Molecular Staging in Stage II and III Melanoma Patients and Its Effect on Long-Term Survival
http://www.100md.com
《临床肿瘤学》
the Department of Dermatology of the Charite, Humboldt University
Department of Medicine III, Campus Benjamin Franklin, Charite Berlin, Berlin
Department of Biometry and Medical Documentation and Department of Dermatology, University of Ulm, Ulm
Department of Dermatology, University of Mainz, Mainz, Germany
Department of Radiology, Memorial Sloan-Kettering Cancer Center, New York, NY
ABSTRACT
PATIENTS AND METHODS: During routine follow-up of American Joint Committee on Cancer stage II and III melanoma patients, serial testing for tyrosinase transcripts in peripheral blood was performed by RT-PCR. The PCR results were compared with the clinical data collected during the follow-up.
RESULTS: Over a period of 3 years, 111 patients (78 stage II and 33 stage III patients) were enrolled, and tyrosinase determinations were carried out. The 6-year disease-specific survival probability was 97% for patients always showing negative RT-PCR results and 67% for patients who tested positive at least once. In a Cox proportional hazards model, the prognostic value of sex, age, site of primary tumor, histologic subtype, stage, Breslow's tumor thickness, Clark level, and the time-dependent variable PCR result was assessed. Patients with a positive RT-PCR test had a distinctly higher risk of dying from melanoma, with a hazard ratio of 12.6 (95% CI, 3.4 to 46.3; P < .001).
CONCLUSION: Our study shows a strong association between PCR and disease-specific survival time. Detection of tyrosinase mRNA in peripheral blood may be of similar importance for the clinical course of melanoma as the detection of micrometastatic disease in the sentinel lymph node. Whether a combination of these two factors leads to a better definition of the prognosis of melanoma patients is under investigation in current studies.
INTRODUCTION
Since the original description of a reverse transcriptase polymerase chain reaction (RT-PCR) method for detection of tyrosinase transcripts in peripheral blood,4 controversial data have been reported in terms of sensitivity of this method and its clinical relevance.5-12 The use of a multiple marker assay was introduced to improve the sensitivity.13 Recurrence-free and disease-specific survival depending on RT-PCR test results have been rarely addressed in homogenous patient cohorts.12,14-16 Therefore, large prospective studies with long-term follow-up in stage II and III melanoma patients are necessary to determine the prognostic value of these assays.16-19 We report on serial tyrosinase RT-PCR measurement in 111 consecutive stage II and III melanoma patients during routine follow-up examination to investigate the prognostic relevance of this assay.
PATIENTS AND METHODS
All blood sampling was performed concomitantly to tumor staging procedures as patients presented for routine follow-up examinations at scheduled intervals. Patients with primary tumors with a tumor thickness of up to 1.5 mm presented every 6 months, and patients with primary tumors thicker than 1.5 mm and stage III patients presented every 3 months; therefore, the time intervals between consecutive tests performed on patients differed, which is the normal course in follow-up. In an earlier study, we reported our experience with tyrosinase RT-PCR from peripheral blood,23 but the patients of the study reported here are completely different and no subset of the former study.
Blood Sample Processing, RNA Preparation, and Reverse Transcription
Samples of 2 x 2.7 mL EDTA blood were taken and processed within 2 hours. For RNA preparation, both samples were pooled and treated with erythrocyte lysis buffer. Extraction of total RNA was performed by means of Qia shredder columns for purification and Rneasy TM total RNA kit (Qiagen GMBH, Hilden, Germany). All precaution was taken to ensure integrity and purity of RNA. RNA concentration and purity were determined by UV spectrophotometry. RNA was quantified by UV spectrophotometry at 260 and 280 nm and stored at –80°C. Total RNA (1.5 μg) was transcribed by means of 50 ng of random hexamers, 10 mmol/L of deoxynucleotide triphosphate mix, and 200 U of superscript II RT (GIBCO BRL, Inc, Grand Island, NY).
Nested Tyrosinase PCR
Primers for a nested tyrosinase PCR were used as previously described.4 The outer primers of HTYR1 (TTGGCAGATTGTCTGTAGCC) and HTYR2 (AGGCATTGTGCATGCTGCTT) amplify a product of 284 base pairs (bp), whereas the inner primers of HTYR3 (GTCTTTATGCAATGGAACGC) and HTYR4 (GCTATCCCAGTAAGTGGACT) amplify a product of 207 bp. Thirty cycles were run using a schedule described elsewhere.22,23 To avoid contamination, the reaction mixtures were prepared in a fume hood. The second amplified product was analyzed on a 3% agarose gel followed by ethidium bromide staining (80 minutes, 80 V). A 100-bp ladder (GIBCO) was used as standard. PCR samples free of cDNA served as negative controls, and PCR samples with cDNA of the SK-mel-28 melanoma cell line served as the positive control.
Quality Control Experiments
To test the sensitivity, 10-fold serial dilutions of SK-mel-28 melanoma cells (0 to 105 melanoma cells/10 mL blood) in peripheral blood from a healthy donor (spiking experiments) were performed. One SK-mel-28 melanoma cell in a background of 106 peripheral-blood mononuclear cells could be detected. This can also be mimicked by aspiration of a single cell by means of a heat-drawn glass capillary and by performing a single-cell PCR as previously described.24
We had participated in an interlaboratory trial of the European Organisation for Research and Treatment of Cancer for interlaboratory quality assurance.25 Several steps have been taken to detect any cross-over contamination. Before cDNA synthesis, RNAs were purified via silica-gel columns and treated with DNAse. cDNA has always been transcribed in duplicate, with and without superscript enzyme. Each PCR run was controlled by a second run starting from RNA preparation to cDNA synthesis and PCR procedure to confirm results. PCR from RNA as template (ie, PCR without reverse transcription) was carried out to exclude amplification of genomic cDNA. Negative and positive controls were run on each agarose gel in addition to patient samples. If the results of patient samples were discordant, the whole procedure starting from RNA preparation was rerun twice until two concordant results were seen. To ensure that RNA in the tyrosinase gene-negative samples had not been degraded, a parallel PCR run for the detection of glyceraldehyde-3-phosphatase dehydrogenase (housekeeping enzyme) using an equal amount of RNA was performed. The specificity of PCR products was examined by sequencing in all positive samples. cDNA was amplified by using HTYR3/4 with previously marked universal primer M13 (forward and reverse). Sequencing was performed via dye primer sequencing (Perkin Elmer, Norwalk, CT). The analysis of each specimen was carried out on a 373 DNA sequencer (ABI; Applied Biosystems, Foster City, CA). In 20 blood samples of healthy donors and of patients with nonmelanoma cancers, tyrosinase transcripts were absent.
Statistics
To describe the association between the RT-PCR result on the one hand and sex, primary tumor site, histologic subtype, Clark level, and tumor thickness on the other hand, odds ratios (ORs) with 95% CIs were calculated. Fisher's exact test was performed to analyze the association between PCR result and recurrences during follow-up.
