Introduction

Historic papers confirm the lethal potential of metastatic testis cancer. There were improvements in prognosis in Ireland in the mid 80s with the introduction of chemotherapy and surgery but survival rates did not match the favourable prognosis now reported with modern staged multidisciplinary management. Considering the young population affected and the high potential for cure, there should be clear guidelines relating to management of metastatic testis cancer not least when considering the associated morbidity with oncological and surgical intervention. The role of post-chemotherapy RPLND in the management of metastatic testis cancer has become clearly defined in recent decades. The bulk of data on the topic comes from large series from major international specialist centres.

Testis cancer metastases follow a predictable path to the retroperitoneal lymph nodes [1]. The primary treatment for non-seminoma clinical Stage 1 and low volume retroperitoneal nodal disease remains controversial. The choice of active surveillance, primary chemotherapy or primary RPLND (laparoscopic or open) varies between countries and is influenced by tradition and available expertise and resources [2]. For sizeable or advanced metastatic disease (>Stage T2b) there is consensus that in seminoma, non-seminoma and combined tumours, initial treatment should involve platinum based chemotherapy [3]. While primary chemotherapy can achieve complete remission in low volume metastatic disease, partial remission is more common for more sizeable lesions. Optimal management of residual retroperitoneal mass lesions after primary chemotherapy in non-seminoma involves retroperitoneal lymph node dissection at appropriate interval. Such surgery involves complete extirpation of residual masses and the appropriate nodal drainage field. The rationale of post-chemotherapy RPLND is to achieve histopathological diagnosis of residual masses. The presence of active tumour, residual teratoma, or fibrosis/necrosis is of prognostic value. The early histopathological identification of residual cancer allows prompt intervention with salvage chemotherapy. Retroperitoneal lymph node dissection, by achieving complete remission, is therapeutic when pathology confirms teratoma. Such tumour is chemo-resistant but has potential to enlarge (“growing teratoma”) and present later as germ cell cancer [4]. Post chemotherapy RPLND also facilitates follow-up, in that relapse should not occur within the resection field. Any subsequent potential relapse in the thorax or pelvis is easier to detect with CT scanning than the retroperitoneum.

In America the majority of RPLND surgeries are performed by a small number of surgeons [5]. In the UK 18 centres have been identified to perform this complex surgery including one in Belfast [6]. Less well described is the role for this operation in Ireland, a country with a small population, and combined with the low incidence of testis cancer (7.3 per 100,000; 176 cases per year) [7], a small number of patients requiring RPLND.

In this study, we describe the experience of a single Irish centre with the performance of post-chemotherapy RPLND in the management of metastatic testis cancer. We aim to characterise the patient cohort with advanced testis cancer and their characteristics of disease in terms of histological sub-types, staging, sites of spread and response to chemotherapy.

Initial data from Ireland were reported by Casey et al. [8]. This follow-up study has increased duration of follow-up and narrowed inclusion criteria to obtain a more homogeneous patient cohort, by including only patients undergoing post-chemotherapy surgery.

Methodology

Data in this retrospective series were prospectively collected including follow-up data in line with the STROBE guideline for observational studies [9].

All patients undergoing RPLND for metastatic testis cancer of all histological sub-types between January 1996 and December 2011 in a single centre by a single surgeon were included. All patients must have undergone neo-adjuvant primary chemotherapy prior to RPLND. Data were collected by review of medical records, clinical correspondence, operative records, radiology and histology reports. Follow-up data were accrued from the initial referring oncology service. Where unavailable, the patient’s primary care practitioner was contacted. Finally, when required, individual patients were contacted personally. Follow-up data were available on 85 % (n = 66) of patients.

Staging was by the Royal Marsden staging system [10] with stage migration defined as change of pre-chemotherapy stage by at least one sub-classification following completion of chemotherapy (Table 1). Complications were defined as early (less than 30 days) and late (more than 30 days), with severity classified by the Clavien Dindo Classification of morbidities [11]. Recurrence of disease was defined as confirmed evidence of new metastases at a time remote from RPLND. Finally, sub-group analyses were performed to identify differences in outcomes relating to disease stage and template of resection.

