In the second half of the past century, dramatic improvements were made in the management of advanced cancers of the genitourinary tract, with particular progress in the management of germ cell tumors. The management of metastatic testicular germ cell tumors has become one of the paradigms of successful treatment, reflecting advances in chemotherapy and an improved understanding of the principles of tumor biology and of the importance of multidisciplinary management. Nevertheless, with increasing experience, we have come to recognize that these approaches to treatment have some flaws, and that we must be careful if we are to maintain the high cure rates that have been achieved. An increased understanding of risk factors has allowed us to tailor our treatment to the level of risk. We have come to realize that some of our treatment strategies are associated with significant late effects, and we have attempted to avoid these by modifying some treatment approaches. However, such attempts to improve treatment outcomes have failed, largely because of a reduction in the initial cure rate. This serves to illustrate a critically important principle- when effective treatment is available, modifications must be introduced in a thoughtful and structured manner to ensure that there are no hidden costs associated with the innovations under consideration (Table 1). Pathobiology Most germ cell tumors arise from tissues derived from primordial cells that originate within the genital ridges and usually migrate in the midline to the testicles, and less frequently, to the retroperitoneum, mediastinum, and pineal region. These cells are capable of differentiating along two major histogenetic lines, forming seminomas and nonseminomatous germ cell tumors (NSGCT).[1,2] All of these tumors have a common germ cell origin and, in the testicles, arise from precursor cells described histologically as carcinoma in situ.[3] Less than 5% of testicular cancers are lymphomas and other non-germ cell tumors.[ 1,4] The ability to differentiate along different pathways is of particular importance with regard to late relapse (see below). Seminomas
The most common type of seminoma is the classical variant, composed of uniform, round, or polygonal cells with abundant cytoplasm and a centrally placed nucleolus.[1] Less common spermatocytic and anaplastic variants have been described, although it should be noted that the anaplastic variant sometimes represents a misdiagnosed NSGCT.[5] Seminomas are characterized by the presence of syncytiotrophoblastic giant cells in about 30% of cases, and these may produce a marker protein, human chorionic gonadotropin (HCG). We have identified a subgroup of seminomas, morphologically resembling a solid variant of yolk sac carcinoma that is associated with a worse prognosis when treated with conventional radiotherapy or standard chemotherapy regimens.[5] Studies are in progress to define whether the biochemical determinants of outcome identified in the molecular revolution may explain the differences in the natural history of these tumors- for example, aberrations in expression of c-kit, other oncogenes, or tumor-suppressor genes. Nonseminomatous Germ Cell Tumors
The group of NSGCTs includes several histologic subtypes: embryonal carcinoma, mature and immature teratoma, endodermal sinus tumor, and choriocarcinoma.[1] Frequently, NSGCT may consist of elements of undifferentiated cancer, trophoblastic tissue, and varying components of somatic differentiation, such as cartilage, glandular tissue, or hair.[1,2] Another tumor marker, alpha-fetoprotein (AFP) is classically produced by endodermal sinus tumor, although it is also associated with embryonal carcinoma. The presence of AFP in the circulation signifies the presence of nonseminomatous elements, even if pure seminoma has been diagnosed in the primary tumor. This constitutes an indication to manage the patient for a nonseminomatous germ cell malignancy. Common Features
Testicular seminomas and NSGCTs share many features in common[ 1,2]: (1) Both occur predominantly in males aged 18 to 35 years; (2) both usually follow an orderly pattern of spread, from the testis to the surrounding supportive tissues and/or up the spermatic cord, to regional and distant lymphatic channels, and sometimes to visceral sites via blood-borne metastasis; (3) both are characterized by the elaboration of tumor markers and by common etiologic associations, including a characteristic marker of the short arm of chromosome 12,[6,7] testicular maldescent, carcinoma in situ of the testis,[3] and a less clearly explained association with the syndrome of multiple atypical nevi[8]; (4) both histologic patterns have been associated with a susceptibility gene localized to chromosome Xq27[9]; and (5) both are highly sensitive to the effects of chemotherapy.