PROGNOSTIC UTILITY OF 24 HOUR URINARY 5-HIAA DOUBLING TIME IN PATIENTS WITH NEUROENDOCRINE TUMORS

ENDOCRINE PRACTICE Rapid Electronic Article in Press Rapid Electronic Articles in Press are preprinted manuscripts that have been reviewed and accepted for publication, but have yet to be edited, typeset and finalized. This version of the manuscript will be replaced with the final, published version after it has been published in the print edition of the journal. The final, published version may differ from this proof. DOI:10.4158/EP-2018-0022 2017 AACE. Original Article EP-2018-0022 PROGNOSTIC UTILITY OF 24 HOUR URINARY 5-HIAA DOUBLING TIME IN PATIENTS WITH NEUROENDOCRINE TUMORS Amit Tirosh, MD 1,2, Naris Nilubol, MD 1, Dhaval Patel, MD 1, Electron Kebebew, MD 1,3 Running title: Urinary 5-HIAA Doubling Time in NET, Endocr Pract. 2018;24(No. 8) Submitted for publication date? Accepted for publication May 8, 2018 From the 1 Endocrine Oncology Branch, National Cancer Institute, National Institutes of Health, Bethesda, Maryland, 2 NET Service, Sheba Medical Center, Tel Hashomer and Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel, and 3 Department of Surgery, The George Washington University, School of Medicine and Health Sciences, Washington, DC. Address correspondence to Dr. Electron Kebebew, Endocrine Oncology Branch, National Cancer Institute, National Institutes of Health, Building 10-CRC, Bethesda, MD 20892. Email: kebebewe@mail.nih.gov Published as a Rapid Electronic Article in Press at http://www.endocrinepractice.org. DOI: 10.4158/EP-2018-0022

Copyright 2018 AACE.

ABSTRACT Objective: New clinical prognostic tools are needed to select the population of patients with neuroendocrine tumors (NETs) that have a high risk of disease progression and diseasespecific mortality (DSM). Biochemical biomarker doubling time (DT) is used clinically for prognosis prediction in several solid malignancies. The aim of the current study was to determine whether 24-hour urinary 5-hydroxyindoleacetic acid (5-HIAA) level DT has any prognostic utility in patients with NETs. Methods: Patients with NETs were enrolled in a prospective study with comprehensive biochemical analysis. The current analysis included 90 subjects with increasing 5-HIAA levels in two consecutive measurements. DT was calculated using the Schwartz equation. The primary outcome measures were DSM and disease progression. Results: 5-HIAA DT of <434 days was associated with a higher rate of DSM (P =.02), with positive and negative predictive values for DSM of 75 and 77%, respectively. The difference in DSM was accounted for mainly by patients with small intestine or unknown primary NET (P =.007). In addition, a shorter 5-HIAA DT in patients with small intestine or unknown primary NET was associated with a higher risk of disease progression both in univariate (P =.001) and multivariable analyses (hazard ratio, 19.9; 95% confidence interval, 1.6 to 248.5; P =.02). Conclusion: 5-HIAA DT may be used as a risk stratification tool in patients with small intestine NET or NET of unknown primary and can assist in identifying patients with a high risk for disease progression and DSM. (Endocr Pract. 2018;24:xxx-xxx)

