| ObjectiveGerm cell tumors (GCTs) from the central nervous system (CNS), usually called intracranial germ cell tumors(ICGCTs) are traditionally classified into germinoma and nongeminomatous germ cell tumors (NGGCTs). Although the pathologic subtype is the most important factor to predict prognosis, normalized diagnostic and therapeutic algorithm as early as possible is also very important. A multiple-disciplinary diagnostic platform is the basis for early diagnosis (in half a year from onset of disease), and histopathological or cytological confirmation by surgical or cerebrospinal fluid (CSF) management is the most important part of it. Tumor markers are indispensible in this platform, however, their clinical values have not been comprehensively investigated. The current several kinds of diagnostic criterior of "secreting" tumors is not established on the basis of histopathology or cytology from lumber CSF, and the tumor markers are only limited to non germinoma specific β subunit of human chorionicgonadotropin (β-hCG) and alpha-fetoprotein (αFP), while the value of some germinoma-specific tumor markers, such as placental alkaline phosphatase (PLAP) and CD117(a stem cell factor receptor with several variants:the transmembrane form as "c-kit" and the soluble form as "s-kit"), has not been routinely measured and investigated further. Therefore, a multiple-disciplinary platform is needed to be established to evaluate tumor markers (β-HCG, αFP, PLAP and s-kit) in lumber CSF and serum in the diagnosis of CNS GCTs based on histopathology and CSF cytology.Methods(1) The study cohort was divided into two groups according to the date of their diagnoses. The PAST group included32patients with intracranial GCTs diagnosed by histopathology or CSF cytology between March1991and September2011. The PRESENT group included26patients, whose diagnoses were made based on the multiple-disciplinary diagnostic platform between September2011and December2012. Demographic information, diagnosis, duration from first symptom to diagnosis, immunohistology of tumor markers and lumber CSF cytology and level of tumor markers in CSF and serum were compared between the two groups.(2)To investigate the synchronized relationship of0-hCG and a FP between CSF and serum; to evaluate several different diagnostic criteria in current use; a receiver operating characteristic (ROC) curve was made by using patients with langerhans cell histiocytosis (LCH)(n=12) and other intracranial tumors(n=5) as controls. Based on this curve, new cutoff value for β-hCG and a FP in diagnosis of GCTs were investigated; to evaluate tumor markers in early diagnosis of GCTs in male patients with incomplete precocious puberty/latent hypogonadotropic hyperandrogenemia.(3) The levels of PLAP and s-kit in CSF and serum were measured by ELISA. The difference between PRESENT group, LCH group, other intracranial tumors group and hippocampal sclerosis group (n=5) were compared. ROC curve was used to calculate the reasonable cutoff value of PLAP and s-kit for diagnosis and the new diagnostic criteria was evaluated.(4) All data were analyzed with SPSS11.5. Independent t test or One-Way ANOVA were used to compare differences between two groups for normal distributed variable and non-normal distributed variable (logarithmic transformed). Wilcoxon signed-ranks test or chi-square test were used for categorical variables; ROC curves were used to calculate the cutoff value for diagnosis; P<0.05is considered as statistically significant.Results1. In PAST group and PRESENT group, the ratios of male:female were1.1:1and1.9:1respectively, the average ages of diagnosis were19+7years and13±4years respectively. Incidence rates of isolated sella mass were59.4%and38.5%, and incidence rates of isolated pineal mass were3.1%and15.4%respectively. The rates of diagnosis confirmed by surgical biopsy were71.9%and96.2%respectively. In the final pathological diagnosis of all patients, the incidence of germinomas decreased from90.6%to65.4%. Immunohistochemistry for four tumor markers was stained in9.4%-46.9%patients in PAST group, while it was conducted in100%of PRESENT group; In PAST group, β-hCG and