Disease-specific survival time was defined as the time from first blood sampling to the time of death from melanoma. Recurrence-free survival time was defined as the time from first blood sampling to a recurrence or to tumor-related death. Survival data was analyzed by the Kaplan-Meier method.26 Survival curves were compared using the log-rank test. All P values presented are two-sided. Multiple Cox regression was applied with backward elimination (selection level of 0.05) for selection of prognostic factors. The prognostic factors used for variable selection were age, sex, site of primary tumor, stage, histologic subtype, Clark level, and tumor thickness (tumor thickness was forced into the model as a well-known important prognostic factor). These factors were considered as non-time-dependent variables because they were included in the model with their value at time of first blood sampling. Additionally, the RT-PCR result was considered a time-dependent prognostic factor because two to eight RT-PCR results per patient were available, and RT-PCR results changed over time.27
Because we used as survival time the time from first blood sampling instead of time from first diagnosis of the primary tumor, we examined the data for lead time bias. The hazard ratios for the covariate of time from first diagnosis of the primary tumor to first blood sampling in the Cox models for each of our survival times were close to 1 with large P values, thus showing no lead time bias in our data. However, we circumvented a length time bias (ie, slowly progressing lesions with less aggressive potential will preferably be detected) because disease progression was not palpated but detected by ultrasound independent of palpability, and therefore, size of lesion is not that important.28 The statistical analyses were performed with the Statistical Analysis System Version 8.2 (SAS Institute, Cary, NC) and StatXact-3 (Cytel, Cambridge, MA).
RESULTS
Tyrosinase RT-PCR Results
Overall, 107 (28.2%) of 380 blood samples were positive for tyrosinase RT-PCR. Positive results were obtained in 23.3% of stage II patients (60 of 258 samples) and 38.5% of stage III patients (47 of 122 samples). In Table 2, RT-PCR results are shown in relationship to sex, primary tumor site, histologic subtype, Clark level, and tumor thickness with respect to the clinical stage at time of first blood sampling. In both clinical stages, data showed a slight tendency towards higher proportions of positivity for tyrosinase RT-PCR for male compared with female patients, but 95% CIs were wide and included the value 1.0. Overall, a positive RT-PCR was observed in 24 (48%) of 50 male patients compared with 26 (43%) of 61 female patients.
Relationship of RT-PCR Result to Tumor Site, Histologic Subtype, Clark Level, and Tumor Thickness
No clear differences were observed in RT-PCR positivity for different primary tumor sites. ORs listed in Table 2 have wide 95% CIs, including the value of 1.0. Overall, positivity was observed in 25 (45%) of 55 patients with limb tumors compared with 25 (45%) of 56 patients with other tumor sites.
Histologic subtype results were different for stage II and III patients (Table 2). Overall, positivity was observed in 13 (36%) of 36 patients with superficial spreading melanoma (SSM) or lentigo maligna melanoma (LMM) compared with 24 (45%) of 53 patients with nodular melanoma (NM) and nine (53%) of 17 patients with acrolentiginous melanoma (ALM) or other histologic subtypes. Stage II patients with NM showed a higher probability of being tyrosinase positive than stage II patients with SSM or LMM (OR, 2.1). Although 17 (43%) of 40 patients with NM tumors had tyrosinase transcripts in their peripheral blood, only seven (26%) of 27 melanoma patients with SSM or LMM tumors had positive test results. The contrary could be seen for stage III patients. Proportions of positivity were at 67% for SSM and LMM tumors in stage III patients and at 54% for patients with NM tumors. Because of the small numbers, however, no noticeable differences could be seen.
A strong association was seen between Clark level and RT-PCR positivity in both clinical stages. Overall, a positive RT-PCR was observed in 37 (51%) of 72 patients with Clark level IV or V compared with five (19%) of 27 patients with Clark level II or III (OR, 4.7; 95% CI, 1.6 to 13.6).
In clinical stage II patients, data showed a strong effect towards higher proportions of positive test results for tumors thicker than 4 mm compared with tumors thinner than 4 mm; however, this effect was not seen in stage III patients. Overall, positivity occurred in 17 (71%) of 24 patients with tumors ≥ 4 mm compared with 27 (35%) of 78 patients with tumors less than 4 mm (OR, 4.6; 95% CI, 1.7 to 12.4).
Association of RT-PCR Result With Recurrence-Free Survival and Disease-Specific Survival
Recurrences were detected in 40 patients during the follow-up after the first blood sampling (22 of 78 stage II patients and 18 of 33 stage III patients). Among 40 patients presenting with recurrences, nine stage III patients had recurrences without progress, five stage II patients had progressed to stage III, and 26 patients (17 stage II patients and nine stage III patients) had progressed to stage IV during the follow-up period. In 18 (45%) of 40 patients with recurrence, at least one of the RT-PCRs before clinical detection of the recurrence was positive; whereas in 18 (25%) of 71 patients without recurrence, at least one RT-PCR result was positive (P = .034). Differences in positivity rates for patients with recurrence compared with patients without recurrence were more pronounced among stage III patients than among stage II patients (Table 3). Recurrence-free survival probabilities are shown in Figure 1. The 6-year recurrence-free survival probability in patients with tyrosinase RT-PCR that was always negative was higher than in patients with at least one positive RT-PCR (72% v 52%, respectively; P = .025). During the follow-up period, 20 patients died of melanoma (11 of 78 stage II patients and nine of 33 stage III patients). Two deaths other than disease-specific death occurred in our cohort. At least one of the RT-PCR tests was positive in 85% of patients who died (17 of 20 patients), whereas only 36% of patients alive (33 of 91 patients) were positive for tyrosinase RT-PCR (P < .001). Differences in positivity rates for patients with tumor-related death compared with patients who were alive were more pronounced among stage II patients than among stage III patients (Table 4). Disease-specific survival probabilities are shown in Figure 2. The 6-year survival probability in patients with tyrosinase RT-PCR that was always negative was higher than in patients with at least one positive RT-PCR (97% v 67%, respectively; P < .001).
Cox Proportional Hazards Regression
To further model the influence of RT-PCR testing on the prognosis (ie, depending on which RT-PCR result was derived at which point in time), RT-PCR testing was used as a time-dependent prognostic factor in a proportional hazards model. Therefore, a RT-PCR result was considered correct as long as it could not be replaced by the result of the subsequent RT-PCR. This regression analysis for all 111 melanoma patients showed that RT-PCR was a prognostically relevant factor for disease-specific survival, with a hazard ratio of 15.8 (95% CI, 5.2 to 49.0; P < .001), which means that the risk of dying became 15.8 times greater when RT-PCR turned to be positive during follow-up. The regression analysis for all 111 melanoma patients also showed that RT-PCR was a prognostically relevant factor for recurrence-free survival, with a hazard ratio of 4.5 (95% CI, 2.4 to 8.5; P < .001).
In a multiple Cox regression analysis, sex, age at time of first blood sampling (in years), site of primary tumor (limb v trunk, neck, and scalp), stage (III v II), Clark level (IV and V v II and III), histologic subtype (NM v SSM and LMM v ALM and others), and, as a time-dependent covariate, RT-PCR result (positive v negative) were considered for variable selection using backward elimination, whereas tumor thickness (in millimeters) was forced into the model as a well-known important prognostic factor. All together, 96 patients were included in the Cox regression analyses because five patients with missing values for histologic type and 10 patients with different histologic types others than NM, SSM, ALM, or LMM were excluded. For disease-specific survival, RT-PCR result and age at time of first blood sampling were selected as important prognostic factors (Table 5). For recurrence-free survival, RT-PCR result, stage, and Clark level were selected as important prognostic factors (Table 6). Patients with stage III tumors and patients with Clark level 4 or 5 had an increased risk of a recurrence. Most striking was that both the probability of dying from the primary tumor and the probability of a recurrence increased dramatically if RT-PCR turned to be positive during follow-up, even though these effects were adjusted for tumor thickness.
