The Agency for Healthcare Research and Quality (AHRQ) defines a patient registry as “an organized system that uses observational study methods to collect uniform data (clinical and other) to evaluate specified outcomes for a population defined by a particular disease, condition, or exposure, and that serves one or more predetermined scientific, clinical, or policy purposes” . Registry use occurs on a variety of levels, ranging from clinic logs to rare disease entities to international disease databases. Data collected often include demographic information such as sex and age, medical history, diagnostic information, procedure and device specifics, and clinical outcomes. The uniform collection of these data creates a better understanding of practice characteristics, treatments, diseases, and outcomes. In their most practical use, registries can help identify which patients have a certain condition or disease, or follow patients receiving certain devices or treatments. The RSSearch™ Registry fulfils this ideal for patients treated with SRS and SBRT.
Multi-institutional, observational patient registries can be powerful tools to describe disease presentation, patterns of care, treatment effectiveness, safety and quality of care in a real-world setting [33–35]. The Society of Thoracic Surgery Cardiac Surgery National Database, which began in 1989 and includes over five million patient records, has been used to report treatment patterns, support quality improvement initiatives, develop clinical treatment guidelines, support comparative effectiveness studies, government collaborations, regulatory compliance and reimbursement strategies for cardiac and thoracic patients [34, 41–43]. The American Society of Breast Surgeons (ASBS) MammoSite™ Registry began in 2002, as a manufacturer-sponsored registry developed to provide information on the optimal use of the device in clinical practice . The ASBS assumed responsibility of the database in 2003, and have reported outcomes from the registry demonstrating efficacy, safety and improved quality of life for early-stage breast cancer patients treated with breast-conserving therapy.[44, 45] The ASBS have been consistently following-up and reporting on the use of MammoSite™ as the registry matures and continues to report on clinical outcomes of partial breast irradiation even before formal prospective data from randomized clinical trials becomes available [46, 47]. Clinical outcome data from registries have also been used to determine coverage (payment) policies. The Centers for Medicare & Medicaid Services expanded its coverage for PET scans in diagnosis of certain cancers because of information obtained through the National Oncologic PET Registry . These reports demonstrate the potential impact registries can provide to the medical community, patient care and treatment practices.
In this report, we demonstrate that patterns of care for SRS/SBRT can be collected in a multi-institutional database and patterns of care research can be performed. We describe the initial patient demographics, lesion characteristics and SRS/SBRT treatment for 11,457 patients enrolled in the RSSearch™ Registry. To our knowledge, this is the largest aggregate report of SRS/SBRT-treated patients in a published registry. As one might expect, lesions in the brain/cranial nerve/spinal cord were the single most prevalent treatment location in RSSearch™, which correlates with the long history of SRS experience for intracranial lesions. The median SRS dose and fractionation schedule for brain lesions reported in RSSearch™ was 24 Gy delivered in 1 fraction. A single fraction of 24 Gy is the recommended dose used in to treat 1–3 metastatic brain lesions up to 2 cm as described in RTOG 9508 and RTOG 9005 [37, 48] and SRS delivered as a single fraction (15–24 Gy) to individual intracranial lesions has been established as a safe alternative to surgical resection [39, 49]. In a subset analysis, we reported a median overall survival rate of 11 months for patients with brain metastases with a good performance score (KPS > 70) compared to 4 months for patients with KPS ≤ 70. Outcomes for patients with brain metastases are generally poor with a median survival following whole brain radiation therapy (WBRT) alone of 3–6 months [37, 39, 50]. The addition of SRS following WBRT and SRS alone have improved overall survival and local control rates. Rades et al. showed a 13 month overall survival rate for SRS alone compared to 7 months for WBRT . One-year local control rate for SRS alone was 64% compared to 26% for WBRT. Li et al. compared WBRT vs. SRS vs. WBRT plus SRS and showed a median survival rate of 5.7 months, 9.3 months and 10.6 months, respectively . KPS has been shown to be a predictor of survival for patients with brain metastases . Using recursive partitioning analysis of three RTOG trials, the best survival (median 7.1 months) was observed in patients < 65 years, KPS of at least 70 and a controlled primary tumor with brain as the only site of metastases. The worst survival (median 2.3 months) was seen in patients with KPS < 70. In our study, overall survival was significantly greater in patients with KPS > 70 compared to KPS ≤ 70, correlating with previous reports.
When using high doses of radiation with SRS/SBRT, there is critical importance to minimize the volume of normal tissues receiving high dose per fraction in order to protect normal tissues from adverse radiation effects. The American Association of Physicists in Medicine (AAPM) task group and others have defined normal dose limits for organs at risk for SBRT from published literature and randomized clinical trials [53–55]. We calculated the median maximum point dose for organs at risk from the treatment plans in RSSearch™ in order to compare to reported treatment guidelines. For the brainstem, eye/lens, optic chiasm and optic nerve, the median max point dose in RSSearch™ was 5, 0.4, 2 and 1 Gy, respectively, which is below the recommended max point dose ranges of 15–31 Gy for the brainstem, 2–3 Gy for the eye/lens, 10–15 Gy for optic chiasm and 8–15 Gy for optic nerve when delivered in 1 fraction [53, 54]. Furthermore, with automated data entry from treatment platforms, more stringent dose-volume histogram analysis of all targets and organs at risk can be performed. These results indicate that most participating centers treating intracranial lesions adhere to recommended guidelines for tolerance limits for normal organs.
