Abstract
Background:
The immunosuppressive cytokine IL-8, produced by tumor cells and some myeloid cells, promotes inflammation, angiogenesis, and metastasis. In our discovery work, elevated serum IL-8 at androgen deprivation therapy (ADT) initiation portended worse overall survival (OS). Leveraging serum samples from the phase 3 CHAARTED trial of patients treated with ADT +/− docetaxel for metastatic hormone-sensitive prostate cancer (mHSPC), we validated these findings.
Methods:
233 patients had serum samples drawn ≤28 days from ADT initiation. The samples were assayed using the same Mesoscale multiplex ELISA platform employed in the discovery cohort. After adjusting for performance status, disease volume, and de novo/metachronous metastases, multivariable Cox proportional hazards models assessed associations between IL-8 as continuous and binary variables on OS and time to castration-resistant prostate cancer (CRPC). The median IL-8 level (9.3 pg/mL) was the a priori binary cutpoint. Fixed-effects meta-analyses of the discovery and validation sets were performed.
Results:
Higher IL-8 levels were prognostic for shorter OS (continuous: HR 2.2, 95% CI: 1.4–3.6, p=0.001; binary >9.3: HR 1.7, 95%CI: 1.2–2.4, p=0.007) and time to CRPC (continuous: HR 2.3, 95%CI: 1.6–3.3, p≤0.001; binary: HR 1.8, 95%CI: 1.3–2.5, p<0.001) and independent of docetaxel use, disease burden, and time of metastases. Meta-analysis including the discovery cohort, also showed that binary IL-8 levels >9.3pg/mL from patients treated with ADT alone was prognostic for poorer OS (HR 1.8, 95% CI: 1.2–2.7, p=0.007) and shorter time to CRPC (HR: 1.4, 95%CI: 0.99–1.9, p=0.057).
Conclusions:
In the phase 3 CHAARTED study of men with mHSPC at ADT initiation, elevated IL-8 portended worse survival and shorter time to castration-resistant prostate cancer independent of docetaxel administration, metastatic burden, and metachronous versus de novo metastatic presentation. These findings support targeting IL-8 as a strategy to improve mHSPC outcomes.
Keywords: hormone sensitive prostate cancer, IL-8, cytokines, ADT, TNF-a, MCP
Introduction
Chronic inflammation, whether innate or adaptive due to infection, has been implicated in the development of prostate cancer.1–4 It is likely that various inflammatory pathways contribute to the transition to the androgen insensitive state known as castration-resistant prostate cancer (CRPC), the lethal form of prostate cancer. A lead candidate in this process is the nuclear factor-κB (NF-κB) pathway, which induces a complex immune response including upregulation of pro-survival anti-apoptotic pathways and pro-inflammatory cytokines that recruit immune cells to areas of inflammation [e.g. tumor necrosis factor-α (TNF-α), interleukin-8 (IL-8), IL-1, and IL-6].5
Activation of NF-κB depends on the REL family members. Nuclear RelB can be elevated in high grade prostate cancers and in aggressive PC-3 androgen insensitive cells lines as compared to androgen-sensitive lines like LNCaP.6,7 Inhibiting RelB in aggressive androgen-resistant PC-3 cells significantly decreased tumor growth and incidence and corresponded with decreased IL-8 levels, while stabilizing RelB expression in androgen-sensitive LNCaP tumors increased IL-8 levels.7 Blocking NF-κB in aggressive PC-3 models decreased expression of IL-8, vascular endothelial growth factor (VEGF) and matrix metalloproteinase-9 (MMP-9) suggesting a potential strategy to inhibit angiogenesis, invasion and metastasis.8
IL-8 (CXCL-8) is a pro-inflammatory cytokine produced by a number of cell types including tumor cells, macrophages and stromal cells. Secretion is increased by many factors including NF-κB, TNF-α, IL-1β, environmental stressors, hypoxia, chemotherapy agents, steroids (e.g., dexamethasone), androgens, and estrogens.9 Two cell surface receptors, CXCR1 and CXCR2, mediate IL-8’s wide-ranging functions including induction of HIF-1, NF-κB, AP-1, AR, STAT-3, and B-catenin signaling pathways. The result is a broad footprint across a spectrum of pro-cancer pathways including immune surveillance, angiogenesis, inflammation, cell cycle interruption, migration, invasion, and anti-apoptosis.9–11 A variety of studies employing cell lines and human specimens have characterized IL-8 as low or undetectable in androgen-sensitive cell lines (e.g., LNCaP, LAPC-4) and localized, non-metastatic specimens (e.g., hormone sensitive prostate cancer (HSPC) and higher in more advanced stages or cell lines not responsive to androgens (e.g., CRPC; DU-145 and PC-3 cell lines).10–14
Prior data generated by our group revealed that higher levels of the inflammatory cytokines IL-8, TNF-α, and monocyte chemoattractant protein-1 (MCP-1) at ADT initiation for metastatic HSPC were associated with poorer prostate cancer outcomes.15 Specifically, serum samples of 122 men from two institutional biorepositories were interrogated. With a median follow-up of greater than 3 years and accounting for ECOG performance status (PS), age, PSA at ADT initiation and extent of metastases, men with elevated IL-8 and TNF-α had decreased overall survival (OS) with a trend to lower OS with elevated MCP-1 (IL-8: HR 1.9, 95% CI: 1.0,3.5, p=0.04; TNF-α: HR 2.0, 95% CI: 1.1, 3.5, p=0.02; MCP-1: HR 1.7 95% CI: 1.7, 3.0 p=0.08). A similar but not statistically significant trend to shorter time to CRPC was observed with elevated protein levels (IL-8: HR 1.4, 95% CI: 0.9, 2.2, p=0.13; TNF-α: HR 1.3, 95% CI: 0.8,2, p=0.18; MCP-1: 1.0, 95% CI: 0.7,1.6, p=0.95). No associations were found with IL-6, IL-2 or IL-1β. We postulated that an active, pro-cancer inflammatory infiltrate present at ADT initiation may thwart the efficacy of initial ADT explaining the poorer OS.
