Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Lynch syndrome testing of colorectal cancer patients in a high-income country with universal healthcare: a retrospective study of current practice and gaps in seven australian hospitals

Lynch syndrome testing of colorectal cancer patients in a high-income country with universal... Background: To inform effective genomic medicine strategies, it is important to examine current approaches and gaps in well-established applications. Lynch syndrome (LS) causes 3–5% of colorectal cancers (CRCs). While guidelines commonly recommend LS tumour testing of all CRC patients, implementation in health systems is known to be highly variable. To provide insights on the heterogeneity in practice and current bottlenecks in a high-income country with universal healthcare, we characterise the approaches and gaps in LS testing and referral in seven Australian hospitals across three states. Methods: We obtained surgery, pathology, and genetics services data for 1,624 patients who underwent CRC resections from 01/01/2017 to 31/12/2018 in the included hospitals. Results: Tumour testing approaches differed between hospitals, with 0–19% of patients missing mismatch repair deficiency test results (total 211/1,624 patients). Tumour tests to exclude somatic MLH1 loss were incomplete at five hospitals (42/187 patients). Of 74 patients with tumour tests completed appropriately and indicating high risk of LS, 36 (49%) were missing a record of referral to genetics services for diagnostic testing, with higher missingness for older patients (0% of patients aged ≤ 40 years, 76% of patients aged > 70 years). Of 38 patients with high-risk tumour test results and genetics services referral, diagnostic testing was carried out for 25 (89%) and identified a LS pathogenic/likely pathogenic variant for 11 patients (44% of 25; 0.7% of 1,624 patients). * Correspondence: julia.steinberg@nswcc.org.au The Daffodil Centre, The University of Sydney, a joint venture with Cancer Council NSW, 153 Dowling St, NSW 2011 Woolloomooloo, Australia Full list of author information is available at the end of the article © The Author(s). 2022 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Steinberg et al. Hereditary Cancer in Clinical Practice (2022) 20:18 Page 2 of 8 Conclusions: Given the LS testing and referral gaps, further work is needed to identify strategies for successful integration of LS testing into clinical care, and provide a model for hereditary cancers and broader genomic medicine. Standardised reporting may help clinicians interpret tumour test results and initiate further actions. Keywords: Lynch syndrome, Mismatch repair, Tumour testing, Genetics services referral, Bottleneck, Gap, Heterogeneity in practice, Medical records Background carriers among CRC cancer patients also leads to missed Colorectal cancer (CRC) is one of the most common opportunities for cascade testing of relatives, with analo- cancers, with ~ 1.9 million new diagnoses and > 935,000 gous potential of late cancer diagnoses and deaths that deaths globally in 2020 [1]. Approximately 3–5% of could have been prevented. CRCs are due to Lynch syndrome (LS) [2], an inherited Australia is a high-income country with universal predisposition to cancer (previously called “hereditary healthcare aiming to move towards routine genomic non-polyposis colorectal cancer”). Early detection of LS medicine in practice [13]. Knowledge of current LS test- is key, as it gives individuals access to cancer risk man- ing practice in Australia can therefore provide insights agement strategies including colonoscopic surveillance, on gaps that can occur even in high-income countries which reduces CRC mortality [3]. Moreover, it allows and highlight key areas to address for future genomic for cascade testing of relatives and risk management for medicine approaches. To the best of our knowledge, those who also have LS [4]. only four studies have assessed LS tumour testing prac- Guidelines in many countries including the US, UK tice in the Australian healthcare context [14–17], and and Australia recommend screening of all CRC patients only included 1–2 hospitals or genetics services in one for LS, generally using a step-wise process (see Add- state during different study periods (see Additional File itional Files 1 and 2)[4–9]. The availability of long- 1), limiting insights on the heterogeneity of practice. standing recommendations and cost-effective, evidence- Moreover, only one [17] of these studies included more based testing strategies make LS an important example recent data up to 2017. for genomic medicine [10]. In this study, we characterise the tumour testing and LS is mostly due to pathogenic germline genetic vari- referral approaches in seven large hospitals located in ants in DNA mismatch repair genes (MLH1, MSH2, three different states in Australia in 2017–2018, MSH6, PMS2), or a deletion of the EPCAM gene causing highlighting key gaps and variation in practice. epigenetic MSH2 silencing [7]. Deficient mismatch re- pair (dMMR) often leads to microsatellite instability Methods (MSI) [7]; however, MSI/dMMR also occurs in ~ 15% of Study population non-LS colorectal tumours due to somatic hypermethy- This study was carried out in conjunction with an imple- lation of the MHL1 promoter [11]. mentation trial to improve LS testing [18]. Patients with Individuals with high risk of LS based on tumour CRC resection in the two calendar years prior to the tests (here and below, dMMR/MSI, plus BRAF V600E start of the trial were included, ensuring none of the and/or MLH1 promoter hypermethylation tests where tumour testing and referral rates would be influenced by needed) should be referred for genetic counselling the trial itself. Specifically, this study included seven and germline genetic testing as appropriate. Tumour public hospitals (H1-H7 in the following) from three testing practice varies between contexts (see Add- states in Australia: New South Wales (four hospitals), itional File 1); in some settings, germline genetic test- Victoria (two hospitals), and Western Australia (one ing for LS can also be done directly by surgeons/ hospital). Ethics approval for this study was granted by oncologists (e.g. in Australia since 2020) [12]. the Royal Prince Alfred Hospital Human Research Ethics A key aim of the testing and referral guidelines is to Committee (reference HREC/17/RPAH/542); the ethics improve health outcomes for LS carriers and their rela- committee approved a waiver of consent from individual tives. Missed opportunities to identify LS carriers patients to collect the clinical data. Site-specific govern- through gaps in tumour testing or referrals for genetic ance approval was obtained from each hospital site. counselling and germline genetic testing where appropri- In each hospital, we identified all patients who under- ate lead to missed opportunities in prevention and early went a CRC resection from 01/01/2017 to 31/12/2018. detection of potential metachronous cancers for these Depending on the hospital system, patients were identi- individuals, with a subsequent burden of more aggressive fied from electronic hospital databases or surgeons’ re- treatment or even cancer deaths that could have been cords (see Additional File 3 for details). For each patient, prevented. Similarly, missed opportunities to identify LS pathology data were obtained from electronic patient Steinberg et al. Hereditary Cancer in Clinical Practice (2022) 20:18 Page 3 of 8 medical records or using linkage to a pathology data- After patients’ resection, pathology, and genetics ser- base. Throughout this study, data were obtained using vice data were linked at each hospital, all patient infor- structured queries of electronic databases where pos- mation was de-identified and securely transmitted to the sible, supplemented with manual checks and additional central study team for analysis. data extraction (see Additional File 3). We excluded patients with neuroendocrine, nerve Data analysis sheath, granular cell, yolk-sac, stromal or appendiceal All analyses were done in R v3.6.0. mucinuous tumours, pseudomyxoma peritonei, squa- We calculated the number and percentage of patients mous cell carcinoma, goblet cell carcinoid, acellular with specific tumour test results: (1) both dMMR and mucin (H1: n = 8; H2: n = 6; H2: n = 0; H7: n = 1; H4: MSI test missing; (2) dMMR/MSI detected; (3) MLH1 n = 16; H3: n = 0; H6: n = 3). While for H3, tumour type loss; (4) MLH1 loss with no BRAF V600E nor MLH1 could only be checked for patients without immunohis- promoter hypermethylation test; (5) MLH1 loss with tochemistry test results (see below), all of these were BRAF V600E or MLH1 promoter hypermethylation adenocarcinoma (so that tumour type would not cause detected. missing test results). To ensure that tumour testing was For H1, testing guidelines changed in late 2017, so possible, patients with a medical record note of no/low the number of patients with dMMR/MSI tests missing tumour tissue at resection were excluded (H1: n = 23; was also quantified separately based on 2017 and H2: n = 18; H2: n = 22; H7: n = 5; H4: n = 21; H3: n =8; 2018 data. H6: n = 11). Finally, we excluded patients with known fa- We calculated the number of patients with tumour milial adenomatous polyposis (n < 5; exact number sup- testing completed and the results indicating high risk of pressed to preserve confidentiality). LS (i.e. at least one