Cancer of unknown primary: ESMO Clinical Practice Guideline for diagnosis, treatment and follow-up

•This ESMO Clinical Practice Guideline provides key recommendations for managing cancer of unknown primary (CUP).•The guideline covers CUP diagnosis, risk assessment, classification, treatment and follow-up.•Differential diagnostic algorithms for discrimination between CUP and other cancers are provided.•The author group includes multidisciplinary experts from different institutions in Europe, Australia and the United States.•Recommendations are based on available scientific data and the authors’ collective expert opinion.•In clinical practice, all recommendations provided need to be discussed with patients in a shared decision-making approach. Cancer of unknown primary (CUP) is defined as a carcinoma or undifferentiated neoplasm for which a standardised diagnostic work-up fails to identify the primary tumour responsible for metastatic seeding. CUP accounts for <5% of cancers but, because of its high mortality rate, its relative contribution to cancer deaths is higher.1Brewster D.H. Lang J. Bhatti L.A. et al.Descriptive epidemiology of cancer of unknown primary site in Scotland, 1961-2010.Cancer Epidemiol. 2014; 38: 227-234Crossref PubMed Google Scholar The incidence of CUP has been declining, probably due to improving success in localising primary tumours.2Binder C. Matthes K.L. Korol D. et al.Cancer of unknown primary-Epidemiological trends and relevance of comprehensive genomic profiling.Cancer Med. 2018; 7: 4814-4824Crossref PubMed Scopus (22) Google Scholar The incidence increases with age and is higher in men compared with women. Adenocarcinoma is the most common histology. Approximately 50% of CUP cases can be categorised as well-differentiated to moderately differentiated adenocarcinomas, ∼30% as poorly differentiated adenocarcinomas or undifferentiated carcinomas, ∼15% as squamous-cell carcinomas and ∼5% as undifferentiated neoplasms.2Binder C. Matthes K.L. Korol D. et al.Cancer of unknown primary-Epidemiological trends and relevance of comprehensive genomic profiling.Cancer Med. 2018; 7: 4814-4824Crossref PubMed Scopus (22) Google Scholar,3Pavlidis N. Briasoulis E. Hainsworth J. et al.Diagnostic and therapeutic management of cancer of an unknown primary.Eur J Cancer. 2003; 39: 1990-2005Abstract Full Text Full Text PDF PubMed Scopus (369) Google Scholar Sarcomas, melanomas, germ cell tumours, neuroendocrine tumours and haematological malignancies whose exact site of origin is not established are not included in the CUP definition. Many patients present with metastases in multiple organs, such as the liver (most common), respiratory system, lymph nodes, abdominal cavity, bone and brain.1Brewster D.H. Lang J. Bhatti L.A. et al.Descriptive epidemiology of cancer of unknown primary site in Scotland, 1961-2010.Cancer Epidemiol. 2014; 38: 227-234Crossref PubMed Google Scholar The decrease in CUP incidence has been noted for most metastatic locations and histologies.1Brewster D.H. Lang J. Bhatti L.A. et al.Descriptive epidemiology of cancer of unknown primary site in Scotland, 1961-2010.Cancer Epidemiol. 2014; 38: 227-234Crossref PubMed Google Scholar CUP diagnostics include a meticulous search for the hidden primary cancer, which explains why no new primaries are diagnosed soon after a CUP diagnosis.4Hemminki K. Liu H. Heminki A. et al.Power and limits of modern cancer diagnostics: cancer of unknown primary.Ann Oncol. 2012; 23: 760-764Abstract Full Text Full Text PDF PubMed Scopus (23) Google Scholar The majority of patients with CUP will not have a primary lesion identified during the course of the disease. However, some survivors of CUP may develop (i) initially hidden primary tumours responsible for their metastatic disease or (ii) second primary cancers. Elevated risks for the development of several types of second primaries have been reported,5Shu X. Liu H. Ji J. et al.Subsequent cancers in patients diagnosed with cancer of unknown primary (CUP): etiological insights?.Ann Oncol. 