In this study, we leveraged the existing serum samples and prospectively annotated clinical outcomes from the Phase 3 E3805 CHAARTED study of 790 patients who had received ADT +/− docetaxel for hormone-sensitive disease16 to test whether higher levels of the pro-inflammatory markers IL-8, TNF-α and MCP-1 were associated with shorter OS and time to CRPC. Given the robust preliminary data from the primary analysis and the published biology, our primary analysis [as documented in grant application (R01 CA208254–01)] focused on IL-8 to avoid false discovery by multiple testing.
Patients and Methods
Study Design and Participants
Patients from the Phase 3 CHAARTED study (NCT00309985) formed the foundation of the study group. Patients samples and data were included if they had available serum samples drawn within 28 days of starting ADT +/− docetaxel. The discovery cohort, derived from biorepositories at Dana-Farber Cancer Institute and Hoosier Oncology Group,15 was incorporated in meta-analyses of ADT monotherapy patients. REMARK guidelines were observed.17 The study was conducted in accordance with the Declaration of Helsinki for human subject protection.
Procedures
Baseline clinicodemographic data at ADT initiation was obtained such as age, ECOG PS, PSA, extent of metastases (high versus low based on CHAARTED criteria), and de novo versus metachronous disease presentation (based on whether or not patients had prior local therapy, e.g., radical prostatectomy, radiation). Outcomes data was captured including follow-up time, time to CRPC, and OS. The Fichorova Laboratory at Brigham and Women’s Hospital assessed all serum cytokine levels using the same Mesoscale Discovery Multiplex ELISA assay employed in the discovery cohort and is detailed previously.15
Study assessments
The primary analysis focused on the lead candidate of IL-8 with secondary analyses on TNF-α and MCP-1. We hypothesized that baseline IL-8 above the median in men commencing ADT for mHSPC conferred shorter OS and time to CRPC. Continuous and binary evaluations of levels were performed. The median protein levels in the discovery set served as the a priori cutoff points for each protein (e.g., 9.3 pg/mL for IL-8). The key secondary analysis was to understand the impact of docetaxel, which was not administered to patients in the discovery cohort, and IL-8 levels on OS and time to CRPC. OS was defined as time from randomization to death from any cause. Time to CRPC was defined from randomization until progression by PSA, imaging or symptoms.
Statistical analysis
The Kaplan-Meier method estimated OS and time to CRPC. Hazard ratios were estimated using Cox proportional hazards models. Log-rank tests compared time to event distributions (e.g., OS and time to CRPC) between two groups. Effects of IL-8 as a binary or continuous variable on OS and time to CRPC were assessed in multivariable Cox proportional models that adjusted for baseline clinical variables robustly associated with poor clinical outcome on multivariable analysis in the overall study population and included ECOG PS, disease volume (high versus low) and type of metastatic presentation (de novo versus metachronous). We tested for interaction between treatment and baseline IL-8 levels using a Cox proportional hazards model that included treatment arm (ADT alone vs. ADT+docetaxel), baseline IL-8 level (above/below median) and the interaction between the two as covariates. Fisher’s exact test compared categorical variables between two groups while Wilcoxon test compared continuous variables. Fixed-effects meta-analysis of the discovery and validation sets was performed among the ADT monotherapy patients to increase power.
Results
From CHAARTED, clinical data and samples from 233 patients were included in the analysis (Figure 1). Patient clinico-demographics were balanced between the high and low IL-8 subsets with the exception of performance status for which the low IL-8 subset had a higher percentage of 0–1 (Table 1).
