SEARCH

SEARCH BY CITATION

Keywords:

  • postchemotherapy;
  • retroperitoneal lymph node dissection;
  • viable tumor;
  • predictors;
  • risk categorization

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

BACKGROUND

The presence of viable tumor in the surgical specimen after postchemotherapy retroperitoneal lymph node dissection (PC-RPLND) is associated with an increased risk of disease progression. The objective of this study was to determine whether the presence of viable tumor in the surgical specimen could be predicted.

METHODS

Between 1980 and 2003, 236 patients underwent PC-RPLND for clinical Stage IIA or III nonseminomatous germ cell tumors (NSGCT). The authors retrospectively reviewed the medical records of those patients for pertinent clinical and treatment-related outcomes. A multivariate logistic regression analysis was used to evaluate whether clinical parameters were capable of predicting the presence of viable tumor in the surgical specimen.

RESULTS

International Germ Cell Consensus Classification (IGCCC) risk categories could be assigned to 218 patients, with 101 patients in the good-risk category, 32 patients in the intermediate-risk category, and 85 patients in the poor-risk category. The incidence of viable tumor in the good-risk, intermediate-risk, and poor-risk categories was similar (17.8%, 15.6%, and 15.3%, respectively); however, the risk categories predicted disease-specific and recurrence-free survival (P = .022 and P < .0001, respectively). On multivariate analysis, an elevated serum α-fetoprotein (AFP) level prior to PC-RPLND (P = .05) and the size of the retroperitoneal mass on pathology review (P = .02) were predictive of viable tumor in the surgical specimen.

CONCLUSIONS

Although IGCCC risk categories were correlated with disease-related outcomes, the risk groups had similar incidences of viable tumor. Elevated serum AFP levels prior to surgery and the size of the retroperitoneal mass in the resected specimen may help to predict viable tumor in the PC-RPLND specimen. Cancer 2006. © 2006 American Cancer Society.

An estimated 8000 men in the U.S. will be diagnosed with testicular cancer this year, with 390 deaths expected from this disease.1 In 1997, the International Germ Cell Consensus Classification (IGCCC) enabled patients with nonseminomatous germ cell tumors (NSGCT) to be stratified into 3 risk groups (good risk, intermediate risk, and poor risk) with the objectives of selecting risk-adapted treatment strategies for them and allowing investigators more appropriate comparisons of regimens among clinical trials.2 Currently, most centers treat advanced NSGCT with systemic chemotherapy followed by postchemotherapy retroperitoneal lymph node dissection (PC-RPLND). On final pathology review, the surgical specimen contains fibrosis and/or necrosis in 40% to 50%, of patients, teratoma in 35% to 40% of patients, and viable tumor in 10% to 15% of patients.3 PC-RPLND may be beneficial only for patients who have teratoma or viable tumor in their surgical specimen; therefore, up to 50% of patients who undergo surgery may not derive any therapeutic benefit from it. Furthermore, patients who have viable tumor in their PC-RPLND specimen are at greater risk of disease progression and death4–6 such that, if we were able to predict the presence of viable tumor in the surgical specimen, then these patients potentially could benefit from more intensive preoperative chemotherapy.

Several studies have attempted to identify clinical parameters that may help predict PC-RPLND pathology. The results from a study of 16 patients with NSGCT suggested that an elevated pretreatment serum α-fetoprotein (AFP) level and the size of the residual mass may be potential predictors of viable tumor in the PC-RPLND specimen; however, that study was severely underpowered.7 In another series, Oldenburg et al. were not able to identify any prechemotherapy or postchemotherapy clinical or radiologic parameters that could help predict the histology of the residual mass in patients with clinical Stage IIA NSGCT, and those authors proposed that PC-RPLND remained necessary for their patients.8

The objective of the current study was to determine whether select clinical parameters could be used to predict the pathology of the surgical specimen in a cohort of patients who underwent PC-RPLND. We also assessed the impact of the IGCCC risk category on the incidence of viable tumor in the surgical specimen.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

