Men who present with clinical stage I nonseminomatous germ cell tumors (NSGCTs) have excellent cure rates approaching 99% to 100%. This excellent cure rate is regardless of the initial management option chosen, which includes active surveillance, 1 cycle of bleomycin, etoposide and cisplatin (BEP), or primary retroperitoneal lymph node dissection (RPLND). Retroperitoneal lymph node dissection has been a foundation in the treatment of NSGCT patients for numerous decades. However, its utilization has steadily trended downward as surveillance has become more recognized as an acceptable and safe alternative. This trend in management was highlighted recently by Weiner et al in 4,080 men with clinical stage (CS) IA and 2,580 men with CS IB NSGCT diagnosed from 2004-2013 using the National Cancer Database. ¹ Over this 10-year period, the use of RPLND decreased from 22% in 2004–2005 to 14% in 2012–2013 (p <0.001). This was largely due to the increased utilization of surveillance during this time period. During the last 30 years at Indiana University, we have performed primary RPLND for clinical stage I disease for fewer and fewer patients (fig. 1). Likewise, while we performed 46 primary RPLNDs in 2019, only 7 of these were for clinical stage I disease.

Figure 1. Volume of primary RPLND procedures performed at Indiana University (IU) over 30 years.

The recent American Urological Association (AUA) guidelines for early stage testicular cancer advocate for patients with CS IA to preferably be managed with surveillance, and those with CS IB can be managed with any of the previously stated treatment options. ² This discrepancy in management recommendations by substage is due to the differences in relapse rates between CS IA and IB patients. It is well established that lymphovascular invasion in the orchiectomy specimen portends a higher risk of relapse. Additionally, cases with embryonal dominant tumors are also more likely to recur while on surveillance. Nayan et al recently demonstrated the conditional risk of relapse between patients with CS IA and IB with or without pure embryonal carcinoma in the orchiectomy specimen. ³ For example CS IA without pure embryonal carcinoma had only a 17% risk of relapse, while patients with CS IA with pure embryonal or CS IB had risks of relapse as high as 40% to 50%. However, it must be acknowledged that even though the risk of relapse on surveillance is higher with pure embryonal carcinoma and/or lymphovascular invasion, the ability to cure these patients is similar compared to those without these higher risk features. Thus, surveillance remains an option for patients with CS IB or pure embryonal carcinoma as well.

So, why the shift away from primary RPLND for patients with clinical stage I over the years? First, as mentioned previously, the safety of surveillance has become more widely recognized. Second, most historical reports of RPLND find only around 30% of patients have pathological stage II disease. Therefore, through this lens, there are 70% of patients going through an unnecessary operation. Next, like any surgery, there are procedure-related complications, which can and do occur. However, like most surgical procedures, the risk of complications when performed at high volume centers is lower. Baniel et al reported a 10% complication rate following 478 primary RPLNDs at our center in a historical series from the 1980s and 1990s. ⁴ These complications do not seem to be diminished even with a robotic approach, which recently showed complications of 32% in 1 of the larger series with only 58 patients. ⁵ Complications notwithstanding, our institution recently reported on the value of this surgery in pathological stage II patients, with over 80% cured with surgery alone, mitigating the need for chemotherapy and its long-term side effects. ⁶

The surgical volume of primary RPLND and its direct relation to outcomes warrant further discussion. Figure 2 emphasizes how few primary RPLND procedures are performed annually in the United States compared to other common cancer surgeries. Procedure totals for kidney, lung, prostate and colon resections represent only Medicare patients and thus underrepresent these annual totals. It is not surprising that high volume centers that perform this rare procedure might have improved surgical outcomes. Given these small numbers, there will not be many “high volume” centers to go around. The rare need for this surgery coupled with the high complexity of the procedure itself creates the perfect scenario to reduce its use when other acceptable, less morbid options exist, such as surveillance.

Figure 2. Representation of selected cancer surgeries in the United States annually by cancer type. Adapted from Li J et al: Association of delivery system integration and outcomes for major cancer surgery. Ann Surg Oncol 2018; 25: 856, and Woldu SL et al: Impact of hospital case volume on testicular cancer outcomes and practice patterns. Urol Oncol 2018; 36: 14.e7.

Finally, considering the low incidence of the disease, resulting in low surgical volumes for primary RPLND, as well as the fact that approximately 70% of patients with pathological stage I disease do not benefit from the surgery, who really needs this operation? The debate about optimal treatment for clinical stage I disease has been long and ongoing, and is tired. The problem has been a lack of good prognostic abilities, from either an imaging or a biomarker perspective. Thus, we have relied on pathological factors as mentioned previously, such as lymphovascular invasion or embryonal dominant tumors, which yield a predictive value similar to a coin flip. Some men choose primary RPLND in the hopes of limiting their need for future chemotherapy, as most with microscopic disease found during RPLND are cured with surgery alone. Lastly, some urologists might strongly suggest surgery if the patient seems unreliable to follow through with surveillance, such as someone who is underinsured or uninsured. Fortunately, this tide seems to be turning as a new biomarker continues to gain traction. mi-RNA 371a-3p is poised to help guide decision making, not only in clinical stage I disease, but also in other clinical scenarios of germ cell tumor management. ⁷ Additionally, there are now 2 open clinical trials collecting this biomarker (Children’s Oncology Group study AGCT 1531 and SWOG 1823), and others are or will be in development in the coming years. If the prospective clinical trial data support the growing body of literature on this biomarker, gone will be the days of debates about which of the 3 treatment options, surveillance, RPLND or chemotherapy, is “best.” Patients with miRNA negative results will be placed on surveillance or maybe even discharged from clinic. Patients with positive miRNA results can be managed with primary RPLND. However, the debate will still need to be settled for clinical stage I cases with miRNA positivity regarding the role of chemotherapy vs primary RPLND. However, this topic, we are sure, will fill the gap of future panel debates until forthcoming trial data become available.