Diagnostic and health service pathways to diagnosis of cancer-registry notified cancer of unknown primary site (CUP).

BACKGROUND: Cancer of unknown primary (CUP) is a late-stage malignancy with poor prognosis, but we know little about what diagnostic tests and procedures people with CUP receive prior to diagnosis. The purpose of this study was to determine how health service utilisation prior to diagnosis for people with cancer-registry notified CUP differs from those notified with metastatic cancer of known primary.
METHODS: We identified people with a cancer registry notification of CUP (n = 327) from the 45 and Up Study, a prospective cohort of 266,724 people ≥45 years in New South Wales, Australia, matched with up to three controls with a diagnosis of metastatic cancer of known primary (n = 977). Baseline questionnaire data were linked to population health data to identify all health service use, diagnostic tests, and procedures in the month of diagnosis and 3 months prior. We used conditional logistic regression to estimate adjusted odds ratios (ORs) and 95% confidence intervals (CIs).
RESULTS: After adjusting for age and educational attainment, people with a cancer-registry notified CUP diagnosis were more likely to be an aged care resident (OR = 2.78, 95%CI 1.37-5.63), have an emergency department visit (OR = 1.65, 95%CI 1.23-2.21), serum tumor marker tests (OR = 1.51, 95%CI 1.12-2.04), or a cytology test without immunohistochemistry (OR = 2.01, 95%CI 1.47-2.76), and less likely to have a histopathology test without immunohistochemistry (OR = 0.43, 95%CI 0.31-0.59). Neither general practitioner, specialist, allied health practitioner or nurse consultations, hospitalisations, nor imaging procedures were associated with a CUP diagnosis.
CONCLUSIONS: The health service and diagnostic pathway to diagnosis differs markedly for people notified with CUP compared to those with metastatic cancer of known primary. While these differences may indicate missed opportunities for earlier detection and appropriate management, for some patients they may be clinically appropriate.

Introduction

Cancer of unknown primary site (CUP) is the 14th most commonly diagnosed cancer and the 6th most common cause of cancer death in Australia.[1] It is characterized by metastatic disease with an unidentified primary site, and extremely poor survival;[2-5] in Australia, the 5-year survival rate is 13%.[6] Population-based CUP incidence rates have declined over time,[2,6,7] a trend that can most likely be attributed to advances in diagnostic investigations. While a diagnosis of CUP ideally should be limited to people with a histological confirmation of metastatic cancer, in whom thorough testing has failed to identify the site of the primary tumor, people registered with a diagnosis of CUP in population-based cancer registries are heterogeneous, with many people receiving a diagnosis based on clinical examination only.[4,5,8-11] Thus, it is difficult to compare population-based studies of registry-notified CUP to cases series of “true CUP” cases, who have received exhaustive investigations.

Studies of how people diagnosed with CUP interact with the health system prior to their definitive diagnosis, and what investigations they receive, are limited.[8,12] Although there are no Australian CUP clinical practice guidelines, the US National Comprehensive Cancer Network (NCCN) and the European Society for Medical Oncology (ESMO) CUP-specific guidelines support the judicious use of diagnostic investigations, in keeping with a patient’s prognosis and treatment options.[13,14] Yet, the diagnostic pathway of patients with a cancer-registry notified CUP diagnosis is highly variable,[8,15,16] which may reflect heterogeneity in clinical presentation and patient characteristics.

Previous studies suggest that the diagnostic pathway for people with CUP differs compared to people diagnosed with other cancers, including late-stage cancers of known primary site.[8] CUP appears more likely to be diagnosed following an emergency department visit compared with all other cancers, and diagnosis may involve less specialist input and fewer invasive diagnostic procedures, including histopathological diagnosis of cancer.[8,15] In this prospective Australian cohort study, we sought to compare the pathway to diagnosis for people diagnosed with a cancer-registry CUP diagnosis, compared with people diagnosed with metastatic cancer of known primary site.

Materials and methods

Data sources

The Sax Institute’s 45 and Up Study[17] is a prospective cohort study with comprehensive information on self-reported lifestyle behaviors and a range of health, functional and social measures at baseline. New South Wales (NSW) residents aged at least 45 years were randomly sampled from the Department of Human Services (formerly Medicare Australia) enrollment database, which provides near complete coverage of the population. People 80+ years of age and residents of rural and remote areas were oversampled. A total of 266 933 individuals joined the study by completing a postal questionnaire between 2006 and 2009. The date of completion of the questionnaire was considered the baseline period. Around 18% of those invited participated, and the cohort included 11% of the NSW population aged 45 years or more.

Australia’s publicly funded health care system provides all citizens and permanent residents with a range of health services including treatment in public hospitals, subsidized treatment in private hospitals, subsidized outpatient services including consultations, procedures and tests, and subsidized medicines. Records of these transactions are made available for ethically approved health research. The 45 and Up Study cohort was probabilistically linked to population-based administrative health datasets by the Centre for Health Record Linkage to identify incident cancers, comorbid health conditions, subsidized health services and deaths. The datasets were: (i) the NSW Cancer Registry, a population-based registry of incident invasive cancer diagnoses (excluding basal and squamous cell carcinoma of the skin) in NSW 1994–2012; (ii) the NSW Admitted Patients Data Collection 2001–2015; (iii) the NSW Emergency Department Data Collection 2005–2016; (iv) the Medicare Benefits Scheme (MBS) 2001–2015; (v) the Pharmaceutical Benefits Scheme (PBS) 2004–2015; and (vi) the NSW Registry of Births, Deaths and Marriages 2006–2016. We excluded 209 cohort participants because they did not have a linked MBS or PBS record.

