Tumor markers at the time of recurrence in patients with germ cell tumors

Tumor Markers at the Time of Recurrence in Patientswith Germ Cell Tumors

Jose M. Trigo, M.D.1

Jose M. Tabernero, M.D.2

Luis Paz-Ares, M.D., Ph.D.1

Jose L. Garcıa-Llano, M.D.4

Jose Mora, M.D.3

Pilar Lianes, M.D., Ph.D.1

Emilio Esteban, M.D.4

Ramon Salazar, M.D.2

Juan J. Lopez-Lopez, M.D.2

Hernan Cortes-Funes, M.D.l

1 Servicio de Oncologıa Medica Hospital Universi-tario “Doce de Octubre,” Madrid, Spain.

2 Servicio de Oncologıa Medica Clınica HospitalSanta Creu i Sant Pau, Barcelona, Spain.

3 Servicio de Bioquımica Clınica Hospital SantaCreu i Sant Pau, Barcelona, Spain.

4 Servicio de Oncologıa Medica Hospital Central deAsturias, Oviedo, Spain.

Presented in part at the 22nd European Society forMedical Oncology (ESMO) Congress, Vienna, Aus-tria, November 1996.

Address for reprints: Jose M. Trigo, M.D., CRCCentre for Cancer and Therapeutics, The Instituteof Cancer Research, 15 Cotswold Road, E-Block,Belmont, Sutton, Surrey, SM2 5NG, United King-dom.

Received March 29, 1999; revision received Au-gust 9, 1999; accepted August 9, 1999.

BACKGROUND. a-fetoprotein (AFP), human chorionic gonadotropin (HCG), and

lactate dehydrogenase (LDH) closely follow the course of germ cell tumors (GCTs)

and are widely used for diagnosis, prognosis, and follow-up purposes. The objec-

tive of this study was to assess the concordance of tumor markers at the time of

diagnosis and recurrence.

METHODS. The authors reviewed the records of 794 patients with GCTs treated in

three Spanish hospitals from 1977–1996 and analyzed the concordance between

AFP, HCG, and LDH levels at diagnosis and first and second recurrence. A positive

marker was defined as a level of AFP . 10 ng/mL, HCG . 5 IU/L, or LDH . the

upper limit of normal. One hundred twenty-five patients were identified who

developed a first recurrence (123 had marker levels recorded). The median age was

27 years (range, 14 –78 years). Histology was seminoma in 36 patients (29%) and

nonseminomatous GCT (NSGCT) in 87 patients (71%).

RESULTS. Seventy-nine patients (64%) had elevated tumor markers at diagnosis

and 76 (62%) at first recurrence. An elevated marker was present at first recurrence

in 58 of 79 patients (73%) with initially positive markers and in 18 of 44 patients

(41%) with initially negative markers. In 84 of 123 patients (68%), the same marker

pattern (positive or negative) was present at the time of diagnosis and at first

recurrence, 78% in seminomas and 64% in NSGCTs. The earliest indicator of

recurrence was an elevated marker in patients with NSGCTs and a radiologic

finding in patients with seminomas. Thirty patients developed a second recur-

rence, 27 of whom (90%) had the same marker pattern as at first recurrence.

CONCLUSIONS. Tumor marker pattern at diagnosis is not a good predictor of the

pattern at recurrence, particularly in patients with NSGCTs. Marker assessment

should be included in the follow-up schedule regardless of levels at the time of

diagnosis. Early detection of recurrence should not rely only on marker levels, even

in patients with elevated levels at presentation. Cancer 2000;88:162– 8.

© 2000 American Cancer Society.

KEYWORDS: germ cell tumor, a-fetoprotein, human chorionic gonadotropin, lactatedehydrogenase, tumor markers, recurrence, follow-up.

