Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
(CANCER RESEARCH 53. 61-66. January 1. 19931 Serum Fragment of Cytokeratin Subunit 19 Measured by CYFRA 21-1 Immunoradiometric Assay as a Marker of Lung Cancer Jean-Louis Pujol,1, Jean Grenier, Jean-Pierre Daurès,Alain Daver, Henri Pujol, and François-Bernard Michel Clinique lÃ-esMaladies Respiratoires. Universitétie Mtmtpellier. Hôpital Aiguelonfìue.3405V Montpellier Cedex ¡J-L P.. F-B. M.]; Centre Régionalde Lune contre le Cancer, Laboratoire de Raditi-Anal\ses, ÃŒ4W4Montpellier Cedex ¡J.G.. H. P.¡:Départementde l'Information Médicale.Hôpital Gaston Doumergue. 30000 Nîmes¡J-P.D./: and Centre Paul Papin. 2. Rue Moll. 4W36 Angers ¡A.D.I. France and prognosis. The three main prognostic factors able to help treat ment decision are histology (mainly SCLC versus NSCLC), stage of disease, and performance status (8, 9). The search for new clinical or biological variables as putative independent prognostic factors may potentially improve the management of lung cancer patients. Cytokeratins are intermediate filaments which are part of the cy- ABSTRACT Cytokeratin 19 is a subunit of cytokeratin intermediate filament ex pressed in simple epithelia and their malignant counterparts. Therefore, it is expressed by respiratory epithelium cells and has been detected in lung cancer specimens. An immunoradiometric assay was used to detect a fragment of the cytokeratin 19, referred to as CYFRA 21-1, in the serum of 165 patients with histologically proved lung cancer ( 128 non-small cell toskeleton of both normal epithelia and their malignant counterparts (10). Immunohistochemical studies using broad-spectrum cytokeratin and 37 small cell lung cancers). This prospective study was conducted to evaluate the reliability of this immunoradiometric assay and to identify the relationship between serum CYFRA 21-1 and different features of lung antibodies have demonstrated the expression of cytokeratins in both SCLC and NSCLC suggesting a common endodermal lineage (11, 12). In contrast to other intermediate filaments, cytokeratins are a family of polypeptides, some of them having cell type specificity (10). Cytokeratin 19 is an acidic (type I) subunit expressed in all simple epithelia and in carcinomas such as lung cancer which arise from them. A fragment of cytokeratin subunit 19 can be measured in serum by a new immunoradiometric assay, the CYFRA 21-1 using two mouse MoAb KS 19-1 and BM 19-21. Thus, this cytokeratin 19 fragment is referred to as CYFRA 21-1. The aims of the present prospective study are (a) to evaluate the reliability of serum CYFRA 21-1, (b) to identify cancer including prognosis. The minimal detectable concentration de tected by this assay was 0.06 ng/ml. The reliability of the immunoradio metric assay was demonstrated by the linear relationship between CYFRA 21-1 measurement and dilution of the serum, the reproducibility of the dosage in intraassay and interassay, and the high sensitivity of the method in discriminating low CYFRA 21-1 concentrations. Using a threshold of 3.6 ng/ml, sensitivity and specificity were 0.52 and 0.87, respectively. The sensitivity of the marker was highest in squamous cell carcinoma and lowest in small cell carcinoma. In non-small cell lung cancer patients, the marker varied significantly according to both stage of the disease (Kruskal-Wallis, 13.7; P < 0.005) and performance status (KruskalWallis, 9.16; P < 0.05) inasmuch as a high serum CYFRA 21-1 level was any relationship between the marker and other clinical and biological parameters in both SCLC and NSCLC and (c) to analyze the prog nostic significance of a high CYFRA 21-1 level at time of diagnosis. associated with advanced stages, mediastinal lymph nodes, and poor per formance status. Consequently, the marker was significantly lower in patients who were operated upon when compared with unresectable ones. Lung cancer patients with serum CYFRA 21-1 over 3.6 ng/ml proved to have a significantly shorter overall survival than those with a normal serum level (log rank, P = 0.007; Wilcoxon, /' - 0.001). The negative MATERIALS prognostic effect of CYFRA 21-1 was highly significant in squamous cell carcinomas whereas it was nonsignificant for the other histologies. In Cox's model analysis, performance status, stage grouping, and CYFRA 21-1 were the only significant determinants AND METHODS Patients. One hundred sixty-five consecutive lung cancer patients referred to the Montpellier University