the role of surgical management in pediatric germ cell tumors (gcts), nci case series

ABSTRACT

Purpose: To review the experience of a tertiary referralcenter in pediatric germ cell tumors (GCTs) in the last 8years and to investigate the impact of surgery and site ofdisease on prognosis.

Patients and Methods: We retrospectively analyzedthe cases of pediatric germ cell tumors at National CancerInstitute over an 8 years period. Data concerning diagnosis,surgery and medical decisions were reviewed and analyzedfor all patients. A total of 34 children with (GCTS) werefound, with a mean age, at presentation, of 6.7 years anda follow-up period ranging from 3-52 months. One patientwith benign GCT was excluded during analysis of theresults.

Results: Among the 34 patients, there were 14 malesand 20 females with mean age of 6.7 years (range: 9months-15 years), with male to female ratio 1:1.4. Allpatients were symptomatic at presentation, most commonlywith abdominal swelling (18 patients; 52.9%). Anatomicdistribution of GCTs according to sex organ involvementwas either gonadal in 21 patients (61.8%) or extragonadalin 13 patients (38.2%).

All patients had surgery either in the form of curativeresection or biopsy after formal exploration and evidenceof irresectability. No significant surgical morbidity ormortality were encountered in our patients. Yolk sac tumorand malignant teratoma were the commonest histologicsubtypes in our series. Metastatic disease was encounteredin nine out of 33 patients (27.2%). Adjuvant chemotherapywas administered in 28 out of 33 patients (84.8%), fol-lowing surgery, including all patients with extragonadaldisease.

Our patients were followed-up to 52 months. Twenty-two patients (66.7%) had no recurrence while two patients(6.1%) died from disease. Pelvic extragonadal site wasthe worst site regarding resectability. Complete surgicalresection showed better disease free survival, while those

Journal of the Egyptian Nat. Cancer Inst., Vol. 20, No. 1, March: 70-79, 2008

The Role of Surgical Management in Pediatric Germ Cell Tumors(GCTs), NCI Case Series

IBRAHIM M.Y. FAKHR, M.D., M.R.C.S.*; EL-SAYED ASHRAF KHALIL, M.D., F.R.C.S.I.*;TAREK S. EL-BARADIE, M.D.*; MOHAMED A. SHAALAN, M.D.*; LOBNA M. SHALABY, M.D.**;SHIMAA L.A. NASSIF, M.Sc.** and IMAN G. FARAHAT, M.D.***

The Departments of Surgical Oncology*, Pediatric Oncology** and Pathology***,National Cancer Institute (NCI), Cairo University.

70

with irresectable disease had comparable overall survivalwhile none could be rendered disease free with chemo-therapy.

Conclusion: The initial surgical approach to malignantGCTs at all sites should be complete resection whenpossible; the morbidity of extensive surgical resectionshould be weighed carefully against the good tumor controlwith chemotherapy. Surgical staging does not precludepreservation of fertility, which should always be consideredin this young age. The site of primary disease plays a rolein the prognosis of pediatric germ cell tumors with theextragonadal pelvic tumors being the worst regardingresectability. Good tumor response can be achieved withsurgery and chemotherapy even for advanced stage andmetastatic disease.

Key Words: Germ cell tumors – Childhood solid tumors– management – Surgery.

INTRODUCTION

Gonadal and extragonadal germ cell tumorsare infrequent in children representing approx-imately 3% of cancers diagnosed in personsyounger than 15 years [1,2]. Germ cell tumorsare presumed to share a common cell of origin,the primordial germ cell, yet they remain aheterogeneous group of tumors. Variations re-garding age, sites of presentation, histopathologyand malignant potential stem from the differ-ences in the stage of germ cell development attumorigenesis, the differences in the tumorenvironment secondary to the gender of thepatient and to the location of the clone and theoccurrence of specific genetic aberrations [3].Sacrococcygeal teratomas are the most commongerm cell tumors of childhood, accounting for40% of all and up to 78% of extragonadal germcell tumors. They also are the most frequentlyrecognized neoplasm of fetuses, with approxi-

Correspondence: Ibrahim Fakhr, MD, [email protected]

71

mately 75% of patients are female [4]. Gonadaltumors come next, followed by the rest of theextragonadal sites, namely: mediastinal, retro-peritoneal, brain and multiple other rare sitessuch as neck, oral cavity and stomach [5].

Germ cell tumors show numerous histologicsubtypes. The histologic features of each subtypeare independent of presenting clinical charac-teristics; tumor biology and clinical behaviorvary with site of origin, stage and age of thepatient. For example, mature teratoma in theinfants and in the ovary are almost invariablydiploid and benign, whereas those in the adulttestis that show the same histologic features areaneuploid and potentially malignant [6].