2012; 23: 269-275Abstract Full Text Full Text PDF PubMed Scopus (0) Google Scholar with the highest risks observed for cancers of the small intestine, male genital organs and aerodigestive tract. Significant risks have also been observed for the development of non-Hodgkin’s lymphoma and squamous-cell skin cancer, which are known hallmarks of dysregulated immunity, suggesting a contribution of suppressed immune function as a feature of CUP. Smokers are at risk of developing CUP and this risk correlates with the level of tobacco exposure: from 1.8-fold for smokers of 1-15 cigarettes/day up to 3.5-fold for 16-25 cigarettes/day and 4.1-fold for >25 cigarettes/day.6Kaaks R. Sookthai D. Hemminki K. et al.Risk factors for cancers of unknown primary site: results from the prospective EPIC cohort.Int J Cancer. 2014; 135: 2475-2481Crossref PubMed Scopus (34) Google Scholar The likelihood of being a smoker was higher in patients with CUP and respiratory system metastases (4.9-fold) than in those with CUP and liver metastases (2.0-fold).7Hemminki K. Chen B. Melander O. et al.Smoking and body mass index as risk factors for subtypes of cancer of unknown primary.Int J Cancer. 2015; 136: 246-247Crossref PubMed Scopus (0) Google Scholar Type 2 diabetes (1.8-fold)8Hemminki K. Försti A. Sundquist K. et al.Cancer of unknown primary is associated with diabetes.Eur J Cancer Prev. 2016; 25: 246-251Crossref PubMed Scopus (11) Google Scholar and autoimmune disorders are also associated with an increased risk of CUP; the relative risks were 3.5 for polymyositis/dermatomyositis, 1.8 for primary biliary cirrhosis and 1.7 for Addison disease.9Hemminki K. Sundquist K. Sundquist J. et al.Risk of cancer of unknown primary after hospitalization for autoimmune diseases.Int J Cancer. 2015; 137: 2885-2895Crossref PubMed Scopus (14) Google Scholar Familial predisposition to CUP is another established risk factor.10Hemminki K. Ji J. Sundquist J. et al.Familial risks in cancer of unknown primary: tracking the primary sites.J Clin Oncol. 2011; 29: 435-440Crossref PubMed Scopus (0) Google Scholar High body mass index, waist circumference, low socioeconomic status and black ethnic background may be additional risk factors.6Kaaks R. Sookthai D. Hemminki K. et al.Risk factors for cancers of unknown primary site: results from the prospective EPIC cohort.Int J Cancer. 2014; 135: 2475-2481Crossref PubMed Scopus (34) Google Scholar,11Urban D. Rao A. Bressel M. et al.Cancer of unknown primary: a population-based analysis of temporal change and socioeconomic disparities.Br J Cancer. 2013; 109: 1318-1324Crossref PubMed Scopus (54) Google Scholar The probability of survival after a diagnosis of CUP has remained at ∼20% at 1 year and has not improved much over time.1Brewster D.H. Lang J. Bhatti L.A. et al.Descriptive epidemiology of cancer of unknown primary site in Scotland, 1961-2010.Cancer Epidemiol. 2014; 38: 227-234Crossref PubMed Google Scholar,2Binder C. Matthes K.L. Korol D. et al.Cancer of unknown primary-Epidemiological trends and relevance of comprehensive genomic profiling.Cancer Med. 2018; 7: 4814-4824Crossref PubMed Scopus (22) Google Scholar Around half of the observed deaths occur within the first 3 months following diagnosis, i.e. median survival is ∼3 months. Survival is worse for adenocarcinoma and undifferentiated carcinoma compared with squamous-cell carcinoma (1-year survival of <20% and 36%, respectively).12Hemminki K. Bevier M. Hemminki A. et al.Survival in cancer of unknown primary site: population-based analysis by site and histology.Ann Oncol. 2012; 23: 1854-1863Abstract Full Text Full Text PDF PubMed Scopus (87) Google Scholar Increasing age is associated with a survival disadvantage. Patients with CUP manifestations restricted to lymph nodes have a better prognosis than those with extranodal disease.1Brewster D.H. Lang J. Bhatti L.A. et al.Descriptive epidemiology of cancer of unknown primary site in Scotland, 1961-2010.Cancer Epidemiol. 2014; 38: 227-234Crossref PubMed Google Scholar,12Hemminki K. Bevier M. Hemminki A. et al.Survival in cancer of unknown primary site: population-based analysis by site and histology.Ann Oncol. 