[10,11] However, important differences exist between testicular seminomas and NSGCT, including a somewhat older age range for patients with seminoma, a slightly higher prevalence of second primary tumors among males with seminoma, different patterns of metastasis, and a substantial radiosensitivity in seminoma vs marked radioresistance in NSGCTs. Although the data are relatively preliminary, it appears that the c-kit ligand is expressed more heavily in seminomas than in NSGCTs,[12,13] and this leads to the potential for an increased chance of clinical improvement in response to modulators of these ligands, such as imatinib mesylate (Gleevec) or more specific developmental compounds, when used for seminoma as opposed to NSGCT. When presenting at extragonadal sites, the biology of these tumor types is less similar: The seminomatous tumors tend to retain sensitivity to treatment with chemotherapy or radiotherapy, and patients with these tumors are relatively easily cured. By contrast, the extragonadal NSGCTs are associated with much higher relapse rates and inferior cure rates.[14-16] The reasons for these differences are unknown, as the testicular and extragonadal tumors share virtually identical histologies and etiologic associations and similar patterns of spread. Presentation Primary testicular cancer commonly presents as a painless enlargement of the testis, often noted on self-examination.[17] There may be local pain,[18] which is often associated with hemorrhage within the tumor. Occasionally, a large tumor will drag on the spermatic cord, causing referred pain in the region of the flank, which, therefore, does not necessarily indicate the presence of retroperitoneal lymph node metastases. Unless the patient has previously undergone surgery for testicular maldescent or scrotal violation as part of the initial management of the primary tumor (a cardinal error), inguinal lymph nodes are usually not involved. Drainage of the testis is typically to the lymph nodes at the level of the ipsilateral renal hilum.[19] Symptoms
The symptoms of metastatic germ cell tumors are protean and depend on the sites of involvement.[20,21] Most commonly, the retroperitoneal lymph nodes are involved early, which can be associated with lumbar backache or central/lower abdominal pain. Occasionally, renal colic is due to obstruction of the ureters by advanced lymph node metastases. Pulmonary metastases may be asymptomatic or, if extensive, may be associated with dyspnea, cough, hemoptysis, or chest pain. Brain metastases, although uncommon, may manifest as headache, confusion, dementia, or focal neurologic syndromes, and occasionally may be detected in the routine staging workup.[ 22,23] Liver or bone involvement is now distinctly uncommon at presentation but may be associated with relapse. When these features are seen at presentation, choriocarcinoma should be considered as the most likely histologic subtype. It should be kept in mind, however, that seminoma will sometimes metastasize to bone. Germ cell tumors may be associated with nonspecific or constitutional symptoms, including weakness, sweats, fevers, malaise, and asthenia, especially in the case of advanced disease. Gynecomastia may indicate the presence of testicular cancer, especially in patients with a dominant element of choriocarcinoma either in a primary tumor, metastatic deposits, or both. Extragonadal germ cell tumors manifest symptoms similar to those described above, although they are more heavily influenced by the site of origin. For example, pineal germ cell tumors may be associated with headache, confusion, visual changes, stroke-like syndromes, and Parinaud's syndrome.[14] The presentation of retroperitoneal germ cell tumors is identical to that of testicular tumors with retroperitoneal metastases, with the exception of an obvious testicular primary. Mediastinal germ cell tumors are commonly associated with dyspnea, cough, chest, or back pain, and occasionally with superior vena cava syndrome. Mediastinal NSGCTs are commonly associated with metastases at presentation, and the symptoms reflect the sites of those metastases (Table 2).[24,25] Monitoring Treatment The efficacy of treatment is routinely monitored by clinical assessment, using physical examination and appropriate imaging studies. Plain radiographs and computerized axial tomographic (CT) scans are the standard modalities, with CT having almost completely replaced lymphangiography.[ 26] For brain metastases, magnetic resonance imaging (MRI) is superior to CT scanning, but the role of MRI in the assessment of other sites of systemic disease remains controversial. Although preliminary evidence supports the use of positronemission tomography (PET) in assessing the presence of viable cancer in residual deposits after chemotherapy for germ cell tumors,[27] this technology has not yet secured a defined place in management, given conflicting reports regarding specificity and sensitivity. Tumor Markers