Figure 1:
REMARK Diagram of Patient and Sample Inclusion
Table 1:
Clincodemographic characteristics of patients in the CHAARTED cohort
| Variable | Category | IL-8 High (>9.3 pg/ml) | IL-8 Low (≤9.3 pg/mL) | Total | P-value |
|---|---|---|---|---|---|
| Total | 140 | 93 | 233 | – | |
| Freq. of Missing | 0 | 0 | 0 | ||
| Unknown/Missing | 5 | 3 | 8 | ||
| Unknown/Missing | 0 | 0 | 0 | ||
| Freq. of Missing | 0 | 0 | 0 | ||
| Unknown/Missing | 0 | 0 | 0 | ||
| Freq. of Missing | 0 | 0 | 0 | ||
| Freq. of Missing | 0 | 0 | 0 | ||
| Freq. of Missing | 0 | 0 | 0 | ||
| Freq. of Missing | 55 | 29 | 84 | ||
| Unknown/Missing | 0 | 0 | 0 | ||
| Unknown/Missing | 0 | 0 | 0 | ||
| Follow-up (months) | Median (Q1, Q3) | 51.7 (37.4, 59.7) | 49.1 (41.8, 71.7) | 49.6 (40.4, 60.5) | 0.310 |
| OS (months) | Median (Q1, Q3) | 44 (23.5, 44) | 72.2 (32.4, 72.2) | 49.5 (27.7, 49.5) | 0.009 |
| CRPC (months) | Median (Q1, Q3) | 15.7 (8, 15.7) | 20.7 (11.8, 20.7) | 17.9 (9, 17.9) | 0.001 |
Confirmed complete PSA response per CHAARTED protocol.
Overall Survival in the CHAARTED cohort
As both continuous and binary variables, higher IL-8 was prognostic for shorter OS independent of docetaxel administration or disease volume. On multivariable analysis using a binary cutoff of IL-8 >9.3 pg/mL adjusted for ECOG PS, disease volume and prior local therapy as a surrogate for de novo versus metachronous disease presentation, high IL-8 was prognostic for significantly shorter OS with an HR 1.7 (95%CI: 1.2–2.4, p=0.007). The corresponding HRs for the ADT and ADT+docetaxel cohorts were 1.7 (95%CI: 0.94–3.2, p=0.08) and 1.6 (95%CI 0.95–2.6, p =0.08) respectively (Table 2, Figure 2). No significant interaction between IL-8 level and treatment arm (HR 0.88, 95%CI: 0.41–1.9, p=0.74; Table S9) was observed.
Table 2:
Multivariable analyses* of IL-8 for OS and CRPC
| Continuous IL-8 (log10) | Binary IL-8 ≤ or >9.3 | |||||||
|---|---|---|---|---|---|---|---|---|
| n (events) | HR | 95%CI | p | n (events) | HR | 95%CI | p | |
| OS | ||||||||
| All | 233 (129) | 2.2 | 1.4–3.6 | 0.001 | 233 (129) | 1.7 | 1.2–2.4 | 0.007 |
| ADT alone | 116 (65) | 2.9 | 1.2–7.0 | 0.02 | 116 (65) | 1.7 | 0.94–3.2 | 0.08 |
| ADT +Docetaxel | 117 (64) | 2.1 | 1.2–3.8 | 0.014 | 117 (64) | 1.6 | 0.95–2.6 | 0.08 |
| CRPC | ||||||||
| All | 233 (174) | 2.3 | 1.6–3.3 | <0.001 | 233 (174) | 1.8 | 1.3–2.5 | <0.001 |
| ADT alone | 116 (96) | 3.0 | 1.5–6.1 | 0.003 | 116 (96) | 1.5 | 0.95–2.5 | 0.08 |
| ADT +Docetaxel | 117 (78) | 2.0 | 1.2–3.4 | 0.007 | 117 (78) | 1.8 | 1.2–2.9 | 0.009 |
Controlled for ECOG performance status, disease volume, and prior local therapy (radiation/prostatectomy vs. none).
Figure 2. Kaplan-Meier Curves of IL-8 Levels and Outcomes in the E3805 Cohort.
A. Time to CRPC by high/low IL-8 in the ADT monotherapy patients
B. Time to CRPC by high/low IL-8 in the ADT +docetaxel patients
C. Time to CRPC by treatment arm in patients with baseline low IL-8
D. Time to CRPC by treatment arm in patients with baseline high IL-8
E. Overall survival by high/low IL-8 in the ADT monotherapy patients
F. Overall survival by high/low IL-8 in the ADT +docetaxel patients
G. Overall survival by treatment arm in patients with baseline low IL-8
H. Overall survival by treatment arm in patients with baseline high IL-8
The corresponding multivariable analysis for overall survival using IL-8 as a continuous variable produced an HR of 2.2 (95% CI: 1.4–3.6, p=0.001) with worse prognosis for higher levels. Among the ADT and ADT+docetaxel cohorts, the corresponding HRs were 2.9 (95% CI:1.2–7.0, p=0.02) and 2.1 (95%CI 1.2–3.8, p=0.014) respectively (Table 2). There was no significant interaction between IL-8 level and treatment arm (HR 1.1, 95%CI 0.41–2.64, p=0.93, Table S10).