Table 1 The royal marsden staging system for testis cancer
Full size table

Data were coded to ensure confidentiality. Statistical analysis was performed using Stata 12 Statistical Package (StatCorp, Texas, version 12.0). Mean comparison was performed using T test for parametric variables. Categorical variable comparison was performed using Chi Squared testing. All P values were reported as two tailed. Survival analysis was performed using Kaplan–Meier survival functions and expressed as 1-, 5- and 10-year survival rates. Data were then censored for recurrence, allowing calculation of time to recurrence and disease free survival. A hospital ethics approval was granted for this study.

Results

Seventy-eight patients were identified for inclusion. Mean age was 28.5 ± 7.5 years at time of diagnosis. Initial presenting complaints included testis mass (74.3 %), testis pain (38.4 %) and back pain (10.2 %). Less common presentations included neck or supra-clavicular mass, weight loss and haematuria.

Radical orchiectomy was performed as primary diagnostic and therapeutic management in 89.7 % of patients (n = 70). 12 alternative pathological diagnostic procedures other than orchiectomy included supra-clavicular or cervical lymph node biopsy, retroperitoneal mass biopsy and testis biopsy. In 10.3 % (n = 8) of cases orchiectomy was delayed until the time of post-chemotherapy RPLND. A single patient underwent bilateral orchiectomy for bilateral synchronous tumour. The pathology of primary tumour demonstrated non-seminomatous germ cell (NSGCT) (62.8 %, n = 49), seminoma (6.4 %, n = 5) and combined tumour (19.2 %, n = 15).

The remainder includes nine patients with presumed burned out primary testis tumour because histology demonstrated undifferentiated tissue, fibrotic tissue or calcification.

Disease stage at presentation by Royal Marsden staging system was Stage 1 (16.67 %, n = 13), Stage 2a (1.28 %, n = 1), Stage 2b (20.52 %, n = 16), Stage 2c (21.79 %, n = 17), Stage 3 (14.10 %, n = 11) and Stage 4 (25.64 %, n = 20). All 13 patients with Stage 1 disease at presentation progressed while on surveillance. Most common sites of extraretroperitoneal lymph node metastases included supra-clavicular or cervical (n = 8) and mediastinal (n = 3). Sites of visceral metastases included lungs (n = 16), liver (n = 3), brain (n = 1) and spine (n = 1). Primary chemotherapy in 53 patients (68 %) involved standard four cycle drug regime with Bleomycin, Epirubicin and Cisplatin (BEP). Twelve patients (15.4 %) were given Epirubicin and Cisplatin alone to prevent Bleomycin toxicity. Another 11.5 % of patients (n = 9) were unable to complete full course of BEP, with Bleomycin omitted during latter cycles. The remaining four patients (5.1 %) received alternative individualised chemotherapy regimes.

Following completion of chemotherapy, partial remission was achieved in 29.5 % of cases (n = 23), with stable disease in 55.1 % (n = 43) and disease progression in 15.4 % (n = 12). Clinical re-staging post-chemotherapy revealed Stage 1 (0 %), Stage 2a (23.1 %, n = 18), Stage 2b (30.77 %, n = 24), Stage 2c (16.67 %, n = 13), Stage 3 (7.69 %, n = 6) and Stage 4 (21.79 %, n = 17).

All patients underwent RPLND via an anterior trans-abdominal approach. The extent of resection template was dependent on the pattern of metastatic spread delineated on imaging, and gross appearance at exploration. The boundaries of resection followed those templates described by Donohue et al. [12].

Intra-operative complications (n = 4) included ureteric injury (n = 2) and significant bleeding (n = 2).

No peri-operative deaths occurred. Re-operation was required in five patients, three for early haemorrhage within 6 h, one for wound dehiscence at 1 week and one patient who required fasciotomy under local anaesthetic for lower limb compartment syndrome at 1 day post-op.