Abbreviations: CT = computed tomography; DSM = disease-specific mortality; DT = doubling time; 5- HIAA = 5-hydroxyindoleacetic acid; MRI = magnetic resonance imaging; NET = neuroendocrine tumor; NETUP = neuroendocrine tumor of unknown primary; PET = positron emission tomography; PFS = progression-free survival; PNET = pancreatic neuroendocrine tumor; ROC = receiver operating characteristic; SINET = small-intestine neuroendocrine tumor INTRODUCTION Neuroendocrine tumors (NETs) are a group of neoplasms originating from neuroendocrine cells, located mainly in the gastrointestinal and respiratory tracts. NETs may be categorized by tumor grade according to the Ki-67 index and mitotic rate (1) and by disease stage (2,3). However, the natural history of patients with NETs is varied even when having the same tumor grade and extent of disease (3); this heterogeneity presents a challenge to the clinician in establishing an optimal surveillance and management plan, which should be individualized for each patient. Biochemical biomarkers are a pivotal clinical tool in the follow-up of patients with NETs. Biomarkers in NETs are commonly divided into specific and nonspecific. The former are associated with clinical syndromes, such as insulin, gastrin, 24-hour urinary 5- hydroxyindoleacetic acid (5-HIAA), and vasoactive intestinal peptide in endogenous hypoglycemia, Zollinger-Ellison syndrome, carcinoid syndrome, and Verner-Morrison syndrome, respectively. The latter, nonspecific biomarkers are used based on the anatomic location of the primary tumor. For example, chromogranin A levels are mainly used for patients with pancreatic and small intestine NETs (4,5).

Serotonin (5-hydroxytryptophan) is synthesized by enterochromaffin cells in the gastrointestinal tract and metabolized and excreted in the urine as 5-HIAA. Serotonin secretion is the main cause for the carcinoid syndrome; hence, it is not surprising that urinary 5-HIAA has 70 to 90% sensitivity and specificity among patients with carcinoid syndrome (6). Serum and plasma 5-HIAA levels have been shown to correlate with 24-hour urinary 5-HIAA levels (7). However, further validation in a larger patient populations is still required (6). The utility of biomarker doubling time (DT) in predicting clinical outcome was initially demonstrated for plasma calcitonin levels in patients with medullary thyroid cancer (8,9) and subsequently for many other cancer types and using various biomarkers (10-12). Measuring urinary 5-HIAA is recommended by the European NET Society guidelines in patients with small-intestine NET (SINET) (5) and also for follow-up of patients with lung, colon, and appendiceal NET if elevated at diagnosis (13). Urinary 5-HIAA was also reported as a potential measure for disease burden in patients with SINET and NET of unknown primary lesion (14). Hence, in the current study, we evaluated the utility of urinary 5-HIAA DT as a prognostic factor in patients with gastroenteropancreatic NETs. METHODS Study Population This was a post hoc analysis of a prospective study on 68 Ga-DOTATATE positron emission tomography/computed tomography (PET/CT) conducted at the National Institutes of Health Clinical Center between the years 2013 and 2017 (NCT01967537). Patients known to have NETs based on anatomic imaging (CT, magnetic resonance imaging [MRI]), 68 Ga- DOTATATE PET, and biochemical evidence and/or a pathologically confirmed NET were enrolled in the study. No patients with poorly differentiated NET were included in this analysis.

The study protocol was as previously described (15). Briefly, patients had annual physical examination, biochemical and imaging assessment, or more frequently if clinically indicated (e.g., for disease progression or treatment response assessment). During each assessment, the patient underwent imaging with 68 Ga-DOTATATE PET/CT, in addition to threephase CT, MRI, and/or 18 F-fluorodeoxyglucose (FDG) PET/CT, and measurement of several biochemical biomarkers, including chromogranin A, pancreatic polypeptide, neuron-specific enolase, gastrin, glucagon, vasoactive intestinal polypeptide, and 24-hour urinary 5-HIAA levels (14). It should be emphasized that this evaluation was performed as part of a research protocol, aiming to evaluate the utility of these imaging modalities and biochemical biomarkers, in addition to the required evaluation according to the current guidelines. The current analysis included only subjects with 24-hour urinary 5-HIAA level with at least two consecutive measurements. All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. The study was approved by the National Cancer Institute Institutional Review Board. Written informed consent was obtained from all study participants. NETs were classified according to the primary tumor location based on anatomic imaging, 68 Ga-DOTATATE PET/CT localization, and/or pathologic diagnosis. A major limitation of the current analysis was the availability of histopathologic grading in 51/90 patients. Primary tumor locations were subdivided into pancreatic NETs (PNETs) or SINETs, whereas subjects with NETs of duodenal (n = 8), rectal (n = 3), lung, thyroid, gastric, large intestine, or appendiceal subtypes (n = 1 each) were grouped as NET of other primary location (NETOP, n = 16) due to their small numbers. Subjects with metastatic NETs with no identified primary lesion were defined as NET of unknown primary (NETUP, n = 8). Disease progression was defined as a new lesion as detected by any imaging modality or a growth of a known lesion