Study Design

Prior to conducting this study, a retrospective chart review protocol was approved by our Institutional Review Board. From our tumor registry, 408 males with primary testicular cancer were identified who were diagnosed between January 1980 and July 2003. Of 377 patients for whom complete medical charts were available, 141 patients had clinical Stage I or IS disease and were excluded, so that our study population consisted of 236 patients, all of whom had undergone PC-RPLND. Clinical staging was assigned according to the 1997 Tumor, Lymph Node, Metastasis (TNM) staging system. Patients were categorized into 3 prognostic groups (good-risk, intermediate-risk, and poor-risk) according to the 1997 IGCCC guidelines.2 According to those guidelines, the good-prognosis IGCCC group was defined as patients who had a testis/retroperitoneal primary tumor, no nonpulmonary visceral metastases, and favorable tumor markers (AFP < 1000 ng/mL, β-human chorionic gonadotropin [HCG] < 5000 IU/L, and lactate dehydrogenase [LDH] < 1.5 × the upper limit of normal [at our center, the upper limit of normal LDH is 618 IU/L]). The intermediate-prognosis IGCCC group was defined as patients who had a testis/retroperitoneal primary tumor, no nonpulmonary visceral metastases, and intermediate-risk tumor markers (AFP ≤ 10,000 ng/mL, HCG ≤ 50,000 IU/L, and LDH ≤ 10 × the upper limit of normal). The poor-prognosis IGCCC group was defined as patients who had a mediastinal primary tumor, or nonpulmonary visceral metastases, or high-risk tumor markers (AFP > 10,000 ng/mL, HCG > 50,000 IU/L, and LDH > 10 × the upper limit of normal). IGCCC risk categories were assignable to 218 patients; the remaining 18 patients could not be assigned to any risk group because of missing information regarding their tumor markers. In 88 patients, the specific serum LDH level at diagnosis was not available; however, in all of those patients, the serum LDH level was stated as normal in the medical records.

For all patients, a complete medical evaluation was performed prior to the PC-RPLND, including serum tumor markers (AFP, HCG, and LDH) and radiologic evaluation, which included chest and abdominal/pelvic computed tomography (CT) scans. The greatest circumferential dimension of the retroperitoneal mass, as determined by abdominal imaging prior to surgery, was noted. Clinical variables, including diagnostic findings, treatment type, and patient outcomes, were recorded along with many intraoperative and post-operative surgical variables. All patients received systemic chemo therapy prior to PC-RPLND (mean, 6.6 preoperative cycles; median, 5 preoperative cycles). The chemotherapeutic regimens that were selected for patients varied, depending on the protocols used around the time of diagnosis. During the 1980s, we administered alternating cycles of combined cyclophosphamide, doxorubicin, and cisplatin (CISCA) and combined vinblastine and bleomycin (VB) and gave 2 cycles beyond the normalization of serum tumor markers for all patients with NSGCT, irrespective of the risk category (patients received a minimum of 4 chemotherapy cycles).9 In the 1990s, we continued administering the alternating CISCA and VB regimen with a 20% dose reduction and achieved equivalent survival rates.10 In patients with high-volume disease, which corresponds to the poor- risk IGCCC category, we administered alternating dose-dense chemotherapy.11 Before 1997, patients who had what we defined as small-volume disease received the carboplatin, etoposide, and bleomycin (CEB) regimen;12 however in recent years, 3 cycles of the bleomycin, etoposide, and cisplatin (BEP) regimen or 4 cycles of the etoposide and cisplatin (EP) regimen were given to patients who had good-risk NSGCT. First-line chemotherapy was received by 185 patients (78.4%), and second-line chemotherapy was received by 51 patients (21.6%) prior to surgery. Using a definition of elevated preoperative serum tumor markers as a serum AFP > 15.0 ng/mL and/or HCG > 2.2 IU/mL, as described previously,13 33 patients (14%) underwent PC-RPLND in the context of elevated serum tumor markers. The criteria for performing surgery in the presence of elevated preoperative tumor markers included 1) stabilization of tumor markers prior to surgery, 2) elevation of tumor markers believed to be secondary to nontumor-related factors (e.g., liver dysfunction), or 3) if the patient had received the maximal dose of chemotherapy as assessed by the treating physician. A full bilateral template PC-RPLND was performed in all patients, and all pathologic specimens from the testis primary tumor and from the retroperitoneum or other metastatic sites were reviewed by 1 of our institution's genitourinary pathologists. The pathologic findings of the PC-RPLND specimens were categorized as either complete necrosis/ fibrosis, pure teratoma (immature or mature), or viable malignant tumor.8 Surgical specimens that contained any component of viable cancer were classified as a viable malignant tumor. Most patients who had viable malignant tumor in their PC-RPLND specimen received adjuvant chemotherapy: The regimen and the number of cycles were determined by the treating genitourinary medical oncologist.

