Fax: (916) 743-8094
Follow-up surveillance strategies for genitourinary malignancies
Version of Record online: 23 MAY 2002
Copyright © 2002 American Cancer Society
Volume 94, Issue 11, pages 2892–2905, 1 June 2002
How to Cite
Evans, C. P. (2002), Follow-up surveillance strategies for genitourinary malignancies. Cancer, 94: 2892–2905. doi: 10.1002/cncr.10525
- Issue online: 23 MAY 2002
- Version of Record online: 23 MAY 2002
- Manuscript Accepted: 31 DEC 2001
- Manuscript Revised: 30 NOV 2001
- Manuscript Received: 13 AUG 2001
- urologic cancer follow-up;
- prostate cancer;
- bladder cancer;
- renal cell cancer;
- testicular cancer;
- cancer surveillance
Genitourinary cancers account for more than 20% of all malignancies in the United States. These cancers do not usually yield rapid mortality, thereby necessitating longer-term surveillance strategies.
A review and analysis of relevant studies were performed. Follow-up strategies are proposed to reflect effective methods to detect recurrent prostate, bladder, renal, and testicular cancers. Cost analysis was performed using Medicare reimbursement rates.
For genitourinary tumors, follow-up tests can be planned rationally based on detection rates and patterns. Tumor grade and stage drive follow-up strategies, along with therapeutic implications of detecting a recurrence. Symptomatic recurrences often obviate the need for radiographic tests and can minimize costs. Stage- specific plans for these four urologic malignancies are outlined specifically.
Not all surveillance approaches have been critically tested for follow-up of genitourinary tumors, but ample data are available to propose sound medical and economic strategies. Cancer 2002;94:2892–905. © 2002 American Cancer Society.
The posttherapy follow-up and management of patients with genitourinary malignancies is shared by urologists, medical oncologists, and primary care physicians. Genitourinary cancers accounted for approximately 22% of the 1.22 million new cancer cases in the United States in 2000. 1 For men in the United States, genitourinary cancers account for 40% of all new cancer diagnoses and result in 16% of all cancer deaths. However, between 1989 and 1995, the 5-year relative survival rates for all stages of prostate, bladder, kidney, and testicular cancers diagnosed between those years were 92%, 81%, 60%, and 95%, respectively.2 These statistics indicate that genitourinary malignancies occur frequently but are not immediately lethal. As a result, these patients can require long periods of follow-up. Strategies for the detection of genitourinary recurrences must be sound in the selection of tests and the frequency of their administration.
This article develops follow-up strategies for the four most common genitourinary malignancies: prostate adenocarcinoma, transitional cell carcinoma of the bladder, renal cell carcinoma, and testicular cancer. Factors directing the follow-up approach should take into account the likelihood of tumor recurrence, which primarily depends on cancer grade and stage. This is uniformly so for these four cancers. The pattern of recurrence is also important with regard to determining when and where in the body a tumor may reappear. Although it is important to know the likelihood and pattern of recurrence, the ability to treat it must also be taken into account. For example, one must address whether early detection or localization of the cancer has any impact on the outcome. The strategies developed in this article incorporate not only detection patterns, but related treatment implications as well.
MATERIALS AND METHODS
A literature search of the past 15 years was performed using MEDLINE, HealthStar, and PubMed. In addition, the journals Urology, Cancer, Journal of Urology, European Urology, World Journal of Urology, Journal of the National Cancer Institute, New England Journal of Medicine, and Cancer Detection and Prevention were targeted for specific information related to this review. Cost analysis was performed using clinic, laboratory, and imaging Medicare reimbursement fees (not charges) at the University of California, Davis, Medical Center. The American Society of Clinical Oncology (ASCO) has developed evidence-based clinical practice guidelines for some areas of oncology such as breast cancer surveillance, but no guidelines have been established for genitourinary tumors. 3 This review analysis is derived using studies of evidence levels 1–4 as defined by ASCO. All TNM staging is the 1997 version of the American Joint Committee on Cancer staging classification unless specifically stated otherwise.4
Oncologic follow-up for all genitourinary malignancies should include a generalized review of symptoms and physical examination. General constitutional symptoms should be queried, including loss of appetite, cachexia, lethargy, night sweats, and bone pain. In addition, site-specific symptoms and signs such as localized pain, palpable masses, and hematuria should be explored. Physical examination should correlate with this. Tumor grade, stage, and other factors affecting the likelihood of cancer recurrence should drive follow-up strategies.
Posttreatment surveillance strategies should be linked closely to the likelihood for tumor recurrence. Catalona and Smith 5 reported a large series of almost 1800 patients treated with radical prostatectomy. They found a 19% recurrence rate with a mean time to recurrence of 30 months. Of the 339 patients who had recurrence, 260 were detected by an increase in prostate-specific antigen (PSA) level only, 73 by an increase in PSA level plus local or distant recurrence, and 6 patients by local or distant recurrence without PSA relapse. Two of these six patients had recurrence in the pre-PSA era. Nonprogression probability 7 years after radical prostatectomy based on preoperative PSA level showed a strong correlation with preoperative PSA level. Nonprogression probability was greater than 88% for a PSA level less than 4.0, decreasing to 76% between 4.1 and 9.9 and to 49% with a preoperative PSA level greater than 10. Pathologic stage also correlated, with 81% of T1 N0 M0 (Prostate Cancer Stage B: tumor identified by needle biopsy due to elevated PSA level [T1], no regional lymph node metastasis [N0], no distant metastasis [M0]) and T2 N0 M0 tumors (Prostate Cancer Stage B: palpable but confined tumor involving one or both lobes [T2], no regional lymph node metastasis [N0], no distant metastasis [M0]) not progressing at 8 years compared with 57% of T3 N0 M0 margin-positive patients (Prostate Cancer Stage C: tumor extending through the prostatic capsule or into seminal vesicles [T3], no regional lymph node metastasis [N0], no distant metastasis [M0]), 26% of T3 N0 M0 margin-negative patients, and 19% of lymph node-positive patients (classification N1). Tumor grade also correlated with nonprogression probability 7 years after radical prostatectomy, with 84% of well differentiated tumors not progressing compared with 68% of moderately differentiated and 48% of poorly differentiated cancers. Based on this, a multivariate Cox proportional hazards model predicts that patients at highest risk for recurrence have a preoperative PSA level greater than 10, poorly differentiated tumors, and pathologic stage more advanced than pT3 N0 M0. Follow-up strategies should be based on these risks for recurrence.
