Significant variability in 10-year cumulative radiation exposure incurred on different surveillance regimens after surgery for pT1 renal cancers: yet another reason to standardize protocols?

Authors


Abstract

What's known on the subject? and What does the study add?

  • The topic of radiation safety has been hotly debated not only in the mainstream media, but also in the urological literature. Radiation exposure has been examined in urological diseases such as testicular cancer and urinary stone disease, with resultant recommendations for modifying surveillance imaging.
  • Radiation risk with respect to surveillance regimens after RCC surgery has yet to be examined. We consider this largely to be a result of RCC typically affecting older patients in whom cumulative radiation exposure may be less of a consideration. However, current population data emphasize that RCC diagnosis and therapy have an increasing impact upon younger patients with a longer life expectancy after treatment. Therefore, radiation considerations in this cohort of patients may be significant.

Objective

  • To determine the 10-year cumulative radiation exposure incurred on different surveillance imaging protocols after surgery for pT1 renal cell carcinoma (RCC).

Materials and Methods

  • The PubMed database was queried for surveillance protocols after surgery for RCC.
  • There were two index lesions that were selected: (i) pT1a 3 cm, Fuhrman 2, clear cell and (ii) pT1b 5 cm, Fuhrman 3, clear cell.
  • Exposure for single-phase chest computed tomography (CT), abdominal CT and chest X-ray were 7, 8 and 0.1 mSV, respectively.
  • Calculations assumed biphasic CT scans, negative surgical margins and an Eastern Cooperative Oncology Group status of ≤1.

Results

  • In total, 12 published surveillance regimens were identified.
  • For the first lesion (pT1a, clear cell, Fuhrman 2), we observed significant variability in the proposed regimens, ranging from no imaging to several CT scans of both chest and abdomen.
  • Cumulative incurred radiation exposure for this index patient was in the range 0–102 mSv (mean, 34 mSv).
  • When considering the second tumour (pT1b, clear cell, Fuhrman 3), all studies recommended some form of follow-up imaging, although regimens once again varied from annual chest X-ray to multiple CT scans of chest and abdomen.
  • Cumulative incurred radiation exposure in this scenario was in the range 0.5–450 mSv (mean, 89 mSV).

Conclusions

  • Surveillance protocols after surgery for early-stage RCC result in widely divergent levels of radiation exposure.
  • Such considerations are increasingly paramount given concerns of radiation-induced secondary malignancies and present another reason to standardize follow-up protocols.
Abbreviations
CXR

chest X-ray

ECOG

Eastern Cooperative Oncology Group

SRM

small renal mass

Introduction

The increased use of abdominal imaging (CT and/or MRI) has contributed to a greater detection of incidental small renal masses (SRMs) [1, 2]. When compared with clinically detectable tumours, these asymptomatic SRMs are smaller in size, present at an earlier stage and are almost universally confined to the kidney at diagnosis [2]. Furthermore, population-based registries, such as the Surveillance, Epidemiology and End Results database, indicate that 25% of patients with newly-diagnosed renal tumours will be aged <55 years [3]. With a projected anticipated life expectancy of more than 20 years [4], such patients will probably undergo surgical therapy with resultant follow-up for disease recurrence via surveillance protocols [5]. Indeed, contemporary studies report durable long-term outcomes after surgery for early-stage RCC [6]. However, ionizing radiation exposure incurred on surveillance protocols after partial or radical nephrectomy may be significant in younger patients with earlier stage tumours followed for a longer duration.

Concern with radiation exposure has been explored with respect to urinary lithiasis [7] and testicular cancer [8]. Such studies implicate significant cumulative radiation exposure in patients after therapy and advocate vigilance given public concerns and the risk of secondary malignancy. To date, limited information is available regarding renal malignancies. Current National Comprehensive Cancer Network clinical practice guidelines for kidney cancer suggest performing abdominal and chest CT scans 2–6 months after surgery and then as clinically indicated [9]. Multiple other proposed surveillance algorithms similarly lack uniformity. The lack of a standardized protocol for RCC implies that, depending on the preferred protocol of the institution or individual urologist, patients would undergo a variable number of imaging studies with differing levels of radiation exposure. To better quantify such potential differences, we evaluated the 10-year cumulative radiation exposure incurred on different published surveillance regimens after surgery for pT1 RCC.

Materials and Methods

Surveillance Protocol Selection

The National Institute of Health PubMed database (http://www.pubmed.gov) was queried to identify surveillance protocols after surgery for early-stage RCC. The search terms employed to identify studies published in the English language included ‘renal cell carcinoma’, ‘kidney neoplasm’, ‘partial nephrectomy’, ‘radical nephrectomy’, ‘follow-up’, ‘surveillance’, ‘imaging’ and ‘recurrence’. Studies included retrospective or prospective single-institution series or review articles. If multiple publications sequentially originated from a single institution, the most contemporary protocol was used in the present analysis unless proposed regimens varied significantly. In the latter instance, both protocols were considered. Follow-up regimens proposed in practice guidelines and/or editorial comments were not included in the present study. Only extirpative surgical therapy via partial or radical nephrectomy was considered for analysis.