Clinically, ovarian tumors account for only1% of childhood malignancies with peak of agearound 19 years. They present with an abdom-inal pain in 80% of patients (most of the latterhave associated ovarian torsion) followed byabdominal mass or vaginal bleeding. On theother hand pediatric testicular tumors accountfor 2% of solid malignant neoplasm in boyswith almost all identified as irregular non tenderscrotal masses [7].

The outcome for patients with malignantGCTs was poor before the advent of systemicchemotherapy, with 3-year survival rates ofapproximately 20% [8,9]. The introduction ofcisplatin-based therapy improved the outcomefor patients with localized tumors, but patientswith advanced disease continued to have a verypoor prognosis [10-13].

Surgical resection is the therapy of choicein benign tumors, such as teratomas. With ma-lignant lesions surgical resection is indicated ifpossible. However, given the availability ofeffective chemotherapy, resection should notbe undertaken to the point of sacrificing vitalstructures. In such situation only debulking orbiopsy is appropriate. After initial chemotherapysecond-look surgery with intent of completeresection serves to assist in achieving completeresponse in selected patients [14].

PATIENTS AND METHODS

This retrospective study included all pediat-ric patients with malignant germ cell tumors(GCTs) recorded at the National Cancer Insti-tute, Cairo University from January 2000 toDecember 2007.

A total of thirty four patients were foundwith a follow-up period ranging between 10and 52 months. Data was retrieved from NCIhospital-based registry. Demographic and tumor-specific data was reviewed and analyzed. Itincluded clinical data examination (age, symp-toms and signs at diagnosis) and tumor markers.Diagnostic imaging evaluation included chestradiograph, ultrasonography and trunk computedtomography, or magnetic resonance imaging toassess the anatomical site of the disease and aspart of the metastatic work up.

The type of chemotherapy delivered to pa-tients differed with time and stage of the diseaseat diagnosis: In low risk disease (stage I testic-ular and ovarian tumors), no chemotherapy wasgiven with postoperative observation only. Pa-tients with moderate risk gonadal tumors orprogression of disease in untreated tumors weremanaged adequately with three to four cyclesof a platinum containing regimen. For higherrisk patients (higher stage testicular or ovariantumors and extragonadal tumors), four to sixcycles of a platinum based chemotherapeuticregimen were indicated. The type of chemother-apy delivered to patients with favorable tumorsvaried with time and included one of theseprotocols: cisplatin (platinol), vinblastine andbleomycin (PVB); cisplatin, etoposide and ble-omycin (PEB) and carboplatin, etoposide andbleomycin (JEB). Pediatric patients with unfa-vorable tumors received chemotherapy combi-nation etoposide, ifosfamide, carboplatin orcisplatin. We classified chemotherapy as optimalif its delivery was scheduled each 3 weeks.

Statistical method:All patients were evaluated for the following

response criteria: Event-free survival (EFS) andoverall survival (OS) at 3 years. The EFS wascalculated from the date of diagnosis to the dateof first relapse or the date of last free follow-up. The OS was calculated from the day ofinitial diagnosis until death or last date of news.Survival curves were estimated by the Kaplan-Meier method.

RESULTS

Demographic features:A total of 34 children diagnosed with germ

cell tumors were enrolled in the present study.Their mean age at presentation was 6.7 years(range: 9 months-15 years). Among the 34 en-

The Role of Surgical Management in Pediatric Germ Cell Tumors (GCTs)

Ibrahim M.Y. Fakhr, et al.72

rolled patients; there was a female predominancewith a male to female ratio of 1:1.4 with 14(41.2%) males and 20 (58.8%) females. Maleshad a mean age at presentation of 5.9 years(range 9 months-13 years) versus a mean ageof 7.8 years (range 1-15 years) in females.Difference was statically not significant (pvalue=0.30).

Clinical manifestation at presentation:The most common presentation was abdom-

inal swelling and/or abdominal pain whichoccurred in 21 patients. Less common clinicalmanifestations included: testicular swelling in9 patients, respiratory symptoms (cough anddyspnea) in 4 patients. The least common clin-ical presentations were: proptosis, a visiblelump or mass at the buttocks and vaginal bleed-ing found in a single patient each. None of ourpatients presented with manifestations of hor-mone overproduction or precocious puberty.

For the whole group, the median lag timefrom first complaint to accurate diagnosis ofour patients was 4 months (range 1-11 months).Table (1) shows patients clinical presentation(some patients presented with more than onesymptom).

Diagnosis and initial management:Diagnosis was confirmed after full

preoperative evaluation and work up includingimaging techniques which were used accordingto the suspected site either gonadal or extragonadal, with 21 patients (61.8%) with gonadalinvolvement and only 13 patients (38.2%) atthe extragonadal sites.