A survey has evaluated the strategies of U.S. urologists for prostate cancer follow-up based on tumor stage. 6 In Johnson et al.,6 the most commonly used surveillance strategies were office visits with digital rectal examination (DRE), serum PSA levels, and urinalysis. Other tests routinely used were complete blood count (CBC), multichannel blood chemistry measurements, and chest X-rays. Computed tomography (CT), magnetic resonance imaging (MRI) scanning, bone scans, and monoclonal antibody (MoAb) scans were less commonly used. Surprisingly, 55% of urologists who responded to the survey did not modify their follow-up protocols according to patients' TNM stages. In a survey of the members of the American Urological Association, the members reported that the most common universal strategies for follow-up of patients after radical prostatectomy were office visits with DRE, PSA, and urinalysis. The PSA levels were obtained most commonly every 3 months for the first year, every 6 months in Years 2–5, and annually thereafter.7 The same database was used to determine whether the age of the urologist influenced follow-up strategy practices.8 Older urologists used prostatic acid phosphatase and CT scans more often than younger urologists, who more frequently used bone scans.
Strohmaier et al. 9 evaluated follow-up parameters in 80 patients after radical prostatectomy for localized tumors in 44 patients, with the remainder having metastasis to lymph nodes, distant sites, or both. Every 3 months, the patients had their levels of serum PSA, prostatic acid phosphatase (PAP), and alkaline phosphatase measured. They were also given a DRE and ultrasound of the bladder and kidneys. Every 6 months, patients had a bone scan and a CT scan of the abdomen and pelvis. The results showed that no recurrences were detected by PAP, alkaline phosphatase, CT, or bone scan that were not also detected by elevated serum PSA levels. Ultrasound detected hydronephrosis in six patients, all a result of treatment, not disease progression.
Following radical prostatectomy, a DRE should demonstrate an empty prostatic fossa. Three studies have assessed the utility of DRE compared with PSA in detecting recurrences after radical prostatectomy. 10–12 All patients with local recurrences palpable on DRE and all patients also had elevated PSA levels. None of the studies found utility for cancer detection using DRE when PSA was undetectable. Following radical prostatectomy, a DRE cannot be advocated for a patient on an office visit whose PSA level is undetectable. Following radiation therapy, DRE has been considered to be a more useful adjunct to PSA monitoring, as PSA does not reach the undetectable level. However, Johnstone et al.13 reported a nonrandomized study of 235 patients post–radiation therapy. The majority of abnormal rectal examinations resulted from bleeding or hemorrhoids. Forty-nine cases had tumor recurrences, all initially diagnosed by an increasing PSA level. Eight patients with prostate nodules on DRE had their cancer detected by an elevated PSA level and no biopsies were performed. For patients who received salvage radiotherapy for a recurrence after radical prostatectomy, a multivariant analysis suggests that a DRE can indicate patients who will have a positive biopsy.14 However, only preradiation PSA and postrecurrence PSA doubling time before radiation were predictive of outcome. These data suggest that DRE may not be as useful as once believed for follow-up after radiation therapy, but this single report should be validated with other studies prior to discontinuing this practice.
An increasing PSA level following radical prostatectomy can represent either a local or distant recurrence. Recurrences more than a year after surgery with a slow rate of increase (doubling time > 10 months) would be more representative of a local recurrence compared with more rapid PSA elevations occurring in the first year after surgery. 10 The role of bone scans in detecting distant metastases has been evaluated by Cher et al.15 They found the probability of a positive bone scan to be less than 5% until the PSA level increases to more than 40 ng/mL. It is unlikely that many U.S. urologists wait until this level is reached. The caveat is that all patients with an increasing PSA value following radical prostatectomy and new onset of bone pain should have a bone scan performed regardless of PSA level. In addition, posttherapy recurrences are often a higher grade, which can express less PSA compared with well or moderately differentiated tumors.
Assessment of postprostatectomy PSA recurrences by reverse transcription- polymerase chain reaction (RT-PCR) was investigated in 101 patients, with nine recurrences at 22 months. 16 Of 50 patients tested, 11 were RT-PCR positive for PSA preoperatively, increasing to 24 of 50 patients (48%) 12 weeks postoperatively. However, only 4 of 40 (10%) patients in this group were positive at 1 year postoperatively. Of 65% evaluated at 1 year, 23% were positive preoperatively and only 9% were positive at 1 year. Statistically, there is no correlation between positive RT-PCR and biochemical PSA failure. The ultrasensitive PSA assay has also been evaluated in 442 patients after radical prostatectomy.17 Although a standard serum PSA assay is defined as relapse occurring when the PSA level is higher than 0.1 ng/mL, the ultrasensitive PSA assay is a fourfold lyoconcentrated assay that defines relapse as levels lower than 0.025 ng/mL. At a mean follow-up of 449 days, 88 patients (20.8%) relapsed. Of these, 31% were positive on the ultrasensitive PSA test and later on the standard examination. Forty-two percent of patients were positive simultaneously on both assays and 26% were positive only on the ultrasensitive assay and negative on the standard PSA assay. The relapse detection rate within the first year by standard PSA assay was 25% compared with 85.7% using the ultrasensitive assay.17 Although these data are encouraging that the ultrasensitive PSA assay can detect recurrences earlier, the clinical implication of this remains to be determined. For example, The American Society for Therapeutic Radiation Oncology recommends that radiation therapy to the prostate bed for a local recurrence is most effective when the PSA level is still lower than 1.5 ng/mL. However, whether the ultrasensitive PSA test impacts this detection and has therapeutic benefit is unknown.
