Correlations between presence of spontaneous lesions of the pituitary (adenohypophysis) and plasma prolactin concentration in aged Wistar rats

The predictive value of elevated plasma prolactin concentrations for the presence of spontaneous pituitary lesions was studied in 40 male and 38 female Wistar (Cpb:WU) rats, all 30 months old. The pituitaries were examined light microscopically and stained for prolactin using immunohistochemical methods. Plasma prolactin concentrations were measured by radioimmunoassay. Pituitary lesions were classified on the basis of their morphology in hematoxylin and eosin-stained sections as foci of hypertrophic or hyperplastic cells and hemorrhagic, pleomorphic, or spongiocytic adenomas; no carcinomas were found. There were significantly (P = 0.001) more female than male rats with pituitary adenomas (58% females, 33% males) or without any pituitary lesions (21% females, 5% males); however, there were less female (21%) than male rats (63%) with foci of hyperplastic and/or hypertrophic cells but no adenomas in the pituitary (P = 0.001). Elevation of plasma prolactin concentration above the upper 99th percentile value in age-matched rats without lesions was predictive, but not conclusively, of the presence of pituitary hemorrhagic adenomas in both sexes. It was, however, not predictive of the presence of foci of hypertrophic or hyperplastic cells. Elevation of plasma prolactin concentration above 10 ng/ml in male and 60 ng/ml in female rats was conclusive for the presence of hemorrhagic adenomas. Using multivariate analysis, significant positive correlations (P less than 0.01) were found between plasma prolactin concentration and presence and size of hemorrhagic adenomas and their prolactin staining intensity (correlation coefficients between 0.392 and 0.652). Foci of hyperplastic cells stained positively for prolactin, whereas hypertrophic cell foci and pleomorphic and spongiocytic adenomas did not stain for prolactin. There were no correlations (coefficients of less than +/- 0.189) between plasma prolactin concentration and the presence of hypertrophic or hyperplastic cell foci and pleomorphic or spongiocytic adenomas in the pituitary. The morphologic criteria developed to distinguish spontaneous hypertrophic, hyperplastic, and neoplastic lesions of the rat pituitary corresponded well with their prolactin immunoreactivity and/or ability to elevate plasma prolactin concentration. These criteria constitute a biologically meaningful classification system for these rat pituitary lesions.

Does extragonadal presentation impart a worse prognosis to abdominal germ-cell tumours?

The prognostic significance of extragonadal rather than gonadal presentation of germ-cell tumour in 51 patients presenting between 1979 and 1988 with abdominal tumours was compared with that of 51 control patients with testicular primary tumours matched for bulk fo disease, serum tumour marker concentration, age and year of treatment. Very large volume tumour was found at initial staging in 24 extra-gonadal cases (47%) and high tumour markers in 29 (57%). Actuarial survival at 2 and 5 years was 82% and 70% for cases and 78% and 63%, respectively, for controls. These outcomes were not significantly different and the relative hazard of death for cases compared with controls was 0.7 (95% confidence intervals 0.3-1.5). Thus the presentation of germ-cell tumours with a retroperitoneal mass does not itself adversely influence prognosis compared with testicular presentation with equivalent disease extent. However it is rare for extragonadal presentation to be associated with small volume disease.

Is postchemotherapy retroperitoneal surgery necessary in patients with nonseminomatous testicular cancer and minimal residual tumor masses?

PURPOSE

At least one third of the patients with metastatic testicular cancer are rendered tumor-free by cisplatin-based chemotherapy. One may question, therefore, the routine use of postchemotherapy retroperitoneal lymph node dissection (RLND), especially if the residual masses are less than 20 mm in diameter. To define the role of such surgery, we analyzed the postchemotherapy histology in testicular cancer patients with minimal residual disease.

PATIENTS AND METHODS

Seventy-eight patients with advanced nonseminomatous testicular cancer underwent RLND after three to four cycles of cisplatin- or carboplatin-based chemotherapy. In all patients, the largest diameter of the residual retroperitoneal mass was less than 20 mm.

RESULTS

Complete fibrosis/necrosis was found in 51 patients, mature teratoma in 22, and vital malignant germ cell tumor in five. In two of the latter five patients, alphafetoprotein (AFP) had increased immediately before RLND. In the 76 patients with normal pre-RLND tumor markers, the presence of undifferentiated malignant teratoma (MTU) in the primary tumor and normal prechemotherapy tumor markers were independent parameters predicting complete fibrosis/necrosis, which was demonstrated in all 15 patients with these two pretreatment parameters.

