the role of pretreatment squamous cell carcinoma

7/25/2019 The Role of Pretreatment Squamous Cell Carcinoma

1/8

The role of pretreatment squamous cell carcinomaantigen level in locally advanced squamous cellcarcinoma of the uterine cervix treated by

radiotherapyI. OGINO*, H. NAKAYAMAy, N. OKAMOTOz, T. KITAMURA & T. INOUE**Department of Radiology, Yokohama City University, School of Medicine, Yokohama, Japan; and DepartmentsofyGynecology,zEpidemiology, and Radiation Oncology, Kanagawa Cancer Center, Yokohama, Japan

Abstract. Ogino I, Nakayama H, Okamoto N, Kitamura T, Inoue T. The role of pretreatment squamous cellcarcinoma antigen level in locally advanced squamous cell carcinoma of the uterine cervix treated by radio-therapy.Int J Gynecol Cancer 2006;16:10941100.

The purpose of this study was to determine the pretreatment serum squamous cell carcinoma antigen(SCC-ag) level as a generally applicable measurement in predicting and estimating the treatment outcomeof patients with locally advanced SCC of the cervix. Three hundred fifty-two patients with stage IIBIVASCC of the cervix were managed with both external irradiation and highdose rate intracavitary brachy-therapy. A significantly higher median SCC-ag was seen in association with increasing stage, tumor size,and lymph node involvement. The difference in disease-free survival (DFS) between stages IIB and IIIpatients was not statistically significant with SCC-ag level ,2 ng/mL. In multivariate analysis, medianSCC-ag level (6.0 ng/mL) and lymph node metastases had significant independent effects on absolute sur-vival and DFS. A direct linear relationship (y 22.932x 1 84.896) existed between the median SCC-ag ofgroups distributed by pretreatment prognostic factors and the 5-year DFS rate. The 5-year DFS rate asa function of SCC-ag level defined by cervix size, lymph node status, and hydronephrosis was obtainedfrom a formula combining risk scores and the baseline survival function. From the obtained formulas, wecan objectively estimate the treatment outcome in patients with locally advanced squamous cell cervicalcancer.

KEYWORDS: cervical carcinoma, highdose rate intracavitary brachytherapy, prognostic factors, radiationtherapy, squamous cell carcinoma antigen.

The serum squamous cell carcinoma antigen (SCC-ag)is a tumor-related antigen for which Kato and Torigoe(1)

developed a radioimmunoassay method for humancervical SCC. This antigen, which was obtained fromSCC of the cervix or liver metastasis tissue, hasa molecular weight of approximately 45 kd and is

almost a pure glycoprotein(2). The release of SCC-aginto the serum depends on the infiltrative growth andmass of the tumors(3). A recent description of two ofits genes indicates that SCC-ag may function as a ser-ine/cysteine proteinase inhibitor; its components are

known to be involved in the degradation of the extra-cellular matrix and in tumor invasion andmetastasis(4).

Many publications have confirmed its value as a sen-sitive and specific tumor marker in invasive SCCs ofthe cervix(3,58). Several studies have indicated serum

levels of SCC-ag to be correlated with stage, tumorsize, lymph node involvement, residual tumor aftertreatment, recurrent or progressive disease, and sur-vival in cervical carcinoma(4,920). Although an ele-vated pretreatment SCC-ag level with different cutoffsusing the available positive data was proposed, thepositive data were generally not used to estimate thetreatment outcomes.

The purpose of our study was to determine the pre-treatment SCC-ag level as a generally applicable

Address correspondence and reprint requests to: Ichiro Ogino, MD,Department of Radiology, Yokohama City University, School ofMedicine, 3-9 Fukuura, Kanazawa-ku, Yokohama 236-0004, Japan.Email: [email protected]

#2006, Copyright the AuthorsJournal compilation#2006, IGCS and ESGO

Int J Gynecol Cancer 2006,16, 10941100


2/8

measurement in predicting and estimating the treat-ment outcome in patients with locally advanced squa-mous cell cervical cancer.

