Evaluating immune responses after sipuleucel-T therapy

Abstract

Following FDA approval of sipuleucel-T in 2010 for metastatic castration resistant prostate cancer (mCRPC), several studies have described the effect of sipuleucel-T on peripheral immune responses. Retrospective associations have also been made with immune responses and survival. A recently published study by Fong et al. was the first to characterize the immune response of sipuleucel-T in the tumor microenvironment. The findings of this study have been hypothesis generating, yet it remains unclear whether the peri-tumor immune response described is predictive of survival. Increasing evidence suggests that radiographic or PSA progression does not accurately reflect survival with sipuleucel-T and other immunotherapies. Finding an immune biomarker which can accurately reflect clinical benefit and validating it prospectively offers the potential for a predictive indicator of response in an area where none currently exists.

Prostate cancer is the most common cancer type among men in the United States, and the second leading cause of cancer death. It was predicted that 27,540 men will die from prostate cancer in the United States in 2015.¹ At present, androgen deprivation therapy is the cornerstone of initial treatment for metastatic disease. Ultimately most men with metastatic disease will progress and develop castration resistant prostate cancer (mCRPC). Until 2010, docetaxel was the only therapy shown to prolong survival in this patient population. This finding was reported in the TAX 327 study, in which men with mCRPC who were randomized (with potential for cross-over) to docetaxel 75 mg/m² every 3 weeks, plus prednisone 5 mg twice daily, showed a 2.9 month survival benefit, as compared to patients treated with mitoxantrone 12 mg/m² every 3 weeks, plus prednisone 5 mg twice daily (19.2 months versus 16.3 months; P = 0.004).² Since 2010 the FDA has approved a number of additional therapies for mCRPC. Of particular interest is sipuleucel-T, the first approved therapeutic cancer vaccine. This approval was based upon the results of the IMPACT trial, where 512 men with mCRPC were randomly assigned (in a 2:1 ratio) to receive sipuleucel-T every 2 weeks, for a total of 3 doses, or placebo. It was found that patients receiving sipuleucel-T benefited from a 4.1 month improvement in median overall survival (OS) (25.8 months vs. 21.7 months; HR 0.78; 95% CI 0.61 to 0.98; P = 0.03).³

Sipuleucel-T is an autologous vaccine developed from peripheral mononuclear cells that have been cultured ex vivo with PA2024, a fusion protein composed of prostatic acid phosphatase (PAP) and granulocyte monocyte colony stimulating factor (GM-CSF). Sipuleucel-T is thought to cause an antigen-presenting cell (APC) mediated T-cell immune response against PAP, an antigen which is highly expressed by most prostate cancers. An analysis of 3 phase III clinical trials of sipuleucel-T (n = 737) showed a humoral or cellular immune response to PA2024 (PAP/GM-CSF fusion protein) in 78.2% (122/156) of sipuleucel-T treated patients, versus 10.5% (8/76) of control patients. Additionally, 39.5% (60/152) of sipuleucel-T treated patients showed a response to PAP, vs. 5.7% (4/70) of control patients.⁴ The humoral response was still detectable 6 months after treatment, with anti PA2024 and anti PAP IgG antibodies continuing to increase after 26 weeks. Sipuleucel-T treated patients who developed at least one peripheral immune response (T-cell proliferation, IFNγ ELISPOT or antibody production) to PA2024 or PAP were also associated with significantly better OS (HR = 0.47) (95% CI: 0.29–0.78; P = 0.003).⁴

To understand the effect of sipuleucel-T not just peripherally, but on the tumor microenvironment, Fong et al. conducted a phase II clinical trial of sipuleucel-T in the neoadjuvant setting for 37 patients with localized prostate cancer, who were eligible for radical prostatectomy.⁵ In this study, they found that prostatectomy specimens of patients treated with sipuleucel-T showed more than a 3 fold increase in infiltrating CD3+, CD4+FOXP3- and CD8+ T cells, compared with pretreatment biopsies. Interestingly, these responses were limited to the tumor margins. The authors hypothesized that the lack of a broader immune response including within the tumors may have been due to the immunosuppressive nature of the tumor microenvironment or to the timing of the biopsy following therapy (only 2–3 weeks following the last sipuleucel-T dose). Although the reason remains unclear it is noteworthy that sipuleucel-T was able to cause a local immune response in these patients. The question remains whether or not this immune response is reflective of a clinically meaningful outcome when seen in patients after radical prostatectomy.