Over the last 30 years a dramatic change in the natural history ofgerm cell tumors (GCTs) has occurred.1 Major contributions to

the advances achieved in the management of this disease include thediscovery and development of effective chemotherapy agents andregimens, the successful design and conduction of clinical trials insingle institutions or as cooperative efforts, the treatment of patientsby experienced multidisciplinary teams, and the availability of threeserum tumor markers: a-fetoprotein (AFP), the B subunit of humanchorionic gonadotropin (HCG), and lactate dehydrogenase (LDH).2

AFP and/or HCG are elevated in up to 80% of patients with nonsemi-nomatous GCTs (NSGCTs),3-6 HCG is elevated in 20 –30% of semi-

162

© 2000 American Cancer Society

noma patients at the time of diagnosis,6-9 and LDH iselevated in up to 80% of patients with advanced met-astatic seminoma.10

Because of the availability of highly sensitive ra-dioimmunoassays for AFP and HCG,11 as well as therecognition that their serum levels closely follow thecourse of the disease, a series of clinical applicationshave emerged. These include the presumptive diagno-sis of a testicular mass,12 staging refinement,13,14 prog-nostic risk allocation,15 surveillance for recurrent dis-ease,4,16,17 and definition of chemotherapy durationand response.18,19

Concordance between tumor marker levels at di-agnosis and recurrence might be clinically relevant inthe follow-up strategies to detect recurrent disease,and also may be of biologic interest as a guide to theclonal selection of metastatic or drug-resistant pheno-types. The aim of our retrospective study was to assessthe concordance of tumor marker levels at the time ofdiagnosis and first and second recurrence.

MATERIALS AND METHODSPatient PopulationParticipating investigators provided retrospective dataregarding the 794 consecutive male patients with aseminoma or NSGCT of gonadal or extragonadal ori-gin who were referred to 3 Spanish hospitals betweenJanuary 1977 and December 1996. Of these, 125 pa-tients (16%) developed recurrent disease after achiev-ing disease free status, 123 of whom had AFP and HCGlevels recorded at the time of diagnosis and recur-rence; 30 patients (4%) developed a second recur-rence. The serum LDH level was available in only 98cases (80%) and its correlation was analyzed sepa-rately. These 123 cases were the subject population ofthe current study. Their characteristics at presentationare presented in Table 1. Patients with histologicallypure seminoma and an elevated AFP level were con-sidered to have NSGCT. Patients who developed asecond GCT or an increase in marker levels not due torecurrent disease were excluded from this study.False-positive elevations of tumor markers were notrecorded.

TreatmentAll tumors were confirmed by histologic examinationof the testis or other primary site at diagnosis andclassified as seminoma if only seminomatous tissuewas present in the specimen. Tumors were classifiedas nonseminomatous if they had one or more non-seminomatous histologic elements with or withoutseminoma. The extent of the disease was classifiedaccording to the Royal Marsden Hospital (RMH) sys-tem after standard workup explorations (medical

history and physical examination; full blood counts;serum biochemistry; AFP, HCG, and LDH determina-tions; and radiologic exploration).

Initial management included a radical orchidec-tomy in the majority of patients. Subsequent manage-ment varied over the study period. Before 1990, pa-tients with Stage I NSGCT either underwent a stagingretroperitoneal lymph node dissection or were treatedby a surveillance protocol policy. Since 1990, thosepatients with adverse local factors in the surgical spec-imen after orchidectomy were treated with adjuvantcisplatin-based chemotherapy. Patients with NSGCTof Stages II–IV were treated with cisplatin-based che-motherapy plus residual mass surgery if present. Pa-tients with Stage I or IIA–IIB seminoma initially weretreated with paraaortic lymph node irradiation. Since1990, Stage I seminoma patients were submitted to asurveillance policy protocol, with the exception ofthose with adverse local factors, who were treated withadjuvant cisplatin or carboplatin-based chemother-

TABLE 1Patient Characteristics at Presentation (794 Cases Reviewed)

Characteristics

Firstrecurrence

Secondrecurrence

No. % No. %

No. of patients with markers recorded 123 16 30 4Age (yrs)

Median 27 27.5Range 14–78 15–57

Stage at presentation (RMH)I 52 42 8 27II 26 21 10 33III 10 8 2 7IV 35 29 10 33

HistologySeminomaa 36 29 11 37Nonseminomatous GCT 87 71 19 63

Embryonal carcinoma 26 21 5 17Teratocarcinoma 19 15 3 10Choriocarcinoma 7 6 3 10Endodermal sinus tumor 7 6 2 7Teratoma 9 7 4 13Mixed 16 13 2 7Undifferentiated 3 2 0 0

Previous treatment†Surveillance 44 36 7 23Chemotherapy 66 54 18 60Radiotherapy 5 4 1 3Surgery 2 2 0 0Combination 6 5 4 13

Extragonadal 7 6 3 10

RMH: Royal Marsden Hospital; GCT: germ cell tumor.a a-fetoprotein not included.b Orchiectomy not included.