Many contemporary studies have shown equivalent oncological efficacy for partial and radical nephrectomy [5]. Additionally, most contemporary published studies [10-13] do not distinguish between the nephrectomy type but, instead, base risk assessment exclusively on pathological characteristics. Therefore, calculated radiation exposure was solely based on pathological characteristics.

Lesion Selection

Clear cell RCC was selected as the histological subtype of interest because it represents the most common renal malignancy. Furthermore, the index lesions were further selected to represent two relatively common pathological tumour stages and grades encountered in therapy for early-stage kidney cancer. The two index lesions were defined as: (i) pT1a 3 cm, Fuhrman grade 2, clear cell RCC and (ii) pT1b 5 cm, Fuhrman grade 3, clear cell RCC. The index lesions were chosen based on three observations. First, Hollingsworth et al. [2] noted that the increased incidence of kidney cancer and resultant surgery is attributable to the detection of more T1 renal tumours. Second, protocols often vary surveillance strategies based on pT1a vs pT1b. Lastly, we wanted to examine radiation exposure in early-stage kidney cancer, for which the risk of recurrence is low and the risk of radiation exposure may be paramount.

Estimating Radiation Exposure

Radiation exposure from diagnostic imaging is typically reported in Gray (Gy) units, which are defined as the energy absorbed per unit of body mass (J/kg). However, because each body organ absorbs radiation differently, the effective dose, millisieverts (mSv), is used to account for the biological sensitivity of different organs. It represents an estimate of the overall harm.

The reported effective radiation dose (mSv) varies widely in the literature for each of the examinations [14]. In the present study, the radiation exposures for a single phase 5-mm thickness slice chest CT, abdominal CT, and chest X-ray (CXR) were determined to be 7, 8 and 0.1 mSv, respectively. These values were based on study by Metter et al. [14], who reviewed over 150 published articles on radiation dosimetry to determine the respective values. We further compared these values with other references in the radiological literature [15], and noted that there were negligible differences. Therefore, these referent exposures were considered to be both representative and accurate estimations of radiation exposure attributable to radiation studies. Using effective radiation dose standards for each of the imaging modalities, we calculated the total effective radiation dose (mSv) exposure for each protocol. To create uniformity within the published guidelines, the calculated exposures assumed that all CT scans were two phase (non-contrast + iodinated contrast), all surgical margins were negative and performance status was Eastern Cooperative Oncology Group (ECOG) 0 or 1.

Results

Twelve surveillance regimens originating from 11 institutions over a period of 15 years (1994–2009) were included. Table 1 summarizes the surveillance protocols delineated in different regimens over a 10-year period with an accompanying calculation of the total effective radiation dose exposure. Of note, several studies only outline a 5-year surveillance plan, with no recommended imaging thereafter [16-18].

Table 1. Surveillance protocols and cumulative 10-year radiation incurred after surgery for specified index lesions
Reference3 cm, grade 2, clear cell5 cm, grade 3, clear cell
ProtocolExposure (mSv)ProtocolExposure (mSv)
  1. Abd, abdominal; CXR, chest X-ray.
  2. aSiddiqui et al. [10] also recommended five abdominal ultrasonographic examinations at years 3, 4, 6, 8 and 9.
  3. bBoth Klatte et al. [12] and Lam et al. [11] recommended alternating CXR with chest CT after 3 years of follow-up for the second index lesion.
  4. cFive-year surveillance protocols.
  5. dThe study by Gofrit et al. [19] has a discrepancy between the abstract and the discussion section regarding the surveillance protocol.
Siddiqui et al. [10]a6 Abd CT96