Twenty-three patients (29.5 %) underwent bilateral infra-hilar nodal template resection, 36 cases (46.2 %) underwent modified unilateral resection. Additional supra-hilar dissection was performed in 16 patients (20.5 %), while the template also included pelvic dissection in 3 patients (3.8 %). Disease volume and location mandated 34 further procedures be performed in 30 patients including; nephrectomy (n = 16), ureteric repair (n = 3), nephroureterectomy, aortic replacement, caval resection, ureterolysis, Boari flap reconstruction and lower lung lobectomy (n = 1 each). Another eight patients underwent orchiectomy at time of post-chemotherapy RPLND having had initial pathological diagnosis by other means. Complete surgical abdominal remission was achieved in all but one case in whom the extent of local invasion precluded a complete resection. In this case of seminoma the procedure was abandoned and the patient referred for further oncological measures. This patient is excluded from relapse calculations.

Median hospital stay was 10.5 days (range 5–50 days). Four patients had prolonged admission of 50, 47, 34 and 27 days. There were no peri-operative deaths. Early post-operative complications were observed in 29.5 % cases (n = 23). Using the Clavien Dindo classification of complication severity [11] these included grade 1 10.3 %, n = 8, grade 2 10.3 %, n = 8, grade 3a 1.3 %, n = 1, grade 3b 5.1 %, n = 4, grade 4 2.5 %, n = 2 and grade 5 0 %. Three patients underwent early abdominal re-exploration within 24 h for suspected haemorrhage. One case required control of a lumbar artery and another control of a mesenteric vein. The third case did not have haemorrhage confirmed. One patient developed partial wound dehiscence, requiring repair. One patient developed lower limb compartment syndrome requiring fasciotomy. This patient had undergone iliac artery resection because of densely adherent tumour necessitating an aorto-iliac graft during RPLND. One patient with chylous ascites and pleural effusions required prolonged ventilation for 10 days, a chest drain (removed early) and peritoneal drain (removed after 12 weeks) [13]. Minor complications included atelectasis, respiratory tract infection, minor wound infection and prolonged ileus.

Other early complications included DVT (n = 3), transient lower limb oedema (n = 3), transient lower limb paresthesia (n = 2) and pulmonary embolism (n = 1).

Owing to peripheral follow-up of the majority of patients, the incidences of late complications may not be comprehensive. One patient with recurrent adhesional bowel obstruction required five subsequent laparotomies elsewhere. Three patients were reported to develop small bowel obstruction elsewhere which settled with conservative measures. One patient underwent repair of a small incisional hernia elsewhere. One patient describes ongoing chronic back pain.

Histopathological analysis of RPLND specimens revealed; fibrosis/necrosis (n = 37, 47.4 %), mature teratoma (n = 32, 41.1 %) and active cancer (n = 9, 11.5 %).

Thirteen patients with Stage 1 disease at initial presentation progressed to metastatic disease while on surveillance. They are assumed to have harboured occult micrometastatic disease at the time of orchiectomy. In this subgroup, testis histology was NSGCT in 11 (84.6 %) and mixed in 2 (15.4 %). RPLND showed fibrosis in 5 (38.4 %), mature teratoma in 6 (46.2 %) and active disease in 2 (15.4 %). One patient had progression during primary chemotherapy, one patient had stable disease while all others had a partial response.

Follow-up data were available on 66 patients (85 %). Median duration was 100 months (range 11–207 months).