Tumor Markers in Germ Cell Tumors/Trigo et al. 163

apy. Patients with Stage IIC seminoma initially weretreated with radiotherapy, and since 1985 with cispla-tin or carboplatin-based chemotherapy. Patients withStage III–IV seminomas were treated with cisplatin-based chemotherapy.

Follow-UpFollow-up was performed according to standardguidelines. During the first 3 months of follow-up,physical examination, blood tests, tumor marker de-terminations, and chest radiography were performedmonthly. During the first year, patients underwentthese studies together with ultrasound or computedtomography (CT) scan (since 1981), which were per-formed every 3 months during the first year, every 4months during the second year, every 6 months dur-ing the third year, and annually thereafter. A physicalexamination of the contralateral testis was performedat every visit and a testicular ultrasound scan wasperformed annually.

Marker DeterminationsSerum AFP and HCG levels were determined by stan-dard radioimmunoassay techniques. All values . 10ng/mL for AFP and . 5 IU/L for HCG were consideredpositive.

Statistical AnalysisStandard statistical descriptive parameters (enumera-tions, proportions, medians, and ranges) were used tocharacterize the data. The significance of the differ-ences between proportions was assessed by the Fisherexact test. Differences with a P value , 0.05 wereconsidered statistically significant. Disease free sur-vival (DFS) and overall survival (OS) were calculatedfrom the date of diagnosis using the Kaplan–Meierproduct-limit method20 and the statistical analysis ofobserved differences was assessed by the log ranktest.21

RESULTSFirst RecurrenceThe diagnosis of recurrence was made after a mediandisease free interval of 6 months (range, 1–185months). The most frequent site of recurrence was theretroperitoneum. The earliest indicator of recurrencewas an increase in tumor markers in patients withNSGCTs and abnormal radiologic explorations in pa-tients with seminomas. An elevated marker (AFP,HCG, or LDH) was the only initial evidence of recur-rent disease in 35 of 87 NSGCT patients (40%) and in9 of 36 seminoma patients (25%) (Table 2).

Among the 123 recurrent patients, 79 (64%) pre-sented with increased levels of AFP or HCG at diagno-sis and 76 (62%) at first recurrence. These tumormarkers were present at the time of first recurrence in58 of 79 patients (73%) who had positive markers atdiagnosis and in 18 of 44 patients (41%) who hadnegative markers at diagnosis. In 84 of 123 patients(68%) the same marker pattern (positive or negative)was present at the time of diagnosis and at first recur-rence (Table 3).

Among the 36 patients with seminoma, 12 (33%)had elevated HCG levels at diagnosis (Table 3). Ten ofthese patients (83%) had an elevated HCG level at firstrecurrence. AFP (4 patients) and/or HCG (4 patients)were found to be elevated in 6 of 24 patients (25%)who initially had negative markers (Table 4). In 28 of36 patients (78%) the same marker pattern (positive ornegative) was present at the time of diagnosis and atfirst recurrence.

Among patients with NSGCTs, 48 of 67 (72%) whoinitially had positive and 12 of 20 (60%) who initiallyhad negative AFP or HCG levels presented with ele-vated markers at first recurrence (Table 3). Only 56 of87 patients (64%) showed the same marker pattern(positive or negative) at diagnosis and at first recur-rence. The correlation of marker levels at the time ofdiagnosis and recurrence for each specific tumormarker is shown in Table 4. An elevated AFP level at

TABLE 2Earliest Indicator of First Recurrence

Indicator

All cases Seminoma NSGCT

No. % No. % No. %

Increased marker 44 36 9 25 35 40Radiologic explorations 38 31 15 42 23 26Clinical signs or symptoms 10 8 4 11 6 7Combination 31 25 8 22 23 26

NSGCT: nonseminomatous germ cell tumor.

164 CANCER January 1, 2000 / Volume 88 / Number 1

first recurrence was observed in 30 of 62 patients(48%) and in 10 of 25 patients (40%) with elevated andnormal AFP levels at the time of diagnosis, respec-tively. HCG levels at recurrence were elevated in 21 of47 patients (45%) and in 11 of 40 patients (28%) withelevated and normal HCG levels at diagnosis, respec-tively.

