| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Editorials |
Department of Biochemistry, Faculty of Medicine Building, The University of Hong Kong, Hong Kong SAR, Fax 852-2712-2719, E-mail ihnwong{at}hkucc.hku.hk
The biological cascade of cancer progression starts from uncontrolled proliferation of cells to generate a tumor mass. The enlarging tumor acquires vasculature through which malignant cells detach from the primary site and disseminate into blood, invade adjacent tissues, and ultimately form metastases at distant sites. The circulating tumor load in blood can thus provide an index of tumor progression to identify the risk for developing malignancies. Since Smith et al. (1) reported the detection of circulating melanoma cells (CMCs) in blood by reverse transcription-PCR (RT-PCR), this line of investigation has been pursued by researchers aiming at early detection of cancer cells in blood, bone marrow, lymph nodes, or sentinel nodes for predicting cancer progression (2). Unexpectedly, contradictory results were obtained relating to the diagnostic and prognostic implications of circulating tumor cells (CTCs).
RT-PCR positivity for tyrosinase (TYR) mRNA in lymph nodes, but not that in blood, predicts relapse and correlates with the Breslow thickness, a well-documented prognosticator (3)(4)(5). The CMCs are possibly intermittently shed into blood, or hematogenous dissemination occurs in a small proportion of melanoma patients. Although tumor cells transiently persist in blood, the rate of CMC detection in blood has been associated with tumor stage, relapse, and disease-free survival (6)(7).
In this issue of Clinical Chemistry, Szenajch et al. (8) have focused on the prognostic implication of multiple RT-PCR testing for TYR mRNA in blood of melanoma patients. TYR is a key enzyme involved in the melanin biosynthetic pathway; thus TYR mRNA is presumably a tissue-specific marker for detecting CMCs or healthy melanocytes in blood. However, cells in other tissues also express TYR mRNA, and the assumption of the entire absence of healthy melanocytes in blood of melanoma patients may not be true. RT-PCR results are extremely variable and sometimes negative in advanced melanoma patients. An intermittent pattern of release of CMCs from the primary or secondary tumor into blood may lead to such discrepancies. In addition, this variability could possibly be related to densities of CMCs in blood that are lower than the detection limit of the assay. The cellular and expression heterogeneity in blood highlights the importance of quantitative assessment of target mRNAs (9)(10)(11)(12)(13). In this regard, quantitative RT-PCR is an indispensable tool to be applied for confirming this notion (9)(10)(11)(12)(13). Indeed, the development of real-time quantitative RT-PCR is valuable for monitoring cancer progression (14). The quantity, timing, and duration of cell shedding in melanoma patients should also be addressed, as has previously been demonstrated in liver cancer patients (9)(10)(11).
The frequency of TYR mRNA positivity during repeated testing was significantly associated with melanoma stage, recurrence, and overall survival (8). Although a single positive TYR test could indicate the risk of relapse, multiple RNA diagnostic testing on multiple blood samples may predict cancer progression more accurately. Szenajch et al. (8) have carefully applied statistical analyses, using a logistic regression model to interpret apparently inconsistent RT-PCR data. The frequency of positivity after multiple TYR testing appeared to increase significantly with melanoma stage, independent of the time interval between consecutive TYR RT-PCR tests. The majority of melanoma patients with visceral metastases were systemically treated with chemotherapy, immunotherapy, and chemoimmunotherapy (8). Previously, Wong et al. (11) have demonstrated that systemic chemotherapy could apparently remove the CTCs in blood of liver cancer patients. Although systemic chemotherapy, immunotherapy, and chemoimmunotherapy could possibly eradicate the CMCs, Szenajch et al. (8) did not notice any critical effects of systemic treatment on the persistence of CMCs in melanoma patients. Apparently, multiple TYR mRNA diagnostic testing can be useful for molecular cancer staging for melanoma patients and hence for prediction of relapse or development of metastases.
Detection of prostate-specific antigen (PSA) mRNA in bone marrow of prostate cancer patients has been associated with recurrence and disease-free survival (15). In these patients, the rate of CTC detection in blood was lower than that in bone marrow (16). Previously, the PSA RT-PCR assay has been applied for prostate cancer staging or progression assessment. In another report in this issue of Clinical Chemistry, Schamhart et al. (17) have raised the intriguing idea of applying multiple PSA RT-PCR testing on blood samples to make possible prostate cancer diagnoses, awaiting further investigation and confirmation. Regarding their original statistical approach in the interpretation of multiple RT-PCR testing results, intra- and interlaboratory variation should also be evaluated with reference to serially diluted total RNA from the positive control cell line.