6 Abd CT

12 CXR

97.2
Klatte et al. [12]b5 Chest CT70

5 Abd CT

8 Chest CT

2 CXR

192.2
Skolarikos et al. [13]13 CXR1.313 CXR1.3
Lam et al. [11]b

2 Abd CT

5 Chest CT

102

5 Abd CT

8 Chest CT

2 CXR

192.2
Stephenson et al. [27]10 CXR1.010 CXR1.0
Mickisch et al. [17]c8 CXR0.88 CXR0.8
Gofrit et al. [19]dNone0

15 Abd CT

15 Chest CT

450
Ljungberg et al. [20]None09 CXR0.9
Levy et al. [16]c5 CXR0.55 CXR0.5
Hafez et al. [23]

10 CXR

5 Abd CT

81

10 CXR

5 Abd CT

81
Sandock et al. [22]13 CXR1.313 CXR1.3
Montie [18]c

10 CXR

3 Abd CT

49

10 CXR

3 Abd CT

49
Mean33.689
Median1.325.2
Range0–1020.5–450

For the first index lesion (pT1a, clear cell, Fuhrman 2), we observed significant variability in proposed regimens. For example, Gofrit et al. [19] and Ljungberg et al. [20] did not recommend radiographic follow-up for these lesions, citing minimal risk of recurrent disease. Conversely, other protocols advocate for a more rigorous surveillance strategy. Specifically, Siddiqui et al. [10] recommended six abdominal CT scans (total 96 mSv), whereas Lam et al. [11] proposed two abdominal CTs and five chest CTs (total 102 mSv) over a 10-year period. Of note, within a relatively short 3-year period, the University of California Los Angeles Integrated Staging System protocol was updated [12] to not include any abdominal imaging, thereby lowering the exposure to 70 mSv. When considering all studies, the cumulative incurred radiation exposure for this index lesion was in the range 0–102 mSv (mean, 33.6 mSv; median, 1.3 mSV).

With respect to the second index tumour (pT1b, clear cell, Fuhrman 3), because it represents a higher recurrence risk, all studies recommended some form of follow-up imaging. Nonetheless, the regimens once again varied from annual CXR to multiple CT scans of the chest and abdomen. Several protocols, such as those proposed by Levy et al. [16] (five CXR, 0.5 mSv) and Mickisch et al. [17] (eight CXR, 0.8 mSv), favoured the use of chest radiography as the sole imaging follow-up for these patients. Conversely, a few protocols advocate more rigorous surveillance. Specifically, the Mayo clinic protocol from Siddiqui et al. [10] proposes six abdominal CT scans and 12 CXR (97.2 mSV), whereas the University of California Los Angeles Integrated Staging System protocol from Lam et al. [11] and Klatte et al. [12] advocates for a combination of five abdominal CT, eight chest CT and two CXR (192 mSv). Collectively, the cumulative incurred radiation exposure in this scenario was in the range 0.5–450 mSv (mean, 89 mSv; median, 25.2 mSv).

Discussion

Advances in medical imaging technology have contributed to an increased diagnosis of SRMs. The mean age at presentation for these renal masses is 64 years of age, with 25% of diagnoses occurring in patients aged <55 years [3]. The clinical management of these lesions must weigh the risk of progression against the potential morbidity of treatment. Contemporary guidelines show that surgery in the form of partial or radical nephrectomy is the standard of care owing to durable oncological outcomes with acceptable perioperative morbidity [5, 6, 21]. After surgical treatment, patients are typically enrolled in surveillance regimes that incorporate imaging studies at certain intervals of follow-up. More regimes are incorporating pathological tumour characteristic to better tailor the follow-up schedule. Such practice is attributable to studies reporting that disease relapse and cancer-specific survival are highly related to the pathological features of the primary tumour [22-24]. Furthermore, early detection of disease relapse allows for intervention by either salvage surgery or systemic-targeted immunotherapy geared to improve survival [12, 13].

In the present study, we have performed a comprehensive review of the published surveillance protocols after surgery for localized RCC. To avoid redundancy, we only referenced the earliest protocol for summary when multiple identical surveillance algorithms originating from the same institution existed [25]. In certain instances, if subsequent protocols deviated significantly from the original proposed regimen, both algorithms were included for consideration [11, 12]. Our search also identified that some institutions, such as the Memorial Sloan Kettering Cancer Center [26], employed nomograms aiming to determine the risk of cancer recurrence after RCC surgery. We elected to not include this study and similar publications on risk stratification because many failed to identify the type, duration and frequency of specific imaging studies.

In 1994, Montie [18] proposed a fairly homogenous surveillance schedule after RCC surgery, showing that early detection results in improved treatment options. Sandock et al. [22] refined this follow-up algorithm by defining a patient population that requires more intensive surveillance based on the pathological stage of their cancer. Subsequently, new protocols have increasingly attempted to incorporate the TNM staging system, as well as other clinical and pathological features, to optimize surveillance. Nonetheless, as shown in Table 1, there remains significant variability in the proposed regimes. Many studies classify our first index renal tumour (pT1a, Fuhrman grade 2, clear cell RCC) as indicating low risk for recurrence or progression after surgery. Despite this designation, there exists significant discrepancy with the type (plain radiography vs CT scan) and frequency of surveillance imaging studies. For example, several protocols advocate no imaging studies [19, 20], some suggest periodic chest radiography [13, 16, 17, 22, 27], whereas others call for a more intensive follow-up schedule involving multiple CT scans [10-12, 18, 23]. This variability with respect to the proposed regimens translates to an mean (range) effective radiation dose exposure (cumulative at 10 years) of 33.6 mSv (0–102) mSv.