The circulating tumor markers AFP and HCG should also be measured as part of the initial diagnostic work-up and when monitoring therapy. The first specimens should be obtained prior to primary surgery. AFP has a normal half-life in the circulation of 5 to 7 days, and HCG has a half-life of 24 to 36 hours.[28] Prolongation of a circulating marker's half-life after orchiectomy usually denotes occult metastatic disease, and indicates the need for further investigation and treatment.[28] In patients with metastatic disease, the absolute levels of tumor markers constitute independent prognostic determinants. During chemotherapy, there may be a transient release of markers from dying cancer cells, causing a transient elevation of blood levels before they begin to decline according to normal half-life gradients. Thus, serial measurements should be taken to determine whether the patient is responding adequately to treatment. If these repeated measurements are not taken during the period of chemotherapy, tumor marker release will not be identified and the half-life calculation will be incorrect, suggesting a prolonged tumor marker decline (incorrectly implying the presence of resistant disease). Treatment of Stage I/A Disease One of the most controversial issues in the management of germ cell tumors is the treatment of tumors without evidence of metastatic disease (stage I or A).[29] Traditionally, stage I seminomas were treated by inguinal orchiectomy and adjuvant radiotherapy (doses of around 35-40 Gy, later reduced to 25-30 Gy). With this approach, cure rates approached 90% to 95%, especially after the introduction of CT and gallium scanning as part of the diagnostic work-up. More recently, an approach involving close observation after inguinal orchiectomy, with serial measurement of tumor markers and repeated CT scans, has been implemented. The paradigm for this approach was initially developed for NSGCTs (see below), but recent studies have suggested that it may be safely applied to the management of seminomas.[ 30-32] It is important to note that the pattern of relapse of seminoma may be more insidious than that of NSGCT, and relapses with a later time of onset have been documented in many series. Thus, most active surveillance protocols for stage I seminoma require close follow-up for at least 5 years and include continued abdominal CT scans. Seminoma is also less commonly associated with tumor marker production, and thus, radiologic imaging assumes greater importance. Surgical Approach
For stage I/A NSGCT, the situation is also complex. Traditionally, such tumors have been managed by inguinal orchiectomy, followed by a radical retroperitoneal lymph node dissection. This approach has produced cure in more than 90% of cases, and has even afforded cure in some patients with histologic more extensive the involvement of the lymph nodes by cancer, the greater the likely need for chemotherapy. evidence of micrometastases, without the need for chemotherapy; ie, the more extensive the involvement of the lymph nodes by cancer, the greater the likely need for chemotherapy. The major disadvantage of this surgical approach is the risk of surgical complications, including pulmonary emboli, hemorrhage, and perioperative pain, and the occurrence of incompetent ejaculation secondary to transection of retroperitoneal nerves. The potential for recurrence is predicated on the T stage (or extent of tumor within the testes and surrounding tissues), with adverse prognostic factors including local extension to the spermatic cord and surrounding structures, undifferentiated histology, and lymphatic-vascular invasion.[33,34] In addition, failure of tumor markers to normalize with appropriate half-life times is associated with an increased risk of relapse[33] and constitutes an indication for chemotherapy. Active Surveillance
Studies from the Royal Marsden Hospital[11] and the Danish Testicular Cancer Group[35] have clearly shown that adjuvant radiotherapy does not prolong survival after orchiectomy for stage I/A NSGCT, and this modality is no longer a treatment option in this context. However, studies from the United Kingdom forged the development of the policy of active surveillance for stage I/A NSGCT.[36,31] This approach, predicated on the high cure rates among patients with good-risk metastatic disease, requires meticulous initial diagnostic work-up and the exclusion of patients with established lymph node or visceral, small-volume metastases. For patients without major risk factors and with normal imaging tests, the relapse rate after orchiectomy alone is less than 25%. Thus, appropriate surveillance allows 75% of patients to be spared the need for adjuvant therapy (lymphadenectomy or radiotherapy) after orchiectomy.[37] However, a higher percentage of patients will potentially relapse following surveillance than after retroperitoneal lymph node dissection, and will thus be candidates for chemotherapy.