DISCUSSION
The present study extends the previous analysis of our group23 in terms of patient numbers and follow-up duration. It also differs from the previous reports in the literature because it systematically addresses the potential informative value of serial blood tests. Serial blood tests, until recently, were only seen as a pure reflection of the stage of disease at diagnosis, but now, they have been proposed to define the prognostic relevance of a change in circulating melanoma cell status dynamically, and not.37
We report here on the clinical relevance of RT-PCR-based detection of circulating tumor cells considering 380 serial tyrosinase RT-PCR analyses in 78 stage II and 33 stage III patients with a median follow-up of 6.3 years. In a Cox proportional hazards model with RT-PCR result as a time-dependent covariate, patients becoming RT-PCR positive during follow-up had an increased risk of dying from melanoma, with a hazard ratio of 12.6 (95% CI, 3.4 to 46.3; P < .001), and had an increased risk for a recurrence, with a hazard ratio of 4.0 (95% CI, 1.9 to 8.3; P < .001). Both hazard ratios are adjusted for tumor thickness.
Tumor thickness was forced into the Cox models because it is a known important prognostic factor. However, in both models, it had a P > .05. Therefore, tyrosinase results seem to be more important for prognosis than baseline measurements of tumor thickness; however, this might be a result of serial measurement of RT-PCR during follow-up and consideration of RT-PCR results at the particular time of measurement.
At the time when our study was initiated, none of the previous reports by other groups had investigated the prognostic value of serial RT-PCR assays as the main focus of their studies, although four studies did analyze serial data. The results reported in our study are in line with both the findings by Mellado et al,36 which revealed an association between RT-PCR assay results during adjuvant interferon treatment and both overall survival and disease-free survival in 120 melanoma patients across all stages, and the preliminary results reported by Curry et al.19 But our results contrast with the reports by Hanekom et al34 and Brownbridge et al,35 who did not find a considerable RT-PCR conversion rate in melanoma patients before clinical disease progression. The reasons for the discrepancies among the reports are difficult to assess; they may be methodologic in nature, related to the patient populations, or related to the various follow-up intervals. Recently, Szenajch et al24 showed that patients with a more advanced disease more frequently test positive for tyrosinase, which is in line with our results. Osella-Abate et al38 conducted their trial by using fixed intervals of blood samples, independent of stage or the urgency of follow-up visits, and performing baseline blood samples (within 3 to 6 weeks after surgical treatment of regional nodal metastases in all patients). This was not performed in our subset because we wanted to evaluate minimal residual disease in patients who were already in follow-up and not starting from the beginning of diagnosis. Osella-Abate et al, 38 like us, found an association between positive tyrosinase results and disease recurrence during follow-up, and thus, they encouraged the clinical use of serial tyrosinase testing because of reliability and reproducibility.38 However, they did not correctly model the time dependence of the RT-PCR results, as we did in our Cox model. Our report has an advantage over previous studies because we only included patients with stage II and III disease in this study. We also had a major advantage because of the structure of our clinic, which allowed a complete and long-term follow-up (Table 7). In our clinic, all data, including the follow-up information, are available to the principal investigator of the study (C.V.). Because of this, we did not face the situation of other departments in which patients are lost to oncology or surgical services and in which it is difficult to obtain complete data. Another aspect of the study reported here is the finding that the serial RT-PCR assay result is an important prognostic factor for disease-specific survival. Therefore, RT-PCR results may be useful for patient stratification in clinical trials. Thus, stage II and III patients repeatedly presenting with a positive tyrosinase RT-PCR result from peripheral blood by serial testing can be considered as patients at high risk for relapse, and adjuvant clinical trials can be carried out in this patient subset. Further studies are necessary to address whether the RT-PCR assay adds information to serologic assays (S100 protein and melanoma inhibitory factor MIA), which also have been shown to predict survival of melanoma patients in several studies.39-43 The relationship of RT-PCR analysis in peripheral blood and histologic, immunohistologic, and PCR analysis of sentinel lymph nodes has not been systematically investigated to date. Theoretically, the lymph node analysis should yield the relevant information on regional spread, and the RT-PCR analysis of peripheral blood should predict hematogenous spread. The ultimate goal of all prognostic factors (including serologic and mRNA tumor markers), however, is to serve as a basis for treatment decisions. At present, there exists no promising chemotherapy or chemoimmunotherapy for melanoma, and in addition, data concerning adjuvant immunotherapy with interferon do not indicate improvement of disease-free or even overall survival.44 With the currently available data, however, there is no rationale for treatment decisions based on RT-PCR assay results or serologic markers because large prospective trials are necessary (and currently underway) to investigate whether a specific treatment modality may have different effects in patient cohorts with or without PCR evidence for circulating melanoma cells. One interesting aspect could be the recent progress in the field of melanoma vaccines, which are currently entering phase III testing in stage III melanoma patients. Apparently, multiple tyrosine RT-PCRs can be useful for molecular melanoma staging and, hence, for prediction of recurrence or development of metastases.
Authors' Disclosures of Potential Conflicts of Interest
Acknowledgment
We thank Heidi Hainzl for her continuous support and helpful discussions.
NOTES
Supported by grant No. 70-2459-Ke2 from Deutsche Krebshilfe.
Both C.V. and M.K. contributed equally to this work.
Presented in part at the Oral Melanoma Session of the 38th Annual Meeting of the American Society of Clinical Oncology in Orlando, FL, May 18-21, 2002.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
1. Balch CM, Buzaid AC, Atkins MB, et al: A new American Joint Committee on Cancer staging system for cutaneous melanoma. Cancer 88:1484-1491, 2000
2. Balch CM, Soong SJ, Gershenwald JE, et al: Prognostic factors analysis of 17,600 melanoma patients: Validation of the American Joint Committee on Cancer melanoma staging system. J Clin Oncol 19:3622-3634, 2001
3. Balch CM, Buzaid AC, Soong SJ, et al: Final version of the American Joint Committee on Cancer staging system for cutaneous melanoma. J Clin Oncol 19:3635-3648, 2001
4. Smith B, Selby P, Southgate J, et al: Detection of melanoma cells in peripheral blood by means of reverse transcriptase and polymerase chain reaction. Lancet 338:1227-1229, 1991
5. Brossart P, Keilholz U, Willhauck M, et al: Hematogenous spread of malignant melanoma cells in different stages of disease. J Invest Dermatol 101:887-889, 1993
6. Brossart P, Keilholz U, Scheibenbogen C, et al: Detection of residual tumor cells in patients with malignant melanoma responding to immunotherapy. J Immunother 15:38-41, 1994
7. Foss AJ, Guille MJ, Occleston NL, et al: The detection of melanoma cells in peripheral blood by reverse transcription-polymerase chain reaction. Br J Cancer 72:155-159, 1995
8. Pittman K, Burchill S, Smith B, et al: Reverse transcriptase-polymerase chain reaction for expression of tyrosinase to identify malignant melanoma cells in peripheral blood. Ann Oncol 7:297-301, 1996