No significant association between OS and MCP-1 and TNF-α levels were seen (Online supplementary (S) data: Table S2–3, S5–6, Figures S1–S2). Time to OS for continuous serum markers based on high versus low disease volume was also performed with no significant difference (Table S7).
Time to CRPC in the CHAARTED cohort:
Higher IL-8 levels, as both continuous and binary variables, were prognostic for shorter time to CRPC. On multivariable analysis using a binary IL-8 cutoff >9.3 pg/mL adjusted for ECOG PS, disease volume and prior local therapy, the HR for time to CRPC by IL-8 protein level >9.3 pg/mL was highly significant at 1.8 (95%CI: 1.3–2.5, p<0.001) (Table 2). By treatment arm, the HR in the ADT cohort was 1.5 (95%CI 0.94–2.5, p=0.08) and 1.8 (95%CI 1.2–2.9, p =0.009) in the ADT+docetaxel cohort (Table 2, Figure 2). There was no significant interaction between IL-8 level and treatment arm (HR 1.3, 95%CI: 0.67–2.38, p=0.46, Table S9).
On multivariable analysis using continuous marker values (log10 scale) and adjusted for ECOG PS, disease volume and local therapy, the HR for time to CRPC was significant at 2.3 (95%CI: 1.6–3.3, p≤0.001). By treatment arm, the HR was 3.0 in ADT monotherapy patients (95%CI 1.5–6.1, p=0.003) and 2.0 (95%CI 1.2–3.4, p =0.007) in ADT+docetaxel cohort (Table 2). No significant interaction between IL-8 level and treatment arm (HR 1.1, 95%CI: 0.5–2.4, p=0.83, Table S10) was observed.
No significant associations between time to CRPC and TNF-α or MCP-1 levels were found (Table S2, S4–6, Figures S1–2). Analyses of time to CRPC for continuous serum markers based on high versus low disease volume status were underpowered to detect an interaction if present (Table S8).
Comparisons and Meta-analyses of the Discovery15 and CHAARTED validation cohorts
The median IL-8 levels of the CHAARTED patients were in a similar range at 10 pg/dL, 8.6pg/mL, and 11.2 pg/mL across all patients, the ADT monotherapy and ADT+docetaxel cohorts respectively. When comparing the binary data of the ADT monotherapy patients in the discovery and validation cohorts (Table 3), both cohorts trend in the same direction towards worse OS with higher IL-8. Upon Fixed-effects meta-analysis of the binary protein levels using the median levels defined from discovery set of all three cytokines, higher protein levels portended worse OS (Table 4). Specifically, the HR for protein level above the median was 1.8 (95% CI: 1.2–2.7, p=0.007) for IL-8; 1.5 (95% CI: 1.0–2.2, p=0.06) for TNF-α; and 1.6 (95%CI: 1.1–2.3, p=0.02) for MCP-1. Upon meta-analysis using binary thresholds, there was a trend towards longer time to CRPC (HR: 1.4, 95%CI: 0.99–1.9, p=0.057) for IL-8 but not for TNF-α or MCP-1 (Table 4).
Table 3:
Univariable analysis of binary marker levels (above/below median) and outcomes in E3805 and the Sharma et al cohorts.
| E3805ADT HR | Sharma HR | E3805ADT p value | Sharma p value | |
|---|---|---|---|---|
| Overall Survival | ||||
| IL-8 | 1.680 | 1.900 | 0.078 | 0.040 |
| TNF-α | 1.140 | 2.000 | 0.620 | 0.020 |
| MCP1 | 1.480 | 1.700 | 0.120 | 0.080 |
| CRPC | ||||
| IL-8 | 1.310 | 1.400 | 0.240 | 0.130 |
| TNF-α | 0.780 | 1.300 | 0.270 | 0.290 |
| MCP1 | 1.200 | 1.014 | 0.370 | 0.950 |
Table 4:
Fixed-effects meta-analysis for univariable analysis of binary markers (above/below median) and outcomes in E3805 ADT monotherapy arm and Sharma et al cohorts.
