The driving forces for early observation studies in clinical stage I patients were the infertility resulting from RPLND (due to retrograde ejaculation) and the apparent absence of therapeutic benefit (i.e., orchiectomy was a curative procedure or systemic disease occurred in the absence of retroperitoneal disease). The ability of cisplatin-based chemotherapy to cure systemic disease directly permitted observation studies, because cure of low-burden disease was more than 95% and treatment of relapse would not compromise survival. Relapse occurs in 25% to 30% of patients who are observed. A higher likelihood of retroperitoneal or systemic relapse was associated with T2-4 tumors and lymphatic or vascular invasion in T1 tumors. Some studies suggested that a high percentage of embryonal carcinoma and other histologic features also predicted a higher likelihood of relapse. However, the correlation between lymphatic-vascular invasion and the presence of embryonal carcinoma is high, and general agreement on histologic criteria for relapse independent of vascular or lymphatic invasion does not exist. Therefore, vascular-lymphatic invasion is the critical pathologic predictor for relapse in tumors confined to the testis. The retroperitoneum is the site of relapse in approximately two thirds of patients; the lungs or markers alone, in approximately one third; and other visceral sites, much less frequently. There is a slightly higher likelihood that both chemotherapy and modified bilateral RPLND ( not nerve sparing) are needed in order to achieve the same cure rate.
Patients with clinical stage I nonseminomatous GCT with a T1 tumor without vascular-lymphatic invasion and serum tumor markers that are normal or declining at half-life should be offered both surgical and observation options. If RPLND is chosen, it should be of the nerve-sparing type, thereby preserving ejaculatory capacity in the majority of patients. Frequent CT scans of the abdomen are unnecessary once an RPLND has been performed. If surveillance is chosen, then a possibly unnecessary RPLND is avoided, limiting therapy to orchiectomy alone in at least 70% of the patients (i.e., those who never relapse). Patient compliance cannot be overemphasized. A physical examination, chest x-ray, and determinations of AFP and hCG levels are required at monthly intervals in the first year, every other month in the second year, quarterly in the third year, and less frequently thereafter. An abdominal CT scan is required quarterly in the first year, every 4 months in the second year, and every 6 months beginning in the third year. Visits and evaluations should be annual in the fifth year and beyond. In both situations, relapses are extremely uncommon after 2 years and have only very rarely been observed after 5 years, in contrast to seminoma.
There are few data regarding chemotherapy as initial treatment of clinical stage I disease when the risk of retroperitoneal disease is high. In three reports of patients receiving two cycles of cisplatin-based chemotherapy, fewer than 5% relapsed and about 1% died of GCT. Although this approach avoids RPLND and the duration of therapy is brief, a majority of these patients would be exposed to the transient (e.g., myelosuppression), permanent (e.g., neuropathy), and delayed (e.g., Raynaud’s phenomenon, acute leukemia) toxicities of chemotherapy. The data are not yet mature and follow-up is short; this approach should be considered investigational.
Rarely, patients with clinical stage I disease are found to have persistently elevated serum concentrations of AFP or hCG after orchiectomy. If these markers increase or plateau at an elevated level after a period of observation (4 weeks or less), metastatic disease is present. This group of patients should receive initial systemic chemotherapy, because the disease is often not limited to the retroperitoneum. An RPLND should be done only if clinical studies at the conclusion of therapy demonstrate new disease.