CONCLUSIONS

Postchemotherapy RLND can be omitted in patients with MTU in the primary tumor who have normal AFP/human chorionic gonadotropin (AFP/HCG) before chemotherapy and whose residual retroperitoneal mass is less than 20 mm in diameter. If the pre-RLND tumor markers are normal, RLND should be performed in all other patients with small residual masses, even in the presence of a normal computed tomography (CT) and particularly if regular follow-up of the patients is not guaranteed.

Current perspectives on the role of adjunctive surgery in combined modality treatment for patients with germ cell tumors

The role of surgery in patients with advanced GCT after chemotherapy has evolved substantially in the era of combined modality therapy. In evaluating patients for surgery after chemotherapy, the clinician must consider carefully the histology (seminoma v NSGCT) of the primary as well as the extent of the residual disease. In patients with seminoma, the size of residual disease (greater than or equal to 3 cm) permits selection of patients with a high incidence of residual malignancy. In contrast, criteria designed to select NSGCT patients in whom surgical intervention after chemotherapy can be avoided are associated with substantial error. A normal radiographic evaluation in patients with NSGCT does not indicate a negative pathology and the treating physician must consider the approximately 20% risk of residual teratoma or carcinoma despite evidence of a radiographic CR. Continued research is needed to improve the sensitivity and specificity of case selection for patients requiring surgery after chemotherapy in order to limit the toxicity of curative therapy in this patient population.

Immunoreactive plasma inhibin levels in men after polyvalent chemotherapy of germinal cell cancer

In vitro studies have shown that the Sertoli cell is the primary source of inhibin in the male. We measured immunoreactive inhibin with a new two-site immunoenzymatic assay in the plasma of 92 men: 40 normal men, 7 patients with germinal cell cancer after unilateral orchidectomy and 45 patients with the same disease following unilateral orchidectomy and subsequent chemotherapy based on cisplatin. Normal men had inhibin levels of 1.77 +/- 0.09 U/l x 10(-3) (mean +/- SEM). Seven patients after unilateral orchidectomy had inhibin concentrations within the lower normal range (1.23 +/- 0.22 U/l x 10(-3)). Forty-five patients were investigated in a cross-sectional study up to 102 months after completion of chemotherapy. Inhibin levels were within the normal range in 25 patients (1.76 +/- 0.14 U/l x 10(-3)); 18 patients had significantly lower inhibin levels (0.48 +/- 0.05 U/l x 10(-3), p less than 0.005) when compared to patients after unilateral orchidectomy. Two patients had elevated inhibin levels (4.4 and 5.6 U/l x 10(-3)). The proportion of patients with normal and subnormal inhibin was not dependent on the time that elapsed after completion of chemotherapy or on the chemotherapy combination. There was no correlation between immunoreactive plasma inhibin and LH, FSH, testosterone or sperm count. The decrease in inhibin concentrations after chemotherapy may indicate long-term damage to Sertoli cells in some of the patients.

Predictive factors for outcome in treatment of metastatic nonseminomatous germ cell tumors

In recent years less intensive chemotherapy programs for patients with metastatic nonseminomatous germ cell tumors with high likelihood of cure have been proposed, and the use of innovative more intensive treatments for patients with less favorable prognosis is being explored. The development of validated prognostic classifications has thus become important. In 77 patients with metastatic nonseminomatous germ cell tumors treated with chemotherapy, the ability of various prognostic factors to predict outcome of treatment was assessed. The multifactorial prognostic classification (Indiana classification) and a mathematical predictive formula correctly allocated patients to low- or high-risk groups in 84.4 percent and 87.0 percent of cases. The multifactorial classification system (M.D. Anderson system) correctly allocated patients in 61 percent of cases. The presence of serum beta HCG levels over 1,000 mg/mL, a pure choriocarcinoma histology and possibly an extragonadal primary origin of tumor were found to predict an adverse outcome in a small number of patients. It is concluded that use of the Indiana classification or mathematical predictive formula is an accurate means of allocating patients with metastatic germ cell tumors to high- or low-risk groups and that allocation of patients with pure choriocarcinoma histology, very high beta HCG levels, or extragonadal primary origin of tumor to the poor prognosis category will improve the accuracy of prediction in a few cases.