| HR | 95% CI lower | p | |
|---|---|---|---|
| Overall Survival | |||
| IL-8 | 1.780 | 1.170–2.709 | 0.007 |
| TNF-α | 1.460 | 0.991–2.151 | 0.056 |
| MCP1 | 1.566 | 1.071–2.290 | 0.021 |
| CRPC | |||
| IL-8 | 1.356 | 0.991–1.854 | 0.057 |
| TNF-α | 0.983 | 0.709–1.362 | 0.916 |
| MCP1 | 1.111 | 0.828–1.490 | 0.482 |
Discussion
Leveraging the well-annotated, prospective clinical outcomes data from the phase 3 CHAARTED study, we validated the prior findings of our discovery cohort that elevated levels of IL-8 are prognostic for worse prostate cancer outcomes. The findings were most consistent when IL-8 levels were employed as a continuous factor rather than binary levels based on the discovery set’s median protein level. Recognizing a binary level is more viable for implementation as a biomarker in clinical trial or routine use, we combined the binary data from both the discovery and validation subsets of ADT monotherapy patients and found IL-8 levels >9.3pg/mL conferred worse OS (HR 1.8, 95% CI: 1.2–2.7, p=0.007) and shorter time to CRPC (HR: 1.4, 95%CI: 0.99–1.9, p=0.057).
The role of immunotherapy in prostate cancer is evolving. While the first anticancer vaccine, Sipuleucel-T, an autologous dendritic cell-based immunotherapy,18 was approved for CRPC, the broad success of immune checkpoint blockade across multiple solid tumors has not been realized in prostate cancer, except among a subset of patients.19,20 However the complicated interplay between the immune system and cancer likely far surpasses the PD-1 and CTLA-4 immune checkpoints. Prime among alternative targets are inhibitory or pro-inflammatory chemokines and cytokines prevalent in prostate cancer, such as IL-8, which can be manipulated to induce a more hospitable immune milieu.
IL-8 is differentially expressed in early androgen sensitive disease and more advanced castration-resistant cancers.10–14 Aalinkeel et al observed that IL-8 was significantly elevated in prostate cancer cell lines prone to metastasize, PC-3 and DU-145, compared to LNCaP, which has less invasive capacity.10 Lehrer et al found significant elevations of circulating IL-8 in men with bone metastases compared to those with localized disease (n=39, p=0.007).13 Mechanistically, IL-8 can increase androgen-independent cell proliferation in both androgen responsive and insensitive cells lines.14,21 Signaling through IL-8 can increase AR expression, potentiate receptor activity and upregulate AR-responsive genes such as PSA and Cdk2 providing support for its role in the transition to the ligand-independent CRPC state.21 Conversely, IL-8 inhibition enhanced AR blockade with bicalutamide in androgen sensitive LNCaP cells.21 To investigate whether elevated IL-8 was a “cause or consequence” of androgen independence and test their hypothesis that IL-8, whether constitutively produced or stimulated by inflammatory agents, triggers downregulation of the androgen-mediated cell proliferation pathways and induces androgen independence, Araki et al continuously exposed androgen responsive cell lines (e.g. LNCaP and LAPC-4 cells) to IL-8.11 They observed AR downregulation, decreased dependence on androgens for proliferation, increased survival factors such as AKT, SRC, and NF-κB, and increased resistance to antiandrogens and docetaxel. Further, these cells had significantly increased motility, invasion, and production of VEGF and matrix metalloproteinases, the latter being integral in invasion through the extracellular matrix. IL-8 secreting cells with higher levels of pro-survival factors SRC and NF-κB were more resistant to docetaxel, which could be reversed by SRC and NF-κB inhibitors.
Complementary work by Singh et al illustrated that depletion of endogenous IL-8 using small interfering RNA (siRNA) in androgen insensitive cell lines PC3 and DU-145, which constitutively secrete IL-8, diminished NF-κB activity, induced cell cycle arrest, decreased motility, increased apoptosis and enhanced the cytotoxicity of docetaxel.22 The CHAARTED samples provided a rare opportunity to assess differences in docetaxel sensitivity based on IL-8 levels. The docetaxel cohort had higher median IL-8 levels than the ADT monotherapy patients (11.2 vs. 8.6 pg/mL, p<0.001) (Table S1). Analysis allowing treatment-marker interaction did not show differences in treatment effect across marker levels.23 However, we cannot exclude that subsequent docetaxel for CRPC in the ADT monotherapy group may have confounded OS differences, as it was not captured in CHAARTED. Docetaxel administration was also most effective in CHAARTED patients with high volume disease.24 The sample sizes were too small to draw meaningful conclusion from an interaction term analysis based on disease volume in this study.
