hypochlorhydria and achlorhydria are … and achlorhydria are associated with false-positive...

Hypochlorhydria and Achlorhydria Are AssociatedWith False-Positive Secretin Stimulation Testing for

Zollinger-Ellison SyndromePari Shah, MD, MSCE,* Maneesh H. Singh, MD,Þ Yu-Xiao Yang, MD, MSCE,Þþ

and David C. Metz, MDÞþ

Objectives: Secretin stimulation testing (SST) is used to evaluate pa-tients with hypergastrinemia in the diagnosis of Zollinger-Ellison syn-drome. Case series have documented false-positive SST in patients withachlorhydria. This study reviews our experience with SST in hypochlo-rhydric and achlorhydric patients.Methods: We examined 27 patients with hypochlorhydria or achlor-hydria based on a predefined basal acid output (BAO) measurementof less than 5.0 mEq/h who also underwent SST for diagnosis ofZollinger-Ellison syndrome. We report the frequency of false-positiveSST results in this setting.Results: Three hundred thirty patients underwent gastric analysis ofwhich 27 had BAO of less than 5.0 mEq/h and SST conducted. The mean(SD) fasting gastrin level was 247 (304) pg/mL, and the mean (SD) BAOmeasurement was 1.6 (1.8) mEq/h. Twenty patients were off, and 7 wereon antisecretory therapy at time of testing. Four patients had false-positive SST results: 3 with gastric atrophy (BAO = 0 mEq/h) and 1 withdrug-induced hypochlorhydria (BAO = 0.5 mEq/hr). These false-positivetest results were confirmed by structural and functional imaging studies.Conclusions: We have identified a 14.8% false-positive rate in SSTin patients with hypochlorhydria or achlorhydria. Growing literaturehas identified severe consequences associated with discontinuing anti-secretory treatment for testing; therefore, SST will require interpreta-tion in the setting of gastric acid suppression and needs to be interpretedin this context.

Key Words: Zollinger-Ellison syndrome, secretin, achlorhydria,gastrin, proton-pump inhibitor

(Pancreas 2013;42: 932Y936)

Zollinger-Ellison syndrome (ZES) is a condition defined bythe presence of a gastrin-secreting neuroendocrine tumor

resulting in gastric acid hypersecretion (GAH)1,2; these patientsoften present with complications of GAH including peptic ulcerdisease, dyspepsia, gastroesophageal reflux, abdominal pain, anddiarrhea.3 The hallmark of this condition is hypergastrinemiain the presence of low gastric pH4,5 (ie, inappropriate hypergastri-nemia). Gastrin is normally secreted by G-cells in the gastricantrum and regulated by feedback inhibition via somatostatinfrom adjacent D-cells in the presence of a low gastric pH.6 Pa-tients with exogenous sources of gastrin are unresponsive to

these inhibitory mechanisms leading to unopposed gastrin re-lease and subsequent GAH.7,8

Inappropriate hypergastrinemia is the mainstay of ZES di-agnosis. This is generally defined as an elevated serum gastrinin the presence of a gastric pH less than 3.0. Gastric pH canbe measured during routine upper endoscopy by aspiration ofgastric secretions; however, if the gastric pH is greater than3.0 and less than 6.0, formal testing with gastric analysis tocalculate true basal acid output (BAO) is used for further eval-uation.7 Basal acid output is measured by the passage of a naso-gastric tube into the dependent portion of the stomach withaspiration and quantification of gastric juice production over a1-hour period. A BAO of greater than 15 mEq/h is highly sug-gestive of ZES. However, BAO testing is not available at allcenters, and patients with ZES can have variable levels of acidoutput.7,9

In an effort to improve the diagnosis of ZES, several pro-vocative tests have been developed including the secretin stimu-lation test (SST).10Y14 The SST is based on the observation thatserum gastrin rises in an unopposed fashion in patients with ZESin response to secretin injection.15 This is theorized to occurdue to a lack of inhibitory somatostatin-releasing D cells adja-cent to tumor cells and the presence of secretin receptors onthe gastrinoma cells themselves.16 Various diagnostic cutoffvalues have been proposed for the determination of a positive testfor ZES. In a recent study, of 830 patients with ZES, Berna et al17

proposed a cutoff of 120 pg/mL or greater as the new standardcriterion for diagnosis with an associated sensitivity and speci-ficity of 94% and 100%, respectively. The McGuigan criteria,now less commonly used, proposed an increase of 200 pg/mL orgreater as a positive result,14 although Berna et al17 found thiscutoff to be associated with a sensitivity and specificity of 83%and 100%.

