Review and Comparison of Cancer Biomarker Trends in Urine as a Basis for New Diagnostic Pathways

Cancer is one of the major causes of mortality worldwide and its already large burden is projected to increase significantly in the near future with a predicted 22 million new cancer cases and 13 million cancer-related deaths occurring annually by 2030. Unfortunately, current procedures for diagnosis are characterized by low diagnostic accuracies. Given the proved correlation between cancer presence and alterations of biological fluid composition, many researchers suggested their characterization to improve cancer detection at early stages. This paper reviews the information that can be found in the scientific literature, regarding the correlation of different cancer forms with the presence of specific metabolites in human urine, in a schematic and easily interpretable form, because of the huge amount of relevant literature. The originality of this paper relies on the attempt to point out the odor properties of such metabolites, and thus to highlight the correlation between urine odor alterations and cancer presence, which is proven by recent literature suggesting the analysis of urine odor for diagnostic purposes. This investigation aims to evaluate the possibility to compare the results of studies based on different approaches to be able in the future to identify those compounds responsible for urine odor alteration.


Introduction
Cancer is one of the major causes of mortality worldwide [1]. The International Agency for Research on Cancer (IARC), focusing on geographic variability across 20 world regions, estimated 18.1 million new cancer cases and 9.6 million cancer deaths in 2018 [1].
This large burden is projected to increase significantly in the near future with predicted 22 million new cancer cases and 13 million cancer-related deaths occurring annually by 2030 [2]. This increasing magnitude of cancer is a consequence of population growth and aging, but societal, economic, and lifestyle changes-linked to increasing human development-are likely to additionally increase the scale and alter cancer trends in the next decades [2].
Given low accuracies of current diagnostic procedures of some cancer types [3], researchers are rising to the challenge of developing innovative diagnostic tools able of identifying cancer during its early stage, that is the most curable one [4]. Effective screening tests should be non-invasive, easily accessible, quickly quantifiable, reliable, and reproducible. They should have high sensitivity, high specificity, low financial burden on patients and the lowest possible risk level [4].
In recent years, the deep understanding of biological structures and processes has revolutionized several aspects of cancer research, offering new opportunities for early cancer diagnosis before symptoms appear, when the tumor is at early stages.
Diagnostic methods, currently involved for LC diagnosis, are chest x-ray, computed tomography (CT), positron emission tomography (PET), sputum cytology, and bronchoscopy [29]. They are time-consuming, expensive, and quite dangerous due to radiation exposure [30]. Moreover, they do not provide exhaustive diagnostic information, thus biopsy or lung resection must be carried out to confirm the diagnosis and define the treatment plan [28,31,32].
Carcinoembryonic antigen (CEA), squamous cell carcinoma (SCC) antigen and neuron specific enolase (NSE) are serum markers, commonly used to detect lung cancer, monitor its progression and disease recurrence. However, for localized tumors, their diagnostic capabilities are relatively poor. In fact, most LC cases are found at an advanced stage, reducing the number of effective treatments and increasing the cancer related mortality [27,29,31].
Early-stage disease may allow a curative surgery with an adequate tumor extirpation, but no screening process, adopted up to now, are capable to detect the disease at a stage which improves the overall survival [28]. Indeed, the last 30 years have seen little improvement in the overall five-year survival rate for LC, with only 15% of patients living for at least five years after their initial diagnosis [27].
In recent years, researchers have proposed the combination of imaging tests with metabolomics in order to improve the specificity of the actual procedure [27]. Therefore, biological fluids have been investigated with the aim of identifying metabolites capable of providing information related to lung cancer presence and tumor development. Table 1 summarizes literary papers investigating urinary biomarkers specific for lung cancer, focusing on the main aspects of the studies presented (i.e., population involved, experimental methods, and results). Hexanal concentration is higher in LC patients than in controls, while heptanal concentration is not so different between LC patients and controls Hanai et al. (2012) [35] 20 LC patients (59-77 years) at different stages 20 controls: (38-62 years) Sample preparation: thawing, centrifugation, filtration and extraction by SPME Analytical technique: GC-TOF MS Tetrahydrofuran 2-chloroethanol 2-pentanone 2-methylpyrazine Cyclohexanone 2-ethyl-1-hexanol 2-phenyl-2-propanol Isophorone All of these biomarkers are up-regulated in LC patients than controls, the most significant are six of them. 2-pentanone is important for the differentiation between patients with adenocarcinoma and those with squameous cell, because in the first one 2-pentanone is higher.

