Is full bilateral retroperitoneal lymph node dissection always necessary for postchemotherapy residual tumor?




Traditionally, postchemotherapy (PC) surgery for metastatic nonseminomatous germ cell tumor (NSGCT) has used a full bilateral retroperitoneal lymph node dissection (RPLND) from the crus of the diaphragm to the bifurcation of the common iliac arteries, from ureter to ureter. With the primary landing zone well defined in low-volume retroperitoneal disease, the authors performed modified dissections in the PC setting in a select population; and, herein, they report disease outcome.


From 1991 to 2004, a retrospective review of the testicular cancer database at the authors' institution was performed to identify patients with NSGCT, normal serum tumor markers after cisplatin-based chemotherapy, and residual retroperitoneal tumor who underwent modified PC-RPLND. All patients had metastatic disease at initial presentation that was limited to the primary landing zone (left or right).


One hundred patients were identified, including 43 who underwent a right modified template, 18 patients who underwent a left full modified template, and 39 patients who underwent a left modified template. Pathology revealed cancer in 2% of patients, teratoma in 62% of patients, and necrosis in 36% of patients. The 2- and 5-year disease-free survival rate was 95%, and the median follow-up was 31.9 months (range, 1–152 months). Four patients developed recurrent disease with a median time to recurrence of 8.25 months (range, 6–11 months). All recurrences were outside the boundaries of a full bilateral RPLND.


Selected patients at PC surgery can be managed with modified PC-RPLND. Cancer 2007. © 2007 American Cancer Society.

Retroperitoneal mapping studies have accurately documented the predictable lymphatic spread of testicular cancer, allowing for the use of more limited surgical templates in low-volume (<5 cm) metastatic disease.1 In the surgical management of clinical stage A disease, surgical boundaries over the last 30 years have evolved from a full bilateral, suprahilar dissection to a unilateral nerve-sparing template without compromising surgical cure.2 In the management of postchemotherapy retroperitoneal masses, there has been no such consistent reduction in the retroperitoneal surgical boundaries, and a full bilateral dissection is considered standard therapy by many investigators.

Treatment options for patients with low-volume, metastatic germ cell cancer consist of either primary retroperitoneal lymph node dissection (RPLND) or 3 or 4 courses of induction chemotherapy, followed by surgery if the tumor persists in the retroperitoneum. Because of the acute toxicity of cisplatin-based chemotherapy in the 1970s and 1980s, low-volume retroperitoneal tumors were managed more commonly with primary RPLND in an attempt to avoid chemotherapy. With the advent of effective antiemetics, growth factors, and supportive care, theacute toxicity of cisplatin-based chemotherapy has decreased. Subsequently, a greater number of patients with low-stage disease currently are treated primarily with chemotherapy.3

Many patients from the 1970s and 1980s who underwent postchemotherapy surgery had high-volume residual disease; thus, the decision to perform a full bilateral RPLND was rational and appropriate. Because chemotherapy currently is being administered for relatively low-volume retroperitoneal disease, and because these tumors typically are restricted to the primary landing zone of the affected testicle, a question has arisen about the appropriateness of full bilateral RPLND in this population.

The potential benefits to the patient from a reduction in the postchemotherapy RPLND (PC-RPLND) template include shorter operative times, reduced fluid requirements, and maintenance of emission and ejaculation without using nerve-sparing techniques. Of course, any modification in the surgical template must not be accompanied by higher local recurrence rates or decreased disease-free survival (DFS). The objective of the current study was to determine whether these goals have been achieved in this retrospective cohort.


The testis cancer database at our institution was queried to identify all patients who underwent modified PC-RPLND from 1991 to 2004. The charts of these patients were reviewed retrospectively to identify patients who presented with metastatic, nonseminomatous germ cell cancer with normal serum tumor markers and residual retroperitoneal disease after initial cisplatin-based chemotherapy. Only patients with tumor limited to the primary landing zone of the affected testis, both before and after chemotherapy, were included. Patients with late recurrent disease (disease recurrence >2 years after the last treatment), desperation surgery (rising serum tumor makers during or after chemotherapy), or redo-resection were excluded.