Statistical Analysis

The methods of Kaplan and Meier14 were used to estimate the median disease-specific survival (DSS) and recurrence-free survival (RFS) rates. For the analysis of DSS, disease-specific death was the only event noted, with patients who survived censored at the date of their last follow-up. Disease-specific death was defined as death resulting from testicular cancer or its treatment. Patients who died of other causes were censored at the date of their death.

For the analysis of RFS, recurrence was noted as an event. A logistic regression model was used to estimate the probability (P) that a patient had viable tumor in their PC-RPLND specimen.15 The model predicted the natural log of the odds ( equation image) that a patient would be in 1 category (having viable tumor) or another (not having viable tumor). The model was done in a univariate fashion. From this model, we estimated the odds ratio for each potential prognostic factor with a 95% confidence interval. Then, all potential prognostic factors with a P value < .25 from the univariate analysis were included in a saturated model, and backward elimination was used to remove factors from the model based on the likelihood-ratio test in the multiple regression analysis. All statistical analyses were performed by using SAS software (version 9.1; SAS Institute, Cary, NC) with the statistical significant level set at .05.

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

Patient Characteristics

In total, 236 patients with testis cancer underwent PC-RPLND. Patient characteristics and clinical variables are listed in Table 1. The majority of patients (86%) presented with local signs and symptoms of testis cancer, and a painless or painful testicular mass was the most common clinical presentation. Prior to orchiectomy, the median serum tumor markers were AFP, 105.8 ng/mL (normal range, 0–5.0 ng/mL); HCG, 30.6 IU/L (normal range, 0–1 IU/L); and LDH, 485 IU/L (normal range, 313–618 IU/L). The most common primary testis tumor histology was mixed NSGCT (76.3%), with pure embryonal carcinoma (14.8%), pure teratoma (8.5%), and pure yolk sac (0.4%) the remaining tumor histologic types. The clinical stage of the primary tumor most frequently was T1 (63.1%) or T2 (23.2%). Most patients who underwent PC-RPLND had advanced clinical stage testis cancer (Stage IIC in 41.1% of patients and Stage III in 25.9% of patients). On preoperative abdominal imaging, the mean and median greatest dimensions of the retroperitoneal mass were 5.7 cm and 4.0 cm, respectively. Prior to PC-RPLND, the median serum tumor markers were AFP, 3.2 ng/mL; HCG, 1.0 IU/L; and LDH, 436 IU/L. The majority of patients (42.8%) were in the good-risk group according to the IGCCC risk categorization, and the remaining patients were in the intermediate-risk group (13.6%) or the poor-risk group (36.0%). The median number of lymph nodes removed was 19 per PC-RPLND specimen, and the most common pathology in the surgical specimen was teratoma (41.5%) or fibrosis (41.1%). Viable tumor was present in the remaining 17.4% of specimens. The majority of patients (85%) with viable tumor in their surgical specimen received adjuvant chemotherapy.

Table 1. Descriptive Statistics for Patients with Testicular Cancer after Postchemotherapy Retroperitoneal Lymph Node Dissection
VariableNo. of patients (%)MeanMedian (Range)
  • AFP indicates α fetoprotein; HCG, β human chorionic gonadotropin; LDH, lactate dehydrogenase; PC-RPLND, postchemotherapy retroperitoneal lymph node dissection; NSGCT, nonseminomatous germ cell tumor; IGCCC, International Germ Cell Consensus Classification; N/A, not available.

  • *

    For 3 patients with testicular cancer, the clinical stage of the primary tumor was not available.