2012; 23: 1854-1863Abstract Full Text Full Text PDF PubMed Scopus (87) Google Scholar In a comparison of survival outcomes for patients with CUP versus those with metastatic cancer of known primary and matched location of metastases, in general, patients with CUP had a poorer survival, with the exception of those with brain and respiratory system metastases.13Riihimäki M. Thomsen H. Hemminki A. et al.Comparison of survival of patients with metastases from known versus unknown primaries: survival in metastatic cancer.BMC Cancer. 2013; 13: 36Crossref PubMed Scopus (0) Google Scholar Histology and immunohistochemistry (IHC) on good quality tissue specimens are required. A morphological pattern-based approach is first applied to differentiate between epithelial, round, spindle-shaped and anaplastic cancers to identify the pattern of tissue organisation regarding entity and tissue of origin. For undifferentiated neoplasms or cells of unclear lineage, an initial IHC screening is carried out,14Selves J. Long-Mira E. Mathieu M.C. et al.Immunohistochemistry for diagnosis of metastatic carcinomas of unknown primary site.Cancers (Basel). 2018; 10: 108Crossref PubMed Scopus (77) Google Scholar typically comprising a broad-spectrum keratin to identify an epithelial phenotype (e.g. AE1/AE3, OSCAR), cluster of differentiation 45 (CD45) for haematolymphoid origin (be aware of downregulation of CD45 expression in immature B-cell neoplasms) and SOX10 and/or S100 for melanoma. In case of a triple-negative screen, a mesenchymal origin must be considered. There is no single screening marker for sarcoma (see Supplementary Table S1, available at https://doi.org/10.1016/j.annonc.2022.11.013). After lineage classification, a stepwise approach, using additional marker assessments navigated by the clinical work-up results, must be undertaken (see Supplementary Table S2, available at https://doi.org/10.1016/j.annonc.2022.11.013). For carcinomas, cytokeratin (CK)7 and CK20 staining patterns may provide an indication of primary localisation (see Supplementary Table S3, available at https://doi.org/10.1016/j.annonc.2022.11.013). For male patients, metastatic prostate cancer must be ruled out using prostate-specific membrane antigen (PSMA) and/or NKX3.1 as markers. For female patients, GATA3 should be used to screen for breast cancer and SOX10 for triple-negative breast cancer. Only ∼60% of poorly differentiated and metastatic lung adenocarcinomas stain positive for thyroid transcription factor 1 (TTF1).15Noh S. Shim H. Optimal combination of immunohistochemical markers for subclassification of non-small cell lung carcinomas: a tissue microarray study of poorly differentiated areas.Lung Cancer. 2012; 76: 51-55Abstract Full Text Full Text PDF PubMed Scopus (0) Google Scholar In the setting of CK7 positivity and TTF1 negativity but suspicion of a lung primary, SWI/SNF-related, matrix-associated, actin-dependent regulator of chromatin, subfamily A, member 4 (SMARCA4) staining should be considered as many TTF1-negative lung adenocarcinomas show loss of SMARCA4 nuclear staining.16Herpel E. Rieker R.J. Dienemann H. et al.SMARCA4 and SMARCA2 deficiency in non-small cell lung cancer: immunohistochemical survey of 316 consecutive specimens.Ann Diagn Pathol. 2017; 26: 47-51Crossref PubMed Google Scholar Napsin A can be useful in a panel together with TTF1 in the diagnostic work-up of lung adenocarcinoma but it has limited value when TTF1 is negative (see Supplementary Table S3, available at https://doi.org/10.1016/j.annonc.2022.11.013). For analysis of biopsies including an adenocarcinoma in the liver, the initial IHC panel should include CK7, CK20, caudal type homeobox 2 (CDX2) and TTF1 (plus GATA3 and/or SOX10 in women) to screen for metastatic tumours of breast, lung, gastrointestinal (GI) and/or pancreaticobiliary origin (see Supplementary Table S3, available at https://doi.org/10.1016/j.annonc.2022.11.013). At least 80% of colorectal cancers (CRCs) show the classic CK7-negative, CK20-positive, CDX2-positive immunophenotype, with CK20 and CDX2 staining usually being diffuse and strong. Occasional upper GI and rare pancreaticobiliary adenocarcinomas also demonstrate a colorectal immunophenotype. In this setting, special AT-rich sequence-binding protein 2 (SATB2) positivity is fairly specific for tumours of lower GI origin.17Lugli A. Tzankov A. Zlobec I. et al.Differential diagnostic and functional role of the multi-marker phenotype CDX2/CK20/CK7 in colorectal cancer stratified by mismatch repair status.Mod Pathol. 