Materials and methods

Patient population

Between April 1986 and January 2001, 352 patientswith FIGO stage IIBIVA invasive SCC of the intactuterine cervix were managed with both external irra-diation of the pelvis and highdose rate intracavitary

brachytherapy at Kanagawa Cancer Center. SCC-agwas measured in all patients before radiotherapy, andwe conducted a retrospective review of the patient re-cords. The primary treatment protocol was uniform,and there was no protocol for recurrent patients afterprimary treatment. Treatment other than radiotherapy

was not performed in any patients before recurrence.

Initial clinical assessment

Patients were initially evaluated with a complete his-tory and physical examination as well as a pelvic andrectal examination, blood counts, chemistry profile,chest X-ray, intravenous pyelogram, and computedtomography (CT). Cystoscopy was performed rou-tinely. All patients underwent a biopsy.

CT images were acquired for all patients. Theanteriorposterior (AP) cervix diameter was measured

objectively by CT on a section at the largest portion ofthe cervix(21). Magnetic resonance imaging (MRI) wasalso used to confirm the location of the cervix in somepatients. Lymph nodes greater than 1 cm in greatestdiameter were interpreted as being involved.

Serum samples were collected from each patient forserum SCC-ag level determination within 2 weeks

before treatment initiation. Serum specimens werestored at 220C until the assay. The SCC-ag assay wasperformed within a week after sample collection(1).Serum SCC-ag levels were determined by radioimmu-noassay using an SCC-RIABEAD kit from DainabotCo., Ltd. (Tokyo, Japan). The sensitivity limit of the

assay was 0.1150 ng/mL.

Radiation treatment

The radiotherapy techniques used in these patientshave been reported previously in detail and are de-scribed here only briefly(21,22). All the patients receivedcombination treatment with intracavitary brachytherapyand external irradiation. External irradiation was de-livered to the pelvis using a 10-MV linear accelerator.

The patient was treated through parallel opposed an-teroposterior portals. The external beam daily dosecalculated along the central axis at the midplane of thepelvis was 1.82 Gy. The median total dose given bythe external irradiation was 50 Gy (range: 23.466 Gy),and 328 (93%) patients received external doses be-

tween 45 and 50.4 Gy. A small block was placed toprotect the bladder and rectum after irradiation of 1051.4 Gy (median 28 Gy) when the cervical tumor bur-den was small enough to be cured by intracavitary

brachytherapy. The treatment schedule included aprogram of four external beam fractions per weekcombined with one highdose rate intracavitary bra-chytherapy treatment using 60Co per week. The doseper fraction planned at point A was 56 Gy per weekfor 56 fractions, and the total dose was adjusted byclinical tumor regression. The median intracavitary

brachytherapy dose was 29 Gy (range: 536 Gy).

Follow-up

The median follow-up for all patients was 53 months(range: 2191 months). Clinical examination, Papsmears, serum SCC-ag, and other blood analyses wereperformed every 3 months for 5 years after the com-pletion of therapy and once a year thereafter. A CTscan of the whole abdomen was done before and afterradiotherapy, every year for 2 years, and then indivi-dually scheduled according to clinical findings orelevated serum SCC-ag levels. Chest X-rays were per-formed every 6 months.

Statistical analysis

Statistical analyses were performed using KruskalWallis one-way analysis of variance and the MannWhitney test to examine the relationship betweenSCC-ag levels and other pretreatment factors. Thesenonparametric tests were used because they do notrely on the normality of data distribution and do notrequire equality of variance. Because a cutoff level of2 ng/mL was the most commonly used level in otherreports(4,912,15), we selected this level to compare theclinical outcome of other studies. Median data were

also chosen to identify the optimal cutoff to discrimi-nate low- and high-probability categories. KaplanMeiercurves were used to assess pretreatment variables asa risk predictor and compared using the log-rank test(23).A stepwise Cox proportional hazards model was usedto determine which parameters influenced the proba-

bility of disease-free survival (DFS) and absolutesurvival (AS) in carcinoma of the cervix treated by radi-ation therapy(24). The correlation between the medianSCC-ag of groups distributed by pretreatment prognostic

SCC-ag in locally advanced cervical cancer 1095

#2006 IGCS, International Journal of Gynecological Cancer 16, 10941100


3/8

factors and the 5-year DFS rates was examined graphi-cally and also using Pearsons correlation coefficients(R). Factors with a difference at the 0.05 level wereconsidered significant. Statistics were analyzed usingSPSS 11.0.

