Renal cancer. Contemporary management. Editor John A. Libertino. Springer New York 2013.
4. Active surveillance
Обоснование для AS
Between 1983 and 2002, RCC tumors identiﬁed between 2 and 4 cm in size have increased in incidence from 1.0 to 3.3 per 100,000. Resected tumor size dropped from a maximum diameter of 7.8–5.3 cm between 1989 and 1998. The incidental diagnosis of RCC increased from 7% to 13% in the early 1970s to 48–66% of kidney cancer cases currently; incidental tumors are most commonly found in patients older than 65 years, a group more prone to the adverse effects of surgery due to the increased presence of comorbidities. In an examination of a cohort of 26,618 individuals treated surgically for localized kidney cancer, the relative beneﬁt of therapy is notably diminished by competing causes of mortality in older patients, with nearly one-third of patients with RCC aged 70 years and older succumbing to unrelated comorbid disease within 5 years of receiving curative RCC surgery. The current epidemiology of RCC suggests a marked increase in the incidence of cases, and despite a matching increase in therapy for incidentally detected RCC, the overall RCC mortality rates across the population have not decreased. Taken together, these data suggest that many early stage I RCCs are often clinically indolent and current treatment algorithms may overemphasize the beneﬁts of surgery compared to less aggressive treatment strategies.
As described in this and previous sections of this chapter, an appropriate algorithm for management of the SRMs would include a pretreatment renal mass biopsy to conﬁrm the diagnosis and to consider AS and expectant management in appropriately selected persons. The evidence to supporting this protocol includes the following:
- Not all renal masses are RCC. Review of the literature indicates that approximately 15% of SRMs are benign lesions that do not demand or beneﬁt from any intervention.
- SRMs are frequently detected in elderly patients with comorbidities. The risk of perioperative morbidity and possible mortality is likely higher in these patients and may markedly exceed the anticipated risk of impact from RCC progression or metastasis.
- The majority of SRMs conﬁrmed as RCC have nonaggressive pathologic features, with histology suggestive of low-grade appearance and anticipated to demonstrate a slow growth rate and a low metastatic potential, early in their natural history. Predictive tools exist to help quantify the likelihood of aggressive versus indolent disease and to quantify the risk of competing comorbidities on longevity to make informed treatment decisions.
- A delay in treatment does not appear to lessen the effectiveness of standard surgical intervention. The outcome of RCC therapy may not be compromised if progression is detected early and curative treatment performed. Progression to advanced stage is rare in well-selected patients managed by active surveillance. As techniques to monitor and predict RCC growth and behavior evolve, this risk may be further minimized.
Показания для AS
Paramount to the evaluation of a patient with a newly diagnosed SRMs is an assessment of the patient’s comorbid conditions with the goal of stratifying risk of treatment prior to choosing a treatment strategy. As with nephron-sparing surgery, we tend to categorize the indication for AS into absolute, relative, and elective indication. Patients with severe comorbidities in which surgical treatment would impart an immediate and unacceptable risk of mortality are considered to have an absolute indication for AS. Those with a second and potentially more aggressive malignancy, the potential need for renal replacement therapy, and other signiﬁcant medical comorbidities that make surgery high risk but not intolerable are considered to have a relative indication for observation. Elective indications include low-risk surgical candidates that choose to pursue AS as an alternative to active treatment. In a recently published review of contemporary AS series, the indications were elective (60.9%), relative (12.5%), and absolute (26.6%) in the eight studies (n = 312 patients) reporting the reason for AS enrollment.
Предиктивные инструменты и использование в клинических условиях
The primary goal of AS is to balance the risks of treatment versus the risks of disease progression and the development of metastatic disease. A number of posttreatment nomograms have been developed to predict risk of cancer-speciﬁc death or disease recurrence which is beyond the scope of this review. However, several preoperative predictive models have been developed which one can use to quantify risks based on commonly available preoperative parameters. Initial efforts to predict benign versus malignant disease and indolent versus aggressive tumors using clinical characteristics such as tumor size, age, gender, and smoking history were met with limited success.
