Retroperitoneal lymph node dissection for non-seminomatous germ cell tumour

Abstract

Background: The Swedish and Norwegian Testicular Cancer Group (SWENOTECA) is a multidisciplinary collaboration with an aim to improve the testicular cancer care for patients in Sweden and Norway by publishing guidelines, include patients in the SWENOTECA register and promote research. In 2007, SWENOTECA initiated RETROP, a multicenter study on nonseminomatous germ cell tumour (NSGCT) patients that are operated with post-chemotherapy retroperitoneal lymph node dissection (PC-RPLND). PC-RPLND, a cornerstone in the therapy for metastatic NSGCT, is a challenging procedure with a high risk of post-operative complications. However, in up to 50% of the patients that undergo PC-RPLND, the histopathology result is benign. No viable cancer or teratoma was found and the surgery performed was unnecessary. Paper I, II and III focus on strategies to reduce the negative effects of PC-RPLND using patients included in RETROP. Patients with NSGCT and lymph nodes on CT-scan sized < 20 mm and normalized tumour markers following orchiectomy is defined as CS IIA Mk (-). The survival rate for this patient group is high, and the recommended therapy is either initial surveillance, chemotherapy or primary Retroperitoneal Lymph Node Dissection (RPLND). The evaluation of a surveillance strategy is the focus in Paper IV.

Aims: Paper I) To compare the intra- and post-operative complications between a unilateral and a bilateral PC-RPLND; Paper II) To describe the distribution of retroperitoneal metastases on pre-operative CT scans for patients with a residual tumour of 10 – 49 mm following chemotherapy and to correlate it to the histopathological finding for the different areas resected. To assess the rate and location of retroperitoneal recurrences; Paper III) To validate a statistical model that predicts benign histopathology at PC-RPLND; Paper IV) To evaluate a surveillance plan with the aim of reducing overtreatment for patients with NSGCT in CS IIA Mk (-). To assess the RPLND histopathology results and describe the distribution of the metastatic retroperitoneal lymph nodes.

Methods: In paper I, patients in the RETROP database that underwent PC-RPLND during 2007–2014, were included (n = 318). A bilateral PC-RPLND was compared to a unilateral PCRPLND for intra- and post-operative complications and retrograde ejaculation using X2-test. In paper II, patients in the RETROP database with a residual tumour of 10-49 mm and normalized or low and declining tumour markers following chemotherapy were included (n =215). Patients were classified depending on the affected testicle (right- or left-sided NSGCT) and the lymph node spread to retroperitoneal lymph nodes on CT-scans (unilateral or bilateral spread) with reference to the aorta. The pathologic lymph nodes detected on CT-scans preand post-chemotherapy were correlated to the findings in the resected lymph nodes. Paper III includes patients in the RETROP database with normalized tumour markers following chemotherapy (n = 284). Variables used to predict outcome were: levels of alpha fetoprotein (AFP), ß-human chorionic gonadotropin (ß-HCG) and lactate dehydrogenase (LDH) before receiving chemotherapy, teratoma component in primary testicular tumour, change in lymph node size pre- and post-chemotherapy and maximal size of residual tumour. Discrimination and calibration analyses including Hosmer-Lameshow test were used to validate the prediction model and clinical utility expressed as net benefit was calculated. Paper IV CS IIA Mk (-) patients diagnosed 2008-2019 were identified in the Swedish part of the SWENOTECA database. The surveillance plan consisted of repeated CT-scans and tumour markers every 6th week up to 18 weeks before therapy decision. Information obtained from the SWENOTECA register and chart review were: Royal Marsden clinical stage, CTscan results, presence of lymphovascular invasion (LVI) and teratoma in the orchiectomy histopathology, number and type of chemotherapy courses, areas dissected at RPLND, histopathology for each area dissected, lymph node yield and complications, operating time, hemorrhage and length of stay following RPLND, recurrence and cause of death. Follow-up was continued until August 2021.

Results: Paper I. 318 patients underwent PC-RPLND (73% unilateral and 27% bilateral PC-RPLND). Patients that were operated with a bilateral PC-RPLND had a larger metastatic burden and received more chemotherapy courses than patients operated with a unilateral PC-RPLND. Bilateral PC-RPLND compared to unilateral PC-RPLND had more intra-operative complications (14% vs. 4%), post-operative complications (45% vs. 25%) and a higher incidence of retrograde ejaculation (59% vs. 32%). Paper II. For 65% of the patients, the retroperitoneal metastases were located unilateral to the aorta on CT-scans. For these patients, the risk of finding a retroperitoneal teratoma or viable cancer contralateral to the aorta were 2% and 3% for a right-sided and a left-sided testicular NSGCT respectively. Furthermore, the risk of recurrence for that group were 0% and 4% for right- and left-sided testicular NSGCT respectively. Paper III. Validation was completed with good reproducibility (AUC of 0.82 [95% CI 0.77 –0.87]) and calibration [Hosmer-Lemeshow test (p = 0.37)]. Using a decision threshold of 70%, the net benefit was 0.098 and 21% patients could have been spared surgery. Surgery would have been correctly avoided in 39% of all patients with a benign histopathology. On the other hand, teratoma or cancer would have been left untreated in 6% out of all patients with teratoma or cancer. Paper IV includes 57 CS IIA Mk (-) patients. During the observational period, 12% were downgraded to CS I and 4% progressed to CS IVA. The remaining patients received either chemotherapy (47%) or primary RPLND (37%). The RPLND histopathology results revealed that 24% had benign lymph nodes, 5% had teratoma, and 71% of the patients had viable cancer. Only one patient with unilateral cancer spread on CT-scan had a contralateral lymph node containing viable cancer.

Conclusions: Paper I. A unilateral PC-RPLND is more favourable than a bilateral PC-RPLND concerning the intra- and post-operative complications. However, patients selected for a bilateral PCRPLND had a larger metastatic burden and received more chemotherapy courses. Paper II. It was concluded that patients with unilateral retroperitoneal metastases on CTscans are the most suitable to select for a unilateral PC-RPLND. These patients have a low risk of contralateral metastases and retroperitoneal recurrences. Paper III. The prediction model was validated with good reproducibility and calibration. If the model is used in a clinical setting, the patients that are spared surgery need thorough follow-up to find recurrences at early stages. Paper IV. This study supports an initial surveillance period for NSGCT patients in CS IIA Mk (-), to find patients in CSI and avoid overtreatment. A primary unilateral RPLND is recommended in case of unilateral spread on pre-operative CT-scan.