of: 1) high MSI, 2) MSH2 loss, 3) MSH6 loss, 4) PMS2 loss but no MLH1 loss, or 5) Data on tumour testing and referral to genetics services MLH1 loss with no BRAF V600E nor MLH1 promoter Reports for the following tests were obtained using hypermethylation detected), and the proportion of these pathology data and patient medical records as above: patients with a record of referral to genetics services (in- immunohistochemistry (IHC) tests for dMMR (i.e. cluding any available records of referrals before or after loss of at least one MMR protein), PCR tests for MSI, the CRC resection). “High LS risk” here does not include BRAF V600E tests, and MLH1 promoter hypermethy- patients with tumour MLH1 loss but a BRAF V600E lation tests. Where no report for testing of the variant or MLH1 promoter hypermethylation detected, tumour sample was available, electronic medical re- as they were not generally considered to require a genet- cords for tests undertaken on biopsy samples were ics services referral. checked manually. Data on discussion of patients at We assessed pairwise differences between hospitals for multidisciplinary team (MDT) meetings was available three gaps in testing and referral (Fisher’s test): missing for all hospitals except H2. (1) both dMMR and MSI test results; (2) BRAF V600E To identify patients referred to genetics services, and MLH1 promoter methylation test results with we obtained a list of CRC patients from the genetics tumour MLH1 loss; (3) record of referral to genetics ser- service (or familial cancer centre, hereafter included vices for patients with all tumour tests complete and in- in “genetics services”) working with each of the hos- dicating high LS risk. We also compared the number of pitals, linking data based on patient ID, name, and patients with dMMR/MSI between hospitals (Fisher’s date of birth or surgery (see Additional File 3). This test). To account for multiple testing, significance was included records for patients referred to the genetics defined as p < 0.0024 (Bonferroni correction for 21 pair- service by a clinician, general practitioner (GP) or wise comparisons). who self-referred; we separately noted where pa- We assessed differences between the ages of patients tients’ hospital records stated referral to a different who had tumours with versus without dMMR/MSI genetics service. For each patient included in the (two-sample Wilcoxon test); in a post hoc analysis, this study, presence or absence of a genetics service re- test was repeated for patients aged < 70 years. ferral record, attendance of the patient at the genet- For patients who had all tumour tests complete and ics service, and results of any relevant diagnostic results indicating high LS risk (see above), we also tested genetic tests were noted, including available records for (1) differences between ages of those who did versus of these events that occurred prior to the CRC re- did not have a record of referral to genetics services section (to allow for e.g. prior referral based on fam- (two-sample Wilcoxon test); (2) association between a ily history or cascade testing). To allow for a lag record of referral and discussion of patients at MDT between the resection and referral, genetics service meetings (Fisher’s test); (3) association between a record data were extracted to 28/02/2019. of referral to genetics services and age, sex, MDT Steinberg et al. Hereditary Cancer in Clinical Practice (2022) 20:18 Page 4 of 8 discussion, and hospital in a joint model (logistic regres- Of the 1,413 patients with a recorded dMMR/MSI test sion, complete case analysis). result, 16% had dMMR tumours (13% MLH1 and PMS2 Finally, for patients with all tumour tests complete and loss, 1% MSH2 and MSH6 loss, 1% PMS2 loss only, < indicating high LS risk who had a diagnostic genetic test, 1% MSH6 loss only, none MSI only). Overall, patients we tested for differences in age between those with and with tumour dMMR were older (Fig. 2 A; Wilcoxon p = − 13 without a relevant pathogenic/likely pathogenic variant 3.7 × 10 ). However, a post-hoc analysis showed this (two-sample Wilcoxon test). was mainly due to higher dMMR rates in patients aged 70 + years (Wilcoxon p = 0.75 when restricting the ana- Results lysis to patients aged < 70 years). We included 1,624 patients who underwent a CRC re- In total, 22% (42/187) of patients with tumour MLH1 section from 01/01/2017 to 31/12/2018, with 116–382 loss had missing results for both BRAF V600E and patients per hospital (Table 1). The majority of patients MLH1 promoter hypermethylation tests. BRAF V600E were male (57.1%), and 28.9% were aged ≤ 60 years at test results were recorded for all patients with tumour resection. MLH1 loss at two hospitals (H1, H7), of which one (H7) The seven hospitals had different approaches to LS further included MLH1 promoter hypermethylation test tumour testing (Fig. 1): five hospitals used dMMR IHC results where no somatic BRAF V600E variant was tests, while two (H1, H3) also used MSI PCR tests for found. At the other five hospitals, the follow-up testing some patients. For patients aged 60 + years at H1, was missing for 7–59% of patients with MLH1 loss dMMR/MSI testing was only undertaken upon explicit (Fig. 1; Additional File 4). Where present, the follow-up clinician request in 2017, with testing of all patients even testing detected likely somatic MLH1 inactivation in without explicit requests (“universal testing”) introduced 79% (115/145) of patients with MLH1 loss, underscoring in 2018. All other hospitals introduced universal testing the importance of this test to identify patients with low before 2017. To check whether MLH1 loss (where LS risk. present) is likely somatic only, four hospitals used BRAF Of all patients with full tumour testing completed V600E tests only, while three also used MLH1 promoter (dMMR/MSI and where needed, a BRAF V600E or hypermethylation tests. MLH1 hypermethylation test), a result indicating high Overall, we detected pervasive tumour testing gaps in LS risk was present in 5% (74/1371). However, we could the majority of hospitals. dMMR/MSI test results were not find evidence of a referral to genetics services for present for all patients in only one of seven hospitals 49% (36/74) of these patients. Lack of referral was (H7), closely followed by another hospital (H6; missing strongly associated with older age (Fig. 2B; none missing results for 2% of patients). Testing gaps were more sub- in patients aged ≤ 40 years, one in three missing for pa- stantial at other hospitals: dMMR/MSI test results were tients aged > 40 to 60 years, one in two missing for pa- missing for 8–14% of patients at H2-H5, and 37% of pa- tients aged > 60 to 70 years, and three in four missing tients at H1 (2017: 8% and 79% of patients aged < 60 and for patients aged > 70 years), but not with discussion of 60 + years, respectively; 2018: 19% of patients). In total, patients at multidisciplinary team meetings (see Add- 13% (211/1624) patients had no dMMR/MSI test result itional File 1). Of the 38 patients with a referral record, recorded. For these and other missing tests described records of genetics services consultations were available below, p-values for differences between hospitals are for 28 (74%), 25 (89% of 28) had a diagnostic genetic shown in Additional File 4, with generally significant dif- test, and 11 (44% of 25) had a pathogenic/likely patho- ferences between the hospitals with the highest and low- genic variant in MLH1/PMS2/MSH2/MSH6, thus quali- est missingness rates. fying for a LS diagnosis (see Additional Files 1 and 5). Table 1 Characteristics of colorectal cancer patients included in the study Hospital Number of patients Number of female patients (%) Mean age of patients (range) Number of patients aged ≤ 60 years (%) H1 271 114 (42.7%) 67.9 (30–92) 74 (27.3%) H2 382 171 (44.8%) 67.6 (26–94) 106 (27.7%) H3 123 51 (41.5%) 66.5 (30–90) 40 (32.5%) H4 311 127 (40.8%) 64.5 (22–94) 109 (35.0%) H5 251 102 (40.6%) 69.1 (29–94) 61 (24.3%) H6 170 81 (47.6%) 67.6 (32–92) 44 (25.9%) H7 116 51 (44.0%) 67.9 (26–95) 35 (30.2%) Total 1,624 697 (42.9%) 67.2 (22–95) 469 (28.9%) Steinberg et al. Hereditary Cancer in Clinical Practice (2022) 20:18 Page 5 of 8 Fig. 1 LS tumour testing and referral to genetics services at seven Australian hospitals in 2017–2018. The percentages in each box are calculated relative to the number of patients in the previous testing step. The numbers of patients with (i) MSH2 or MSH6 loss, or (ii) PMS2 loss only, together account for < 10 patients per hospital and are not shown separately to protect patient confidentiality. * test used for at least some patients Discussion with tumour testing complete and indicating high LS In summary, we have analysed LS tumour testing and risk. referral to genetics services for 1,624 CRC patients in In view of the increasing interest in integrating gen- seven Australian hospitals in 2017–2018 (24 months). omic technologies into clinical practice [18–20], the We found three areas with pervasive gaps: missing identification of wide-spread gaps in the context of well- dMMR/MSI test records at five of seven hospitals established LS testing shows that additional research is (overall ~ 1 in 10 patients); missing follow-up testing needed to identify best-practice approaches to genomic of patients with MLH1 loss at five hospitals (overall testing and determine how to support their implementa- ~ 1 in 5 of patients with MLH1 loss); missing