Serum LDH was elevated at the time of diagnosisin 16 of 30 patients (53%) and 18 of 68 patients (26%)with seminoma and NSGCTs, respectively. SerumLDH was elevated at the time of recurrence in 14 of 16patients (87%) and in 14 of 18 patients (78%) who hadelevated LDH levels at the time of diagnosis, and in 10

of 14 patients (71%) and 17 of 50 patients (34%) whohad normal levels at presentation, respectively.

We analyzed the influence of other variables onthe marker status at the time of recurrence in NSGCTs.Elevated levels of AFP and/or HCG at recurrence wereobserved more frequently in those patients with ad-vanced disease stage (Stages II–IV) at diagnosis (41 of53 patients; 77%) versus those with Stage I disease (19of 34 patients; 56%) (P 5 0.034), and in those with ashorter disease free interval (# 6 months) (34 of 43patients; 79%) versus those with a longer disease freeinterval (26 of 44 patients; 59%) (P 5 0.044).

The extension of the disease at the time of recur-rence was influenced by the marker pattern (positiveor negative) at the time; patients with positive markersat recurrence were detected at earlier clinical stagesthan patients with negative markers. Among the 76patients with positive markers at recurrence, 12 (16%)had marker elevations as the only evidence of recur-rence, 46 (61%) were found to have Stage II disease,and 18 (24%) were found to have Stage III or IV dis-ease. Among the 47 patients with negative markers atthe time of recurrence, 32 (68%) were found to haveStage II disease, and 15 (32%) were found to haveStage III or IV disease (P 5 0.014). These stage discrep-ancies were higher in the 41 patients whose recur-rence first was diagnosed only by tumor markers ver-sus the remaining 82 patients (elevated tumor markersonly: 29% vs. 0%; Stage II disease: 56% vs. 67%; andStage III or IV disease: 15% vs. 33%; P , 0.0001). Weobserved no significant differences in DFS or OS basedon the marker pattern at recurrence. However, recur-rences in NSGCTs were detected earlier in patientswhose first indicator of recurrence was an increase intumor markers compared with the remaining patients(5 vs. 8 months; P 5 0.022).

Second RecurrenceA diagnosis of second recurrence was established in 30of 123 patients (24%) who developed a first recurrenceafter a median DFS of 10 months (range, 0 – 48months). The most frequent site of recurrence alsowas the retroperitoneum. An increase in tumor mark-ers was the only initial evidence of recurrent disease in11 of 19 NSGCT patients (58%) and in 3 of 11 patientswith seminoma (27%) and was a major contributor tothis diagnosis in 19 of 30 patients (63%).

AFP and/or HCG were elevated in 24 patients(80%) at second recurrence, including 22 of 23 pa-tients (96%) who had positive markers at the time offirst recurrence and only 2 of 7 patients (29%) who hadnegative markers at first recurrence (data not shown).In 10 of 11 patients with seminomas (91%) and 17 of19 patients with NSGCTs (89%), the same marker pat-

TABLE 3Correlation between Marker Status (AFP and HCG) at Diagnosis andFirst Recurrence for All Cases, Seminomas, and NSGCTs

Markers at first recurrence

Positive Negative Overall

All casesMarkers at diagnosis Positive 58 21 79

Negative 18 26 44Overall 76 47 123

SeminomasMarkers at diagnosis Positive 10 2 12


NSGCTsMarkers at diagnosis Positive 48 19 67


AFP: a-fetoprotein; HCG: human chorionic gonadotropin; NSGCTs: nonseminomatous germ cell tu-

mors.

TABLE 4Correlation between Specific Positive Tumor Markers (AFP and HCG)at Diagnosis and First Recurrence for Seminomas and NSGCTs

Positive markers at firstrecurrence

AFP HCG Both None

SeminomasPositive markers at diagnosis AFP 0 0 0 0

HCG 0 10 0 2AFP 1 HCG 0 0 0 0None 2 2 2 18

NSGCTsPositive markers at diagnosis AFP 10 1 2 7

HCG 1 3 1 0AFP 1 HCG 13 12 5 12None 4 4 4 8

AFP: a-fetoprotein; HCG: human chorionic gonadotropin; NSGCTs: nonseminomatous germ cell tu-

mors.


tern (positive or negative) was present at the time ofboth first and second recurrence. AFP and HCG atsecond recurrence were elevated in 82% (9 of 11) and80% (12 of 15) of the patients with increased levels ofeach marker at first recurrence, respectively.