Single agent activity of carboplatin in patients with previously untreated non-seminomatous germ cell tumours

The response to a single course of carboplatin has been investigated in 12 patients with previously untreated non-seminomatous testicular germ cell tumours. Patients received one course of carboplatin at a dose calculated to achieve a target area under the free carboplatin plasma concentration versus time curve (AUC) of 7 mg/ml x mins using the formula: dose (mgs) = target AUC x (GFR + 25). Response to carboplatin was assessed after a single course and treatment was then continued on the POMB/ACE schedule. Ten of 12 patients had either a greater than 50% decrease in serum HCG and/or AFP levels or a greater than 50% decrease in tumour volume after a single course of carboplatin. No patient had evidence of disease progression after carboplatin. This study demonstrates that single agent carboplatin is highly active in patients with non-seminomatous testicular germ cell tumours and thus provides evidence to justify its inclusion in chemotherapy combinations.

Serum lactate dehydrogenase isoenzyme 1 and tumour volume are indicators of response to treatment and predictors of prognosis in metastatic testicular germ cell tumours

44 patients with metastatic testicular germ cell tumours treated with cisplatin-based chemotherapy were evaluated for prognostic implications of clinical characteristics. 22 obtained complete remission by the initial chemotherapy, and 30 are disease-free. S-LDH-1 had an overall predictive value regarding the response of 80%, S-LDH of 64%, S-AFP of 62%, and S-hCG of 62%. In multivariate analysis regarding response, only tumour volume classified according to the Royal Marsden system (P = 0.0036) and S-LDH-1 (P = 0.0069) yielded information. Regarding survival, S-LDH-1 (P = 0.0141) and an estimate of total tumour mass (P = 0.0171) had most impact with additional information from S-hCG only (P = 0.0536). We conclude that S-LDH-1 may be used as a tumour marker in addition to S-hCG and S-AFP in patients with metastatic testicular germ cell tumour.

Sanctuary site relapse in chemotherapy-treated testicular cancer

The testis and central nervous system (CNS) may act as sanctuary sites for testicular germ cell tumours, as cytotoxic drugs penetrate these areas less well than systemic sites. We describe three patients who relapsed in the testis (one patient) or CNS (two patients) after receiving chemotherapy for responsive systemic disease. All three were asymptomatic at relapse, which was first manifest by rising tumour marker levels. These sanctuary site relapses were managed locally with surgery +/- radiotherapy. Two patients were rendered disease-free; one died of progression of his local disease only. Sanctuary site tumour should be considered when relapse occurs in the setting of otherwise chemosensitive disease; local therapy may be curative.

The Second Medical Research Council study of prognostic factors in nonseminomatous germ cell tumors. Medical Research Council Testicular Tumour Working Party

PURPOSE

To assess prognostic factors in a large population of patients with metastatic nonseminomatous germ cell tumors (NSGCT) arising in gonadal or extragonadal sites.

PATIENTS AND METHODS

Data from 795 patients treated with chemotherapy between 1982 and 1986 in 13 centers were analyzed. Particular emphasis was placed on exact tumor measurements (eg, size of nodal masses, number of lung metastases), and the diagnostic pathology was also reviewed. Cox regression analysis was performed on these data. The patients were treated with a variety of cisplatin-containing chemotherapy regimens, 86% of which included etoposide.

RESULTS

With median follow-up of 45 months, overall 3-year survival is 85%. The independently adverse features proved to be (1) the presence of liver, bone, or brain metastases; (2) raised marker levels (alpha-fetoprotein [AFP] level greater than 1,000 kU/L or beta subunit of human chorionic gonadotropin [HCG] greater than 10,000 IU/L [corrected]); (3) the presence of a mediastinal mass greater than 5 cm in diameter; (4) the presence of 20 or more lung metastases; (5) increasing age; and (6) absence of undifferentiated teratoma (embryonal carcinoma) or fibrous tissue from the primary tumor.

CONCLUSIONS

The first four factors were used to define a simple prognostic classification. A good-prognosis group having none of these features comprised 67% of our patient population and had a 3-year survival of 93%. The remaining 33% of patients having at least one of these features had a 3-year survival rate of 68%. These patient groups are currently the subjects of international randomized clinical trials.