record tion. At the seven hospitals in this study, a clinical trial of referral to genetics services for ~ 1 in 2 patients to design interventions addressing the above gaps in LS Fig. 2 dMMR/MSI test results and referral to genetics services by patient age group. a dMMR/MSI tumours were more common in older patients − 13 (Wilcoxon p = 3.7 × 10 for difference in ages of patients with and without dMMR/MSI). b Older patients with tumour test results complete and − 5 indicating high LS risk were less likely to be referred to genetics services (Wilcoxon p = 9.8 × 10 for difference in ages of patients with and without referral record) Steinberg et al. Hereditary Cancer in Clinical Practice (2022) 20:18 Page 6 of 8 testing and referral is currently underway [18]. This in- genomic testing into clinical practice. One would ex- cludes the identification of barriers to appropriate test- pect such training to play an even more important ing and referral as perceived by a wide range of different role in the future, with, for example, recent changes stakeholders at each hospital, which may also be inform- enabling surgeons or oncologists in Australia to dir- ative for other applications of genomic medicine. In par- ectly order germline LS genetic testing without prior ticular, there are likely mutliple psychosocial and referral to genetics services. contextual factors contributing to tumour testing and re- While we did not observe an association between a ferral gaps, such as mixed perceptions of healthcare staff record of referral to genetics services and discussion of around the utility of tumour testing and referral, high patients with molecular results indicating high risk of LS workload and competing priorities, lack of clarity around at an MDT meeting, this could be due to limited sample processes and roles, difficulty interpreting tumour test size. For most hospitals, we were also not able to obtain results and challenges around remembering criteria for the dates of MDT meetings and thus could not deter- referral to genetics services. Ultimately, the clinical trial mine whether the tumour test results were available for aims to test which interventions could support health- that discussion. For example, for hospital H1, the care staff for relevant steps of the tumour testing and re- complete tumour test results were available at the time ferral pathway. of the MDT discussion for 3/5 patients with high-risk As an example of excellent practice, LS tumour testing results who were referred to genetics services, but only of all CRC patients with highly standardised processes for 1/5 patients without a record of referral. Conse- was demonstrated by one hospital in this study (H7). quently, further data are needed to determine whether While extracting the exact wording of all pathology re- and how systematic discussion of patients with high-risk ports and whether they used synoptic reporting was out- tumour test results at MDT meetings could help im- side the scope of this work, we obtained some examples prove risk-appropriate referrals to genetics services. of pathology reports (see Additional File 6). These sug- A particular challenge for identifying and applying gest the reports for H7 were highly standardised, with best practice is the availability and accessibility of up- more variability in phrasing at some other hospitals, es- to-date clinical data (detailed discussion see Add- pecially for results indicating high LS risk. itional File 1). The results presented here are based As suggested by the Australian Gastrointestinal Path- on data for 2017–2018, noting that pathology practice ology Society [8], standardised reporting might help cli- since then may have changed and impacted the extent nicians interpret test results and initiate further actions. of testing gaps. International guidelines to recommend Electronic medical records with standardised forms, easy tumour testing of CRC patients have emerged since access to reliable knowledge, and clinical decision sup- 2009 [4, 5, 7], and universal testing was thought to port systems may also assist effective integration of gen- be in place for six of the seven hospitals in this study omics into clinical workflows [21]. However, appropriate by 2017 (and in all seven by 2018). However, testing referral and germline genetic testing also involves practice could have been further improved by the obtaining informed consent from patients, requiring Australasian Gastrointestinal Pathology Society con- both genomics knowledge and communication/counsel- sensus practice guidelines [8], which endorsed univer- ling skills. sal testing of all CRC patients. Thus, it will be A previous study in the USA also suggested that important to continue examining LS tumour testing the involvement of genetics services earlier in the and referral practice in Australia beyond 2018. In testing and referral pathway could substantially im- clinical practice, improved data management and re- prove referral rates: in that context, referral to genet- search governance could facilitate internal hospital ics services was both faster and more comprehensive audits and encourage learning health systems with during a period when tumour test results were imme- better integration of research and clinical practice diately sent to both colorectal surgeons and genetic [25]. Meanwhile, research data as presented here can counsellors, and the genetic counsellors contacted pa- help showcase best-practice achievements. tients directly [22]. By contrast, when only surgeons A limitation of our study is that patients with meta- received the tumour test results, about 1 in 2 CRC static cancer and no CRC resection were not included. patients with tumour dMMR or MSI in that hospital While we employed a consensus approach to ensure were not referred to genetics services, similar to the between-hospital data comparability (see Methods, Add- average rate observed in our study. In alignment with itional Files 3, 7 and 8), some information for patients past reviews [23] and the Australian Medical Associ- undergoing CRC resections may have been missed due ation 2020 statement on Genetic Testing and Genom- to data extraction errors or incomplete records. To miti- ics in Medicine [24], this also suggests that training gate this, data extraction was carried out by healthser- of clinicians is key to successful integration of vice professionals employed within the healthcare Steinberg et al. Hereditary Cancer in Clinical Practice (2022) 20:18 Page 7 of 8 system, extensively consulting hospital stakeholders to alternative to identify colorectal resections (in conjunction with ICD optimise data access and coverage, and combining infor- codes for CRC diagnosis). mation from multiple systems where necessary. Finally, we could not identify where referral to genetics services Acknowledgements Not applicable. was declined by a patient or not possible due to their ill- health/death (more likely among older patients, for Authors’ contributions whom referral records were less common). While these Conceptualization: N.T., J.S. Data curation: J.S., P.C., G.T., A.M., R.V., L.T., L.S., A.P., C.N., S.M., A.H., N.E. Formal analysis, methodology, and visualization: J.S. factors could preclude some referrals, missed LS diagno- Project administration: E.H., P.C., G.T. Funding acquisition: N.T., K.T., F.M., A.G., ses also reduce prevention and surveillance opportunities K.C., N.P., B.P., M.S., J.K. Writing – original draft: J.S. Writing – review & editing: for patients’ relatives, which was identified as particularly all authors. The author(s) read and approved the final manuscript. important by a patient representative who was part of Funding the study team. This study is funded by Cancer Institute NSW (2017/CDF005) and Cancer Our work also has notable strengths: inclusion of mul- Australia (1123924). The contents of this publication are solely the responsibility of the authors and do not reflect the views of Cancer Australia. tiple hospitals in different Australian states, use of clin- ical data and electronic health records where available, Availability of data and materials and identification of referrals using data from genetics Individual-level patient data are not publicly available to preserve patients’ services. privacy. Reasonable requests for data can be submitted to the corresponding author and will be subject to ethics committee and individual hospital governance requirements. Conclusions Declarations Tumour testing for LS is widely adopted at hospitals in Australia, but we have found that gaps in practice re- Ethics approval and consent to participate Ethics approval for this study was granted by the Royal Prince Alfred Hospital main at many hospitals in different states. Further work Human Research Ethics Committee (reference HREC/17/RPAH/542); the is needed to identify the procedures, funding structures, ethics committee approved a waiver of consent from individual patients to and targeted implementation strategies that can ensure collect the clinical data. Site-specific governance approval was obtained from each hospital site. successful implementation of universal testing in the LS context, and inform approaches for future broader inte- Consent for publication gration of genomic testing pathways into clinical Not applicable. practice. Competing interests KC is co-PI of unrelated investigator-initiated trial of cervical screening in Abbreviations Australia (‘Compass’) conducted by the Victorian Cytology Service, which has LS: Lynch syndrome; CRC: Colorectal cancer; dMMR: Deficient mismatch received funding contribution from Roche Molecular Systems and Ventana, repair; MMR: Mismatch repair; MSI: Microsatellite instability; USA. The other authors declare no potential conflicts of interest. IHC: Immunohistochemistry; PCR: Polymerase chain reaction; MDT: Multidisciplinary team; GP: General practicioner; ICD: International Author details Classification of Diseases; MBS: Medicare Benefit Schedule; ACHI: Australian The Daffodil Centre, The University of Sydney, a joint venture with Cancer Classification of Health Interventions Council NSW, 153 Dowling St, NSW 2011 Woolloomooloo, Australia. Surgical Outcomes Research Centre (SOuRCe), Royal Prince Alfred Hospital, Camperdown, NSW, Australia. NSW Health, Western Sydney Local Health Supplementary information District, Westmead, NSW, Australia. Monash Health, Melbourne, Victoria, The online version contains supplementary material available at https://doi. 5 6 Australia. Cancer Council NSW, Sydney, NSW, Australia. Genetic Services of org/10.1186/s13053-022-00225-1. Western Australia, King Edward Memorial Hospital, Perth, Western Australia, Australia. The Royal Melbourne Hospital, Melbourne, Victoria, Australia. School of Medicine and Public Health, University of Newcastle, Newcastle, Additional file 1: Supplementary Notes. NSW, Australia. The University of Sydney, Northern Clinical School Royal Additional file 2: Conceptual map of a step-wise tumour testing ap- North Shore Hospital, Sydney, NSW, Australia. Personalised Oncology proach for LS. Division, Walter and Eliza Hall Institute, Melbourne, Victoria, Australia. Additional file 3: Supplementary Methods. Cancer Genetics Department, Royal Prince Alfred Hospital, Camperdown, NSW, Australia. Department of Tissue Pathology & Diagnostic Oncology, Additional file 4: P-values for pairwise differences for missing test NSW Health Pathology, Royal Prince Alfred Hospital, Camperdown, NSW, results and referrals to genetics services between hospitals. Australia. Faculty of Medicine and Health, The University of Sydney, Sydney, Additional file 5: Diagnostic genetic testing among patients with NSW, Australia. School of Medicine and Pharmacology, University of complete, high-risk tumour test results who had a referral to genetic Western Australia, Perth, Western, Australia. Centre for the Health Economy, services. Macquarie University, Sydney, NSW, Australia. Westmead Familial Cancer Additional file 6: Examples of different terminology for LS tumour test Services, The Crown Princess Mary Cancer Centre, Westmead Hospital, results used in pathology reports. Westmead, NSW, Australia. Department of Colorectal Surgery, Royal Prince Alfred Hospital, Camperdown, NSW, Australia. Department of Colorectal Additional file 7: Consensus list of Medicare Benefit Schedule (MBS) Surgery, The University of Sydney, Westmead Hospital, Westmead, NSW, procedure codes used to identify colorectal resections (in conjunction Australia. Fiona Stanley Hospital, South Metropolitan Health Service, with ICD codes for CRC diagnosis). Murdoch, Western Australia, Australia. Sydney Medical School, University of Additional file 8: Australian Classification of Health Interventions (ACHI) Sydney, Sydney, NSW, Australia. Colorectal Medicine and Genetics, procedure codes mapped to consensus list of MBS items as an Department of Medicine, The Royal Melbourne Hospital, Melbourne, Victoria, Steinberg et al. Hereditary Cancer in Clinical Practice (2022) 20:18 Page 8 of 8 Australia. Hereditary Cancer Clinic, Prince of Wales Hospital, Sydney, NSW, 19. Khoury MJ. No Shortcuts on the Long Road to Evidence-Based Genomic Australia. Prince of Wales Clinical School, UNSW Sydney, NSW, Australia. Medicine. JAMA. 2017;318(1):27–8. School of Population Health, Faculty of Medicine, UNSW Sydney, NSW, 20. Roberts MC, Kennedy AE, Chambers DA, Khoury MJ. The current state of Australia. implementation science in genomic medicine: opportunities for improvement. Genet Sci. 2017;19(8):858–63. Received: 2 February 2022 Accepted: 10 April 2022 21. Williams MS, Taylor CO, Walton NA, Goehringer SR, Aronson S, Freimuth RR, et al. Genomic Information for Clinicians in the Electronic Health Record: Lessons Learned From the Clinical Genome Resource Project and the Electronic Medical Records and Genomics Network. Front Genet. 2019;10: References 1. Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, et al. 22. Heald B, Plesec T, Liu X, Pai R, Patil D, Moline J, et al. Implementation of Global cancer statistics 2020: GLOBOCAN estimates of incidence and universal microsatellite instability and immunohistochemistry screening for mortality worldwide for 36 cancers in 185 countries. CA: A Cancer Journal diagnosing lynch syndrome in a large academic medical center. J Clin for Clinicians; 2021. Oncol. 2013;31(10):1336–40. 2. U.S National Library of Medicine. Lynch Syndrome 2021. 2020. https://ghr. 23. Crellin E, McClaren B, Nisselle A, Best S, Gaff C, Metcalfe S. Preparing Medical nlm.nih.gov/condition/lynch-syndrome. Accessed 20 Apr 2022. Specialists to Practice Genomic Medicine: Education an Essential Part of a 3. Barrow P, Khan M, Lalloo F, Evans DG, Hill J. Systematic review of the Broader Strategy. Front Genet. 2019;10:789. impact of registration and screening on colorectal cancer incidence and 24. Australian Medical Association. Genetic Testing and Genomics in Medicine mortality in familial adenomatous polyposis and Lynch syndrome. Br J Surg. 2020. 2020. https://www.ama.com.au/articles/genetic-testing-and-genomics- 2013;100(13):1719–31. medicine-2020. Accessed 20 Apr 2022. 4. Evaluation of Genomic Applications in P, Prevention Working Group. 25. Clay-Williams R, Taylor N, Braithwaite J. Potential solutions to improve the Recommendations from the EGAPP Working Group: genetic testing governance of multicentre health services research. Med J Aust. 2018;208(4): strategies in newly diagnosed individuals with colorectal cancer aimed at 152–4. reducing morbidity and mortality from Lynch syndrome in relatives. Genet Med. 2009;11(1):35–41. 5. Provenzale D, Gupta S, Ahnen DJ, Bray T, Cannon JA, Cooper G, et al. Publisher’sNote Genetic/Familial High-Risk Assessment: Colorectal Version 1.2016, NCCN Springer Nature remains neutral with regard to jurisdictional claims in Clinical Practice Guidelines in Oncology. J Natl Compr Canc Netw. 2016; published maps and institutional affiliations. 14(8):1010–30. 6. Tognetto A, Michelazzo MB, Calabró GE, Unim B, Di Marco M, Ricciardi W, et al. A Systematic Review on the Existing Screening Pathways for Lynch Syndrome Identification. Front Public Health. 2017;5:243. 7. Vasen HF, Blanco I, Aktan-Collan K, Gopie JP, Alonso A, Aretz S, et al. Revised guidelines for the clinical management of Lynch syndrome (HNPCC): recommendations by a group of European experts. Gut. 2013;62(6):812–23. 8. Yozu M, Kumarasinghe MP, Brown IS, Gill AJ, Rosty C. Australasian Gastrointestinal Pathology Society (AGPS) consensus guidelines for universal defective mismatch repair testing in colorectal carcinoma. Pathology. 2019; 51(3):233–9. 9. National Institute for Health and Care Excellence. Molecular testing strategies for Lynch syndrome in people with colorectal cancer 2017. 2017. https://www.nice.org.uk/guidance/dg27. Accessed 20 Apr 2022. 10. Rahm AK, Cragun D, Hunter JE, Epstein MM, Lowery J, Lu CY, et al. Implementing universal Lynch syndrome screening (IMPULSS): protocol for a multi-site study to identify strategies to implement, adapt, and sustain genomic medicine programs in different organizational contexts. BMC Health Serv Res. 2018;18(1):824. 11. Lynch HT, Lynch PM, Lanspa SJ, Snyder CL, Lynch JF, Boland CR. Review of the Lynch syndrome: history, molecular genetics, screening, differential diagnosis, and medicolegal ramifications. Clin Genet. 2009;76(1):1–18. 12. Australian Government Department of Health. Medicare Benefits Schedule - Item 73354. 2020. http://www9.health.gov.au/mbs/fullDisplay.cfm?type= item&q=73354&qt=item. Accessed 20 Apr 2022. 13. Australian Government Department of Health. Genomics Health Futures Mission 2021. 2021. https://www.health.gov.au/initiatives-and-programs/ genomics-health-futures-mission. Accessed 20 Apr 2022. 14. Schofield L, Grieu F, Amanuel B, Carrello A, Spagnolo D, Kiraly C, et al. Population-based screening for Lynch syndrome in Western Australia. Int J Cancer. 2014;135(5):1085–91. 15. Brennan B, Hemmings CT, Clark I, Yip D, Fadia M, Taupin DR. Universal molecular screening does not effectively detect Lynch syndrome in clinical practice. Th Adv Gastroenterol. 2017;10(4):361–71. 16. Long JC, Debono D, Williams R, Salisbury E, O’Neill S, Eykman E, et al. Using behaviour change and implementation science to address low referral rates in oncology. BMC Health Serv Res. 2018;18(1):904. 17. Loh Z, Williams DS, Salmon L, Dow E, John T. Impact of universal immunohistochemistry on Lynch syndrome diagnosis in an Australian colorectal cancer cohort. Intern Med J. 2019;49(10):1278–84. 18. Morrow A, Hogden E, Kang YJ, Steinberg J, Canfell K, Solomon MJ, et al. Comparing theory and non-theory based implementation approaches to improving referral practices in cancer genetics: a cluster randomised trial protocol. Trials. 2019;20(1):373. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Hereditary Cancer in Clinical Practice Springer Journals

Loading next page...