Study population

As previously described, we defined cases as cohort participants with a cancer registry diagnosis of CUP (ICD-10-AM codes C80, C76, C26 or C39) more than 3 months after baseline.[18,19] We included all people with a registry diagnosis of CUP, regardless of whether it was histopathologically confirmed. We randomly selected a control group with a cancer registry diagnosis of solid metastatic cancer of known primary site; for the controls, the first manifestation of this cancer was metastatic disease, with a recorded extent of disease spread that was either distant or regional, on the basis of the highest degree of spread (extent of spread of cancer from its point of origin) classified by the NSW Cancer Registry. Individuals with another registered cancer diagnosis in the same month were also eligible for inclusion as a case or control.

We matched the controls to cases by month and year of completion of the baseline questionnaire in the cohort and by month and year of cancer diagnosis. We selected up to three controls per case using incidence density sampling with replacement.[20] We did not match on age and sex as we wished to examine the effect of these factors.

Health service use

We quantified health services accessed in the month of diagnosis and the three months prior to diagnosis, including hospitalisations, emergency department (ED) visits, and consultations with general practitioners (GP), specialists, allied health practitioners, and nurses, and residence in an aged care facility. As the NSW Cancer Registry only provides the month and year of cancer diagnosis, we ascertained all health services in the entire month of diagnosis. We focused on the month of diagnosis and the three months prior to diagnosis based on our previous work showing that health service utilisation increased around this time [8], and clinical advice that this was the most relevant time period for diagnostic investigation. We confirmed that these observations held true for our cohort (data not shown).

We also identified the subset of GP consultations in the home, an institution or hospital (MBS item numbers 4, 24, 37, 47, 58, 59, 60, 65, 5003, 5023, 5043, 5063, 5220, 5223, 5227, 5228) as a marker of patient frailty; we also distinguished GP consultations that occurred in the doctors’ rooms (doctor’s offices) (MBS item numbers 2, 3, 23, 35, 36, 44, 52, 53, 54, 57, 2501, 2504, 2517, 2521, 2525, 2546, 2552, 2721, 2725, 5000, 5020, 5040, 5060). We further distinguished GP consultations for the preparation, contribution or review of a GP management plan or multidisciplinary/team care plan (MBS item numbers 721, 723, 729, 731, 732, 735, 739, 743, 747, 750, 758, 820, 822, 823, 825, 826, 828, 830, 832, 834, 835, 837, 838, 900). We separately identified specialist consultations for the initial assessment or review of patients with at least two comorbidities, classified as complex cases (MBS item numbers 132, 133). All health service use variables were dichotomized (any vs none).

Diagnostic tests and procedures

We ascertained all cancer-related imaging, endoscopy, medical procedures and pathology tests performed in tertiary and community settings during the month of diagnosis and the three months prior. We identified X-rays, computerized tomography (CT) scans, single photon emission computerized tomography (SPECT), ultrasound, magnetic resonance imaging (MRI), nuclear imaging and endoscopy. We classified medical procedures as exploratory surgery, non-surgical resection (such as fine needle aspiration, biopsy), and surgical resection (both non-cutaneous and cutaneous). The pathology tests of interest were one or more serum tumor markers (beta-2 microglobulin, alpha-fetoprotein, ca-15.3 antigen, ca-125 antigen, ca-19.9 antigen, cancer associated serum antigen, carcinoembryonic antigen, human chorionic gonadotrophin, neuron specific enolase, thyroglobulin, and prostate specific antigen; MBS item numbers 66629, 66650, 66651, 66652, 66653, 66655), cytology and histopathology alone or with immunostaining (immunocytochemistry or immunohistochemistry), and cytogenetics. These tests are most commonly used to help identify the tissue of origin for CUP.[14,21]

Statistical analysis

We used conditional logistic regression to estimate the odds of a cancer-registry notified CUP diagnosis associated with health service use and diagnostic tests and procedures. We first modelled each factor individually adjusted by age and sex only, and those variables with p<0.20 were considered for inclusion in the fully adjusted multivariable model. We used the 45 and Up Study baseline questionnaire to identify potential confounding factors, such as self-reported overall health and thus fitness for diagnostic investigation. To avoid multicollinearity between similar measures, we assessed the association between pairs of factors using Cramér's V statistic and those with a correlation coefficient ≥0.25 were considered correlated. This approach was also used to identify tests and procedures that were on the pathway between consultations and visits and a cancer diagnosis, to avoid over-adjustment.

We built conditional logistic regression models using backward elimination, stopping when the remaining variables in the model were all significantly associated with CUP (p<0.05). We built multiple multivariable models with all possible combinations of non-correlated variables, and then selected the model with the lowest Akaike Information Criterion as the final model.

Ethics

The study was approved by the NSW Population and Health Services and Human Research Ethics Committee (2012/11/428) and the 45 and Up Study was approved by the University of New South Wales Human Research Ethics Committee (HREC 15408). All participants provided written informed consent at the time of recruitment for follow-up and linkage of their information to administrative health databases. All procedures were in accordance with the ethical standards of the ethics committees mentioned above and with the 1964 Helsinki declaration and its later amendments.

