Retrospective evaluation of neutropenic admission events in metastatic or high-risk hormone-sensitive prostate cancer (HSPC) patients having docetaxel chemotherapy upfront or for castrate-resistant prostate cancer (CRPC) in STAMPEDE

Harriet P. Mintz^1,2,3, Claire Amos³, Rachael Brannan⁴, Christopher Brawley³, Joanna Calvert³, Melissa R. Gannon⁵, Luke Hounsome⁴, Fiona C. Ingleby³, Sean McPhail⁴, Mahesh KB Parmar³, Prashant Patel^2,6, Helen Parsons¹, Mary Rauchenberger³, Matthew R. Sydes³, Haiyan Wu⁷, Nicholas D. James^2,6, the STAMPEDE trial group
¹ Warwick Medical School, University of Warwick, Coventry, England; ² University Hospitals Birmingham NHS Foundation Trust, Birmingham, England; ³ MRC Clinical Trials Unit at UCL, London, England; ⁴ Public Health England, London, England; ⁵ London School of Hygiene and Tropical Medicine, London, England;
⁶ Institute of Cancer and Genomic Sciences, University of Birmingham, England, ⁷ HDR UK, UCL, London, England

Background

• Docetaxel (Doc) was 1st licenced for CRPC (TAX-327 trial)
  Recent RCTs showed docetaxel at HSPC diagnosis ↑ survival
• But ↑ HSPC sepsis rates were also reported
• - Was this due to docetaxel timing or case-mix differences?
• No RCTs have compared sepsis rates by docetaxel timing
• We compared upfront (HSPC) & relapse (CRPC) sepsis rates in STAMPEDE trial patients

 Table 1: Acronym descriptions.

Abstract

Background

Docetaxel (Doc) was initially licenced for CRPC (TAX327) but more recently trials showed Doc at HSPC diagnosis improved survival, shifting patterns of use. Higher neutropenic toxicity rates were reported in the HSPC trials, but it is unclear if this was due to the Doc timing or differences in case-mix. We compared sepsis rates for Doc at HSPC and CRPC using routine NHS data for men randomised in STAMPEDE in England.

Methods

STAMPEDE patient data were linked to routine NHS data (Hospital Episode Statistics: HES; Systemic Anti-Cancer Therapy: SACT). Patient note review (ref) linked to NHS data assessed admission rates by HSPC & CRPC Doc at 1 site (N=44) and were used to develop and validate algorithms for detecting sepsis events across the entire data set. Algorithms were restricted to sepsis-only & sepsis + neutropenia (S+N) (N=3642). Missing HES CRPC Doc regimens were imputed (inferred) with HES (N=3642) or enhanced with SACT (N=1573). Odds ratios (OR) were calculated for risk of sepsis (OR < 1 = lower risk for HSPC).

Results

Sepsis rates varied by method; for most, rates at CRPC were higher than in TAX327 but similar to or higher than reported STAMPEDE HSPC data.

Conclusions

This analysis does not support the hypothesis that HSPC Doc has a higher sepsis risk than Doc use in CRPC. Sepsis rates found using routine data were higher than in the TAX327 trial but similar to reported “real world” CRPC data. Rates varied by data-identification method used; for most, CRPC sepsis rates were higher or similar to HSPC rates & overall a little higher than the reported STAMPEDE HSPC rate. These data suggest CRPC Doc has a similar or higher sepsis rate than use in HSPC & this should be factored into discussions for men with newly-diagnosed metastatic HSPC and supports Doc use in this setting.

Methods

Events of interest & data sources

Sepsis defined as:

Severe events: admission with infection e.g. febrile neutropenia, neutropenic sepsis, infection with neutropenia, neutropenia grade 3+, pyrexia, pneumonia and infection

Table 2: Events of interest identified during the analyses. * = In the absence of STAMPEDE data, routine data was used to identify HSPC and CRPC sepsis admissions & CRPC chemo regimens.

Patients analysed

STAMPEDE patients in England
4 cohorts (N=44, 113, 1573, 3642) (1-2: single-site, 3-4: multi-site)

Figure 1: The STAMPEDE trial (treatment arms A-G). Figure adapted from the STAMPEDE protocol, for further details see the protocol [3].

Results

Routine Data Accuracy

Chemotherapy use

Figure 2: Accuracy of detecting HSPC and CRPC chemo regimens compared to the clinical noting.

Sepsis admission events (all events)

Figure 3: Accuracy of detecting sepsis admissions. A) vs. STAMPEDE; B) vs. clinical noting.

• Routine data did not accurately identify upfront trial chemo but did identify relapse chemo
• SACT accuracy improved over time (after April 2012)
• Models were successfully created to account for missing data, improving detection of relapse chemo and sepsis events
• Routine data identified 92% of case report form reported STAMPEDE upfront chemo sepsis events but appeared to identify more (trial under-reporting?)

Rates

Table 3: Rates of sepsis events identified. NR = not reported; % difference = HSPC – CRPC rate.

1. Sepsis-only rate: HSPC 2% less than CRPC (but regimens missed)
2. All events: HSPC 3% more than CRPC (but regimens missed)
3. All events + missing events inferral: HSPC 20% less than CRPC
4. All event rate + SACT: similar

Figure 4: Forest plot graphically displaying table 3. Cohort 3, N=1573; cohort 4, N=3642.

Conclusion

Routine data limitations mean each individual variable should be validated for accuracy & completeness prior to use

Rates varied by method (different rates identified using different models) but for most, relapse sepsis rates were ↑ or similar to upfront rates

The upfront routine data sepsis rates were overall ↑ than case report form upfront rates (under-reporting?)

Routine data relapse sepsis rates were ↑ than TAX-327 but similar to ‘real world’ relapse data [4]

Overall, analysis does not support a ↑ upfront (HSPC) docetaxel sepsis risk vs. the CRPC setting

This evidence supports upfront HSPC docetaxel use. This should factor into discussions for men with newly-diagnosed mHSPC

Contact

Hattie Mintz

@HattieMintz on Twitter

warwick.ac.uk/hattiemintz

H.P.Mintz@warwick.ac.uk

COI, SPONSOR & REFS

Harriet Mintz has no COI to declare & the study was funded by Warwick Medical School

[1] Hospital Episode Statistics - NHS Digital [ONLINE] Available: http://content.digital.nhs.uk/hes [Accessed: 04 Sept. 2019]

[2] Systemic Anti-Cancer Therapy dataset - National Cancer Registration and Analysis Service [ONLINE] Available: http://www.ncin.org.uk/collecting_and_using_data/data_collection/chemotherapy [Accessed: 04 Sept. 2019]

[3] Systemic therapy in Advancing or Metastatic Prostate Cancer: Evaluation of Drug Efficacy, A multi-arm multi-stage randomised controlled trial. The protocol [ONLINE] Available: http://www.stampedetrial.org

[4] Templeton A, Vera-Badillo F, Wang L, Attalla M, De Gouveia P, Leibowitz-Amit R, et al. Translating clinical trials to clinical practice: outcomes of men with metastatic castration resistant prostate cancer treated with docetaxel and prednisone in and out of clinical trials. Annals of oncology. 2013;24(12):2972-7.

N.B: 3 further pts may be eligible for inclusion pending review of consent