2008; 21: 1403-1412Abstract Full Text Full Text PDF PubMed Scopus (0) Google Scholar The differential diagnosis of intrahepatic cholangiocarcinomas (CCAs) by IHC remains difficult due to the lack of specific markers. Immunohistochemical loss of BRCA1-associated protein 1 (BAP1) or AT-rich interactive domain-containing protein 1A (ARID1A) can support the diagnosis but the final decision can only be made in conjunction with clinical and radiological findings.18Pauli C. Bochtler T. Mileshkin L. et al.A challenging task: identifying patients with cancer of unknown primary (CUP) according to ESMO guidelines: the CUPISCO trial experience.Oncologist. 2021; 26: e769-e779Crossref PubMed Scopus (15) Google Scholar In order to identify neuroendocrine tumours, a synaptophysin and/or INSM1 staining must be carried out in tumours with a solid, trabecular, gyriform or regular glandular growth pattern, uniform nuclei and coarsely stippled (‘salt and pepper’) chromatin. Likewise, synaptophysin and/or INSM1 staining should also be carried out in high-grade tumours that resemble small-cell carcinomas or large-cell neuroendocrine tumours of the lung. Positivity for CDX2 and ISLET 1 may hint towards primary locations of neuroendocrine tumours in the GI tract and pancreas, respectively (see Supplementary Table S3, available at https://doi.org/10.1016/j.annonc.2022.11.013). Specific caveats in the CUP work-up exist for mesotheliomas, which are typically positive for keratins and therefore might be misclassified as carcinomas. Mesothelioma should be considered in biopsies originating from the pleura, pericardium and peritoneum. Immunostaining with calretinin should be carried out in these cases and, upon positivity, should be complemented with Wilms tumour 1 (WT1), CK5/6, D2-40 and BAP1 (loss) (see Supplementary Table S1, available at https://doi.org/10.1016/j.annonc.2022.11.013). Expression of broad-spectrum epithelial markers by mesenchymal tumours is focal in most cases. However, in cases with an epithelioid morphology, expression of these markers can be diffuse, and strong keratin positivity (e.g. synovial sarcoma, epithelioid sarcoma) often leads to erroneous classification as carcinoma. Keratin positivity might also be seen in small round blue cell sarcomas (e.g. desmoplastic round cell tumour, Ewing’s sarcoma).19Bahrami A. Gown A.M. Baird G.S. et al.Aberrant expression of epithelial and neuroendocrine markers in alveolar rhabdomyosarcoma: a potentially serious diagnostic pitfall.Mod Pathol. 2008; 21: 795-806Abstract Full Text Full Text PDF PubMed Scopus (122) Google Scholar Regardless of broad-spectrum epithelial marker positivity, sarcoma should always be considered in the mediastinum, retroperitoneum and soft tissue, particularly in cases with spindle cell morphology (see Supplementary Table S1, available at https://doi.org/10.1016/j.annonc.2022.11.013). Broad-spectrum keratins can be expressed by haematolymphoid tumours such as plasma cell neoplasms, anaplastic large-cell and mantle-cell lymphomas.20Adams H. Schmid P. Dirnhofer S. et al.Cytokeratin expression in hematological neoplasms: a tissue microarray study on 866 lymphoma and leukemia cases.Pathol Res Pract. 2008; 204: 569-573Crossref PubMed Scopus (31) Google Scholar Useful immunohistochemical markers for screening are listed in Supplementary Table S1, available at https://doi.org/10.1016/j.annonc.2022.11.013. The minimal mandatory work-up for all patients should comprise the following tests:•Thorough patient history and physical examination.•Blood draw with basic blood and biochemical analyses.•Either computed tomography (CT) with intravenous (i.v.) contrast agent infusion or magnetic resonance imaging (MRI) scans of the neck, thorax, abdomen and pelvis.