Prospective comparison of [18F]fluorodeoxyglucose positron emission tomography with conventional assessment by computed tomography scans and serum tumor markers for the evaluation of residual masses in patients with nonseminomatous germ cell carcinoma




To assess the ability of [18F]fluorodeoxyglucose (F-18 FDG) positron emission tomography (PET) to predict the viability of residual masses after chemotherapy in patients with metastatic nonseminomatous germ cell tumors (GCT), PET results were compared in a blinded analysis with computed tomography (CT) scans and serum tumor marker changes (TUM) as established methods of assessment.


Independent reviewers who were blinded to each other's results evaluated the PET results and corresponding CT scan and TUM results in 85 residual lesions from 45 patients. All patients were treated within prospective clinical trials and received primary/salvage, high-dose chemotherapy with autologous blood stem cell support for primary poor prognosis disease or recurrent disease. PET results were assessed both visually and by quantifying glucose uptake (standardized uptake values). Results were validated either by histologic examination of a resected mass and/or biopsy (n = 28 lesions) or by a 6-month clinical follow-up after evaluation (n = 57 lesions).


F-18 FDG PET showed increased tracer uptake in 32 of 85 residual lesions, with 29 true positive (TP) lesions and three false positive (FP) lesions. Fifty-three lesions were classified by PET as negative (no viable GCT), 33 lesions were classified by PET as true negative (TN), and 20 lesions were classified by PET as false negative (FN). In the blinded reading of the corresponding CT scan and TUM results, 38 residual lesions were assessed correctly as containing viable carcinoma and/or teratoma. Forty-six lesions were classified as nonsuspicious by CT scan/TUM (33 TN lesions and 14 falsely classified lesions). PET correctly predicted the presence of viable carcinoma in 5 of these 14 and the absence of viable carcinoma in 3 of these 14 lesions. Resulting sensitivities and specificities for the prediction of residual mass viability were as follows: PET, 59% sensitivity and 92% specificity; radiologic monitoring, 55% sensitivity and 86% specificity; and TUM, 42% sensitivity and 100% specificity. The positive and negative predictive values for PET were 91% and 62%, respectively. The diagnostic efficacy of PET did not improve when patients with teratomatous elements in the primary tumor were excluded from the analysis. In patients with multiple residual masses, a uniformly increased residual F-18 FDG uptake in all lesions was a strong predictor for the presence of viable carcinoma.


F-18 FDG PET imaging performed in conjunction with conventional staging methods offers additional information for the prediction of residual mass histology in patients with nonseminomatous GCT. A positive PET is highly predictive for the presence of viable carcinoma. Other useful indications for a PET examination include patients with multiple residual masses and patients with marker negative disease. Cancer 2002;94:2353–62. © 2002 American Cancer Society.

DOI 10.1002/cncr.10494

The development of cisplatin-based combination chemotherapy has dramatically improved the prognosis of patients with metastatic nonseminomatous germ cell tumors (NSGCT), resulting in a long-term cure rate of 70–80%.1, 2 After the completion of chemotherapy, residual tumor lesions are found in up to 40% of patients with advanced NSGCT, despite the normalization of serum tumor markers.3 The differentiation between viable carcinoma, mature teratoma, and necrosis is possible only with histologic examination of the resected specimen. Thus, the surgical resection of all residual masses, if technically possible, is the recommended standard of care. Approximately 50% of residual lesions consist of necrosis only, whereas mature teratoma is found in 30% of residual masses, and the remaining 20% of residual lesions contain viable immature tumor cells.4 In the latter group of patients, secondary resection may improve the prognosis of the patient depending on the biologic aggressiveness of the disease, the completeness of surgery, and the effectiveness of additional therapy. The resection of mature teratoma also is important, because subsequent problems caused by local tumor growth or late disease recurrence with possible malignant, nongerminal transformation can be avoided. However, in the case of necrotic tissue, patients do not benefit from secondary surgery yet are exposed to surgery-related morbidity and mortality.5, 6 Therefore, it is of great interest to identify those patients who require a secondary postchemotherapy resection and those who do not in order to optimize individual treatment. To date, no diagnostic tool has been developed that reliably predicts in vivo the viability of the residual mass. Unfortunately, radiologic criteria derived from computed tomography (CT) scans or magnetic resonance imaging (MRI) have failed to differentiate reliably between viable carcinoma, mature teratoma, and necrosis/scar tissue from residual lesions in patients with NSGCT. Serum tumor markers also may be misleading, because viable tumor may be found in some patients with normalized serum tumor markers, and necrosis may be found in some patients with a prolonged elevation of serum tumor markers.4, 7