 
/lp/springer-journals/lynch-syndrome-testing-of-colorectal-cancer-patients-in-a-high-income-sAw8B00CNB

References (19)

Publisher
Springer Journals
Copyright
Copyright © The Author(s) 2022
eISSN
1897-4287
DOI
10.1186/s13053-022-00225-1
Publisher site
See Article on Publisher Site

Abstract

Background: To inform effective genomic medicine strategies, it is important to examine current approaches and gaps in well-established applications. Lynch syndrome (LS) causes 3–5% of colorectal cancers (CRCs). While guidelines commonly recommend LS tumour testing of all CRC patients, implementation in health systems is known to be highly variable. To provide insights on the heterogeneity in practice and current bottlenecks in a high-income country with universal healthcare, we characterise the approaches and gaps in LS testing and referral in seven Australian hospitals across three states. Methods: We obtained surgery, pathology, and genetics services data for 1,624 patients who underwent CRC resections from 01/01/2017 to 31/12/2018 in the included hospitals. Results: Tumour testing approaches differed between hospitals, with 0–19% of patients missing mismatch repair deficiency test results (total 211/1,624 patients). Tumour tests to exclude somatic MLH1 loss were incomplete at five hospitals (42/187 patients). Of 74 patients with tumour tests completed appropriately and indicating high risk of LS, 36 (49%) were missing a record of referral to genetics services for diagnostic testing, with higher missingness for older patients (0% of patients aged ≤ 40 years, 76% of patients aged > 70 years). Of 38 patients with high-risk tumour test results and genetics services referral, diagnostic testing was carried out for 25 (89%) and identified a LS pathogenic/likely pathogenic variant for 11 patients (44% of 25; 0.7% of 1,624 patients). * Correspondence: julia.steinberg@nswcc.org.au The Daffodil Centre, The University of Sydney, a joint venture with Cancer Council NSW, 153 Dowling St, NSW 2011 Woolloomooloo, Australia Full list of author information is available at the end of the article © The Author(s). 2022 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Steinberg et al. Hereditary Cancer in Clinical Practice (2022) 20:18 Page 2 of 8 Conclusions: Given the LS testing and referral gaps, further work is needed to identify strategies for successful integration of LS testing into clinical care, and provide a model for hereditary cancers and broader genomic medicine. Standardised reporting may help clinicians interpret tumour test results and initiate further actions. Keywords: Lynch syndrome, Mismatch repair, Tumour testing, Genetics services referral, Bottleneck, Gap, Heterogeneity in practice, Medical records Background carriers among CRC cancer patients also leads to missed Colorectal cancer (CRC) is one of the most common opportunities for cascade testing of relatives, with analo- cancers, with ~ 1.9 million new diagnoses and > 935,000 gous potential of late cancer diagnoses and deaths that deaths globally in 2020 [1]. Approximately 3–5% of could have been prevented. CRCs are due to Lynch syndrome (LS) [2], an inherited Australia is a high-income country with universal predisposition to cancer (previously called “hereditary healthcare aiming to move towards routine genomic non-polyposis colorectal cancer”). Early detection of LS medicine in practice [13]. Knowledge of current LS test- is key, as it gives individuals access to cancer risk man- ing practice in Australia can therefore provide insights agement strategies including colonoscopic surveillance, on gaps that can occur even in high-income countries which reduces CRC mortality [3]. Moreover, it allows and highlight key areas to address for future genomic for cascade testing of relatives and risk management for medicine approaches. To the best of our knowledge, those who also have LS [4]. only four studies have assessed LS tumour testing prac- Guidelines in many countries including the US, UK tice in the Australian healthcare context [14–17], and and Australia recommend screening of all CRC patients only included 1–2 hospitals or genetics services in one for LS, generally using a step-wise process (see Add- state during different study periods (see Additional File itional Files 1 and 2)[4–9]. The availability of long- 1), limiting insights on the heterogeneity of practice. standing recommendations and cost-effective, evidence- Moreover, only one [17] of these studies included more based testing strategies make LS an important example recent data up to 2017. for genomic medicine [10]. In this study, we characterise the tumour testing and LS is mostly due to pathogenic germline genetic vari- referral approaches in seven large hospitals located in ants in DNA mismatch repair genes (MLH1, MSH2, three different states in Australia in 2017–2018, MSH6, PMS2), or a deletion of the EPCAM gene causing highlighting key gaps and variation in practice. epigenetic MSH2 silencing [7]. Deficient mismatch re- pair (dMMR) often leads to microsatellite instability Methods (MSI) [7]; however, MSI/dMMR also occurs in ~ 15% of Study population non-LS colorectal tumours due to somatic hypermethy- This study was carried out in conjunction with an imple- lation of the MHL1 promoter [11]. mentation trial to improve LS testing [18]. Patients with Individuals with high risk of LS based on tumour CRC resection in the two calendar years prior to the tests (here and below, dMMR/MSI, plus BRAF V600E start of the trial were included, ensuring none of the and/or MLH1 promoter hypermethylation tests where tumour testing and referral rates would be influenced by needed) should be referred for genetic counselling the trial itself. Specifically, this study included seven and germline genetic testing as appropriate. Tumour public hospitals (H1-H7 in the following) from three testing practice varies between contexts (see Add- states in Australia: New South Wales (four hospitals), itional File 1); in some settings, germline genetic test- Victoria (two hospitals), and Western Australia (one ing for LS can also be done directly by surgeons/ hospital). Ethics approval for this study was granted by oncologists (e.g. in Australia since 2020) [12]. the Royal Prince Alfred Hospital Human Research Ethics A key aim of the testing and referral guidelines is to Committee (reference HREC/17/RPAH/542); the ethics improve health outcomes for LS carriers and their rela- committee approved a waiver of consent from individual tives. Missed opportunities to identify LS carriers patients to collect the clinical data. Site-specific govern- through gaps in tumour testing or referrals for genetic ance approval was obtained from each hospital site. counselling and germline genetic testing where appropri- In each hospital, we identified all patients who under- ate lead to missed opportunities in prevention and early went a CRC resection from 01/01/2017 to 31/12/2018. detection of potential metachronous cancers for these Depending on the hospital system, patients were identi- individuals, with a subsequent burden of more aggressive fied from electronic hospital databases or surgeons’ re- treatment or even cancer deaths that could have been cords (see Additional File 3 for details). For each patient, prevented. Similarly, missed opportunities to identify LS pathology data were obtained from electronic patient Steinberg et al. Hereditary Cancer in Clinical Practice (2022) 20:18 Page 3 of 8 medical records or using linkage to a pathology data- After patients’ resection, pathology, and genetics ser- base. Throughout this study, data were obtained using vice data were linked at each hospital, all patient infor- structured queries of electronic databases where pos- mation was de-identified and securely transmitted to the sible, supplemented with manual checks and additional central study team for analysis. data extraction (see Additional File 3). We excluded patients with neuroendocrine, nerve Data analysis sheath, granular cell, yolk-sac, stromal or appendiceal All analyses were done in R v3.6.0. mucinuous tumours, pseudomyxoma peritonei, squa- We calculated the number and percentage of patients mous cell carcinoma, goblet cell carcinoid, acellular with specific tumour test results: (1) both dMMR and mucin (H1: n = 8; H2: n = 6; H2: n = 0; H7: n = 1; H4: MSI test missing; (2) dMMR/MSI detected; (3) MLH1 n = 16; H3: n = 0; H6: n = 3). While for H3, tumour type loss; (4) MLH1 loss with no BRAF V600E nor MLH1 could only be checked for patients without immunohis- promoter hypermethylation test; (5) MLH1 loss with tochemistry test results (see below), all of these were BRAF V600E or MLH1 promoter hypermethylation adenocarcinoma (so that tumour type would not cause detected. missing test results). To ensure that tumour testing was For H1, testing guidelines changed in late 2017, so possible, patients with a medical record note of no/low the number of patients with dMMR/MSI tests missing tumour tissue at resection were excluded (H1: n = 23; was also quantified separately based on 2017 and H2: n = 18; H2: n = 22; H7: n = 5; H4: n = 21; H3: n =8; 2018 data. H6: n = 11). Finally, we excluded patients with known fa- We calculated the number of patients with tumour milial adenomatous polyposis (n < 5; exact number sup- testing completed and the results indicating high risk of pressed to preserve confidentiality). LS (i.e. at least one of: 1) high MSI, 2) MSH2 loss, 3) MSH6 loss, 4) PMS2 loss but no MLH1 