California Cancer Registry 2010 Gastrointestinal Stromal Tumor (GIST) Recoding Audit CCR Project ID #300.66 Final Report Revised July 11, 2011 The Data Standards and Quality Control Unit (DSQC) of the California Cancer Registry (CCR) initiated an audit of Gastrointestinal Stromal Tumors (GIST) cases due to an identified issue on a Patterns of Care (POC) Study performed in 2010. The 2010 POC Study included malignant GIST as one of the sites in the study. During the data collection portion of the study, regional staff performing the medical record review noted that there were no malignant terms in the medical records at one hospital. Investigation into why these cases were coded malignant without documented malignant terminology revealed that the physicians at this facility consider all GISTs malignant and therefore there is no need to document GIST tumors as malignant. The Surveillance Epidemiology and End Results (SEER) Program has stated that the only GIST cases that are reportable are those cases in which the pathology report states “malignant GIST” or clinician states “malignant GIST”. If one of these two elements is not in place, the case is deemed not reportable by SEER. Internet research as well as review of the American College of Surgeons (ACoS) Inquiry and Response (I&R) System indicates that GIST, NOS (no statement of malignancy) is not assumed to be benign by many physicians. Mitotic activity, aggressive behavior, the presence or absence of metastatic disease, and regional lymph node involvement are all factors that appear to be influencing abstractors’ decision when determining reportability at the time of original case completion. It was decided to review the entire California Cancer Registry (CCR) database to determine the extent of this issue. Audit Sampling A data miner query was written that identified all GIST cases in the Eureka database. The sample was then randomized by region and diagnosis year. This produced a total of 2,908 GIST cases. There were 744 GIST cases diagnosed prior to 2001 and 2,164 GIST cases diagnosed 2001 to present. The most recent cases for each region were reviewed as part of this audit. A total of 320 cases were recoded, 40 cases from each region. Audit Performance Methodology Each case was recoded based on the text documentation submitted on the abstract. The original code must match the recode by the auditor. A discrepancy resulted if these codes did not match. The audit was performed utilizing the peer review method. Each case was audited blindly and independently by two separate auditors. When both the primary and secondary auditors completed their assigned region, the primary auditor then compared the discrepancies between 1 California Cancer Registry 2010 Gastrointestinal Stromal Tumor (GIST) Recoding Audit CCR Project ID #300.66 Final Report Revised July 11, 2011 the auditors. The two auditors then reconciled their differences and discussed any issues that they may have discovered during the audit process. Data Items Audited Behavior was the only data item recoded and was based on the text included for each case. If the elements that SEER required for a case to be considered reportable were present, then the behavior was coded to /3 (malignant), otherwise, it was recoded /1 (borderline). Audit Results There were several cases that did not use the word “malignant” but did use the term “metastatic” or there was clear metastatic disease. While it is widely understood that metastatic disease would indicate that the tumor in question is in fact malignant, SEER’s interpretation of what constitutes a malignant GIST has caused significant hesitation in coding a GIST to a behavior /3. Furthermore, there were other factors the auditors discovered that may indicate a malignant GIST such as invasion, mitotic rate, chemotherapy treatment administered, or terminology such as “high malignant potential” or “malignant risk.” These factors were documented in published articles in various medical journals. There were 16 different examples of ambiguous terminology and vague text from real cases sent to SEER via SEER SINQ (20100014) for their input. All but one was returned as “not reportable” and clearly supported SEER’s finite definition of what constitutes a diagnosis of malignant GIST. See Appendix A for the details of SEER SINQ 20100014, 20091021, and 20021151. Discrepancies Of the 320 cases audited in this audit, 211 (65.9%) met SEER’s requirements for reportability. One-hundred nine (109 or 34.1%) were considered not reportable according to SEER’s guidelines. This is a significant result. Discrepancies by Regional Registry Region 6 had the least number of discrepancies with 6 (5.5%) discrepancies identified. Region 8 had the highest number, with 23 (21.1%) discrepancies. Region 3 had the second highest number, with 16 (14.7%). This was followed by Region 5 with 16 (14.7%), Region 10 with 16 (14.7%), and Region 9 with 12 (11%) discrepancies. Region 2 and Region 4 had 10 discrepancies (9.2%) each. Graph 1 illustrates the distribution of the discrepancies across the regional registries. 2 California Cancer Registry 2010 Gastrointestinal Stromal Tumor (GIST) Recoding Audit CCR Project ID #300.66 Final Report Revised July 11, 2011 Conclusions There is a clear difference between SEER’s reportability requirement and current clinical medical practice. It is apparent that at some reporting facilities, current clinical medical practice is the exact opposite of SEER’s GIST reporting requirement. Some physicians consider all GISTs to be malignant and do not necessarily document the terms “malignant” and “GIST” in the same sentence. If the CCR applied SEER’s current reporting guideline, a significant number of the GIST cases in the Eureka database would need to be deleted. This is a significant impact to the CCR’s database. Until the CCR receives further direction from SEER in response to the audit findings, it is recommended that these cases remain in the database. Recommendations 1. Inform SEER of the results of this audit along with all researched documents including medical journal articles documenting current medical practice. 