9. Glaser R, Rass K, Seiter S, et al: Detection of circulating melanoma cells by specific amplification of tyrosinase complementary DNA is not a reliable tumor marker in melanoma patients: A clinical two-center study. J Clin Oncol 15:2818-2825, 1997
10. Jung FA, Buzaid AC, Ross MI, et al: Evaluation of tyrosinase mRNA as a tumor marker in the blood of melanoma patients. J Clin Oncol 15:2826-2831, 1997
11. Farthmann B, Eberle J, Krasagakis K, et al: RT-PCR for tyrosinase-mRNA-positive cells in peripheral blood: Evaluation strategy and correlation with known prognostic markers in 123 melanoma patients. J Invest Dermatol 110:263-267, 1998
12. Ghossein RA, Coit D, Brennan M, et al: Prognostic significance of peripheral blood and bone marrow tyrosinase messenger RNA in malignant melanoma. Clin Cancer Res 4:419-428, 1998
13. Hoon DS, Wang Y, Dale PS, et al: Detection of occult melanoma cells in blood with a multiple-marker polymerase chain reaction assay. J Clin Oncol 13:2109-2116, 1995
14. Kunter U, Buer J, Probst M, et al: Peripheral blood tyrosinase messenger RNA detection and survival in malignant melanoma. J Natl Cancer Inst 88:590-594, 1996
15. Mellado B, Colomer D, Castel T, et al: Detection of circulating neoplastic cells by reverse-transcriptase polymerase chain reaction in malignant melanoma: Association with clinical stage and prognosis. J Clin Oncol 14:2091-2097, 1996
16. Ghossein RA, Carusone L, Bhattacharya S: Molecular detection of micrometastases and circulating tumor cells in melanoma prostatic and breast carcinomas. In Vivo 14:237-250, 2000
17. Hoon DS, Bostick P, Kuo C, et al: Molecular markers in blood as surrogate prognostic indicators of melanoma recurrence. Cancer Res 60:2253-2257, 2000
18. Ghossein RA, Bhattacharya S, Coit DG: Reverse transcriptase polymerase chain reaction (RT-PCR) detection of melanoma-related transcripts in the peripheral blood and bone marrow of patients with malignant melanoma. What have we learned Recent Results Cancer Res 158:63-77, 2001
19. Curry BJ, Myers K, Hersey P: Polymerase chain reaction detection of melanoma cells in the circulation: Relation to clinical stage, surgical treatment, and recurrence from melanoma. J Clin Oncol 16:1760-1769, 1998
20. Buzaid AC, Ross MI, Balch CM, et al: Critical analysis of the current American Joint Committee on Cancer staging system for cutaneous melanoma and proposal of a new staging system. J Clin Oncol 15:1039-1051, 1997
21. Voit C, Schoengen A, Weber L, et al: Identification of melanoma metastases by tyrosinase-reverse transcription-polymerase chain reaction of fine needle aspirates. J Am Acad Dermatol 39:1030-1032, 1998
22. Voit C, Schoengen A, Schwurzer M, et al: Detection of regional melanoma metastases by ultrasound B-scan, cytology or tyrosinase RT-PCR of fine-needle aspirates. Br J Cancer 80:1672-1677, 1999
23. Proebstle TM, Jiang W, Hogel J, et al: Correlation of positive RT-PCR for tyrosinase in peripheral blood of malignant melanoma patients with clinical stage, survival and other risk factors. Br J Cancer 82:118-123, 2000
24. Szenajch J, Jasinski B, Synowiec A, et al: Prognostic value of multiple reverse transcription-PCR tyrosinase testing for circulating neoplastic cells in malignant melanoma. Clin Chem 49:1450-1457, 2003
25. Keilholz U, Willhauck M, Rimoldi D, et al: Reliability of reverse transcription-polymerase chain reaction (RT-PCR)-based assays for the detection of circulating tumour cells: A quality-assurance initiative of the EORTC Melanoma Cooperative Group. Eur J Cancer 34:750-753, 1998
26. Kaplan EL, Meier P: Nonparametric estimation from incomplete observations. J Am Stat Assoc 53:457-481, 1958
27. Lawless JF: Statistical Models and Methods for Lifetime Data. New York, NY, Wiley, 1982
28. Autier P, Coebergh JW, Boniol M, et al: Management of melanoma patients: Benefit of intense follow-up schedule is not demonstrated. J Clin Oncol 21:3707-3708, 2003
29. Reinhold U, Ludtke-Handjery HC, Schnautz S, et al: The analysis of tyrosinase-specific mRNA in blood samples of melanoma patients by RT-PCR is not a useful test for metastatic tumor progression. J Invest Dermatol 108:166-169, 1997
30. Tessier MH, Denis MG, Lustenberger P, et al: Detection of circulating neoplastic cells by reverse transcriptase and polymerase chain reaction in melanoma. Ann Dermatol Venereol 124:607-611, 1997
31. Schittek B, Blaheta HJ, Florchinger G, et al: Increased sensitivity for the detection of malignant melanoma cells in peripheral blood using an improved protocol for reverse transcription-polymerase chain reaction. Br J Dermatol 141:37-43, 1999
32. Schittek B, Bodingbauer Y, Ellwanger U, et al: Amplification of MelanA messenger RNA in addition to tyrosinase increases sensitivity of melanoma cell detection in peripheral blood and is associated with the clinical stage and prognosis of malignant melanoma. Br J Dermatol 141:30-36, 1999
33. Palmieri G, Strazzullo M, Ascierto PA, et al: Polymerase chain reaction-based detection of circulating melanoma cells as an effective marker of tumor progression: Melanoma Cooperative Group. J Clin Oncol 17:304-311, 1999
34. Hanekom GS, Stubbings HM, Johnson CA, et al: The detection of circulating melanoma cells correlates with tumour thickness and ulceration but is not predictive of metastasis for patients with primary melanoma. Melanoma Res 9:465-473, 1999
35. Brownbridge GG, Gold J, Edward M, et al: Evaluation of the use of tyrosinase-specific and melanA/MART-1-specific reverse transcriptase-coupled-polymerase chain reaction to detect melanoma cells in peripheral blood samples from 299 patients with malignant melanoma. Br J Dermatol 144:279-287, 2001
36. Mellado B, Del Carmen Vela M, Colomer D, et al: Tyrosinase mRNA in blood of patients with melanoma treated with adjuvant interferon. J Clin Oncol 20:4032-4039, 2002
37. Dutcher JP: The prognostic role of detection of circulating melanoma cells in the blood. J Clin Oncol 21:757-759, 2003
38. Osella-Abate S, Savoia P, Quaglino P, et al: Tyrosinase expression in the peripheral blood of stage III melanoma patients is associated with a poor prognosis: A clinical follow-up study of 110 patients. Br J Cancer 89:1457-1462, 2003
39. Bosserhoff AK, Kaufmann M, Kaluza B, et al: Melanoma-inhibiting activity, a novel serum marker for progression of malignant melanoma. Cancer Res 57:3149-3153, 1997
40. Stahlecker J, Gauger A, Bosserhoff A, et al: MIA as a reliable tumor marker in the serum of patients with malignant melanoma. Anticancer Res 20:5041-5044, 2000
41. Henze G, Dummer R, Joller-Jemelka HI, et al: Serum S100-a marker for disease monitoring in metastatic melanoma. Dermatology 194:208-212, 1997
42. Juergensen A, Holzapfel U, Hein R, et al: Comparison of two prognostic markers for malignant melanoma: MIA and S100 beta. Tumour Biol 22:54-58, 2001
43. Djukanovic D, Hofmann U, Sucker A, et al: Comparison of S100 protein and MIA protein as serum marker for malignant melanoma. Anticancer Res 20:2203-2207, 2000
44. Hancock BW, Wheatley K, Harris S, et al: Adjuvant interferon in high-risk melanoma: The AIM HIGH Study—United Kingdom Coordinating Committee on Cancer Research randomized study of adjuvant low-dose extended-duration interferon alfa-2a in high-risk resected malignant melanoma. J Clin Oncol 22:53-61, 2004(Christiane Voit, Martina )
Department of Medicine III, Campus Benjamin Franklin, Charite Berlin, Berlin
Department of Biometry and Medical Documentation and Department of Dermatology, University of Ulm, Ulm
Department of Dermatology, University of Mainz, Mainz, Germany
Department of Radiology, Memorial Sloan-Kettering Cancer Center, New York, NY
ABSTRACT
PATIENTS AND METHODS: During routine follow-up of American Joint Committee on Cancer stage II and III melanoma patients, serial testing for tyrosinase transcripts in peripheral blood was performed by RT-PCR. The PCR results were compared with the clinical data collected during the follow-up.