Schamhart et al. (17) have modified the PSA RT-PCR assay by setting a new numeric definition for precise assessment of the circulating tumor load in blood. According to their statistical analyses of the expected Poisson frequency distributions of positive results within RT-PCR repetitions, they have determined the required number of positive results to distinguish between true- and false-positive results. However, PSA mRNA is not a prostate cancer-specific or prostate tissue-specific molecular marker. Multiple repetitions of PSA RT-PCR as conducted by the four participating centers gave rise to extremely divergent results (17).
With different detection limits, Schamhart et al. (17) have compared two nested RT-PCR protocols using 25 and 35 PCR cycles, respectively. The false positivity may be attributed to illegitimate transcription derived from the weak basal promoter activity in virtually all cell types, expressing as low as one mRNA molecule of a tissue-specific gene per 100/1000 cells. In addition, PSA mRNAs may originate from nonprostate cells or circulating healthy/nontumor/nonmalignant prostate cells in blood of healthy individuals or patients with benign prostatic hyperplasia. The fundamental question is how we may distinguish between these different cell types expressing the target mRNAs. Because blood samples were collected from prostate cancer patients before radical prostatectomy or at least 3 weeks after rectal examination or prostate needle biopsy (17), shedding of nontumor cells should not have been expected, as demonstrated in liver cancer patients (10)(11). However, the assumption of the total absence of PSA-expressing prostate or nonprostate cells in the blood of patients with benign prostatic hyperplasia and healthy individuals may not practically hold true. The determination of a cutoff mRNA concentration as a reference is needed for the adjustment of RT-PCR positivity without loss of the diagnostic sensitivity (9)(10)(11)(12)(13).
Widely disparate or inconsistent results might stem from enormous variations in the RT-PCR assays. Positivities could vary tremendously among patients with similar pathologic stages. In terms of RNA diagnostics, one has to consider several crucial parameters, including sensitivity, specificity, false positivity, and reproducibility as reflected by the numbers of PCR tests and blood samples required to predict true-positive or -negative results. The detection rate for PSA mRNA could decrease on multiple testing of the same blood sample, highlighting the stochastic effects of RT-PCR testing and the significance of PCR repetitions (18). Reducing assay sensitivity by decreasing the cycling repetitions may lower the rate of false positivity. Nevertheless, target mRNA markers would still preferably be cancer-specific molecules.
Technically, the RT-PCR assay has to be rapid with anticontamination properties for routine clinical applications. Although multiple nested RT-PCR testing can enhance the sensitivity and specificity for detecting CTCs, this technique has posed a major problem of carryover contamination with the first-step PCR products (8)(17). The sporadic detection of PSA mRNA may possibly reflect its low basal concentration in blood, leading to false negativity or stochastic effects of multiple RT-PCR testing in relation to sampling or pipetting errors. Of relevance, incorporation of a rapid and nonradioactive Southern blot analysis can indeed enhance both the sensitivity and specificity of the RT-PCR assay (9)(10)(11)(12)(13).
A major concern is the assay accuracy for predicting recurrence/metastasis and the prognosis of cancer patients at different stages. Multicentric standardization of the RT-PCR methodology is strongly recommended for interlaboratory comparisons. In the investigation by Schamhart et al. (17), four participating centers were operating multiple PSA RT-PCR testing. However, only 10 prostate cancer patients were recruited from each center for the analyses. Further investigation of a larger cohort of prostate cancer patients is required.
Although the detection of CTCs might direct the selection of high-risk cancer patients for more advanced treatment, PSA RT-PCR may not always predict pathologic stage or biochemical recurrence in prostate cancer patients (19). Unless one can identify the morphologic and phenotypic characteristics of CTCs, the CTCs inferred by the target mRNAs could probably be circulating noncancerous cells derived from specific tissues where the primary tumor or the secondary metastasis is located. Molecular staging based on RT-PCR positivity is attractive, but the possibility of qualitative instead of quantitative RNA diagnostics of CTCs becoming surrogate endpoints for metastasis/recurrence awaits further characterization of CTCs and a longitudinal follow-up of a larger series of cancer patients.
The clinical relevance of RT-PCR results is still controversial. The critical issues include the proportion of CTCs that would subsequently form metastases, the critical mass of CTCs that the host immune system could eliminate, the association of CTCs with tumor recurrence, and the source/extent of illegitimate transcription in cells isolated from blood. Very importantly, we need to confirm the association of CTC detection at diagnosis with cancer stage, an important prognosticator related to recurrence and overall survival. Certainly we should optimize and standardize the RT-PCR assay for routine clinical applications and interlaboratory comparisons. The mere detection of CTCs cannot shed light on their potential to complete the metastatic cascade. Characterization of CTCs in terms of their biological capabilities of developing metastases can provide insights into the metastatic mechanisms.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