Subsequent efforts to determine renal grade preoperatively were also unsuccessful with limited predictive accuracy. In contrast, a number of clinical tools have recently been developed to determine tumor malignant potential and risk of death based on pretreatment characteristics with acceptable predictive accuracies facilitating use in the clinical setting. To facilitate their use, we have recently operationalized clinical nomograms with predictive accuracies greater than 70% to expedite their use (www.cancernomograms.com).
In 2011, Kutikov et al. developed a (Фиг. 8.2) tool to predict the probabilities of harboring malignant and high-grade pathology based on anatomic variables which was described in more detail earlier in this review (103). For example, an 80-year-old male with an enhancing renal mass with a nephrometry score of 1 + 3 + 1 + a + 2 = 7a has only a 26% chance of lesion malignancy using Kutikov’s model. If the mass is malignant, the chance of a high-grade malignancy (Fuhrman grade III or IV) is approximately 30%. Therefore, the probability of harboring high-grade malignancy is 7.8% (0.26 × 0.30 = 0.078). In contrast, the chance of malignancy in an 80-year-old female with a nephrometry score of 2 + 2 + 2 + a + 3h = 9ah is 92% with a 59% chance of high-grade disease should malignancy be present (0.92 × 0.59 = 0.542 or 54.2% chance of a high-grade malignancy). Using readily available clinical information, this validated model has allowed the physician to differentiate between two seemingly similar patients with clear clinical management implications.
Фиг. 8.2. Номограммы для оценки риска ренального образования быть злокачественным и высокозлокачественным. Total point values are independently calculated for the cancer and the high-grade models and then applied to the corresponding probability scale at the bottom of the ﬁgure
Kutikov and colleagues have also developed clinical tools to predict overall mortality, cancerspeciﬁc death, and death from other malignancies. Using SEER data, the authors developed a comprehensive nomogram incorporating race, gender, age, and tumor size to calculate competing risks of death and help facilitate clinical trade-off decisions (Фиг. 8.3). Whereas the initial effort was criticized for lack of comorbidity information, the authors recently updated this tool incorporating the Charlson comorbidity index (CCI) based on claims available in linked SEER-Medicare data. Using this nomogram, an 80-year-old African American male with a history of a myocardial infarction, moderate renal insufﬁciency (CCI of 3), and a 4 cm renal mass is expected to have a 5-year mortality of 5% from RCC versus 48% from non-RCC causes. Meanwhile, a 75-year-old Caucasian female with no signiﬁcant comorbidities (CCI of 0) and a 7 cm renal mass is predicted to have a 5-year mortality of 13% from RCC and 7.5% from other causes. Although these tools are limited by use of only treated patients for model development, with further reﬁnement, these and other predictive models show signiﬁcant potential for counseling patients newly diagnosed with SRMs, particularly elderly individuals with signiﬁcant competing risks.
Фиг. 8.3. Nomogram evaluating 5-year competing risks of death in patients with localized renal cell carcinoma. Total point values are independently calculated for each cause of death and then applied to the corresponding probability scale at the bottom of the ﬁgure
These predicted probabilities can then be objectively incorporated into treatment planning accounting for risks of comorbid medical conditions and the morbidity of treatment itself. As part of the initial workup, each physician must attempt to quantify life expectancy, assess the patient’s performance status and operative risk, and compare these factors against the potential for morbidity and mortality of an untreated SRMs after calculating the probability that an aggressive RCC is present. This optimally would be a multidisciplinary approach that includes the urologist; primary care provider; cardiac, pulmonary, and nephrology specialists; and an anesthesiologist. In patients that are elderly and/or have diabetes, hypertension, and other systemic diseases that predispose to chronic kidney disease (CKD), the potential need for postoperative dialysis must be taken into consideration. It is well known that end-stage renal disease carries signiﬁcant adverse morbidity and mortality. Furthermore, increased risks of death, cardiovascular events, and hospitalization have been demonstrated in patients with mild renal insufﬁciency in recent large population-based cohort data. At our intuition, all consultations for SRMs include a determination of the creatinine clearance and GFR allowing for stratiﬁcation into CKD stages. Patients with CKD stage IV or V are typically referred to nephrology for further evaluation functional risk preoperatively. In all situations where patients choose AS over active treatment, in-depth counseling as to the limitations of radiologic surveillance and growth kinetics and the possibility of disease progression including metastases and death is performed. Patients must consider and accepted the calculated risk involved due to the occasionally unpredictable behavior of RCC prior to proceeding with AS.