Breast Cancer
Breast cancer (BC) is the second most common cancer form and the second leading cause of cancer related death in women [36]. In the Global Cancer Statistics for 2018 [1], about 2,088,849 BC new cases, mainly invasive tumors (i.e., about the 80%), and about 626,679 related deaths were estimated worldwide.
Breast cancer can start from different parts of the breast. However, most breast cancers start developing in the ducts that carry milk to the nipple (i.e., ductal cancer) or in the glands where breast milk is formed (i.e., lobular cancer). Although many types of breast cancer can cause lumps in the breast, most breast lumps may be related to non-cancerous breast tumors, which do not spread outside of the breast and are not life threatening [36].
The first step of current diagnostic procedure involves breast exam and imaging tests (mammogram, breast ultrasound, breast magnetic resonance MRI scan) to identify any lumps or abnormalities [37,38].
Since several researchers proved the higher efficacy of cancer treatments if breast cancer is detected at an early stage [39,40], women over 40 years old are recommended to undergo mammograms every year with the aim of detecting cancers before physical symptoms appear and treatments alternative to mastectomy and chemotherapy might be adopted.
Despite substantial increases in the number of cases of early-stage BC detected, screening mammography has only marginally reduced the rate of detection of advanced cancers. The imbalance suggests that there is substantial overdiagnosis, accounting for nearly a third of all newly diagnosed breast cancers, and that screening is having, at best, only a small effect on the rate of death from breast cancer [41].
In case of positive results of screening tests, the breast biopsy is carried out to confirm the cancer presence through the direct examination of the cell cytomorphology [42]. The choice of the type of the biopsy (i.e., core needle biopsy, fine needle aspiration (FNA) biopsy, surgical/large core biopsy, vacuum-assisted biopsy) to be performed depends on several factors: how suspicious the abnormality appears, the size, the shape and the location of the abnormality, the number of abnormalities present and the patient's medical history.
Even with its high accuracy, breast biopsy is an invasive test, which carries some risks for patients and, since the consequential risk increases as the procedure becomes more invasive, core needle biopsy and fine needle aspiration (FNA) biopsy are preferred respect to surgical/large core biopsy and vacuum-assisted biopsy [43].
In recent years, many research groups started to investigate alternative methods to biopsy with the aim to develop non-invasive diagnostic tools [44][45][46][47][48][49]. Therefore, they started looking for novel BC biomarkers capable to improve current diagnostic procedure in biological fluids. Table 2 summarizes the main aspects of recent literary studies proposing the investigation of urine samples aimed at the discovery of specific BC biomarkers.

Prostate Cancer
Prostate cancer (PrC) represents the most common cancer in men globally [50] and the fifth most frequent cancer in the world. In the Global Cancer Statistics for 2018 [1], about 1,276,106 PrC new cases, and about 358,989 related deaths were estimated worldwide.
PrC is an asymptomatic and slow growing tumor, and, thus, it is a perfect candidate for screening programs, whose purpose is the detection of cancer during its early stage, when the tumor is localized in the prostate and clinical treatments result more effective [4].
The current screening procedure for PrC detection involves the measurement of Prostate-Specific Antigen (PSA) serum level, the digital rectal examination (DRE), and the prostate biopsy.
The Serum Prostate-Specific Antigen (PSA) is currently the most important biomarker for the detection, follow-up, and therapeutic monitoring of prostate cancer. Although its use has been associated with a significant reduction in prostate cancer mortality, it has also resulted in the over-diagnosis and overtreatment of indolent prostate cancer [51]. Indeed, PSA test is characterized by a poor diagnostic accuracy (i.e., 30%) in terms of specificity and by a high false positive rate [52]: Many positive results are related to urinary tract infections or benign prostatic hyperplasia and not to prostate cancer. Additionally, serum PSA levels are affected by biologic variability that may be related to differences in androgen levels or prostate manipulation and may have distinct racial variation [53].
Therefore, patients with increased PSA values undergo prostate biopsy to confirm the cancer presence. The biopsy is an invasive test, which consists in the analysis under a microscope of tissue samples taken from the suspicious mass by surgery. It may carry out significant risks for patients and longer-term health issues such as bleeding, inflammatory, or infectious complications [54][55][56]. It also entails a low level of accuracy (i.e., only 30% detection rate at the first biopsy).
In addition to its invasiveness, the actual diagnostic procedure does not allow detecting the tumor before symptoms appear and when it is localized in the prostate. This reduces patients' chances of surviving and increases patients' management costs (e.g., need for extensive treatments and frequent hospital admissions) [4]. There is, thus, a need for more reliable, non-invasive method to diagnose prostate cancer.
Many studies regarding the investigation of novel PrC biomarkers have been published in the scientific literature. Table 3 summarizes literary works, focusing on the population considered, the experimental method involved, and the biomarkers proposed.
The schematization of the main aspects of literary studies highlighted that only partial results have been obtained until now, because, although different papers proposed the same metabolites as potential urine markers, results relevant for the quantitative urine characterization are discordant.
Sarcosine is the most debated biomarker among the ones proposed. Indeed, many authors [57,58] reported that its level in urine from PrC patients is higher than in control samples and that its classification performance is good, whereas other researchers [59,60] showed that changes in its concentrations between PrC and healthy men were not statistically significant.  In PrC samples, higher concentrations of arginine both before (P = 0,018) and after (P = 0,009) prostate massage and higher levels of proline only after prostate massage (P = 0,032) were detected. Higher levels of proline and homoserine and tyramine correlate with GS7 with respect to GS 6 and GS 5.