Positron emission tomography (PET) imaging using 2-[18F]fluoro-2-deoxy-D-glucose (F-18 FDG) is a new diagnostic technique that allows the visualization and semiquantitative calculation of regional glucose metabolism within the body.8, 9 Because tumor cells are characterized by a higher glucolytic rate than normal tissue cells, PET exploits this difference by assessing the rate and quantity of F-18 FDG uptake by the tumor. PET has been demonstrated as a valuable imaging technique for the accurate staging of lymphoma.10–12 In patients with testicular carcinoma, PET has demonstrated a high sensitivity and specificity for the detection of metastases at initial diagnosis.13, 14

Thus, the primary objective of the current study was to evaluate prospectively the ability of PET to predict the presence of viable carcinoma, mature teratoma, or necrosis in residual masses from patients with metastatic NSGCT. Because both tumor marker decline and radiologic imaging currently serve as established methods to assess the response to chemotherapy and the viability of residual masses, the potential value of PET in evaluating residual masses was compared with those methods.


Patients and Treatment

Patients with either newly diagnosed, metastatic, poor prognosis NSGCT according to the International Germ Cell Cancer Collaborative Group classification or patients with recurrent disease after cisplatin-based chemotherapy and at least one residual mass measuring ≥ 1 cm in greatest dimension on a CT scan were eligible for inclusion in the PET protocol. All patients were treated within one of two prospective German multicenter, high-dose chemotherapy trials between September 1995 and October 1999.15–17 Eligibility criteria were similar for both high-dose chemotherapy trials and consisted of the following: germ cell carcinoma of any primary tumor site, Karnofsky performance status > 50%, normal kidney function, absence of severe heart or liver disease, and written informed consent. Patients with poor prognosis germ cell carcinoma at initial diagnosis received sequential, first-line, high-dose chemotherapy plus autologous stem cell support. The treatment protocol consisted of one cycle of standard-dose etoposide, ifosfamide, and cisplatin (VIP) chemotherapy (cisplatin 20 mg/m2, etoposide 75 mg/m2, and ifosfamide 1200 mg/m2 daily for 5 days) followed by three cycles of high-dose VIP chemotherapy (cisplatin 20 mg/m2, etoposide 300 g/m2, and ifosfamide 2000–2400 mg/m2 daily for 5 consecutive days every 3 weeks for a total of 3 cycles). Patients with recurrent disease were treated with three cycles of standard-dose paclitaxel, ifosfamide, and cisplatin (TIP) chemotherapy followed by one cycle of thiotepa, etoposide, and carboplatin (TEC) high-dose chemotherapy. Standard-dose TIP chemotherapy consisted of paclitaxel 175 mg/m2 given on Day 1 and ifosfamide 1200 mg/m2 and cisplatin 20 mg/m2, both administered on Days 2–6 of a 22-day cycle. The TEC high-dose regimen contained thiotepa 150 mg/m2, etoposide 600 mg/m2, and carboplatin 500 mg/m2, with all drugs given daily over 3 days. All patients received autologous peripheral blood stem cell support and granulocyte-colony stimulating factor after high-dose chemotherapy according to treatment protocol and institutional practice.

All patients were treated at Tuebingen University Medical Center. The current study was approved by the Ethical Committee of the University of Tuebingen. All patients were required to provide written informed consent.

Tumor Response Evaluation

All patients underwent extensive staging procedures that included either a spiral CT scan with orally and intravenously administered contrast medium or an MRI of the chest, abdomen, and brain; determination of serum tumor marker levels (β-human chorionic gonadotrophin, α-fetoprotein, and lactate dehydrogenase) as well as a baseline PET image prior to the start of chemotherapy. Serum tumor marker levels were determined prior to each chemotherapy cycle, and a CT scan or MRI of the tumor lesions was performed after every second cycle. After the completion of chemotherapy, all patients underwent a PET examination as well as a CT scan or MRI to assess residual masses. The PET examination and the CT scan were performed at least 3 weeks after treatment. All staging procedures were performed within 3 weeks of one another to allow adequate comparisons. If it was feasible technically, all residual masses were resected after the completion of chemotherapy. Follow-up examinations, which included spiral CT scans and serum tumor marker level evaluations, were performed in 3-month intervals after the completion of therapy.