measured by anatomic imaging modalities, during the interval between baseline 68 Ga- DOTATATE PET/CT scan and follow-up imaging. Tumor growth and/or new lesion detection was defined according to the interpretation of an expert radiologist, and results were reviewed and discussed in a dedicated meeting of a multidisciplinary team including endocrinologists, endocrine surgeons, a radiologist, and a nuclear medicine expert. Biochemical Analysis Urinary 5-HIAA levels were measured using liquid chromatography tandem mass spectrometry (Mayo Medical Laboratories, Rochester, MN) (16), with 8 mg/24 hours being the upper limit of the reference range. Patients were instructed to refrigerate their urine sample during collection and avoid consumption of food high in serotonin for 48 hours before or during collection, as well as medications that could affect the results. Urinary collections of insufficient volume were reported by the lab and were not included in the analysis. To further validate our measurements, we calculated the ratio between the two urinary 5-HIAA measurements of each patient (second/baseline), with an average increase of 1.7 and a low standard deviation of 0.9, reflecting a low variability, further supported by a high correlation coefficient between the two measurements (r = 0.94; P<.001). Plasma chromogranin A was measured using immunochemiluminometric assays (Mayo Medical Laboratories). The 24-hour urinary 5-HIAA level DT was calculated using the Schwartz formula (log102 [time interval between tests]/[log10(second value) log10(first value)]) (17,18). Statistical Analyses Statistical calculations were performed using the SPSS 20.0 software (SPSS Inc, Chicago, IL). Results are expressed as mean and SD unless otherwise indicated. For group comparisons, the independent Student s t test was used to analyze differences in parametric

variables, and the chi-square test was used to analyze differences in categorical variables. Nonparametric tests were used as appropriate. Receiver operating characteristic (ROC) curve analysis was performed to assess the discriminative efficacy of 5-HIAA DT for prognosis of patients with gastroenteropancreatic NETs and to determine the optimal DT cut-off for identifying high-risk patients. The Kaplan-Meier estimator with a log-rank (Mantel-Cox) test was used to estimate and compare progression-free survival (PFS) and disease-specific mortality (DSM) rates by 5-HIAA DT. Cox regression analysis was used to estimate the hazard ratio (HR) and 95% confidence interval (CI) for PFS and DSM during follow-up. The regression model included variables that may have impact on the clinical outcome, in addition to plasma chromogranin A levels that were used to adjust for fluctuations in 5-HIAA levels due to possible dietary incompliance (19). The covariates included in the multivariable model were treatment with somatostatin analogues, postinclusion surgery, functional tumor status, and baseline plasma chromogranin A levels, and were added by the enter approach. None of the covariates were significant in univariate analysis. The 2-tailed P value for statistical significance was set at <.05. RESULTS Ninety patients, including 45 (50%) females, were followed for a median of 25 months (range, 10 to 35 months). Forty-one patients (45.6%) had PNETs, 26 patients (28.9%) had SINETs, 7 patients had NETUPs, and 16 patients had NETOPs (Table 1). The peak 24-hour urinary 5-HIAA levels among patients with different NET subtypes are presented in Supplementary Table 1. ROC curve analysis showed discriminative efficacy for 5-HIAA DT for DSM (area under the curve, 0.76), with a cutoff of 434 days giving maximal sensitivity and specificity (Fig. 1). The ROC curve analysis for disease progression showed high accuracy as well, with an area under the curve of 0.67.