In humans, Hawley et al evaluated cytokines changes in men receiving ADT for biochemically recurrent prostate cancer finding.25 Significantly elevated levels of innate cytokines IL-8 and IL-6 at PSA progression were observed in patients who did not achieve a complete PSA response to 6 months of chemical castration. IL-8 levels strongly correlated with immunosuppressive TNF-α, higher levels of which were associated with shortened time to PSA progression. It will be important to evaluate the impact of IL-8 levels in the more recent studies investigating the addition of the latest generation of androgen antagonists (e.g., enzalutamide, apalutamide, darolutamide) and biosynthesis inhibitors (e.g., abiraterone acetate). We plan to assess the reproducibility of these findings in the recently reported global phase 3 ENZAMET study, which revealed a significant improvement in time to CRPC and OS with the addition of enzalutamide to ADT in mHSPC.26 A significant number of patients (53%) received early docetaxel in this study, but its concurrent use did not appear to improve survival over enzalutamide alone.
Ongoing therapeutic strategies are targeting IL-8. HuMax IL-8 (BMS-986253), a fully human IgG1 anti-IL-8 monoclonal antibody, binds directly to IL-8 preventing binding to its receptors CXCR1 and CXCR2. In a phase 1 study of advanced solid tumors, HuMax was well tolerated with no dose-limiting toxicities, and all dose levels induced significant reductions in serum IL-8 by day 3 (p=0.0004).27 The two patients with mCRPC experienced disease stabilization ranging from 2–8 months and reductions in IL-8 levels. Combination approaches are also rational given data that response to anti-PD-1 may correlate with decreased IL-8 levels and that rising levels may herald progression.28 Drake and colleagues are evaluating a combinatorial strategy of depleting IL-8 and blocking PD-1 in the ongoing MAGIC-8 trial (NCT03689699). Employing a phase 1b/2 design, the study will assess the rate of PSA progression at 10 months after initiation of ADT plus nivolumab +/− HuMax (BMS-986253) in men with non-metastatic HSPC, who are experiencing biochemical recurrence. Other innovative strategies to silence IL-8 are in development such as nanotherapy approaches, which would dispense siRNA to interfere with IL-8, and may also incur lower toxicity.29
Limitations of this study include those inherent to retrospective analyses of a prospective clinical trial cohort. While dividing a subset of randomized patients into separate cohorts retrospectively could lead to imbalance in baseline prognostic factors, the IL-8 high/low CHAARTED patients appeared similar except for better performance status in the low IL-8 cohort (Table 1). However, performance status was accounted for in the multivariate analysis, and the prognostic impact of IL-8 remained independent. While only 25% of patients had serum available, their survival outcomes were representative of the whole CHAARTED population.24 To limit discrepancies in protein levels due to biologic variation over time and ADT administration, we limited samples to ≤28 days from ADT initiation and employed the same methodology and laboratory for sample analysis as in the training cohort. Similarly, at the discovery stage, identifying the optimal cutpoint of a potential biomarker is challenging. While we decided a priori to employ the median level from the training set, we also investigated the levels as a continuous factor to minimize missing a signal based on an erroneous assumption. Median levels varied minimally between the pilot and validation sets (9.3pg/mL for discovery group, 8.6pg/mL in the CHAARTED ADT monotherapy, cohort and 10pg/mL in all CHAARTED patients). Strengths of the validation cohort include samples collected at fixed time points and prospectively collected outcomes.
In conclusion, interrogation of a prospective phase 3 study of serum cytokines taken from patients with mHSPC at the time of ADT initiation demonstrates that baseline higher levels of IL-8 are prognostic for worse overall survival and shorter time to CPRC independent of docetaxel use, metastatic burden, or metachronous or de novo metastatic disease presentation. This work corroborated the findings of our earlier study comprised of two institutional registries. These collective efforts converge on IL-8, whose diverse spectrum of biologic functions, may adversely impact prostate cancer outcomes. Continued investigation of the measurement of serum IL-8, an assay that could be easily developed for routine use, as a prognostic biomarker in mHSPC and/or for selection of patients for IL-8 inhibition is warranted.
Supplementary Material
Acknowledgements:
We appreciate the patients and their families who contributed to this study as well as the many EA3805 co-investigators.
Funding & Research Support: This study was coordinated by the ECOG-ACRIN Cancer Research Group (Peter J. O’Dwyer, MD and Mitchell D. Schnall, MD, PhD, Group Co-Chairs) and supported by the National Cancer Institute of the National Institutes of Health (NIH) under the following award numbers: U10CA180820, U10CA180794, UG1CA233180, UG1CA233196, R01CA208254-01 and Sanofi. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH, nor does mention of trade names, commercial products, or organizations imply endorsement by the U.S. government.
Footnotes
Ethics approval and consent to participate:
Patients from the Phase 3 CHAARTED study (NCT00309985) formed the foundation of the study group and were consented to the phase 3 study for treatment and gave permission for sample acquisition. The study was conducted in accordance with the Declaration of Helsinki for human subject protection.
Consent for publication: All authors, ECOG-ACRIN, and Sanofi provide consent for publication.
Competing Interests Disclosures:
XVW, GS, RNF, RSD report no COI.