Disease relapse following RPLND was observed in nine patients (13.6 %). Testis histology was NSGCT in 6 (66.6 %), Seminoma in 2 (22.2 %) and burnt out primary in 1 (11.1 %). Disease stage following primary chemotherapy was Stage 2b n = 2 (22.2 %), Stage 2c n = 2 (22.2 %), Stage 3 n = 2 (22.2 %) and Stage 4 n = 3 (33.3 %). RPLND histology showed fibrosis in 4 (44.4 %), teratoma in 2 (22.2 %) and active disease in 3 (33.3 %). Median time to relapse was 15 months (range 8–60 months) (Fig. 1). Sites of relapse included retroperitoneal (outside field of resection) n = 2, supra-diaphragmatic nodal n = 1, pulmonary n = 2, disseminated n = 1, combined pulmonary/retrocrural n = 1, and combined pulmonary/supra-diaphragmatic nodal relapse n = 1 and tumour marker elevation alone n = 1. Median disease free survival was 95 months (range 8–204 months).

Fig. 1
figure 1

Kaplan–Meier for disease free survival

Full size image

Four patients died with median time to death 30 months (11–96 months) post-surgery.

Case 1 who died was aged 50 and had extended bilateral RPLND for non-seminoma clinical Stage 2a disease post-chemotherapy. Complete abdominal surgical remission was achieved but pathology confirmed active germ cell cancer. The patient did not tolerate salvage chemotherapy and died at 11 months due to advancing Stage 4 supra-diaphragmatic and visceral disease.

Case 2 mortality was aged 34 and underwent a partial resection/biopsy of bulky residual seminoma (Stage 2b) post primary chemotherapy. Pathology confirmed active seminoma. This patient died at 36 months despite salvage chemotherapy for advancing widespread disease.

Case 3 mortality was aged 48 and underwent extensive bilateral RPLND for clinical Stage 2c NSGCT post primary chemotherapy. Pathology confirmed active cancer. The patient received salvage chemotherapy but dose modified due to toxicity. He subsequently died at 96 months due to disseminated disease.

Fourth mortality case, age 23, underwent extensive bilateral RPLND for non-seminoma Stage 4 disease including bilateral supra-hilar thoracoabdominal dissection. Pathology confirmed teratoma without active germ cell tumour but the patient died at 24 months despite salvage chemotherapy when he developed advancing lung metastases.

Time specific Kaplan–Meier survival rates (including only patients with follow-up data at that time) were; 98.5 % at 1 year (n = 65), 95.2 % at 5 years (n = 49) and 92.6 % at 10 years (n = 25). The Kaplan–Meier Curve with 95 % confidence interval for the entire cohort demonstrates a plateau at 100 months with no delayed mortalities after 96 months post-operative (Fig. 2). Stratification of survival based on disease stage revealed for Stage 2 disease survival was 98 % at 1 year, 95.4 % at 5 years and 91.2 % at 10 years. For Stage 3 disease, survival was 100 %. For Stage 4 disease, survival was 100 % at 1 year, 91.6 % at 5 years and 91.6 % at 10 years. These variations do not reach statistical significance. There were also no statistically significant associations between complication rates (P = 0.43) or rate of disease relapse (P = 0.89) and resection template.

Fig. 2
figure 2

Kaplan–Meier for overall survival

Full size image

Discussion

The Irish National Cancer Registry indicate an incidence of testicular cancer in Ireland of 7.3 cases per 100,000 per year (176 new cases per year) and a lifetime risk of <0.1 % for all [7]. It constitutes 0.5 % of registered malignancies per year and less than 2 % of male cancers [14]. As expected, most patients were aged 20–39 years at diagnosis [15, 16]. This disease prevalence data therefore yields a small total number of patients who require RPLND post-chemotherapy. The bulk of patients who never require RPLND include those with Stage 1 NSGC testis cancer, those cases of seminoma (with few exceptions) and those who achieve complete remission after primary chemotherapy for metastatic NSGC testis cancer.