[29] The most important risk associated with this approach is lack of compliance by either: (1) a patient who defaults from regular follow-up without receiving the adjuvant therapy (eg, retroperitoneal lymph node dissection) that might have granted him a cure, or (2) a physician who is unaware of the necessary follow-up procedures, waits too long to assess the patient for the development of metastases, and thereby reduces the potential for cure. There is no optimal follow-up schedule.[38] However, successful approaches have all been predicated on meticulous initial diagnosis and work-up, and a structured approach to routine follow-up (with serial clinical assessment as well as serial assessment of tumor markers, chest radiographs, and CT scans). With time, less frequent CT scans are performed, but the need for frequent monitoring of tumor markers and chest x-rays continues. Table 3 illustrates schedules of active surveillance that have been associated with high cure rates.[32,38-40] The common feature of these schedules is meticulous and close follow-up for the first 2 to 3 years. In centers of excellence, where clinicians are experienced in the conduct of surveillance programs and a clear policy of patient selection and follow-up has been defined, this approach appears to be safe, producing cure rates of greater than 90% to 95%.[29,32,38,39] Some patients develop relapses that prove to be intractable to therapy,[32,38] and rarely, patients develop complications of chemotherapy, such as leukemia.[ 41] Nevertheless, cure rates among patients treated with adjuvant retroperitoneal lymph node dissections or radiotherapy protocols are also less than 100%, and thus, active surveillance remains a reasonable option. As noted previously, implementation of this approach by an inexperienced clinician without a defined and validated plan of surveillance and follow-up is potentially dangerous. High-Risk Patients
The appropriate management of patients with stage I/A disease and adverse risk factors is also controversial. Cullen et al suggested that adjuvant chemotherapy after orchiectomy is effective in preventing early relapse in patients with locally extensive tumors, involvement of the spermatic cord, and other high-risk factors.[42] The obvious problem with this approach is the introduction of early and late toxicity in patients who would not relapse in this context if they were simply observed. For example, we have shown that up to 50% of patients with high-risk stage I/A disease do not relapse if they are simply observed.[33,38] Therefore, for high-risk disease, in a highly monitored setting, close observation may be a reasonable approach in the hands of an experienced investigator. Another appropriate option is retroperitoneal lymph node dissection, which can identify lymph node involvement, and thus, patients who are likely to require chemotherapy. In this context, the problem is that some patients with spermatic cord involvement or vascular invasion in the primary tumor will relapse in the lungs or other visceral sites without evidence of retroperitoneal lymph node metastasis, and therefore are not helped by the procedure. Ideally, a patient with high-risk stage I testicular cancer should be referred to a center of excellence for a second opinion before a final plan of management is determined. Treatment of Small-Volume Lymph Node Metastases When evidence of small-volume (≤ 5 cm) lymph node involvement (stage IIA or IIB, stage B) is found in seminoma, patients usually receive radiotherapy to the ipsilateral lymph nodes in the pelvis, with extension to the para-aortic chain (including the involved nodes).[43] Although the trend recently has been toward dose reduction, most clinicians use a relatively standard radiation dose of 30 to 35 Gy to ensure local tumor control. There is clear evidence that prophylactic mediastinal lymph node irradiation does not improve outcome and may, in fact, compromise any subsequent chemotherapy. The preliminary experience with doses of 20 to 25 Gy suggests no increase in the pattern of early relapse, but longer follow-up will be necessary to determine whether such modifications are truly safe. In the case of NSGCT, optimal management of early, stage II disease is controversial, although there is general agreement that radiotherapy has no role. However, proponents of retroperitoneal lymph node dissection view the standard surgical approach as offering both diagnosis and definitive treatment and cite surgical cure rates of up to 50%, particularly in patients with only microscopic evidence of lymph node involvement.[44] With this approach, chemotherapy can be used to salvage most relapsing patients.[45] In many centers, adjuvant chemotherapy is administered routinely to patients with more than three to five involved nodes.