Hospital between February 1990 and February 1992 were prospectively entered in the study (Table 1). All patients had pathologically confirmed lung cancer. Among them were 37 SCLC. 85 SQC, 28 adenocarcinomas, and 15 large cell carcinomas as defined by the WHO classification (13). Performance status was estimated according to the Eastern Cooperative Oncology Group and the percentage of weight loss during the previous 4 months was recorded. Staging was carried out by exhaustive pro cedure according to the 4th edition of the Union Internationale Contre le Cancer tumor-nodes-metastasis classification (14) and the American Thoracic of survival. This study sup ports the use of the serum fragment of cytokeratin subunit 19 CYFRA 21-1 as an independent prognostic marker of squamous cell carcinoma of the lung. INTRODUCTION Treatment of lung cancer is probably one of the great challenges of medical oncology owing to an increasing incidence in both men and women and poor prognosis (1, 2). SCLC2 have neuroendocrine bio Society map of regional pulmonary nodes (15). Staging of NSCLC was es tablished according to Mountain's stage grouping (16); for SCLC, limited disease was defined as disease confined to one hemithorax including medias tinal lymph nodes and/or ipsilateral supraclavicular lymph nodes; extensive disease was defined as a disease more advanced than limited SCLC. For all patients staging procedure included clinical examination, standard chest roent- logical properties usually associated with a high sensitivity to chemo therapy (2-^). Surgery offers the best probability of cure for NSCLC but it cannot be proposed for locally advanced and metastatic stages; therefore, other modality treatments such as radiotherapy-chemother apy association have been proposed (5). Because metastatic disease is the first cause of failure for both SCLC (6) and NSCLC (7), research for new treatment modalities is an important goal. Such an experi mental approach requires an accurate definition of groups of patients genography. computed tomographic scan of chest and upper abdomen, fiberop tic bronchoscopy, liver sonography. and bone scanning. Brain computed to mographic scan was done systematically for SCLC and if clinically required for NSCLC. Controls. The serum CYFRA 21-1 level was measured in 104 patients with nonmalignant pulmonary diseases (20 infectious diseases. 75 chronic obstruc tive pulmonary diseases, and 9 miscellaneous). Treatment Decision. Each patient was discussed by a medical panel com posed of thoracic surgeons, chest physicians, radiologists, and medical oncol ogists. NSCLC patients with stage I or II disease or with moderate locally advanced NSCLC (Ilia with resectable node metastasis) underwent surgery in an attempt to achieve complete resection. Other NSCLC patients with perfor mance status <2 and distant metastasis (stage IV) or gross mediastinal involve ment (stage Illb and stage Ilia with more than 2 ipsilateral mediastinal lymph Received 7/2/92: accepted 10/20/92. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely (o indicate this fact. 1To whom requests for reprints should he addressed. 2 The abbreviations used are: SCLC. small cell lung carcinoma: NSCLC. non-small cell lung carcinoma: MoAb. monoclonal antibodies: SQC. squamous cell carcinomas; MDC. minima] detectable concentration; ROC, receiver operating characteristic; St. cy tokeratin standard; LDH, láclatedehydrogenase; NX. nodal stage \. 61 Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1993 American Association for Cancer Research. SERUM CYFRA 21-1 IN LUNG CANCER Table 1 Patients' characteristics 165 146/19 62 (10,31-88) No. M/F Mean age (SD, range) 128 NSCLC Performance stalus 0-1 2-3 4(1 Slage grouping l und II Illa Illb IV 17 27 35 49 Wt loss (%) Absence Presence 73 55 SCLC 37 Disease extent Limited disease Extensive disease 21 In Wt loss (%) 0 >5 21 16 Performance status 0-1 2-3 28 9 node métastases)were eligible for chemotherapy. Best supportive care, includ ing palliative radiation therapy when needed, was given to patients with ad vanced stage and poor performance status. SCLC patients underwent chemo therapy [usually an alternating regimen containing cisplatin (17)]. Treatment was decided upon according to routine clinical and biological findings and without knowledge of the serum CYFRA 21-1 level. Hence, treatment was not considered as a prognostic variable in this study. CYFRA 21-1 Immunoradiometric Assay. A blood sample was taken from each