Age at diagnosis, y 29.228.0 (15.1–58.9)
Tumor markers at diagnosis
 AFP, ng/mL 3477.8105.8 (0–65,255)
 HCG, IU/L 15,939.630.6 (0–60,006.5)
 LDH, IU/L 888.3485.0 (76–8856)
Tumor markers prior to PC-RPLND
 AFP, ng/mL 22.93.2 (1–1866.6)
 HCG, IU/L 3.71.0 (1–343)
 LDH, IU/L 478.4436.0 (15.1–6000)
No. of chemotherapy cycles prior to PC-RPLND 6.65.0 (2–20)
Greatest dimension of retroperitoneal mass on preoperative imaging, cm 5.74.0 (1–25)
Greatest dimension diameter of retroperitoneal mass on pathology, cm 4.63.5 (0.6–21)
Clinical classification of primary tumor (n = 233 patients)*
 T1147 (63.1)  
 T254 (23.2)  
 T324 (10.3)  
 Tx8 (3.4)  
Vascular invasion (n = 236 patients)
 No163 (69.1)  
 Yes73 (30.9)  
Clinical stage (n = 236 patients)
 Stage IIA19 (8.1)  
 Stage IIB59 (25.0)  
 Stage IIC97 (41.1)  
 Stage III61 (25.9)  
Histology of testis primary tumor (n = 236 patients)
 Viable NSGCT216  
 Mixed180 (76.3)  
 Embryonal35 (14.8)  
 Yolk sac1 (0.4)  
 Teratoma20 (8.5)  
Tumor site (n = 236 patients)
 Retroperitoneum only175 (74.2)  
 Retroperitoneum and others61 (25.8)  
Pathology of retroperitoneal mass
 Viable tumor41 (17.4)  
 Fibrosis97 (41.1)  
 Teratoma98 (41.5)  
IGCCC risk categorization (n = 236 patients)
 Good risk101 (42.8)  
 Intermediate risk32 (13.6)  
 Poor risk85 (36.0)  
 Risk category N/A18 (7.6)  

IGCCC Risk Groups

Table 2 shows that the incidence of viable tumor in the IGCCC good-risk, intermediate-risk, and poor-risk categories was similar (17.8%, 15.6%, and 15.3%, respectively). In addition, the median number of chemo therapy cycles in the good-risk, intermediate-risk, and poor-risk groups was 4 cycles, 6 cycles, and 7 cycles, respectively. There was a statistically significant difference in the number of chemotherapy cycles between the good-risk and intermediate-risk groups (P = .03) and between the good-risk and poor-risk groups (P < .0001). A Kaplan–Meier estimate of DSS stratified by IGCCC risk category demonstrated the prognostic value of these risk groups (P = .022) (Fig. 1). Similarly, a statistically significant difference was noted between the IGCCC risk categories in terms of RFS (P < .0001) (Fig. 2). A subgroup analysis of the incidence of viable tumor in patients who had complete sets of tumor markers (excluding the 88 patients for whom the specific serum LDH value at diagnosis was not available, although the medical record stated that it was normal) was conducted with similar rates of viable tumor in all 3 IGCCC risk groups (good-risk, 19.2%; intermediate-risk, 10.3%; poor-risk, 14.8%; P = .86). There was a statistically significant difference in the number of preoperative chemotherapy cycles between the good-risk and intermediate-risk groups (median, 4 cycles and 6 cycles, respectively; P = .01) and between the good- risk and poor-risk groups (median, 4 cycles and 7 cycles, respectively; P < .0001). However, there was no statistically significant difference in the number of preoperative chemotherapy cycles between the intermediate-risk and poor-risk groups (median, 6 cycles and 7 cycles, respectively; P = .39).

thumbnail image

Figure 1. Kaplan–Meier estimates of disease-specific survival by International Germ Cell Consensus Classification (IGCCC) risk category.

Download figure to PowerPoint

thumbnail image

Figure 2. Kaplan–Meier estimates of recurrence-free survival by International Germ Cell Consensus Classification (IGCCC) risk category.

Download figure to PowerPoint

Table 2. Incidence of Viable Tumor and Amount of Chemotherapy According to International Germ Cell Consensus Classification Risk Group
IGCCC risk categoryTotal no. of patientsNo. of patients with viable tumorIncidence of viable tumor (%)Mean no. of chemotherapy cyclesMedian no. of chemotherapy cycles
  1. IGCCC indicates International Germ Cell Consensus Classification.