Table (2) shows the anatomic distributionof GCTs in all groups. The most common sitewas the ovary in 12 patients (35%), followedby the pelvic in 10 patients (29.2%), then thetestis in 9 patients (26%), mediastinal (Picture1), sacrococcygeal and orbit in a single patienteach (3%).

The tumor markers, AFP and HCG, weremeasured in all cases prior, following surgeryand during the whole period of follow-up.

Histopathology:The most common histologic subtypes were

yolk sac tumor and malignant teratoma, whichwere diagnosed in 13 patients (38.2%) each,followed by mature teratoma, dysgerminomaand embryonal carcinoma in two patients (5.9%)each. Lastly, benign mature cystic teratoma andmixed GCTs in only one patient (2.9%) each.All GCTs were malignant except for the onepatient with benign mature teratoma (2.9%)(Table 3).

Staging and surgical management:All patients, in our study, had surgery either

in the form of curative complete resection (24patients) or a biopsy after formal explorationwith evidence of irresectability (9 patients).The outcomes of the different surgical proce-dures were listed in Table (4). Different surgicalprocedures were done based on the site of originof GCTs being either gonadal or extragonadaland according to the gender of patients.

Gonadal tumors: Twenty patients were in-cluded (one female patient with benign teratomawho had an ovariectomy was excluded fromour statistical analysis). Eight male patientswith testicular tumors underwent radical unilat-eral orchiectomy with high ligation and en blocexcision of the spermatic cord structures andtesticle; alone in 4 patients (20.0%) and withadditional procedures in the other 4 patients(20.0%). Hemi-scrotalectomy was added in 2patients (10.0%) for suspected invasion. A para-aortic lymph node dissection (PALND) wasperformed in one patient (5.0%) with para-

Table (1): Clinical manifestations of pediatric germ celltumors at presentation.

Symptom

Abdominal swelling/painTesticular swellingCough and dyspneaProptosisMass in the buttocksVaginal bleeding

(%)

61.726.511.72.92.92.9

No.

2194111

Table (2): Anatomic distribution of GCTs in all patients.

Site

Gonadal:OvaryTestis

Extragonadal:PelvicSacrococcygealMediastinalOrbit

Total

Males Females Total

No.

21129

1310111

34

(%)

(35.4)(35.4)(0.0)

(23.6)(17.6)(3.0)(0.0)(3.0)

(59.0)

No.

12120

86101

20

(%)

(26.4)(0.0)(26.4)

(14.6)(11.6)(0.0)(3.0)(0.0)

(41.0)

No.

909

54010

14

(%)

(61.8)(35.4)(26.4)

(38.2)(29.2)(3.0)(3.0)(3.0)

(100.0)

73

aortic lymph node enlargement. Lastly, a singlepatient (5.0%) had excision of brain metastasesin addition to orchiectomy for solitary brainmetastases. Only one patient (5.0%) had irre-sectable retroperitoneal nodal disease and or-chiectomy was done as a biopsy.

Regarding the 11 female patients, five pa-tients (25.0%) were staged as FIGO Stage I andthey all had ovariectomy. Two patients (10.0%)had additional appendectomies, for grosslyabnormal appendices. Three patients (15.0%)had an additional infracolic omentectomy andone patient had a superadded formal panhys-trectomy with the infracolic omentectomy(5.0%). Only one patient with ovarian GCTshad peritoneal metastases and ascites (5.0%).None of the 11 female patients had LN metastas-es.

Extragonadal tumors: Thirteen patients wereincluded. A successful complete resection wasachieved in 5 (38.5%) patients: Three (23.1%)patients with pelviabdominal masses (Picture2), among whom one (7.7%) needed a colosto-

my, a single (7.7%) patient with mediastinalmass who needed a lower lobectomy and a lastpatient who had an eye enucleation (7.7%).Eight patients (61.5%) had irrresectable lesions.Six patients (46.1%) with pelvic lesions amongwhom one patient needed a colostomy, onepatient (7.9%) with a lesion at the sacrococ-cygeal area and lastly another patient with apelvic lesion of no definite origin (7.9%). Nosignificant surgical morbidity or mortality wasencountered in our patients. Only one of the 13patients (7.7%) with the extragonadal tumorshad para-aortic lymph node metastases.

Different factors that might be related toresectability were analyzed. Only extragonadalorigin in the pelvis site was statistically signif-icant for irresectability; as tumors of extrago-nadal origin in the pelvis proved to be lessresectable (8/9 patients, 61.5%; 7/9 patients,70%, respectively) than gonadal origin and non-pelvic tumor site at the other sites (p<0.001)(Table 5). Other factors as sex and pathologicalsubtype were not statistically significant.


Table (3): Distribution of histologic subtypes at anatomic sites.