patient at presentation, and Ihe serum was separated and stored at -190°C until tested. CYFRA 21-1 (Centocor Diagnostics. Malvern, PA: Cis Biointernational. Gif/Yvette, France) is a solid-phase immunoradiometric assay based on the two-site sandwich method. In this method the cytokeratin 19 was recognized by two mouse MoAb. KS 19-1 and BM 19-21. directed against two different epitopes of a fragment of cytokeratin subunit 19. which is referred to as serum CYFRA 21-1. These MoAb were obtained by immunization against the MCF7 cell line. MoAb KS 19-1 -coated polystyrene spheres were incubated with 200 ul of patient serum, control serum, or standard curve (composed of the following concentrations of cytokeratin 19: 0, 3, 8, 25, and, 50 ng/ml). for 20 h at 2-8°C. Afterwards, the solid phase was washed with distilled water and then incubated with 0.85 uCi/ml of '"I-labeled BM 19-21 for 3 h at 2-8°C. Finally, the solid phase was washed again with distilled water in order to cancel the nonfixed labeled reagents. Radioactivity was counted in a well-type gamma counter (Auto-Gamma: Packard Instrument Company. Downers Grove. IL) and expressed in cpm. The calculated concentration of cytokeratin 19 was expressed ¡nng/ml. Control of the Immunoradiometric Assay. The sensitivity of the method was assessed by determining the MDC. This method used two cytokeratin 19 standards containing, respectively. 0 (cytokeratin free. StO) and 3 ng/ml (Stl): these 2 standards were measured in 6 different assays (Table 2): in each of the 6 assays, the same total radioactivity was introduced for both standards. The bound/total radioactivity was calculated for StO and Stl. The MDC in ng/ml was determined as: X The control of the reliability of the method of measurement was done using the following tests: (a) test of linearity of the relationship between consecutive dilutions of a serum of known high cytokeratin 19 level and results of the titration; (b) assessment of intraassay reproducibility by measuring 7 different sera twice during the same assay: (c) assessment of interassay reproducibility by measuring 2 different sera in 10 different assays: (a1)analysis of the standard curve in the low values in order to evaluate the difference of signal given by the StO and the St l. Total LDH assays were done following the Deusch Chemical Society recommendations by measuring its activity using pyruvate as a substrate (Bio-Merieux, France). The upper limits of normal values were as follows: LDH, 330 units/liter: alkaline phosphatase. 220 units/liter; leukocytes 8000/ul. The lower limits of normal values were 32 g/liter for albumin and 135 mmol/liter for serum sodium. All sera samples were assayed blind of clinical information. Statistics. ROC curves were constructed using both patient and control subject serum CYFRA 21-1 levels ¡nan attempt to establish a sensitivityspecificity relationship; areas under the ROC curve were calculated (18). The serum tumor marker was not distributed normally: thus, for each patient subset, results were expressed as median, and variation was expressed as interquartile range. Nonparametric statistical analyses were used: differences between two independent groups were determined by means of the Mann-Whitney U test; differences between more than two groups were determined by means of the Kruskal-Wallis one-way analysis of variance: P < 0.05 was considered as significant. Survival was defined as the time from the date of sampling to the date of death. Survival data were updated in February 1992 and none of the patients were lost from sight during follow-up. Probability of survival was estimated by the method of Kaplan and Meier (19). Single variable survival analyses were done by means of Wilcoxon and log rank tests and multivariate regression was done with the Cox's model (20). There were 12 variables in the model. Serum CYFRA 21-1 was examined as normal (<3.6 ng/ml) or elevated (>3.6 ng/ml) as were the other biological variables (LDH, serum sodium, alkaline phosphatase. albumin, and leukocytes) according to their respective upper normal limits. Age groups were defined as less than or equal to and over 50 years old. Other variables were stage of the disease, sex, histology, weight loss, and performance status. Survival was analyzed using the SAS software package. RESULTS Sensitivity and Reliability of the Immunoradiometric Assay. According to the data in Table 2 the calculated MDC was 0.06 ng/ml. The test of linearity showed a linear