Colorectal Cancer
Colorectal cancer (CrC) is the second most commonly diagnosed cancer in females and the third in males worldwide [71].
CrC incidence varies worldwide and is higher in more developed countries including Australia, New Zealand, Western Europe, and North America, revealing a strong contribution of environmental factors in disease development [71]. However, CrC incidence is increasing also in several Asian regions and in Eastern Europe, probably as a result of increases in smoking, adoption of unhealthy dietary habits and sedentary lifestyles [71]. CrC is thought to develop as a result of environmental factors and the accumulation of genetic and epigenetic alterations [72]. It occurs in inherited, familial and sporadic forms [72].
CrC related mortality is also high, with a global estimate of 600,000-700,000 deaths/year [71]. Current CrC screening tests can be broadly distinguished as early detection tools or cancer-prevention tools depending on their modes of action [6].
Detecting CrC at an early stage improves dramatically survival rates: five-year survival rate is 93% for stage I patients, but only 8% for stage IV patients [73].
Early detection tools involve fecal occult blood tests (FOBTs) and fecal immunochemical tests (FITs), which are non-invasive and cost effective [6]. However, the positive results are routinely recommended for endoscopy [6], which is the gold standard technique for CrC detection.
Nevertheless, its invasiveness, associated discomfort, potential risk of complications represent marked disadvantages [74] and lead many researchers to investigate and define novel methods based on specific biomarkers that would improve early CrC detection [75].
Although, in recent years, some CrC biomarkers, as carcinoembryonic antigen (CEA), have already been combined to current diagnostic tests, their sensitivity and specificity are relatively poor [76]. Table 4 summarizes the CrC biomarkers proposed in the scientific literature in the last 10 years, highlighting the analytical techniques adopted for their identification and quantification in urine samples.

Gastric Cancer
Stomach or gastric cancer (GC) was estimated to be responsible for over 1,000,000 new cases in 2018 worldwide and about 783,000 deaths, making it the fifth most frequently diagnosed cancer and the third leading cause of cancer death [1]. Rates are 2-fold higher in men than women. Several studied documented a strong environmental component in explaining the regional variation in stomach cancer incidence rates [1].
Helicobacter pylori is the main risk factor for stomach cancer, with almost 90% of new cases of noncardia gastric cancer attributed to this bacterium. However, also the dietary habits (i.e., low fruit intake, alcohol, and foods preserved by salting consumption, active tobacco smoking) engrave GC incidence risk [1].
GC can generally be distinguished into cardia and noncardia GC. Those two GC forms are characterized by different incidence rates: the number of noncardia GC new cases is decreasing thanks to prevention and the improvements in the preservation and storage of foods; whereas rates of cardia gastric cancer are increasing due to the rise of its risk factors, as obesity and gastroesophageal reflux disease (GERD) [1].
In general, since GC symptoms are non-cancer specific and may be related to other diseases as gastritis [78], GC is often diagnosed late, reducing the efficacy of treatments and patients' chances of surviving [79]. However, for tumors detected when they are confined to the mucosal or submucosal layer, the 5 year survival rate is above 90% after surgical management [79]. This highlights the importance of appropriate screening in higher-risk population.
Currently, the standard diagnostic method for GC early detection involves gastroduodenal endoscopy (GE) as mass screening tool. GE adoption as mass screening tool has resulted in the reduction of GC-specific mortality and improved survival rates of GC patients [80]. However, data on the impact of GE screening programs on gastric cancer mortality are limited [81] and GE is an invasive and expensive tool.
Thus, the interest in the development of non-invasive and reliable tests capable of detecting GC in asymptomatic patients has increased significantly in recent years [78,79,82].
Many serum-and tissue-based biomarkers specific for GC have been identified through genomic and proteomic techniques, but only serum carcinoembryonic antigen (CEA) and carbohydrate antigen 19-9 (CA 19-9) have been proved to being clinically useful. Nevertheless, their sensitivities are poor [79].
Alternatively, urine has been investigated as potential source of specific GC markers that may allow early diagnosis and the discrimination between GC and benign gastric disease (BN). Table 5 summarizes the main aspects of research studies, investigating urinary metabolites related to stomach cancer, reported in the scientific literature.

