analysis was undertaken to assess the relationship between assay type and breakthrough rates at the 1.7 and 0.7 nmol/l thresholds. The quality of
Referência(s)