Re: Andres Lidgren, Ylva hedberg, kjell grankvist, torgny rasmuson, anders bergh and borje ljungberg. Hypoxia-inducible factor 1 alpha expression in renal cell carcinoma analyzed by tissue microarray

Dr. Lidgren and coworkers have to be congratulated far the presentation of their analysis of hypoxia- inducible factor la (HIF-la) expression using tissue microarray in a large series of 216 patients with renal cell carcinoma (RCC) of any T stage, treated with radical nephrectomy [1]. The authors also reported on the possible correlation between HIF- la expression and other well-known clinicopatho- logic variables. It is very important to notice that the percentage of positively stained tumour cells was uniform throughout the study period, confirming that tissue microarray is a useful method that provides stable, feasible, and consistent results also if done on formalin-fixed tumour blocks. In this study, the HIF-la staining was detected using a monoclonal antibody (NB100-131; Novus Biologicals, Littleton, CO); the HIF-la staining was found mainly in the cytoplasm and nuclear staining was only sparsely observed and was disregarded. The level of HIF-la expression did not differ among the RCC subtypes and, interestingly, for patients with clear- cell RCC (cc-RCC), tumour size correlated inversely (p = 0.012) with HIF-la expression and high HIF-la cytoplasm staining correlated with longer survival with a p value that nearly reached significance (p = 0.0553). Intracapsular cc-RCCs are known to have an overall excellent prognosis, but 15-20% of patients with Tl- T2 will have a recurrence and eventually die from the disease [2]. Therefore, to evaluate the prognostic significance of HIF-la cytoplasm staining in intracapsular cc-RCC, given the reported inverse correlation between HIF-la expression and tumour size in the whole series, it would have been interesting if the authors could have provided their results in the 66 patients with cc-RCC limited to the kidney (pT1a, pT1b, pT2). In most cancers and also in reports on cc-RCC, high HIF-la levels have been associated with an unfavourable prognosis [3-5]. On the contrary, Lidgren and coworkers indicate a trend towards better survival for patients with high HIF-la cyto- plasmic protein expression but, given the nuclear transcriptional action of the HIF-la protein [6-8], the evaluation and detection of a high HIF-la nuclear staining in such patients with better prognosis, would have strengthened the authors thesis that "an upregulation of HIF-la is not necessarily all negative, at least not in conventional RCC" and that "a hypoxia-independent pathway, other than through HIF-la, might be present in this RCC type," but unfortunately using the above-mentioned monoclonal antibody and tissue microarray, the nuclear staining was only sparsely detected. In a previous study, the same study group showed that HIF-la mRNA expression represents a favourable independent prognostic factor in RCC, although no inforrnation was available on the subcellular location ofHIF-la [9]. The different subcellular location ofHIF- la might constitute an important confounding factor in expression studies. Moreover, it would have been of great value if the authors could provide also the data about the von Hippel-Lindau tumour suppressor protein (pVHL) expression to confirm the inverse correlation amongpVHL 19, pVHL 30, and HIF-la and therefore to show tha t not only high cytoplasmic HIF- la expression but also low pVHL expression correlate with longer survival in their series. As far as we are concemed, being currently involved with the evaluation of pVHL and HIF-la expression using tissue microarray on formalin- fixed tumour blocks, the positive correlation between high HIF-la cytoplasmic staining and survival, without detecting the same correlationn between high HIF -10' n uclear staining and survival cannot be considered enough to clear the field and further studies are needed to shed light on renal tumourigenesis.