The MoAb scan ProstaScint uses I111-tagged prostate-specific membrane antigen to detect recurrences. The radiolabeled antibody is circulated in the bloodstream before scanning. It is 89% sensitive and 67% specific. 18, 19 One hundred percent of fossa biopsies that were positive for recurrence also had positive ProstaScint scans. The ProstaScint scan has the potential utility to identify fossa, lymph node, and distant recurrences in very select patients. However, the false-negative and false-positive rates affect its usefulness. One must also consider whether the specific location of the recurrence affects the choice of treatment. Another imaging modality, positron emission tomography (PET), has been unable to differentiate benign prostatic hyperplasia (BPH), prostate cancer, and scar.20
The location of the recurrence may have an impact on the options of observation, radiation therapy, androgen ablation, or combined therapy. Data suggest that earlier detection of a recurrence after prostatectomy is relevant to the outcome of salvage therapy. In treating local recurrences after a prostatectomy, a preradiotherapy PSA level greater than 1 ng/mL was the most significant predictor of biochemical failure after therapeutic radiotherapy. 21 Another trial reported that salvage preradiation PSA level influenced outcome only for patients with Gleason grade tumors less than 3.22 The actual therapeutic benefit of salvage radiotherapy for 89 patients treated for local recurrences after prostatectomy was reported by Anscher et al.23 At a median follow-up of 48 months, 50% demonstrated disease-free survival. Whether radiation should be given as adjuvant therapy to a positive surgical margin or in salvage setting of a local recurrence is still being debated.24, 25 For lymph node-positive patients after pelvic lymphadenectomy and radical prostatectomy or radiation therapy, the addition of androgen ablation was beneficial.26, 27 Furthermore, treatment of a rising PSA level with androgen deprivation for patients with classification M0 disease is reported to be of benefit.28 The best PSA testing interval for detection of these local, regional, and distant recurrences is unknown, but 3–6 months is generally practiced. The rationale for this is to detect a recurrence while the tumor burden is low and potentially amenable to aggressive salvage therapy.
Based on the information presented, follow-up recommendations after radical prostatectomy are shown in Figure 1. Essentially, PSA levels should be checked after prostatectomy between Weeks 6 and 12. Following that, the frequency of checks should be based on tumor stage and grade. DRE should be employed after radical prostatectomy with an increasing PSA level and perhaps after radiation therapy, although one retrospective study suggests that rectal examination after radiation therapy does not add to the detection of cancer recurrence. A bone scan should be used for patients with a PSA level greater than 35 ng/mL or for those with bone pain. These recommendations are based on recent data and therefore differ from those proposed by Montie. 29 In his comprehensive review, follow-up intervals were not stratified by tumor stage and bone scans were recommended for patients with PSA level gretaer than 8 ng/mL. Although no studies have evaluated hematuria as a sign of local prostate cancer recurrence, urinalysis is unlikely to be useful and is certainly less sensitive and specific than an increasing PSA level. However, urinalysis is useful for follow-up of patients treated with radiation therapy as there is a delayed, increased risk for developing bladder neoplasms. Therefore, unless patients have symptoms of infection, a routine urinalysis is likely to be of low utility as a routine follow-up test for patients after radical prostatectomy.
The risk for recurrence in patients with transitional cell carcinoma (TCC) of the bladder depends greatly on stage and grade. Patients with all stages and high-grade TCC of the bladder are at risk for recurrence. Patients with a superficial stage Ta N0 M0 (Bladder Cancer Stage 0: noninvasive papillary carcinoma [Ta], no regional lymph node metastasis [N0], no distant metastasis [M0]) tumor recurrence at their first 3-month post-transurethral resection bladder tumor (TURBT) cystoscopy as well as those with tumors larger than 5 cm and those with multifocal tumors are at risk for recurrence. Up to 70% of Stage Ta N0 M0 TCCs of the bladder recur, but 80% of these recurrences are superficial. A post-TURBT–positive cytology also suggests residual tumor that would likely recur. Regarding tumor progression, risk factors include Stage T1 N0 M0 (Bladder Cancer Stage A: tumor invades subepithelial connective tissue [T1], no regional lymph node metastasis [N0], no distant metastasis [M0]) cancers, Grade 3 tumors, carcinoma in situ (CIS), vascular or lymphatic tumor invasion, and p53+ and RB- status. The standard surveillance protocol followed by most urologists includes cystoscopy every 3 months for 2 years, then every 6 months for 2 years, and then annually. Cytologies are used for high-grade lesions or CIS.