  2. *Comparison of amount of chemotherapy cycles between the IGCCC risk categories: Wilcoxon test P = .0001 for good risk vs. intermediate risk; P<.0001 for good risk vs. poor risk; and P = .15 for intermediate risk vs. poor risk.

Good risk1011817.85.74.0
Intermediate risk32515.66.66.0
Poor risk851315.38.07.0

Impact of PC-RPLND Pathology on Outcome

A Kaplan–Meier estimate of DSS by PC-RPLND pathology demonstrated the adverse prognosis associated with the presence of viable tumor in the PC-RPLND specimen (P = .004) (Fig. 3). Similarly, a Kaplan–Meier estimate of RFS by PC-RPLND pathology supported the higher risk of recurrence with the presence of viable tumor in the surgical specimen (P = .0002) (Fig. 4).

thumbnail image

Figure 3. Kaplan–Meier estimates of disease-specific survival according to postchemotherapy retroperitoneal lymph node dissection pathology.

Download figure to PowerPoint

thumbnail image

Figure 4. Kaplan–Meier estimates of recurrence-free survival according to postchemotherapy retroperitoneal lymph node dissection pathology.

Download figure to PowerPoint

Potential Predictors of Viable Tumor in the PC-RPLND Specimen

Univariate and multivariate logistic regression analyses were performed of potential predictors of having viable disease in the surgical specimen (Table 3). In the multivariate analysis, the presence of an elevated serum AFP level prior to PC-RPLND (P = .05) and large size of the retroperitoneal mass in the resected pathologic specimen (P = .02) predicted the presence of viable tumor in the PC-RPLND specimen. There were 14 patients who had an elevated AFP level (> 15 ng/mL) prior to PC-RPLND. Seventy-five percent of large retroperitoneal masses (defined as > 5 cm) were observed in the IGCCC intermediate-risk group (34%) and the poor-risk group (33%). In contrast, only 13 patients (13%) in the IGCCC good-risk group had large retroperitoneal masses. In the multivariate analysis, because the size of the retroperitoneal mass increased in increments of 1 cm in the surgical specimen, there was a 1.12 increased odds of having viable cancer. In a univariate assessment of the impact of the size of the retroperitoneal mass, patients who had a retroperitoneal mass > 10.3 cm on final pathology had a 25% probability of having viable tumor, whereas patients who had a retroperitoneal mass > 15.8 cm had a 37.5% probability of having viable cancer, and patients with a mass > 20.6 cm had a 50% probability of having cancer.

Table 3. Univariate and Multivariate Logistic Regression Analysis of Potential Predictors of Viable Tumor
VariableNo. of patients (Viable tumor mass)Parameter estimatePOR95% CI for ORMultivariate parameter estimateMultivariate P valueOR95% CI for OR
  1. OR indicates odds ratio; 95% CI, 95% confidence interval; AFP, α fetoprotein; HCG, β human chorionic gonadotropin; LDH, lactate dehydrogenase; IGCCC, International Germ Cell Consensus Classification, N/A, not available; PC-RPLND, postchemotherapy retroperitoneal lymph node dissection.

Clinical stage of primary tumor
 T1147 (30)1.00    
 ≥ T286 (11)−0.2793.140.570.27—1.21    
Tumor markers at orchiectomy
 AFP0.000018.341.001.00—1.00    
 HCG−3.17E—6.511.001.00—1.00    
 LDH−0.00058.190.9990.999—1.00    
Histology of testicular tumor
 Viable tumor211 (37)1.00    
 Nonviable tumor25 (4)−0.055.850.900.29—2.76    
Presence of vascular invasion
 No163 (34)0.4551.042.491.05—5.91    
 Yes73 (7)1.00    
Greatest dimension of retroperitoneal mass (preoperative), cm−0.015.770.990.89—1.09    
Clinical stage
 Stage IIA, IIB78 (13)−0.0371.840.930.45—1.91    
 Stage IIC, III158 (28)1.00    
IGCCC risk categorization
 Good risk101 (18)1.00    
 Intermediate risk32 (5)−0.079.770.850.29—2.52    
 Poor risk85 (13)−0.092.650.830.38—1.82    
 Risk category N/A18 (5)0.2865.331.770.56—5.62    
Tumor site
 Retroperitoneum175 (33)0.2158.311.540.67—3.55    
 Retroperitoneum and other sites61 (8)1.00    
Tumor markers prior to PC-RPLND
 AFP0.0023.071.0021.00—1.0050.00272.051.0031.00—1.005
 HCG−0.0031.770.9970.976—1.02    
 LDH−0.00014.801.000.999—1.001    
Total no. of chemotherapy cycles−0.0259.600.970.88—1.08    
Greatest dimension of retroperitoneal mass (on pathology), cm0.1064.021.111.02—1.210.111.021.121.02—1.23