•Mammography in females. Beyond this minimal diagnostic work-up, further tests are indicated according to the clinical and pathological results. This includes the tumour markers α-fetoprotein (AFP) and β-human chorionic gonadotropin (β-hCG) in males with a suspected germ cell tumour; prostate-specific antigen (PSA) in males with a possible prostate cancer; cancer antigen (CA)15-3 and CA125 in females with a suspected gynaecological primary and carcinoembryonic antigen (CEA), carbohydrate antigen 19-9 (CA19-9) and CA72-4 when a GI primary is suspected and chromogranin A in patients with a possible neuroendocrine malignancy. Despite frequent non-tumour type-specific elevations of CEA, CA19-9, CA15-3 and CA125, these markers may be used to determine the disease course and monitor treatment response. Gastroscopy and colonoscopy are generally recommended whenever a putative GI primary is deemed possible. In contrast, bronchoscopy may be withheld unless IHC or the clinical picture for lung lesions and/or mediastinal lymph nodes implies a lung primary. Diagnostic and staging guidelines for patients with an anticipated CUP diagnosis are summarised in Supplementary Table S4, available at https://doi.org/10.1016/j.annonc.2022.11.013, and include both the minimal mandatory and additional results-driven tests. MRI is recommended for suspected head and neck tumours, brain metastases and for suspected pelvic neoplasms. Dedicated protocols are needed for some primary tumours, such as breast or prostate cancers, or for differentiating adenoma from metastasis in case of enlarged adrenal glands. Whole-body [18F]2-fluoro-2-deoxy-D-glucose (FDG)–positron emission tomography (PET)–CT is optional in the routine CUP diagnostic work-up. Although it is excellent for depicting the true extent of disease and identifying lesions that are otherwise difficult to detect, it is only able to identify a primary in around a third of cases.21Kwee T.C. Kwee R.M. Combined FDG-PET/CT for the detection of unknown primary tumors: systematic review and meta-analysis.Eur Radiol. 2009; 19: 731-744Crossref PubMed Scopus (208) Google Scholar However, it is generally recommended in the following situations that warrant radical locoregional treatment:•For patients with single-site/oligometastatic CUP, FDG–PET–CT should be carried out to rule out additional manifestations.22Lievens Y. Guckenberger M. Gomez D. et al.Defining oligometastatic disease from a radiation oncology perspective: an ESTRO-ASTRO consensus document.Radiother Oncol. 2020; 148: 157-166Abstract Full Text Full Text PDF PubMed Google Scholar•For patients with cervical lymph node metastases suspicious for head and neck cancer.23Albertson M. Chandra S. Sayed Z. et al.PET/CT evaluation of head and neck cancer of unknown primary.Semin Ultrasound CT MR. 2019; 40: 414-423Abstract Full Text Full Text PDF PubMed Scopus (2) Google Scholar,24Maghami E. Ismaila N. Alvarez A. et al.Diagnosis and management of squamous cell carcinoma of unknown primary in the head and neck: ASCO guideline.J Clin Oncol. 2020; 38: 2570-2596Crossref PubMed Scopus (47) Google Scholar In these cases, FDG–PET–CT may be carried out early during the diagnostic work-up, ahead of panendoscopy with biopsies and tonsillectomy, to avoid false-positive findings.23Albertson M. Chandra S. Sayed Z. et al.PET/CT evaluation of head and neck cancer of unknown primary.Semin Ultrasound CT MR. 2019; 40: 414-423Abstract Full Text Full Text PDF PubMed Scopus (2) Google Scholar Furthermore, when clinically suspected, specific tumour entities can be diagnosed if special tracers are used, such as DOTATOC for neuroendocrine tumours or PSMA ligands for prostate cancer. With pathological and clinical diagnostic tests complete, the diagnosis of CUP relies on the multidisciplinary team’s interpretation of the clinical, pathological and radiographic findings in order to decide whether the tumour manifestations represent a primary cancer or metastases compatible with CUP. The most important diagnostic tool for this purpose is sound clinical reasoning. In the absence of a clearly identifiable primary tumour or an entity-specific genomic alteration (see Supplementary Table S5, available at https://doi.org/10.1016/j.annonc.2022.11.013), it must be decided whether one of the visible lesions is likely to represent a primary tumour. Radiological clues in the diagnosis are (i) size and location of lesions and their