All CT scans were reviewed by an independent, board-certified radiologist at the Department of Radiology of the University of Tuebingen who was not aware of the PET findings. Most CT scans were performed at this department using standard, state-of-the-art techniques (spiral CT scanning with a slice thickness of 8 mm, a table feed of 12 mm, and increments of 7 mm; oral and intravenous contrast media). Clinical response was classified according to the modified World Health Organization criteria.18 The criteria for the assessment of viability on CT scans were changes in tumor size compared with the tumor size on initial CT scan and the degree of contrast enhancement. A decrease in tumor size ≥ 50% and/or diminished or absent contrast medium uptake were considered radiologic signs of a nonviable lesion. Progressive lesions, lesions with a reduction < 50% in size, and lesions with persistent/increased contrast medium uptake were rated viable. Compared with baseline values prior to therapy, a serum tumor marker decline ≥ 90% after chemotherapy was classified as a favorable response predicting a nonviable residual lesion. A tumor marker decrease < 90% or increased tumor marker levels were considered unfavorable responses and were predictive of a viable residual mass.

PET Imaging

A dedicated PET scanner (ADVANCE; General Electrics Medical Systems, Milwaukee, WI) was used, providing an axial field of view (FOV) of 14.6 cm and rotating 68Ga/68Ge line sources for measured attenuation correction. Emission data were corrected for random events, attenuation (restricted to FOVs containing residual tumor masses in patients with early-stage tumors), and scattering. Thirty-five slices (4.25 mm) per FOV were reconstructed iteratively or by filtered back projection with a 128 × 128 pixel matrix (4.3 × 4.3 mm pixel size), resulting in a final resolution of about 8 mm (full width at half maximum). The reconstructed images were printed on film (Matrix LR3300 P Laser Imager; Agfa-Gevaert, Mortsel, Belgium) using a black-and-white color table representing standard uptake values (SUVs) of 0–8 (SUV 8, black) to facilitate visual image analysis. In patients who underwent whole body scans, coronal slices (8.6 mm) also were documented.

All patients fasted for a minimum of 12 hours prior to PET imaging. Blood glucose levels were checked for each patient before the intravenous administration of 250–500 megabecquerels of F-18 FDG. Forty-five to 60 minutes after the F-18 FDG injection, static PET scans were recorded for 5–15 minutes per FOV (depending on reconstruction algorithm, injected radioactivity, and patient weight), typically covering the head, the trunk, and the upper legs to the midfemur. Transmission scanning (3–20 minutes per FOV, depending on patient weight, transmission source activity, and whether segmentation was applied; sinogram windowing) was performed prior to F-18 FDG injection or after emission scanning. After tracer injection, patients received 1 L of fluid orally or intravenously and 20 mg furosemide intravenously to minimize image artifacts from residual radioactivity in the urinary tract.

PET Image Analysis

Each PET image was reviewed by an experienced nuclear medicine physician from the Department of Nuclear Medicine at the University of Tuebingen who was blinded to both the serum tumor marker course and the interpretation of CT scan and/or MRI results. PET images were classified as either positive or negative by visual assessment for all patients and were classified subsequently by semiquantitative analysis using SUVs in 75 of 85 patients.8, 19 No SUVs were calculated in 10 patients. Residual tumor lesions with an SUV ≥ 2 were considered positive. On visual assessment, focally increased F-18 FDG uptake exceeding that of the surrounding tissue and/or contralateral body regions was interpreted as viable tumor tissue. In addition, tomographic whole body coronal images were reviewed for F-18 FDG accumulations outside of known lesions.

PET and conventional assessment (CT scan and tumor marker decline) results also were correlated with histologic findings in residual masses from patients who underwent secondary resection after the completion of chemotherapy. If no resection of residual tumor masses was performed, then the clinical course of the patient was used. The absence of tumor progression on CT scans or tumor marker increases within 6 months after chemotherapy were considered indicators of a nonviable residual lesion. Significance levels for differences between median SUVs were calculated using the Fisher exact test.

For the evaluation of the predictive ability of PET imaging, each mass was assessed separately. Similarly, the comparison of PET images with CT scans was based on the comparison of each residual mass. The assessment of serum tumor marker decline was based on each patient, because tumor markers are the same for all residual lesions in a patient. When comparing PET images with the combination of CT scan/serum tumor markers, each residual mass was assessed together with the corresponding serum tumor marker decline to allow an exact comparison.