Survival Analysis

All 236 patients were included in our survival analysis, with 191 patients who remained alive (180 patients were alive without disease, and 11 patients were alive with disease) and 45 patients who died. The median duration of follow-up among survivors was 51 months (range, 6–250 months). Among the patients who died, 30 patients (66.7%) died from testis cancer, and 15 patients (33.3%) died of other causes. The median time to death was 17 months (range, from 1 day to 236 months). The 5-year and 10-year DSS rates were 85% and 81%, respectively. In addition, 53 patients (22.5%) experienced recurrence of tumor (local recurrence in 18 patients and distant metastasis in 35 patients). The 5-year and 10-year RFS rates were 74% and 72%, respectively. The Kaplan–Meier survival curves for DSS and RFS are illustrated in Figure 5.

thumbnail image

Figure 5. Kaplan–Meier estimates of disease-specific survival (DSS) and recurrence-free survival (RFS).

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

In the current study, we reaffirmed the adverse prognosis associated with the presence of viable tumor in the PC-RPLND specimen. The identification of viable cancer in the surgical specimen was associated with inferior DSS and RFS, consistent with previous published studies.4–6, 16 Although the majority of patients with viable disease in their PC-RPLND specimen received adjuvant chemotherapy, they nevertheless had a worse prognosis than patients who exhibited either teratoma or necrosis/fibrosis in their surgical specimen. This places a significant burden on the genitourinary pathologist to evaluate the RPLND specimen adequately and to identify any viable cancerous elements, if present. The ability to predict which patients are at high-risk of viable tumor in the surgical specimen would enable us to select them for more aggressive systemic therapy prior to PC-RPLND and for more stringent surveillance and follow-up. In our multivariate analysis of potential clinical predictors of viable disease in the surgical specimen, we observed that an elevated serum AFP level prior to PC-RPLND and large size of the retroperitoneal mass in the resected pathologic specimen predicted the presence of viable cancer on final pathology. Although previous studies5, 7 have shown an association between elevated preoperative serum tumor markers and the likelihood of viable cancer in the surgical specimen, our study further validated those findings in our patient cohort using a multivariate analysis. However, because no clinical variable either alone or in combination is sufficiently accurate to predict the histology of the surgical specimen, we recommend that all patients who have residual disease after systemic chemotherapy for metastatic NSGCT should undergo PC-RPLND.

In the current study, one of the predictors of viable tumor in the surgical specimen was increasing size of the retroperitoneal mass in the resected pathologic specimen; however, the specimen is not available prior to surgery and cannot help the treating physician make a treatment decision. We were somewhat surprised that increasing size of the retroperitoneal mass as measured on pathology, and not on preoperative imaging, was predictive of the presence of viable tumor in the surgical specimen. We believe that this most likely reflects the large volume of disease and/or chemorefractory behavior of these large resected masses.

Two other groups of investigators reported that the pathology of retroperitoneal masses after chemotherapy could be predicted in part by the absence of a radiologic and tumor marker response to chemotherapy and by the size of the retroperitoneal mass.7, 17 Nevertheless, our findings are consistent with the previous study by Oldenburg et al., who proposed that no prechemotherapy or postchemotherapy clinical or radiologic parameter can predict the pathology of the residual mass in patients with clinical Stage IIA NSGCT.8 The ability to predict PC-RPLND pathology is not sufficiently accurate to be used in any treatment plan, and the failure to predict viable disease accurately may have very detrimental effects on patient outcome.