Results

The AS and DFS rates of all patients at 5 years ob-tained by KaplanMeier product-limited methodswere 61.3% and 65.7%, respectively. Five-year pelviccontrol obtained by radiation therapy was 81.4%. Thedistribution of possible pretreatment prognostic fac-tors is shown in Table 1. A significantly higher medianSCC-ag was seen in association with increasing FIGOstages (3.0 ng/mL for stage IIB vs 8.1 ng/mL for stageIII vs 19.75 ng/mL for stage IVA, KruskalWallis, v2 39.1 [df 2], P , 0.001). Large tumors, AP cervix size

45 mm, were also associated with a significantlyhigher SCC-ag level (median SCC-ag of 8.85 vs 4.2ng/mL, MannWhitney, P , 0.001). The medianSCC-ag was significantly greater in patients withlymph node involvement, LN (1), than in those with-out lymph node involvement, LN (2) (12.7 vs 5.5ng/mL, MannWhitney,P , 0.001).

Of the 352 patients, 82 were found to have SCC-aglevel ,2 ng/mL and 270 were found to haveSCC-ag level 2 ng/mL. DFS in relation to cutoff ofSCC-ag level at 2 ng/mL is shown in Figure 1. The5-year DFS rate was lower in patients with SCC-aglevel2 ng/mL than in patients with SCC-ag level ,2

ng/mL (60.0% vs 83.3%; P 0.0011). The incidence ofFIGO stages distribution correlated significantlyamong those 82 patients with SCC-ag level below 2ng/mL. There were 40 of 99 patients (40%) with stageIIB, 41 of 239 patients (17%) with stage III, and 1 of 14patients (7%) with stage IVA (KruskalWallis, v2 23.2 [df 2], P , 0.001). When only patients withSCC-ag level below 2 ng/mL were considered, the dif-ference in the 5-year DFS rate between stages IIB andIII patients was not statistically significant (83.7% vs82.5%; P 0.49) (Fig. 2). For patients with SCC-aglevel2 ng/mL, the 5-year DFS rate was higher in pa-

tients with stage IIB than in patients with stage III dis-eases (70.6% vs 58.7%;P0.05) (Fig. 3).

The 5-year KaplanMeier estimates of DFS rate ac-cording to pretreatment prognostic factors are out-lined in Table 2. Age, FIGO stage, SCC-ag level, APcervix size, LN (1), and hydronephrosis were identi-fied as having a significant effect on the DFS in uni-variate analysis.

The 5-year DFS rate was lower in patients receivinga larger external dose without block, or larger total in-tracavitary brachytherapy dose. The 5-year DFS ratewas 52% in patients receiving an external dose with-out block of 28Gy, 80% in patients receiving anexternal dose without block of,28 Gy (P , 0.0001),62% in patients with total intracavitary brachytherapydose of 29 Gy, and 70% in patients with total

Table 1. Distribution of patient characteristics (n 352)

Characteristic Distribution of patients

Age (years)Median 61

Range 2890Stage (FIGO),n(%)

IIB 99 (28)III 239 (68)IVA 14 (4)

SCC-ag (ng/mL)Median 6.0Range 0.5450

AP cervix size (mm)Median 45Range 2090

Lymph node involvementn(%)Negative 286 (81)Positive 66 (19)

Hydronephrosisn(%)Negative 307 (87)Positive 45 (13)

Hematocrit (%)Median 35.7Range 15.151.8

External dose without block (Gy)Median 28Range 1051.4

Intracavitary brachytherapy dose (Gy)Median 29Range 536

Time in Month

60483624120

P

robability

1.0

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0.0

SCC-ag < 2ng/ml N = 82

SCC-ag >=2ng/ml N = 270

P = 0.0011

Figure 1. DFS for patients with SCC-ag level 2 ng/mL and thosewith SCC-ag level ,2 ng/mL.