Despite being a well-defined entity, the diagnosis of ZESremains challenging. Patients with ZES may have fluctuatingor modestly increased levels of serum gastrin.3 In addition,other conditions can result in elevated serum gastrin concentra-tions including G-cell hyperplasia and conditions associatedwith achlorhydria such as chronic atrophic gastritis or phar-macologic therapy with proton pump inhibitors (PPIs).10

The now ubiquitous use of PPIs has provided gastroenter-ologists with another notable challenge in the diagnosis ofZES. As described by Corleto et al,18 effective control of gas-tric acid secretion may mask ZES symptoms and potentiallydelay diagnosis allowing an underlying gastrinoma to growunchecked. More directly applicable to this study, testing pa-tients while on PPIs can affect serum gastrin levels and SSTresponse19; however, stopping antisecretory therapy in thesepatients can lead to severe rebound GAH with potentially fatalconsequences.20

To date, there have been a number of case reports andseries describing false-positive rises in gastrin after secretin

ORIGINAL ARTICLE

932 www.pancreasjournal.com Pancreas & Volume 42, Number 6, August 2013

From the *Division of Gastroenterology and Nutrition, Memorial-SloanKettering Cancer Center, New York, NY; †Department of Medicine; and‡Division of Gastroenterology, Hospital of the University of Pennsylvania,Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA.Received for publication September 25, 2012; accepted December 21, 2012.Reprints: Yu-Xiao Yang, MD, MSCE, Department of Medicine Division

of Gastroenterology Hospital of the University of Pennsylvania,1 Convention Ave, 9 Penn Tower Philadelphia, PA 19104(e-mail: [email protected]).

This study was supported by NIH T32-DK007740.The authors declare no conflict of interest.Copyright * 2013 by Lippincott Williams & Wilkins

Copyright © 2013 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

mailto:[email protected]

injection in patients who are achlorhydric from atrophic gastri-tis or PPI therapy.19,21Y24 However, no clinical studies haveevaluated the frequency at which false-positive results occur.In the era of PPI use and with a growing literature citing thehazards of PPI cessation in patients suspected of having ZES,the answer to this question has become increasingly significant.

We performed a retrospective descriptive analysis exam-ining the frequency of false-positive SST results in patients whowere hypochlorhydric or achlorhydric at the time of testing. Wepresent our observed rate of false-positive tests in this subgroupof patients and offer recommendations on how to proceed inthe diagnosis of ZES in this growing class of patients.

MATERIALS AND METHODSThis study was reviewed and approved by the institutional

review board at the Hospital of the University of Pennsylvania.All patients who underwent both gastric analysis with BAOcalculation and SST at this institution between January 1994and September 2009 were identified and reviewed. Patientswere referred for evaluation of ZES because of an elevatedfasting gastrin level.

Basal acid output was calculated after aspiration of gastriccontents via nasogastric tube placed into the dependent portionof the stomach after an overnight fast, as previously described.7

For this study, we predefined hypochlorhydria as a calculatedBAO of less than 5 mEq/h and achlorhydria as a calculatedBAO of 0 mEq/h.

All patients with hypochlorhydria and achlorhydria in thesetting of an intact stomach were included in the study popula-tion, whereas those with prior gastric acidYreducing surgerywere excluded. Clinical and demographic data, BAO measure-ments, and SST results were collected for evaluation.

Fasting serum gastrin determinations were performed bya company contracted with the Hospital of the University ofPennsylvania, Associated Regional and University Patholo-gists, Inc (ARUP, Salt Lake City, Utah). The gastrin serumassay is a double-antibody radioimmunoassay kit manufacturedby Diagnostic Products Corporation (Los Angeles, Calif ). Thisassay assesses only biologically active gastrin with no knowncross-reactivity with NH2 terminal or glycine-extended frag-ments. Intraassay and interassay coefficients of variability are5% and 7% at the upper reference limit. The detection limitis approximately 4.5 pg/mL, and the upper limit of normal forthis assay is 100 pg/mL.