Results

We identified 327 incident cases of CUP and 977 matched incident solid metastatic cancer controls, diagnosed between 2006 and 2012. The median age at diagnosis of CUP was 76 years (interquartile range, IQR: 66–82 years) and 68 years (IQR: 60–76 years) for solid metastatic cancer controls. The median time from completion of the baseline questionnaire to cancer diagnosis was 33 months (IQR: 21–46 months). Of the people with a registry-notified CUP diagnosis, 165 (50.5%) had a histopathogical diagnosis, 49 (15.0%) a cytological diagnosis (including fine needle aspiration, smears, washing and sputum), 88 (26.9%) a clinical diagnosis (including clinical, imaging and biochemical procedures) and 25 (7.6%) were identified by death certificate only. Of the CUP cases (n = 165) registered with a NSW Cancer Registry histopathological diagnosis, 20 (12.1%) tumors were carcinoma (subtype not specified), 81 (49.1%) were carcinoma (subtype specified), 50 (30.3%) were adenocarcinoma, and 14 (8.5%) were other morphological types. Of the 870 people with metastatic cancer of known primary with a histopathological diagnosis, 407 (46.8%) tumors were carcinoma, 395 (45.4%) were adenocarcinoma, and 68 (7.8%) were other. The most common primary sites for the solid metastatic cancer controls were breast (C50; n = 168), bronchus and lung (C34; n = 163), colon (C18; n = 152), prostate (C61; n = 123) and rectum (C20; n = 57).

In the month of diagnosis and the three months prior, people with a cancer-registry notified CUP diagnosis were more likely to have an ED visit (odds ratio, OR = 1.84, 95% confidence interval, CI 1.41–2.40) and less likely to have a specialist consultation (OR = 0.66, 95%CI 0.48–0.92), compared to solid metastatic cancer controls (Table 1) in models adjusted for age and sex only. During this period there was no difference between the two metastatic cancer groups in terms of hospitalisation or consultations with a GP, allied health practitioner, nurse, or specialist indicating complex care. People diagnosed with CUP were more than 3-fold more likely to be in aged care (OR = 3.42, 95%CI 1.78–6.56). We found people diagnosed with CUP were less likely than controls to have an endoscopy, a surgical (non-cutaneous) resection, and histopathology (Table 2). Conversely, they were more likely to have serum tumor marker tests and cytology. Overall, 21 (6.4%) people diagnosed with CUP did not have any cancer-related investigations, as defined by imaging procedures, endoscopy, medical procedures (e.g. resection), or pathology tests. Among controls, 60 (6.1%) did not have any cancer-related investigations.

Table 1

Association between recent health service use and diagnosis of CUP compared to metastatic cancer of known primary.

Health service use in three months prior and month of diagnosis	CUP (n = 327)	Metastatic cancer, known primary site (n = 977)	Age- and sex-adjusted OR (95% CI)
	n (%)	n (%)
Tertiary care
Hospitalisation	221 (67.6)	682 (69.8)	0.88 (0.66–1.17)
Emergency department visit	172 (52.6)	316 (32.3)	1.84 (1.41–2.40)
Consultations
General practitioner (GP)	305 (93.3)	932 (95.3)	0.79 (0.44–1.42)
Consulting room (including after-hours visits)a	297 (90.8)	920 (94.2)	0.78 (0.47–1.29)
Home, institution or hospital (including after-hours visits)b	49 (15.0)	79 (8.1)	1.43 (0.96–2.13)
Management and multidisciplinary care plans	64 (19.6)	143 (14.6)	1.27 (0.91–1.78)
Specialist/consultant physician	244 (74.6)	825 (84.4)	0.66 (0.48–0.92)
Complex case	51 (15.6)	121 (12.4)	1.30 (0.89–1.90)
Allied health practitioner	82 (25.1)	245 (25.1)	0.96 (0.71–1.31)
Nurse	39 (11.9)	127 (13.0)	0.84 (0.56–1.26)
Residence in aged care facilityc	32 (9.8)	19 (1.9)	3.42 (1.78–6.56)

a MBS item numbers 2, 3, 23, 35, 36, 44, 52, 53, 54, 57, 2501, 2504, 2517, 2521, 2525, 2546, 2552, 2721, 2725, 5000, 5020, 5040, 5060.

b MBS item numbers 4, 24, 37, 47, 58, 59, 60, 65, 5003, 5023, 5043, 5063, 5220, 5223, 5227, 5228.

c EDDC referral source = 5 (Residential Aged Care facility); APDC: source of referral 6 (Nursing home/Residential Aged Care Facility) or mode of separation = 3 (transferred to nursing home) or peer group = F2 (Nursing home) and MBS item numbers: 20, 35, 43, 51, 92, 93, 95, 96, 5010, 5028, 5049, 5067, 5260, 5263, 5265, 5267, 2125, 2138, 2179, 2220, 82223, 82224, 82225, 73934, 73935, 10984, 903, 731.

Table 2

Association between recent cancer-related diagnostic tests and procedures and diagnosis of CUP compared to metastatic cancer of known primary.