[Insert Table 1 near here] Among patients with SINET (n = 26), 4 had newly diagnosed NET, 11 had stable disease, and 11 had progressive/recurrent NET at enrollment. Patients with SINET had tumor grades G1, G2, and G3 in 13, 7, and 2 patients, respectively (4 patients had incomplete data), and 20 patients (76.9%) had liver metastases. Carcinoid syndrome was reported in 15 of 26 patients (57.7%), and 17 patients (65.4%) were treated with somatostatin analogues at enrollment. None of the patients had symptomatic carcinoid heart disease, and 4 patients underwent transthoracic echocardiography that was normal for all cases. [Insert Figure 1 near here] PFS Comparison of the risk for disease progression among the entire cohort revealed comparable risk between patients with short versus long 5-HIAA DT (log-rank test, P =.5; Fig. 2 A). However, subgroup analysis including patients with SINETs and NETUPs revealed significantly higher risk for progression in patients with 5-HIAA DT <434 days compared with longer DT, with a median time to progression of 2 versus 13 months, respectively (log-rank test, P =.001; Fig. 2 B). This difference was statistically significant also in the multivariable analysis (HR, 15.8; 95% CI, 1.3 to 198.0; P =.03). When comparing the risk for progression among patients with SINET, between those with elevated versus nonelevated baseline 24-hour urinary 5-HIAA levels, we found a trend towards a higher risk among those with high 5-HIAA levels; however, this did not reach statistical significance (log-rank test, P =.08). Similarly, no difference was found when comparing those with elevated versus nonelevated serum

chromogranin A levels at baseline (log-rank test, P =.1). No difference in disease progression by 5-HIAA DT was found in patients with PNETs (log-rank test, P =.5; Fig. 2 C). [Insert Figure 2 near here] DSM Four patients died during follow-up, 2 patients had PNETs, and 2 patients harbored NETUPs. All 4 patients had metastatic disease to lymph nodes, 1 patient had also liver metastases, and 2 patients had liver and bone metastases. In the Kaplan-Meier analysis, patients with a 5-HIAA DT <434 days had a higher risk for DSM during follow-up in the entire study population (log-rank test, P =.02; Fig. 3 A), mainly accounted for by patients with SINETs or NETUPs with 5-HIAA DT <434 days having higher risk of death (log-rank test, P =.01; Fig. 3 B). In contrast to 5-HIAA DT, no association with risk for DSM was found between patients with elevated versus nonelevated baseline serum chromogranin A levels (P =.2) or baseline 24- hour urinary 5-HIAA levels (P =.2). No difference in term of risk for DSM was found among patients with PNETs, between those with 5-HIAA DT <434 days or higher (P =.6; Fig. 3 C). [Insert Figure 3 near here] DISCUSSION In the current study, we found that patients with shorter 5-HIAA DT had a higher risk for DSM, mainly accounted for by the subgroups of patients with SINETs or NETUPs. In addition, a short 5-HIAA DT in patients with SINETs or NETUPs was associated with a higher risk for disease progression, both on univariate and multivariable analyses.