Eighty-five residual lesions in 45 patients who received high-dose chemotherapy for NSGCT (32 patients for recurrent disease and 13 patients for poor prognosis disease at initial diagnosis) were evaluated with F-18 FDG PET before and after treatment. Patient characteristics are listed in Table 1. The most common metastatic sites were the retroperitoneal lymph nodes and the lungs. Eight patients presented with an extragonadal primary tumor. Baseline PET images prior to the start of chemotherapy showed an increased F-18 FDG uptake in the tumors from all patients. After the completion of chemotherapy, 21 patients (47%) showed a marker negative partial response, and 8 patients (18%) showed a partial remission without marker normalization. Fifteen patients (33%) had stable disease or progressive disease within 4 weeks after the completion of chemotherapy. One patient (2%) who was without initial marker elevation had a partial response.

Table 1. Patient Characteristics (N = 45 patients)
CharacteristicNo. (%)
  1. HCG: human chorionic gonadotropin; AFP: α-fetoprotein; LDH: lactate dehydrogenase; PR: partial response; TUM: serum tumor marker.

Median age in yrs (range)33 (21–57)
Primary tumor localization
 Gonadal37 (82)
 Extragonadal8 (18)
Primary histology
 Malignant teratoma6 (13)
 Embryonal carcinoma7 (16)
 Yolk sac2 (4)
 Chorion-carcinoma3 (7)
 Mixed histology24 (53)
  With teratoma13
  Without teratoma11
 Other/not known3 (7)
Location of metastases
 Retroperitoneal lymph nodes28 (62)
 Lungs24 (53)
 Mediastinum12 (27)
 Liver10 (22)
 Other11 (24)
Patients with elevated serum tumor markers
 β-HCG22 (49)
 AFP24 (53)
 LDH15 (33)
 None3 (7)
Treatment situation
 First-line treatment13 (29)
 Treatment for first recurrence32 (71)
Response to therapy
 PR marker negative21 (47)
 PR marker positive8 (18)
 PR (no initially elevated TUM)1 (2)
 Stable disease11 (24)
 Progression4 (9)
Histologic findings/secondary resection (n = 28 localizations, n = 18 patients)
 Viable carcinoma13 (46)
 Mature teratoma3 (11)
 Necrosis12 (43)
Course of disease after the end of therapy
 No disease progression (for at least 6 months)21 (47)
 Disease progression (within 6 months)24 (53)
Median follow-up in months (range)27 (6–62)

A histologic specimen was available after secondary residual mass resection in 8 patients with a total of 28 residual lesions; whereas, in 27 patients with a total of 57 lesions, no resection could be performed. Therefore, the clinical course over the subsequent 6 months after evaluation was monitored in the latter patients. Histologic examination revealed viable carcinoma in 12 lesions (43%), mature teratoma in 3 lesions (11%), and necrosis in 13 lesions (46%). Overall, 21 patients (47%) remained free of disease progression for at least 6 months after high-dose chemotherapy, whereas 24 patients (53%) experienced disease recurrence within the same period. Sixteen patients experienced disease recurrence beyond 6 months follow-up (7 patients after 6–12 months, 7 patients after 12–18 months, and 2 patients after > 18 months). These patients also showed markedly increased tumor markers at the time of recurrence, indicating the presence of viable carcinoma rather than mature teratoma. The median follow-up of all patients was 27 months (range, 6–62 months).

Evaluation of Residual Masses by PET

Overall, viability was assessed correctly by PET imaging in 62 of 85 residual masses (73%). A negative PET was found by visual assessment in 53 residual masses (62%). In 20 of these lesions, either tumor progression was observed within 6 months after treatment (18 lesions), or mature teratoma was found on histologic examination (2 lesions), indicating false negative PET findings. Thirty-two lesions (38%) in 22 patients were considered positive by the PET reviewer. Nineteen patients with a total of 29 lesions either experienced disease recurrence within 6 months after high-dose chemotherapy, or the histology of the resected residual tumor mass after high-dose chemotherapy still revealed the presence of viable carcinoma/mature teratoma. Three residual masses in three different patients were considered positive, but the patients had a favorable outcome. In two patients, the histology of the resected mass revealed inflammation; whereas, in the third patient, necrosis was found. Thus, the sensitivity and specificity of PET for the prediction of residual mass viability after the completion of chemotherapy were 59% and 92%, respectively. The positive predictive value of PET for the presence of viable carcinoma/teratoma was 91%, whereas the negative predictive value of PET was 62% (Table 2).