loss, or 5) Data on tumour testing and referral to genetics services MLH1 loss with no BRAF V600E nor MLH1 promoter Reports for the following tests were obtained using hypermethylation detected), and the proportion of these pathology data and patient medical records as above: patients with a record of referral to genetics services (in- immunohistochemistry (IHC) tests for dMMR (i.e. cluding any available records of referrals before or after loss of at least one MMR protein), PCR tests for MSI, the CRC resection). “High LS risk” here does not include BRAF V600E tests, and MLH1 promoter hypermethy- patients with tumour MLH1 loss but a BRAF V600E lation tests. Where no report for testing of the variant or MLH1 promoter hypermethylation detected, tumour sample was available, electronic medical re- as they were not generally considered to require a genet- cords for tests undertaken on biopsy samples were ics services referral. checked manually. Data on discussion of patients at We assessed pairwise differences between hospitals for multidisciplinary team (MDT) meetings was available three gaps in testing and referral (Fisher’s test): missing for all hospitals except H2. (1) both dMMR and MSI test results; (2) BRAF V600E To identify patients referred to genetics services, and MLH1 promoter methylation test results with we obtained a list of CRC patients from the genetics tumour MLH1 loss; (3) record of referral to genetics ser- service (or familial cancer centre, hereafter included vices for patients with all tumour tests complete and in- in “genetics services”) working with each of the hos- dicating high LS risk. We also compared the number of pitals, linking data based on patient ID, name, and patients with dMMR/MSI between hospitals (Fisher’s date of birth or surgery (see Additional File 3). This test). To account for multiple testing, significance was included records for patients referred to the genetics defined as p < 0.0024 (Bonferroni correction for 21 pair- service by a clinician, general practitioner (GP) or wise comparisons). who self-referred; we separately noted where pa- We assessed differences between the ages of patients tients’ hospital records stated referral to a different who had tumours with versus without dMMR/MSI genetics service. For each patient included in the (two-sample Wilcoxon test); in a post hoc analysis, this study, presence or absence of a genetics service re- test was repeated for patients aged < 70 years. ferral record, attendance of the patient at the genet- For patients who had all tumour tests complete and ics service, and results of any relevant diagnostic results indicating high LS risk (see above), we also tested genetic tests were noted, including available records for (1) differences between ages of those who did versus of these events that occurred prior to the CRC re- did not have a record of referral to genetics services section (to allow for e.g. prior referral based on fam- (two-sample Wilcoxon test); (2) association between a ily history or cascade testing). To allow for a lag record of referral and discussion of patients at MDT between the resection and referral, genetics service meetings (Fisher’s test); (3) association between a record data were extracted to 28/02/2019. of referral to genetics services and age, sex, MDT Steinberg et al. Hereditary Cancer in Clinical Practice (2022) 20:18 Page 4 of 8 discussion, and hospital in a joint model (logistic regres- Of the 1,413 patients with a recorded dMMR/MSI test sion, complete case analysis). result, 16% had dMMR tumours (13% MLH1 and PMS2 Finally, for patients with all tumour tests complete and loss, 1% MSH2 and MSH6 loss, 1% PMS2 loss only, < indicating high LS risk who had a diagnostic genetic test, 1% MSH6 loss only, none MSI only). Overall, patients we tested for differences in age between those with and with tumour dMMR were older (Fig. 2 A; Wilcoxon p = − 13 without a relevant pathogenic/likely pathogenic variant 3.7 × 10 ). However, a post-hoc analysis showed this (two-sample Wilcoxon test). was mainly due to higher dMMR rates in patients aged 70 + years (Wilcoxon p = 0.75 when restricting the ana- Results lysis to patients aged < 70 years). We included 1,624 patients who underwent a CRC re- In total, 22% (42/187) of patients with tumour MLH1 section from 01/01/2017 to 31/12/2018, with 116–382 loss had missing results for both BRAF V600E and patients per hospital (Table 1). The majority of patients MLH1 promoter hypermethylation tests. BRAF V600E were male (57.1%), and 28.9% were aged ≤ 60 years at test results were recorded for all patients with tumour resection. MLH1 loss at two hospitals (H1, H7), of which one (H7) The seven hospitals had different approaches to LS further included MLH1 promoter hypermethylation test tumour testing (Fig. 1): five hospitals used dMMR IHC results where no somatic BRAF V600E variant was tests, while two (H1, H3) also used MSI PCR tests for found. At the other five hospitals, the follow-up testing some patients. For patients aged 60 + years at H1, was missing for 7–59% of patients with MLH1 loss dMMR/MSI testing was only undertaken upon explicit (Fig. 1; Additional File 4). Where present, the follow-up clinician request in 2017, with testing of all patients even testing detected likely somatic MLH1 inactivation in without explicit requests (“universal testing”) introduced 79% (115/145) of patients with MLH1 loss, underscoring in 2018. All other hospitals introduced universal testing the importance of this test to identify patients with low before 2017. To check whether MLH1 loss (where LS risk. present) is likely somatic only, four hospitals used BRAF Of all patients with full tumour testing completed V600E tests only, while three also used MLH1 promoter (dMMR/MSI and where needed, a BRAF V600E or hypermethylation tests. MLH1 hypermethylation test), a result indicating high Overall, we detected pervasive tumour testing gaps in LS risk was present in 5% (74/1371). However, we could the majority of hospitals. dMMR/MSI test results were not find evidence of a referral to genetics services for present for all patients in only one of seven hospitals 49% (36/74) of these patients. Lack of referral was (H7), closely followed by another hospital (H6; missing strongly associated with older age (Fig. 2B; none missing results for 2% of patients). Testing gaps were more sub- in patients aged ≤ 40 years, one in three missing for pa- stantial at other hospitals: dMMR/MSI test results were tients aged > 40 to 60 years, one in two missing for pa- missing for 8–14% of patients at H2-H5, and 37% of pa- tients aged > 60 to 70 years, and three in four missing tients at H1 (2017: 8% and 79% of patients aged < 60 and for patients aged > 70 years), but not with discussion of 60 + years, respectively; 2018: 19% of patients). In total, patients at multidisciplinary team meetings (see Add- 13% (211/1624) patients had no dMMR/MSI test result itional File 1). Of the 38 patients with a referral record, recorded. For these and other missing tests described records of genetics services consultations were available below, p-values for differences between hospitals are for 28 (74%), 25 (89% of 28) had a diagnostic genetic shown in Additional File 4, with generally significant dif- test, and 11 (44% of 25) had a pathogenic/likely patho- ferences between the hospitals with the highest and low- genic variant in MLH1/PMS2/MSH2/MSH6, thus quali- est missingness rates. fying for a LS diagnosis (see Additional Files 1 and 5). Table 1 Characteristics of colorectal cancer patients included in the study Hospital Number of patients Number of female patients (%) Mean age of patients (range) Number of patients aged ≤ 60 years (%) H1 271 114 (42.7%) 67.9 (30–92) 74 (27.3%) H2 382 171 (44.8%) 67.6 (26–94) 106 (27.7%) H3 123 51 (41.5%) 66.5 (30–90) 40 (32.5%) H4 311 127 (40.8%) 64.5 (22–94) 109 (35.0%) H5 251 102 (40.6%) 69.1 (29–94) 61 (24.3%) H6 170 81 (47.6%) 67.6 (32–92) 44 (25.9%) H7 116 51 (44.0%) 67.9 (26–95) 35 (30.2%) Total 1,624 697 (42.9%) 67.2 (22–95) 469 (28.9%) Steinberg et al. Hereditary Cancer in Clinical Practice (2022) 20:18 Page 5 of 8 Fig. 1 LS tumour testing and referral to genetics services at seven Australian hospitals in 2017–2018. The percentages in each box are calculated relative to the number of patients in the previous testing step. The numbers of patients with (i) MSH2 or MSH6 loss, or (ii) PMS2 loss only, together account for < 10 patients per hospital and are not shown separately to protect patient confidentiality. * test used for at least some patients Discussion with tumour testing complete and indicating high LS In summary, we have analysed LS tumour testing and risk. referral to genetics services for 1,624 CRC patients in In view of the increasing interest in integrating gen- seven Australian hospitals in 2017–2018 (24 months). omic technologies into clinical practice [18–20], the We found three areas with pervasive gaps: missing identification of wide-spread gaps in the context of well- dMMR/MSI test records at five of seven hospitals established LS testing shows that additional research is (overall ~ 1 in 10 patients); missing follow-up testing needed to identify best-practice approaches to genomic of patients with MLH1 loss at five hospitals (overall testing and determine how to support their implementa- ~ 1 in 5 of patients with MLH1 loss); missing record tion. At the seven hospitals in this study, a clinical trial of referral to genetics