2. Share these results with the Cancer Surveillance Research Branch (CSRB) to inform them of the potential disparity of GIST cases in the CCR database. 3. Confirm with SEER that a significant number of GIST cases in the Eureka database will be deleted due to their reportability criteria of GIST cases. 4. Write an article outlining audit findings in the CCR newsletter, “The Criterion.” 5. Include information regarding clarifications received from SEER in “The Criterion” and in training modules. 3 California Cancer Registry 2010 Gastrointestinal Stromal Tumor (GIST) Recoding Audit CCR Project ID #300.66 Final Report Revised July 11, 2011 Appendix A Question ID: 20100014 Status: Final Question: Reportability: Per SINQ 20091021 and 20021151, GIST cases are not reportable unless they are stated to be malignant. A Pathologist or clinician must confirm the diagnosis of cancer. We are seeing cases that are not stated to be malignant in the Pathology report nor confirmed as such by a clinician; however, these cases do have information that for other primary sites would typically be taken into consideration when determining reportability. We would like instruction from SEER on how to interpret these cases. See Discussion. Discussion: 1) The Pathology and final diagnosis only state “GIST”. Pathology indicates that the bulk of the tumor is submucosal. It extends through the muscularis propria and abuts the serosa. Is this reportable? 2) Pathology and final diagnosis state “GIST”. Pathology report states tumor extends to serosal surface of transverse colon, but not into muscularis propria. CD 117 and CD 34 are positive. Is this reportable? 3) Pathology and final diagnosis state “GIST”. Pathology indicates that tumor invades through the gastric wall to the serosal surface. Is this reportable? 4) Pathology and final diagnosis state “GIST”. Pathology indicates that tumor invades pericolic fat tissue. Is this reportable? 5) Pathology and final diagnosis state “GIST”. No further information in Pathology report, however, scans indicate omental caking. Reportable? 6) Pathology final diagnosis state “GIST”. No further information in Pathology report, however, scans indicate hepatic metastasis. Hepatic metastases are not biopsied. Is this reportable? 7) Biopsy only performed which states “GIST”. Tumor stated to be unresectable and extends into pancreas. Chemotherapy (administered). Final diagnosis states “GIST”. Is this reportable? 8) Pathology and final diagnosis state “GIST”. Pathology states tumor is low to intermediate grade and involves serosal (visceral peritoneum). Is this reportable? 9) Pathology and final diagnosis state “GIST”. Tumor size is 17.5 cm. Pathology states “malignant risk”. Is this reportable? 10) Pathology and final diagnosis state “GIST”. Pathology report states tumor “into muscularis propria” or tumor “involves muscularis propria” or “infiltrates into muscularis propria”. Is this reportable? 11) Pathology and final diagnosis state “GIST”. Pathology report states: “high malignant potential; omentum invaded by tumor”. It is not stated in Pathology report or final diagnosis to be malignant GIST. Is this reportable? 12) Pathology and final diagnosis state “GIST”. Pathology report states that tumor arises from wall of small bowel and extends into thin serosal surface. Is this reportable? 13) Pathology and final diagnosis state “GIST”. Pathology report states: “minimal invasion of lamina propria. (Tumor) does not penetrate the muscularis propria.” Is this reportable? 14) Pathology and final diagnosis state “GIST”. Pathology report states: “high mitotic activity >10/50 HPF; high risk for aggressive behavior; moderate malignant potential.” Is this scenario reportable? 15) Pathology and final diagnosis state “GIST”. Pathology states tumor size is >5 cm. Intermediate risk for aggressive behavior; CD117+ KIT exon 11+. Is this reportable? 16) Pathology and final diagnosis state “GIST”. Pathology states “high risk of malignancy”. Is this reportable? Answer: For GIST to be reportable, the final diagnosis on the Pathology report must definitively state that the GIST is malignant, or invasive, or in situ. Your case number 6 is the only exception, and would be reportable assuming that the scan actually states "hepatic metastases." Based only on the information provided, none of the other examples 4 California Cancer Registry 2010 Gastrointestinal Stromal Tumor (GIST) Recoding Audit CCR Project ID #300.66 Final Report Revised July 11, 2011 are reportable. Extension and/or invasion are not sufficient to confirm malignancy. Borderline neoplasm’s can extend and invade, but do not metastasize. Only malignant neoplasms metastasize. Question ID: 20091021 Status: Final Question: Behavior/Reportability--GIST: Address the issue of reportability of GIST tumors. Answer: Do not report the case to SEER if it does not satisfy the criteria for reportability. According to the current reportability criteria, malignant GIST (8936/3) is reportable to SEER. GIST coded to 8936/0 or 8936/1 is not reportable. If your pathologist will not indicate "malignant" or "benign," code 8936/1 applies according to ICD-O-3 and, therefore, these are not reportable to SEER. Question ID: 20021151 Status: Final Question: Reportability: A "gastrointestinal stromal tumor" (GIST) is not always stated to be "malignant" in the path report even though the tumor appears to meet criteria for malignancy. Should the tumor be considered SEER reportable? Answer: The case is reportable if a pathologist or clinician confirms a diagnosis of cancer. If there is no such confirmation, do not consider the case to be SEER reportable. 5 California Cancer Registry 2010 Gastrointestinal Stromal Tumor (GIST) Recoding Audit CCR Project ID #300.66 Final Report Revised July 11, 2011 GIST Review Reportable vs. Non Reportable 34 35 30 30 28 30 24 25 24 20 16 24 23 17 16 16 15 12 10 10 10 6 5 0 2 3 4 5 6 Region Reportable Non Reportable Graph 1 6 8 9 10