In this analysis, we also examine patient demographics, treatment practices and outcomes of liver metastases, a common extracranial treatment location. Currently, there is no consensus for the treatment of primary liver tumors and liver metastases with radiation. In 2012, the Liver Cancer Workgroup of the Third International Consensus on Metastases Workshop at the 2010 American Society for Radiation Oncology meeting published an international survey on the status of radiation therapy of liver metastases . The survey indicated there was a 54% increase in the average number of liver referrals over the past 5 years and the majority of referrals were for SBRT. No uniform SBRT treatment dose was identified and there was a wide variation of treatment regimens which were dependent on whether the treatment objective was curative or palliation. The group concluded there is a need for prospective studies and registries for comparison of treatment regimens and identification of parameters to optimize patient selection. Interestingly, liver is the third most common extracranial SBRT treatment location in RSSearch™ and SBRT is being used as both a palliative and curative treatment option for liver lesions. The SBRT dose ranged from 10 – 60 Gy delivered over 1–5 fractions, with the median SBRT dose of 45 Gy delivered in 3 fractions. The median OS was 22 months and OS 6, 12 and 18 months were 91%, 69% and 60%, respectively. LPFS at 6, 12 and 18 months was 94%, 76% and 70%, respectively. In our analysis of this patient group (n = 174) with a wide variety of underlying primary tumors, different systemic and local treatments, lesion sizes and dose/fractionation schedules the OS and local control rates were within the range of published reports[57, 58]. Future studies are planned to examine prognostic factors and the effect of dose/fractionation schedules on OS and local control from the RSSearch™ dataset.
The most prevalent extracranial treatment site in RSSearch™ was lung/bronchus. This correlates with the AHRQ 2011 report which described the current state of SBRT as an emerging technology for the treatment of solid malignant tumors and identified lung/thorax as the most common site treated with SBRT . SBRT is now considered a standard treatment option for medically inoperable patients through prospective, multi-institutional trials [60, 61]. There is a wide variety of SBRT dose and fractionation schedules reported for lung lesions which range from single fractions of 19–34 Gy [62–64] to hypofractionated schedules of 50–66 Gy delivered in 3–5 fractions [13, 65–67]. In RSSearch™, the median SBRT dose to lung lesions was 54 Gy delivered in 3 fractions, which is in line with other published reports. We also reported the median maximum point dose to the esophagus, heart, trachea/bronchus and spinal cord reported in RSSearch™ and demonstrated that the median values all fell below the recommendations for normal dose limits described by the AAPM Task Group and RTOG 0618 . It is reassuring to note that these tolerance doses to organs at risk were more that met without compromising on outcomes. The outcomes reported in Figures 2 and 3 from this study are comparable for this population to other studies. Moreover, these results may be more generalizable than single institution or prospective controlled studies. The results also suggest that the majority of centers participating in RSSearch™ are following the standardized treatment guidelines, published reports and protocols for SBRT for the treatment of lung cancer. Future studies are planned to obtain the entire treatment plan to correlate clinical outcomes with DVHs as well as analyze and report on the treatment management practices and clinical outcomes of lung cancer patients in RSSearch™.
The second most common extracranial site in RSSearch™ was the prostate. SBRT treatment of early stage organ-confined prostate cancer has become increasingly popular, as initial studies have demonstrated excellent biochemical control rates with very minimal toxicity [16, 19, 68]. In April 2013, ASTRO published their SBRT Model Policy stating that SBRT was an appropriate treatment option for select patients with low to intermediate risk prostate cancer and should be included in coverage (payment) policies . Several studies have shown 5-year biochemical disease-free survival rates of 90-97% when using doses of 35 – 36.25 Gy when delivered in 5 fractions for low and intermediate risk disease. The median dose in RSSearch™ was 36.25 Gy in 5 fractions. The median max point dose to the urethra, rectum, penile bulb and femoral heads was 43, 38, 27 and 14 Gy, respectively, and fell below the published recommended normal dose tolerances (max dose of 35 – 47 Gy for urethra, 21 – 38 Gy for rectum, 50 Gy for penile bulb, and 30 Gy for femoral heads) [53, 54]. The median max point dose for the bladder was 41 Gy and slightly higher than published recommendations ranging from 24 – 25 Gy. Future studies from RSSearch™ will be conducted to compare dose volume histograms of target and organs at risk to correlate clinical outcomes and toxicities.
SBRT has been evaluated for the treatment of locally advanced pancreatic with mixed results [20–23]. Variations in SBRT dose and fractionation schedules in conjunction with different chemotherapeutic agents and schedules have made the interpretation of clinical results and optimization of patient selection challenging. SBRT doses have ranged from 15 – 25 Gy in a single fraction to 24 – 45 Gy in 3 fractions with concurrent gemcitibine [20, 21, 23]. The University of Texas Southwest Medical Center, Dallas, TX have initiated a Phase I trial where patients with surgically resectable pancreatic cancer receive 30 Gy SBRT in 1 fraction to the surgically inaccessible at-risk margin prior to surgery, in effort to reduce local failure following surgery . These studies indicate that SBRT can be an effective treatment option for pancreatic cancer, but its optimal use remains to be determined. The median SBRT dose for pancreatic lesions reported RSSearch™ was 29 Gy delivered in 3 fractions. Future studies will be conducted to investigate variations in treatment management practices and clinical outcomes for pancreatic cancer patients enrolled in RSSearch™.
The RSSearch™ registry was feasible as it was nurtured by a SRS/SBRT specialty society. Medical specialty societies are organizations that represent networks of physicians. These organizations often exist to provide services to their members in the areas of advocacy, education, and practice management. Based on their close interaction with their members and their national and international reach, medical specialty societies are uniquely positioned to collect and manage data related to the type of care their members provide. Medical specialty societies are able to connect with the members directly at meetings and symposia, and by using direct-to-member correspondences. The Radiosurgery Society, an independent non-profit organization of professionals dedicated to SRS/SBRT will foster and champion the maintenance and future success of the registry.