The estimation of the growth rate of tumor cells is difficult, as it may be exponential growth and logarithmic in its nature (10). This may be solved by using DT for estimating tumor burden dynamics, either of tumor diameter, tumor volume, or biochemical biomarkers. DT is calculated as log102 divided by the change of the logarithmically transformed biomarker levels over time. This slope may be derived either through a linear regression or using the Schwartz formula (17). The Schwartz formula may be adopted to specific measures (diameter, volume, or biomarker levels) and was shown to be useful both for biochemical biomarker DT (10,18) and for tumor diameter and volume DT calculations (20). Previous studies have shown that 24-hour urinary 5-HIAA levels reflect tumor burden among patients with SINET (14); hence, the use of DT measurement enables reliable estimation of disease dynamics, reflecting the rate of disease burden change. This is the main strength of 5- HIAA DT over any static marker, as it is time dependent and explains the superiority of 5-HIAA DT over baseline 5-HIAA and chromogranin A as a prognostic factor, even though both are considered accurate biomarkers in patients with SINET. Our study utilized 68 Ga-DOTATATE PET/CT for the diagnosis and follow-up of patients with NETs. 68 Ga-DOTATATE PET/CT is a highly sensitive imaging method, allowing the detection of NETs that might not have been identified using anatomic imaging modalities or older somatostatin-receptor based imaging (Octreoscan) (15). The current analysis is based on a larger cohort of patients in a prospective study comparing the utility of 68 Ga-DOTATATE PET/CT to other anatomic (CT and MRI) and functional ( 18 F-FDG PET/CT and Octreoscan) imaging modalities for NET detection. In this analysis, 68 Ga-DOTATATE had higher detection rate compared with Octreoscan (95.1% vs. 30.9%, respectively) and compared with anatomic imaging modalities (95.1% vs. 45.3%; P<.001 for both comparisons) (15). Hence, the use of 68 Ga-DOTATATE PET/CT might have improved our assessment of 5-HIAA DT as a prognostic factor in patients with NETs.

Recently, Dasari et al (21) reported that patients harboring NET with versus without carcinoid syndrome may have shorter survival. However, as patients with carcinoid syndrome usually have advanced disease, this work was criticized for not controlling for disease burden, which might explain the shorter survival (22) and as was shown in multivariable analyses earlier (19). However, other works have shown that low urinary 5-HIAA levels are associated with favorable outcome (23). As was previously shown, SINET disease burden, assessed by 68 Ga- DOTATATE based volumetric measurement, is positively correlated with urinary 5-HIAA levels (14). Hence, the calculation of 5-HIAA DT in the current study reflects tumor burden dynamics and not only serotonin secretion. In addition, in the current analysis, we assessed 5- HIAA dynamics, in contrast to single 5-HIAA measurement or the presence of carcinoid syndrome as a sole predictor for patient prognosis. We found an association between 5-HIAA DT and DSM in patients with NETs of various types. Urinary 5-HIAA has a high specificity and sensitivity for SINETs (5,24). Hence, it is not surprising that 5-HIAA DT in our analysis was associated with DSM and PFS most pronouncedly among patients with SINETs and among those with NETUPs, which most commonly derive from small SINETs. Nevertheless, 5-HIAA DT is possibly a marker for NETs of other origins. Serotonin secretion was also reported in other NET types, including NETs of foregut or hindgut origins (25). Moreover, Soga (26) reported a 23.3% incidence of carcinoid syndrome in patients with PNETs, compared with 35.8% in SINETs and 24.1% in ileocecal NETs. This is in contrast to other reports of an incidence of carcinoid syndrome in <5% of patients with pancreatic NET (27). Interestingly, the prognostic utility of urinary 5-HIAA was already shown in the past in patients with mixed types of NETs (28). Thus, the association between 5-HIAA DT and DSM for the entire cohort might not derive solely from patients with SINETs or NETUPs but also from patients with NETs of other sites in the gastrointestinal tract. It is difficult to estimate the secretion of serotonin by pancreatic NETs, as 5-HIAA measurement

is not usually performed or indicated in patients with pancreatic NETs for diagnosis or follow-up according to current guidelines (6). The current analysis has several limitations. First, the use of urinary 5-HIAA levels might be biased due to medical treatment with somatostatin analogues and might reflect both disease burden and the functional status of the tumor (22,19). Hence, it is important to correct for these factors in the multivariable analysis, as was performed in the current work. Second, the study cohort includes patients with NETs which typically do not secrete serotonin and with localized NETs not associated with carcinoid syndrome. Third, the study sample was relatively small, and this should be taken into consideration, especially regarding the multivariable analysis results. Our study findings may assist in identifying patients with NETs that are more likely to die of their disease or that are more likely to have disease progression. These patients may benefit from early intervention or from more-aggressive therapeutic approaches or closer surveillance and imaging to identify sites of disease that may require treatment. This is especially important in light of the expanding therapeutic arsenal, ranging from active surveillance to potent biologic medical therapy, chemotherapy, and radioactive therapeutic modalities. The current analysis is limited by the heterogeneous NET subtypes and by the limited follow-up duration. Nevertheless, we have clearly shown the association between DT and patient prognosis and opened the opportunity to assess this clinical tool in other NET patient populations. CONCLUSION In conclusion, 5-HIAA DT may be used as a risk stratification tool in patients with NETs, mainly among those with SINETs or NETUPs, and can be used to assist in identifying patients with a high risk of disease progression or death due to their disease. DISCLOSURE