LCH reports consulting fees from Genentech, Dendreon, Pfizer, Medivation/Astellas, Exelixis, Bayer, Kew Group, Corvus, Merck, Novartis, Michael J Hennessy Associates (Healthcare Communications Company and several brands such as OncLive and PER), Jounce, EMD Serrano, and Ology Medical Education; Research funding from Bayer, Sotio, Bristol-Myers Squib, Merck, Takeda, Dendreon/Valient, Jannsen, Medivation/Astellas, Genentech, Pfizer, Endocyte (Novartis), and support for research travel from Bayer and Genentech. Currently employed by Surface Oncology.
AAH reports consulting fees from Merck Sharp & Dohme; honoraria from Bayer.
CGD reports this role as a co-inventor on patents licensed from JHU to BMS and Janssen, has served as a paid consultant to AZ Medimmune, BMS, Pfizer, Roche, Sanofi Aventis, Genentech, Merck, and Janssen, and has received sponsored research funding to his institution from the Bristol-Myers Squibb International Immuno-Oncology Network.
MAC reports consulting fees from Pfizer, Roche-Genentech, Astellas; Research funding: EMD Serrano, Pfizer, Merck, Bristol-Myers Squibb
CJS reports consulting or Advisory Role: Sanofi, Janssen, Astellas Pharma, Bayer, Genentech, AstraZeneca, Pfizer, Celgene Research Funding: Janssen Biotech (Inst), Astellas Pharma (Inst), Sanofi (Inst), Bayer (Inst), Sotio (Inst), Dendreon (Inst); Patents, Royalties, Other Intellectual Property: Pathenolide (Indiana University): dimethylaminoparthenolide (Leuchemix); Exelixis: Abiraterone plus cabozantinib combination. Stock or Other Ownership: Leuchemix.
References:
- 1.Bhavsar NA, Bream JH, Meeker AK, et al. A peripheral circulating TH1 cytokine profile is inversely associated with prostate cancer risk in CLUE II. Cancer epidemiology, biomarkers & prevention : a publication of the American Association for Cancer Research, cosponsored by the American Society of Preventive Oncology 2014;23:2561–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.De Marzo AM, Platz EA, Sutcliffe S, et al. Inflammation in prostate carcinogenesis. Nature reviews Cancer 2007;7:256–69. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Kwon OJ, Zhang L, Ittmann MM, Xin L. Prostatic inflammation enhances basal-to-luminal differentiation and accelerates initiation of prostate cancer with a basal cell origin. Proceedings of the National Academy of Sciences of the United States of America 2014;111:E592–600. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Putzi MJ, De Marzo AM. Morphologic transitions between proliferative inflammatory atrophy and high-grade prostatic intraepithelial neoplasia. Urology 2000;56:828–32. [DOI] [PubMed] [Google Scholar]
- 5.Hoesel B, Schmid JA. The complexity of NF-kappaB signaling in inflammation and cancer. Mol Cancer 2013;12:86. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Lessard L, Begin LR, Gleave ME, Mes-Masson AM, Saad F. Nuclear localisation of nuclear factor-kappaB transcription factors in prostate cancer: an immunohistochemical study. British journal of cancer 2005;93:1019–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Xu Y, Josson S, Fang F, et al. RelB enhances prostate cancer growth: implications for the role of the nuclear factor-kappaB alternative pathway in tumorigenicity. Cancer research 2009;69:3267–71. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Huang S, Pettaway CA, Uehara H, Bucana CD, Fidler IJ. Blockade of NF-kappaB activity in human prostate cancer cells is associated with suppression of angiogenesis, invasion, and metastasis. Oncogene 2001;20:4188–97. [DOI] [PubMed] [Google Scholar]
- 9.Waugh DJ, Wilson C. The interleukin-8 pathway in cancer. Clinical cancer research : an official journal of the American Association for Cancer Research 2008;14:6735–41. [DOI] [PubMed] [Google Scholar]
- 10.Aalinkeel R, Nair MP, Sufrin G, et al. Gene expression of angiogenic factors correlates with metastatic potential of prostate cancer cells. Cancer research 2004;64:5311–21. [DOI] [PubMed] [Google Scholar]
- 11.Araki S, Omori Y, Lyn D, et al. Interleukin-8 is a molecular determinant of androgen independence and progression in prostate cancer. Cancer research 2007;67:6854–62. [DOI] [PubMed] [Google Scholar]
- 12.Inoue K, Slaton JW, Eve BY, et al. Interleukin 8 expression regulates tumorigenicity and metastases in androgen-independent prostate cancer. Clinical cancer research : an official journal of the American Association for Cancer Research 2000;6:2104–19. [PubMed] [Google Scholar]