DISCUSSIONAFP, HCG, and LDH are highly specific tumor markersand are of incalculable value in monitoring the clinicalcourse of GCTs, especially those of nonseminomatoushistology, from the time of diagnosis to posttreatmentfollow-up. We believe there is no doubt regarding theability of these markers to detect recurrent disease insome patients, and all follow-up schedules for diseasefree patients include serial and frequent serum deter-minations of AFP, HCG, and LDH. Surprisingly, to ourknowledge there are very little data in the literaturedefining the correlation of tumor markers in this set-ting, particularly in patients treated with chemother-apy with or without adjunctive surgery. The currentstudy shows that tumor marker elevation is a majorcontributor to the detection of recurrent GCT in .60% of the patients. Moreover, it was the only sign ofrecurrence in 33% of the patients and was present in ahigher proportion of those patients with NSGCTs.Only 68% of the patients in the current study (84 of123) had the same marker pattern (positive or nega-tive) at the time of diagnosis and at recurrence. Thiscorrelation is not strong enough to reject continuousmarker monitoring in initially marker negative pa-tients or to avoid other follow-up explorations (CTscans, chest X-rays, etc.) in those patients with ele-vated AFP and/or HCG levels at diagnosis.

The predictive value of AFP and HCG in the diag-nosis of testicular masses is well established.6-12 Pos-torchidectomy determinations are of great signifi-cance in patients with early and advanced disease,and the management of clinical Stage I NSGCTs withnormal postoperative markers should be chosen de-pending on the policy and expertise of the treatingcenter, histologic prognostic factors, and patient pref-erences.16,17 Conversely, the treatment of those pa-tients with persistently elevated markers (clinicalStage IM) should be selected with the certainty thatmicrometastatic disease exists in the retroperito-neum.13,17 The levels of AFP, HCG, and LDH haveunquestioned prognostic relevance, which had led toa new staging system for the disease and an interna-tionally accepted consensus regarding risk alloca-tion.15 During treatment, the monitoring of tumormarkers is essential for response assessment andmanagement decision-making.18,19

To our knowledge the correlation of tumor mark-ers in the follow-up of GCTs after achievement of

disease free status, although proven, has not beenevaluated as exhaustively as in the contexts men-tioned earlier. In the current series of 123 patients,62% had elevated AFP and/or HCG levels at first re-currence, with a higher proportion of these patientscomprised of NSGCT patients (69%) compared withseminoma patients (44%), as was expected.3-9 Overall,these figures are very similar to those at the time ofdiagnosis, with 77% of NSGCT patients and 33% ofseminoma patients having elevated markers. In an-other large series from the RMH with 166 recurrentNSGCT patients, 73% had abnormal markers at thetime of diagnosis and 63% at first recurrence.4 Tumormarkers were increased more frequently at the time ofsecond recurrence (80%) than at first recurrence(62%); this may reflect the higher proportion of NS-GCTs, but also treatment selection of clones withmarker production potential.

GCT is one of the few malignancies in which re-current patients still have a reasonable chance of cure.Furthermore, factors directly related to an early diag-nosis of recurrent disease such as tumor bulk, diseasesite, and AFP or HCG levels have a clear impact onsurvival, regardless of the stage at diagnosis and thetreatment previously given.22,23 In the current study,the levels of AFP and HCG were determinants of anearly diagnosis of tumor recurrence in 56% of thecases (Table 2). More important, in 36% of the patientsthe only sign of first recurrence was an elevation inmarker levels (40% of NSGCTs and 25% of semino-mas). Two surveillance series of patients with Stage INSGCT after orchidectomy offer similar rates of recur-rence based on tumor markers, alone or combinedwith other methods.24,25 In the current series, patientswith recurrent disease diagnosed only by tumor mark-ers had earlier disease stages than those diagnosed byother methods (77% vs. 68%) as pointed out in otherclinical contexts.26

Given that not all patients have a positive markerat recurrence, one of the objectives of this study was toassess the predictive value of several variables at thetime of diagnosis, including AFP and HCG levels, inthe marker status (positive or negative) at the time ofrecurrence. Overall, 73% of the patients with positivemarkers at diagnosis and 59% of those with negativemarkers at diagnosis had the same pattern at firstrecurrence. The marker concordance was better inseminoma patients (78%) because only 25% of the 24patients with initially normal HCG developed a recur-rence with elevated levels of HCG or AFP, and 10 of 12patients (83%) with an elevated HCG level at diagnosishad the same marker elevated at the time of recur-rence. Mirimanoff et al.27 reported increased HCG lev-els at the time of recurrence in 5 of 6 patients (83%)


with abnormal pretreatment marker values. Thesedata support the ability of tumor markers, particularlyHCG, to detect seminomatous activity in patients dur-ing follow-up. Its discriminatory power appears to berelated strongly to the marker status at diagnosis.