Table 2. Comparison of the Sensitivity, Specificity, and Negative and Positive Predictive Values of Positron Emission Tomography (Visual assessment), Computed Tomography Scans/Magnetic Resonance Imaging, and Serum Tumor Markers for the Assessment of Viability of Residual Masses
ParameterPET (n = 85 lesions)CT scans/MRI (n = 85 lesions)Tumor markers (n = 42 patients with a total of 78 lesions)a
%95% CI%95% CI%95% CI
  • 95% CI: 95% confidence interval.

  • a

    Three patients with a total of 7 residual lesions and without serum tumor marker elevation at the time of initial diagnosis were excluded.

Negative predictive value6248–75%5844–72%5233–70%
Positive predictive value9175–98%8467–95%10071–100%

Quantification of F-18 FDG uptake by calculating SUVs was performed for direct comparison in 75 residual lesions (Table 3). The average SUV of true positive lesions was 2.6 (range, 0.9–10), and the average SUV of true negative lesions was 1.2 (range, 0.9–5.8; P < 0.05). The median SUV of lesions that contained necrosis was 1.5 (range, 0.9–5.8), and, in the two resected lesions that contained mature teratoma, the median SUVs were 1.5 and 1.7 (P > 0.05). Thus, the sensitivity (68%) and negative predictive value (67%) improved slightly by the estimation of SUVs compared with visual assessment alone (sensitivity, 59%; negative predictive value, 62%). In contrast, the specificity (83%) and positive predictive value (84%) decreased slightly.

Table 3. Comparison of the Sensitivity, Specificity, and Negative and Positive Predictive Values of Visual Assessment and Quantitative Analysis (Standard uptake value) of Positron Emission Tomography in Patients with Residual Masses after High-Dose Chemotherapy for Nonseminomatous Germ Cell Tumorsa
ParameterVisual assessmentQuantitative analysis
%95% CI%95% CI
  • 95% CI: 95% confidence interval.

  • a

    Seventy-five of 85 lesions were evaluated in this comparative analysis.

Negative predictive value6248–75%6747–83%
Positive predictive value9175–98%8464–95%

Twenty-eight of 45 patients presented with residual masses at multiple localizations. Whereas, in 24 of these patients, the residual masses within each patient displayed a uniform clinical behavior (either regression or progression), the residual masses in 4 patients developed differently from one another. PET indicated uniform behavior of all residual masses in 19 of these patients: Four patients had truly positive results, 10 patients had truly negative results, and 5 patients had false negative results. No false positive overall behavior was suggested by PET in patients with a uniform glucose uptake in all residual masses.

Comparison of PET Imaging with CT Scan/MRI for the Prediction of Therapy Response

Comparisons of PET imaging with CT scan/MRI performed at the same time interval after chemotherapy were available for all 45 patients, who had a total of 85 residual lesions (Table 2). Overall, CT scan/MRI correctly predicted viability in 58 residual lesions (68%). CT scan/MRI correctly indicated viable carcinoma/teratoma in 27 residual lesions (20 in patients with stable disease and 7 in patients with tumor progression during chemotherapy). Fifteen of these lesions were also PET positive; whereas, in 12 residual masses, PET results were considered negative by the reviewers.

In 31 residual lesions without progression during follow-up and/or with decreased contrast medium uptake, CT scan/MRI correctly predicted a necrotic residual mass. Whereas only 28 of these lesions were also PET negative, an increased F-18 FDG uptake was found in 3 lesions. Histologic examination after resection of these three lesions showed inflammation in two lesions and necrosis in the third lesion (see above).

CT scans/MRI after therapy were not able to predict correctly the viability of 27 lesions (32%). Twenty-two of these residual masses regressed more than 50% without contrast medium enhancement on CT scan/MRI, indicating a nonviable lesion, although they still progressed during follow-up. In 14 of these lesions, PET results correctly predicted a viable residual mass. However, eight lesions also had false negative PET results. Five lesions remained progression free or showed necrosis on histologic examination, despite the fact that CT scans showed no tumor lesion response. PET correctly assessed all five lesions as necrosis. Thus, the respective sensitivity and specificity of CT scans/MRI (stable disease or progressive disease predicting viable residual masses) were 55% and 86%. The positive and negative predictive values of CT scans/MRI were 84% and 58%, respectively (Table 2).