relationship between the titrations of CYFRA 21-1 and the dilutions of a serum sample with a high cytokeratin 19 concentration (Fig. 1). Seven sera were measured twice during the same assay (intraassay). The results of the first and second measurements were, respectively (ng/ml): 12, 11.4; 27, 30; 6, 7.5; 3, 2; 4.4, 4.8; 45, 41 ; 3.4, 3.2. The coefficients of variation of 2 sera tested in 10 different assays (interassays) were 7 and 9.8%, respectively. Analysis of the standard curve in the low values demonstrated that the difference between 0 and 3 ng/ml was 700 cpm, a value sufficient to prevent overlap of results. Tumor Marker Distribution. The median and interquartile range interquartile range, of serum CYFRA 21-1 was significantly higher in Table 2 Results of 6 different inuminoradiotnetric assays performed in order to determine the minimal detectable concentratitm StO Stl Assay123456T" (cpm)226.473219.399208.757241.441207.643198.408cpm358976718364B/T0.020.040.040.030.0 x where A is MDC in ng/ml: X is [standard Stl] - [standard StOJ = 3 ng/ml: Y is mean Stl (cpm) - mean StO (cpm): and y = 2 SD of StO (cpm). " B, bound radioactivity; T, total radioactivity. Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1993 American Association for Cancer Research. SERUM CYFRA 21-1 IN LUNG CANCER >SO level differed when the histológica! type was considered (KruskalWallis test KW = 7.69; P < 0.05). Serum CYFRA 21-1 level was y •¿ »8.78 X - 0.17 •¿ 0.998; p •¿ 0.002 significantly higher in SQC when compared with each of the other histologies (Mann-Whitney U test, P < 0.05). CYFRA 21-1 and Clinical Variables at Presentation. In NSCLC, the CYFRA 21-1 level, was significantly higher in patients with a metastatic disease (7.4; interquartile range, 3.0-21.0 ng/ml) in com parison with the limited disease patient subgroup (3.8; interquartile range. 1.9-8.5 ng/ml; Mann-Whitney U test, P < 0.001). When the marker was analyzed according to stage grouping, a significant in crease of the serum level was observed from stage I to stage IV (Kruskal Wallis. 13.7; P < 0.005; Fig. 4) but there was no significant difference when stage Ilia and stage Illb were compared (MannWhitney U test, 500; nonsignificant). CYFRA 21-1 levels were lower 1:16 1:32 1:64 Serum dilutions Fig. l. Dilution test of a serum with a high level of cytokeratin I1). in patients who were operated upon when compared with inoperable patients (2.4; interquartile range. 1.25-5.0 versus 5.8; interquartile range, 2.7-13.0 ng/ml; Mann-Whitney U test, P < 0.001). In addition. Table 3 Sensitivity and specificity of CYFRA 21-1 using a 3.ft-n\>/ml upper limit of normal mines 1.0 0.76 Subsets Sensitivity Specificity Non-small cell SQC Small cell All 0.56 0.63 0.46 0.52 0.89 0.91 0.89 0.87 knuka Willi! p no'. 0.6 g 0.28 3.6 ADE SQC LCC SCLC Fig. 3. CYFRA 21-1 distribution according to histology. . median value: columns. interquartile range. ADE. adenocarcinoma: LCC. large cell carcinoma. 0.26 False-Positive 0.6 0.76 1.0 Rate (1 - Specificity) Fig. 2. Receiver operating eharaeteristie eurve for serum CYFRA 2I-1. O. squanmus cell carcinoma, area under Ihe concentration curve = 0.83 ±0.03 (SE); •¿ non-small cell lung cancer, area under the concentration curve = 0.80 ±0.03. cancer patients (4.3; interquartile range, 2.3-9.5) when compared with control subjects (1.05; interquartile range, 0.8-1.8; Mann-Whitney U test, P < 0.001). Areas under ROC curves were 0.83 ±0.03, 0.80 ±0.03, and 0.74 ±0.05 for the SQC, NSCLC. and SCLC groups, respectively (Fig. 2). Using ROC curves, a threshold of 3.6 ng/ml was chosen as the upper limit of normal values. The sensitivity and specificity of this cutoff in the different patient subsets are shown in Table 3. CYFRA 21-1 and Histology. The median (interquartile range) serum CYFRA 21-1 for SCLC, SQC. adenocarcinomas, and large cell carcinomas were 3.4 (1.8-7.0), 5.6 (2.6-13.0), 2.9 (1.85-7.10), and 3.8 (1.8-7.0) ng/ml, respectively (Fig. 3). The serum CYFRA 21-1 24222018161412108642n-Krutk.l .0.006-----nn W.lli. p .-3.6 TTq_ji tÃ-— I & II lila 1Mb IV Fig. 4. CYFRA 21-1 distribution according to stage grouping in non-small eel lung cancer. , median value; columns, interquartile range. Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1993 American Association for Cancer Research. SERUM CYFRA 21-1 IN LUNG CANCER Table 5 Significant prognostic factors in the entire population patients who presented mediastinal lymph nodes (N2 or N3) demon strated higher serum CYFRA 21-1 levels (5.0; interquartile range, 2.6-12.5 ng/ml), when compared with patients without mediastinal lymph nodes (NO or Nl, 3.2; interquartile range, 1.75-8.0 ng/ml; Mann-Whitney U test, P < 0.05). In NSCLC, the distribution of CYFRA 21-1 varied significantly according to the performance status of NSCLC patients (KruskalWallis 9.16; P < 0.05; Table 4). In SCLC, CYFRA 21-1 did not significantly differ between limited (3.0; interquartile range, 1.8-5.4 ng/ml) and extensive disease (4.55; interquartile range, 2.1-11.5 ng/ ml; Mann-Whitney U test, P = 0.07). Survival. Univariate analysis showed that lung cancer patients with serum CYFRA 21-1 over 3.6 ng/ml proved to have a significantly shorter overall survival than those with a normal serum level (log rank, P = 0.007; Wilcoxon. P = 0.001 ; Fig. 5). The negative prognostic effect of CYFRA 21-1 was highly significant in SQC (log rank, P < 0.001; Wilcoxon, P < 0.001) whereas it was nonsignificant for the other histologies. Separate survival analyses showed a significant effect of stage group, weight loss, performance status, serum albumin, LDH, and CYFRA 21-1 (Table 5). With Cox's model analysis, per (dayslStageI Factor and level Median surviva 1Wilcoxon rank0.0001 IIIliaHlbIVPerformance and Log 0.00010.0001 status9-12-4Wt 0.00010.0018 lossNo loss0-5%>5%Albumin wt 0.00150.0049 (g/liter)>32<32Láclate 0.0370.0041 (units/liter)<330>330CYFRA dehydrogenase 0.01770.001 formance status, stage grouping, and CYFRA 21-1 were the only significant determinants of survival (Table 6). (ng/ml)<3.6>3.6420:?4155407148402159191367191398252402213f 21-1 0.007 DISCUSSION Table 6 Regression analvsis results: estimated relative risk for significant variables in = 165) New strategies of chemotherapy intensification are under evalua tion in SCLC patients in an attempt to circumvent secondary chemoresistance (3, 4). NSCLC is a group of different histologies having a lower chemosensitivity when compared with SCLC. Surgery of stage I and II NSCLC offers the best probability of cure (5, 21) but in the more advanced stages combined modality treatments are required Table 4 CYFRA 21-! distribution according to performance in non-small cell lung cancers 78 319 Log-rink p -0.007 Wilcoxon p -0.001 £ O.B t "o 0.6 i 0.4 Z 5 s 0.2 400 600 800 days CYFRA <• 3.6 nQ/mL CYFRA > 3.6 ng/mL median «urvlval <• 3.6 ng/HL > 3.6 ng/mL % c«n«ortd •¿0 40« SS 86 fit 34 index Stage of disease CYFRA 21-1Regression ±0.22 0.90 ±0.29 0.49 ±0.21P<0.0001 risk2.81 0.0024 0.021Relative 2.47 1.64 Thus, the treatment of both SCLC and NSCLC lung cancer is still an experimental approach. In this setting, an awareness of significant variables able to predict metastatic spread and poor prognosis is recommended (25). Some important prognostic factors of lung cancer have already been reported. For NSCLC, the two main variables which determine sur vival are the stage of the disease and the performance status (8). In several studies, tumor markers such as tissue polypeptide antigen (26, 27), LDH (8), or carcinoembryonic antigen (26, 27) have been demonstrated as significant prognostic factors in univariate analyses. Other biological properties of NSCLC have been investigated as pu tative indicators of prognosis. In different multivariate studies aneuploidy (28), lack of expression of blood group A antigen (29) or expression of the neural cell adhesion molecule (30) have been inde pendently demonstrated as prognostic factors; however, correlation and interactions between these variables are not known. For SCLC, several studies have demonstrated the univariate influence of disease extent, performance status, LDH, and alkaline phosphatase on sur vival (9, 31-33); moreover, some authors have suggested that the poorest prognosis is seen in SCLC patients with bone marrow métas tases (34), elevated serum neuron specific enolase (35), or carcino embryonic antigen (26). Intermediate filaments are part of the cytoskeleton which provide important information on the cell origin (36). They are classified into five tissue specific protein families. The cytokeratin family is ex pressed by all epithelial cells (10) including endocrine cells of the dispersed neuroendocrine system (37) and, it appears, therefore, to be a general specific and useful marker of the epithelial differentiation. Interestingly, the expression of cytokeratins is not lost by epithelial cells during malignant transformation (10, 36), a phenomenon which (ng/ml)1 .3-5.0 2.2-9.5 2.7-16.0 4.6-84.0n10 coefficient (ß±SE)1.03 owing to the frequency of microscopic metastatic disease (7, 22-24). status (PS) range PSO PS1 4.36.0 PS2 10.0Interquartile PS3Median2.7" 1Kruskal-Wallis = 9.16; P < 0.05. VariablePerformance Fig. 5. Probability of survival of all patients with normal and elevated pretreatment serum CYFRA 21-1 level. 