RESULTS: Over a period of 3 years, 111 patients (78 stage II and 33 stage III patients) were enrolled, and tyrosinase determinations were carried out. The 6-year disease-specific survival probability was 97% for patients always showing negative RT-PCR results and 67% for patients who tested positive at least once. In a Cox proportional hazards model, the prognostic value of sex, age, site of primary tumor, histologic subtype, stage, Breslow's tumor thickness, Clark level, and the time-dependent variable PCR result was assessed. Patients with a positive RT-PCR test had a distinctly higher risk of dying from melanoma, with a hazard ratio of 12.6 (95% CI, 3.4 to 46.3; P < .001).
CONCLUSION: Our study shows a strong association between PCR and disease-specific survival time. Detection of tyrosinase mRNA in peripheral blood may be of similar importance for the clinical course of melanoma as the detection of micrometastatic disease in the sentinel lymph node. Whether a combination of these two factors leads to a better definition of the prognosis of melanoma patients is under investigation in current studies.
INTRODUCTION
Since the original description of a reverse transcriptase polymerase chain reaction (RT-PCR) method for detection of tyrosinase transcripts in peripheral blood,4 controversial data have been reported in terms of sensitivity of this method and its clinical relevance.5-12 The use of a multiple marker assay was introduced to improve the sensitivity.13 Recurrence-free and disease-specific survival depending on RT-PCR test results have been rarely addressed in homogenous patient cohorts.12,14-16 Therefore, large prospective studies with long-term follow-up in stage II and III melanoma patients are necessary to determine the prognostic value of these assays.16-19 We report on serial tyrosinase RT-PCR measurement in 111 consecutive stage II and III melanoma patients during routine follow-up examination to investigate the prognostic relevance of this assay.
PATIENTS AND METHODS
All blood sampling was performed concomitantly to tumor staging procedures as patients presented for routine follow-up examinations at scheduled intervals. Patients with primary tumors with a tumor thickness of up to 1.5 mm presented every 6 months, and patients with primary tumors thicker than 1.5 mm and stage III patients presented every 3 months; therefore, the time intervals between consecutive tests performed on patients differed, which is the normal course in follow-up. In an earlier study, we reported our experience with tyrosinase RT-PCR from peripheral blood,23 but the patients of the study reported here are completely different and no subset of the former study.
Blood Sample Processing, RNA Preparation, and Reverse Transcription
Samples of 2 x 2.7 mL EDTA blood were taken and processed within 2 hours. For RNA preparation, both samples were pooled and treated with erythrocyte lysis buffer. Extraction of total RNA was performed by means of Qia shredder columns for purification and Rneasy TM total RNA kit (Qiagen GMBH, Hilden, Germany). All precaution was taken to ensure integrity and purity of RNA. RNA concentration and purity were determined by UV spectrophotometry. RNA was quantified by UV spectrophotometry at 260 and 280 nm and stored at –80°C. Total RNA (1.5 μg) was transcribed by means of 50 ng of random hexamers, 10 mmol/L of deoxynucleotide triphosphate mix, and 200 U of superscript II RT (GIBCO BRL, Inc, Grand Island, NY).
Nested Tyrosinase PCR
Primers for a nested tyrosinase PCR were used as previously described.4 The outer primers of HTYR1 (TTGGCAGATTGTCTGTAGCC) and HTYR2 (AGGCATTGTGCATGCTGCTT) amplify a product of 284 base pairs (bp), whereas the inner primers of HTYR3 (GTCTTTATGCAATGGAACGC) and HTYR4 (GCTATCCCAGTAAGTGGACT) amplify a product of 207 bp. Thirty cycles were run using a schedule described elsewhere.22,23 To avoid contamination, the reaction mixtures were prepared in a fume hood. The second amplified product was analyzed on a 3% agarose gel followed by ethidium bromide staining (80 minutes, 80 V). A 100-bp ladder (GIBCO) was used as standard. PCR samples free of cDNA served as negative controls, and PCR samples with cDNA of the SK-mel-28 melanoma cell line served as the positive control.
Quality Control Experiments
To test the sensitivity, 10-fold serial dilutions of SK-mel-28 melanoma cells (0 to 105 melanoma cells/10 mL blood) in peripheral blood from a healthy donor (spiking experiments) were performed. One SK-mel-28 melanoma cell in a background of 106 peripheral-blood mononuclear cells could be detected. This can also be mimicked by aspiration of a single cell by means of a heat-drawn glass capillary and by performing a single-cell PCR as previously described.24
We had participated in an interlaboratory trial of the European Organisation for Research and Treatment of Cancer for interlaboratory quality assurance.25 Several steps have been taken to detect any cross-over contamination. Before cDNA synthesis, RNAs were purified via silica-gel columns and treated with DNAse. cDNA has always been transcribed in duplicate, with and without superscript enzyme. Each PCR run was controlled by a second run starting from RNA preparation to cDNA synthesis and PCR procedure to confirm results. PCR from RNA as template (ie, PCR without reverse transcription) was carried out to exclude amplification of genomic cDNA. Negative and positive controls were run on each agarose gel in addition to patient samples. If the results of patient samples were discordant, the whole procedure starting from RNA preparation was rerun twice until two concordant results were seen. To ensure that RNA in the tyrosinase gene-negative samples had not been degraded, a parallel PCR run for the detection of glyceraldehyde-3-phosphatase dehydrogenase (housekeeping enzyme) using an equal amount of RNA was performed. The specificity of PCR products was examined by sequencing in all positive samples. cDNA was amplified by using HTYR3/4 with previously marked universal primer M13 (forward and reverse). Sequencing was performed via dye primer sequencing (Perkin Elmer, Norwalk, CT). The analysis of each specimen was carried out on a 373 DNA sequencer (ABI; Applied Biosystems, Foster City, CA). In 20 blood samples of healthy donors and of patients with nonmelanoma cancers, tyrosinase transcripts were absent.
Statistics
To describe the association between the RT-PCR result on the one hand and sex, primary tumor site, histologic subtype, Clark level, and tumor thickness on the other hand, odds ratios (ORs) with 95% CIs were calculated. Fisher's exact test was performed to analyze the association between PCR result and recurrences during follow-up.