Comparison of PET with Serum Tumor Marker Decline for the Prediction of Residual Mass Viability

The comparison of PET imaging with serum tumor marker decline during therapy included 42 patients with a total of 78 lesions. Because three patients with a total of seven lesions never exhibited elevated markers during the course of their disease, they were not included in the analysis. Residual mass viability was predicted correctly by serum tumor marker decline alone in 27 patients (64%) with 56 lesions (72%). In 11 patients with 27 lesions, tumor markers remained elevated or even increased after therapy, and all of these patients failed treatment. In 8 of these patients with a total of 20 residual lesions, increased F-18 FDG uptake was found in at least one residual lesion after treatment. Conversely, PET falsely predicted a necrotic residual mass in three patients with seven lesions. Tumor marker decline correctly indicated the absence of viable tumor in the other 16 patients with 29 lesions. PET also was normalized in 27 of these 29 residual masses yet remained positive at the time of restaging after treatment in 2 lesions (2 patients). The decline (7 patients with 9 lesions) or normalization (8 patients with 13 lesions) of tumor markers indicated the presence of necrosis in 15 other patients with a total of 22 residual masses, but all of these patients experienced disease progression. In comparison, PET correctly predicted the viability in 9 patients with a total of 13 lesions and falsely indicated the absence of viable tumor in the remaining 6 patients with 9 lesions. There was no false positive serum tumor marker elevation. The sensitivity and specificity using tumor marker elevation alone (based on the number of patients) for the indication of residual tumor viability were 42% and 100%, respectively. The positive predictive value was 100%, and the negative predictive value was 52% (Table 2).

Comparison of PET with the Combined Assessment by CT Scan, MRI, and Serum Marker Decline

For the blinded comparison of PET results with the combined assessment by CT scan/MRI plus serum tumor marker decline, the data from all 45 patients with a total of 85 lesions were included (Table 4). Of 85 residual lesions, viability was predicted correctly by the combination of CT scan and MRI plus tumor marker decline in 71 masses (83%). Thirty-eight lesions progressed or were stable on CT scans/MRI during therapy, with serum tumor markers remaining elevated or even increasing in these patients. Either these patients experience disease recurrence during follow-up, or histology of the resected mass revealed viable carcinoma/teratoma. PET imaging performed at the time of restaging after the end of therapy was also positive in 24 of these lesions, whereas it falsely indicated the absence of viable carcinoma/teratoma in 14 residual lesions. CT scans/MRI findings as well as tumor marker levels showed a complete or partial remission after chemotherapy in all 33 residual lesions that contained necrosis or remained in remission during follow-up. PET also was predictive in 30 of 33 lesions yet remained markedly positive in 3 residual lesions at the time of restaging (2 with inflammation and 1 with necrosis; see above). However, in 14 residual masses, the combination of CT scan/MRI and tumor marker decline falsely suggested a nonviable mass. In five of these lesions, a positive PET scan after chemotherapy correctly predicted the presence of viable carcinoma/teratoma.

Table 4. Comparison of Positron Emission Tomography with Computed Tomography/Magnetic Resonance Imaging Scan plus Serum Tumor Markers (N = 45 patients; 85 residual lesions)
PET (no. of lesions)CT/MRI and tumor markers (no. of lesions)
Correct positiveCorrect negativeNot correctTotal
  1. CT/MRI: computed tomography/magnetic resonance imaging; PET: positron emission tomography.

Correct positive240529
Correct negative030333
False positive0303
False negative140620

Evaluation of Residual Masses by PET Imaging in Patients with Primary Tumors Without Teratomatous Elements

Because it was shown that two of three histologically confirmed, false negative PET findings were due to the presence of mature teratoma, a separate evaluation that excluded patients with primary tumors containing teratomatous elements was performed. The residual lesions from patients with primary tumors containing teratomatous elements had a greater probability of containing mature teratoma in residual lesions after chemotherapy. Twenty-three patients with a total of 48 residual lesions were assessable for this analysis (Table 5). Thirty-three lesions were assessed correctly by PET: 15 lesions were assessed correctly as positive, and 18 lesions were assessed correctly as negative. In three patients, the PET image was considered positive, but the patients had a favorable outcome (see above): CT scans/MRI showed remission of these three lesions. Moreover, there were still 12 residual masses without an increased F-18 FDG uptake, yet these patients experienced disease progression within the first 6 months after therapy. CT scan/MRI correctly predicted the presence of viable carcinoma/teratoma in 8 of these 12 lesions; whereas, in the remaining 4 residual masses, CT scans/MRI also produced false negative results. Thus, with a sensitivity of 56% and a specificity of 86% and with a negative predictive value of 60% and a positive predictive value of 83%, the diagnostic efficacy of PET imaging did not improve when patients with primary tumors containing teratomatous elements were excluded from the analysis. Similar to the whole group of patients, quantification of F-18 FDG uptake did not provide additional information (sensitivity, 59%; specificity, 79%; negative predictive value, 61%; positive predictive value, 77%).