Histologic Subtype

Yolk sac tumor

Malignant Teratoma

Mature Teratoma

Dysgerminoma

Embryonal Carcinoma

Mixed GCT

Benign Mature Cystic Teratoma

Total

Males FemalesTotal

Anatomic site

7331

8521

11

22

11

0

11

20

No.

OvaryPelvicSacrococcygeal

OvaryPelvicOrbit

Ovary

Ovary

Pelvic

Ovary

OrganNo.

651

5221

11

0

11

11

0

14

Organ

TestisPelvic

TestisPelvicMediastinal

Testis

Testis

Pelvic

No. (%)

13 (38.2)

13 (38.2)

2 (5.9)

2 (5.9)

2 (5.9)

1 (2.9)

1 (2.9)

34 (100.0)


Lymph node (LN) metastases were encoun-tered in only 6 out of 33 patients, patients(18.2%). Five patients (83.4%) had testicularGCTs and one patient (86.6%) had pelvic mass.Paraaortic LNs were the most commonly affect-ed (5 patients; 86.6%), however, inguinal (3patients; 50.0%), iliac (3 patients; 50.0%) andmediastinal (1 patient; 16.7%) lymph nodesaffection was also encountered. Moreover 4 outof 6 patients (66.6%) with LN metastases hadalso associated distant metastases.

Distant metastases were encountered in 9patients (27.2%). Five (55.5%) patients hadgonadal and 4 (44.5%) had extragonadal tumors.Lung metastases occurred in 6 out of 9 meta-

static patients (66.6%) either alone or associatedwith spleen and adrenal gland deposits or evenwide spread metastases. One ovarian GCT hadascites and another one with testicular GCT hadbrain metastases (Table 6).

Adjuvant chemotherapy was given to 28patients (84.9%) with advanced tumors aftersurgery including 13 patients with extragonadaldisease (Picture 3). Chemotherapy related mor-bidity was encountered in two patients (7.2%).The first patient had holoxan encephalopathyand the patient eventually died; the secondpatient had avascular necrosis of the head ofthe femur and was treated conservatively.

Table (4): Surgical procedures and outcome in our series.

Surgical Procedure(s)

Complete resection

In Males:Isolated OrchiectomyOrchiectomy + hemi-scrotalectomyOrchiectomy + Para-aortic LNDOrchiectomy + Brain metastatectomyWLE + Lower lung lobectomyWLE + Stoma (colostomy)WLE

In Females:Isolated OvariectomyOvariectomy + Infracolic omentectomyOvariectomy + AppendectomyOvariectomy + PanhystrectomyWLE + Eye enucleationWLE

Irresectable:In MalesIn Females

Total

WLE = Wide local excision. LND = Lymph node dissection.

Extragonadal

5 (38.5)

3 (23.1)0 (0.0)0 (0.0)0 (0.0)0 (0.0)1 (7.7)1 (7.7)1 (7.7)

2 (15.4)0 (0.0)0 (0.0)0 (0.0)0 (0.0)1 (7.7)1 (7.7)

8 (61.5)2 (15.4)6 (46.1)

13 (100.0)

No. (%)

Gonadal

19 (95.0)

8 (40.0)4 (20.0)2 (10.0)1 (5.0)1 (5.0)0 (0.0)0 (0.0)0 (0.0)

11 (55.0)5 (25.0)3 (15.0)2 (10.0)1 (5.0)0 (0.0)0 (0.0)

1 (5.0)1 (5.0)0 (0.0)

20 (100.0)

No. (%)

Total

No. (%)

24 (72.7)

11 (33.3)

13 (39.4)

9 (27.3)3 (9.1)6 (18.2)

33 (100.0)

Table (6): Anatomical site of metastases.

Site(s) of distant metastases

PulmonaryLung & spleenLung & adrenal glandLung & wide spread diseasePeritoneal implants & ascitesBrain

Total

Number

311121

9

Percent (%)

(33.3)(11.1)(11.1)(11.1)(22.2)(11.1)

(100.0)

Table (5): Pelvic (extragonadal) site as a determinant forirresectability.

Factors

Tumor site:ExtragonadalPelvic tumors

Non-ExtragonadalPelvic tumors*

*This group includes all gonadal tumors and extrapelvic extrag-onadal tumors.

ResectabilityNo.

10

23

Yes (n=24)

3 (30.0%)

21 (91.3%)

No (n=9)

7 (70.0%)

2 (8.7%)

p-value

<0.001

75

Most of the patients received Bleomycin,Etoposide & Cisplatin (BEP) (25 patients;89.3%), 3 of them (10.7%) received, 2nd lineVinblastin, Ifosfamide & Cisplatin (VIP), while2 others (7.1%) received Ifosfamide, Carboplatin& Etoposide (JEB). Two other patients (7.1%)received Ifosfamide, one with additional Eto-poside and the other with Cisplatin. The latterhad VIP as 2nd line. A single patient (3.6%)received Carboplatin, Etoposide & Bleomycin(JEB) (Table 7).