Disease-specific survival time was defined as the time from first blood sampling to the time of death from melanoma. Recurrence-free survival time was defined as the time from first blood sampling to a recurrence or to tumor-related death. Survival data was analyzed by the Kaplan-Meier method.26 Survival curves were compared using the log-rank test. All P values presented are two-sided. Multiple Cox regression was applied with backward elimination (selection level of 0.05) for selection of prognostic factors. The prognostic factors used for variable selection were age, sex, site of primary tumor, stage, histologic subtype, Clark level, and tumor thickness (tumor thickness was forced into the model as a well-known important prognostic factor). These factors were considered as non-time-dependent variables because they were included in the model with their value at time of first blood sampling. Additionally, the RT-PCR result was considered a time-dependent prognostic factor because two to eight RT-PCR results per patient were available, and RT-PCR results changed over time.27
Because we used as survival time the time from first blood sampling instead of time from first diagnosis of the primary tumor, we examined the data for lead time bias. The hazard ratios for the covariate of time from first diagnosis of the primary tumor to first blood sampling in the Cox models for each of our survival times were close to 1 with large P values, thus showing no lead time bias in our data. However, we circumvented a length time bias (ie, slowly progressing lesions with less aggressive potential will preferably be detected) because disease progression was not palpated but detected by ultrasound independent of palpability, and therefore, size of lesion is not that important.28 The statistical analyses were performed with the Statistical Analysis System Version 8.2 (SAS Institute, Cary, NC) and StatXact-3 (Cytel, Cambridge, MA).
RESULTS
Tyrosinase RT-PCR Results
Overall, 107 (28.2%) of 380 blood samples were positive for tyrosinase RT-PCR. Positive results were obtained in 23.3% of stage II patients (60 of 258 samples) and 38.5% of stage III patients (47 of 122 samples). In Table 2, RT-PCR results are shown in relationship to sex, primary tumor site, histologic subtype, Clark level, and tumor thickness with respect to the clinical stage at time of first blood sampling. In both clinical stages, data showed a slight tendency towards higher proportions of positivity for tyrosinase RT-PCR for male compared with female patients, but 95% CIs were wide and included the value 1.0. Overall, a positive RT-PCR was observed in 24 (48%) of 50 male patients compared with 26 (43%) of 61 female patients.
Relationship of RT-PCR Result to Tumor Site, Histologic Subtype, Clark Level, and Tumor Thickness
No clear differences were observed in RT-PCR positivity for different primary tumor sites. ORs listed in Table 2 have wide 95% CIs, including the value of 1.0. Overall, positivity was observed in 25 (45%) of 55 patients with limb tumors compared with 25 (45%) of 56 patients with other tumor sites.
Histologic subtype results were different for stage II and III patients (Table 2). Overall, positivity was observed in 13 (36%) of 36 patients with superficial spreading melanoma (SSM) or lentigo maligna melanoma (LMM) compared with 24 (45%) of 53 patients with nodular melanoma (NM) and nine (53%) of 17 patients with acrolentiginous melanoma (ALM) or other histologic subtypes. Stage II patients with NM showed a higher probability of being tyrosinase positive than stage II patients with SSM or LMM (OR, 2.1). Although 17 (43%) of 40 patients with NM tumors had tyrosinase transcripts in their peripheral blood, only seven (26%) of 27 melanoma patients with SSM or LMM tumors had positive test results. The contrary could be seen for stage III patients. Proportions of positivity were at 67% for SSM and LMM tumors in stage III patients and at 54% for patients with NM tumors. Because of the small numbers, however, no noticeable differences could be seen.
A strong association was seen between Clark level and RT-PCR positivity in both clinical stages. Overall, a positive RT-PCR was observed in 37 (51%) of 72 patients with Clark level IV or V compared with five (19%) of 27 patients with Clark level II or III (OR, 4.7; 95% CI, 1.6 to 13.6).
In clinical stage II patients, data showed a strong effect towards higher proportions of positive test results for tumors thicker than 4 mm compared with tumors thinner than 4 mm; however, this effect was not seen in stage III patients. Overall, positivity occurred in 17 (71%) of 24 patients with tumors ≥ 4 mm compared with 27 (35%) of 78 patients with tumors less than 4 mm (OR, 4.6; 95% CI, 1.7 to 12.4).
Association of RT-PCR Result With Recurrence-Free Survival and Disease-Specific Survival
Recurrences were detected in 40 patients during the follow-up after the first blood sampling (22 of 78 stage II patients and 18 of 33 stage III patients). Among 40 patients presenting with recurrences, nine stage III patients had recurrences without progress, five stage II patients had progressed to stage III, and 26 patients (17 stage II patients and nine stage III patients) had progressed to stage IV during the follow-up period. In 18 (45%) of 40 patients with recurrence, at least one of the RT-PCRs before clinical detection of the recurrence was positive; whereas in 18 (25%) of 71 patients without recurrence, at least one RT-PCR result was positive (P = .034). Differences in positivity rates for patients with recurrence compared with patients without recurrence were more pronounced among stage III patients than among stage II patients (Table 3). Recurrence-free survival probabilities are shown in Figure 1. The 6-year recurrence-free survival probability in patients with tyrosinase RT-PCR that was always negative was higher than in patients with at least one positive RT-PCR (72% v 52%, respectively; P = .025). During the follow-up period, 20 patients died of melanoma (11 of 78 stage II patients and nine of 33 stage III patients). Two deaths other than disease-specific death occurred in our cohort. At least one of the RT-PCR tests was positive in 85% of patients who died (17 of 20 patients), whereas only 36% of patients alive (33 of 91 patients) were positive for tyrosinase RT-PCR (P < .001). Differences in positivity rates for patients with tumor-related death compared with patients who were alive were more pronounced among stage II patients than among stage III patients (Table 4). Disease-specific survival probabilities are shown in Figure 2. The 6-year survival probability in patients with tyrosinase RT-PCR that was always negative was higher than in patients with at least one positive RT-PCR (97% v 67%, respectively; P < .001).
Cox Proportional Hazards Regression
To further model the influence of RT-PCR testing on the prognosis (ie, depending on which RT-PCR result was derived at which point in time), RT-PCR testing was used as a time-dependent prognostic factor in a proportional hazards model. Therefore, a RT-PCR result was considered correct as long as it could not be replaced by the result of the subsequent RT-PCR. This regression analysis for all 111 melanoma patients showed that RT-PCR was a prognostically relevant factor for disease-specific survival, with a hazard ratio of 15.8 (95% CI, 5.2 to 49.0; P < .001), which means that the risk of dying became 15.8 times greater when RT-PCR turned to be positive during follow-up. The regression analysis for all 111 melanoma patients also showed that RT-PCR was a prognostically relevant factor for recurrence-free survival, with a hazard ratio of 4.5 (95% CI, 2.4 to 8.5; P < .001).
In a multiple Cox regression analysis, sex, age at time of first blood sampling (in years), site of primary tumor (limb v trunk, neck, and scalp), stage (III v II), Clark level (IV and V v II and III), histologic subtype (NM v SSM and LMM v ALM and others), and, as a time-dependent covariate, RT-PCR result (positive v negative) were considered for variable selection using backward elimination, whereas tumor thickness (in millimeters) was forced into the model as a well-known important prognostic factor. All together, 96 patients were included in the Cox regression analyses because five patients with missing values for histologic type and 10 patients with different histologic types others than NM, SSM, ALM, or LMM were excluded. For disease-specific survival, RT-PCR result and age at time of first blood sampling were selected as important prognostic factors (Table 5). For recurrence-free survival, RT-PCR result, stage, and Clark level were selected as important prognostic factors (Table 6). Patients with stage III tumors and patients with Clark level 4 or 5 had an increased risk of a recurrence. Most striking was that both the probability of dying from the primary tumor and the probability of a recurrence increased dramatically if RT-PCR turned to be positive during follow-up, even though these effects were adjusted for tumor thickness.