1096 I. Oginoet al.



4/8

intracavitary brachytherapy dose of ,29 Gy (P 0.044). Radiotherapy doses were excluded from themultivariate analysis because patients with larger tu-mors had poorer prognosis and were more likely toreceive a higher radiation dose. Stepwise Cox multi-variate analysis performed with the covariates out-lined in Table 2 is shown in Table 3. Parameterscarrying significant association with poor outcome interms of DFS included, in decreasing order of signifi-cance, presence of LN (1) (P , 0.001), SCC-ag 6.0ng/mL (P0.004), and hydronephrosis (P0.017).

For AS, FIGO stage (P 0.0013), SCC-ag level (P ,

0.0001), AP cervix size (P 0.0058), LN (1) (P 0.0004), hydronephrosis (P 0.0002), and hematocrit(P 0.0058) were identified as having a significant

effect in univariate analysis. SCC-ag6.0 ng/mL (P0.002) and LN (1) (P 0.012) significantly contrib-uted to AS in stepwise Cox multivariate analysis(Table 4). Multivariate analysis significant testing con-firmed the power of the risk associated with SCC-ag6.0 ng/mL in AS and DFS.

A direct linear relationship existed between medianSCC-ag of groups distributed by pretreatment prog-nostic factors and 5-year DFS rates was obtained fromTable 2 and is shown in Figure 4. The equation of thedirect linear relationship is as follows:

y 22:932x184:896 R 0:965;P , 0:001

In these expressions, y denotes the percentage of5-year DFS rate andxdenotes median SCC-ag.

Time in Month

60483624120

Probability

1.0

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0.0

Stage IIb N = 40

Stage III N = 41

P = 0.49

SCC-ag=2ng/ml

Figure 3. DFS for stages II and III patients with SCC-ag level 2ng/mL. Five-year rates: stage II, 71% and stage III, 59%.

Table 2. DFS at 5 years after radiotherapy

CharacteristicNo. ofpatients

SCC-aglevel (ng/mL)median

5-yearsDFS (%) Pvalue

Age (years),61 164 5.6 61 0.045

61 188 6.95 70Stage

IIB 99 3 76 0.0019III 239 8.1 63IVA 14 19.75 30

SCC-ag (ng/mL),6.0 175 77 ,0.00016.0 177 53

AP cervixsize (mm),45 172 4.2 76 0.000245 180 8.85 56

Lymph nodeinvolvement

Negative 286 5.5 72 ,0.0001Positive 66 12.7 39

HydronephrosisNegative 307 5.5 70 ,0.0001Positive 45 18.3 33

Hematocrit (%),35.7 175 7.2 63 0.1035.7 177 5.4 69

Table 3. Significance of pretreatment parameters affecting DFSin stepwise Cox multivariate analysis

Hazardratio

95% confidenceinterval Pvalue

SCC-ag in ng/mL(,6.0 vs6.0)

1.791 1.2012.671 0.004

AP cervix size in mm(,45 vs45)

1.442 0.9652.154 0.074

Lymph nodeinvolvement

2.075 1.3753.133 0.001

Hydronephrosis 1.742 1.1022.752 0.017




5/8

Stepwise Cox multivariate regression analysis ofDFS was performed with the previously described co-variates, and along with the continuous SCC-ag levelthe results are shown in Table 5. The risk scores r(x)obtained by stepwise Cox multivariate analysis werecombined with the baseline survival functionS(t). The

formula of survival probabilities is as follows:

y Stexprx

In these expressions, y denotes the percentage of5-year DFS rate andx denotes the SCC-ag level.S(t) at5 years was 0.675. The graphs showing the 5-year sur-vival probabilities as a function of SCC-ag withineight groups defined by AP cervix size, lymph nodestatus, and hydronephrosis are shown in Figure 5.