Secretin stimulation testing was performed at our insti-tution as previously described by Frucht et al.10 Before 2001,biologically derived porcine secretin (formerly marketed byFerring Pharmaceuticals, Tarrytown, NY; No longer availablein the United States) was used and administered at a dose of2 clinical units/kg per standard protocol. After 2001, a syntheticsecretin produced by ChiRhoClin, Inc, (Silver Spring, Md) re-placed the biologically derived peptide and was administered atan equivalent dose of 0.4 Kg/kg. The performance of this agentin the diagnosis of ZES was examined by our principal inves-tigator at the time of crossover.25 Blood draws for serumgastrin determination were obtained immediately before intra-venous secretin administration and then 1, 2, 5, 10, 15, and30 minutes later. The primary end point for a positive diag-nosis of ZES was an increase in serum gastrin concentration by120 pg/mL or greater, as defined by Berna et al.17 Secondaryend points consisted of a 50% increase in serum gastrin con-centration and 200 pg/mL or greater increase in serum gastrinconcentration as defined by McGuigan and Wolfe.14

All patients with positive SST result underwent radio-graphic and functional imaging testing with computed tomo-graphic scan and/or octreotide scanning to further evaluate forthe presence of gastrinoma. Patients with positive SST resultbut negative structural and functional imaging and documentedhypochlorhydria or achlorhydria were considered false-positivefor SST.

Statistical analysis was performed using summary statisticsin Stata software (StataCorp LP, College Station, Tex).

RESULTSA total of 330 patients underwent gastric analysis with

BAO calculation at our institution between January 1, 1994,and December 31, 2009. A total of 40 patients had BAO valuesof less than 5 mEq/h and were reviewed for inclusion in thestudy cohort. Of these, 27 patients met the inclusion criteriaof no prior surgery and SST completed and were included inthe final analysis (see Fig. 1). The clinical and demographicinformation for these patients is presented in Table 1. Themean (SD) age of these patients was 47.9 (15.4) years (range,17Y74 years). Nineteen patients were female (70%), and 8 weremale. The mean (SD) BAO was 1.63 (1.75) mEq/h (range,0Y4.9 mEq/h). The mean (SD) basal serum gastrin level atthe time of SST testing was 318 (413) pg/mL (range, 6Y1670pg/mL). Twenty patients were hypochlorhydric or achlorhydricbecause of chronic atrophic gastritis, whereas 7 patients (26%)had decreased acid production due to antisecretory therapy

FIGURE 1. Patient flow chart of all patients undergoing BAO andSST. *Positive SST is defined as an increase in serum gastrinconcentration by 120 pg/mL or greater as discussed in the‘‘Materials and Methods’’ section. †Patients determined to haveZES based on positive SST, symptoms of GAH, and evidence ofgastrinoma on cross-sectional imaging as discussed in the‘‘Materials and Methods’’ section.

Pancreas & Volume 42, Number 6, August 2013 False-Positive SST in Achlorhydria

* 2013 Lippincott Williams & Wilkins www.pancreasjournal.com 933


including omeprazole, lansoprazole, esomeprazole, and rabepra-zole at the time of testing. These patients were kept on antisecre-tory therapy because of severe symptoms of GAH.

The results of all patients with positive SSTs are pre-sented in Table 2. Of 27 patients, 7 patients had positive SSTresults as follows: 3 of these patients had achlorhydria dueto pharmacologic therapy with positive SST result and a con-firmed diagnosis of gastrinoma on imaging. These patientswere true-positive patients with ZES. Two patients were notedto have a BAO of less than 5 mEq/h and positive SST resultwith an increase in gastrin after secretin injection of greaterthan 200 pg/mL. The first was noted to have a BAO of 0.5 mEq/hand was hypochlorhydric because of pharmacologic therapywith rabeprazole dosed at 20 mg twice daily. The second wasnoted to have a BAO of 0 mEq/h and was achlorhydric becauseof atrophic gastritis. These patients had no evidence of gastri-noma on subsequent testing and were considered false posi-tives by SST. Two additional patients were captured as positivefor SST when an increase in serum gastrin after secretin injec-tion of more than 120 pg/mL was used as the diagnostic criteria.Both of these patients were achlorhydric with BAOs of 0 mEq/hbecause of atrophic gastritis and were diagnosed as havingfalse-positive SST result after a negative imaging workup forZES. In total, in our small cohort, 4 (14.8%) of 27 patients(95% confidence interval, 4.74%Y37.38%) were identifiedwith a BAO of less than 5 mEq/h and false-positive SST result.