A similar diverse pattern can be appreciated when considering the second index lesion (pT1b, Fuhrman grade 3, clear cell RCC). Certain studies classify this tumour as indicating low risk, whereas others consider it to indicate an intermediate risk for recurrence. Accordingly, follow-up protocols are more divergent. Specifically, although all studies advocate some form of imaging, certain protocols proposed only periodic CXR [13, 16, 17, 20, 22, 27], whereas others regularly incorporate CT scans in their surveillance algorithms [10-12, 23]. The mean (range) total effective radiation dose exposure over a 10-year follow-up in this case is 89 (0.5–450) mSv.

The topic of radiation safety has been hotly debated not only in the mainstream media, but also in the urological literature. Radiation exposure has been examined in urological diseases such as testicular cancer [8] and urinary stone disease [7]; however, it has not been examined in RCC. We consider that this is largely a result of RCC typically affecting older patients in whom cumulative radiation exposure may be less of a consideration. However, the current population data show that RCC diagnosis and therapy is increasingly impacting younger patients with a longer life expectancy after treatment [3]. Currently, most of the data supporting claims of increased cancer risk from low-dose radiation exposure are extrapolated from publications on outcomes of the survivors of the atomic bombing at Nagasaki and Hiroshima [28]. By combining the atomic bomb data with the analysis of a large-scale study of radiation works in the nuclear industry [29], Brenner et al. [15] assessed the cancer risk from radiation exposure from CT by estimating the organ dose involved and applying organ-specific cancer incidence. They concluded that up to 2% of all cancers in the USA may be attributed to the radiation from CT studies [15]. The BEIR VII (Biologic Effect of Ionizing Radiation) report from the National Academy of Sciences concludes that there is a linear dose–response relationship between exposure to ionizing radiation and the development of solid cancers. Their lifetime risk model predicts that, with an exposure of 100 mSv, one out of 100 individuals would be expected to develop cancer (solid vs leukaemia) [30]. Lastly, based on data from the radiation effects research foundation [28], it is estimated that the attributable risk of developing a solid tumour cancer from radiation exposure is 7.6% for an exposure in the range 100–200 mSv and 15.7% for an exposure in the range 200–500 mSv. These studies stress the need for achieving a balance between the risk of radiation exposure and the risk of delaying recurrence detection. A review of our data (Table 1) emphasizes that some protocols incur cumulative radiation doses well in excess of 100 mSv, thereby potentially increasing the risk of secondary malignancy substantially over the baseline population.

The present study has several limitations. First, to summarize each proposed surveillance regimen, we elected to have a broad classification based on the TNM stage and Fuhrman nuclear grade of each protocol. We also assumed an ECOG performance status of 0 or 1 and negative surgical margins. Of interest, only the University of California Los Angeles protocols [11, 12] incorporated the ECOG performance status when evaluating the recurrence risk. We did not focus on specific tumour characteristics such as tumour necrosis, DNA ploidy and microvascular invasion, which played a role in calculating the risk factors in protocols from Siddiqui et al. [10], Ljungberg et al. [20] and Sorbellini et al. [26]. Because most of the reviewed protocols utilized the TNM stage and Fuhrman grade without specifying other tumour characteristics, we consider that our data analysis represents an accurate portrayal of the decision algorithm broadly utilized by treating urologists. Second, we intentionally refrained from extrapolating cancer risk attributable to radiation exposure. Although we know that atomic bomb survivors who were exposed to high radiation doses have an increased risk of malignancy, it is unclear whether chronic lower radiation exposure over a longer period of time translates to the same risk of cancer. This remains a topic for future exploration.

Despite these limitations, we hope to show that there is tremendous variability in the surveillance protocols after surgery for early-stage RCC. These disparate regimes have widely divergent levels of effective radiation exposure to the patient, which could result in the development of secondary malignancies. We acknowledge that the radiation exposure risk only encompasses one aspect of the decision-making process. As physicians, we need to balance patient concerns and legal ramifications regarding the missed detection of cancer recurrence with the risk of excess radiation exposure. With more attention focused on the risk of radiation exposure in the news media, there will be probably be greater discussion between patients and providers regarding the risks and benefits of subsequent imaging studies. It is our hope that highlighting the radiation exposure incurred on surveillance regimens may further encourage the standardization of protocols, allowing for the early detection of recurrence and, at the same time, minimizing the amount of radiation exposure. Furthermore, such information may encourage the increased use of radiation-limiting imaging strategies, such as low dose CT, MRI and/or ultrasonography, which can provide imaging information at a fraction of the incurred radiation dose.

In conclusion, a review of 12 surveillance protocols after surgery for early-stage RCC uncovers disparate regimens with widely divergent levels of radiation exposure to patients. Such considerations are increasingly paramount given contemporary concerns of radiation-induced secondary malignancies and present another potential reason to standardize follow-up protocols.

Conflict of Interest

None declared.

Ancillary