Only one patient (3.0%) received adjuvantradiotherapy as first-line treatment after excisionof a solitary brain metastasis. Five stage I pa-tients, (15.1%) were only watched (3 stage Iaand 2 stage Ic) without any adjuvant treatment.

Outcome and prognostic factors:The duration of follow-up ranged from 3 to

52 months (Mean 18.5 months; median 14months). Among the 33 patients, 18 (54.5%)remained in first complete remission; all ofthem were with surgically resectable disease.Ten patients (30.3%) had controlled residualdisease. Three patients (9.1%) had progressivedisease. Finally we had two (6.1%) disease-specific mortalities, one in the surgically resect-

ed group and the second in the non-resectablegroup. The overall cumulative survival at 3years was 92.0% and the event free survival at3 years was 28.2%. The resectable group tumorshad a borderline better EFS (p=0.118) than theirresectable ones; this might be explained bythe small number of patients. As for the site ofdisease (pelvis Vs. extra-pelvic) it had no effecton EFS (Table 8).

As a final outcome in the surgically resectedpatients’ group, 95.8% of patients were alive;75.0% being free of disease. The disease per-sisted in 5 patients (20.8%), being controlledin 3 patients (12.5%) and progressive in theother 2 patients (8.3%). One patient (4.2%) dieddue to holoxan induced encephalopathy (Table9).

As a final outcome in the surgically resectedpatients’ group, 90.0% of patients were alive.However, none of them could be rendered freeof disease by chemotherapy; the disease beingstable in 8 patients (80.0%) and progressive inonly one patient (10.0%). A single mortality(10.0%) from the disease was recorded (Table9).


Table (7): Chemotherapy: Protocols used and outcome of treatment.

Chemotherapy

BEPBEP → VIPBEP → ICEJEBIFX-Plat → VIPIFX-VP16

Total

Total CR PR SD

(%)

(30.0)(33.3)(50.0)(0.0)(100.0)(100.0)

(35.7)

No.

611011

10

(%)

(10.0)(66.7)*(50.0)(0.0)(0.0)(0.0)

(17.9)

No.

221000

5

(%)

(60.0)(0.0)(0.0)(100.0)(0.0)(0.0)

(46.4)

No.

1200100

13

(%)

(71.4)(10.7)(7.1)(3.6)(3.6)(3.6)

(100)

No.

2032111

28

IFXPlatVP16JEB*Dead patients.

= Ifosfamide.= Cisplatin.= Etoposide.= Carboplatin, Etoposide & Bleomycin.

CRPRSDBEPVIPICE

= Complete remission.= Progression of disease.= Stable disease.= Bleomycin, Etoposide & Cisplatin.= Vinblastin, Ifosfamide & Cisplatin.= Ifosfamide, Carboplatin & Etoposide.

Table (8): Patients’ survival at 3 years.

Factor

EFS = Event-free survival.

Overall survival

EFS*:ResectableIrresectablePelvic siteExtra-pelvic site

p-value

0.118

0.498

95% CI

–

(12.45 - 75.55)–

(13.79 - 22.21)(13.78 - 30.22)

–

Cumulativesurvival

92.0%

28.2%74.0%

–40.0%65.5%

No.

33

332491023

Mediansurvival ± SE

–

44.00±16.1044 ± –18.00±2.1522.00±4.20

–


Table (9): Outcome of patients in relation to surgical resectability.

Patients outcome

Living:Free of diseaseWith stable diseaseWith progressive disease

Death

No. (%)

9 (90.0)0 (0.0)8 (80.0)1 (10.0)

1 (10.0)

Resectable (n=24)

23 (95.8)18 (75.0)3 (12.5)2 (8.3)

1 (4.2)

No. (%)

Non-Resectable (n=10)

Picture (1): CT of the mediastinal mass at the level of the main bronchi.

Picture (2): CT of an abdominal GCT mass; (A) Preoperative; (B) Post-resection.

Picture (3): CT of a para-aortic LNs mass; (A) Pre-chemotherapy; (B) Post-chemotherapy residual.

77

DISCUSSION

Malignant germ cell tumors (GCTs) areamong the potentially curable malignancies inchildhood. They have special characteristics;they possibly share the same tissue of originbut they involve different anatomic sites. Al-though they are not so common tumors theyform a group of tumors of variable histologicalsubtypes that needs collaborative work betweensurgeons and pediatricians working in the fieldof pediatric oncology to face the challengepresented by these neoplasms which is mainlyto control the tumor while minimizing toxicityand maintaining future fertility.

