The optimal treatment strategy for this subtype of prostate cancer is unknown since there is no clinical trial specifically designed for SRPC. In this study, we indentified significant age-and racial-related disparities as an important factor for selecting prostatectomy or radiation therapy, which is consistent with previous findings [11, 13, 14]. For example, younger age was the strongest predictor for receiving radical prostatectomy rather than external beam radiotherapy (EBRT) and younger age as a predictor of aggressive local therapy, while older prostate cancer patients were often treated less aggressively [15, 16]. Efforts should be made to indentify high risk individuals for potential intervention in order to reduce these disparities.

Consistent with single institution studies, the cancer-specific survival of SRPC is poor. In our study, the 3- and 5-year survival rates of SRPC were 89.6% and 83.8% (Table 5), which are significantly lower, in comparison with 5- and 10-year survival rates of 99.9% and 92%, respectively, in patients with prostate cancer as a whole [17].

Despite receiving more aggressive therapy, younger patients in this study had a poorer prognosis (Figure 1(c)). Controversy exists regarding the importance of patient age in disease behavior of prostate cancer. In several series, younger age at diagnosis has been correlated with more aggressive tumor types and subsequent mortality [18, 19]. Conversely, other studies have shown survival rates among younger patients to be equivalent or even superior to those of elderly patients [20–22]. Lin et al. [22] examined the association between age at diagnosis and grade, stage, treatment, and survival outcomes in men who were diagnosed during the era of PSA testing. Younger men were more likely to undergo prostatectomy, have lower grade cancer, and, as a group, to have better overall and equivalent cancer-specific survival at 10 years compared with older men. However, among men with high grade and locally advanced prostate cancer, the younger men had a poorer prognosis compared to older men. The subgroup of men with high grade and locally advanced prostate cancer described in Lin's study are very similar to our subjects. 82.8% of our entire study cohort (90% of the patients with known histology grade) had either poorly or undifferentiated histology. The findings collectively, from this study and others, support the view that prostate cancer is a heterogeneous disease, different subtypes may represent different diseases [22–25]. In practical terms, these findings suggest that better diagnosis and therapy of prostate cancer are likely to be achieved by investigating each subtype of prostate cancer separately rather than grouping them all together [25]. Our finding of subtype-related differences in survival has significant therapeutic implications with regard to patient selection, trial design, and therapy recommendations and warrants further study to offer this poor-prognosis group of men better preventative and therapeutic options.

Cancer incidence is known to be increased with age [12, 17]. Paradoxically, tumor growth and metastasis often occurred at a slower rate in aged human and animal populations. For example, tumors grow slower and metastasize less in old patients with cancer [26–28]. In experimental models, tumor growth has frequently been shown to be slower and to display a reduced aggressiveness in aged as compared to young animals [29]. The mechanisms responsible for these phenomenons have not yet been established. Decreased proliferative capacity [30], decline in growth factors with age [31], and age-related changes in antitumor immunity [32] have been suggested. Recently, an increased apoptotic cell death has also been linked to the reduced malignant behavior of tumors in the aged [33].

Currently, there is no literature addressing the impact of age on survival, specifically on SRPC. Our finding of age as an independent prognostic factor for this disease is interesting. The survival difference among different age groups might be explained by a difference in the biologic behavior of SRPC in the younger patients and elderly. Molecular research may provide additional insights into these questions. Similarly, breast cancer arising in young women is correlated with inferior survival and higher incidence of negative clinicopathologic features. A large-scale genomic analysis illustrates that breast cancer arising in young women is a unique biologic entity driven by unifying oncogenic signaling pathways [34]. Age-specific differences in oncogenic pathway dysregulation have also been investigated in patients with acute myeloid leukemia [35]. The age-related differential biological behavior of tumors also implies the necessity of a differential therapy for cancer patients of different ages. In fact older patients are now more frequently treated with radiation therapy than surgery modulated by projected patient longevity and biomarkers predicting cancer recurrence 7 years following initial biopsy or resected tumor specimen using Systems Pathology [35, 36].

Similar to the findings of increasing in 5-year survival in men with prostate cancer in post-PSA era, we also observed a significant improvement in the outcome of patients with SRPC during this period (Figure 1(d)). These improvements were consistent in all SEER stage groups and are likely due to widely application of PSA screening, early diagnosis, advance in local and systemic therapy, and increasing adoption of multidisciplinary prostate cancer care [14, 17].

Our findings should be interpreted with caution. First, this is a nonrandomized study; therefore, selection bias might have been present because patients undergoing surgery tend to be healthier. Although we adjusted for differences in demographic and tumor factors, residual confounding might still exist. Second, the pathological diagnoses in SEER were based on local pathologists' reports, and there was no central review of pathology reports. In addition, SEER data did not allow us to examine receipt of hormone or chemotherapy and patients' comorbidities. However, the analysis reported here attempted to overcome this data limitation by measuring prostate cancer-specific survival, rather than overall survival. Finally, the sample size in our study may still not be large enough to fully describe the factors that affect the incidence, treatment choice, and survival of this rare prostate cancer subtype. Our findings may not necessarily be generalized to patients with prostate adenocarcinoma as a whole.