Tests or procedures in month of diagnosis and three months prior	CUP (n = 327)	Metastatic cancer, known primary (n = 977)	Age- and sex-adjusted OR (95% CI)
	n (%)	n (%)
Imaging procedures
X-ray	165 (50.5)	493 (50.5)	0.99 (0.76–1.29)
Computerized tomography (CT)	213 (65.1)	618 (63.3)	1.17 (0.89–1.55)
Single photon emission computerized tomography (SPECT)	49 (15.0)	159 (16.3)	1.17 (0.81–1.69)
Ultrasound	141 (43.1)	436 (44.6)	1.24 (0.94–1.62)
Magnetic resonance imaging (MRI)	14 (4.3)	38 (3.9)	1.37 (0.71–2.62)
Nuclear imaging	50 (15.3)	174 (17.8)	1.07 (0.74–1.53)
Endoscopy	53 (16.2)	277 (28.4)	0.55 (0.39–0.77)
Medical procedures
Exploratory surgerya	c	c	1.36 (0.21–8.71)
Resection, non-surgicalb	290 (88.7)	882 (90.3)	1.05 (0.69–1.60)
Resection, surgical (non-cutaneous)	20 (6.1)	158 (16.2)	0.46 (0.28–0.75)
Resection, surgical (cutaneous)	27 (8.3)	68 (7.0)	1.12 (0.68–1.85)
Pathology tests
Serum tumor markers	126 (38.6)	327 (33.5)	1.36 (1.04–1.78)
Cytology without immunohistochemistry	112 (34.3)	256 (26.2)	1.78 (1.33–2.38)
Cytology with immunocytochemistry	25 (7.6)	36 (3.7)	2.54 (1.45–4.46)
Histopathology without immunohistochemistry	101 (30.9)	561 (57.4)	0.40 (0.29–0.53)
Histopathology with immunohistochemistry	59 (18.0)	252 (25.8)	0.77 (0.55–1.08)
Cytogenetics	c	c	5.11 (0.79–33.1)

a Exploratory surgery includes laparotomy, thoracotomy, and cervical exploration of mediastinum

b Non-surgical resection includes fine needle aspiration, biopsy or excision, other than regional or radical excision which has been classified as surgical resection

c Cell size <5; exact cell size suppressed for privacy reasons

In our fully adjusted model, the only confounding factors were age and educational attainment, with those diagnosed with CUP more likely to be older and to have no school certificate (less than 4 years of secondary education) (Table 3).[18] After adjustment for these factors, and the mutual adjustment for health service use, diagnostic tests and procedures, the only variables that remained associated with an increased probability of a CUP diagnosis were being in an aged care facility (OR = 2.78, 95%CI 1.37–5.63), ≥1 ED visit (OR = 1.65, 95%CI 1.23–2.21), and the following pathology tests: serum tumor marker tests (OR = 1.51, 95%CI 1.12–2.04), and cytology without immunohistochemistry (OR = 2.01, 95%CI 1.47–2.76). People diagnosed with CUP were less than half as likely to have had a histopathology test without immunohistochemistry (OR = 0.43, 95%CI 0.31–0.59).

Table 3

Association between recent health service use, cancer-related diagnostic tests and procedures, and diagnosis of CUP compared to metastatic cancer of known primary.

Tests or procedures in month of diagnosis and three months prior	Fully adjusted
	OR (95% CI)a
Tertiary care
≥1 emergency department visit	1.65 (1.23–2.21)
Aged care facility	2.78 (1.37–5.63)
Pathology tests
Serum tumor markers	1.51 (1.12–2.04)
Cytology without immunohistochemistry	2.01 (1.47–2.76)
Histopathology without immunohistochemistry	0.43 (0.31–0.59)

a Adjusted for age and educational attainment

Discussion

In a contemporary cohort of Australian adults, we observed differences in the pathways to diagnosis for people with a cancer-registry notified CUP diagnosis compared with people notified with metastatic cancer of known origin. People with CUP were more likely to be in aged care, and their pathway to diagnosis was more likely to involve an emergency presentation and the use of less invasive diagnostic tests. Many ED visits in people with cancer are avoidable;[22] however, the use of less invasive diagnostic tests, for example imaging and fine needle aspiration, may be appropriate for people who are frail (including many aged care residents [23]), or those with a poor prognosis, and this diagnostic pathway is supported by clinical guidelines for CUP [14,21].

This is one of the first studies to identify an increased risk of a CUP diagnosis in aged care residents.[8] Very few cohort studies have examined the association between aged care residency and cancer diagnosis as this information is not typically available at the population-level. A US cancer-registry study of women insured by Medicaid observed a 2.5-fold excess risk of late breast cancer diagnosis in women who were nursing home residents or were in a long-term care facility.[24] A more recent US study[25] found that very few Medicaid-insured nursing home patients received cancer services, and they exhibited a high prevalence of late or unstaged common cancers. While these studies may not be generalizable to the Australian context, it is known that on average, aged care residents are likely to be frail with potentially complex multimorbidity and health care needs. In addition, they may not have close family members to advocate on their behalf regarding changes in their health status. Frail patients may not tolerate cancer treatment due to the increased risk of toxicity and mortality [26], decreasing the need for comprehensive testing to identify the tissue of origin. Late diagnosis may not necessarily be a marker of inadequate care in this setting, as it is possible that functional impairment, care dependency, prognosis, and patient preferences were taken into account when considering the most appropriate diagnostic management of people with metastatic cancer of unknown origin, in keeping with clinical guidelines[13,21].