Serum markers in germ cell neoplasms

Innovations in the treatment of testicular cancer, including surveillance of clinical stage I patients and curative chemotherapy for disseminated disease, have increased the need for sensitive ways to stage and monitor patients, both during and after therapy. Serum tumor markers, in combination with radiographic studies, have significantly improved our ability to evaluate and treat patients with seminomas and NSGCT. Elevated AFP and BHCG levels provide prognostic information at diagnosis, indicate persistent disease following orchiectomy or RPLND, and signal a recurrence after chemotherapy. Significantly delayed clearance of markers during chemotherapy often indicates persistent disease. Serum markers help define the duration of therapy, thus minimizing the substantial toxicities often associated with curative chemotherapy. Despite these advances, areas of concern remain. A small percentage of patients with NSGCT and the majority of patients with seminoma have undetectable levels of AFP and BHCG. The search for additional sensitive and specific serum markers in these cases has not been wholly successful. LDH, PLAP, and BFP occasionally serve as useful markers in seminoma but suffer lack of specificity. In addition, normal postoperative or postchemotherapy serum marker levels do not always ensure complete remission. This is difficult clinically when residual masses persist following therapy. Resection is always required to rule out persistent disease. The next decade may reveal additional useful serum tumor markers and potentially new imaging techniques incorporating antimarker antibodies to differentiate necrotic tissue from active disease.

Single institutional experience with non-seminomatous germ cell tumours of the testis: local perspectives on a curable cancer

We have reviewed 77 patients with Non-seminomatous germ cell tumour of the testis (NSGCTT) treated at a single institution. A residual mass following definitive treatment occurred in 16 patients (35%), 13 of whom had a resection of the mass, yielding active tumour in only one patient. Nine patients (12%) relapsed including four of the 14 with Stage I disease who were treated by orchidectomy alone. Four relapses occurred at more than two and a half years after primary treatment. Relapse prior to the development of clinical symptoms or signs was evident in three of nine patients; in two patients by routine imaging and one with elevated routine serum markers. Three of the nine patients who relapsed had elevated serum markers. Two patients died from disease but there were four treatment-related deaths (7%). Overall, 64 patients (83%) remain disease free at the time of follow-up. A further seven (9%) have been lost to follow-up but were disease-free at a minimum of 26 months after diagnosis. This study confirms features of this disease including the excellent prognosis when adequately treated. However, it also reveals the problems of late or marker negative relapses, the implementation of an observation policy in Stage I disease and treatment related mortality in young men.

Serum tumor markers and patient allocation to good-risk and poor-risk clinical trials in patients with germ cell tumors

The allocation of patients with advanced germ cell tumors (GCT) to different treatment programs based on clinical characteristics is standard in the design of clinical trials today. Studies have shown that substantial differences exist between entry criteria and that these differences could influence the outcome of clinical trials. The factors contributing to these differences are not clear due to patient selection biases. Two hundred five unselected and consecutive patients allocated to and treated in good-risk and poor-risk treatment programs at Memorial Sloan-Kettering Cancer Center (MSKCC) were reassigned risk status by the Indiana University (IU) Classification. The results were compared with those of the Southeastern Cancer Study Group (SECSG). The results using both criteria indicated substantial agreement in total end results and the identification of good-risk patients. The results in poor-risk patients differed substantially, with 39 patients (19%) classified as poor-risk by MSKCC criteria and 66 (32%) by Indiana criteria. The major discrepancy occurred in IU Stage 7, in which 26 of 32 patients (81%) achieved a complete response. The major factor contributing to this difference in risk assignment was the use of serum tumor markers. Serum tumor markers must be incorporated into risk assignment criteria for GCT clinical trials to minimize the number of good-risk GCT patients in poor-risk trials.

Nonseminomatous germ cell tumor with very high serum human chorionic gonadotropin

Most patients with disseminated nonseminomatous germ cell tumor (NSGCT) have an excellent prognosis with modern chemotherapy, although certain subgroups with a worse prognosis have been described. One such subgroup includes patients with high serum levels of the tumor marker, human chorionic gonadotropin (HCG). Sixteen patients of 104 treated for NSGCT at the CRC Wessex Medical Oncology Unit (Southampton, UK) presented with serum HCG greater than 25,000. Most of these patients exhibited features of the "choriocarcinoma syndrome" with bulky, rapidly progressive disease; frequent pulmonary, hepatic, and central nervous system complications; and a generally poorer response to standard NSGCT chemotherapy. Histologic identification of trophoblastic tumor was not made in all patients and is not essential for the diagnosis of the syndrome; indeed, closed biopsy may be contradicted in some circumstances because of the risk of hemorrhage. The NSGCT patients with poor prognosis, including patients with the choriocarcinoma syndrome, must be clearly identified in order to improve management and, eventually, cure rates.