Although most of the GCTs in childhoodperiod are benign, this was not the case withour reviewed series as 33 out of 34 patients(97%) could definitely be classified as havingmalignant neoplasms. This discrepancy in thefrequency of malignancies could be attributedto the fact that common benign tumors werebeen managed by the pediatric surgeon whilepatients who had uncommon disease or thoseviewed as possibly requiring multimodal ther-apies were referred to the NCI as a tertiaryreferral center. Prior to recent chemotherapeuticsuccesses the 10-year survival rate for malignantgerm cell tumor ranged between 25% for em-bryonal carcinoma to 75% for dysgerminoma[15,16]. Now the overall prognosis of malignantgerm cell tumors has improved dramaticallywith the advent of platinum-based therapy withoverall survival rates exceeding 80% [17,18].

Clinical presentation of GCTs depends onthe site of origin of the tumor. Symptoms aremainly swelling either abdominal or testicularin gonadal GCTs, or pain and pressure symptomsin extragonadal sites. In many series; sacrococ-cygeal teratomas is reported to be the mostcommon site for GCTs in childhood. It may bediagnosed prenatally as an incidental ultrasono-graphic finding at delivery, few weeks afterbirth or later [19-21].

In our series gonadal and extragonadal tu-mors were encountered in 61% and 39% of thecases respectively. Of the latter group one case(3%) had sacrococcygeal teratoma; this uncom-mon relative frequency in the pediatric agegroup could be attributed to the fact that thosechildren were managed by the general pediatricsurgeon.

Different imaging modalities are used forthe diagnosis of GCTs including ultrasonography(US), computed tomography (CT) and magneticresonance imaging (MRI). These tools allowsfor accurate diagnosis, proper staging and fol-low-up of the tumors. Also the development ofsensitive assays for the tumor markers; AFPand HCG, has improved diagnostic accuracyand monitoring of therapy for their relatedsubtypes of malignant germ cell tumors [22].Positron emission tomography (PET) has beenrecently introduced to the management of GCTswhere it has been particularly valuable in de-tection of relapses [23].

Surgery remains the main stay in the initialmanagement of GCTs in different sites of origin.Accurate diagnosis and staging has great impacton the decision and extent of surgery. They alsoavoid unnecessary and extensive resections withpotential risks of morbidity and mortality. Totalgross resection of the tumor has always beenpursued. The tumor and the involved adjacentstructure should be resected en block if this ispossible and does not lead to disfigurement,major morbidity, or permanent disability [24].

In our study, we followed the same surgicalprinciples in the treatment of patients. Attemptat surgical resection was done in all patientswhich was feasible and complete in 24 patients,and not feasible in 9 patients where only biopsywith intra-operative staging was performed.The main issue withholding the surgeon fromperforming complete resection was the necessityfor cystectomy or pelvic exentration whichcould not be justified by the surgeon weighingthe resultant lifelong morbidity against thepotential response of tumor to chemotherapyand possible subsequent salvage surgery. Inaddition impaction in the pelvis and invasionof the lateral pelvic walls obscuring the majorpelvic vessels were other determinant pointsfor irresectability. In this series none of thepatients had salvage surgery after chemotherapyfor irresectable disease but most of the patientshad stable disease with chemotherapy. Fewpatients in this group suffered from progressivedisease which included development of me-tastases; a condition that reflects the aggressivebiological behavior that may have also lead toirresectability. Another critical issue in themanagement of GCTs in the pediatric age groupis the preservation of fertility. This issue logi-



cally concerns young females more than youngmales. Because most girls with malignant ova-rian GCTs are young and the prognosis for earlydisease is excellent preservation of childbearingcapacity is desired. Unilateral salpingo-oopherectomy with preservation of the uterusand contralateral ovary, if these organs appearnormal, is strongly advised provided that thiswill not compromise a curative resection of thedisease. In our series fertility was preserved in10 out of 11 female patients with gonadal GCTswhile only one patient had panhystrectomy forbilateral gonadal affection. On the other handnone of the 9 male patients with gonadal tumorshad bilateral orchiectomy.