imaging features, (ii) associated phenomena (see Supplementary Tables S6 and S7, available at https://doi.org/10.1016/j.annonc.2022.11.013), (iii) invasion patterns into adjacent structures and (iv) the distribution of haematogenic and lymphogenic metastases. Specifically, tumours that metastasise along the preformed lymphatic pathways may be surrounded by lymphatic metastases in typical locations, with the closest adjacent lymph node groups usually being the most heavily involved. In the presence of widespread disease, the absence of lymph node metastases in typical sites may refute the presence of a primary tumour in a suspected location. In order to standardise the interpretation of findings in ambiguous cases, following established diagnostic algorithms to delineate CUP from cancer entities with a known primary is generally recommended.18Pauli C. Bochtler T. Mileshkin L. et al.A challenging task: identifying patients with cancer of unknown primary (CUP) according to ESMO guidelines: the CUPISCO trial experience.Oncologist. 2021; 26: e769-e779Crossref PubMed Scopus (15) Google Scholar The respective decision algorithms are based on histology and IHC, metastatic pattern and radiographic criteria, and may be applied, particularly in clinical trials, to ensure the integrity of the study cohorts and to harmonise the eligibility process among different trials for the sake of comparability. The differential diagnosis between CUP and non-small-cell lung cancer (NSCLC) poses a recurrent diagnostic dilemma. Since 40% of lung cancers are negative for TTF1, patients presenting with neoplastic pulmonary lesions may either have metastases from an unknown primary tumour to the lung, or one of the lung lesions may be the primary itself, usually an NSCLC, with extrapulmonary as well as pulmonary metastases. The respective decision algorithms are based on pathological and radiographic features of the lung mass, hilar and mediastinal lymph node involvement and the pattern of distant metastases (see Figures 1 and 2). Radiographic features to support the discrimination are shown in Supplementary Table S6, available at https://doi.org/10.1016/j.annonc.2022.11.013.Figure 2Differential diagnostic algorithm to discriminate between CUP and TTF1-positive NSCLC.Show full captionPurple: general categories or stratification; white: other aspects of management.CK, cytokeratin; CUP, cancer of unknown primary; LN, lymph node; NSCLC, non-small-cell lung cancer; TTF1, thyroid transcription factor 1.View Large Image Figure ViewerDownload Hi-res image Download (PPT) Purple: general categories or stratification; white: other aspects of management. CK, cytokeratin; CUP, cancer of unknown primary; LN, lymph node; NSCLC, non-small-cell lung cancer; TTF1, thyroid transcription factor 1. The presence of intrahepatic lesions and histological proof of adenocarcinoma constitute a recurrent problem in discriminating between primary CCA (with or without additional intrahepatic metastases) and hepatic metastases due to an unknown extrahepatic primary tumour (with or without additional extrahepatic metastases).25Bridgewater J. Galle P.R. Khan S.A. et al.Guidelines for the diagnosis and management of intrahepatic cholangiocarcinoma.J Hepatol. 2014; 60: 1268-1289Abstract Full Text Full Text PDF PubMed Scopus (907) Google Scholar,26Conway A.M. Morris G.C. Smith S. et al.Intrahepatic cholangiocarcinoma hidden within cancer of unknown primary.Br J Cancer. 2022; 127: 531-540Crossref PubMed Scopus (1) Google Scholar The decision algorithm to differentiate CUP with liver metastases from intrahepatic CCA relies on (I)HC, radiological morphology, size and number of hepatic lesions and the overall metastatic pattern (see Figure 3). Radiologically, the criteria shown in Supplementary Table S7, available at https://doi.org/10.1016/j.annonc.2022.11.013, suggest intrahepatic CCA. Further algorithms have been established for the differential diagnosis between CUP and ovarian, renal, salivary gland and breast primaries (see Figure 4, Figure 5, Figure 6, Figure 7).Figure 5Differential diagnostic algorithm to discriminate between CUP and RCC.Show full captionPurple: general categories or