The operative report for each patient was reviewed to identify the location of tumor, extent of surgery, and whether additional procedures or ipsilateral, efferent, sympathetic, nerve-sparing was performed. A right-sided, modified PC-RPLND included removal of the right common iliac (medial to the ureter), paracaval, precaval, retrocaval, and interaortocaval (IAC) lymph nodes. A left modified PC-RPLND included removal of the preaortic, periaortic, and left common iliac lymph nodes. With the additional removal of IAC lymph nodes superior to the inferior mesenteric artery (IMA), the procedure was denoted as “full left modified.” The right and left iliac lymph nodes medial to the ureter were removed and sent with the paracaval and periaortic lymph node packages, respectively. The decision to perform a left modified or a full left modified dissection was surgeon preference.

Blood loss, perioperative complications, and hospital stay also were evaluated. Length of follow-up was calculated from the date of surgery to the date of last clinic visit or the date of recurrence. SPSS version 13.0 software was used for statistical analysis.


The study cohort included 100 patients. Ninety-eight patients had good-risk disease, as defined by the International Germ Cell Consensus Group,4 and received induction chemotherapy only. Two patients with poor-risk disease (defined by marker elevation) had received second-line chemotherapy before surgery because of persistent marker elevation. Forty-three patients underwent a right modified dissection, 18 patients underwent a left full modified, and 39 patients underwent a left modified template. Ninety-four patients had clinical stage B1 disease (lymph node <2 cm) or B2 disease (lymph node <5 cm) before and after chemotherapy. Six patients presented with clinical stage B3 disease (lymph nodes >5 cm). Pathology revealed viable germ cell cancer in 2% of patients, teratoma in 62% of patients, and fibrosis in 36% of patients. Pathology for each surgical template is specified in Table 1.

Table 1. Retroperitoneal Pathology After Modified Retroperitoneal Lymph Node Dissection
TemplatePathology: No. of patients (%)
Right modified (n = 43)030 (69.8)13 (30.2)
Left full modified (n = 18)1 (5.6)9 (50)8 (44.4)
Left modified template (n = 39)1 (2.6)23 (59)15 (38.4)
Total (n = 100)2 (2)62 (62)36 (36)

Thirty of 43 patients (69.8%) who underwent a right-sided dissection had teratoma in the surgical specimen. Location of teratoma for right-sided dissection included the IAC in 20 patients (67%), with 14 patients having teratoma only in the IAC region. No patient had pathologically confirmed cancer. Of the 18 patients who underwent a left full modified dissection, 10 patients (55.6%) had nonnecrosis pathology (9 with teratoma and 1 with active germ cell cancer). In the left modified template group (n = 39 patients), teratoma was identified in 23 patients (59%), and cancer was identified in 1 patient (2.6%). The distribution of lymph node metastases is provided in Table 2.

Table 2. Distribution of Lymph Node Metastasis (Teratoma/Cancer) After Modified Dissection
LocationRight modified (n = 30)Left full modified (n = 10)Left modified (n = 24)
  1. NA indicates not applicable.

Right gonadal vein1NANA
Left gonadal veinNA11

Nephrectomy was performed in 7 patients, all undergoing a left modified template dissection. Three patients had complete ureteral encasement by tumor, 3 patients had tumor adherent to the renal vein, and 1 patient with a deep venous thrombosis and inferior vena caval filter had suspected tumor thrombus in the left renal vein. Sympathetic efferent ipsilateral nerve sparing was performed in 16 patients, on the right in 10 patients and for a full and modified left template in 3 patients each.

Intraoperative blood loss ranged from 175 mL to 1000 mL (mean, 318 mL). Seventy percent of patients were managed without a nasogastric tube. The mean and median time to discharge was 4.4 days and 4.0days, respectively (range, 2–56 days). Eighty-one patients were discharged by Day 4. One patient with previously undiagnosed antiphospholipid antibody syndrome had a prolonged ventilatory course and was discharged on Day 56. There were 16 complications, and the most common was an ileus presenting with nausea and/or vomiting in 6 patients, all of whom were managed conservatively (Table 3). Two patients required re-exploration, including 1 patient with a symptomatic lymphocele 1 month postoperatively and 1 patient who developed a small bowel obstruction 6 weeks postoperatively.

Table 3. Perioperative Complications in 100 Patients Who Underwent a Modified Postchemotherapy Retroperitoneal Lymph Node Dissection
ComplicationNo. of patients
Prolonged ventilation1
Ureteral injury1
Small bowel obstruction1
Decreased oxygen saturation1
Renal artery injury (repaired)1
Incisional skin bleed1

The median follow-up for all 100 patients was 31.9 months (range, 1–152 months). The 2- and 5-year DFS rates both were 95%. The mean and median follow-up for the 62 patients who had teratoma were 33.6 months and 25.5 months, respectively; and, for the 36 patients who had necrosis, they were 30.5 months and 31 months, respectively. The mean and median follow-up for the 2 patients who had cancer were both 7.5 months: Both patients received adjuvant chemotherapy and currently are free of disease.