DISCUSSION
The present study extends the previous analysis of our group23 in terms of patient numbers and follow-up duration. It also differs from the previous reports in the literature because it systematically addresses the potential informative value of serial blood tests. Serial blood tests, until recently, were only seen as a pure reflection of the stage of disease at diagnosis, but now, they have been proposed to define the prognostic relevance of a change in circulating melanoma cell status dynamically, and not.37
We report here on the clinical relevance of RT-PCR-based detection of circulating tumor cells considering 380 serial tyrosinase RT-PCR analyses in 78 stage II and 33 stage III patients with a median follow-up of 6.3 years. In a Cox proportional hazards model with RT-PCR result as a time-dependent covariate, patients becoming RT-PCR positive during follow-up had an increased risk of dying from melanoma, with a hazard ratio of 12.6 (95% CI, 3.4 to 46.3; P < .001), and had an increased risk for a recurrence, with a hazard ratio of 4.0 (95% CI, 1.9 to 8.3; P < .001). Both hazard ratios are adjusted for tumor thickness.
Tumor thickness was forced into the Cox models because it is a known important prognostic factor. However, in both models, it had a P > .05. Therefore, tyrosinase results seem to be more important for prognosis than baseline measurements of tumor thickness; however, this might be a result of serial measurement of RT-PCR during follow-up and consideration of RT-PCR results at the particular time of measurement.
At the time when our study was initiated, none of the previous reports by other groups had investigated the prognostic value of serial RT-PCR assays as the main focus of their studies, although four studies did analyze serial data. The results reported in our study are in line with both the findings by Mellado et al,36 which revealed an association between RT-PCR assay results during adjuvant interferon treatment and both overall survival and disease-free survival in 120 melanoma patients across all stages, and the preliminary results reported by Curry et al.19 But our results contrast with the reports by Hanekom et al34 and Brownbridge et al,35 who did not find a considerable RT-PCR conversion rate in melanoma patients before clinical disease progression. The reasons for the discrepancies among the reports are difficult to assess; they may be methodologic in nature, related to the patient populations, or related to the various follow-up intervals. Recently, Szenajch et al24 showed that patients with a more advanced disease more frequently test positive for tyrosinase, which is in line with our results. Osella-Abate et al38 conducted their trial by using fixed intervals of blood samples, independent of stage or the urgency of follow-up visits, and performing baseline blood samples (within 3 to 6 weeks after surgical treatment of regional nodal metastases in all patients). This was not performed in our subset because we wanted to evaluate minimal residual disease in patients who were already in follow-up and not starting from the beginning of diagnosis. Osella-Abate et al, 38 like us, found an association between positive tyrosinase results and disease recurrence during follow-up, and thus, they encouraged the clinical use of serial tyrosinase testing because of reliability and reproducibility.38 However, they did not correctly model the time dependence of the RT-PCR results, as we did in our Cox model. Our report has an advantage over previous studies because we only included patients with stage II and III disease in this study. We also had a major advantage because of the structure of our clinic, which allowed a complete and long-term follow-up (Table 7). In our clinic, all data, including the follow-up information, are available to the principal investigator of the study (C.V.). Because of this, we did not face the situation of other departments in which patients are lost to oncology or surgical services and in which it is difficult to obtain complete data. Another aspect of the study reported here is the finding that the serial RT-PCR assay result is an important prognostic factor for disease-specific survival. Therefore, RT-PCR results may be useful for patient stratification in clinical trials. Thus, stage II and III patients repeatedly presenting with a positive tyrosinase RT-PCR result from peripheral blood by serial testing can be considered as patients at high risk for relapse, and adjuvant clinical trials can be carried out in this patient subset. Further studies are necessary to address whether the RT-PCR assay adds information to serologic assays (S100 protein and melanoma inhibitory factor MIA), which also have been shown to predict survival of melanoma patients in several studies.39-43 The relationship of RT-PCR analysis in peripheral blood and histologic, immunohistologic, and PCR analysis of sentinel lymph nodes has not been systematically investigated to date. Theoretically, the lymph node analysis should yield the relevant information on regional spread, and the RT-PCR analysis of peripheral blood should predict hematogenous spread. The ultimate goal of all prognostic factors (including serologic and mRNA tumor markers), however, is to serve as a basis for treatment decisions. At present, there exists no promising chemotherapy or chemoimmunotherapy for melanoma, and in addition, data concerning adjuvant immunotherapy with interferon do not indicate improvement of disease-free or even overall survival.44 With the currently available data, however, there is no rationale for treatment decisions based on RT-PCR assay results or serologic markers because large prospective trials are necessary (and currently underway) to investigate whether a specific treatment modality may have different effects in patient cohorts with or without PCR evidence for circulating melanoma cells. One interesting aspect could be the recent progress in the field of melanoma vaccines, which are currently entering phase III testing in stage III melanoma patients. Apparently, multiple tyrosine RT-PCRs can be useful for molecular melanoma staging and, hence, for prediction of recurrence or development of metastases.
Authors' Disclosures of Potential Conflicts of Interest
Acknowledgment
We thank Heidi Hainzl for her continuous support and helpful discussions.
NOTES
Supported by grant No. 70-2459-Ke2 from Deutsche Krebshilfe.
Both C.V. and M.K. contributed equally to this work.
Presented in part at the Oral Melanoma Session of the 38th Annual Meeting of the American Society of Clinical Oncology in Orlando, FL, May 18-21, 2002.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
1. Balch CM, Buzaid AC, Atkins MB, et al: A new American Joint Committee on Cancer staging system for cutaneous melanoma. Cancer 88:1484-1491, 2000
2. Balch CM, Soong SJ, Gershenwald JE, et al: Prognostic factors analysis of 17,600 melanoma patients: Validation of the American Joint Committee on Cancer melanoma staging system. J Clin Oncol 19:3622-3634, 2001
3. Balch CM, Buzaid AC, Soong SJ, et al: Final version of the American Joint Committee on Cancer staging system for cutaneous melanoma. J Clin Oncol 19:3635-3648, 2001
4. Smith B, Selby P, Southgate J, et al: Detection of melanoma cells in peripheral blood by means of reverse transcriptase and polymerase chain reaction. Lancet 338:1227-1229, 1991
5. Brossart P, Keilholz U, Willhauck M, et al: Hematogenous spread of malignant melanoma cells in different stages of disease. J Invest Dermatol 101:887-889, 1993
6. Brossart P, Keilholz U, Scheibenbogen C, et al: Detection of residual tumor cells in patients with malignant melanoma responding to immunotherapy. J Immunother 15:38-41, 1994
