We read with interest this article by Berdjis et al.[1]. , which aimed to investigate the Ki-67 labelling index (LI) as a marker of disease progression in squamous cell carcinoma (SCC) of the penis. Whereas we commend their effort, we wish to address several key issues that warrant scrutiny.

Conventionally, clinicians use primary penile histology to assess the risk of disease progression. Tumour grade, stage, depth of invasion and lymphovascular invasion have been used to select patients suitable for lymph node dissection [2]. Solsona et al.[3]. prospectively combined tumour stage and grade to subdivide patients with clinically negative groins into low-, intermediate- and high-risk groups for disease progression. This has practical value in stratifying patients into treatment cohorts.

In their study the authors stated that, “a high Ki-67 LI was significantly associated with poorly differentiated tumours (P < 0.005)”. We are concerned that this statement may be misleading, given that only four of the 44 patients had grade 3 tumours. Furthermore, it would be interesting to know the nodal status and primary histology of the six patients who had disease progression, as these patients may have had confounding factors such as tumour type (e.g. basaloid tumours) and lymphovascular invasion that could have influenced their outcome.

Current methods of predicting outcomes are clearly suboptimal. Invasive techniques such as ultrasonography-guided fine-needle aspiration cytology and dynamic sentinel lymph-node biopsy are currently under investigation. However, the focus appears to have shifted onto noninvasive techniques. Several groups found optimism in the role of other isolated biomarkers for predicting disease progression in penile cancer, e.g. p53 [4,5], proliferating cell nuclear antigen (PCNA) [4] and SCC antigen (SCCAg) [6,7]. The role of a combined panel of biomarkers can potentially increase the predictability of disease progression. This is partly exemplified by the study reported by Tsuji et al.[8]. , who showed that Ki-67 and p53 are useful prognostic indicators in patients with urinary bladder cancer after radical cystectomy.

It would be useful to explore the potential cumulative role of multiple biomarkers in a combined approach using all the proven individual markers such as p53, PCNA and SCCAg in conjunction with Ki-67. Given the relative rarity of the disease, a meaningful conclusion of the prognostic value of this combined approach can only be made in the form of a well-structured multicentre large-scale study. We feel that this may be the best way of achieving a predictive model based on biomarkers alone.