64 Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1993 American Association for Cancer Research. SI-RUM CYI-RA 21-1 IN l.UNG contrasts with the well-known phenotypic instability of cancer cells (38). Antibodies raised against some common antigenic determinants of all cytokeratins are therefore useful in typing malignant tumors with poor histologie-aïfeatures of differentiation (36). Immunohistochemical studies using these broad-spectrum cytokeratin antibodies have clearly demonstrated that all four histológica! types of lung cancer expressed cytokeratins (II, 12, 39); in particular, the classic SCLC type which have been shown to coexpress both cytokeratin and neurofilament whereas the variant type only express cytokeratin (40). This supports the hypothesis of a common epithelial lineage of all histológica! types of lung cancer. A cytokeratin is a heterotypic tetratner of protofiluments composed of two polypeptidcs: one acidic type I subunit; and one basic type II subunit. Each type of epithelia and their malignant counterpart express a specific cytokeratin polypeptide pattern (10, 41); simple epithelia, including pseudostratified epithelia. express cytokeratins 7, 8. 18, and 19. Selective anti bodies raised against simple epithelium type of cytokeratin have been shown to react with all histologies of lung cancers (42, 43). Although cytokeratins are part of the cytoskeleton some fragments might be released in the serum owing to cell lysis or tumor necrosis, which gives support to the present research on cytokeratin subunit 19 frag ment as a tumor marker of lung cancer. The immunoradiometric assay described in this study is able to detect the serum fragment of cytokeratin subunit 19, referred to as CYFRA 21-1, with both sensitivity and reliability. The threshold of 3.6 ng/ml was determined using the ROC curve analysis which showed that this threshold allowed the best sensitivity-specificity relationship. Our study has demonstrated a correlation between serum CYFRA 21-1 and stage of the disease in NSCLC but not in SCLC. Moreover, in NSCLC, the marker varied significantly according to nodal status. These data suggest that CYFRA 21-1 might be consid ered as a marker of tumor mass. Thus, patients who present a high level of serum CYFRA 21-1 might require a careful search of métas tases. Despite these results it is not possible to predict operability of NSCLC using the serum level of the marker owing to the important overlap of the distribution in stage Ilia (usually marginally resectable) and stage Illb- (unresectable) NSCLC groups. The reason for this result might be that the distinction between stage Ilia and stage Illb mainly depends on anatomic extension rather than tumor mass. The low incidence of a high serum CYFRA 21-1 level in the stage I-II group clearly indicates that the marker is of no help in screening or early diagnosis of the disease. Because CYFRA 21-1 is correlated with both disease extent and performance status it is not surprising that univariate analysis showed that patients with a high serum level proved to have a significantly shorter survival. The negative effect of CYFRA 21-1 on survival was particularly significant in SQC whereas this effect was not demonstrated in the other histologies. Interestingly. Cox's model demonstrated that the prognostic significance brought by CYFRA 21-1 is independent and adds information to the well-known and above-mentioned prognosis factors. This study supports the use of the serum fragment of cytokeratin subunit 19 CYFRA 21-1 as an independent prognostic marker of squamous cell carcinoma of the lung. 3. Klastersky. J. Commentary: intensive chemotherapy in small cell lung cancer. Eur. J. Cancer Clin. Oncol., 25: 933-937, 1989. 4. Hansen. H. H., and Kristjansen. P. E. G. Chemotherapy of small cell lung cancer. Eur. J. Cancer. 27: 342-349, 1991. 5. Mulshine, J. L.. Glatstein, E.. and Ruckdeschel. J. C. Treatment of non-small cell lung cancer. J. Clin. Oncol., 4: 1704-1715. 1986. 6. lannuzzi. M. C.. and Scoggin. C. H. State of the art: small cell lung cancer. Am. Rev. Rcspir. Dis.. 134: 593-608. 1986. 