In the current study serum LDH levels provideduseful information and complemented the routinemeasurement of AFP and HCG in the management ofpatients with GCT, especially those with seminomas.

The lower concordance of marker status ob-served between the time of diagnosis and first re-currence (64%) in NSCGT patients most likely is dueto a similar prevalence of elevated markers at recur-rence among patients with normal levels (60%) andthose with elevated levels (72%) at diagnosis. In thecurrent analysis AFP and HCG equally contributedto the inconsistency of marker levels during theevolution of NSGCTs. A possible explanation for thisfinding is the uncertainty regarding the pathogene-sis of this disease and its heterogeneity, particularlyin NSGCTs, in which marker-producing clonogensmay form a highly variable component of diseasevolume in patients who respond to or develop arecurrence after treatment. Similarly, drug-resistantcells may be a small proportion of the total tumorburden at the start of chemotherapy. The pattern ofmarker change may be determined by the resistantcomponent, and this might explain why marker sta-tus at the time of second recurrence correlates bet-ter with the status at the time of previous recur-rence. However, changes in tumor marker patternbetween the time of diagnosis and recurrence hadno significant impact on clinical outcome.

The current study identified other subsets of pa-tients with a higher likelihood of elevated AFP and/orHCG levels at the time of recurrence, regardless oftheir marker status at the time of diagnosis. Thesepatient subsets were comprised of those with an ini-tially advanced stage of disease (Stages II–IV by theRMH classification) and those with a shorter diseasefree interval (# 6 months).

Although an increase in tumor markers usually isthe first indication of recurrence, marker status at thetime of diagnosis appears to be a poor indicator ofmarker status at the time of recurrence. In those pa-tients with elevated tumor markers at the time ofdiagnosis, radiologic exploration (CT scan and/or ul-trasound scan) should continue to be performedthroughout the follow-up period because positivemarker values are presumptive evidence of recur-rence. However, a negative marker value should notbe used to rule out recurrent disease. In contrast, inthose patients with negative tumor markers at diag-

nosis, AFP and HCG levels should continue to bemonitored to detect recurrent disease.

REFERENCES1. Einhorn LH. Karnofsky Memorial Lecture. Treatment of tes-

ticular cancer: a new and improved model. J Clin Oncol1990;8:1777– 81.

2. Kurman RJ, Scardino PT, McIntire KR, Waldmann TA, Ja-vadpour N. Cellular localization of alpha-fetoprotein andhuman chorionic gonadotropin in germ cell tumors of thetestis using an indirect immunoperoxidase technique. Anew approach to classification utilizing tumor markers.Cancer 1977;40:2136 –51.

3. Bosl GJ, Geller NL, Cirrincione C, Nisselbaum J, Vugrin D,Whitmore WF Jr., et al. Serum tumor markers in patientswith metastatic germ cell tumors of the testis. A 10-yearexperience. Am J Med 1983;75:29 –35.

4. Price A, Nicholls J, Norman A, Dearnaley D, Horwich A.Tumour markers at relapse in patients with non-seminoma-tous germ cell tumors. Ann Oncol 1994;5(Suppl 8):63.

5. Lange PH, McIntire KR, Waldmann TA, Hakala TR, FraleyEE. Serum alpha fetoprotein and human chorionic gonad-otropin in the diagnosis and management of nonsemino-matous germ-cell testicular cancer. N Engl J Med 1976;295:1237– 40.

6. Bates SE, Longo DL. Use of serum tumor markers in cancerdiagnosis and management. Semin Oncol 1987;14:102–38.

7. Ball D, Barret A, Peckham MJ. The management of meta-static seminoma testis. Cancer 1982;50:2289 –94.

8. Weissbach L, Bussar-Maatz R. HCG-positive seminoma. EurUrol 1993;23(Suppl 2):29 –32.

9. Mauch P, Weichselbaum R, Botnick L. The significance ofpositive chorionic gonadotropins in apparently pure semi-noma of the testis. Int J Radiat Oncol Biol Phys 1979;5:887–9.