Four patients, all presenting with clinical stage B2 disease, developed recurrent disease at a median of 8.25 months (range, 6–11 months) (Table 4). Patient 1, who had poor-risk disease defined by tumor marker elevation, underwent a left full modified dissection with teratoma in the periaortic lymph node package. He subsequently developed recurrent teratoma in the chest at 8 months. The remaining 3 patients (all with good-risk disease) underwent a right modified dissection, all with teratoma in the IAC lymph nodes. One patient developed recurrent disease at 8 months over the right common iliac artery, lateral to the right ureter, out of the field of dissection, and underwent primary resection, which revealed teratoma. Patient 3 developed recurrent disease at 6 months with tumor in the retrocrural and right parapsoas region lateral to the ureter. It was suspected that this was recurrent teratoma, and treatment was primary resection. Pathology revealed teratoma and yolk sac tumor. Adjuvant chemotherapy was not administered. The fourth patient developed recurrent disease at 11 months and had teratoma identified in the right inguinal canal.

Table 4. Characteristics of Patients Who Underwent Modified Retroperitoneal Lymph Node Dissection With Recurrent Disease
PatientTemplatePathologyInitial locationTime to recurrence, moLocation of recurrenceTreatmentPathologyCurrent status
  1. NED indicates no evidence of disease; IAC, interaortocaval.

1Left full modifiedTeratomaPeriaortic8ChestResectionTeratomaNED
2Right modifiedTeratomaIAC8Right iliac arteryResectionTeratomaNED
3Right modifiedTeratomaIAC6Retrocrural right parapsoasResectionYolk sac and teratomaNED
4Right modifiedTeratomaIAC11Right inguinal canalResectionTeratomaNED

Twenty patients had follow-up <12 months. Excluding those 20 patients, the median follow-up for the remaining 76 patients who did not develop recurrent disease was 40.9 months, and the 2- and 5-year DFS rates both were 95%.


There is general agreement among testis cancer experts that residual retroperitoneal tumor after chemotherapy should be resected. The rationale for this is that, even in patients with serum tumor marker normalization, residual retroperitoneal tumor may contain teratoma or carcinoma, which ultimately may have a deleterious effect on long-term outcome. This resection of residual retroperitoneal tumor is not merely a biopsy or sampling. Indeed, complete resection is necessary for an optimal outcome.5 If, by the employment of a modified template in the postchemotherapy setting, the local recurrence rate increases, then a more limited template should not be used.

In a series of 113 patients who underwent full bilateral PC-RPLND, investigators at the Memorial Sloan-Kettering Cancer Center reported an 8% incidence of disease (cancer/teratoma) identified in the contralateral landing zone.6 All of the men in their cohort presented with “bulky disease” and were treated with cisplatin or carboplatin-based chemotherapy. The investigators concluded that a modified dissection should be considered in patients with 1) no palpable residual tumor mass, 2) left primary tumors, and 3) right primary tumors that have no evidence of cancer/teratoma on frozen section analysis of the residual mass. Rabbani et al. reported on 39 patients who underwent a bilateral PC-RPLND to determine the incidence of teratoma/cancer identified in the contralateral landing zone if a modified dissection was performed. One patient (2.6%) with a left-sided primary tumor had teratoma outside the boundaries of a modified template in the area of the contralateral landing zone.7 This cohort differs from the current series, in that 1) the boundaries of a right-sided template included the periaortic lymph nodes above the IMA, and 2) also included in this series were patients with residual tumor masses after chemotherapy on the contralateral landing zone of the primary tumor who would have been excluded from the current series.

Oldenberg et al. reported the results of 87 patients who underwent PC-RPLND in a group of patients with residual masses <20 mm.8 Pathology revealed necrosis in 67% of patients, teratoma in 26% of patients, and cancer in 7% of patients. There were 5 recurrences, and the 5-year DFS rate was 94% at a medium follow-up of 80 months. Fifty of the 87 patients underwent a “unilateral RPLND,” and 3 recurrences were reported, all outside the retroperitoneum, indicating that the template of dissection was adequate. Those authors did state concerns that a modified dissection may not entirely avoid the risk of a growing teratoma near the area of resection.