- 13.Lehrer S, Diamond EJ, Mamkine B, Stone NN, Stock RG. Serum interleukin-8 is elevated in men with prostate cancer and bone metastases. Technol Cancer Res Treat 2004;3:411. [DOI] [PubMed] [Google Scholar]
- 14.Murphy C, McGurk M, Pettigrew J, et al. Nonapical and cytoplasmic expression of interleukin-8, CXCR1, and CXCR2 correlates with cell proliferation and microvessel density in prostate cancer. Clinical cancer research : an official journal of the American Association for Cancer Research 2005;11:4117–27. [DOI] [PubMed] [Google Scholar]
- 15.Sharma J, Gray KP, Harshman LC, et al. Elevated IL-8, TNF-alpha, and MCP-1 in men with metastatic prostate cancer starting androgen-deprivation therapy (ADT) are associated with shorter time to castration-resistance and overall survival. The Prostate 2014;74:820–8. [DOI] [PubMed] [Google Scholar]
- 16.Sweeney CJ, Chen YH, Carducci M, et al. Chemohormonal Therapy in Metastatic Hormone-Sensitive Prostate Cancer. The New England journal of medicine 2015;373:737–46. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.McShane LM, Altman DG, Sauerbrei W, et al. Reporting recommendations for tumor marker prognostic studies (REMARK). J Natl Cancer Inst 2005;97:1180–4. [DOI] [PubMed] [Google Scholar]
- 18.Kantoff PW, Higano CS, Shore ND, et al. Sipuleucel-T immunotherapy for castration-resistant prostate cancer. The New England journal of medicine 2010;363:411–22. [DOI] [PubMed] [Google Scholar]
- 19.Hansen AR, Massard C, Ott PA, et al. Pembrolizumab for advanced prostate adenocarcinoma: findings of the KEYNOTE-028 study. Annals of oncology : official journal of the European Society for Medical Oncology / ESMO 2018;29:1807–13. [DOI] [PubMed] [Google Scholar]
- 20.Kim JW SR, Massard C, Powles T, Harshman LC, Braiteh FS, Conkling PR, Sarkar I, Kadel EE, Mariathasan S, O’Hear C, Schiff C, Fasso M, Carroll S, Petrylak DP. A phase Ia study of safety and clinical activity of atezolizumab (atezo) in patients (pts) with metastatic castration-resistant prostate cancer (mCRPC). Journal of Clinical Oncology 36, no 6_suppl (February 20, 2018) 187–187. [Google Scholar]
- 21.Seaton A, Scullin P, Maxwell PJ, et al. Interleukin-8 signaling promotes androgen-independent proliferation of prostate cancer cells via induction of androgen receptor expression and activation. Carcinogenesis 2008;29:1148–56. [DOI] [PubMed] [Google Scholar]
- 22.Singh RK, Lokeshwar BL. Depletion of intrinsic expression of Interleukin-8 in prostate cancer cells causes cell cycle arrest, spontaneous apoptosis and increases the efficacy of chemotherapeutic drugs. Mol Cancer 2009;8:57. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Polley MY, Freidlin B, Korn EL, Conley BA, Abrams JS, McShane LM. Statistical and practical considerations for clinical evaluation of predictive biomarkers. J Natl Cancer Inst 2013;105:1677–83. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Kyriakopoulos CE, Chen YH, Carducci MA, et al. Chemohormonal Therapy in Metastatic Hormone-Sensitive Prostate Cancer: Long-Term Survival Analysis of the Randomized Phase III E3805 CHAARTED Trial. Journal of clinical oncology : official journal of the American Society of Clinical Oncology 2018;36:1080–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Hawley JE, Pan S, Figg WD, et al. Association between immunosuppressive cytokines and PSA progression in biochemically recurrent prostate cancer treated with intermittent hormonal therapy. The Prostate 2020;80:336–44. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Davis ID, Martin AJ, Stockler MR, et al. Enzalutamide with Standard First-Line Therapy in Metastatic Prostate Cancer. The New England journal of medicine 2019;381:121–31. [DOI] [PubMed] [Google Scholar]
- 27.Bilusic M, Heery CR, Collins JM, et al. Phase I trial of HuMax-IL8 (BMS-986253), an anti-IL-8 monoclonal antibody, in patients with metastatic or unresectable solid tumors. J Immunother Cancer 2019;7:240. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Sanmamed MF, Perez-Gracia JL, Schalper KA, et al. Changes in serum interleukin-8 (IL-8) levels reflect and predict response to anti-PD-1 treatment in melanoma and non-small-cell lung cancer patients. Annals of oncology : official journal of the European Society for Medical Oncology / ESMO 2017;28:1988–95. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Aalinkeel R, Nair B, Chen CK, et al. Nanotherapy silencing the interleukin-8 gene produces regression of prostate cancer by inhibition of angiogenesis. Immunology 2016;148:387–406. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.