Hepatic Cancer
In 2018, GLOBOCAN 2018 estimated that the liver or hepatic cancer (HC) would be the sixth most commonly diagnosed cancer and the fourth leading cause of cancer-related death worldwide, with about 841,000 new cases and 782,000 deaths annually [1].
Rates of both incidence and mortality are 2 to 3 times higher among men in most world regions; so liver cancer ranks sixth in terms of global cases and second in terms of deaths for males [1]. Liver cancer is diffused throughout the world, but in particular in Northern and Western Africa (Egypt, the Gambia, Guinea) and Eastern and South-Eastern Asia (Mongolia, Cambodia, Vietnam) [1].
Liver cancer incidence is increasing in Western Europe and in United States, maybe because of chronic alcohol use and chronic hepatitis C infection [84]. Diabetic and metabolite diseases of the liver have been known to contribute to this increasing incidence trend too.
Current methods for liver cancer diagnosis rely on imaging techniques (i.e., radiographic and ultrasound) that are not practical for mass screening tools aiming at early HC diagnosis [85], being costly and time-consuming [86].
Moreover, the current diagnostic procedure allows the detection of HC only at advanced stages. Consequently, treatments effectiveness is reduced, resulting in high mortality rate [87].
Because of the late diagnosis of liver cancers, the rate of postoperative recurrence at 5 years is significant (about 70%) [88].
Early detection of liver cancer and advances in surgical techniques might greatly improve short-term survival for those diagnosed with HC. Nevertheless, the identification of novel and more accurate biomarkers remains a very challenging task primarily due to the heterogeneity of disease development and progression [84].
Although alpha-fetoprotein (AFP), des-γ-carboxyrothrombin, the AFPL3 and the midkine (MDK) has been reported as specific HC biomarkers, researchers are still working on the identification of novel biomarkers because of low sensitivity and specificity of the actual procedures. Indeed, the AFP often leads to high rates of false positives and false negatives [89]: About 30% of primary liver cancer patients are AFP negative [87].
Novel biomarkers should be capable both to detect HC and monitor its progression, because methods currently available to investigate HC metastatic potential, such as cell adhesion, migration, invasion, and angiogenesis, are time-consuming, costly and unsuitable for clinical application [88]. Hence, a new rapid, cost-effective, and accurate prognostic method of invasion is needed. Table 6 summarizes the main aspects of literary studies about urinary liver cancer biomarkers, highlighting population involved, techniques adopted and differences between healthy people and HC patients.

Bladder Cancer
Bladder cancer (BlC) is the seventh most common cancer and ninth leading cause of cancer-related death, with an estimated 549,000 new cases and 200,000 deaths in 2018 worldwide [1]. Bladder cancer is more common in men than women, with respective incidence and mortality rates of 9.6 and 3.2 per 100,000 in men: about 4 times those of women globally [1]. Incidence rates in both sexes are highest in Southern Europe (Greece, Spain, Italy), Western Europe (Belgium and the Netherlands), and Northern America. Considering only women, the highest rates are estimated in Lebanon.
Although NMIBC and MIBC both originate from the urothelium in the urinary bladder, they have distinct clinical characteristics [92]. NMIBC is associated with good survival compared to other malignancies, although 30-50% of patients with NMIBC will eventually experience recurrence after transurethral resection (TUR) of the primary tumor, and 10-20% will progress to muscle-invasive bladder cancer (MIBC) [2]. In the case of MIBC, instead, patients often have poor outcomes despite systemic treatments, although radical cystectomy, radiation therapy, and chemotherapy are considered to be effective therapies [92].
Therefore, early diagnosis of BlC is essential to properly manage BlC and improve the efficacy of treatments and patients' chances of surviving [92].
The current standard procedure for BlC detection and monitoring tumor progression and recurrence involves urine cytology, cystoscopy, and biopsy [92,93].
Urine cytology allows the detection of cancer or pre-cancer cells through a microscope screening of urine samples. However, its low sensitivity towards low-grade tumors reduces test reliability, although it is a perfect tool for the detection of high-grade bladder cancers [92][93][94].
Cystoscopy, i.e., the endoscopy of bladder via urethra, is an expensive, invasive and painful test, which may easily miss high-grade tumors [92,93]. Indeed, carcinoma in situ (CIS) looks like red mucosal spots typical of inflammatory lesions and may be confused [94]. Therefore, if an abnormal area is seen during cystoscopy, the patient undergoes biopsy to confirm the diagnosis. Although recent advances in biopsy technology, this exam is not a perfect tool and it may miss especially small tumors.
Thus, new diagnostic approaches that will improve the diagnostic accuracy of current procedure and the discrimination between non-malignant conditions and MIBC from NMIBC are needed [92]. In recent years, many efforts have been made to develop a non-invasive and less expensive tool for diagnosis, prognosis and monitoring of the treatment efficacy [93].
For this purpose, different analytical techniques, such as LC-MS, GC-MS, and NMR, have been proposed in the scientific literature for the chemical characterization of urine aimed at the identification of potential markers [92]. Table 7 proposes a schematization of literary studies investigating potential urinary BlC biomarkers, highlighting the main aspects of research studies presented in the literature.