Currently, there is no data to support any speciﬁc AS protocol (frequency and type of radiographic follow-up). Unfortunately no studies comparing the effectiveness of active surveillance/delayed intervention with traditional surgical therapies or ablative techniques have been performed. Performing such trials poses tremendous logistical challenges under current practice patterns/ incentives. In addition, a high degree of patient adherence is required to participate in such trials due to the implicit risk involved with AS and, for some, the demanding follow-up schedule. Studies must also examine the costs of surgical morbidity and mortality in such these cohorts. To minimize the risks of undetected disease progression, current recommendations call for repeat imaging utilizing a consistent modality at deﬁned intervals (initially 3–6 months). The choice of imaging interval should be based on clinical risk factors speciﬁc to the renal mass and the patient’s overall health status. We typically obtain imaging at 3–6-month interval following initiation of AS with the goal of establishing baseline growth kinetics (time zero to point one). Once these are established, the timing of further imaging studies is determined. Tumor size comparisons should be performed using the same lesion characteristics (e.g., maximum tumor diameter or estimated tumor volume) obtained from consistent imaging modalities at the same tumor level. Most importantly, in the event that their tumor exhibits a rapid growth rate, a new lesion appears, or the onset of clinical symptoms occurs, patients must be appropriately counseled objectively regarding the risks of continued AS versus immediate treatment in their individual circumstances.
Рентгенографические прогностические факторы темпа роста опухоли и злокачественного потенциала
The majority of localized renal tumors exhibit slow radiographic growth with low metastatic potential while under an initial period of observation as shown from pooled published observations. Deﬁnitive radiographic characteristics associated with rapid growth rate or aggressive malignant potential have yet to be identiﬁed. There has been no correlation documented between tumor growth and patient age, initial MTD, tumor size >4 cm, development of clinical symptoms versus incidental detection, multifocality, or solid/cystic appearance. Initial assumptions that larger renal masses demonstrated faster growth rates have been proven incorrect. In fact, smaller tumors have been shown to grow at proportionally faster rates than larger tumors based on annual percent change in tumor size and volume. The theory behind this observation is that a tumor’s growth rate is initially exponential and then decreases with increasing size (Gompertzian theory of growth kinetics). Some series have reported on the observation of larger tumors (clinical T1b and T2) in select patients with signiﬁcant medical comorbidity signifying that the indications for surveillance may be expanding. However, the biology of these lesions must be distinguished from the infrequent case of a localized mass with aggressive malignant potential whose disease progresses during a period of AS.
Efforts to predict the malignant potential/ growth rate of SRMs have yielded conﬂicting results and often lack complete pathologic assessment. Studies examining Fuhrman grade on ﬁnal pathology and growth rate during surveillance showed that grade 3 lesions grew faster than grade 2 lesions (0.93 vs. 0.28 cm/year; p = 0.01); however, these ﬁndings are limited by small sample size (n = 18). In addition, grade 1 lesions grew faster than grade 2 lesions (0.37 vs. 0.28 cm/year) although this trend was not statistically signiﬁcant (p = 0.47). Others have retrospectively compared patients with proven RCC (n = 10) versus oncocytoma (n = 6), reporting no statistical differences in tumor growth rate between groups (0.71 vs. 0.52 cm/year). Data from one of the largest single institution experiences to date (154 patients, 173 SRMs followed for a minimum of 12 months) showed no differences in growth rates when stratiﬁed by Fuhrman grade or presence of benign histologic. Chawla et al. reported no difference between initial MTD (2.0 vs. 2.2 cm; p = 0.59) and mean growth rate (0.1 vs. 0.4 cm/year; p = 0.15) in oncocytomas versus RCC. This ﬁnding is supported by the observation from two studies that percutaneously biopsied oncocytomas have displayed positive growth rates with observation suggesting that a positive growth rate is not always indicative of malignant histology [116, 117]. Kawaguchi et al. observed a yearly linear growth rate of 0.2 cm, which is not too dissimilar from the growth rates of SRMs of variable histology reported in other series . Only eight of the 45 oncocytomas underwent extirpation, with one of the eight lesions harboring chromophobe RCC. These data highlight the need for the identiﬁcation of characteristics that better predict aggressive malignant potential.
Малые образования почек показывают “нулевой чистый рост”, когда находятся под наблюдением
The range of linear growth rates of SRMs on surveillance in contemporary series is between 0.06 and 0.86 cm/year [19–35]. Two recent publications summarizing the available data reported mean linear growth rates ranging from 0.28 to 0.31 cm/year. However, within these reported series of SRMs on AS, a subset of SRMs that demonstrated no interval growth on serial imaging has been identiﬁed. When comparing radiographic characteristics of zero net growth lesions (n = 35) and those exhibiting growth (n = 70), no differences were seen with respect to patient age (p = 0.96), initial MTD (p = 0.41), solid/cystic appearance (p = 1.0), or incidental detection rate (p = 0.38) . As expected, lesions demonstrating positive growth rates underwent higher rates of active treatment (51 vs.
17%, p = 0.001) yet revealed similar malignancy rates (83 vs. 89%, p = 0.56). This observation has been conﬁrmed in other small series [19, 35]. Among the studies with available data [19–21, 24, 26–30, 32, 34–36], 22.9% of SRMs exhibited zero net growth over time, and no difference in initial MTD (2.3 ± 1.3 cm vs. 2.5 ± 1.3 cm; p = 0.21) or pathologic malignancy rate (88.2% vs. 92.3%, p = 1.0) was observed between lesions exhibiting positive and zero growth when the available data were pooled. While the lack of growth under surveillance did not correlate with benign histology, all of these zero net growth lesions remained localized radiographically with no patients developing measureable metastatic disease.
Наблюдаемое SRM прогрессирование в метастазы
Fortunately, progression to metastatic disease in patients with SRMs under AS has been an uncommonly observed event. Of 880 patients with SRMs under AS identiﬁed in a systematic review, only 18 (2.1%) patients progressed to metastatic disease. In the 13 patients with reported indications for AS, indications were absolute in 61.5% and elective in 38.5%. Distant visceral or bony disease with or without positive lymphadenopathy (eight patients; 73%) and lymph node involvement only (three patients; 27%) was identiﬁed in the patients with available information. Histology was predominantly clear cell (66.7%) and papillary (22.2%), with one lesion exhibiting mixed clear cell and papillary features (11.1%.). Fortunately, the mean time to detection of metastasis, on average, occurred later in the course of AS (mean of 40.2; range 12–132 months).
Таблица 8.3. Comparison of clinical and cross-sectional imaging characteristics in patients who did not progress to metastasis (pooled cohort series data) and patients who demonstrated evidence of progression (case series data) during periods of observation
Comparing patients with metastatic disease to those that remained on AS (Таблица 8.3), there were signiﬁcant differences in mean patient age (75.1 vs. 66.6 years; p = 0.03), but the duration of observation was similar between groups (40.2 vs. 33.3 months; p = 0.47). Larger tumor size (4.1 vs. 2.3 cm; p < 0.0001) and estimated tumor volume (66.4 vs. 15.1 cm3; p < 0.0001) at diagnosis as well as mean linear (0.80 vs. 0.30 cm/year; p = 0.0001) and volumetric growth rate (27.1 vs. 6.2 cm3/year; p < 0.0001) were greater in patients that progressed to metastasis. Lesions progressing were predominantly high grade at the time of histologic conﬁrmation. Those that progressed were more common in elderly patients with absolute indications for surveillance with higher risk tumors. This group included some individuals who were lost to follow-up, and it is conceivable that a proportion of these patients would have undergone deﬁnitive treatment if more closely followed.
AS remains an underutilized and evolving management strategy, and the interpretation of these data involves signiﬁcant limitations including the level of evidence (all £ level III) and lack of centralized pathologic evaluation. These studies may contain signiﬁcant selection bias, and therefore, it is especially important to exclude rapidly growing (if serial imaging available at presentation) and clinically high-risk lesions. Despite the limitations inherent to AS, the available data show that metastasis tended to occur late in the course of AS (>3 years following diagnosis), almost all lesions that progressed to metastasis were >3 cm when metastases were detected and demonstrated positive growth rates, and no lesion exhibiting zero net growth while under surveillance has developed metastases while under observation. The most accurate available predictor of potential for disease progression among readily available metrics signaling the need for deﬁnitive intervention appears to be positive growth rate. Based on the best available data, lesions demonstrating zero net growth have not metastasized and appear most appropriate for prolonged AS. Only one case (2.4 cm renal mass) progressing to bony metastases (after 5 months) with no change in tumor size has been reported. Although this tumor may have been systemic at its initial diagnosis, this one case reinforces the need for careful patient selection for entry onto an AS protocols.
Стоимость-эффективность AS относительно активного лечения
With the increasing costs of healthcare globally, cost-effectiveness relative to other treatment modalities has become an increasingly signiﬁcant component in clinical decision making. This may be especially true in clinical scenarios where the treatment of choice has questionable effect on disease biology, such as the treatment of low-risk early-stage cancers. Using decision analytical modeling, a means to evaluate evidence from multiple sources and evaluate the impact of uncertainty on clinical outcomes, several recently published studies have evaluated the cost-effectiveness of various approaches for treatment of SRMs. Evaluating the costs associated with diagnosis, Heilbrun et al. performed a cost-effectiveness analysis of percutaneous biopsy and AS versus active treatment in a hypothetical cohort of 2 cm renal masses in 60-year-old healthy men . Immediate treatment was the highest cost but was the “most effective” diagnostic strategy and provided the longest overall survival of 18.53 lifeyears. AS was the lowest cost, “least effective” diagnostic strategy. On cost-effectiveness analysis using a societal willingness to pay threshold of $50,000, active surveillance was the preferred choice at a $75,000 willingness to pay threshold, while biopsy and treatment were acceptable ($56,644 and $70,149 per life-year, respectively). When analysis was adjusted for quality of life, biopsy dominated immediate treatment as the most cost-effective diagnostic strategy at $33,840 per quality adjusted life-year gained. Using the base case of a SRMs in a healthy 65-year-old male to evaluate the cost-effectiveness of various nephron-sparing treatment approaches, Chang et al. found that observation was the least costly approach but that immediate laparoscopic partial nephrectomy was the most cost-effective approach among the strategies that treated the tumor with an incremental cost-effectiveness ratio of $36,645 per quality adjusted life-year gained . It should be noted that laparoscopic partial nephrectomy has largely been supplanted by the more expensive robotic approach.
Inherent to all decision analytic models, these studies are limited by the validity of the data used to develop them. The data on observation and even ablation strategies are limited to shortand intermediate-term follow-up, making the development of lifetime models incorporating these treatment options difﬁcult. Furthermore, the model cannot answer the question of which patients are best observed. Future advancements to improve the identiﬁcation of clinically signiﬁcant tumors using markers or imaging techniques will be important factors in the costeffectiveness analysis of the treatment of SRMs.