It was our hypothesis that a modified postchemotherapy dissection would adequately manage the retroperitoneum without an increased incidence of local recurrence in patients who presented with metastatic disease limited to the primary landing zone of the affected testicle. The reason for the generation of this hypothesis was the increasing use of chemotherapy to treat patients with low-volume, retroperitoneal-only, metastatic tumors.3 By selecting patients for this modified template who had unilateral tumor both before and after chemotherapy, we hoped to identify patients with no residual, contralateral, microscopic tumor. In this series of a 100 patients, at a median follow-up of 32 months, there were only 4 recurrences, all of which were outside the boundaries of a bilateral dissection and of which 3 were subdiaphragmatic. Specifically, the iliac recurrence was lateral to the ureter and, thus, was outside the boundaries of a full bilateral template. The parapsoas recurrence also was lateral to the ureter and likely was secondary to aberrant lymphatic or venous drainage of the gonadal vein. Chang et al. and others have hypothesized that such recurrences may be reduced with wide excision of the spermatic cord and surrounding lymphatic tissue.9 The third recurrence was along the inguinal canal and may have represented an inadequate radical orchiectomy, although the boundaries of the RPLND would have no impact on this recurrence. Our study suggests that in select patients with residual retroperitoneal masses after chemotherapy, a modified dissection is not associated with a higher recurrence rate than would be expected.

The current series included only patients who presented with low-volume, metastatic disease limited to the primary landing zone. This series of 100 patients was highly selective and represented <10% of PC-RPLND performed during the study period. The majority of patients who undergo postchemotherapy surgery at our university have bulky disease both before and after chemotherapy and are not candidates for a modified template. Likewise, because low-volume, stage B disease often is cured with primary chemotherapy alone, at Indiana University, we do not advocate PC-RPLND for patients who have normal computed tomographic scans after chemotherapy. In this population, retroperitoneal pathology rarely will be cancer; thus, a 2% incidence of cancer in the current series is not surprising. Furthermore, a higher incidence of teratoma would be expected in a group of patients with small-volume disease with minimal change postchemotherapy.

One concern against adopting a modified postchemotherapy template is the risk of residual microscopic teratoma in the contralateral landing zone with subsequent development of late recurrent disease or malignant transformation. It is unknown whether the contralateral side at initial presentation harbored micrometastatic disease or not. Nevertheless, the ipsilateral tumor dictates therapy and is treated (along with the radiographically normal, contralateral side) with 3 or 4 courses of platinum-based chemotherapy. It is generally agreed that residual ipsilateral tumor should be resected, but how best to manage the contralateral side?

This dilemma is analogous to the management of high-risk, clinical stage A disease with adjuvant chemotherapy. This population, which has a nearly 50% chance of harboring micrometastatic disease (much higher than would be observed in the “normal” contralateral side in the current series), almost universally is cured with 2 courses of chemotherapy (with no subsequent resection). The primary criticism to adjuvant chemotherapy in clinical stage A disease typically is not directed toward the efficacy of chemotherapy in managing micrometastatic disease but, rather, toward the toxicity of treatment for patients without metastatic disease. Cullen et al. reported on outcomes of 114 patients with high-risk stage A NSGCT who were treated with 2 courses of combined bleomycin, etoposide, and cisplatin.10 At a median follow-up of 4 years, there were 2 recurrences, 1 at 7 months in the retroperitoneum, liver, and lung and the other at 18 months in the groin. Other series have reported similar success with long-term follow-up.11, 12

Thus, by extrapolating data from clinical stage A studies, along with emerging data from other modified postchemotherapy series,13–15 a cogent argument can be made to limit the boundaries of dissection in highly select patients undergoing postchemotherapy surgery. Using a postchemotherapy modified template should result in the same decrease in patient morbidity that was produced in the primary setting and should include greater preservation of antegrade emission and ejaculation with avoidance of the contralateral efferent sympathetic fibers; decreased operative time; potentially reduced perioperative ileus, negating the need for nasogastric tubes; and shorter hospital stay. These factors are significant for the patient, and attempts should be made to continue to reduce morbidity while maintaining similar oncologic efficacy.

In select patients with residual masses after chemotherapy, a modified dissection appears to remove retroperitoneal tumor adequately. Patients with low-volume postchemotherapy tumors are candidates for modified PC-RPLND.