Table 5. Comparison of the Sensitivity, Specificity, and Negative and Positive Predictive Values of Positron Emission Tomography (Visual assessment) of All 85 Residual Masses with a Cohort of All Residual Masses and of Patients Without Teratomatous Elements in their Primary Histology (N = 48 lesions)a
ParameterAll massesWithout primary teratomatous histology
%95% CI%95% CI
  • 95% CI: 95% confidence interval.

  • a

    The patients without teratomatous elements in their primary tumors had an extremely low risk of developing mature teratomas postchemotherapy.

Negative predictive value6248–75%6041–77%
Positive predictive value9175–98%8359–96%


The accurate assessment of residual masses after chemotherapy for patients with NSGCT remains a diagnostic problem, because the differentiation between viable carcinoma, mature teratoma, or scar tissue/necrosis has been possible only after patients undergo surgical resection and histologic work-up. Patients with residual masses that contain viable carcinoma have a worse prognosis compared with patients who have residual necrosis only, even if the residual masses have been resected completely.20, 21 Because approximately 50% of all residual masses contain necrosis only, many patients do not benefit from postchemotherapy surgical resection.4, 21–23 Although surgical morbidity and mortality rates have been reduced substantially during the past decade, the ability to identify these patients with necrotic residual lesions by noninvasive methods would allow the avoidance of unnecessary surgery.24–26 However, there are no established criteria to identify patients in whom secondary surgery can be avoided. Both CT scan as well MRI parameters have not been good predictors of the viability of residual masses.27–30 Other studies have investigated clinical prognostic factors for the prediction of residual mass viability after chemotherapy. Clinical criteria, such as a residual mass measuring > 3 cm in greatest dimension and several others, have been investigated that decide in favor of adjunctive surgical resection rather than observation.20, 22, 29, 31–36 Combining different criteria, Steyerberg et al. developed a statistical model for the prediction of histologic findings in residual retroperitoneal masses based on the following factors: presence of teratomatous elements in the primary tumor, prechemotherapy levels of serum tumor markers, size of the residual mass, and percentage of shrinkage in greatest mass dimension. A good discriminative ability was achieved for the prediction of necrosis, whereas the model only showed poor discriminative power for the distinction between carcinoma and mature teratoma.22, 34

It was the objective of the current study to determine whether PET, as a method for the visualization and quantification of regional glucose metabolism within the body, can be used for the evaluation of residual masses in patients with NSGCT. In addition, the value of PET imaging was compared with the commonly used means of assessment, such as CT scans, MRI, and changes in serum tumor marker levels.

The results of the current study show that a positive PET image after treatment was a strong predictor for the presence of viable carcinoma/teratoma, because 9 of 10 patients with high F-18 FDG uptake levels either had a histology of viable carcinoma/teratoma or had a recurrence within 6 months after treatment (positive predictive value, 91%). However, there were also three false positive PET results. Histology revealed inflammation in two patients and necrosis in the third patient. F-18 FDG is not a tumor specific agent, and it also may accumulate in tissue macrophages. This phenomenon has been reported as a source of false positive PET examinations in patients with malignant disease.14, 37–40 To reduce the rate of false positive findings due to inflammation, the interval between the end of treatment and the PET examination was required to be at least 3 weeks. In addition, it was helpful to correlate the PET results with both the morphologic information derived from the CT scan and with the decline of previously elevated tumor markers. Conversely, a negative PET after the end of treatment does not allow one to avoid surgery, because 37% of all negative PET masses either progressed during 6 months of follow-up, or histologic examination revealed mature teratoma. Semiquantitative calculation of glucose uptake by using SUVs increased the sensitivity of PET from 59% to 68%, which, nevertheless, remains too low to alter definitive treatment decisions. Moreover, neither visual judgment nor the calculation of SUVs allowed a reliable differentiation between mature teratoma and necrosis, because SUVs for necrosis and teratoma were essentially the same. Similar results were obtained by Stephens et al. and Cremerius et al., who found that the presence of mature teratoma was the most common cause for false negative PET results in residual lesions from patients with NSGCT.14, 38 In the current study, the sensitivity and specificity of F-18 FDG PET alone for the prediction of residual mass viability were 59% and 92%, respectively, which are similar to the results obtained by other authors.14, 38, 39, 41, 42 Based on the assessment of sensitivity and specificity of either the radiologic response alone or the changes in tumor marker levels alone, the corresponding values were 55% and 42% for sensitivity and 86% and 100% for specificity, respectively. This indicates that no method in itself appears sufficiently accurate to predict the viability of residual masses.