Different studies proved that the introductionof cisplatinum chemotherapy and current ad-vances in the surgical treatment have resultedin a dramatic improvement of the prognosis ofchildren with malignant germ cell tumors(GCTs). Cisplatinum chemotherapy generallyresults in sufficient systemic tumor control butlocal relapses may still occur in patients whodid not receive adequate local treatment. There-fore the therapeutic consideration must takeinto account age, primary site of the tumor andits histology [24,25]. In gonadal tumors there isa high chance of primary complete resectionsince these tumors tend to be encapsulated.Moreover testicular GCTs are particularly moreoften detected at a low tumor stage. In contrasta primary complete resection may be impossiblein large non-gonadal tumors such as sacrococ-cygeal or mediastinal GCTs. In these tumors aneoadjuvant or preoperative chemotherapy afterclinical diagnosis by imaging and evaluationof tumor markers significantly facilitates com-plete resection during a delayed surgery. Inaddition the impact of chemotherapy on localtumor control may be enhanced by locoregionalhyperthermia [26]. In most intracranial GCTscomplete resection is impossible and may beassociated with significant morbidity. Neverthe-less biopsy is essential for diagnosis in non-secreting tumors. In intracranial GCTs radio-therapy significantly contributes to local tumorcontrol and doses are stratified according to thehistology [27]. These general considerationshave been integrated into many national andinternational cooperative treatment protocols.In most current protocols treatment is stratifiedaccording to an initial risk assessment thatincludes the parameters age, site, histology,

stage, completeness of resection and the tumormarkers alpha-1-fetoprotein (α-1-FP) and beta-human chorionic gonadotropin (β-HCG) [21-25,28].

In our study, adjuvant chemotherapy wasgiven to patients with malignant or irresectabletumors. Follow-up and statistical analysis ofour data from the first time each patient wasclinically examined at the NCI until the firstoccurrence of disease progression, relapse, ordeath, or until last reported contact if no eventsoccurred, showed an overall survival of 92%.This is comparable to the reported rates foroverall survival although our series includedmore patients with malignant tumors and tumorsat difficult anatomical locations. It should betaken into consideration that survival drops withtime and a larger group of patients is requiredfor a longer follow-up period to match theresults of international series. The current re-sponse to chemotherapy in irresectable diseasehighlights the possibility of a further survivalimprovement through the application of thenewly emerging protocols with their elaboratedneoadjuvant chemotherapy role. Hopefully, thiswill facilitate and allow for a planned subsequentsalvage complete resection.

Conclusion:The initial surgical approach to malignant

GCTs at all sites should be complete resectionwhen possible; the morbidity of extensive sur-gical resection should be weighed carefullyagainst the good tumor control with chemother-apy. Surgical staging does not preclude preser-vation of fertility which should always be con-sidered in this young age. The site of primarydisease plays a role in the prognosis of pediatricgerm cell tumors with the extragonadal pelvictumors being the worst regarding resectability.Good tumor response can be achieved withsurgery and cisplatinum-based combinationchemotherapy even for advanced stage andmetastatic disease. On the other hand low-riskpatients are treated expectantly after curativesurgery and spared the toxicity of chemotherapy.More thought and future investigation shouldbe directed to the role of salvage surgery afterchemotherapy for locally advanced GCTs.

REFERENCES1- Bernste in L, Smith MA, Liu L. Germ cell, Tropho-

blastic and other Gonadal Neoplasms, in Ries LAGSM, Gurney JG, Linet M, Tamra T, Young JL, Bunin

79

GR (eds): Cancer Incidence and Survival AmongChildren and Adolescents: United States SEER Pro-gram 1975-1995, National Cancer Institute, SEERProgram. Bethesda, Cancer Statistics Branch CancerSurveillance Research Program Division of CancerControl and Population Sciences National CancerInstitute, Bethesda, MD. 1999, pp 125-38.

2- MiIler RW, Young JL Jr, Novakrovic B. ChildhoodCancer. Cancer. 1995, 75: 395-405.

3- Mulligan RM. Pathogenesis of Teratoid Tumors ofthe Ovary and Testis. Pathol Annu. 1975, 10: 271-98.

4- Dehner LP. Gonadal and Extragonadal Germ CellNeoplasia of Childhood. Hum Pathol. 1983, 14: 493-511.

5- Marina N, Fontanesi J, Kun L. Treatment of ChildhoodGerm Cell Tumors. Review of the St. Jude Experiencefrom 1979 to 1988. Cancer. 1992, 70: 2568-75.

6- Mann JR, Pearson D, Barrett A. Results of the UnitedKingdom Children Cancer Study Group’s MalignantGerm Cell Tumor Studies. Cancer. 1989, 63: 1657-67.

7- Chretien PB, Milam JD, Foote FW. Embryonal Ade-nocarcinoma (a Type of Malignant Teratoma) of theSacrococcygeal Region: Clinical and Pathologic As-pects of 21 Cases. Cancer. 1970, 26: 522-35. DeborahFB: Germ cell tumors, Surg Clin N Am. 2006, 86:489-503.

8- Kurman RJ, Norris HJ. Endodermal Sinus Tumor ofthe Ovary: A clinical and pathologic analysis of 71cases. Cancer. 1976, 38: 2404-19.

9- Slayton RE, Hreshchyshyn MM, Sliverberg SC.Treatment of Malignant Ovarian Germ Cell Tumors:Response to Vincristine, Dactinomycin and Cyclo-phosphamide (preliminary report). Cancer. 1978, 42:390-8.