Pancreatic Cancer
Pancreatic cancer (PaC) is the eighth leading cause of cancer death in both males and females, with 432,000 deaths and 459,000 cases worldwide [1]. Incidence rates are higher in countries with high human development index (HDI), mainly in Europe, North America, and Australia/New Zealand [1].
The risk of developing pancreatic cancer goes up as people age: about 80% of PaC patients are at least 60 years old and 71 is the average age at the time of diagnosis [97].
The most frequent pancreatic malignant tumor is the pancreatic ductal adenocarcinoma, representing the 85% of all reported cases [98], which could be associated to nodal metastasis and hepatic, bone or pulmonary metastasis [97]. PaC risk may be related to genetics factors (family history) and/or environmental factors, as smoking, alcohol consumption, chronic pancreatitis, obesity, and diabetes [97,99].
Rarely pancreatic cancer manifests specific symptoms when the tumor is at an early stage [100], thereby resulting in late diagnosis. Indeed, when symptoms like asthenia, jaundice, abdominal pain and weight loss, appear, patients already have an advanced pancreatic neoplasia [97].
Surgery is still the only curative therapy for pancreatic cancer [101]. However, although successful surgery of the pancreas is possible, only the 20-25% of patients are diagnosed at early disease stages when resection is effective [102]. Moreover, even in case of successful surgery, the median survival after surgery is only 17-23 months, due both to the resistance of pancreatic cancer to chemotherapy [98] and to the advanced stage diagnosis [102].
Current PaC diagnostic techniques are imagine techniques as computed tomography (CT), positron emission tomography-CT, magnetic resonance imaging (MRI), endoscopic retrograde cholangiopancreatography (ERCP), and endoscopy ultrasonography (EUS), which is sometimes associated to biopsy used for grading tumor histology [98,101,103]. However, these techniques are not so effective in detecting tumor of less than 2 cm in diameter [102].
The measurement of the Carbohydrate antigen 19-9 (CA 19-9) serum level is commonly used as a complementary test in the current diagnostic protocol [101]. However, although its sensitivity can reach 80%, the CA 19-9 has low specificity and it is unsuitable for the detection of PC at resectable stages [102].
In the scientific literature, metabolomics, genomics and post-genomic technologies have been proposed for the identification of potential PaC markers in biological fluids with the aim to define novel non-invasive and accurate diagnostic tool to enhance the early PaC detection [100]. Many authors proved the possibility to discriminate between healthy and PaC subjects through the investigation of blood, tissue and saliva [105][106][107][108][109][110][111][112][113][114] by means of NMR, GC/MS, HPLC/MS, UPLC/MS and CE. However, these studies involved relatively small populations and considered a small number of early stage or resectable cancers [102]. Other authors proposed the characterization of bile and pancreatic fluids, but their handling requires costly and invasive techniques [104].
Only few research groups proposed the investigation of urine samples and proposed specific PaC urinary biomarkers (Table 8), whereas Arasaradnam et al. [115], proposed the adoption of the ion mobility spectrometry for the characterization of the urine volatiles, achieving a sensitivity of 91%, a specificity of 83% and an accuracy of 92%. The model based on those 9 metabolites and CA19-9 achieved a diagnostic accuracy of 96%.

Renal Cancer
The renal or kidney cancer (RC) is one of the ten most common cancers in both men and women, although it is more frequent in men than in women [1]. For 2019, the American Cancer Society estimated about 73,820 new cases of kidney cancer and about 14,770 deaths related to this form of tumor [121].
Actual diagnostic procedure involves urinalysis and blood tests to look for blood traces and measure the levels of calcium and liver enzymes, which might be altered due to kidney cancer presence.
Urine cytology is carried out on urine samples to identify cancer cells eventually present. In case of abnormal results, patients undergo imaging tests, capable to provide useful information about tumor size, shape, and location, producing detailed cross-sectional images of suspected areas.
Computed tomography (CT) scan is the most common imaging tool involved for renal cancer diagnosis, but also magnetic resonance imaging (MRI) scan, ultrasound, positron emission tomography (PET) scan, intravenous pyelogram or angiography can be used [122,123].
Unlike other cancer types, biopsy is not often used for kidney cancer diagnosis, since imaging tests provide enough information to evaluate the need of surgery. However, in case of small tumors, biopsy is carried out to evaluate alternative treatments to surgery [122].
Since the RC is an asymptomatic tumor at early stages, in general current diagnostic procedure allows detection when symptoms appear and prognosis is poor [124]. Therefore, the identification of a screening biomarker has the potential for substantial health benefit.
In recent years, many researchers have proposed urine chemical characterization for metabolic profiling and identification of specific RC biomarkers. Literary works reported in Table 9 proposed novel urinary biomarkers specific for RC. Some authors [124,125] proposed an innovative approach, without attempting to identify all detected peaks, but rather than focusing on evaluation of the use of mass spectrometric and peak processing techniques for the development of innovative diagnostic tests for RC. Their approach was based on the idea that a large group of potential biomarkers was more likely to evolve patterns for disease recognition. The decisional models proposed by both research groups [124,125] achieved diagnostic accuracies above 88%. Isobutyrylcarnitine, Suberoylcarnitine and Acetylcarnitine levels were higher in RC patients than in controls. Acylcarnitines levels in urine increased as function of tumor grade.