7. Foss AJ, Guille MJ, Occleston NL, et al: The detection of melanoma cells in peripheral blood by reverse transcription-polymerase chain reaction. Br J Cancer 72:155-159, 1995
8. Pittman K, Burchill S, Smith B, et al: Reverse transcriptase-polymerase chain reaction for expression of tyrosinase to identify malignant melanoma cells in peripheral blood. Ann Oncol 7:297-301, 1996
9. Glaser R, Rass K, Seiter S, et al: Detection of circulating melanoma cells by specific amplification of tyrosinase complementary DNA is not a reliable tumor marker in melanoma patients: A clinical two-center study. J Clin Oncol 15:2818-2825, 1997
10. Jung FA, Buzaid AC, Ross MI, et al: Evaluation of tyrosinase mRNA as a tumor marker in the blood of melanoma patients. J Clin Oncol 15:2826-2831, 1997
11. Farthmann B, Eberle J, Krasagakis K, et al: RT-PCR for tyrosinase-mRNA-positive cells in peripheral blood: Evaluation strategy and correlation with known prognostic markers in 123 melanoma patients. J Invest Dermatol 110:263-267, 1998
12. Ghossein RA, Coit D, Brennan M, et al: Prognostic significance of peripheral blood and bone marrow tyrosinase messenger RNA in malignant melanoma. Clin Cancer Res 4:419-428, 1998
13. Hoon DS, Wang Y, Dale PS, et al: Detection of occult melanoma cells in blood with a multiple-marker polymerase chain reaction assay. J Clin Oncol 13:2109-2116, 1995
14. Kunter U, Buer J, Probst M, et al: Peripheral blood tyrosinase messenger RNA detection and survival in malignant melanoma. J Natl Cancer Inst 88:590-594, 1996
15. Mellado B, Colomer D, Castel T, et al: Detection of circulating neoplastic cells by reverse-transcriptase polymerase chain reaction in malignant melanoma: Association with clinical stage and prognosis. J Clin Oncol 14:2091-2097, 1996
16. Ghossein RA, Carusone L, Bhattacharya S: Molecular detection of micrometastases and circulating tumor cells in melanoma prostatic and breast carcinomas. In Vivo 14:237-250, 2000
17. Hoon DS, Bostick P, Kuo C, et al: Molecular markers in blood as surrogate prognostic indicators of melanoma recurrence. Cancer Res 60:2253-2257, 2000
18. Ghossein RA, Bhattacharya S, Coit DG: Reverse transcriptase polymerase chain reaction (RT-PCR) detection of melanoma-related transcripts in the peripheral blood and bone marrow of patients with malignant melanoma. What have we learned Recent Results Cancer Res 158:63-77, 2001
19. Curry BJ, Myers K, Hersey P: Polymerase chain reaction detection of melanoma cells in the circulation: Relation to clinical stage, surgical treatment, and recurrence from melanoma. J Clin Oncol 16:1760-1769, 1998
20. Buzaid AC, Ross MI, Balch CM, et al: Critical analysis of the current American Joint Committee on Cancer staging system for cutaneous melanoma and proposal of a new staging system. J Clin Oncol 15:1039-1051, 1997
21. Voit C, Schoengen A, Weber L, et al: Identification of melanoma metastases by tyrosinase-reverse transcription-polymerase chain reaction of fine needle aspirates. J Am Acad Dermatol 39:1030-1032, 1998
22. Voit C, Schoengen A, Schwurzer M, et al: Detection of regional melanoma metastases by ultrasound B-scan, cytology or tyrosinase RT-PCR of fine-needle aspirates. Br J Cancer 80:1672-1677, 1999
23. Proebstle TM, Jiang W, Hogel J, et al: Correlation of positive RT-PCR for tyrosinase in peripheral blood of malignant melanoma patients with clinical stage, survival and other risk factors. Br J Cancer 82:118-123, 2000
24. Szenajch J, Jasinski B, Synowiec A, et al: Prognostic value of multiple reverse transcription-PCR tyrosinase testing for circulating neoplastic cells in malignant melanoma. Clin Chem 49:1450-1457, 2003
25. Keilholz U, Willhauck M, Rimoldi D, et al: Reliability of reverse transcription-polymerase chain reaction (RT-PCR)-based assays for the detection of circulating tumour cells: A quality-assurance initiative of the EORTC Melanoma Cooperative Group. Eur J Cancer 34:750-753, 1998
26. Kaplan EL, Meier P: Nonparametric estimation from incomplete observations. J Am Stat Assoc 53:457-481, 1958
27. Lawless JF: Statistical Models and Methods for Lifetime Data. New York, NY, Wiley, 1982
28. Autier P, Coebergh JW, Boniol M, et al: Management of melanoma patients: Benefit of intense follow-up schedule is not demonstrated. J Clin Oncol 21:3707-3708, 2003
29. Reinhold U, Ludtke-Handjery HC, Schnautz S, et al: The analysis of tyrosinase-specific mRNA in blood samples of melanoma patients by RT-PCR is not a useful test for metastatic tumor progression. J Invest Dermatol 108:166-169, 1997
30. Tessier MH, Denis MG, Lustenberger P, et al: Detection of circulating neoplastic cells by reverse transcriptase and polymerase chain reaction in melanoma. Ann Dermatol Venereol 124:607-611, 1997
31. Schittek B, Blaheta HJ, Florchinger G, et al: Increased sensitivity for the detection of malignant melanoma cells in peripheral blood using an improved protocol for reverse transcription-polymerase chain reaction. Br J Dermatol 141:37-43, 1999
32. Schittek B, Bodingbauer Y, Ellwanger U, et al: Amplification of MelanA messenger RNA in addition to tyrosinase increases sensitivity of melanoma cell detection in peripheral blood and is associated with the clinical stage and prognosis of malignant melanoma. Br J Dermatol 141:30-36, 1999
33. Palmieri G, Strazzullo M, Ascierto PA, et al: Polymerase chain reaction-based detection of circulating melanoma cells as an effective marker of tumor progression: Melanoma Cooperative Group. J Clin Oncol 17:304-311, 1999
34. Hanekom GS, Stubbings HM, Johnson CA, et al: The detection of circulating melanoma cells correlates with tumour thickness and ulceration but is not predictive of metastasis for patients with primary melanoma. Melanoma Res 9:465-473, 1999
35. Brownbridge GG, Gold J, Edward M, et al: Evaluation of the use of tyrosinase-specific and melanA/MART-1-specific reverse transcriptase-coupled-polymerase chain reaction to detect melanoma cells in peripheral blood samples from 299 patients with malignant melanoma. Br J Dermatol 144:279-287, 2001
36. Mellado B, Del Carmen Vela M, Colomer D, et al: Tyrosinase mRNA in blood of patients with melanoma treated with adjuvant interferon. J Clin Oncol 20:4032-4039, 2002
37. Dutcher JP: The prognostic role of detection of circulating melanoma cells in the blood. J Clin Oncol 21:757-759, 2003
38. Osella-Abate S, Savoia P, Quaglino P, et al: Tyrosinase expression in the peripheral blood of stage III melanoma patients is associated with a poor prognosis: A clinical follow-up study of 110 patients. Br J Cancer 89:1457-1462, 2003
39. Bosserhoff AK, Kaufmann M, Kaluza B, et al: Melanoma-inhibiting activity, a novel serum marker for progression of malignant melanoma. Cancer Res 57:3149-3153, 1997
40. Stahlecker J, Gauger A, Bosserhoff A, et al: MIA as a reliable tumor marker in the serum of patients with malignant melanoma. Anticancer Res 20:5041-5044, 2000
41. Henze G, Dummer R, Joller-Jemelka HI, et al: Serum S100-a marker for disease monitoring in metastatic melanoma. Dermatology 194:208-212, 1997
42. Juergensen A, Holzapfel U, Hein R, et al: Comparison of two prognostic markers for malignant melanoma: MIA and S100 beta. Tumour Biol 22:54-58, 2001
43. Djukanovic D, Hofmann U, Sucker A, et al: Comparison of S100 protein and MIA protein as serum marker for malignant melanoma. Anticancer Res 20:2203-2207, 2000
44. Hancock BW, Wheatley K, Harris S, et al: Adjuvant interferon in high-risk melanoma: The AIM HIGH Study—United Kingdom Coordinating Committee on Cancer Research randomized study of adjuvant low-dose extended-duration interferon alfa-2a in high-risk resected malignant melanoma. J Clin Oncol 22:53-61, 2004(Christiane Voit, Martina )