Some data is available from similar disease settings which may help put these findings into context. In colorectal cancer, T cell intra-tumor infiltration has repeatedly been associated with improved disease free and OS. However, the association with peri-tumor infiltration remains unclear. Immunohistochemical (IHC) analysis on 109 resected colon carcinomas from patients with stage II/III colon cancer by Guidoboni et al.⁶ found that higher densities of CD8+ T cells within the tumor was independently associated with longer OS, with HR being 0.33 (95% CI 0.15–0.73) after 5 years of follow up. Ropponen et al.⁷ performed a histopathologic analysis of resected colorectal cancer specimens from 276 patients with stage I-IV colorectal cancer. These patients were followed for a median of 14 years and it was found that a higher density of tumor infiltrating lymphocytes was independently associated with improved OS (P < 0.01) on multivariate analysis. A study conducted by Pages et al.,⁸ where IHC analysis of tissue microarrays was prepared from 415 resected colorectal cancers and analyzed using multivariate analysis, found that a high density of CD45RO+ memory T cells inside the tumor was independently associated with longer OS (P = 0.02). Patients with tumors which had a high density of T cell infiltration had a median disease-free survival of 36.5 months, and a median OS of 53.2 months, as compared with 11.1 months and 20.6 months, respectively, in patients whose tumors had a low density of T cell infiltration (P < 0.001 for both comparisons).

Experimental identification of T cell infiltration at the tumor margin has led to conflicting results, with some studies indicating the presence of T cells as an independent prognostic factor for improved disease free survival, while others show no correlation with survival. Using IHC techniques, Menon et al. ⁹ analyzed resected colorectal cancer samples from 93 patients with stage II/III colon cancer, and found that both higher CD 8 + (P = 0.04) and CD 57+ (P = 0.05) cell infiltration within the tumor margin correlated with prolonged disease free survival in multivariate analysis. After an immunohistochemical analysis of 843 colorectal cancer specimens, Ogino et al.¹⁰ found that intra-tumor as well as peri-tumor lymphocytic infiltration were both associated with longer OS (P = 0.005 and P = 0.014 respectively). However, it should be noted Ogino et al. did not assess peri-tumor lymphocytic infiltration independent of intra-tumor infiltration, as nearly all specimens with peri-tumor lymphocytic involvement also had intra-tumor involvement. After performing IHC analysis on colorectal cancer specimens from 131 patients, Natio et al.¹¹ concluded that CD8+ T cell infiltration was only independently associated with survival when there was infiltration within cancer cell nests (P = 0.016) but not when there was CD 8+ T cell infiltration of tumor stroma (P = 0.56) or the tumor margin (P = 0.68). In addition, McMullen et al.¹² conducted IHC analysis on 40 rectal cancer specimens from patients with stage I-IV rectal cancer and found that CD3+ T cell infiltration of the tumor margin was not independently associated with survival (P = 0.52).

Given this information it is difficult to interpret the clinical significance of the peri-tumor immune response seen in Fong et al.⁵ However, regardless of the known clinical impact of this response, the presence of T cells in the tumor microenvironment is important because it speaks to the pharmacodynamic and immunologic impact of sipuleucel-T.  In addition we should be cautious in interpreting this immune data as whole mounts from prostatectomy specimens may have allowed for more tissue sampling than biopsies pretreatment.