7. Cox. J. D.. and Yesner, R. A. Causes of treatment failure and death in carcinoma of the lung. Yale J. Biol. Med.. 54: 201-207. 1981. 8. O'Connel. J. P.. Kris. M. G.. Gralla. R. J.. Groshen, S., Trust, A., Fiore. J. J.. Kelsen. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. ACKNOWLEDGMENTS 31. The authors wish to thank Jo Baissus for help in preparing Ihe manuscript. REFERENCES 32. 1. Stjemswurd. J.. and Stanley, K. Etiology, epidemiology and prevention of lung cancer. Lung Cancer, 4: 11-24. 1988. 2. Hansen. H. H. Lung cancer. In: H. M. Pinedo, D. L. Longo, and B. A. Chabner (eds.). Cancer Chemotherapy and Biological Response Modifiers, pp. 222-240. Annual 10. Amsterdam: Elsevier/North-Holland BiomédicalPress, 1988. CANCI-R 33. D. P., Heelan, R. T., and Golhey. R. B. Frequency and prognostic importance of pretrcatment clinical characteristics in patients with advanced non-small cell lung cancer treated with combination chemotherapy. J. Clin. Oncol.. 4: 1604-1614. 1986. Osterlind. K.. [hde, D. C.. Ettinger. D. S.. Gralla. R. J.. Karrer. K.. Krauss. S., Maurer. L. H.. Rorth. M-, Sorenson. S., and Vincent. R. Staging and prognostic factors in small cell carcinoma of the lung. Cancer Treat. Rep.. 67: 3-9, 1983. Moll, R.. Franke. W. W.. Schiller. D. L.. Geiber. B., and Krepier, R. The catalog of human cytokeratins: patterns of expression in normal epithelia. tumors and cultured cells. Cell, 31: 11-24, 1982. Broers, J. L. V.. Rot, M. K., Oostendorp. T., Huysmans, A., Wagenaar. S. S., Wiersmavan Tilburg. A. J. M.. Vooijs. P, and Ramaekers, F. C. S. Immunohistochemical detection of human lung heterogeneity using antibodies to epithelial, neuronal and neuroendocrine antigens. Cancer Res., 47: 3225-3234, 1987. Pujol, J. L.. Simony, J., Laurent. J. C., Richer, G., Mary. H.. Bousquet, J.. Godard. Ph.. and Michel, F. B. Phenotypic heterogeneity studied by immunohistochemistry and aneuploidy in non-small cell lung cancers. Cancer Res., 4V: 2797-2802, 1989. World Health Organization. The World Health Organization Histológica! Typing of the Lung Tumors. Ed. 2. Am. J. Clin. Pathol., 77: 123-136. 1982. Sobin. L. H.. Hermanek. P.. and Hutter. R. V. P. TNM classification of malignant tumours. Ed. 4. Geneva: International Union Against Cancer. 1987. Tisi. G. M.. Friedman, P. J.. Peters. R. M.. Pearson. G.. Carr. D.. Lee, R. E.. and Selawry, O. American Thoracic Society: clinical staging of primary lung cancer. Am. Rev. Respir. Dis., 125: 659-64. 1982. Mountain. C. F. A new international staging system for lung cancer. Chest, 4(s): 225s-233s, 1986. Pujol. J. L., Demoly, P., Gaulier. V.. Romieu. G.. Dan Aouta. M., Stenger. R.. Parrai. IL, Tadlaoui, O., Marcenac, A. M., Benahmed. S. S.. Michel, F. B.. and Godard. P. Phase II study of alternating combination chemotherapy in small cell lung cancer. Lung Cancer, 7: 285-294, 1991. Beck, J. R.. and Shultz. E. K. The use of relative operating characteristic (ROC) curves in test performance evaluation. Arch. Pathol. Lab. Med.. Ill): 13-20, 1986. Kaplan. E. L.. and Meier. P. Nonparametric estimation from incomplete observations. J. Am. Slat. Assoc., 53: 457^81. 1958. Cox. D. R. Regression models and life tables. J. R. Statist. Soc. B.. 34: 187-202, 1972. Gail. M. H., Eagan, R. T, Feld, R.. Ginsberg, R.. Goodell, B., Hill, L.. Holmes, E. C, Lukeman. J. M.. Mountain. C. F.. Oldhani, R. K., Pearson, F. G., Wright, P. W., and Lake. W. H. Prognostic factors in patients with resected stage I non-small cell lung cancer. Cancer (Phila.), 54: 1802-1813. 1984. Pearson, F. G.. DeLarue, N. C.. lives, R.. Todd. T. R. J., and Oxiper. J. D. Significance (if positive superior mediastinal nodes identified at mediastinoscopy in patients with resectable cancer of the lung. J. Thorac. Cardiovasc. Surg., 83: 1-11, 1982. Pujol, J. L.. Rossi, J. F.. Le Chevalier, T.. Daurès.J. P.. Rouanet. P.. Douillard, J. Y., Dubois. J. B.. Amagada. R.. Mary, H.. Godard, P., and Michel. F. B. Phase 11 pilot study of neoadjvam ifosfamide, cisplatin. and etoposide in locally advanced nonsmall cell lung cancer. Eur. J. Cancer. 26: 798-801, 1990. Tubiana. M. The growth and progression of human tumors: implications for man agement strategy. Radiothcr. Oncol.. 6: 167-184. 1986. World Health Organization. WHO Handbook for Reporting the Results of Cancer Treatment, WHO Offset Publication No. 48. Geneva: WHO. 1979. Buccheri. G. F. Violante, B., Sartoris, A. M., Ferrigno, D., Curcio, A., and Vola. F. Clinical value of a multiple biomarker assay in patients with hronchogcnic carcinoma. Cancer (Phila.), 57: 2389-23%. 