Testicular Cancer
The Global Cancer Statistics 2018 [1] estimated for testicular cancer (TC) 71,105 new cases and 9507 related deaths. TC is more common cancer in young adults and potential risk factors include undescended testis (cryptorchidism), personal or family history of testicular cancer, age, ethnicity, and infertility [129].
About 1% of testicular cancers are neuroendocrine tumors, commonly known as carcinoid tumors, which are mostly primary tumors of the testes and rarely are metastasis to the testes from other organs. As opposed to common testicular cancer, carcinoid tumors can affect men of all ages [130].
As first steps of the current diagnostic procedure for testicular cancer, the patient undergoes a physical exam to check testicles for lumps, swelling, hardening, or tenderness and identify abnormal masses and his history and health habits are examined [131].
In case of abnormal results of the physical exams, an ultrasound exam of the testes and blood tests for specific-tumor markers are performed to evaluate if abnormalities are related to benign conditions (i.e., hydrocele or varicocele) or to testicular cancer [131].
Alpha-fetoprotein (AFP), human chorionic gonadotropin (HCG) and lactate dehydrogenase (LDH) are used as serum markers of testicular cancers. Indeed, rises in levels of AFP and HCG may allow the detection of the tumor and the identification of its type: non-seminomas or pure seminomas; while high LDH levels indicate a widespread disease [132].
Biopsy is rarely done for diagnosing testicular cancer because of the high risk of spreading cancers. Thus, based on ultrasound and blood marker tests, the doctor will very likely recommend inguinal orchiectomy, which consists in the removal of the entire testicle through an incision on the groin [131].
Although the scientific literature investigating testicular cancer urinary biomarkers is not so huge, this kind of tumor is reported for completeness of this review paper. Table 10 reports the main aspects of the scientific papers.

Design of the Experiment
In the scientific literature, it is commonly known that cancers are caused by uncontrolled growth of abnormal or mutated cells. Mutated cells contain broken genetic information, which make them resistant to apoptosis and capable of undergoing faster replication [135]. Due to genetic alterations, the pathways through which cancer cells acquire and replenish their metabolic needs are different from those of normal cells. This results in qualitative and quantitative alterations of the metabolomics profile of urine samples from healthy subjects and cancer patients, which can be investigated for early cancer diagnosis. Therefore, with the purpose to identify specific metabolites relevant for those metabolic alterations, most of the scientific papers about novel biomarkers investigation propose the comparative analysis of urine samples from healthy subjects and cancer patients and, in general, the population involved is divided into control and sick groups.
In many cases, the control group includes only healthy subjects with no cancer-related pathologies. However, some authors [47,61,63,67,68,78,84,[90][91][92]125,126,136] investigated also the specificity of proposed novel biomarkers towards the cancer type of interest, including in the study also other tumor types or interfering diseases (such as benign prostatic hypertrophy BHP, stones, cysts, oncocytoma, adenoma).
In almost all papers, the cancer patients were divided into subclasses according to their cancer stage with the aim to identify urinary specific-cancer metabolites useful for both tumor detection and prognosis, since one of the main issues related to current screening tools is the over-diagnosis of patients [137][138][139][140]. Indeed, due to their low specificity, actual diagnostic tests may provide positive results also for non-cancerous diseases and do not allow the discrimination between aggressive and indolent cancers, especially for prostate, bladder, liver cancers. This entails the overtreatment of patients affected by non-life-threatening tumors and consequently the increase of patients' management costs (e.g., hospitalization, frequent control tests) [4].
Some authors [62,84] studied also the disease recurrence, analyzing urine samples from patients who developed early biochemical recurrence within two years of surgery and those who remained recurrence-free after more than five years. These researches had the aim to identify specific compounds capable of predicting tumor recurrence with high diagnostic accuracy and thus guide the post-surgery therapeutic plan.
This review focuses also on the experimental protocols suggested in the scientific literature for urine sample preparation and its chemical characterization.
Diverse analytical techniques, involving the analysis of liquid urine or its gaseous headspace, were proposed. The most common ones were the gas or liquid chromatography combined to mass spectroscopy (GC-MS and LC-MS), but also other techniques, as IELC, IMS, or NMR, allowed to achieve very promising results in terms of diagnostic sensitivity and specificity.
In almost all papers, the experimental method involved the freezing and the pre-treatment of urine samples before analysis. The freezing allowed to prevent the alteration of urine composition during the storage due to bacterial activity, while pre-treatments were carried out on liquid urine samples to optimize the following characterization of urine composition.
The type of pre-treatments strictly depends on the analytical technique involved. In general, the pre-treatments consist in the centrifugation of liquid urine just after the collection or after thawing and the addition of chemicals (e.g., NaOH, HCl) to maximize the headspace enrichment in case of GC-MS or facilitate the detection of target compounds (supposed to be indicators of cancerous metabolic alterations) in case of liquid urine analysis.
In the scientific literature, both the qualitative and quantitative characterizations of urine composition were proposed. Some authors [63,66,69,70,78,124,127,141] suggested also the adoption of data processing techniques of the multivariate statistics, as ANOVA, PLS-DA, LDA, for the elaboration of analytical results to investigate the urine metabolic fingerprint of different cancer types with the aim of identify specific metabolites capable of both diagnosing tumors and monitoring their progression.
The critical investigation of the scientific literature in this field highlighted that the chemical characterization of urine composition has the potential to provide various cancer biomarkers, which could contribute to the development of new tests to allow early detection and avoid invasive diagnostic procedure, thereby reducing the economic burden of unnecessary or ineffective treatments.
Unfortunately, since this research field is new and continuously moving, results published until now in the scientific literature are not exhaustive and the discoveries of literary researches cannot yet be applied in everyday clinical practice. Indeed, results need to be validated, and method uncertainty evaluated.
Most existing studies have been compromised due to small number of samples analyzed and the lack of well-defined control groups [142]. Indeed, to achieve the ultimate goal of practicality for clinical applications, the relationship between the presence in urine of specific metabolites and the presence of cancer needs to be validated for large sample sizes with appropriate control groups. For this purpose, the experts in the field strongly recommend the inclusion in the control group of patients with other diseases or disorders with clinical and metabolic profiles close to those of the cancer of interest, in order to assess the specificity of the potential new diagnostic tool towards the cancer of interest [142].
Moreover, in the validation phase, the results' transferability between different experimental protocols should be assessed with the aim to resolve disparities among the findings pointed out by different studies investigating the same disease.