Only three studies were identified that assessed when cystoscopies can be discontinued in patients who have no recurrences of their bladder cancer. Cookson and colleagues at Memorial Sloan-Kettering Cancer Center (MSKCC) reported 15-year outcomes for 86 patients with high-risk bladder cancer, of whom 81% had CIS and 44% had T1 N0 M0 lesions. 30 Endpoints were pathologically confirmed stage progression and disease-specific survival. Thirty-one percent of patients progressed at 5 years, but 5% of patients progressed after 10 years. A Swedish study reviewed 176 patients at lower risk for recurrence, predominantly Ta N0 M0 and some T1 N0 M0 tumors.31 Eighty percent had a recurrence after 1 year. Patients with recurrences 4 years after their first bladder tumor or patients with more than 10 recurrences total required life-time cystoscopy. Solitary low-grade tumors will very rarely recur beyond 5–10 years after the last bladder tumor resection. Fujii et al.32 reported on 100 patients who were followed for 4.1–20 years (median 9.7). These patients were at risk for late recurrence and progression of their Ta-T1 N0 M0 tumors if one of the following variables existed: intravesical chemotherapy, three or more tumors, bladder neck involvement, or positive urine cytology. These studies suggest that patients with low-grade and superficial tumors may not need cystoscopies 5–10 years after their last tumor, but patients with high-risk tumors (T1 classified lesions and CIS) should have life-time cystoscopies.
Two of these studies also assessed the need for long-term upper tract imaging to rule out TCC of the upper tracts. 30, 31 In the MSKCC study with 86 high-risk patients, 21% had upper tract tumors at a median of 7.3 years from their initial diagnosis. Six patients had recurrence in the first 5 years, seven patients between 5 and 10 years, and five patients between 10 and 15 years. Certainly these high-risk patients should have upper tract imaging for life at some interval. In the Swedish study of lower-risk patients with Ta N0 M0 and T1 N0 M0 lesions, there was one upper tract recurrence. As a result, high-risk patients should have an intravenous pyelogram (IVP) annually for life and low-risk patients should have upper tract imaging for 5 years. Currently, helical CT is being evaluated as a replacement for IVP. However, outcomes are unknown and the costs vary at different institutions. Renal ultrasound is employed by some physicians to evaluate the upper tracts. Although the distal ureter is not visualized with ultrasound, it can be argued that there are few primary and solitary distal ureteral lesions. If these lesions occur and grow, hydronephrosis will be present and detectable by ultrasound. However, curable ablation or resection at that point is unknown and monitoring with ultrasound remains unproven.
Tests less invasive than cystoscopy have been sought for monitoring the recurrence of bladder cancer. Many articles report on experimental markers, which are not clinically tested or approved. Standard existing tests were reviewed and large representative studies or reviews are cited. Cells for cytology can be collected from either a voided specimen or a bladder washing. Although cytology has 99% specificity, its sensitivity ranges widely, from 20% to 90%. 33, 34 It is quite sensitive for CIS and high-grade tumors but lacks sensitivity for low-grade papillary tumors. Newer tests, such as the bladder tumor antigen (BTA) and nuclear matrix protein-22 (NMP-22), have improved upon cytology.34–36 The BTA assay measures basement membrane protein antigen and has a 52–71% specificity for detecting bladder cancer and a 26–85% sensitivity. Newer renditions of the BTA test measuring human complement factor H-related protein demonstrate slight improvements in sensitivity and specificity. The NMP-22 test measures urinary nuclear mitotic apparatus protein and has a similar sensitivity (70.5%). However, it is not highly specific for bladder cancer and appears most useful in detecting disease recurrence. When compared with cytology, the BTA and NMP-22 tests have greater sensitivity than voided cytology for low-grade bladder cancer, but not for high-grade tumors.34 Cytology demonstrates superior specificity to either bladder marker test. There is no ideal test to replace cystoscopy to rule out recurrences. A combination of cytology and cystoscopy is useful in identifying high-grade tumors and ruling out CIS.
Based on the available literature, follow-up for superficial bladder cancer should include cystoscopy for low-grade tumors. Cystoscopy should be performed every 3 months for 2 years, then every 6 months for 2 years, and then annually. Other frequency schemas have been suggested as described in the Discussion. Cytology in conjunction with cystoscopy should be performed for high-grade tumors or CIS. The use of a bladder marker alone at this point is unconventional.
The goal for careful follow-up to detect recurrences in nonmuscle invasive bladder cancer is to select patients with progression, before the onset of metastasis. Herr and Sogani 37 reported 15–20-year follow-up for 307 patients at risk for progression as defined by requiring one or more courses of bacillus Calmette–Guerin therapy (BCG). Patients were followed with cystoscopy and cytology at 3 or 6-month intervals. Of 90 patients treated with cystectomy for BCG failure, patients survived significantly longer if they underwent cystectomy within 2 years rather than after 2 years. This suggests that monitoring for recurrences impacts outcomes following subsequent treatment if detected and treated early.
The likelihood of cancer recurrence following radical cystectomy for invasive TCC is stage dependent. Slaton and colleagues at University of Texas, MD Anderson Cancer Center evaluated 382 patients who underwent cystectomy. 38 The results were based on pathologic stage. Of the 382 patients, 97 had metastatic recurrences. Seventy-four percent of patients with these recurrences were asymptomatic and recurrences were detected primarily by either chest X-ray or blood test. The utility of a CT scan was found to be low, with only 10% of patients having asymptomatic abdominal recurrences, the majority of these occurring in patients with pT3 N0 M0 (Bladder Cancer Stage B2 and C: tumor invades perivesical tissue [T3], no regional lymph node metastasis [N0], no distant metastasis [M0]) disease. Patients with pT1 N0 M0 bladder cancer treated with cystectomy are at very low risk for recurrence. Based on the recommendations of Slaton et al., these patients should be seen annually for history, physical examination, review of systems, chest X-ray, and laboratory studies. The latter should include a comprehensive metabolic panel with liver function tests, alkaline phosphatase, and CBC. Twenty percent of pT2 N0 M0 (Bladder Cancer Stage B1 and B2: tumor invades muscle [T2], no regional lymph node metastasis [N0], no distant metastasis [M0]) patients developed recurrences, most commonly in the chest. Ninety percent of those with pulmonary metastases were asymptomatic and metastases were detected only by chest X-ray, supporting the use of this follow-up test. Patients with bone, brain, or pelvic recurrences were predominantly symptomatic, but those with lymph node recurrences were detected predominantly on physical examination. CT scan detected only one recurrence in an asymptomatic patient with lymph node metastases. A logical algorithm for following these patients, as presented by Slaton et al., is to perform the same studies as for patients with pT1 N0 M0 lesions but at 6-month intervals for the first 3 years and annually thereafter. CT scans are not recommended based on their low likelihood of detecting a pelvic recurrence.