Discussion

FIGO stage, tumor size, and lymph node status havebeen recognized as important predictors of patient

survival. Pretreatment SCC-ag levels have been shownto correlate with clinical FIGO stage, tumor volume,and risk of lymph node metastases.

Strong relationships between pretreatment SCC-aglevels and clinical FIGO stage were reported. In associ-ation with increasing stage, a significantly increased

proportion of SCC-ag levels divided by different cutoffranging from 1.5 to 5 ng/mL was reported(12,13,18). Asignificantly higher median or mean SCC-ag was alsoseen in association with increasing stage(10,11).

A good correlation between the pretreatment SCC-aglevels and the tumor size was reported(12,13,17,18,20).Cervical tumor size was measured using differentmodalities such as bimanual examination, colposcopy,ultrasonography, and surgically removed specimens.Using a multiple logistic model, Takeshimaet al.(20) re-ported that pathologic tumor size and pelvic lymphnode metastasis were found to have a significant effect

Table 4. Significance of pretreatment parameters affecting ASin stepwise Cox multivariate analysis

Hazardratio


Age in years(,61 vs61)

1.327 0.9601.834 0.087

Stage (II vs III vs IV) 1.393 0.9961.947 0.053SCC-ag in ng/mL

(,6.0 vs6.0)1.717 1.2232.412 0.002

Lymph node involvement 1.628 1.1132.382 0.012Hematocrit

(,35.7 vs35.7)0.747 0.5381.037 0.081

3020100

100

90

80

70

60

50

40

30

20

Age

StageAP cervix sizeLymph node involvementHydronephrosis

Hematocrit

y = -2.932x + 84.896R = 0.965P < 0.001

Figure 4. Scatter plot demonstrating high correlation betweenmedian SCC-ag of groups distributed by pretreatment prognostic fac-tors and 5-year DFS rates. Parameters with statistical significance andnonsignificance are depicted by open and solid symbols, respectively.

Table 5. Significance of pretreatment parameters affectingDFS in stepwise Cox multivariate analysis along with the con-tinuous SCC-ag

Hazardratio


SCC-ag in ng/mL

(continuous)

1.004 1.0001.007 0.035

AP cervix size in mm(,45 vs45)

1.541 1.0342.297 0.034

Lymph node involvement 2.072 1.3673.141 0.001Hydronephrosis 1.738 1.0682.828 0.026

4035302520151050

100

90

80

70

60

50

40

30

20

10

0

Group 1:AP45,LN(-),HN(-)Group 3:AP45,LN(+),HN(-)Group 7:AP45,LN(+),HN(+)

Figure 5. The correlation between SCC-ag level and 5-year DFS ratewithin each group defined by AP cervix size, lymph node status,and hydronephrosis.LN(1), with lymph node involvement; LN (2), without lymph nodeinvolvement; AP 45, AP cervix size 45 mm; AP , 45, AP cervixsize ,45 mm; HN (1), with hydronephrosis; HN (2), withouthydronephrosis.

1098 I. Oginoet al.



6/8

on the ratio of serum levels above 4 ng/mL in patientswith stage IB squamous cell cervical cancer undergo-ing hysterectomy. Ohara et al.(17) reported that therewas a significant positive correlation between SCC-agand tumor volumes measured by T2-weighted MRIsin patients with stage IBIVA squamous cell cervical

cancer treated primarily by radiotherapy.Elevated pretreatment serum SCC-ag levels were re-

ported as independent predictors for the presence oflymph node metastases(10,12,16,18,20). Duk et al.(12) re-ported that elevated pretreatment serum SCC-ag lev-els2.0 ng/mL, lesion size40 mm, and presence ofvascular invasion by tumor cells were independentpredictors for the presence of lymph node metastasesin stage IBIIA squamous cell cervical cancer patientstreated by radical hysterectomy. Bolger et al.(10) re-ported the positive predictive value for lymph nodemetastases at .8.6 ng/mL SCC-ag is 100% in stage