Recurrent Cancer Biomarkers and Their Levels in Urine
Although the critical aspects above mentioned, the in-depth analysis of literary works about urinary cancer biomarkers highlighted that some metabolites were proposed by different authors as qualitative and, in some cases, quantitative indicators of the presence of different cancer types. Table 11 reports a schematization of recurrent urinary cancer biomarkers proposed in the investigated literature and information regarding their concentration trends in urine samples from cancer patients with respect to controls. As already mentioned, it is important to highlight that this research area is continuously moving and evolving, so the data here reported are susceptible of variations with future discoveries in this field. Thus, current data shall be considered as uncertain, although precise data about their uncertainty range are not yet available. Table 11. Concentration levels of recurrent cancer biomarkers in urine samples from cancer patients with respect to controls (> higher concentrations; < lower concentrations; nd: no information about concentrations; -no correlation).

Concentration Levels in Urine from Cancer Patients with Respect to Controls
Glycine, serine, threonine, alanine, and phenylalanine have been proposed as biomarkers related to colon [77], prostate [64,65,70,143], liver [84,89,90], breast [48], bladder [93], lung [33], and gastric [83] cancers. For all cancer types investigated, authors reported that threonine and glycine levels were higher in cancer patients than in healthy subjects, whereas no unique trend has been proposed for serine and alanine. Phenylalanine concentration was higher in cancer patients than controls for all tumors considered, except for prostate cancer, for which the concentration trend in cancer patients with respect to controls was the opposite.
Those recurrent biomarkers are involved in metabolic pathways associated with cells demand and production of energy, as glycolysis, Krebs cycle, oxidative phosphorylation (OXPHOS), glutamine metabolism, fatty acid oxidation, nucleic acid synthesis, lipid synthesis, and amino acid metabolism, which are altered in case of cancer presence. Indeed, many researchers proved that cancer modifies these metabolic pathways involved in the production of energy [143], to satisfy the uncontrolled growth of mutated cells, although mechanisms on which these phenomena are based on have not been fully understood yet.