Of 163 patients with Stage pT3 N0 M0 or higher bladder cancer, 65 (40%) had a recurrence. Eighty-two percent of patients with Stage pT3 N0 M0 tumors and pulmonary recurrences were detected by chest X-ray. Fifty percent of patients with bone metastases presented with symptoms, whereas the other 50% were detected by elevated levels of alkaline phosphatase. CT scanning was useful to detect two of six patients with hepatic metastases, 3 of 19 patients with pelvic recurrences, and four of six patients with lymph node recurrences. Although the majority of tumor recurrences are not detected by CT scanning, it has increased utility in this category. Follow-up guidelines remain the same as for patients with Stage pT2 N0 M0 lesions, with the addition of abdominal and pelvic CT scanning at 6, 12, and 24 months after cystectomy. The number of patients with symptomatic recurrences shows the importance for review of systems, history, and physical exams on each visit. There were no asymptomatic recurrences after 24 months for any stage of bladder cancer after cystectomy. Bladder cancer follow-up strategies are summarized in Figure 2.
Renal Cell Carcinoma
The likelihood that renal cell carcinoma will recur is related much more to stage than to grade. Greater than 90% of patients with pathologic Stage T1 N0 M0 tumors (Renal Cell Carcinoma Stage I: tumor 7 cm in greatest dimension, limited to the kidney [T1], no regional lymph node metastasis [N0], no distant metastasis [M0]) with negative surgical margins will not have recurrence following radical nephrectomy. The stage-dependent recurrence of renal cell cancer has been reported by Levy et al. 39 Two hundred eighty-six patients with Stage pT1-pT3 N0 or NX renal cell cancer (Renal Cell Carcinoma Stage I-IIIA: tumor extending into major veins or invading adrenal gland or perinephric tissues but not beyond Gerota fascia [T3], no regional lymph node metastasis [N0], no distant metastasis [M0]) were followed for interval and site of tumor metastases and method of diagnosis correlating with detection of the recurrence. Of these, 68 patients developed 92 metastases at a median interval of 23 months after nephrectomy. Eight of 113 Stage pT1 N0 M0 patients, 17 of 64 Stage pT2 N0 M0 patients (Renal Cell Carcinoma Stage II: tumor more than 7 cm in greatest dimension, limited to the kidney [T2], no regional lymph node metastasis [N0], no distant metastasis [M0]), and 43 of 109 Stage pT3 N0 M0 patients had recurrences. Overall, 64% of patients with recurrences were asymptomatic. In patients with Stage pT1 N0 M0 disease, all four with pulmonary recurrences were asymptomatic and recurrence was detected on chest X-ray. Recurrence in all other patients was detected with symptoms (two with bone, one with brain, and one with uvular metastases). For patients with Stage pT2 N0 M0 disease, recurrences in eight of nine patients were detected on chest X-ray but two of three with osseous metastases had symptoms, as did both patients with brain and pancreatic metastases. Hepatic metastases in patients were detected by elevated liver function tests and adrenal and lymph node metastases in two patients were detected only by CT scan. Only three patients with Stage pT3 N0 M0 disease had recurrences detected by CT scan; an adrenal metastasis, a local recurrence, and a hepatic metastasis. All other patients had their recurrences detected by other symptoms, physical examination, chest X-ray, or blood work.
Levy et al. 39 recommended that patients with Stage pT1 N0/NX disease be seen at 12-month intervals with a history, physical examination, chest X-ray, and laboratory studies, but no abdominal imaging. Stage pT2 N0/NX tumors should have the same studies performed at 6-month intervals for 5 years with the addition of an abdominal CT scan at 24 and 60 months. Patients with Stage pT3 N0/NX disease should have the same studies as those patients with Stage pT2 N0/NX disease, with the addition of a visit at 3 months for history, physical examination, chest X-ray, and laboratory studies. The rationale for no CT scans in the first 2 years is based on the finding that lesions in asmptomatic patients were not detected in the first 2 years by CT scan alone. However, site-specific symptoms may prompt CT scans at other intervals. In addition, site-specific symptoms should prompt the appropriate test for ruling out any metastasis in the brain, bone, and extremities.
Others have recommended that imaging studies should not be performed in follow-up of patients with renal cell carcinoma. 40 In that series, no patients with Stage T1 N0 MX disease had recurrence, but those with pT2 and pT3 disease had a high likelihood of recurrence to the primary metastatic sites of lung, bone, abdomen, local, and liver. These authors recommended 6-month visits for 3 years and then annually, with visits consisting of history, physical examination, and blood work without any imaging tests. This conflicts with Levy et al.'s data that demonstrate that most patients with pulmonary metastases are asymptomatic. If untreated, patients with pulmonary lesions will become symptomatic. Whether earlier detection and intervention is of benefit is still being debated. Renal cell carcinoma has a 2% renal contralateral recurrence rate. Although urinalysis has not been evaluated in these studies, it is a logical test to perform. Follow-up recommendations are shown in Figure 3.