DISCUSSIONDespite the evolution of provocation testing, ZES remains

a difficult syndrome to diagnose. The introduction of the SSThas aided clinicians in the diagnosis of ZES because of its highsensitivity and specificity; however, there are clinical circum-stances under which a false-positive test result can occur. Severalcase series have reported false-positive SST results in patientswho are hypochlorhydric or achlorhydric because of atrophicgastritis or PPI use.19,21Y23

This study aimed to quantify the frequency of false-positive SST results in 1 cohort of patients at a single institu-tion to provide an assessment of how important it may or maynot be to ensure that SST is only done in combination witha measurement of acid secretory capability. In our cohort ofpatients with hypochlorhydria or achlorhydria, approximately15% have false-positive SST results.

The mechanism by which patients with hypochlorhydriaor achlorhydria develop false-positive SST results is unclear.One proposed mechanism by Goldman et al19 suggests thatpatients with hypochlorhydria and achlorhydria may developa decrease in D-cell density with a relative paucity of the po-tent inhibitor, somatostatin. We support an alternative mecha-nism proposed by others22,26Y28 and shown in Figure 2. Patientswith hypochlorhydria or achlorhydria may develop a relativeG-cell hyperplasia from decreased acid secretion and lack ofinhibitory signaling. G-cells have been shown to release gas-trin in response to secretin29 with serum levels shown to be afunction of antral G-cell mass.30 Thus, we believe that G-cellhyperplasia likely functions in a similar fashion as a gastrinomain response to secretin stimulation resulting in a positive rise ingastrin and false-positive SST result.

This study has significant clinical implications. Protonpump inhibitor therapy is currently one of the most commonlyprescribed classes of medications worldwide with more than20 million prescriptions written annually. In addition, patientswith clinical conditions that trigger the suspicion of ZES arefrequently placed on PPI therapy before SST can be completedas part of the evaluation. Atrophic gastritis can result fromautoimmune conditions such as pernicious anemia or fromchronic gastritis secondary to long-standing Helicobacter pyloriinfection. These conditions frequently go undetected becauseof their insidious and relatively asymptomatic nature. Thus, pa-tients on PPI therapy or with atrophic gastritis may present forSST because of documented hypergastrinemia with unknownunderlying hypochlorhydria or achlorhydria.

In patients on PPI therapy, it is recommended to withholdmedication and recheck fasting serum gastrin before proceed-ing with SST. However, with a developing literature describingthe hazards of gastric acid rebound in patients with ZES takenoff these medications, an understanding of the potential limita-tions of SST in the setting of active PPI use is necessary. Anovel biomarker, pancreastatin, has recently been shown to risein the setting of neuroendocrine tumors independent of PPI useand may become an option in these patients that would precludethe need for PPI cessation; however, its use in the setting ofgastrinoma requires further testing.31,32 For now, to improve di-agnostic accuracy in the setting of PPI use, it is our recommen-dation that patients who are hypochlorhydric or achlorhydric

TABLE 1. Demographic and Clinical Data of all Patients WhoUnderwent BAO and SST

Mean age, y 47.9Female, n (%) 19 (70)Basal gastrin level, mean (SD), mEq/h 318 (431)BAO, mean (SD), mEq/h 1.6 (1.8)On PPI at time of SST, n (%) 7 (26)History of acid-limiting surgery, n 0

TABLE 2. Clinical Data of All Patients With Positive SST Result

Patient No. Age, y SexOn PPI atTime of SST

BAO,mEq/h

Basal GastrinLevel, pg/mL

Delta,pg/mL

Cross-SectionalImaging for ZES Interpretation

1 47 F N 0 1049 131 Negative Atrophy2 37 F N 0 191 213 Negative Atrophy3 51 F N 0 1025 170 Negative Atrophy4 31 M Y 0.5 1670 1300 Negative Drug-induced5 69 F Y 1.3 513 1897 Positive ZES6 27 F Y 0.4 877 341 Positive ZES7 24 F Y 1.2 205 154 Positive ZES

F, female; M, male; N, no; Y, yes.

Shah et al Pancreas & