Cost-effectiveness of Contrast-enhanced Ultrasound for Diagnosis and Active Surveillance of Complex Cystic Renal Lesions

Abstract

Purpose:

The aim of this study was to determine the cost-effectiveness of contrast-enhanced ultrasound for the active surveillance of complex renal masses compared to the more established imaging modalities of CT and MRI.

Methods:

A decision-analytic Markov state microsimulation model was constructed in TreeAge Pro. We simulated independent cohorts of 100,000 60-year-old individuals with either a Bosniak IIF or Bosniak III complex renal mass who were followed for 10 years or until death. The model compared 3 imaging strategies: (1) contrast-enhanced ultrasound, (2) contrast-enhanced magnetic-resonance imaging, and (3) contrast-enhanced computed tomography for active surveillance of a complex renal mass.

Results:

For 60-year-old patients with either Bosniak IIF or III renal masses, contrast-enhanced ultrasound was the most cost-effective strategy even after varying rates of active surveillance from 10%-100%.

Conclusion:

Contrast-enhanced ultrasound is a viable and cost-effective option in the active surveillance of Bosniak class IIF and III renal cysts. Even after varying the rates of active surveillance usage, contrast-enhanced ultrasound was robust and remained the most dominant strategy. For patients who have impaired kidney functions, contrast-enhanced ultrasound is can be a safer alternative than noncontrast enhanced CT or MRI in the management of patients with Bosniak III renal cysts.

Cystic renal lesions are often found incidentally during various imaging procedures as they become more common in people over the age of 50.¹ While some incidental lesions can be clearly characterized at the time of encounter, others are considered indeterminate due to their small size and unclear features. The Bosniak classification system is used to categorize cystic renal lesions into one of 5 groups, originally based on cross-sectional images.² Bosniak class I and II are considered benign, while class IV are considered malignant. It is Bosniak class IIF and III where many of the indeterminate cystic renal lesions (<4 cm) fall. Of class IIF and III lesions, approximately 7% and 55% will be malignant, respectively.³ Precise stratification of these renal cysts is important since misdiagnosis can lead to under- or overtreatment, improper financial resource distribution, and harm to the patient. Compared to CT and MRI, contrast-enhanced ultrasound (ceUS) is a promising, newer imaging modality for the evaluation of indeterminate cystic renal lesions. CeUS has several advantages over CT and MRI including the ability to image patients with very low and even zero glomerular filtration rate and reporting fewer adverse events.^1,4,5 Zhou et al showed that ceUS had a sensitivity and specificity of 0.95 and 0.85 respectively, compared to MRI (sensitivity 0.92; specificity 0.91) for the diagnosis of complex cystic renal masses.⁶ Recently, active surveillance (AS) has been considered in contemporary guidelines for the management of small renal masses.^7,8

While recommendations generally address renal masses in elderly patients and those with significant comorbidities, there is literature supporting the slow growing and indolent nature of many small renal masses that encourage the broader application of AS.^8,9 Today, surgery remains the treatment of choice for patients with Bosniak III, however AS can be a rational treatment plan.¹⁰ While there have been studies regarding the diagnostic viability of ceUS, there have been no investigations on its applications in diagnosis plus AS of indeterminate renal masses. The aim of this study is to determine the cost-effectiveness of ceUS for the AS of complex renal masses compared to the more established imaging modalities of CT and MRI.

Methods

Model Design

A decision-analytic Markov state microsimulation model was constructed in TreeAge Pro (TreeAge 2020, Williamstown, Massachusetts). We simulated a hypothetical cohort of 100,000 60-year-old individuals with either Bosniak IIF or III complex renal masses who were followed for 10 years or until death. A start age of 60 was chosen due to significant increases in incidence of renal masses during this period.¹¹ Each cycle in the model was 1 month. The model compared 3 imaging strategies: (1) ceUS, (2) contrast-enhanced magnetic-resonance imaging (ceMRI), and (3) contrast-enhanced computed tomography (ceCT) for AS of a complex renal mass.

The model initiates with a decision tree that represents the clinical pathways based on the frequency of malignancy of Bosniak IIF and III cysts, and the diagnostic performance of the respective imaging modality (Figure 1, A). The initial imaging distributes patients into the possible resulting clinical situations (true-positive, false-negative, true-negative, and false-positive). Patients in the true- and false-negative cohorts were on AS, while patients in the true- and false-positive cohorts could either elect for primary nephron-sparing surgical intervention (NSS) or AS. While AS usage rates among eligible patients range between 10%-20%,⁸ we ran multiple base-case scenarios where we varied AS usage rates in the true- and false-positive cohorts from 10%, 20%, 50%, and 100%. In AS, patients would have follow-up imaging every 6 months for the first 2 years, and then once annually from years 3 to 5.⁹ During surveillance periods, if there were findings concerning for definitive malignancy or a significant change from prior imaging, patients would discontinue surveillance and receive NSS. The health states in our model included “no malignancy,” “localized malignancy,” “metastatic malignancy,” and “death” (Figure 1, B). Patients could remain or move between the “no malignancy” and “localized malignancy” health states; however, once patients entered the “metastatic malignancy” health state, they would remain there until death. Patients could move from any health state to death due to all-cause mortality, surgical mortality, or cancer mortality.

Figure 1.

A, Simplified model schematic for 60-year-old base-case analysis. ceCt indicates contrast-enhanced computed tomography; ceMRI, contrast-enhanced magnetic resonance imaging; CEUS, contrast-enhanced ultrasound. B, Health states within the Markov model.

Outcomes

The primary outcomes of interest were total cost, quality-adjusted life years (QALYs), and incremental cost-effectiveness ratios (ICERS). QALYs are a composite measure of the value of health outcomes that can be used to compare medical interventions. Calculating QALYs requires the utility value associated with a given health state, where 0 represents death and 1 represents a year of perfect health, as well as the length of time patients are in that specific health state. ICERs were calculated as follows:

$$\text{ICER}=\frac{\left({\text{Costs}}_{\text{Strategy} y}-{\text{Costs}}_{\text{Strategy} x}\right)}{\left({\text{QALYs}}_{\text{Strategy} y}-{\text{QALYs}}_{\text{Strategy} x}\right)}$$

The resulting ICER between strategies demonstrates the costs per additional QALY gained. A willingness to pay (WTP) threshold of $100,000 per additional QALY was used to determine cost-effectiveness. Any strategy with an ICER that was below this threshold was still considered to be cost-effective.

Parameter Estimates and Model Assumptions

Model parameters were based on estimates from the literature and are summarized in Table 1. Due to the paucity of data regarding the accuracy of the separate imaging modalities while surveilling small renal masses over time, it was assumed the respective sensitivities and specificities were constant for each subsequent surveillance period. Annual probabilities and risks incorporated from the literature were converted to monthly probabilities to accommodate for the model’s monthly cycle.

Table 1.

Model Inputs

Parameter	Base-case estimate	Range used in sensitivity analysis	Distribution for PSA	Reference No./source
Start age	60
Probabilities
All-cause mortality—general	Life table			18
All-cause mortality—stage 3b CKD	Life table			18,19
ceCT sensitivity	0.90	(0.85-0.95)	β	20
ceCT specificity	0.85	(0.85-0.90)	β	20
ceUS sensitivity	0.95	(0.90-1)	β	6
ceUS specificity	0.84	(0.79-0.89)	β	6
ceMRI sensitivity	0.92	(0.87-0.97)	β	6
ceMRI specificity	0.91	(0.86-0.96)	β	6
CT w/o contrast sensitivity	0.62	(0.57-0.67)	β	21
CT w/o contrast specificity	0.744	(0.694-0.794)	β	21
MRI w/o contrast sensitivity	0.629	(0.579-0.679)	β	22
MRI w/o contrast sensitivity	0.829	(0.779-0.879)	β	22
Monthly risk of progression of localized cancer on AS	0.0060293	(0.0045220-0.0075366)	β	23
Monthly risk of metastasis on AS	0.0002437	(0.0001827-0.0003046)	β	23
Monthly risk of local recurrence after NSS	0.0001668	(0.0001251-0.0002085)	β	23
Monthly risk of metastasis after NSS	0.0001443	(0.0001082-0.0001804)	β	23
Monthly risk of metastasis after local recurrence	0.0115381	(0.0086536-0.0144226)	β	23
Monthly risk of metastasis with positive margins	0.0025612	(0.0019209-0.0032014)	β	14
Monthly risk of mortality from localized RCC	0.0012087	(0.0009066-0.0015110)	β	SEER*Explorer
Monthly risk of mortality from metastatic RCC	0.0318954	(0.0239216-0.0398693)	β	SEER*Explorer
Mortality from NSS for 60-year-old	0.0066	(0.00495-0.00825)	β	24
Mortality from NSS for 80-year-old	0.021	(0.0158-0.0263)	β	25
Risk of non-R0 resection	0.057	(0.0428-0.0713)	β	14
Risk of surgical complications	0.18	(0.135-0.225)	β	23
Utilities
Healthy	1
Localized cancer	0.75	(0.94-1)	β	14, 26, 27
Metastatic cancer	0.66	(0.56-0.76)	β	15, 26
First month post NSS	0.75	(0.65-0.85)	β	23
4 Mos after NSS with major complications	0.50	(0.40- 0.60)	β	24
Costs, $
Cost of ceCT	449.21	(224.60-898.42)	γ	12, 13 CPT: 71460
Cost of first ceUS	528.81	(264.40-1,057.62)	γ	12, 13 CPT: 76978
Cost of additional ceUS	426	(213.00-852.00)	γ	12, 13 CPT: 76979
Cost of ceMRI	769.76	(384.88-1,539.52)	γ	12, 13 CPT: 74183
Cost of CT w/o contrast	283.04	(141.52-566.08)	γ	12, 13 CPT: 74150
Cost of MRI w/o contrast	462.97	(231.49-925.94)	γ	12, 13 CPT: 74181
Cost of NSS w/o major complications	16,360.38	(8,180.19-32,720.76)	γ	28
Cost of NSS with major complications	31,276.89	(15,638.45-62,553.78)	γ	28
Monthly costs for first year of metastatic cancer	8,324.78	(4,162.39-16,649.56)	γ	29
Monthly costs for subsequent years of metastatic cancer	2,538.19	(1,269.10-5,076.38)	γ	29

Abbreviations: AS, active surveillance; ce, contrast-enhanced; CKD, chronic kidney disease; CPT, Current Procedural Terminology code; CT, computed tomography; MRI, magnetic resonance imaging; NSS, nephron-sparing surgery; PSA, probabilistic sensitivity analysis; RCC, renal cell carcinoma; US, ultrasound; w/o, without.

Costs

The present study was performed from a third-party payer perspective. The model included direct medical costs of the respective imaging modalities, cost of NSS (with or without major complications), and associated costs for metastatic renal cancer treatment. Costs for the various imaging modalities were estimated using the Centers for Medicare & Medicaid Services Physician Fee Schedule and Medicare Procedure Lookup tool.^12,13 Renal cancer treatment costs were divided into first year treatment costs and subsequent year treatments costs. Costs were accrued from time of screening until death. Published cost estimates from prior years were converted to 2021 dollars using the Consumer Price Index (U.S. Bureau of Labor Statistics), and all costs were discounted an annual rate of 3%.

Utilities

Quality of life utility values relating to health, localized malignancy, and metastatic malignancy were incorporated in our model. Utility decrements for the first month post NSS, as well as for major surgical complications, were also applied. QALYs were discounted at an annual rate of 3%.

Sensitivity Analyses

For sensitivity analyses, the range of values was assumed to be ±25% of the base-case estimate. We performed 1-way deterministic sensitivity analyses by altering individual variables across a range of values to investigate the key parameters that had the greatest bearing on the outcomes of the model. In addition, a probabilistic sensitivity analysis (PSA) was performed to address model and parameter uncertainty. For PSAs, 10,000,000 reiterations of the model were run with each variable parameter being assigned a random value within a predefined distribution of values. β distributions were fitted for transition probabilities and utilities, while γ distributions were fitted for cost parameters.

Results

60-Year-old Analysis—Bosniak IIF

Our base-case analysis demonstrated that ceUS was the most cost-effective strategy for Bosniak IIF cysts, compared to ceCT and ceMRI (Table 2). When AS was performed 10%-50% of times, ceUS was the most effective strategy, but costlier than ceCT. When AS was increased to 100%, ceUS was undominated, meaning it was the most effective and cheapest strategy. For all rates of AS from 10%-100%, ceMRI was consistently dominated as it yielded the greatest costs and lowest effectiveness.

Table 2.

Base-case Analysis for 60-Year-old Patient—Bosniak IIF

	Strategy	Total cost, $	QALYs	ICERs
10% Active surveillance	ceCT	6,460.13	8.1897	–
	ceUS	6,576.51	8.1924	43,168.99
	ceMRI	7,703.56	8.1907	Dominated
20% Active surveillance	ceCT	6,200.30	8.1853	–
	ceUS	6,301.35	8.1882	34,762.32
	ceMRI	7,540.17	8.1837	Dominated
50% Active surveillance	ceCT	5,429.69	8.1707	–
	ceUS	5,475.65	8.1739	14,559.42
	ceMRI	7,055.89	8.1601	Dominated
100% Active surveillance	ceUS	4,098.25	8.1498	–
	ceCT	4,138.99	8.1460	Dominated
	ceMRI	6,250.88	8.1215	Dominated

Abbreviations: ceCT, contrast-enhanced computed tomography; ceMRI, contrast-enhanced magnetic resonance imaging; ceUS, contrast-enhanced ultrasound; ICER, incremental cost-effectiveness ratio; QALY, quality-adjusted life year.

Figure 2 shows the results of the 1-way sensitivity analysis. When ceUS was compared to ceCT, cost of ceCT, cost of ceUS and ceUS sensitivity and specificity had the largest effect on the model. Threshold analysis revealed that when cost of ceCT was $772.66, cost of additional ceUS was $290.53, ceUS specificity was 0.86, or cost of initial ceUS was $620.27, the ceUS strategy was no longer cost-effective due to exceeding the $100k WTP threshold. When comparing ceUS and ceMRI, the cost of ceMRI and the cost of additional ceUS had the greatest effect on the model. The threshold analysis revealed that when the cost of ceMRI exceeded $1,143.48, or the cost of additional ceUS exceeded $500, ceUS ceased to be cost-effective, compared to ceMRI.

Figure 2.

A, One-way sensitivity analysis results of contrast-enhanced ultrasound (ceUS) vs contrast-enhanced magnetic resonance imaging (ceMRI), Bosniak IIF. B, One-way sensitivity analysis results of ceUS vs contrast-enhanced computed tomography (ceCT), Bosniak IIF. C, One-way sensitivity analysis results of ceUS vs ceMRI, Bosniak III. D, One-way sensitivity analysis results of ceUS vs ceCT, Bosniak III.

The results of the PSA showed that after 10,000,000 iterations of the model, at a WTP of $100K per QALY, ceUS was cost-effective in 49.49% of iterations. CeCT and ceMRI were cost-effective in 46.77% and 3.74% of iterations, respectively.

60-Year-old Analysis—Bosniak III

The results of our base-case analysis for a 60-year-old with a Bosniak III cyst revealed that ceUS was the most cost-effective strategy for all rates of AS (10%-100%; Table 3). Similar to the Bosniak IIF analysis, ceCT was the always the cheapest strategy, while ceMRI was consistently the costliest with lowest effectiveness.

Table 3.

Base-case Analysis for 60-Year-old Patient—Bosniak III

	Strategy	Total cost, $	QALYs	ICERs
10% Active surveillance	ceCT	14,394.00	7.9141	–
	ceUS	14,647.06	7.9436	8,596.91
	ceMRI	15,072.11	7.8953	Dominated
20% Active surveillance	ceCT	13,988.96	7.8651	–
	ceUS	14,234.95	7.8941	8,491.19
	ceMRI	14,632.95	7.8286	Dominated
50% Active surveillance	ceCT	12,734.68	7.7099	–
	ceUS	12,959.54	7.7370	8,304.19
	ceMRI	13,268.73	7.6188	Dominated
100% Active surveillance	ceCT	10,818.87	7.4537	–
	ceUS	11,009.96	7.4782	7,784.41
	ceMRI	11,159.57	7.2732	Dominated

Abbreviations: ceCT, contrast-enhanced computed tomography; ceMRI, contrast-enhanced magnetic resonance imaging; ceUS, contrast-enhanced ultrasound; ICER, incremental cost-effectiveness ratio; QALY, quality-adjusted life year.

Figures 2, C and 2, D show the results of the 1-way sensitivity analysis. When ceUS and ceCT were compared, cost of ceCT, cost of initial and additional ceUS, and cost of ceUS specificity had the greatest effects on the model. However, even after varying these variables across their respective ranges, ceUS still remained cost-effective with an ICER below the $100K WTP threshold. When ceUS and ceMRI were compared, the cost of ceMRI, cost of additional ceUS, ceMRI specificity, and ceUS specificity most affected the outcomes of the model, but ceUS always remained cost-effective below the $100K WTP threshold.

Figure 3.

A, Probabilistic sensitivity analysis cost-effectiveness (CE) acceptability curve, Bosniak IIF. B, Probabilistic sensitivity analysis CE acceptability curve, Bosniak III. AS indicates active surveillance; ceCT, contrast-enhanced computed tomography; ceUS, contrast-enhanced ultrasound; MRI, magnetic resonance imaging.

The results of the PSA showed at a WTP of $100K per QALY, ceUS was the most cost-effective strategy for 80.81% of iterations (Figure 3, A). CeCT and ceMRI were the most cost-effective for 18.57% and 0.62% of iterations, respectively. Figure 3, B shows that when the WTP threshold varied from $20K to $300K, ceUS remained the most cost-effective strategy for the majority of times.

Discussion

In this study, we compared the cost-effectiveness of ceUS to CT and MRI (both with and without contrast) for the surveillance of Bosniak class IIF and III renal masses. Based on our results, ceUS was the most cost-effective surveillance option for 60-year-olds at the prescribed WTP of $100,000 per QALY.

Our study findings are similar to the conclusions of prior cost-effectiveness analyses that highlight the diagnostic strength and validity as a surveillance of ceUS for the management of indeterminate renal masses in 60-year-old patients.^14,15 While previous cost-effectiveness studies focused mostly in a European context, or without a direct comparison with ceCT, our study achieved similar results while also considering different rates of AS and patient profiles to more closely represent clinical reality. Contrary to the results of Spiesecke et al,¹⁵ we found that ceMRI resulted in lower total QALYs when compared to ceCT, as the rates of elective AS increased for patients who tested true- and false-positive. These results are due to our implementation of varying AS rates on positive findings, as well as due to the inferior specificity of ceCT that opportunistically provide curative surgeries for patients with malignancies on AS. Due to our model set-up, patients who have pretest malignancies and either initially tested as false-negative or did not elect initial surgery at the start of the model were treated with AS. During surveillance periods, these patients with malignancies would test “false-positive” more frequently on ceCT and ceUS strategies due to their lower specificities and be treated with curative NSS. While the ceUS and ceCT strategies may confer greater QALYs due to opportunistic curative surgeries, we implemented disutilities of surgery and costs of complications as to not overly favor strategies with lower specificities. While AS is an alternative to primary surgical intervention, imaging modalities with lower specificities present a challenge as patients can still receive unnecessary delayed surgeries after false-positive results during surveillance years. Due to the growing evidence of the indolent nature of small malignant renal masses, immediately treating them with surgery, despite low mortality rates,¹⁶ may still result in overtreatment and expose patients to unnecessary harm.⁹

Although ceUS has the lowest specificity of 0.84 compared to ceMRI and ceCT, its high effectiveness as an imaging modality is due to its superior sensitivity which reduces the number of false-negative cases.⁶ CeUS is also a suitable alternative to CT or MRI as it can avoid nephrotoxic contrast agents for patients with impaired kidney function.^1,17 While AS is endorsed at varying levels by different medical societies, currently, only 10%-20% of eligible patients undergo AS.⁸ Unfortunately, the current guidelines by the American Urological Association regarding AS only carries a moderate recommendation with an evidence strength of low certainty.⁷ We believe our study contributes to the medical literature validating the use of AS, especially with ceUS.

Our study does have several limitations. Cost-effectiveness analyses are simplifications of complex clinical paradigms that use data and estimates combined from multiple sources. While there are established data regarding the diagnostic performance of the individual imaging modalities for detecting malignancy, there is a paucity of data regarding their repeated accuracy while surveilling small renal masses over time. We assumed the respective diagnostic specificities and sensitivities of each modality would remain constant throughout the duration of surveillance periods, which may lead to an overestimation of detected malignancies. Another limitation of our study regarded the rate of metastasis for patients under AS. The minority of patients in AS who could have consecutive false-negative results during subsequent surveillance periods should theoretically have an altered cancer growth kinetic, however in the model these patients had the same rate of metastasis as those who only had 1 false-negative result. Furthermore, patients in metastatic states of cancer remained in that state and were assumed incurable for model simplicity. Our model also used U.S. Medicare cost data to estimate cost-effectiveness, and payment systems and health care systems may differ between other countries and the U.S.

Conclusion

CeUS is a viable and cost-effective option in the AS of Bosniak class IIF and III renal cysts. Even after varying the rates of AS usage, ceUS was robust and remained the most cost-effective strategy.