Renal parenchymal-sparing surgery is occurring more frequently. Historical indications for partial nephrectomy have included a solitary kidney, bilateral tumors, renal insufficiency, as well as renal stone formers and indeterminate masses. However, many urologists perform partial nephrectomies in the context of a normal contralateral kidney when lesions are polar and less than 4 cm. 41–43 The risk of multifocality in the remaining kidney was best assessed by Kletscher et al.,44 who analyzed 100 radical nephrectomy specimens pathologically. Sixteen specimens demonstrated multifocal disease. However, it was suspected by preoperative imaging in seven of these (44%) and confirmed by gross examination in three. The true incidence of unsuspected multifocality is 6%. The local tumor recurrence rates following partial nephrectomy for pathologic Stages T1,T2, T3a, and T3b N0 M0 wre reported to be 0%, 2%, 8.2%, and 10.6%, respectively.45 The metastatic tumor recurrence rates following partial nephrectomy for pathologic Stages T1, T2, T3a, and T3b N0 M0 were 4.4%, 5.3%, 11.5%, and 14.9%, respectively. However, modifications of the 1987 International Union Against Cancer Staging criteria were referenced as applied, separating classification T1 from T2 tumors at a size cutoff of 2.5 cm. Due to biologic behavior, Stage T1 N0 M0 tumors are now defined as being smaller than 7 cm and therefore these groups should be grouped using present criteria. Based on these findings, it is recommended that tumors smaller than 2.5 cm do not need additional monitoring. It also recommended that patients with tumors larger than 2.5 cm but confined to the renal capsule should have an annual chest radiograph and an abdominal CT scan every 2 years. Patients with Stage T3 N0 M0 tumors should have an annual chest radiograph and an abdominal CT every 6 months for the first 2 years and every 2 years thereafter. Ten pT3 N0 M0 patients (18.8%) had a local recurrence, 7 within the first 2 years, which is the basis for the abdominal imaging. However, the data do not strongly support abdominal CT imaging every 2 years for patients with Stage T2 N0 M0 disease. Only 3 of 151 Stage T2 N0 M0 patients (2%) had local recurrence, at a mean of more than 5 years after surgery. Eight Stage T2 N0 M0 patients had metastatic recurrences, but in only one of these patients a pancreatic recurrence at 30 months after surgery might have been detected. Further studies of this patient population would provide clinical utility. Recommendations for follow-up are shown in Figure 4.
Monitoring for recurrent testicular carcinoma must be separated into (1) surveillance following orchiectomy for presumed understaging of localized disease and (2) follow-up after secondary therapeutic modalities looking for recurrence. Strategies for the follow-up of testicular tumors are well defined based on the pathology of the orchiectomy specimen. Nonseminomatous germ cell tumors usually recur in the first 2 years after orchiectomy and very uncommonly after that. Conversely, seminomatous germ cell tumors also recur commonly in the first 2 years, but 16% of recurrences can develop after the 2-year mark.
Due to the chemotherapy and radiation sensitivity of seminomas, most clinically localized seminomas are treated with radiation to the retroperitoneum. The low doses required to treat understaging have a very acceptable side-effect profile and surveillance for these patients is not standard. Conversely, only 30% of patients with nonseminomatous germ cell tumors will recur in the retroperitoneum. Risk factors for these recurrences were identified and include a predominance of embryonal cell carcinoma in the primary orchiectomy specimen, vascular invasion, and a serum α-fetoprotein level higher than 70. 46 Patients with these pathologic findings will be considered at higher risk, with up to 30% having microscopic retroperitoneal metastasis. In this scenario, surveillance is not endorsed, and patients best undergo either a retroperitoneal lymphadenectomy, preferably using a nerve-sparing template or two cycles of cisplatin, etoposide, and bleomycin (PEB) chemotherapy. Placing high-risk patients on a surveillance protocol based on the rationale that detection of a recurrence followed by PEB chemotherapy will yield a nearly 100% cure rate is not conventional. Low-risk patients are more suitable candidates for surveillance.
Surveillance strategies for Stage T1 N0 M0 nonseminomatous germ cell tumors (Testis Cancer Stage I: tumor limited to the testis and epididymis and no vascular/lymphatic invasion. Tumor may invade the tunica albuginea but not the tunica vaginalis [T1], no regional lymph node metastasis [N0], no distant metastasis [M0]) should be intense in the first 2 years following orchiectomy. As reviewed by Koch, 47 this entails monthly visits with a history, physical examination, and tumor markers. Chest X-rays are performed every month and CT scanning of the abdomen and pelvis are performed every third month. Motivated compliant patients are critical to the successful detection of recurrent disease. Sharir et al.48 proposed a slightly different follow-up strategy for these patients, beginning Month 2 after orchiectomy and continuing every 2 months for the first 2 years, every 3 months in the third year, biannually in the fourth year, and annually thereafter. In the critical first 2 years, follow-up includes history, physical examination, tumor markers, and chest X-ray every other month, abdominal and pelvic CT scan at Months 4, 8, and 12, and this strategy should be repeated at those intervals in the second year. CT scans are not performed thereafter. Using this protocol, one less abdominal and pelvic CT scan was performed per year in the first 2 years. The authors also noted that no patient had pulmonary metastases detected by chest X-ray as a single modality of surveillance. Although Sharir et al. advocated removal of chest radiography from the surveillance protocol, others have appropriately disagreed.49 A proposed surveillance strategy is provided in Figure 5.
