Skewing towards neuroendocrine phenotype in high grade or high stage androgen-responsive primary prostate cancer: Commentary