To reduce the likelihood of the presence of mature teratoma as the assumed main source of false negative findings in residual masses, a separate analysis was performed that included only patients without teratomatous elements in the primary histology.14, 38 However, there were still 12 false negative PET results, suggesting that false negative findings may not only occur due to the presence of mature teratoma (Table 5).

In contrast to the previous PET investigations in patients with NSGCT by Stephens et al. and Nuutinen et al., which primarily included patients with retroperitoneal metastases and a high proportion of postchemotherapy necrosis/teratoma, our study focused on a specific high-risk population.38, 39 Many patients had advanced disease with several metastatic sites, and all underwent high-dose chemotherapy due to their overall poor prognosis. Thus, viable carcinoma and tumor progression after high-dose chemotherapy occurred in about 50% of patients.

Furthermore, to our knowledge, this is the first study to compare prospectively F-18 FDG PET with established criteria for the assessment of response in patients with NSGCT. In addition, this study was performed with blinded readings of CT scans, MRIs, and PET images, with none of the diagnostic physicians aware of the corresponding test results. Thus, the differences in sensitivity and specificity of all methods can be compared directly. Overall, the diagnostic efficacy was similar for all three diagnostic methods. Compared with CT scans/MRI and tumor markers, PET seems to offer additional information. A potential benefit of PET was observed in patients with stable disease or with disease in remission on CT scans/MRI and/or declining or normalized tumor markers, in whom elevated F-18 FDG uptake correctly predicted the presence of viable carcinoma/teratoma for five residual lesions. Also, PET may offer valuable information in addition to CT scans in patients with serum tumor marker negative disease, particularly when tumor remission is diagnosed using CT scans. In contrast, persisting or even increasing tumor markers as well as progressive disease on CT scans/MRI during chemotherapy are strong predictors for the presence of viable carcinoma/teratoma, and PET does not provide additional information in these patients.

PET also appears to yield important information regarding patients with multiple residual masses. The management of several residual masses of NSGCT at different localizations within a single patient is particularly difficult, because complete resection of all lesions often is impossible. When PET results were uniformly positive or negative for all lesions within a single patient, the clinical behavior of the multiple residual masses with respect to remission or progression also was uniform in our study. In these patients, a uniformly positive PET appears to be a strong predictor for viable carcinoma, because no false positive PET results were observed. Hartmann et al. found dissimilar histologic results at different anatomic localizations in approximately 30% of patients who underwent resection at multiple localizations.4 In patients who underwent resection of a necrotic lesion at one localization and had PET images that were uniformly negative for all lesions, the probability of necrosis at the remaining nonresected lesions clearly was > 90% (based on the negative predictive value of a quantitatively assessed PET of 67%). Thus, PET may add important information in patients with multiple, incompletely resectable, residual masses.

However, the limitations of the current study also must be considered. The limited number of patients included may have given rise to misleading results. Larger studies are necessary to confirm our results. A possible second limitation may be the method used to confirm the prediction of PET. Although it would have been ideal to have a histologic diagnosis of all tumors at the time of PET assessment, this was feasible in only 33% of all lesions. The results are based largely on the patient follow-up over a 6-month period after final assessment. However, because recurrent or progressive germ cell carcinoma in high-risk patients, it is once present, usually occurs rapidly after the end of treatment, the data obtained in this study still seem valid. Considering the median follow-up of 27 months, we are confident that the results would not change significantly with longer follow-up.

In conclusion, PET performed in conjunction with conventional staging methods appears to offer additional information for the prediction of residual mass histology in patients with NSGCT in several treatment situations. However, the limitations of PET imaging must be considered. A negative PET after chemotherapy may represent either necrosis or mature teratoma, whereas a positive PET image is highly predictive for the presence of viable carcinoma. Other useful indications for PET imaging in patients with NSGCT include the assessment of patients with multiple residual lesions and patients with tumor marker negative disease.