Patients at high risk for microscopic retroperitoneal dissemination who have undergone retroperitoneal lymphadenectomy are unlikely to have recurrence in the retroperitoneum following surgery. Only 3 of 464 patients (0.6%) in an Indiana University series with lymph node-positive disease following retroperitoneal lymphadenectomy experienced late relapse, and all recurrent disease was in the chest. 50 Late retroperitoneal relapses following retroperitoneal lymphadenectomy are related to an inadequate initial operation.51 History, physical examination, chest radiography, and tumor markers should be performed at 6- month intervals for 1–5 years and then annually. It is not unreasonable to add CT scanning on an annual basis for some period of time such as 5 years; however, there is no clear-cut data to substantiate these recommendations. Similarly, patients who have undergone retroperitoneal radiation for presumed localized testicular seminoma are unlikely to have recurrence in the retroperitoneum. However, as seminomas are less likely to produce tumor markers, annual abdominal and pelvic CT scanning is not unreasonable. Therefore, long-term follow-up should include history, physical examination, tumor markers, chest radiographs, and CT scans as outlined in Figure 5.
PET for testicular cancer is potentially useful for clinical staging. It has been shown to be sensitive for detection of small retroperitoneal lymph nodes. 52 For clinical staging, PET has demonstrated a sensitivity of 87% and a specificity of 94% in a small series.52 Another study found no difference between the sensitivities of PET and CT scanning, but PET scanning was more specific than CT scanning.53 For patients who received PET scans within 2 weeks of completing chemotherapy, false-negative rates were much higher than for patients who received PET scanning after more than 2 weeks of completing chemotherapy. At this time, there is no proven role for PET scanning in the follow-up of patients with testicular cancers.
The likelihood of developing a metachronous testicular tumor in the remaining testicle is between 2% and 3%. The characteristics of this phenomenon were evaluated by Park et al. 54 Fifty percent of first tumors were seminoma compared with 55% of second tumors. The concordance rate for the first and second tumors was 35%. Metachronous testicular cancer can occur years after the primary orchiectomy. Therefore, patients should perform testicular self-examination for life and they should be followed annually by their physicians.
Long-term follow-up is necessary for patients treated for testicular cancer due to the risk of developing a second malignant neoplasm. A large study of more than 28,000 survivors of testicular cancer performed by Travis et al. 55 found second cancers in 1406 men. Statistically significant increases were noted for acute lymphoblastic leukemia, acute nonlymphocytic leukemia, melanoma, non-Hodgkin lymphoma, cancers of the stomach, colon, rectum, pancreas, prostate, kidney, bladder, thyroid, and connective tissues. The risk of developing these tumors was similar whether the initial testicular cancer was pathologic seminoma or nonseminomatous tumor. Adjuvant radiation and chemotherapy were associated with secondary leukemia development whereas radiotherapy alone was associated with increased incidence of carcinoma of the stomach and bladder. In a 4-year follow-up of 802 patients with pure testicular seminoma of whom 758 received radiation, 76 chemotherapy, and 5 surveillance, 32 secondary cancers were recorded.56 The overall relative risk (RR) of secondary cancer was +4.8, with 1.1% of patients developing a nontesticular second tumor. Only the risk for renal cell carcinoma was increased at a statistically significant level. Based on these data, it is evident that patients with testicular cancer should be followed annually for life, with a good history, physical examination, and routine laboratory tests to detect metachronous testicular tumors and secondary malignancies.
The potential costs of follow-up strategies have important implications. Table 1 shows MediCare reimbursements at the University of California, Davis, Medical Center for established clinic visits and individual tests commonly utilized in the follow-up of genitourinary malignancies. Based on values in Table 1, Figure 6 summarizes the costs for 5 years of follow-up for the urologic cancers using algorithms described in the text and in Figures 1–4. The level of established care assigned to each follow-up visit depended on the extent of physical examination and systems review and was between Levels 2 and 3. The surveillance guidelines employed for testicular cancer pT1N0Mx were those in Figure 5.
|CT scan of chest||153.11|
|CT scan of abdomen||215.15|
|CT scan of pelvis||177.15|
|Established patient visit|
Strategic rationale for monitoring genitourinary cancer recurrences is multifactorial and based upon variable amounts of investigated evidence. It is intuitively likely that physicians often perform too many studies, at inappropriate intervals for detecting tumor recurrences. An example of this is the clinical propensity to image Stages T2-3 N0 M0 bladder and Stages T2-3 N0 M0 renal cell cancer with frequent abdominal and pelvic imaging. The frequency can be limited as patients with most recurrences are symptomatic and the recurrences are not detected by CT scans. Furthermore, unless the patient's performance status and personal desire is to receive immunotherapy or protocol-based therapeutics for recurrent renal cell cancer, the need for any significant imaging is minimal. Currently available therapies offer little survival benefit.
Prostate cancer is most simply followed by monitoring serum PSA levels at 3 to 12-month intervals based on stage and time interval from treatment. DRE is useful when PSA levels increase after radical prostatectomy and perhaps after radiation therapy. Urinalysis, serum PAP, CT scans, chest radiographs, bone scans, RT-PCR, and ProstaScint and PET scans are of limited utility. PSA is cost-effective and early detection of recurrent prostate cancer can have implications for salvage treatments and outcomes.