Our series confirms the complexity of RPLND in the post-chemotherapy setting when extensive desmoplastic reaction and tumour itself cause significant distortion of tissue planes. Residual nodal masses post-chemotherapy may invade local structures including ureter, aorta, and vena cava. Thus, a complex dissection may be required, involving additional procedures including nephrectomy, aortic replacement, caval resection or ureteric repair. Djaladat et al. in the University of Southern California demonstrate that 30 % of patients require adjuvant procedures at the time of post-chemotherapy RPLND [17]. In that series 15 % of adjuvant procedures were vascular such as aortic or caval resection/replacement while nephrectomy comprised 14 %. Our overall rate of additional procedures at RPLND (28 %) is in line with this series, although we report a higher rate of concomitant nephrectomy (20 versus 14 %) but lower incidence of vascular procedures (2.6 versus 15 %).

Despite the extensive nature of RPLND surgery, mortality remains low. Five major studies report peri-operative mortality rate less than 1 % [18]. The relative young age and lack of co-morbidities of patients are relevant. There were no peri-operative mortalities in the 16 years of this study, although we acknowledge our cohort is relatively small. Our early complication rate of 29.5 % appears high but when Clavien Dindo grade 1 and 2 minor complications are excluded, the rate drops to 8.9 %, which is lower than reported by Baniel et al. in their large post-chemotherapy cohort [19]. They described a higher complication rate with post-chemotherapy RPLND (18 %) than with primary RPLND (8 %), reflecting the complexity of surgery in the post-chemotherapy setting.

The incidence of viable malignant tissue, teratoma and fibrosis/necrosis reported by histopathology following RPLND is similar in our series to the ranges reported by Winter et al. [20]. Most recent studies have shown the percentage of patients with residual malignant disease decreasing to between 2 and 15 %. Approximately 35–60 % have residual mature teratoma and 35–50 % have fibrotic tissue alone. This beneficial trend in recent decades is largely due to improvements in chemotherapy regimes and their delivery. Thus, 51.2 % of patients in our series (with residual germ cell tumour or teratoma in resected specimens) achieved complete abdominal remission by surgery and as such derived a therapeutic benefit. In the absence of a reliable non-invasive procedure to ascertain the nature of residual masses after chemotherapy, RPLND remains the diagnostic method of choice. Algorithms for predicting residual malignant disease in post-chemotherapy masses using statistical models with combinations of variables report variable success [21–24]. The 75–83 % accuracy achieved in some models still renders false negative rates too high to safely avoid RPLND in this setting. Our relapse rate after RPLND surgery is comparable to that of Flechon et al. [25] at 18.5 % with median follow-up 6.1 months (range 1.3–39 months) post PC-RPLND. Regarding sites of relapse post RPLND, three patients did suffer relapse within the abdomen but these were outside the resection template. Another five of nine patients who relapsed developed disease in the supra-diaphragmatic region. There was no documented isolated pelvic recurrence in this series.

Our overall survival (92.6 % at median 100 months) compares with that in Flechon’s report (87 %, median follow-up 77 months and range 1.3–187 months) [25]. Similarly the 95 % 5-year survival rate in our series is reassuring because overall 5-year survival for all patients with testis cancer is reported at 97 % at 5 years [26]. Considering the relative advanced stage of disease of our patient cohort, these results are encouraging.

Our study suggests a higher survival for Stage 4 disease than less advanced disease at RPLND. This apparent contradiction did not reach statistical significance and could be due to selection bias where a proportion of Stage 4 patients continue with salvage chemotherapy without having RPLND. Also, risk factor analyses in metastatic testis cancer highlight the importance of disease bulk and elevated tumour marker status [27]. Prognosis is not governed by the presence of distal metastases alone.

Rates of testis cancer in the UK closely mirror those in Ireland (7.1 versus 7.3 per 100,000) [28]. Given the low number of patients in Ireland who will require this surgery, as well as its complexity and high rates of secondary procedures required, we advocate this procedure only be performed in one or two supra-regional centres in Ireland, an approach that has been the standard in the UK since 2002 [6].

Conclusion

We present the experience of a single Irish hospital in performing post-chemotherapy retro-peritoneal lymph node dissection for the management of testicular cancer over a 16-year period.

In terms of both peri-operative management and overall prognosis, the results of our series can stand alongside those from the major international urological centres.