Serum tumor marker half-life during chemotherapy allows early prediction of complete response and survival in nonseminomatous germ cell tumors

The prognostic value and therapeutic utility of monitoring the decay of alpha-fetoprotein (AFP) and human chorionic gonadotropin (HCG) after chemotherapy for nonseminomatous germ cell tumors was assessed. Patients treated on successive front line chemotherapy protocols at Memorial Hospital between 1979 and 1988 were studied. Marker values taken within the first 90 days of treatment were reviewed for the 198 patients who had initially abnormal values and serial measurements at Memorial Hospital. Since markers frequently increased in an unpredictable fashion in the first week after chemotherapy, prechemotherapy values would be inaccurate for assessment of subsequent half-life. Therefore, the first two values measured greater than 7 days after the start of treatment were used for all calculations of half-life. Among 38 patients who had the two successive AFP measurements elevated, those who later achieved a complete response (CR) had a median AFP half-life of 6.1 days (n = 20), whereas those not achieving CR had a median AFP half-life of 13.3 days (P = 0.02). Among 37 patients with the two successive HCG values elevated, those who later achieved CR had a median HCG half-life of 4.2 days (n = 10), whereas those not achieving CR had a median HCG half-life of 18.4 days (P = 0.04). Forty-two patients who had an AFP half-life greater than 7 days or an HCG half-life greater than 3 days had significantly shorter overall survival (median, 8 months) than the other 156 patients (median not reached) (P less than 0.0001). These 42 patients also achieved CR in lower proportion (29%) than the other 156 patients (89%) (P less than 0.0001). Cox regression identified prolonged marker half-life as the most significant independent predictor of survival. Lack of appropriate decay of serum tumor markers can identify patients unlikely to achieve CR or prolonged survival and thus can be used to select patients during treatment who may benefit from an early change to more aggressive therapy.

Clinical study of intracranial nongerminomatous germ cell tumors producing alpha-fetoprotein

We present six patients with intracranial nongerminomatous germ cell tumors that produced alpha-fetoprotein (AFP). Their ages ranged from 8 to 20 years (average, 11.5 years old); two were male and four were female. Four of the tumors originated in the pineal region and two in the suprachiasmatic region. One patient treated with only radiation therapy died within 3 months of admission as a result of intraperitoneal metastasis via a ventriculoperitoneal shunt. Another patient, treated with radiation therapy and intrathecal administration of neocarzinostatin, died after 12 months because of tumor progression and subarachnoid dissemination. Two patients who received radiation and combination therapy with cisplatin, vinblastine, and bleomycin died after 13 and 25 months. The remaining two patients treated with radiation therapy and adjuvant chemotherapy (cisplatin and etoposide) are now alive without recurrence after 16 and 19 months from admission. Adjuvant chemotherapy with cisplatin and etoposide appears to be efficacious in the treatment of intracranial nongerminomatous germ cell tumor.

Early clinical stages (CS1, CS1Mk+ and CS2A) of non-seminomatous testis cancer. Value of pre- and post-orchiectomy serum tumor marker information in prediction of retroperitoneal lymph node metastases. Swedish-Norwegian Testicular Cancer Project (SWENOTECA)

During a 5-year period (1981-86) 588 consecutive patients with nonseminatous germ cell tumors of the testis were included into a prospective Swedish-Norwegian multicenter study (SWENOTECA) and clinically staged according to the Royal Marsden system. A total of 370 patients (63%) had early clinical stages (CS) of disease; 295 (50%) had CS1, 32 (5%) had CS1Mk+ (CS1 with pathological serum tumor marker patterns after orchiectomy) and 43 (7%) had CS2A disease. Pathological staging with retroperitoneal lymph node dissection (RPLND) of the retroperitoneum was performed in 345 (93%) of the early CS patients and 128 (37%) had pathological stage 2 (PS2) disease; 27% of the CS1, 100% of the CS1Mk+ and 66% of the CS2A patients. The overall clinical staging accuracy was 75%. All the 40 patients with pathological serum AFP and/or HCG patterns before RPLND had PS2 disease, compared to 81/282 (29%) of patients with normal marker patterns. The PS2 patients with pathological marker patterns had significantly more and larger retroperitoneal metastases than those with normal AFP and HCG values. Elevated pre-orchiectomy AFP level indicated significantly reduced risk of PS2 disease in CS1 patients, but this effect became non-significant if the CS1Mk+ and CS2A cases were included into univariate or multivariate analyses. We suggest that the 'good risk' effect of pre-orchiectomy AFP elevation for CS1 cases may be caused by a selection mechanism during the clinical staging process.