10- Slayton RE, Park RC, Sliverberg SG. Vincristine,Dactinomycin and Cyclophosphamide in The Treat-ment of Malignant Germ Cell Tumors of the Ovary.A Gynecologic Oncology group Study (a final report).Cancer. 1985, 56: 243-8.

11- Canger A, Smith J, van Eys J. Improved Prognosis inChildren with Ovarian Cancers Following ModifiedVAC (vincristine sulfate, dactinomycin and cyclophos-phamide) chemotherapy. Cancer. 1978, 42: 1234-8.

12- Brodeur GM, Howarth CB, Pratt CB. Malignant GermCell Tumors in 57 Children and Adolescents. Cancer.1981, 48: 1890-8.

13- Einhorn LH, Donohue J. Cis–Diamminedichloroplat-inum, Vinblastine and Bleomycin Combination Che-motherapy in Disseminated Testicular Cancer. AnnInt Med. 1977, 87: 293-98.

14- Mann JR, Raafat F, Robinson K. The United KingdomChildren’s Cancer Study Group’s Second Germ CellTumor Study: Carboplatin, Etoposide and Bleomycinare Effective Treatment for Children with MalignantExtracranial Germ Cell Tumors, With AcceptableToxicity. J Clin Oncol. 2000, 18: 3809-18.

15- Gobel U, Schneider DT, Calaminus G. Germ CellTumors in Children and Adolescence. GPOH MAKEIand the MAHO Study Groups. Ann Oncol. 2000, 11:263-71.

16- Cushing B, Giller R, Cullen JW. Randomized Com-parison of Combination Chemotherapy with Etoposide,Bleomycin and Either High-Dose Cisplatinum inChildren and Adolescents with High-Risk MalignantGerm Cell Tumors: A Pediatric Intergroup Study-Pediatric Oncology Group 9049 and Children’s CancerGroup 8882. J Clin Ocol. 2004, 22: 2691-2700.

17- Ciftci AO, Bingol-Kologu M, Senocak ME. TesticularTumors in Children. J Pediatr Surg. 2001, 36 (12):1796-801.

18- Baranzilli MC, Kramer A, Boufet E. Prognostic Factorsin Children With Localized Malignant Non–Semi-nomatous Germ Cell Tumors. J Clin Oncol. 1999, 17:1212.

19- Schmidt B, Habberlik A, Uray E. SacrococcygealTeratoma: Clinical Course and Prognosis with a Spe-cial View to Long Term Functional Results. PediatrSurg Int. 1999, 15: 573-6.

20- Westerberg, B, Fedstein VS, Sandberg PL. SonographicPrognostic Factors in Fetuses with SacrococcygealTeratoma. J Pediatr Surg. 2000, 35 (2): 322-6.

21- Hedrick HL, Flake AW, Cromleholme TM. Sacrococ-cygeal Teratoma: Prenatal Assessment, Fetal Interver-sion and Outcome. J Pediatr Surg. 2004, 39 (3): 430-8.

22- Billmire D, Vinocur C, Rescorla F. Outcome andStaging Evaluation in Malignant Germ Cell Tumorsof The Ovary in Children and Adolescents: An Inter-group Study. J Pediatr Surg. 2004, 39 (3): 424-9.

23- Hain SF, O'Doherty MJ, Timothy AR. Fluorodeoxy-glucose Positron Emission Tomography in the Evalu-ation of Germ Cell Tumors at Relapse. Br J CancerOct. 2000, 83 (7): 863-9.

24- Marina NM, Cushing B, Giller R. Complete SurgicalExcision is Effective Treatment for Children withImmature Teratomas with or Without Malignant Ele-ments: A Pediatric Oncology Group/Children's CancerGroup Intergroup Study. J Clin Oncol Jul. 1999, 17(7): 2137-43.

25- Marina N, Wendy B, Frazier L, Lauer S, Rescorla F,Cushing B, et al. Prognostic Factors in Children withExtragonadal Malignant Tumors: A Pediatric Inter-group Study. J Clin Oncol Jul. 2006, 24 (16): 2544-8.

26- Gobel U, Schnider D, calaminus G, Haas R, SchmidtP, Harms D. Germ Cell Tumors in Childhood andAdolescence. Ann Oncol. 2000, 11: 263-71.

27- Maria E, Eschevarria, Fangusaro J, Goldman S. Pedi-atric Central Nervous System Germ Cell Tumors: AReview. The Oncologist. 2008, 13: 690-99.

28- Sonpavde G, Thomas E, Bruce J. Management ofRecurrent Testicular Tumors. The Oncologist. 2007,12: 51-61.


the role of surgical management in pediatric germ cell tumors (gcts), nci case series

Documents