The importance of this data is driven by the current lack of clinical or immunologic data which predicts response to sipuleucel-T or provides an intermediate marker of response. The resulting dilemma is highlighted by the clinical studies of sipuleucel-T in mCRPC. Data from 3 phase III trials have failed to show a PFS benefit in the setting of a significant OS benefit. In the initial study, 127 patients with asymptomatic mCRPC were assigned in a 2:1 ratio to receive sipuleucel-T (n = 82) or placebo (n = 45).¹³ While there was no difference in time to progression (TTP) between sipuleucel-T (11.7 weeks) and placebo (10.0 weeks) treatment groups, there was a significant increase in median OS for those given sipuleucel-T (25.9 months versus 21.4 months; HR 1.70; 95%CI 1.13 to 2.56; P = 0.01).¹³ These results were integrated with a second similarly designed trial including 98 patients (65 for sipuleucel-T and 33 for placebo)[9]. Again, no difference was noted in TTP between the 2 groups, but median OS was significantly increased for sipuleucel-T-treated patients (19.0 months vs. 15.7 months).¹⁴ In the largest of the 3 trials, the IMPACT trial, 512 patients were randomly assigned in a 2:1 ratio to receive sipuleucel-T (n = 341) every 2 weeks, for a total of 3 doses, or placebo (171).³ Similarly to the previous trials, patients receiving sipuleucel-T benefited from a 4.1 month improvement in median OS (25.8 months versus 21.7 months; HR 0.78; 95% CI 0.61 to 0.98; P = 0.03) while having no significant difference in time to disease progression (14.6 weeks vs. 14.4 weeks). These studies showed an OS benefit with sipuleucel-T, however none of them showed an improvement in short-term progression free survival (PFS). The absence of a validated short-term marker of response such as PFS makes the need to find and validate immune responses which can accurately reflect clinical benefit that much more pressing. For this to be done immune responses need to be better understood and although many markers can be evaluated, the end result, immune response at the tumor is likely the most relevant result. However the greatest hurdle with this approach is acquiring tumor for assessment in clinical trials. Furthermore, it is not clear that a given biopsy represents the immune response to the broader tumor burden. For these reasons future studies will focus on correlating peripheral immune markers with immune responses seen in the tumor microenvironment.

Notably, Fong et al.⁵ reported that in the neoadjuvant setting there was no correlation between peripheral (ELISPOT response to either PA2024 or PAP) and local (CD3+, CD4+FOXP3- and CD8+ T cell infiltration) immune responses 2–3 weeks after completion of sipulecuel-T therapy. These results demonstrate that peripheral immune responses evaluated in this study following therapy may not reflect local responses. Even so, the potential remains that other peripheral immune responses or even these same parameters evaluated at different time points may be associated with an effective localized immune response. Furthermore, a correlation between local and peripheral immune responses may still be found in the metastatic setting where disease is more systemic. For this reason future studies evaluating the immune response of sipuleucel-T in patients with metastatic prostate cancer should ideally evaluate both local and peripheral immune responses.

In addition Fong et al.⁵ found that following sipuleucel-T therapy nearly half of the CD3 T-cells at the tumor interface expressed PD-1 suggesting that activated T cells in this area are highly sensitive to immune checkpoint blockade. Under these circumstances check point inhibitors, such as anti PD-1, when used following sipuleucel-T may facilitate a more effective and clinically meaningful immune response. This rationale has led to the combination of sipuleucel-T and immune checkpoint inhibitors in ongoing trials of mCRPC, including a phase II trial evaluating sipuleucel-T with and without CT-011 (anti -PD1 mAb) [NCT01420965] and a phase II trial evaluating sipuleucel-T followed by immediate or delayed ipulumimab [NCT01804465].

In conclusion, increasing evidence suggests that traditional short term clinical data, such as PFS, may not accurately reflect OS with many immunotherapies. In addition to sipuleucel-T,³ a phase II study of Prostvac-VF for mCRPC,¹⁵ as well as a phase III study of ipilumimab for metastatic melanoma,¹⁶ have both shown an OS benefit in the absence of a short-term PFS benefit. There is therefore a growing need for short term surrogates of survival which accurately reflect long term clinical benefits. Moving forward, studies should continue to look at both local and systemic immune responses with the ultimate goal of correlating these responses and finding clinically relevant results which accurately reflect improved outcomes.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.