The authors have no multiplicity of interest to disclose. ACKNOWLEDGMENT The study was funded by the intramural research program of the Center for Cancer Research, National Cancer Institute. [Insert Supplementary Table here]

REFERENCES 1. Klimstra DS, Modlin IR, Coppola D, Lloyd R V, Suster S. The pathologic classification of neuroendocrine tumors: a review of nomenclature, grading, and staging systems. Pancreas. 2010;39:707-712. 2. Amin MB, Edge SB, American Joint Committee on Cancer. AJCC cancer staging manual. Available at: http://www.springer.com/us/book/9783319406176. Accessed November 6, 2017. 3. Perren A, Couvelard A, Scoazec JY, et al. ENETS consensus guidelines for the standards of care in neuroendocrine tumors: pathology: diagnosis and prognostic stratification. Neuroendocrinology. 2017;105:196-200. 4. Falconi M, Eriksson B, Kaltsas G, et al. ENETS consensus guidelines update for the management of patients with functional pancreatic neuroendocrine tumors and nonfunctional pancreatic neuroendocrine tumors. Neuroendocrinology. 2016;103:153-171. 5. Niederle B, Pape UF, Costa F, et al. ENETS consensus guidelines update for neuroendocrine neoplasms of the jejunum and ileum. Neuroendocrinology. 2016;103:125-138. 6. Oberg K, Couvelard A, Delle Fave G, et al. ENETS consensus guidelines for standard of care in neuroendocrine tumours: biochemical markers. Neuroendocrinology. 2017;105:201-211. 7. Adaway JE, Dobson R, Walsh J, et al. Serum and plasma 5-hydroxyindoleacetic acid as an alternative to 24-h urine 5-hydroxyindoleacetic acid measurement. Ann Clin Biochem. 2016;53:554-560. 8. Miyauchi A, Onishi T, Morimoto S, et al. Relation of doubling time of plasma calcitonin levels to prognosis and recurrence of medullary thyroid carcinoma. Ann Surg. 1984;199:461-466.

9. Barbet J, Campion L, Kraeber-Bodéré F, Chatal JF, GTE Study Group. Prognostic impact of serum calcitonin and carcinoembryonic antigen doubling-times in patients with medullary thyroid carcinoma. J Clin Endocrinol Metab. 2005;90:6077-6084. 10. Kelloff GJ, Coffey DS, Chabner BA, et al. Prostate-specific antigen doubling time as a surrogate marker for evaluation of oncologic drugs to treat prostate cancer. Clin Cancer Res. 2004;10:3927-3933. 11. Rössing RM, Jentzen W, Nagarajah J, Bockisch A, Görges R. Serum thyroglobulin doubling time in progressive thyroid cancer. Thyroid. 2016;26:1712-1718. 12. Masuda T, Beppu T, Horino K, et al. Preoperative tumor marker doubling time is a useful predictor of recurrence and prognosis after hepatic resection of hepatocellular carcinoma. J Surg Oncol. 2010;102:490-496. 13. Knigge U, Capdevila J, Bartsch DK, et al. ENETS consensus recommendations for the standards of care in neuroendocrine neoplasms: follow-up and documentation. Neuroendocrinology. 2017;105:310-319. 14. Tirosh A, Papadakis GZ, Millo C, et al. Association between neuroendocrine tumors biomarkers and primary tumor site and disease type based on total 68 Ga-DOTATATE-avid tumor volume measurements. Eur J Endocrinol. 2017;176:575-582. 15. Sadowski SM, Neychev V, Millo C, et al. Prospective study of 68 Ga-DOTATATE positron emission tomography/computed tomography for detecting gastro-enteropancreatic neuroendocrine tumors and unknown primary sites. J Clin Oncol. 2016;34:588-596. 16. Kroll CA, Magera MJ, Helgeson JK, Matern D, Rinaldo P. Liquid chromatographic tandem mass spectrometric method for the determination of 5-hydroxyindole-3-acetic acid in urine. Clin Chem. 2002;48:2049-2051. 17. Schwartz M. A biomathematical approach to clinical tumor growth. Cancer.