Confirming previous studies, we found that people with a cancer-registry CUP diagnosis were nearly twice as likely to have had an ED visit in the time period immediately preceding diagnosis, compared to people diagnosed with metastatic cancer of known primary. In our study, 53% of people diagnosed with CUP had a recent ED visit, which is similar to previous Australian studies (50–57%)[8,27] and a UK study of a national cancer registry (57%).[28] Cancer diagnosis via ED presentation is considered a marker for late or delayed diagnosis, and is generally associated with older age, greater deprivation, less access to health services, and poorer prognosis.[29-31] However, the reasons for presentation via ED are complex and many patients also have significant interaction with the health care system prior to their ED presentation, which may represent lost opportunities for diagnosis in individuals with good quality of life and life expectancy who are most likely to benefit from treatment.[31]

While the reasons for ED presentation in people diagnosed with CUP are not known to us, the non-specific and variable nature of CUP symptoms may lead to delayed diagnosis, even in patients who regularly interact with the health care system. As such, we saw no difference between CUP and metastatic cancer of known origin in terms of household income, rural residential location, hospitalisations, GP visits, or allied health visits in this cohort. Although this replicates our previous findings in a veterans cohort,[8] we acknowledge that both studies had limited statistical power, and suggest these associations deserve further scrutiny in large, well-annotated cohorts. It is important to understand the reasons for ED presentation prior to diagnosis, and whether there were missed opportunities for earlier diagnosis, specifically through earlier presentation to a GP, and identification of symptoms suggestive of cancer, leading to suitable investigations.

Only a small proportion of cases and controls had no cancer-related investigations (6%); however, the types of investigations between the two groups differed. People diagnosed with CUP had a greater use of less invasive tests such as serum tumor markers and cytology, and less use of histopathology, compared to controls. While the former are less invasive, the routine use of non-specific serum tumor markers is not recommended for patients with suspected cancer because they can be overexpressed in some people without cancer.[13,32] While histopathology is considered part of the diagnostic approach to identify the primary tumor in CUP patients,[32] only 50% of registered CUP cases in our study were histopathologically verified. This is lower than observed in other population-based studies of CUP[2,7,11,15,33] but higher than our previous study in Australian veterans (36%).[8,15] CUP is a heterogeneous diagnosis, and encompasses individuals who may have undergone exhaustive investigation but a primary site cannot be identified (i.e. people with “true CUP”), and also individuals with a clinical diagnosis only, for whom invasive tests are neither warranted nor desired.[13] In a 2019 US study, only 35% of elderly people with CUP received guideline recommended diagnostic evaluation.[33] Patients with a clinical diagnosis only tend to be older, and have poorer outcomes.[7,8] Unfortunately, we are unable to determine whether the diagnostic approach was clinically appropriate; however, it is recommended that investigations considered to have no impact on prognosis should be avoided.

The strength of this study is the comprehensiveness and quality of the data for a large, prospective cohort; incident cancers and deaths were ascertained by high-quality population-based registries, and we had near complete capture of all subsidized health service interactions for our population, including community-based consultations and tertiary care, diagnostic tests and procedures, as well as information on aged care. This allowed us to construct a comprehensive picture of people’s health service use prior to cancer diagnosis. While our sample size is one of the largest population-based studies of people with CUP[8,9], it is modest. As a result, we may have had insufficient power to detect an association with some health services, and where we did, the confidence intervals are wide. While 45 and Up Study participants are healthier on average than the general population, relative estimates calculated from within-cohort comparisons are valid.[34,35] Two data issues are also unlikely to have affected our within-cohort comparisons; firstly, the ED data mainly captured visits to metropolitan and large regional public hospitals (72% and 88% of ED presentations in NSW, respectively).[36,37] Secondly, some pathology services will be under-ascertained as MBS claims data only capture the three most expensive pathology items in an episode of care performed by a GP.[38]

People diagnosed with CUP represent people for whom the primary site could not be identified despite exhaustive testing, as well as those who received minimal diagnostic evaluation; unfortunately we could not distinguish these two groups based on the cancer registry data alone. Ideally, such distinct groups would receive distinguishable diagnoses, but this is not the case. Given the improvement in diagnostics over time, it is likely that the prevalence of “true CUP” will decrease, with most people diagnosed with CUP representing those who did not receive all clinically-indicated tests, and it is important that these people are not ignored in cancer research. Finally, we have made the assumption that the identified procedures and tests were diagnostic for cancer, but some procedures, particularly non-specific tests such as endoscopy and ultrasound, may have been performed for other reasons. Importantly, we could not directly measure frailty in our study. Frailty plays a role in health decisions and outcome, yet is a concept that is difficult to capture in administrative claims data; while many studies have tried they each have their limitations.[39,40] Our findings support the need for further research to elucidate the relationship between frailty, health service utilisation, and CUP diagnosis.

Conclusions

We have shown that people with a cancer-registry notified CUP diagnosis had a different pathway to diagnosis than patients with metastatic cancer of known primary. People diagnosed with CUP were more likely to be aged care residents, to have had an ED presentation immediately prior to diagnosis, and to have received fewer invasive tests. Interestingly, despite CUP patients having poorer self-reported overall health,[18] the two groups exhibited no difference in the likelihood of having at least one hospitalisation or consultation during the month of diagnosis and three months prior. Further research is required to examine whether these results indicate missed opportunities in CUP patients for earlier detection and diagnostic testing to reveal the tissue of origin, or conversely represent clinically appropriate differences in management based on patients’ underlying health and life expectancy.

