Literature Selection criteria
Master table 1a incorporates examples of each of the issues discussed within the review and the discussion below. The second Master table (1b) is the working table, updated regularly from the literature and our own research. It includes data on the three prostate cancer cell line PC3, DU145 and LNCaP, and genes of interested involved in apoptotic and repair mechanisms as well as other unrelated pathways.
Selection criteria;
1. Fresh human prostate tissue samples.
2. Size of the sample population "N"
3. Controls; normal human prostate over BPH or cell line.
4. Cancer grade range and Grading system.
5. Techniques and technology
6. Clinical information.
7. Availability of the literature.
Discussion:
1. Ideally human prostate tissue and unpreserved, as the shorter the period from subject to experiment the more likely expression patterns (mRNA and protein) will be intact. The identity and location of the diseased tissue within the prostate eg transitional or peripheral zone; and the type of majority of the cells ie epithelial vs stroma is also important but often not detailed. In healthy prostate, each zone and cell types has its own unique expression patterns and should dictate the type of control used.
2. Small experimental numbers are the rule rather than exception. Unfortunately there are only 2 truly representative animal models for the human prostate; the African male lion and dogs. African male lions are simply not an option, and due to local ethical mandates dogs are not available.
Xenografts using cell lines derived from prostate cancer are quite common however there are many issues that must be addressed when using prostate cell lines. The three most common PC3, DU145 and LNCaP, are derived from metastatic secondary cancers in the bone, brain and the lymph respectively [ATCC see links] and therefore cannot represent the early stages of oncogenesis.
Further, there is often a mistaken assumption of cell line stability. Our own observation have concurred with other studies that this is not the case and that the changes in molecular profile are not predictable. Much of the apparent disparity in results between studies using the same cell lines can probably be attributed to this instability.
3. Matched controls of normal patient prostate tissue with the patients¹ diseased tissue would obviously be the ideal situation. This is often impractical and unobtainable. Within the literature the use of an unrelated normal donor, a pool of several unrelated normal donor RNA/DNA (for microarray), BPH tissue of any number of the actual patients involved, BPH from an unrelated normal donor etc and in one or two cases; the use of cell lines as controls.
Cases can be made for most of these control alternatives. However, the issue still remains that analysis of tissue between unrelated tissue donors and tissue types will always be difficult to interpret and open to question.
4. Ranges of population sample cancer grades in an ideal setting would be of a statistically significant cross section of clinically matched patients at the same stage, ie the range would be extremely narrow. As has already been discussed, this is rarely the case and generally the result of the difficulty in obtaining significant numbers of samples for research.
The grading systems used by pathology to determine the stage of the disease includes the Gleason score, Gleason Grade, and TMN. It is extremely difficult to not rant and rave at this point at the inability to compare such different clinical grading systems. Suffice to say, they are NOT interchangeable. Comparing results from studies using different staging systems does NOT allow a useful conclusion to be drawn regards stage. Hence why we have included, where available, the grading system used.
5. Molecular techniques and technology can be roughly divided into DNA and RNA/expression studies. Articles where both have been performed on the same patient samples, have been given a higher priority given the other selection criteria.
6. Clinical information such as age, PSA and clinical staging are generally available in most studies. However other information such as prior case history (eg pre-treatment such as radiation or chemotherapy, androgen ablation etc) is relatively rare and there appears to be little in the way of environmental matching of the patient sample. As clinicians, this is of the utmost interest and importance when relating the information to the clinical setting.
7. Availability of the literature is an economic issue common to many. The number of journals, subscription costs and storage issues preclude hospital and university libraries from sharing and accessing all but locally popular journals. On line journals have aided in reducing the time to access and the storage issue, but generally do not alleviate the subscription costs. Purchasing papers of particular interest can rapidly reach astronomical proportions and the general practice is to refrain since an abstract alone cannot guarantee the value of the article in a given situation.
Infrastructure and Technology overview
Literature Selection criteria
Glossary
References
Figure 1: Chromosomal hot spots
Master Table 1a: Review Profile
|
|