services for ~ 1 in 2 patients to design interventions addressing the above gaps in LS Fig. 2 dMMR/MSI test results and referral to genetics services by patient age group. a dMMR/MSI tumours were more common in older patients − 13 (Wilcoxon p = 3.7 × 10 for difference in ages of patients with and without dMMR/MSI). b Older patients with tumour test results complete and − 5 indicating high LS risk were less likely to be referred to genetics services (Wilcoxon p = 9.8 × 10 for difference in ages of patients with and without referral record) Steinberg et al. Hereditary Cancer in Clinical Practice (2022) 20:18 Page 6 of 8 testing and referral is currently underway [18]. This in- genomic testing into clinical practice. One would ex- cludes the identification of barriers to appropriate test- pect such training to play an even more important ing and referral as perceived by a wide range of different role in the future, with, for example, recent changes stakeholders at each hospital, which may also be inform- enabling surgeons or oncologists in Australia to dir- ative for other applications of genomic medicine. In par- ectly order germline LS genetic testing without prior ticular, there are likely mutliple psychosocial and referral to genetics services. contextual factors contributing to tumour testing and re- While we did not observe an association between a ferral gaps, such as mixed perceptions of healthcare staff record of referral to genetics services and discussion of around the utility of tumour testing and referral, high patients with molecular results indicating high risk of LS workload and competing priorities, lack of clarity around at an MDT meeting, this could be due to limited sample processes and roles, difficulty interpreting tumour test size. For most hospitals, we were also not able to obtain results and challenges around remembering criteria for the dates of MDT meetings and thus could not deter- referral to genetics services. Ultimately, the clinical trial mine whether the tumour test results were available for aims to test which interventions could support health- that discussion. For example, for hospital H1, the care staff for relevant steps of the tumour testing and re- complete tumour test results were available at the time ferral pathway. of the MDT discussion for 3/5 patients with high-risk As an example of excellent practice, LS tumour testing results who were referred to genetics services, but only of all CRC patients with highly standardised processes for 1/5 patients without a record of referral. Conse- was demonstrated by one hospital in this study (H7). quently, further data are needed to determine whether While extracting the exact wording of all pathology re- and how systematic discussion of patients with high-risk ports and whether they used synoptic reporting was out- tumour test results at MDT meetings could help im- side the scope of this work, we obtained some examples prove risk-appropriate referrals to genetics services. of pathology reports (see Additional File 6). These sug- A particular challenge for identifying and applying gest the reports for H7 were highly standardised, with best practice is the availability and accessibility of up- more variability in phrasing at some other hospitals, es- to-date clinical data (detailed discussion see Add- pecially for results indicating high LS risk. itional File 1). The results presented here are based As suggested by the Australian Gastrointestinal Path- on data for 2017–2018, noting that pathology practice ology Society [8], standardised reporting might help cli- since then may have changed and impacted the extent nicians interpret test results and initiate further actions. of testing gaps. International guidelines to recommend Electronic medical records with standardised forms, easy tumour testing of CRC patients have emerged since access to reliable knowledge, and clinical decision sup- 2009 [4, 5, 7], and universal testing was thought to port systems may also assist effective integration of gen- be in place for six of the seven hospitals in this study omics into clinical workflows [21]. However, appropriate by 2017 (and in all seven by 2018). However, testing referral and germline genetic testing also involves practice could have been further improved by the obtaining informed consent from patients, requiring Australasian Gastrointestinal Pathology Society con- both genomics knowledge and communication/counsel- sensus practice guidelines [8], which endorsed univer- ling skills. sal testing of all CRC patients. Thus, it will be A previous study in the USA also suggested that important to continue examining LS tumour testing the involvement of genetics services earlier in the and referral practice in Australia beyond 2018. In testing and referral pathway could substantially im- clinical practice, improved data management and re- prove referral rates: in that context, referral to genet- search governance could facilitate internal hospital ics services was both faster and more comprehensive audits and encourage learning health systems with during a period when tumour test results were imme- better integration of research and clinical practice diately sent to both colorectal surgeons and genetic [25]. Meanwhile, research data as presented here can counsellors, and the genetic counsellors contacted pa- help showcase best-practice achievements. tients directly [22]. By contrast, when only surgeons A limitation of our study is that patients with meta- received the tumour test results, about 1 in 2 CRC static cancer and no CRC resection were not included. patients with tumour dMMR or MSI in that hospital While we employed a consensus approach to ensure were not referred to genetics services, similar to the between-hospital data comparability (see Methods, Add- average rate observed in our study. In alignment with itional Files 3, 7 and 8), some information for patients past reviews [23] and the Australian Medical Associ- undergoing CRC resections may have been missed due ation 2020 statement on Genetic Testing and Genom- to data extraction errors or incomplete records. To miti- ics in Medicine [24], this also suggests that training gate this, data extraction was carried out by healthser- of clinicians is key to successful integration of vice professionals employed within the healthcare Steinberg et al. Hereditary Cancer in Clinical Practice (2022) 20:18 Page 7 of 8 system, extensively consulting hospital stakeholders to alternative to identify colorectal resections (in conjunction with ICD optimise data access and coverage, and combining infor- codes for CRC diagnosis). mation from multiple systems where necessary. Finally, we could not identify where referral to genetics services Acknowledgements Not applicable. was declined by a patient or not possible due to their ill- health/death (more likely among older patients, for Authors’ contributions whom referral records were less common). While these Conceptualization: N.T., J.S. Data curation: J.S., P.C., G.T., A.M., R.V., L.T., L.S., A.P., C.N., S.M., A.H., N.E. Formal analysis, methodology, and visualization: J.S. factors could preclude some referrals, missed LS diagno- Project administration: E.H., P.C., G.T. Funding acquisition: N.T., K.T., F.M., A.G., ses also reduce prevention and surveillance opportunities K.C., N.P., B.P., M.S., J.K. Writing – original draft: J.S. Writing – review & editing: for patients’ relatives, which was identified as particularly all authors. The author(s) read and approved the final manuscript. important by a patient representative who was part of Funding the study team. This study is funded by Cancer Institute NSW (2017/CDF005) and Cancer Our work also has notable strengths: inclusion of mul- Australia (1123924). The contents of this publication are solely the responsibility of the authors and do not reflect the views of Cancer Australia. tiple hospitals in different Australian states, use of clin- ical data and electronic health records where available, Availability of data and materials and identification of referrals using data from genetics Individual-level patient data are not publicly available to preserve patients’ services. privacy. Reasonable requests for data can be submitted to the corresponding author and will be subject to ethics committee and individual hospital governance requirements. Conclusions Declarations Tumour testing for LS is widely adopted at hospitals in Australia, but we have found that gaps in practice re- Ethics approval and consent to participate Ethics approval for this study was granted by the Royal Prince Alfred Hospital main at many hospitals in different states. Further work Human Research Ethics Committee (reference HREC/17/RPAH/542); the is needed to identify the procedures, funding structures, ethics committee approved a waiver of consent from individual patients to and targeted implementation strategies that can ensure collect the clinical data. Site-specific governance approval was obtained from each hospital site. successful implementation of universal testing in the LS context, and inform approaches for future broader inte- Consent for publication gration of genomic testing pathways into clinical Not applicable. practice. Competing interests KC is co-PI of unrelated investigator-initiated trial of cervical screening in Abbreviations Australia (‘Compass’) conducted by the Victorian Cytology Service, which has LS: Lynch syndrome; CRC: Colorectal cancer; dMMR: Deficient