1961;14:1272-1294. 18. Enooku K, Tateishi R, Kanai F, et al. Evaluation of molecular targeted cancer drug by changes in tumor marker doubling times. J Gastroenterol. 2012;47:71-78. 19. Zandee W, Kamp K, van Adrichem RCS, Feelders RA, de Herder WW. Limited value for urinary 5-HIAA excretion as prognostic marker in gastrointestinal neuroendocrine tumors. Eur J Endocrinol. 2016;175:361-366. 20. Nakamura R, Inage Y, Tobita R, et al. Epidermal growth factor receptor mutations: effect on volume doubling time of non small-cell lung cancer patients. J Thorac Oncol. 2014;9:1340-1344. 21. Dasari A, Shen C, Halperin D, et al. Trends in the incidence, prevalence, and survival outcomes in patients with neuroendocrine tumors in the United States. JAMA Oncol. 2017;3:1335-1342. 22. Zandee WT, de Herder WW, Jann H. Incidence and prognosis of carcinoid syndrome: hormones or tumour burden? Lancet Oncol. 2017;18:e299. 23. van der Horst-Schrivers ANA, Post WJ, Kema IP, et al. Persistent low urinary excretion of 5-HIAA is a marker for favourable survival during follow-up in patients with disseminated midgut carcinoid tumours. Eur J Cancer. 2007;43:2651-2657. 24. O Toole D, Grossman A, Gross D, et al. ENETS consensus guidelines for the standards of care in neuroendocrine tumors: biochemical markers. Neuroendocrinology. 2009;90:194-202. 25. Meijer WG, Kema IP, Volmer M, Willemse PH, de Vries EG. Discriminating capacity of indole markers in the diagnosis of carcinoid tumors. Clin Chem. 2000;46:1588-1596. 26. Soga J. Carcinoids of the pancreas: an analysis of 156 cases. Cancer. 2005;104:1180-1187. 27. Tsoukalas N, Chatzellis E, Rontogianni D, et al. Pancreatic carcinoids (serotonin-

producing pancreatic neuroendocrine neoplasms) Medicine (Baltimore). 2017;96:e6201. 28. Formica V, Wotherspoon A, Cunningham D, et al. The prognostic role of WHO classification, urinary 5-hydroxyindoleacetic acid and liver function tests in metastatic neuroendocrine carcinomas of the gastroenteropancreatic tract. Br J Cancer. 2007;96:1178-1182.

Fig. 1. Receiver operating characteristic curve analysis for disease-specific mortality during follow-up by 24-hour urinary 5-hydroxyindoleacetic acid (5-HIAA) doubling time (DT). AUC = area under the curve; NPV = negative predictive value; PPV = positive predictive value; Sens = sensitivity; Spec = specificity. Fig. 2. Kaplan-Meier analysis for disease progression according to 24-hour urinary 5- hydroxyindoleacetic acid (5-HIAA) level doubling time (DT) in the entire cohort (n = 49, A) and only patients with small-intestine neuroendocrine tumors (NETs), those with NET of unknown primary location (n = 19, B), and only patients with pancreatic NETs (n = 23, C). Fig. 3. Kaplan-Meier analysis for disease-specific survival according to 5-hydroxyindoleacetic acid (5-HIAA) doubling time (DT) in the entire cohort (n = 90, A), and including only patients with neuroendocrine tumors (NETs) originating from the small intestine or of unknown primary location (n = 33, B), and only patients with pancreatic NETs (n = 41, C).