The UK Children's Cancer Study Group: testicular malignant germ cell tumours 1979-1988

The United Kingdom Children's Cancer Study Group (UKCCSG) malignant germ cell tumour (MGCT) studies were undertaken to establish standard protocols of investigation, staging, and treatment. The efficacy of new drug combinations and the value of serial measurements of serum alphafetoprotein (AFP) and human chorionic gonadotrophin (HCG) were evaluated. Following the initial surgery, staging of the tumour was performed using a variety of investigative approaches. In stage 1 testicular tumours, orchidectomy was performed. In more advanced tumours, and in stage 1 tumours that failed to show the expected decline in AFP or recurred, chemotherapy was used after appropriate surgery. Seventy-three boys, under 14 years of age, with testicular MGCTs have been entered into the UKCCSG studies since 1979. Serum AFP was measured preoperatively, or within 2 weeks of operation, in 70 boys. It was unequivocally elevated in 69. Monitoring by serial AFP measurement proved valuable in assessing response and in early detection of recurrence. HCG was measured in 46 boys, and was raised in three. Sixty-seven (91%) of the tumours were yolk sac (Teilum) tumours, four were immature teratoma, and two were mixed MGCTs. The only non-AFP producing tumour was an immature polydermal teratoma in a 1-year-old boy. Serum HCG was raised in three boys with yolk sac tumours, one with a mixed teratoma, and one 14-year-old boy who had a mixed MGCT. The results of treatment were assessed on April 1, 1989 (median time from diagnosis, 3 years 4 months). Seventy-one boys were alive, 48 of whom had been cured by orchidectomy alone. The remaining 25 patients received chemotherapy.(ABSTRACT TRUNCATED AT 250 WORDS)

[Clinical evaluation on differential quantitation of human chorionic gonadotropin in testicular cancer]

In 113 patients with testicular germ cell tumor, the authors determined human chorionic gonadotropin (hCG) and the free beta subunit levels in sera by differential quantitation using homologous hCG-beta radioimmunoassay (RIA) and hCG enzyme immunoassay (EIA), respectively. In 59 patients with seminoma, intact hCG was positive in 6 patients (10.1%), whereas free hCG-beta was positive in 23 patients (39.0%). Syncytiotrophoblastic giant cells were detected in 4 cases out of 6 patients with positive intact hCG seminoma. Seventeen cases (28.8%) revealed positive free hCG-beta only without elevation of intact hCG. In 54 patients with nonseminoma, both intact hCG and free hCG-beta were positive in 36 cases (66%). In some cases, hCG-beta/hCG ratio increased up to 275% when recurrence of tumor developed. Molecular heterogeneity of hCG is closely related to proliferation and differentiation of hCG producing cells. The differential quantitation of intact hCG and hCG-beta using these two assays is valuable for clinical detection of molecular heterogeneity of hCG in testicular tumor.

Primary mediastinal nonseminomatous germ cell tumors. Results of a multimodality approach

Before cisplatin-based chemotherapy, long-term survival after resection of primary mediastinal nonseminomatous germ cell tumors was unusual. We reviewed the case histories of 48 patients who underwent multimodality treatment for mediastinal nonseminomatous germ cell tumor between 1976 and 1988. Twenty-eight patients received initial therapy at Indiana University and 20 were referred after having had unsuccessful initial therapy elsewhere. In 44 patients (92%) the levels of either one or both serum tumor markers were elevated at the time of diagnosis. Five patients had choriocarcinoma, three embryonal carcinoma, 12 yolk sac carcinoma, four teratocarcinoma, 22 mixed cell type, and two had an unclassified type. Twenty-two of the 28 patients in our initial therapy group had a complete response to treatment, as defined by normal serum tumor markers and absence of residual tumor. In this group, 16 patients had resection of residual disease after chemotherapy, four had total or near total resection before chemotherapy, and only two had chemotherapy alone. Seventeen patients are surviving after this treatment with a median survival of 64 months and a 57% 5-year Kaplan-Meier survival rate. Only two of the 20 patients who were referred for salvage chemotherapy had a complete response. Both required resection of residual disease after salvage chemotherapy. Only one patient survived after this treatment. There was no significant treatment morbidity or mortality. A multimodality approach to primary mediastinal nonseminomatous germ cell tumor with intensive cisplatin-based chemotherapy, emphasis on normalizing serum tumor markers, and aggressive resection of residual disease now offers survival to a significant number of patients.