Investigation of the Correlation between Urine Odour Alteration and Cancer Presence
In recent years, given the criticalities related to the chemical characterization of urine composition and the disparities among different literary works, many researchers started to investigate the possibility to adopt sensorial and sense-instrumental analysis for cancer diagnosis.
In particular, this innovative approach involves the analysis of the odor emanated from urine with the aim detect the presence of the tumors before symptoms appear. This type of analysis provides the characterization of odorous headspaces of urine samples as a whole, providing their "olfactory fingerprints" without identifying the chemical composition of the mixture [4].
This approach has the advantage of simplifying the challenging task of the urine characterization by means of cheaper and faster analytical techniques [4]. Moreover, some of these innovative methods achieved diagnostic sensitivity and specificity higher than the current diagnostic tools [4,16,[144][145][146][147]. As an example, Taverna et al. published in 2015 [26] a pilot study concerning the adoption a rigorous procedure for training two German Shepherd Explosion Detection Dogs to identify a pool of VOCs specific of prostate cancer emanated from urine samples, thereby defining an innovative method for PrC diagnosis. The dogs were taught to sit in front of the cancerous sample after sniffing a set of six urine samples, including one PrC sample and five controls. Urine samples, stored at −20 • C, were defrosted for the analysis and housed in circular perforated metal containers, which were placed in thermally sealed plastic packets to avoid any contamination. Taverna's research involved a huge and multi-faceted population (i.e., 902 participants), including also men and women suffering from different tumors. Diagnostic test performance was evaluated, considering the whole population, after excluding females and considering only control men older than 45 years. In all cases, sensitivity was higher than 98% and specificity was over 96%.
An example of research concerning urine odor analysis by e-nose for cancer diagnosis was published by Horstmann et al. in 2015 [148]. They evaluated the potential of an electronic nose system equipped with MOS sensors for the detection of bladder cancer. Fresh voided urine was collected from 15 patients with the clinical suspicion of primary or recurrent bladder cancer and 21 patients without bladder cancer but with benign urological diseases. The results of this pilot study revealed the high potential of the electronic nose in the detection of bladder cancer with an overall sensitivity of 75% and specificity of 86% necessitating further investigations [7]. This paper does not aim to provide further analysis of the scientific literature regarding the adoption of odor analysis for diagnosing cancer, since comprehensive review papers on this subject have been recently published [7,149]. Nonetheless, the examples here described are representative of the potential of urine odor analysis for diagnostic purposes. Indeed, those very promising results prove the existence of a correlation between urine odor properties and the presence of certain cancer forms. However, this correlation is not yet fully understood and the urine components responsible for these odor alterations have still not been identified.
This review, besides the intent to summarize literary results about novel cancer biomarkers, aims to deeper investigate this aspect. For this purpose, one original aspect of this paper is that it tries to correlate the odor properties of potential urinary biomarkers proposed in the scientific literature for different tumor forms.
The term "odor" refers to the sensation caused by the interaction of some chemical compounds of a gaseous mixture, commonly named "Odorants", with the mammalian olfactory receptors (ISO 5492). The odorants are molecules smaller than 300 amu, characterized by high volatility, which allows them to reach the upper part of the nose and interact with the olfactory receptors.
An odorant is capable to provoke a stimulus in the olfactory system, when it reaches in the atmosphere the "Perception or Odor Threshold" (OT), which is defined as the minimum concentration of the odorant that is perceived by 50% of the exposed population.
In the scientific literature, the OT relevant for diverse odorous pure substances have been proposed [150,151]. However, the tabulated OT values for the same compound are not unique and may differ in order of magnitude, probably due to the different methods adopted for their determination. Nevertheless, the tabulated OT can be considered for a preliminary screening of the odor properties of urine metabolites for which a correlation with various cancer forms has been proved.
This review proposes the investigation of the odor threshold and qualitative description of recurrent biomarkers, fulfilling odorants characteristics (e.g., volatility) with the aim to explore the possibility of identifying those metabolites that tend to alter urine odors, and which may therefore be detectable by trained dogs or electronic noses.
Tabulated OT values refer to human olfaction, and are therefore not directly explicative of the high accuracies in the detection of different types of tumors achieved by trained dogs [152]. Indeed, dogs' olfactory system, which is significantly more powerful than human one, is capable of detecting odor thresholds as low as part per trillion, thanks to the huge dimension of their olfactory epithelium (up to 170 cm 2 vs. 10 cm 2 in humans), the huge number of olfactory receptors (over 200 million vs. nearly 5 million in humans) and the dense innervations of their olfactory mucosa [153]. Table 12 reports the list of cancer biomarkers considered, their odor thresholds and qualitative descriptions and the cancer type for which the correlation has been proven.
Many recurrent biomarkers have characteristic odors, described as unpleasant, pungent or nauseating, and, in some cases (e.g., acetic acid, amine derivatives, pyridine, cresol), their OT are very low. Thus, cancer urinary metabolites can be detected and recognized by the human olfaction at very low concentrations (i.e., ppb level).
These odor properties might confirm the results achieved in recent research studies involving the urine odor analysis by mean of trained dogs or electronic noses to discriminate between control subjects and cancer patients. However, this research field is still growing and, given the disparities among different literary studies about biomarkers concentrations in urine from cancer patients and controls, in this phase no specific considerations aimed at the identification of those molecules detected by trained dogs or electronic nose systems can be made. Nevertheless, these observations prove the possibility to combine the traditional approach based on the chemical characterization of urine composition with the urine odor analysis, with the purpose of adding useful information to the challenging task of cancer biomarker identification. This may provide an innovative pathway for the development of new and more accurate cancer diagnostic tools.
Therefore, despite the difficulties associated with the development of innovative and reliable diagnostic techniques, a significant increase of the research in this field-and hopefully the successful introduction of some of these techniques in clinical diagnosis-in the near future is to be expected due to the high social and economic impact that new technologies for early diagnosis of cancer might have in today's culture [4].

Conclusions
The critical analysis of the scientific literature about urinary cancer biomarkers highlighted the potentialities of metabolomics for the development of innovative cancer diagnostic tool, capable to detect the diseases at early stages and improve the diagnostic accuracy of current procedures. However, results achieved until now are not exhaustive and need to be validated for large sample sizes with appropriate control groups, including also other diseases or disorders with clinical and metabolic profiles close to those of the cancer of interest.
Given the very promising diagnostic accuracies published in recent studies, involving an innovative approach based on urine odor analysis for the development of new diagnostic tool, this paper aimed to deeper investigate the correlation between urine odor alteration and cancer presence. Thus, the investigation of the odor properties of urine biomarkers, for which a correlation with different cancer forms has been discovered, are here reported. This analysis pointed out that some recurrent urinary metabolites can be detected and recognized by the human olfaction at very low concentrations, suggesting the possibility to combine the traditional approach based on the chemical characterization of urine composition with the urine odor analysis with the purpose of simplifying the challenging task of cancer biomarker identification.
Author Contributions: C.B. was responsible of the bibliographic research and conceived the structure and the outline of the paper. B.J.L. contributed to the bibliographic research. S.S. helped with the interpretation of the literature data. L.C. was the responsible of the research project, coordinated the work and revised the paper.
Funding: This research received no external funding.

Conflicts of Interest:
The authors declare no conflict of interest.