10. Fossa A, Fossa SD. Serum lactate dehydrogenase and hu-man choriogonadotrophin in seminoma. Br J Urol 1989;63:408 –15.

11. Waldmann TA, McIntire KR. The use of radioimmunoassayfor alpha-fetoprotein in the diagnosis of malignancy. Cancer1974;34(Suppl 4):1510 –5.

12. Moore MR, Garrett PR Jr., Walton KN, Waldmann TA, McIn-tire RK, Vogel CL. Evaluation of human chorionic gonado-tropin and alpha fetoprotein in benign and malignant tes-ticular disorders. Surg Gynecol Obstet 1978;147:167–74.

13. Klepp O, Flodgren P, Maartman-Moe H, Lindholm CE, Un-sgaard B, Teigum H, et al. Early clinical stages (CS1,CS1Mk1 and CS2) of non-seminomatous testis cancer.Value of pre- and post-orchiectomy serum tumor markerinformation in prediction of retroperitoneal lymph nodemetastases. Swedish-Norwegian Testicular Cancer Project(SWENOTECA). Ann Oncol 1990;1:281– 8.

14. Raghavan D, Sullivan AL, Peckham MJ, Neville AM. Elevatedserum alphafetoprotein and seminoma: clinical evidence fora histologic continuum? Cancer 1982;50:982–9.

15. The International Germ Cell Cancer Collaborative Group.International Germ Cell Consensus Classification: a prog-nostic factor-based staging system for metastatic germ cellcancers. J Clin Oncol 1997;15:594 – 603.

16. Rorth M, Cullen MH, Horwich A, Roth BJ. Management ofpatients with non-seminomatous germ cell tumours stage I.Prog Clin Biol Res 1990;357:295–303.


17. Cullen M. Management of stage I non-seminoma: surveil-lance and chemotherapy. In: Horwich A, editor. Testicularcancer: investigation and management. London: Chapmanand Hall Medical, 1991:149 – 68.

18. Bartlett NL, Freiha FS, Torti FM. Serum markers in germ cellneoplasms. Hematol Oncol Clin North Am 1991;5:1245– 60.

19. Toner GC, Panicek DM, Heelan RT, Geller NL, Lin SY, Ba-jorin D, et al. Adjunctive surgery after chemotherapy fornonseminomatous germ cell tumors: recommendations forpatient selection. J Clin Oncol 1990;8:1683–94.

20. Kaplan EL, Meier P. Nonparametric estimation from incom-plete observations. J Am Stat Assoc 1958;53:457– 81.

21. Peto R, Peto J. Asymptotically efficient rank invariant testprocedures. J R Stat Soc A 1972;35:185–206.

22. Gerl A, Clemm C, Schmeller N, Hartenstein R, Lamerz R,Wilmanns W. Prognosis after salvage treatment for un-selected male patients with germ cell tumours. Br J Cancer1995;72:1026 –32.

23. Horwich A, A’Hearn R, Gildersleve J, Dearnaley DP. Prog-nostic factor analysis of conventional dose salvage ther-apy of patients with metastatic nonseminomatous germcell cancer. Proc Annu Meet Am Soc Clin Oncol 1993;12:232.

24. Williams MP, Husband JE, Heron CW. Stage I nonsemino-matous germ cell tumors of the testis: radiologic follow-upafter orchidectomy. Radiology 1987;164:671– 4.

25. Thompson PI, Nixon J, Harvey VJ. Disease relapse in pa-tients with stage I nonseminomatous germ cell tumor of thetestis on active surveillance. J Clin Oncol 1988;6:1597– 603.

26. Seckl MJ, Rustin GJ, Bagshawe KD. Frequency of serumtumour marker monitoring in patients with non-seminoma-tous germ cell tumours. Br J Cancer 1990;61:916 – 8.

27. Mirimanoff RO, Shipley WU, Dosoretz DE, Meyer JE. Pureseminoma of the testis: the results of radiation therapy inpatients with elevated human chorionic gonadotropin titers.J Urol 1985;134:1124 – 6.


Tumor markers at the time of recurrence in patients with germ cell tumors

Documents

Transcript of Tumor markers at the time of recurrence in patients with germ cell tumors