mismatch received funding contribution from Roche Molecular Systems and Ventana, repair; MMR: Mismatch repair; MSI: Microsatellite instability; USA. The other authors declare no potential conflicts of interest. IHC: Immunohistochemistry; PCR: Polymerase chain reaction; MDT: Multidisciplinary team; GP: General practicioner; ICD: International Author details Classification of Diseases; MBS: Medicare Benefit Schedule; ACHI: Australian The Daffodil Centre, The University of Sydney, a joint venture with Cancer Classification of Health Interventions Council NSW, 153 Dowling St, NSW 2011 Woolloomooloo, Australia. Surgical Outcomes Research Centre (SOuRCe), Royal Prince Alfred Hospital, Camperdown, NSW, Australia. NSW Health, Western Sydney Local Health Supplementary information District, Westmead, NSW, Australia. Monash Health, Melbourne, Victoria, The online version contains supplementary material available at https://doi. 5 6 Australia. Cancer Council NSW, Sydney, NSW, Australia. Genetic Services of org/10.1186/s13053-022-00225-1. Western Australia, King Edward Memorial Hospital, Perth, Western Australia, Australia. The Royal Melbourne Hospital, Melbourne, Victoria, Australia. School of Medicine and Public Health, University of Newcastle, Newcastle, Additional file 1: Supplementary Notes. NSW, Australia. The University of Sydney, Northern Clinical School Royal Additional file 2: Conceptual map of a step-wise tumour testing ap- North Shore Hospital, Sydney, NSW, Australia. Personalised Oncology proach for LS. Division, Walter and Eliza Hall Institute, Melbourne, Victoria, Australia. Additional file 3: Supplementary Methods. Cancer Genetics Department, Royal Prince Alfred Hospital, Camperdown, NSW, Australia. Department of Tissue Pathology & Diagnostic Oncology, Additional file 4: P-values for pairwise differences for missing test NSW Health Pathology, Royal Prince Alfred Hospital, Camperdown, NSW, results and referrals to genetics services between hospitals. Australia. Faculty of Medicine and Health, The University of Sydney, Sydney, Additional file 5: Diagnostic genetic testing among patients with NSW, Australia. School of Medicine and Pharmacology, University of complete, high-risk tumour test results who had a referral to genetic Western Australia, Perth, Western, Australia. Centre for the Health Economy, services. Macquarie University, Sydney, NSW, Australia. Westmead Familial Cancer Additional file 6: Examples of different terminology for LS tumour test Services, The Crown Princess Mary Cancer Centre, Westmead Hospital, results used in pathology reports. Westmead, NSW, Australia. Department of Colorectal Surgery, Royal Prince Alfred Hospital, Camperdown, NSW, Australia. Department of Colorectal Additional file 7: Consensus list of Medicare Benefit Schedule (MBS) Surgery, The University of Sydney, Westmead Hospital, Westmead, NSW, procedure codes used to identify colorectal resections (in conjunction Australia. Fiona Stanley Hospital, South Metropolitan Health Service, with ICD codes for CRC diagnosis). Murdoch, Western Australia, Australia. Sydney Medical School, University of Additional file 8: Australian Classification of Health Interventions (ACHI) Sydney, Sydney, NSW, Australia. Colorectal Medicine and Genetics, procedure codes mapped to consensus list of MBS items as an Department of Medicine, The Royal Melbourne Hospital, Melbourne, Victoria, Steinberg et al. Hereditary Cancer in Clinical Practice (2022) 20:18 Page 8 of 8 Australia. Hereditary Cancer Clinic, Prince of Wales Hospital, Sydney, NSW, 19. Khoury MJ. No Shortcuts on the Long Road to Evidence-Based Genomic Australia. Prince of Wales Clinical School, UNSW Sydney, NSW, Australia. Medicine. JAMA. 2017;318(1):27–8. School of Population Health, Faculty of Medicine, UNSW Sydney, NSW, 20. Roberts MC, Kennedy AE, Chambers DA, Khoury MJ. The current state of Australia. implementation science in genomic medicine: opportunities for improvement. Genet Sci. 2017;19(8):858–63. Received: 2 February 2022 Accepted: 10 April 2022 21. Williams MS, Taylor CO, Walton NA, Goehringer SR, Aronson S, Freimuth RR, et al. Genomic Information for Clinicians in the Electronic Health Record: Lessons Learned From the Clinical Genome Resource Project and the Electronic Medical Records and Genomics Network. Front Genet. 2019;10: References 1. Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, et al. 22. Heald B, Plesec T, Liu X, Pai R, Patil D, Moline J, et al. Implementation of Global cancer statistics 2020: GLOBOCAN estimates of incidence and universal microsatellite instability and immunohistochemistry screening for mortality worldwide for 36 cancers in 185 countries. CA: A Cancer Journal diagnosing lynch syndrome in a large academic medical center. J Clin for Clinicians; 2021. Oncol. 2013;31(10):1336–40. 2. U.S National Library of Medicine. Lynch Syndrome 2021. 2020. https://ghr. 23. Crellin E, McClaren B, Nisselle A, Best S, Gaff C, Metcalfe S. Preparing Medical nlm.nih.gov/condition/lynch-syndrome. Accessed 20 Apr 2022. Specialists to Practice Genomic Medicine: Education an Essential Part of a 3. Barrow P, Khan M, Lalloo F, Evans DG, Hill J. Systematic review of the Broader Strategy. Front Genet. 2019;10:789. impact of registration and screening on colorectal cancer incidence and 24. Australian Medical Association. Genetic Testing and Genomics in Medicine mortality in familial adenomatous polyposis and Lynch syndrome. Br J Surg. 2020. 2020. https://www.ama.com.au/articles/genetic-testing-and-genomics- 2013;100(13):1719–31. medicine-2020. Accessed 20 Apr 2022. 4. Evaluation of Genomic Applications in P, Prevention Working Group. 25. Clay-Williams R, Taylor N, Braithwaite J. Potential solutions to improve the Recommendations from the EGAPP Working Group: genetic testing governance of multicentre health services research. Med J Aust. 2018;208(4): strategies in newly diagnosed individuals with colorectal cancer aimed at 152–4. reducing morbidity and mortality from Lynch syndrome in relatives. Genet Med. 2009;11(1):35–41. 5. Provenzale D, Gupta S, Ahnen DJ, Bray T, Cannon JA, Cooper G, et al. Publisher’sNote Genetic/Familial High-Risk Assessment: Colorectal Version 1.2016, NCCN Springer Nature remains neutral with regard to jurisdictional claims in Clinical Practice Guidelines in Oncology. J Natl Compr Canc Netw. 2016; published maps and institutional affiliations. 14(8):1010–30. 6. Tognetto A, Michelazzo MB, Calabró GE, Unim B, Di Marco M, Ricciardi W, et al. A Systematic Review on the Existing Screening Pathways for Lynch Syndrome Identification. Front Public Health. 2017;5:243. 7. Vasen HF, Blanco I, Aktan-Collan K, Gopie JP, Alonso A, Aretz S, et al. Revised guidelines for the clinical management of Lynch syndrome (HNPCC): recommendations by a group of European experts. Gut. 2013;62(6):812–23. 8. Yozu M, Kumarasinghe MP, Brown IS, Gill AJ, Rosty C. Australasian Gastrointestinal Pathology Society (AGPS) consensus guidelines for universal defective mismatch repair testing in colorectal carcinoma. Pathology. 2019; 51(3):233–9. 9. National Institute for Health and Care Excellence. Molecular testing strategies for Lynch syndrome in people with colorectal cancer 2017. 2017. https://www.nice.org.uk/guidance/dg27. Accessed 20 Apr 2022. 10. Rahm AK, Cragun D, Hunter JE, Epstein MM, Lowery J, Lu CY, et al. Implementing universal Lynch syndrome screening (IMPULSS): protocol for a multi-site study to identify strategies to implement, adapt, and sustain genomic medicine programs in different organizational contexts. BMC Health Serv Res. 2018;18(1):824. 11. Lynch HT, Lynch PM, Lanspa SJ, Snyder CL, Lynch JF, Boland CR. Review of the Lynch syndrome: history, molecular genetics, screening, differential diagnosis, and medicolegal ramifications. Clin Genet. 2009;76(1):1–18. 12. Australian Government Department of Health. Medicare Benefits Schedule - Item 73354. 2020. http://www9.health.gov.au/mbs/fullDisplay.cfm?type= item&q=73354&qt=item. Accessed 20 Apr 2022. 13. Australian Government Department of Health. Genomics Health Futures Mission 2021. 2021. https://www.health.gov.au/initiatives-and-programs/ genomics-health-futures-mission. Accessed 20 Apr 2022. 14. Schofield L, Grieu F, Amanuel B, Carrello A, Spagnolo D, Kiraly C, et al. Population-based screening for Lynch syndrome in Western Australia. Int J Cancer. 2014;135(5):1085–91. 15. Brennan B, Hemmings CT, Clark I, Yip D, Fadia M, Taupin DR. Universal molecular screening does not effectively detect Lynch syndrome in clinical practice. Th Adv Gastroenterol. 2017;10(4):361–71. 16. Long JC, Debono D, Williams R, Salisbury E, O’Neill S, Eykman E, et al. Using behaviour change and implementation science to address low referral rates in oncology. BMC Health Serv Res. 2018;18(1):904. 17. Loh Z, Williams DS, Salmon L, Dow E, John T. Impact of universal immunohistochemistry on Lynch syndrome diagnosis in an Australian colorectal cancer cohort. Intern Med J. 2019;49(10):1278–84. 18. Morrow A, Hogden E, Kang YJ, Steinberg J, Canfell K, Solomon MJ, et al. Comparing theory and non-theory based implementation approaches to improving referral practices in cancer genetics: a cluster randomised trial protocol. Trials. 2019;20(1):373.

Journal

Hereditary Cancer in Clinical PracticeSpringer Journals

Published: May 4, 2022

Keywords: Lynch syndrome; Mismatch repair; Tumour testing; Genetics services referral; Bottleneck; Gap; Heterogeneity in practice; Medical records

There are no references for this article.