6 Jan 2020

PONE-D-19-32621

Diagnostic and health service pathways to diagnosis of cancer-registry notified cancer of unknown primary site (CUP)

PLOS ONE

Dear Dr Schaffer,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please respond point-by-point to each comment raised by the reviewers. In terms of Reviewer 1's comments, I do not think you need both confidence intervals and p-values. Please choose one or the other.

We would appreciate receiving your revised manuscript by Feb 20 2020 11:59PM. When you are ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter.

To enhance the reproducibility of your results, we recommend that if applicable you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). This letter should be uploaded as separate file and labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. This file should be uploaded as separate file and labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. This file should be uploaded as separate file and labeled 'Manuscript'.

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

We look forward to receiving your revised manuscript.

Kind regards,

Erin Bowles

Academic Editor

PLOS ONE

Journal Requirements:

1. When submitting your revision, we need you to address these additional requirements.

Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

http://www.journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and http://www.journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. We note that you have indicated that data from this study are available upon request. PLOS only allows data to be available upon request if there are legal or ethical restrictions on sharing data publicly. For information on unacceptable data access restrictions, please see http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions.

In your revised cover letter, please address the following prompts:

a) If there are ethical or legal restrictions on sharing a de-identified data set, please explain them in detail (e.g., data contain potentially identifying or sensitive patient information) and who has imposed them (e.g., an ethics committee). Please also provide contact information for a data access committee, ethics committee, or other institutional body to which data requests may be sent.

b) If there are no restrictions, please upload the minimal anonymized data set necessary to replicate your study findings as either Supporting Information files or to a stable, public repository and provide us with the relevant URLs, DOIs, or accession numbers. Please see http://www.bmj.com/content/340/bmj.c181.long for guidelines on how to de-identify and prepare clinical data for publication. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories.

We will update your Data Availability statement on your behalf to reflect the information you provide.

3. Please provide additional details regarding participant consent. In the ethics statement in the Methods and online submission information, please ensure that you have specified (1) whether consent was suitably informed and (2) what type you obtained (for instance, written or verbal). If your study included minors under age 18, state whether you obtained consent from parents or guardians. If the need for consent was waived by the ethics committee, please include this information.

4. We noticed you have some minor occurrence(s) of overlapping text with the following previous publication(s), which needs to be addressed:

https://doi.org/10.1016/j.canep.2019.05.001

https://doi.org/10.1016/j.canep.2019.04.004

In your revision ensure you cite all your sources (including your own works), and quote or rephrase any duplicated text outside the Methods section. Further consideration is dependent on these concerns being addressed.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: The authors described their methods well and interpretation was guided by the data presented. I do suggest including p-values in tables 1 and 2 as inclusion of variables in the multivariable regression model was based on the p-value. The authors correct in presenting the 95% CI along with the ORs in the narrative but it would be helpful to see the p-values in the data tables. Overall, the paper is well written and the authors explained the rationale of the study well.

Reviewer #2: This paper is an important study examining how patients are diagnosed with cancer of unknown primary (CUP) compared to patients diagnosed with known solid tumors. This knowledge solidly adds to the literature base of CUP. Identification of ED visits confirmed previous VA findings in the general population but also identified aged care residents receipt of CUP diagnoses- care that may be appropriate. Additionally, the authors identify missed opportunity areas to reduce late diagnoses.

Line 79, pg 4, typo: ‘enrolment.’

18% response rate overall. Do you know what percentage of cancer registry persons responded, was it similar to the overall response rate? It seems the use of the survey results was minimal – do you get the same/similar results when not limited to cancer registry-survey patients?

I may have missed it but were the patients with solid tumors limited to late-stage also (stage III or IV)?

Can you clarify what is meant by “GP consultations in the doctors’ rooms” and how that is different in GP consultations in the institution or hospital - I assume it is a doctor's office visit and that is not captured by the institution category.

Pg 7 lines 175, it would be interesting to note the patients who had a clinical diagnosis It seems the results are not mutually exclusive categories (50% had histo, 49% cytological, and 42.2% clinical). The US SEER has about a third of CUP patients who have undergone guideline-recommended diagnostic tests- however, the US is not comparable to Australia because of the US socioeconomic disparities in access to healthcare. I'm curious to know if Australia does a better job. You have described the extent of diagnostic inquiry in more detail further in the manuscript, can you synthesize the findings?

Please clarify the reporting of carcinoma – does that mean that a more detailed assessment was not completed? Does that correlate with patients who only got cytology or clinical diagnosis? I'm assuming this would be comparable to a “not otherwise specified category,” suggesting incomplete evaluation.

Table 2; at first glance, I was confused about ‘<5^3’; took me two readings to figure that out; possibly identify a way to clarify like using parentheses? Also, type on footnote 3 (reaso).

I’d be curious to know if aged care residents are more likely to visit ED than non-aged care patients and if this was driving the ED finding. Does interaction/effect modification play a role here? The sample size may to too small to assess this, just curious.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: Yes: Julie Smith-Gagen

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files to be viewed.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org. Please note that Supporting Information files do not need this step.