a FP in CSF and serum were measured in50%-57.7%patients, while it increased to100%in PRESENT group; Early diagnosis rate increased from6.3%in PAST group to19.2%in PRESENT group (though chi-square test P>0.05).2. The total sensitivities of CSFand serum β-hCG>50IU/L as diagnostic criteria were18.8%and30.8%in PRESENT group and PAST group respectively, while the specificities were both100%; the sensitivities of α FP>10ng/mL as diagnostic criteria were13.3%and23.1%in each group respectively, while the specificities were66.7%and83.3%respectively; although the sensitivities of αFP>25ng/mL as diagnostic criteria were6.3%and19.2%in each group respectively, while the specificities were both100%.3. In ROC curve analysis, CSFβ-hCG≥8.4IU/L had a sensitivity of44%, serum β-hCG≥2.2IU/L had a sensitivity of33.3%and both had a specificity of100%; although the number of patients with positive a FP was not enough, the diagnostic cretria for CSF a FP could be set at3.8 ng/mL and ICGCTs could be effectively diagnosed, no case was found with serum a FP less than25ng/ml.4. The diagnostic sensitivity of CSF β-hCG increased from26.9%in PAST group to61.5%in PRESENT group, the sensitivity of CSFβ-hCG was higher than that of serum one, which was42.3%(mainly for diagnosis of tumors including STGC); the diagnostic sensitivity of CSF a FP improved from7.7%in PAST group to19.2%in PRESENT group; The total diagnostic sensitivity in PRESENT group increased from38.5%to65.4%, mainly due to CSF β-hCG.5. The incidence rate of male patients with latent hypogonadotropic hayperandrogenemia was17.6%in PAST group and the incidence rate of male patients with precocious puberty or latent hypogonadotropic hayperandrogenemia was52.9%in PRESENT group caused by high serum level of β-hCG.6. In PAST and PRESENT groups,66.7%and16.7%of the patients were finally properly diagnosed as mixed GCTs based on the supplementary information from CSF/serum a FP.7. BBB impairment was demonstrated by higher serum levels of tumor markers than CSF levels. The following diagnostic criteria were used as indication for impaired BBB:serum β-hCG≥2.2IU/L and the concentration in serum is no less than in CSF, or serum a FP≥25ng/mL and the concentration in serum is no less than in CSF; when both criteria were met, impairment of BBB was severe:in PAST group, the incidence of severe impaired BBB was0%; in PRESENT group, the patients of BBB impairment demonstrated by β-hCG and a FP took up18.8%(3cases) and100%(1case) respectively, there was1patient with severely impaired BBB, which took up6.3%of β-hCG positive patients and20%of a FP positive patients. 8. Except in1case of lung cancer metastasized to the sellar area the level of CSF PLAP was significantly elevated (1684pg/mL), the levels were all under the minimum detected value in all studied patients. There was no significant difference of serum s-kit between the studied group and the hippocampus sclerosis group (40.5±9.7ng/mL vs36.1±12.1ng/mL, P>0.05); when CSF s-kit was≥3.0ng/mL, the diagnostic sensitivity and specificity for germinoma was30.8%and100%, respectively, the total diagnosis sensitivity can be increased from64.6%to76.9%with the addition of CSF s-kit in PRESENT group.Conclusions1. A multiple-disciplinary diagnostic platform has been established for the early diagnosis of CNS GCTs, the basis of the platform is histopathology and CSF cytology.2. CSF/serum β-hCG>50IU/L and/or a FP>25ng/mL are confirmed to be more reasonable diagnostic criteria for secreting intracranial GCTs in our study.3. A new diagnostic cutoff point for intracranial GCTs was made:CSF β-hCG≥8.4IU/L, serum β-hCG≥2.2IU/L; CSF a FP cutoff point is decreased to3.8ng/mL and serum a FP unchanged;4. CSF or serum β-hCG is valuable for early diagnosis in precocious boy or latent hypogonadotropic hyperandrogenemic male. CSF and serum a FP may provide additional value to the final diagnosis of mixed GCTs; β-hCG and a FP may help to evaluate the severity of BBB impairment.5. CSF PLAP had no value for diagnosis of intracranial GCTs;6. CSF s-kit was helpful for the diagnosis of pure germinoma and clinical diagnostic cutoff point is3.0ng/mL. |
PDF Full Text Request