Six-Transmembrane Epithelial Antigen of the Prostate 1–4 (STEAP1–4) compose a family of metalloproteinases involved in iron and copper homeostasis and other cellular processes. In prostate cancer, STEAP1, STEAP2, and STEAP4 are overexpressed, while STEAP3 expression is downregulated. STEAP1–4 can be used as a biomarker and therapeutic target for prostate cancer.
1. STEAP1, STEAP2, and STEAP4 as Biomarkers for Prostate Cancer
Serum PSA is currently the gold-standard biomarker for prostate cancer (PCa) screening, detection, and prognostication. However, recent PCa screening trials highlighted the limitations of using PSA—namely, that PSA screening does not significantly reduce mortality in PCa patients and is associated with a high risk of overdiagnosis
[1][2]. This led the United States Preventive Services Task Force to recommend against routine PSA-based PCa screening, particularly in patients who are 70 or older
[3][4]. Since PSA screening does sometimes detect PCa early, the USPSTF recommendation is somewhat controversial. This warrants finding PCa biomarkers with better specificity.
STEAP1, STEAP2, and STEAP4 have been evaluated as possible diagnostic and prognostic biomarkers in glioblastoma, breast cancer, Ewing sarcoma, lung cancer, and PCa
[5][6][7][8][9]. While the literature disagrees about the statistical significance of using STEAP1, STEAP2, and STEAP4 for PCa screening, diagnosis, and prognosis, the evidence suggests some degree of clinical utility.
STEAP1 is a promising diagnostic and prognostic biomarker. One method of using STEAP1 as a biomarker is detecting the number of circulating STEAP1-positive extracellular vesicles (EVs). In two studies, PCa patients were found to have significantly elevated STEAP1 EV levels when compared to healthy males
[10][11]. Khanna et al. concluded that EV-based liquid biopsy may be a useful diagnostic strategy in PCa, but they found no association between total STEAP1 EV levels and disease recurrence or overall survival
[10]. Currently, no EV-based diagnostics for PCa exist. Exploring STEAP1 as a prognostic biomarker, Ihlaseh-Catalano et al. investigated whether STEAP1 overexpression was associated with higher Gleason grades, seminal vesicle invasion, shorter biochemical recurrence-free survival, and higher mortality
[12]. Although they found associations between STEAP1 and these prognostic measures, the only significant association was between STEAP1 overexpression and biochemical recurrence. Meanwhile, Gomes et al. showed that STEAP1 is useful for distinguishing malignant PCa from benign prostatic hyperplasia
[13]. However, they acknowledged that STEAP1 lacks specificity in distinguishing prostatic intraepithelial neoplasia from PCa.
STEAP1-specific antibodies are being tested as diagnostic imaging agents. A notable example is
89Zr-DFO-MSTP2109A, which uses an STEAP1-specific humanized immunoglobulin G1 (MSTP2109A) to target
89Zr-desferrioxamine (DFO), a positron emission tomography radionuclide, to mCRPC
[14][15]. Ongoing phase I/II clinical trials show that
89Zr-DFO-MSTP2109A is well tolerated and demonstrates excellent uptake in bone and soft-tissue mCRPC sites (ClinicalTrials.gov identifier NCT01774071).
STEAP2 and STEAP4 are interesting biomarker candidates because, in the prostate, they are overexpressed in malignant tissue. Burnell et al. showed that STEAP2 overexpression was significantly correlated with the Gleason score, a measure of prognosis
[9]. STEAP2 is particularly promising because, in a series of stratified analyses with possible confounders (age, PSA), STEAP2 was found not to significantly correlate with age or PSA. However, a previous study found no significant association between STEAP2 and the Gleason score
[16]. Burnell et al. also showed that STEAP4 may be useful in predicting PCa relapse
[9]. The number of relapsing patients in the study, though, was small.
Ultimately, further investigation is needed to determine whether STEAP1, STEAP2, and STEAP4 are useful diagnostic and prognostic biomarkers for PCa.
2. STEAP1 as a Therapeutic Target for Prostate Cancer
PCa-specific antigens that are therapeutic targets include prostate stem cell antigen (PSCA), prostate-specific membrane antigen (PSMA), and STEAP1. This section will only discuss STEAP1. Given its localization to surface cell–cell junctions, particularly that of the prostatic secretory epithelium, STEAP1 is a promising target for T-cell and antibody-based immunotherapy
[17][18]. Because STEAP1 is overexpressed in malignant prostate tissue but expressed at low levels in normal prostate tissue, STEAP1 is an ideal therapeutic target (
Table 1)
[14][15][19][20][21][22][23][24][25][26][27][28][29].
Table 1. Therapeutic agents that target STEAP1. mRNA, messenger ribonucleic acid. DNA, deoxyribonucleic acid. CAR-T, chimeric antigen receptor T-cell therapy. MSKCC, Memorial Sloan Kettering Cancer Center. FHCC, Fred Hutchinson Cancer Center.