19 Feb 2020

February 19, 2020

Dear Dr Bowles,

Thank you for the opportunity to revise and resubmit our manuscript, “Diagnostic and health service pathways to diagnosis of cancer-registry notified cancer of unknown primary site (CUP)”. We have ensured that we have cited all previous works using the same cohort, and that there is no overlapping text outside the Methods section.

The data used for this study are from the 45 and Up Study, which is an open research resource managed by the Sax Institute (https://www.saxinstitute.org.au/our-work/45-up-study/for-researchers/). The datasets for this study were created by linkage of the 45 and Up Study baseline survey data to Australian Government and NSW state data sources with support from the NSW Centre for Health Record Linkage (www.cherel.org.au), and permission from the custodians of the datasets under specific ethics approvals for pre-specified analyses only. These data contain sensitive information and are potentially re-identifiable, and we do not have permission to share them with third parties. The data used for this study are available to researchers pending appropriate ethics and data access requirements. Interested researchers can contact the Sax Institute (45andup.research@saxinstitute.org.au) and NSW Centre for Health Record Linkage (cherel.mail@moh.health.nsw.gov.au) for data access approval procedures.

We have also included further information about consent and ethics in the Methods section of the manuscript on Page 7.

We have addressed all reviewers’ comments as described below. A formatted version of our response is also included in the attached cover letter as a pdf.

On behalf of all co-authors,

Andrea Schaffer, MSc, MBiostat, PhD

Centre for Big Data Research in Health

Level 2, AGSM Building (G27)

UNSW Sydney

Sydney AUSTRALIA

******

Reviewer #1: The authors described their methods well and interpretation was guided by the data presented. I do suggest including p-values in tables 1 and 2 as inclusion of variables in the multivariable regression model was based on the p-value. The authors correct in presenting the 95% CI along with the ORs in the narrative but it would be helpful to see the p-values in the data tables. Overall, the paper is well written and the authors explained the rationale of the study well.

RESPONSE: Thank you to the reviewer for their positive comments. The p-value was only one consideration for choosing which covariates to include in the multivariable model, as we considered correlation between covariates and clinical expertise to avoid including variables that were on the same pathway to a CUP diagnosis. We believe that there is limited utility in including p-values in addition to confidence intervals. As advised by the Editors we have chosen to present confidence intervals only.

*******

Reviewer #2: This paper is an important study examining how patients are diagnosed with cancer of unknown primary (CUP) compared to patients diagnosed with known solid tumors. This knowledge solidly adds to the literature base of CUP. Identification of ED visits confirmed previous VA findings in the general population but also identified aged care residents receipt of CUP diagnoses- care that may be appropriate. Additionally, the authors identify missed opportunity areas to reduce late diagnoses.

RESPONSE: Thank you to the reviewer for her comments.

Line 79, pg 4, typo: ‘enrolment.’

RESPONSE: “Enrolment” is the standard spelling in British and Australian English. However, given that PLOS ONE is published by an American organisation we have now used standard American English spellings, including “enrollment”, throughout the manuscript.

18% response rate overall. Do you know what percentage of cancer registry persons responded, was it similar to the overall response rate?

RESPONSE: We do not know the response rate for people with a history of cancer and registered cancer. The decision to participate in the 45 and Up Study and completion of the baseline questionnaire occurred prior to the registry-notified cancer diagnoses in this study.

It seems the use of the survey results was minimal – do you get the same/similar results when not limited to cancer registry-survey patients?

RESPONSE: We did evaluate all survey results and found that information from the survey was minimally predictive of a CUP diagnosis (except for educational attainment) after including health service utilisation in our model. However, this may be because the median time between completion of the baseline questionnaire and the CUP diagnosis was 33 months. Further, while we had information on self-rated health and quality of life, they likely are measuring similar constructs as residence in aged care (which was the strongest predictor of a CUP diagnosis), specifically frailty and poorer health status.

We performed a similar study in Australian veterans (Vajdic et al. Health service utilisation and investigations before diagnosis of cancer of unknown primary (CUP): A population-based nested case-control study in Australian Government Department of Veterans’ Affairs clients. Cancer Epidemiol 2015;39:585-592) and also found that people with a registry-notified diagnosis of CUP were more likely to be in aged care compared with people with metastatic cancer of known primary. We have cited this study in the Discussion when discussing the findings from our current study.

I may have missed it but were the patients with solid tumors limited to late-stage also (stage III or IV)?

RESPONSE: Controls with known primary solid tumours were eligible if the first manifestation of their cancer was metastatic, with either regional or distant metastases, as classified by the NSW Cancer Registry, as described on Page 5. Note that we allowed regional metastases in the control group as there exist “true CUP” cases with regional metastases, e.g. neck node for presumed head and neck primary; axillary lymph node for presumed breast cancer primary. We did not have detailed information on cancer stage in our data.

Can you clarify what is meant by “GP consultations in the doctors’ rooms” and how that is different in GP consultations in the institution or hospital - I assume it is a doctor's office visit and that is not captured by the institution category.

RESPONSE: The reviewer is correct, GP consultations in doctors’ rooms is in the doctor’s office; the former is Australian terminology and is equivalent to “doctor’s office”. We have separated this measure from GP consultations in a home, institution or hospital as we expected that the latter would be a potential marker for frailty or poorer health, given that they represent visits to patients who are hospitalised, in aged care, or who cannot travel to the doctor’s office. We have now clarified in the Methods that “doctors’ rooms” is the same as “doctor’s office” (Page 6).