Cystoscopic follow-up for superficial bladder cancer is practiced universally among urologists. Alternatives to the standard pattern have been proposed. For example, 6 and 9- month cystoscopies can be omitted if results of the 3-month cystoscopy are normal. 57 The rationale is that the progression rate is 1%, even though the recurrence rate is 10%. It remains unknown whether tumors left undetected for up to 9 months might have a higher progression rate. Therefore, it is premature to endorse this modality. No bladder marker test should substitute for cystoscopy, and adjunct usage has limitations as outlined. High-risk patients should have indefinite annual upper tract imaging and this tends to be loosely followed in many clinical practices. CT scans are not warranted in the follow-up of postcystectomy Stage pT1-2 N0 M0 bladder cancer when applying the analysis of Slaton et al.38 This differs from the schema proposed by Montie,58 who advocates CT scanning at 6 and 12 months for all pathologic stage disease. Montie's recommendations were proposed before the M.D. Anderson data.38 However, upper tracts need to be studied annually in this postcystectomy population using either i.v. urography or a loopogram. Sonography and CT scanning have not been compared with these imaging modalities for sensitivity of detecting recurrent upper tract disease and at this point should not be substituted. Chest radiography should be employed routinely as described, as patients with pulmonary metastases are usually asymptomatic until the tumors become significantly enlarged. As mentioned by Montie,58 salvage chemotherapy for patients with metastatic bladder cancer results in many partial responders, but few cures. This suggests that follow-up tests to detect metastatic disease should be limited.59–61 The argument for early detection is to limit morbidity, permit patient enrollment in clinical trials, and allow potential planning for end of life issues.
A clinical tendency exists to image frequently the retroperitoneum for recurrence in patients with Stages T1-3 N0 M0 renal cell cancers. Renal cell carcinoma rarely recurs after nephrectomy for Stage T1 N0 M0 disease, and the proposed follow-up recommendations therefore differ somewhat from those of Montie. 62 Surveillance abdominal CT imaging for Stages T2-3 N0 M0 renal cell cancer should be limited, as most patients are symptomatic and recurrences are not detected by CT scans.39 Furthermore, unless the patient's performance status and personal desire is to receive immunotherapy or protocol-based therapeutics for recurrent renal cell cancer, the need for any significant imaging is minimal. Urinalysis is inexpensive and can detect 2–3% of patients who will develop metachronous contralateral disease, although this recommendation has not been validated clinically. Following partial nephrectomy, 6% of patients may harbor clinically unsuspected multifocality and may be at risk for ipsilateral tumor recurrence. Then, more frequent renal imaging is recommended. Ultrasound has not been compared directly with CT scanning for detecting recurrent renal cell cancer. It might prove a cost-effective alternative for renal or fossa tumor lesions, but this may not be so when the lymph nodes or other structures are involved.
Surveillance for testicular carcinoma differs for patients with localized disease, but risk factors for understaging after orchiectomy versus those undergoing follow-up after secondary treatments. Clear criteria exist to define patients at risk for retroperitoneal metastasis following orchiectomy with negative clinical imaging and tumor marker studies. Placing these patients on surveillance with the concept that salvage rates are high using multimodal therapy is inappropriate. Patient compliance is the major factor in detecting cancer in patients on primary surveillance. After secondary treatment with retroperitoneal radiation therapy for clinically localized seminoma or retroperitoneal lymphadenectomy yielding pathologically negative lymphatic metastasis for nonseminomatous germ cell malignancies, recurrences are unlikely to be in the retroperitoneum and more likely to be in distant sites such as the lungs. Due to the population afflicted with testis cancer, significant patient work attendance and productivity can be lost with intensive office and imaging appointments.
In addition to an effective detection strategy, surveillance protocols have secondary implications. Assessments must be made as to whether testing will result in an earlier diagnosis of cancer recurrence for which intervention and therapy can improve the outcome. For example, following radiation therapy, a recurrence of prostate cancer in a patient with significant comorbidities or advanced age might be treated with androgen ablation therapy regardless of whether it is a local or distant recurrence. A prostate biopsy might be of little utility in addition to an increasing PSA level in altering patient management. A patient with poor performance status who is not a candidate for immunotherapy as secondary treatment for renal cell carcinoma does not need to be followed at close intervals with multiple tests. However, this must be weighed against the patient's desire and right to know of a recurrence in the face of incurable disease. End of life issues are not solely for the physician to determine, but reporting an unalterable course to the patient may lessen the quality of remaining time.
Medical and patient economic factors are impacted by follow-up strategies. Health care costs can be affected significantly by physician surveillance practices. Surprisingly, few guidelines exist in this arena, whereas significant insurance emphasis is placed on reviewing indications for treatments or medications. It is likely that a survey of physician surveillance protocols, adjusted for tumor grade and stage, would demonstrate large practice variability. 63 Within a single physician's practice, tumors with the same stage and grade might be monitored differently due to lack of available follow-up algorithms. Frequent tests, especially those that are time-consuming and complex, affect the patient as well. This can result in numerous medical visits for a patient with a limited prognosis, thereby detracting valuable time away from friends or family. Frequent tests cause patients who have returned to work to miss even more work. Patients with testis cancer are often early in their careers or jobs and work absences are not well received by employers. This might secondarily decrease their quality of life and that of their friends and family members.
The medicolegal implications of follow-up strategies must also be considered. Decreased testing might result in a delay in diagnosis of a cancer recurrence and this must be balanced against the potential benefits that would result from earlier intervention. Physicians are pitted between cost containment pressures from insurance plans and health maintenance organizations and malpractice fears. The balanced solution to this problem has not been defined conclusively by clinical studies, but likely has been evaluated frequently by the legal system. Surveillance guidelines, although not inclusively data driven for every test in each cancer, provide evidence-based structure for the follow-up of genitourinary malignancies.
- 1American Cancer Society. American Cancer Society statistics. Washington, DC: American Cancer Society, 2000.
- 2American Cancer Society. Surveillance research; and surveillance, epidemiology, and end results program 1970–1996, Division of Cancer Control and Population Sciences, National Institutes of Health. Washington, DC: American Cancer Society, 2000.
- 4Manual for staging of cancer. Philadelphia: Lippincott-Raven, 1997., , , et al.