1986. Salvati. F.. Cruciani, A. R., Flore, F.. De Angelis. G.. Pigorini. F.. Antilli. A.. Pau, F.. Munno, R., and Cipri, A. Plasma carcinocmhryonic antigen and tissue polypeptide antigen levels in lung cancer: correlation with cell types and stage. Cancer. Detect. Prev., 8: 111-114, 1985. Zimmerman, P. V. Hawson, G. A. T.. Bint. M. H.. and Parsons. P. G. Ploidy as a prognostic determinant in surgically treated lung cancer. Lancet, 2: 530-533, 1987. Lee. J. S., Ro. J. Y., Sahin. A. A.. Hong. W. K.. Brown. B. W.. Mountain, C. F. and Hittelman. W. N. Expression of blood-group antigen A: a favorable prognostic factor in non-small cell lung cancer. N. Engl. J. Med.. 324: 1084-1090. 1991. Bcrendsen, H. H.. De Leij. L.. Poppema. S.. Postmus. P. E.. Boes. A.. Sluiter. H. J.. and The. H. Clinical characterization of non-small-ccll lung cancer and tumors show ing neuroendocrine differentiation features. J. Clin. Oncol.. 7: 1614-1620. 1989. Rawson, N. S., and Peto. J. An overview of prognostic factors in small cell lung cancer. A report from the subcommittee for the management of lung cancer of the united Kingdom co-ordinating committee on cancer research. Br. J. Cancer, 31: 597-604, 1990. Osterlind. K.. and Andersen. P. K. Prognostic factors in small cell lung cancer: multivariate rmxJel based on 778 patients treated with chemotherapy with or without irradiation. Cancer Res., 46: 4189-4194. 1986. Albain. K. S.. Crowley, J. J., LeBlanc. M., and Livingston, R. B. Determinants of improved outcome in small cell lung cancer: an analysis of the 2580 patient South west Oncology Group data base. J. Clin. Oncol.. K: 1563-1574. 1990. 65 Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1993 American Association for Cancer Research. SERUM CYFRA 21-1 IN LUNG CANCER 34. Van Den Brande. P. Serum láclatedehydrogenase in small cell lung cancer. N. Engl. ¡.Med.. 320: 61, 1989. 35. Jorgensen. L. G.. Osterlind. K., Hansen. H. H., and Cooper, E. H. The prognostic influence of serum neuron specific enolase in small cell lung cancer. Br. J. Cancer. 58: 805-807, 1988. 36. Osbom. M, and Weber, K. Intermediate filaments: cell-type-specific markers in differentiation and pathology. Cell, 31: 303-306, 1982. 37. Hoefler, H., and Denk, H. Immunohistochemical demonstration of cytokeratins in gastrointestinal carcinoids and their probable precursors cells. Virchows Arch. B Cell Palhol., 403: 235-240, 1984. 38. Nicolson. G. L. Tumor cell instability, diversification, and progression to the metastatic phenotype: from oncogene to oncofetal expression. Cancer Res., 47: 14731487, 1987. 39. Van Muijen. G. N. P., Ruiter. D. J., Van Leeuwen, C. Prins, F A.. Reitsema, K., and 40. 41. 42. 43. Wamaar, S. O. Cytokeralin and neurofilament in lung carcinomas. Am. J. Pathol., 116: 363-369. 1984. Broers. J. L. V.. Carney. D. N.. De Ley, L., Vooijs, G. P., and Ramaekers, E C. S. Differential expression of intermediate filament proteins distinguishes classic from varianl small-cell lung cancer cell lines. Proc. Nati. Acad. Sci. USA. 82: 4409^1413. 1985. Moll, R., Krepier, R., Franke, W. W. Complex cytokeratin polypeptide patterns observed in certain human carcinomas. Differentiation, 23: 256-269, 1983. Debus, E., Moll, R.. Franke, W. W., Weber, K., and Osbom, M. Immunohistochemical distinction of human carcinomas by cytokeratin typing with monoclonal antibodies. Am. J. Pathol., 114: 121-130, 1984. Blobel, G. A.. Moll. R., Franke, W. W., and Vogt-Moykopf, I. Cylokeratins in normal and lung carcinomas. I. Adenocarcinomas, squamous cell carcinomas and cultured cell lines. Virchows Arch. B Cell Pathol., 45: 407-^»29,1984. 66 Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1993 American Association for Cancer Research. Serum Fragment of Cytokeratin Subunit 19 Measured by CYFRA 21-1 Immunoradiometric Assay as a Marker of Lung Cancer Jean-Louis Pujol, Jean Grenier, Jean-Pierre Daurès, et al. Cancer Res 1993;53:61-66. Updated version E-mail alerts Reprints and Subscriptions Permissions Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/53/1/61 Sign up to receive free email-alerts related to this article or journal. To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at [email protected]. To request permission to re-use all or part of this article, contact the AACR Publications Department at [email protected]. Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1993 American Association for Cancer Research.