IAIIB cervical cancer patients with primary surgicaltreatment. Linet al.(16) reported that around 86% of thepatients with SCC-ag levels below 8 ng/mL showedno nodal metastasis, while about 65% of the patientswith serum levels above 8 ng/mL exhibited nodalmetastasis in stage IB and IIA squamous cell cervicalcarcinoma undergoing radical hysterectomy and pel-vic lymphadenectomy. Only lymph node metastasishad a significant independent impact on SCC-ag lev-els exceeding 8 ng/mL.

Hong et al.(15) reported that the association betweenincreased SCC-ag level and positive lymph nodes de-tected by CT was significant only when the SCC-aglevel was higher than 10 ng/mL after controlling forstage in stage IIVA squamous cell cervical cancerpatient treated primarily by radiotherapy. Assessmentfor lymph node metastasis usually employs CT or MRIin cervical carcinoma treated primarily by radiother-apy. Because variable sensitivity for lymph node metas-tasis in imaging analysis using CT (range: 2465%) orMRI (range: 2471%) compared with surgicopathologicfindings has been reported(2528), correlation betweenthe levels of SCC-ag and the positive lymph nodes isdifficult to evaluate in patients treated by radiotherapyalone, and no surgicopathologic results were obtained.

In our analysis, median SCC-ag was significantlygreater, in association with increasing FIGO stage, inthose patients with AP cervix size 45 mm and inthose patients with lymph node involvement.

The most important clinical correlation for pretreat-ment SCC-ag levels is its ability to predict clinical out-come. There are several reports showing that theserum SCC-ag value is a good prognostic factor in pa-tients with cervical SCC(12,15,19). Duket al.(12) reportedthat among clinicopathologic factors (serum SCC-ag

level, grade, lesion size, stromal infiltration, vascularinvasion, and node status), only pretreatment serumSCC-ag level 2.0 ng/mL had a significant indepen-dent effect on DFS in stage IB and IIA patients withSCC treated by radical hysterectomy. Strausset al.(19)

reported preoperative SCC-ag, using a cutoff value of

3.0 ng/mL, as an independent prognostic factor, bothfor recurrence-free and overall survival in patientswho underwent radical hysterectomy for stagesIAIIB SCC of the cervix. They analyzed prognosticfactors including maximal tumor diameter, cervicalstroma infiltration, pelvic lymph node invasion, para-metrial spread, and tumor grading. Hong et al.(15)

investigated the prognostic significance of the SCC-aglevels in patients with stages IIVA SCC of the cervixprimarily treated by radiotherapy. They concludedthat SCC-ag level higher than 10 ng/mL had a signifi-cant impact on DFS in a multivariate analysis. Stage,

hemoglobin levels (,

10 g/dL), and positive lymphnode shown by CT scan were also independent prog-nostic factors for patient survival.

We previously reported that both CT-measured APcervical tumor size and CT-detected lymph nodemetastases had significant independent effects on ASand DFS(21). The study was performed without analyz-ing SCC-ag in FIGO stage IIBIVA cervical carcinomapatients who received radiation therapy between 1983and 1992. Nonsquamous carcinoma patients wereincluded in the study. In this study, when medianSCC-ag was included in multivariate analyses, medianSCC-ag and CT-detected lymph node metastases hada significant effect on both AS and DFS. Multivariateanalysis confirmed the power of the risk associatedwith SCC-ag6.0 ng/mL.

Although several authors have reported on theprognostic value of elevated baseline levels of SCC-ag

by citing significantly lower survival rates in patientswith highly elevated values compared with those withnormal baseline values, standard pretreatment SCC-aglevel as a pretreatment factor has not been establishedin clinical practice. Evaluation of the clinical stage by

bimanual examination is usually subjective, and it isdifficult to objectively predict treatment results in pa-

tients with squamous cell cervical cancer treated byradiation therapy alone.