By using the 1997 IGCCC risk categorization,2 we were able to assign 218 patients to a specific risk group, with the majority of patients falling either into the good-risk IGCCC category (n = 101 patients) or into the poor-risk IGCCC category (n = 85 patients). The importance of the IGCCC risk categorization was demonstrated clearly in our patient population, because it allowed the stratification of patients in terms of DSS and RFS. This supports the utility of using the IGCCC criteria at our center and is consistent with previous studies that validated the prognostic value of these risk groups.18, 19 The incidence of viable disease after PC-RPLND was comparable in all 3 IGCCC risk groups (good-risk, 17.8%; intermediate- risk, 15.6%; poor-risk, 15.3%), which we believe resulted from a sequential increase in the number of cycles of preoperative chemotherapy in the intermediate-risk and poor-risk IGCCC groups. We speculate that the larger amounts of chemotherapy administered to patients in higher risk groups may have resulted in comparable incidences of viable tumor in the surgical specimen. Despite comparable incidences of viable tumor in the PC-RPLND specimens, these risk groups clearly differ in terms of their ultimate prognosis because of the higher incidence of visceral metastasis in intermediate-risk and poor-risk patients.

In a previous study by Fizazi et al., the incidence of viable cancer in patients who underwent PC-RPLND was only 9%; however, a cut-off value of at least 10% of the surgical specimen consisting of viable tumor was used to define that group.20 In our study, we defined viable tumor in the PC-RPLND specimen as the presence of any viable cancer in the resected specimen regardless of its proportions. Furthermore, we acknowledge, that the CEB regimen, which we had used in the past for small-volume disease (small volume would be equivalent to a good prognosis), may be inferior to BEP; therefore, it may have resulted in the higher than expected 17% viable tumor rate in the PC-RPLND specimens in patients with good-risk NSGCT. This rate was comparable to the viable tumor incidence rates in the intermediate-risk and poor-risk groups who received more aggressive chemotherapy. Thus, it is possible to speculate that, if we had used 3 cycles of BEP or 4 cycles of EP for good-risk disease as we do now, we may have observed an incidence ≤ 10% of viable tumor in good-risk patients, as noted by Fizazi et al.20

Several limitations to the current study need to be acknowledged. Our study was retrospective and suffers from an inherent selection bias, which is common with most PC-RPLND surgical series.5 Our reporting of surgical specimen pathology was based on a review of pathology reports, and there likely were some differences in the reporting among our genitourinary pathologists. In addition, our surgical series extended over 25 years, during which there were changes in preoperative chemotherapeutic regimens, such that our results will require validation by others using standardized preoperative chemotherapeutic regimens.

In conclusion, the current study highlights the importance of surgical histology on disease-related outcomes. Patients with viable disease in their PC-RPLND specimen are at increased risk of disease progression and require aggressive surveillance and management. Although the serum AFP level prior to PC-RPLND and large size of the retroperitoneal mass on final pathology help predict which patients are at increased risk of having viable cancerous elements in their specimen, these variables are not sufficiently accurate for predicting histology to constitute the basis of any treatment decision. Therefore, PC-RPLND still is required for all patients who have residual disease after chemotherapy for advanced NSGCT.