Table 1 Study Patient Characteristics Compared According to 5-HIAA Doubling Time 5-HIAA doubling time 434 days <434 days Entire cohort n = 66 (73.3%) n = 24 (26.7%) Age at inclusion (years) 53.7 ± 14.7 53.6 ± 15.0 54.2 ± 14.1 NS Female gender n (%) 45 (50.0%) 29 (43.3%) 17 (70.8%).02 Follow-up duration (months) 23.3 ± 7.2 23.8 ± 7.2 22.1 ± 7.3 NS 24-h urinary 5-HIAA doubling time (days) P value 1,361 ± 1,457 1,697 ± 1,522 296 ± 96 <.001 Carcinoid syndrome n(%) 19 (21.1%) 15 (22.7%) 4 (16.7%) NS Primary tumor location PNET 41 (45.6%) 28 (41.8%) 13 (56.5%) SINET 26 (28.9%) 20 (29.9%) 6 (26.1%) Unknown primary location 7 (7.8%) 4 (6.0%) 3 (13.0%) Other 16 (17.8%) 15 (22.4%) 1 (4.3%) Metastatic disease Liver 32 (37.2%) 21 (31.8%) 11 (45.8%) NS Any 48 (55.8%) 35 (53.0%) 13 (54.2%) NS Interventions Somatostatin analogues at inclusion Interventions after inclusion 28 (32.6%) 22 (33.3%) 6 (25.0%) NS Medical therapy 37 (43.0%) 28 (42.4%) 9 (37.5%) NS Surgery 15 (17.4%) 11 (16.7%) 4 (16.7%) NS PRRT/liver directed 6 (7.0%) 4 (6.1%) 2 (8.3%) NS Biochemical biomarkers 24-h urinary 5-HIAA Number of measurements during follow-up 2.9 ± 1.5 2.9 ± 1.4 2.1 ± 1.8 NS Mean value (mg/24 h) 10.1 ± 17.5 8.2 ± 12.8 15.6 ± 26.6.2 Elevated at inclusion (>8 mg/24 h) n (%) Serum chromogranin A levels 22 (25.6%) 13 (19.7%) 9 (37.5%).06 Mean value (ng/ml) 818 ± 2,056 663 ± 1,368 1,286 ± 3,385 NS Elevated at inclusion (>93 ng/ml) n (%) 42 (48.8%) 30 (45.5%) 12 (50.0%) NS WHO 2010 grade G1-2/G3 a n (%) 47/4 (92.2%/7.8%) 32/3 (92.8%/8.1%) 15/1 (93.7%/6.2%) NS Abbreviations: 5-HIAA = 24-hour urinary 5-hydroxyindoleacetic acid; NS = non-significant; PNET = pancreatic neuroendocrine tumor; PRRT = peptide receptor radionuclide therapy; SINET = small-intestine neuroendocrine tumor; WHO = World Health Organization. a Pathology for 3 patients was reported as well differentiated NET, 1 patient in the short 5-HIAA DT and 2 patients in the long 5-HIAA DT groups. Those patients were considered as G1-2 NET. Percentages of total number of patients with available tumor grade. NS

Supplementary Table 1 Peak 24-Hour Urinary 5-HIAA Levels Among Patients With NETs at Various Anatomic Locations PNET SINET Unknown a Other NET b First 24-h urinary 5-HIAA measurement 13 77 121 9.8 Second 24-h urinary 5-HIAA measurement 15 107 229 17 Abbreviations: 5-HIAA = 5-hydroxyindoleacetic acid; NET = neuroendocrine tumor; PNET = pancreatic NET; SINET = small-intestine NET. a NET of unknown primary location b Gastric, duodenal, rectal, lung, or appendiceal subtypes.