Many authors have suggested a cutoff value of 2ng/mL(912,15). In this study, the 5-year DFS rate was83.3% in patients with SCC-ag level ,2 ng/mL. Simi-lar 5-year DFS levels in patients with SCC-ag level ,2ng/mL were reported in stage IB and IIA patientswith SCC treated by radical hysterectomy(12) and inpatients with stage IIVA SCC of the cervix primarilytreated by radiotherapy(15). The 5-year DFS levels in




7/8

their patients with SCC-ag level 2 ng/mL were dif-ferent from ours, but they did not describe the medianSCC-ag of their patients.

In our study, a direct linear relationship was foundbetween median SCC-ag of groups distributed by pre-treatment prognostic factors and the 5-year DFS rate.

We recommend that the median SCC-ag and 5-yearDFS rate of pretreatment factors be reported in addi-tion to the treatment outcome. We can use the graphshown in Figure 5 to read the 5-year survival proba-

bility for individual patients according to SCC-aglevel. From the obtained equations and the graph, wecan objectively estimate treatment outcome in bothgroups of patients and individual patients with locallyadvanced squamous cell cervical cancer treated byradiotherapy.

References1 Kato H, Torigoe T. Radioimmunoassay for tumor antigen of human

cervical squamous cell carcinoma. Cancer 1977;40:16218.2 Ikeda I. Two-site radioimmunometric (sandwich) assay of SCC anti-

gen using monoclonal antibodies. In: Kato H, ed. Squamous cell car-cinoma antigen in the management of squamous cell carcinoma.Princeton, NJ: Excerpta Medica Inc., 1987:21526.

3 Crombach G, Scharl A, Vierbuchen M, Wurz H, Bolte A. Detectionof squamous cell carcinoma antigen in normal squamous epitheliaand in squamous cell carcinomas of the uterine cervix. Cancer1989;63:133742.

4 de Bruijn HW, Duk JM, van der Zee AG et al. The clinical value ofsquamous cell carcinoma antigen in cancer of the uterine cervix.Tumor Biol1998;19:50516.

5 Chou CY, Wang ST, Kuo HC, Tzeng CC, Yao BL. Serum level ofsquamous cell carcinoma antigen and tumor size are useful to iden-tify preoperatively patients at high risk of cervical cancer. Cancer

1994;74:2497501.6 Duk JM, de Bruijn HW, Groenier KH et al. Cancer of the uterine

cervix: sensitivity and specificity of serum squamous cell carcinomaantigen determinations. Gynecol Oncol 1990;39:18694.

7 Maiman M, Feuer G, Fruchter RG, Shaw N, Boyce J. Value of squa-mous cell carcinoma antigen levels in invasive cervical carcinoma.Gynecol Oncol 1989;34:3126.

8 Patsner B, Orr JW Jr, Allmen T. Does preoperative serum squamouscell carcinoma antigen level predict occult extracervical disease inpatients with stage Ib invasive squamous cell carcinoma of the cer-vix? Obstet Gynecol 1989;74:7868.

9 Bae SN, Namkoong SE, Jung JKet al. Prognostic significance of pre-treatment squamous cell carcinoma antigen and carcinoembryonicantigen in squamous cell carcinoma of the uterine cervix. GynecolOncol1997;64:41824.

10 Bolger BS, Dabbas M, Lopes A, Monaghan JM. Prognostic valueof preoperative squamous cell carcinoma antigen level in patientssurgically treated for cervical carcinoma. Gynecol Oncol 1997;65:

30913.

11 Brioschi PA, Bischof P, Delafosse C, Krauer F. Squamous-cell carci-noma antigen (SCC-A) values related to clinical outcome of pre-invasive and invasive cervical carcinoma. Int J Cancer 1991;47:3769.