STEAP1 is immunogenic; STEAP1-derived peptides expressed on MHC class I molecules induce cytotoxic CD8+ T lymphocyte (CTL) activity in PCa ex vivo
[30][31][32]. These studies found that several epitopes of STEAP1 induce CTL activity, e.g., STEAP
86–94, STEAP
262–270, and STEAP
292–300, making them attractive targets for cancer vaccines. In cell-line and mouse models, vaccination with these epitopes conferred protection against new tumor growth; vaccination against some of these epitopes also attenuated the growth of well-established tumors. Since STEAP1 is a self-antigen, one concern about these vaccines is that they may trigger autoimmunity. Recognizing this possibility, Luz Garcia-Hernandez et al. tested whether STEAP1 vaccines induce autoimmunity in mice
[32]. Under physiologic conditions, autoantibodies against STEAP1 were generated at a low but detectable quantity. Notably, though, no pathologic autoimmunity was observed. This may be due to the low basal levels of STEAP1 expression in normal tissues.
Commercially developed PCa vaccines include CureVac’s CV9104, a self-adjuvanted full-length mRNA vaccine that targets the PCa antigens PSA, PSCA, PMSA, STEAP1, prostatic acid phosphatase (PAP), and mucin 1
[33]. Although CV9104 was well tolerated and triggered a robust immune response, its clinical trial (ClinicalTrials.gov identifier NCT01817738) was terminated because CV9104 did not improve overall survival
[19]. As of this writing, no active clinical trials are underway for STEAP1 vaccines.
DSTP3086S (vandortuzumab vedotin) is an antibody–drug conjugate (ADC) that, like
89Zr-DFO-MSTP2109A, contains MSTP2109A
[26]. In DSTP3086S, MSTP2109A is conjugated via a protease–labile linker to the potent antimitotic agent monomethyl auristatin E (MMAE). After antigen-specific binding of DSTP3086S to STEAP1-overexpressing cells, DSTP3086S is endocytosed. MMAE is then released intracellularly, leading to inhibition of cell-cycle progression. DSTP3086S was investigated in a phase I, multicenter, open-label, dose-escalation study (ClinicalTrials.gov identifier NCT01283373). Compared to traditional chemotherapies for mCRPC, DSTP3086S exhibited a relatively mild adverse effect (AE) profile; the most common AEs were fatigue, peripheral neuropathy, and gastrointestinal symptoms. Notably, these were associated with the drug MMAE, not the STEAP1-specific antibody MSTP2109A. Danila et al. noted that, while DSTP3086S requires optimization for further clinical development, STEAP1-targeting ADCs offer promise.
AMG 509 is an Xmab
®2+1 T-cell engager that contains two identical anti-STEAP1 Fab domains, an anti-CD3 scFv domain, and an Fc domain that prolongs serum half-life
[34]. Announcing its development, Li et al. described AMG 509 as having potential treatment utility for mCRPC and Ewing sarcoma. AMG 509 induces CTL-mediated cytotoxicity of STEAP1-positive cancer cells, with a median EC
50 of 37 pM across 19 STEAP1-expressing cancer cell lines. Notably, the presence of two identical anti-STEAP1 Fab domains (versus one) increased CTL activity against cancer cells and decreased off-target activity against normal cells. The bispecificity of AMG 509 ensures that AMG 509 largely binds cells that overexpress STEAP1, resulting in more robust anti-cancer activity and fewer AEs. A phase I, multicenter, open-label study for assessing the safety, tolerability, pharmacokinetics, and efficacy of AMG 509 is currently recruiting mCRPC patients (ClinicalTrials.gov identifier NCT04221542)
[27]. As of this writing, AMG 509 is the only STEAP-related therapeutic that is actively recruiting for clinical trials.
3. Other Considerations and Perspectives on the Role of STEAP1–4 in Prostate Cancer
Other anti-STEAP1 therapeutics are under investigation for renal cell carcinoma, urothelial carcinoma, and Ewing sarcoma
[28][35]. Lin et al. described BC261, a rehumanized STEAP1-IgG that is bispecific for STEAP1 and CD3
[28]. Its design is like that of AMG 509; BC261 consists of two identical anti-STEAP1 Fab domains and two identical anti-CD3 scFv domains. Preliminary results in the Ewing sarcoma, PCa, and canine osteosarcoma cell lines demonstrated significant elevation of T-cell infiltration and tumor ablation. There is, thus, impetus to investigate other anti-STEAP antibody-based agents for therapeutic purposes in cancers other than PCa. Excluding AMG 509 as a possible therapeutic for Ewing sarcoma, these studies have not yet led to clinical trials.
Although STEAP1 is the primary target of current clinical investigation, other STEAP proteins are also the subjects of translational studies. For example, Machlenkin et al. identify STEAP3-derived epitopes as good vaccine candidates for immunotherapy of PCa. In vitro cytotoxicity assays demonstrated that these epitopes induced strong CTL-mediated antitumor activity
[36]. Other approaches for targeting STEAP in PCa include fusion protein vaccines, RNA–lipoplex vaccines, recombinant viral vaccines, DNA vaccines, and CAR-T cells
[20][21][22][23][29]. These studies are promising, and there is much work to do before they can be translated into PCa therapeutics. The literature on targeting STEAP2-4 is limited, but the approaches used to study STEAP1 can also be used to study STEAP2-4.