Pg 7 lines 175, it would be interesting to note the patients who had a clinical diagnosis It seems the results are not mutually exclusive categories (50% had histo, 49% cytological, and 42.2% clinical). The US SEER has about a third of CUP patients who have undergone guideline-recommended diagnostic tests- however, the US is not comparable to Australia because of the US socioeconomic disparities in access to healthcare. I'm curious to know if Australia does a better job. You have described the extent of diagnostic inquiry in more detail further in the manuscript, can you synthesize the findings?

RESPONSE: Thank you for noticing this error, as the categories should in fact be mutually exclusive. In the text we now describe the highest level of diagnostic enquiry:

Page 7: “Of the people with a registry-notified CUP diagnosis, 165 (50.5%) had a histopathogical diagnosis, 49 (15.0%) a cytological diagnosis (including fine needle aspiration, smears, washing and sputum), 88 (26.9%) a clinical diagnosis (including clinical, imaging and biochemical procedures) and 25 (7.6%) were identified by death certificate only."

There are no Australian guidelines for diagnosis of CUP. However, we have now included data on the number of people who did not have any cancer-related diagnostic procedures (i.e. imaging, surgery, pathology tests) in both the CUP cases and controls, and included this information in the Results:

Page 8: “Overall, 21 (6.4%) people diagnosed with CUP did not have any cancer-related investigations, as defined by imaging procedures, endoscopy, medical procedures (e.g. resection), or pathology tests. Among controls, 60 (6.1%) did not have any cancer-related investigations.”

We have also included the following sentence in the Discussion, regarding the disparity in types of procedures that the two groups underwent:

Page 14: “Only a small proportion of cases and controls had no cancer-related investigations (6%); however, the types of investigations between the two groups differed. People diagnosed with CUP had a greater use of less invasive tests such as serum tumor markers and cytology, and less use of histopathology, compared with controls.”

Lastly, we were not aware of the mentioned study at the time of submission of our manuscript; it is highly relevant and we have now cited it in the Discussion.

Please clarify the reporting of carcinoma – does that mean that a more detailed assessment was not completed? Does that correlate with patients who only got cytology or clinical diagnosis? I'm assuming this would be comparable to a “not otherwise specified category,” suggesting incomplete evaluation.

RESPONSE: Of the 165 people with a histopathological diagnosis, 101 (61%) were classified as carcinoma. Of these, 20 were carcinoma not otherwise specified and 81 were a specified carcinoma subtype. We are unable to report specific carcinoma subtypes because most subtypes have less than 5 cases and for privacy reasons we cannot report counts <5. We have modified the text to include the above information:

Page 8: “Of the CUP cases (n=165) registered with a NSW Cancer Registry histopathological diagnosis, 20 (12.1%) tumors were carcinoma (subtype not specified), 81 (49.1%) were carcinoma (subtype specified), 50 (30.3%) were adenocarcinoma, and 14 (8.5%) were other morphological types.”

Table 2; at first glance, I was confused about ‘<5^3’; took me two readings to figure that out; possibly identify a way to clarify like using parentheses? Also, type on footnote 3 (reaso).

RESPONSE: Thank you, we have corrected the typo. For clarity, we have now replaced all table footnote indicators with letters, rather than numbers, which may be confused with numerical content in the tables. “<5^3” in the cells has now been replaced with the letter “c” and we have included relevant information in the footnote.

I’d be curious to know if aged care residents are more likely to visit ED than non-aged care patients and if this was driving the ED finding. Does interaction/effect modification play a role here? The sample size may to too small to assess this, just curious.

RESPONSE: The increased risk of a CUP diagnosis in people with an ED visit is adjusted for residence in an aged care facility, and thus is unlikely to be driven by people in aged care. There was little correlation between residency in aged care and having had an ED visit (Cramér's V statistic=0.12), suggesting that the rate of ED visits was similar between those in aged care and those not. Lastly, detection of significant interactions requires a much larger sample size than for main effects, and as the reviewer correctly points out our sample size is likely too small to detect most interactions.

28 Feb 2020

Diagnostic and health service pathways to diagnosis of cancer-registry notified cancer of unknown primary site (CUP)

PONE-D-19-32621R1

Dear Dr. Schaffer,

We are pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it complies with all outstanding technical requirements.

Within one week, you will receive an e-mail containing information on the amendments required prior to publication. When all required modifications have been addressed, you will receive a formal acceptance letter and your manuscript will proceed to our production department and be scheduled for publication.

Shortly after the formal acceptance letter is sent, an invoice for payment will follow. To ensure an efficient production and billing process, please log into Editorial Manager at https://www.editorialmanager.com/pone/, click the "Update My Information" link at the top of the page, and update your user information. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to enable them to help maximize its impact. If they will be preparing press materials for this manuscript, you must inform our press team as soon as possible and no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

With kind regards,

Erin Bowles

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

4 Mar 2020

PONE-D-19-32621R1

Diagnostic and health service pathways to diagnosis of cancer-registry notified cancer of unknown primary site (CUP)

Dear Dr. Schaffer:

I am pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please notify them about your upcoming paper at this point, to enable them to help maximize its impact. If they will be preparing press materials for this manuscript, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

For any other questions or concerns, please email plosone@plos.org.

Thank you for submitting your work to PLOS ONE.

With kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Erin Bowles

Academic Editor

PLOS ONE