REFERENCES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES
  • 1
    Ries LAG, Eisner MP, Kosary CL, et al. SEER Cancer Statistics Review, 1975–2002 [based on November 2004 SEER data submission; posted to the SEER web site 2005]. Bethesda: National Cancer Institute, 2005.
  • 2
    [No authors listed.] International Germ Cell Consensus Classification: a prognostic factor-based staging system for metastatic germ cell cancers. International Germ Cell Cancer Collaborative Group. J Clin Oncol. 1997; 15: 594603.
  • 3
    Steyerberg EW, Keizer HJ, Fossa SD, et al. Prediction of residual retroperitoneal mass histology after chemotherapy for metastatic nonseminomatous germ cell tumor: multivariate analysis of individual patient data from six study groups. J Clin Oncol. 1995; 13: 11771187.
  • 4
    Sheinfeld J. The role of adjunctive postchemotherapy surgery for nonseminomatous germ-cell tumors: current concepts and controversies. Semin Urol Oncol. 2002; 20: 262271.
  • 5
    Donohue JP, Leviovitch I, Foster RS, Baniel J, Tognoni P. Integration of surgery and systemic therapy: results and principles of integration. Semin Urol Oncol. 1998; 16: 6571.
  • 6
    Hollender A, Stenwig EA, Ous S, Fossa SD. Survival of patients with viable malignant non-seminomatous germ cell tumour persistent after cisplatin-based induction chemotherapy. Eur Urol. 1997; 31: 141147.
  • 7
    Onozawa M, Kawai K, Yamamoto T, et al. Clinical parameters that predict histology of postchemotherapy retroperitoneal lymph node mass in testicular cancer. Int J Urol. 2004; 11: 535541.
  • 8
    Oldenburg J, Alfsen GC, Lien HH, Aass N, Waehre H, Fossa SD. Postchemotherapy retroperitoneal surgery remains necessary in patients with nonseminomatous testicular cancer and minimal residual tumor masses. J Clin Oncol. 2003; 21: 33103317.
  • 9
    Logothetis CJ, Samuels ML, Selig D, Swanson D, Johnson DE, von Eschenbach AC. Improved survival with cyclic chemotherapy for nonseminomatous germ cell tumors of the testis. J Clin Oncol. 1985; 3: 326335.
  • 10
    Fizazi K, Do KA, Wang X, Finn, L, Logothetis CJ, Amato RJ. A 20% dose reduction of the original CISCA/VB regimen allows better tolerance and similar survival rates in disseminated testicular non-seminomatous germ-cell tumors: final results of a Phase III randomized trial. Ann Oncol. 2002; 13: 125134.
  • 11
    Fizazi K, Do KA, Wang X, Finn L, Kim J, Daliani D. Alternating dose-dense chemotherapy in patients with high volume disseminated non-seminomatous germ cell tumors. Br J Cancer. 2002; 86: 15551560.
  • 12
    Prow DM, Amato RJ, Jones DM, Logothetis CJ. Carboplatin, etoposide, and bleomycin (CEB) combination chemotherapy for the treatment of small volume (SV) metastatic non-seminomatous germ cell tumor of the testis (NSGCT). Proc Am Soc Clin Oncol. 1997; 16: 340a. Abstract 1217.
  • 13
    Stephenson AJ, Bosl GJ, Motzer RJ, et al. Retroperitoneal lymph node dissection for nonseminomatous germ cell testicular cancer: impact of patient selection factors on outcome. J Clin Oncol. 2005; 23: 27812788.
  • 14
    Kaplan EL, Meier EP. Nonparametric estimation from incomplete observations. J Am Stat Assoc. 1958; 53: 457481.
  • 15
    Cox D. Regression models and life tables (with discussion). J R Stat Soc B. 1972; 34: 187220.
  • 16
    Stenning SP, Parkinson MC, Fisher C, et al. Postchemotherapy residual masses in germ cell tumor patients: content, clinical features, and prognosis.Medical Research Council Testicular Tumour Working Party. Cancer. 1998; 83: 14091419.
  • 17
    Fossa SD, Aass N, Ous S, et al. Histology of tumor residuals following chemotherapy in patients with advanced nonseminomatous testicular cancer. J Urol. 1989; 142: 12391242.
  • 18
    Schmoll HJ, Souchon R, Krege S, et al. European consensus on diagnosis and treatment of germ cell cancer: a report of the European Germ Cell Cancer Consensus Group (EGCCCG). Ann Oncol. 2004; 15: 13771399.
  • 19
    van Dijk MR, Steyerberg EW, Stenning SP, Dusseldorp E, Habbema JD. Survival of patients with nonseminomatous germ cell cancer: a review of the IGCC classification by Cox regression and recursive partitioning. Br J Cancer. 2004; 90: 11761183.
  • 20
    Fizazi K, Tjulandin S, Salvioni R, et al. Viable malignant cells after primary chemotherapy for disseminated nonseminomatous germ cell tumors: prognostic factors and role of postsurgery chemotherapy—results from an international study group. J Clin Oncol. 2001; 19: 26472657.