12 Duk JM, Groenier KH, de Bruijn HW et al. Pretreatment serumsquamous cell carcinoma antigen: a newly identified prognostic fac-tor in early-stage cervical carcinoma. J Clin Oncol 1996;14:1118.

13 Gaarenstroom KN, Kenter GG, Bonfrer JM et al. Can initial serumcyfra 21-1, SCC antigen, and levels in squamous cell cervical cancerpredict lymph node metastases or prognosis? TPA. Gynecol Oncol

2000;77:16470.14 Holloway RW, To A, Moradi M, Boots L, Watson N, Shingleton

HM. Monitoring the course of cervical carcinoma with the squa-mous cell carcinoma serum radioimmunoassay. Obstet Gynecol1989;74:9449.

15 Hong JH, Tsai CS, Chang JT et al. The prognostic significance ofpre- and post-treatment SCC levels in patients with squamous cellcarcinoma of the cervix treated by radiotherapy. Int J Radiat OncolBiol Phys1998;41:82330.

16 Lin H, ChangChien CC, Huang EY, Tseng CW, Eng HL, Huang CC.The role of pretreatment squamous cell carcinoma antigen in pre-dicting nodal metastasis in early stage cervical cancer. Acta ObstetGynecol Scand 2000;79:1404.

17 Ohara K, Tanaka Y, Tsunoda H, Nishida M, Sugahara S, Itai Y.Assessment of cervical cancer radioresponse by serum squamouscell carcinoma antigen and magnetic resonance imaging. ObstetGynecol2002;100:7817.

18 Scambia G, Benedetti Panici P, Foti E et al. Squamous cell carcinoma

antigen: prognostic significance and role in the monitoring ofneoadjuvant chemotherapy response in cervical cancer. J Clin Oncol1994;12:230916.

19 Strauss HG, Laban C, Lautenschlager C, Buchmann J, Schneider I,Koelbl H. SCC antigen in the serum as an independent prognosticfactor in operable squamous cell carcinoma of the cervix. Eur J Cancer2002;38:198791.

20 Takeshima N, Hirai Y, Katase K, Yano K, Yamauchi K, Hasumi K.The value of squamous cell carcinoma antigen as a predictor ofnodal metastasis in cervical cancer. Gynecol Oncol 1998;68:2636.

21 Ogino I, Okamoto N, Andoh K, Kitamura T, Okajima H, MatsubaraS. Analysis of prognostic factors in stage IIBIVA cervical carcinomatreated with radiation therapy: value of computed tomography. IntJ Radiat Oncol Biol Phys1997;37:10717.

22 Ogino I, Kitamura T, Okamoto N et al. Late rectal complication fol-lowing high dose-rate intracavitary brachytherapy in cancer of thecervix. Int J Radiat Oncol Biol Phys 1995;31:72534.

23 Kaplan EL, Meier P. Non parametric estimation from incomplete

observations.J Am Stat Assoc 1958;53:45781.24 Cox DR. Regression model and life tables. J R Stat Soc (B) 1972;

34:187200.25 Kim SH, Choi BI, Han JKet al. Preoperative staging of uterine cer-

vical carcinoma: comparison of CT and MRI in 99 patients. J ComputAssist Tomogr1993;17:63340.

26 Matsukuma K, Tsukamoto N, Matsuyama T, Ono M, Nakano H.Preoperative CT study of lymph nodes in cervical cancerits corre-lation with histological findings. Gynecol Oncol 1989;33:16871.

27 Subak LL, Hricak H, Powell CB, Azizi L, Stern JL. Cervical carci-noma: computed tomography and magnetic resonance imaging forpreoperative staging [Review]. Obstet Gynecol 1995;86:4350.

28 Yang WT, Lam WW, Yu MY, Cheung TH, Metreweli C. Comparisonof dynamic helical CT and dynamic MR imaging in the evaluationof pelvic lymph nodes in cervical carcinoma. Am J Roentgenol2000;175:75966.

Accepted for publication April 15, 2005

1100 I. Oginoet al.



8/8

the role of pretreatment squamous cell carcinoma

Documents