stratification; white: other aspects of management.CT, computed tomography; CUP, cancer of unknown primary; IHC, immunohistochemistry; MRI, magnetic resonance imaging; RCC, renal-cell carcinoma.View Large Image Figure ViewerDownload Hi-res image Download (PPT)Figure 6Differential diagnostic algorithm to discriminate between CUP and salivary gland carcinoma.Show full captionPurple: general categories or stratification; white: other aspects of management.2D, two dimensional; CT, computed tomography; CUP, cancer of unknown primary; ENT, ear nose and throat; FPS, frames per second; IHC, immunohistochemistry; MRI, magnetic resonance imaging; US, ultrasound.aHigh specification includes: broad band linear array transducer with a frequency range of 5-20 MHz suitable for vascular superficial, superficial small parts and elastography applications; electronic phased array colour Doppler system with minimum 50 000 digital processing channels and ≥256 grey shades for sharp contrast resolution; frame rate of ≥500 FPS; gain control for an additional level of flexibility to image quality control; real-time high-frequency 2D imaging for higher resolution and low-frequency Doppler for higher sensitivity; tissue harmonic imaging in power Doppler imaging mode for improved sensitivity and specificity in differentiating blood from tissue.View Large Image Figure ViewerDownload Hi-res image Download (PPT)Figure 7Differential diagnostic algorithm to discriminate between CUP and breast cancer.Show full captionPurple: general categories or stratification; white: other aspects of management.CUP, cancer of unknown primary; IHC, immunohistochemistry; LN, lymph node; MRI, magnetic resonance imaging.View Large Image Figure ViewerDownload Hi-res image Download (PPT) Purple: general categories or stratification; white: other aspects of management. CT, computed tomography; CUP, cancer of unknown primary; IHC, immunohistochemistry; MRI, magnetic resonance imaging; RCC, renal-cell carcinoma. Purple: general categories or stratification; white: other aspects of management. 2D, two dimensional; CT, computed tomography; CUP, cancer of unknown primary; ENT, ear nose and throat; FPS, frames per second; IHC, immunohistochemistry; MRI, magnetic resonance imaging; US, ultrasound. aHigh specification includes: broad band linear array transducer with a frequency range of 5-20 MHz suitable for vascular superficial, superficial small parts and elastography applications; electronic phased array colour Doppler system with minimum 50 000 digital processing channels and ≥256 grey shades for sharp contrast resolution; frame rate of ≥500 FPS; gain control for an additional level of flexibility to image quality control; real-time high-frequency 2D imaging for higher resolution and low-frequency Doppler for higher sensitivity; tissue harmonic imaging in power Doppler imaging mode for improved sensitivity and specificity in differentiating blood from tissue. Purple: general categories or stratification; white: other aspects of management. CUP, cancer of unknown primary; IHC, immunohistochemistry; LN, lymph node; MRI, magnetic resonance imaging. For the detection of salivary gland carcinoma, additional studies may be needed. Ultrasound may be sufficient, with contrast-enhanced MRI of the neck as a reliable alternative method. CT is less suitable due to its lower soft-tissue contrast. Although a negative imaging result is sufficient to rule out a primary salivary gland tumour, a positive finding may be more difficult to interpret due to the possibility of lymph nodes located inside the gland that may be involved by metastatic spread. In unclear cases, targeted biopsies of the lesion may be carried out. About 25% of patients with presumed CUP have had a prior malignancy.27Bochtler T. Endris V. Leichsenring J. et al.Comparative genetic profiling aids diagnosis and clinical decision making in challenging cases of CUP syndrome.Int J Cancer. 2019; 145: 2963-2973Crossref PubMed Scopus (15) Google Scholar In these cases, a relapse of the prior malignancy should always be considered. In dubious cases, comparative sequencing of tissue from the prior malignancy and presumed CUP is recommended to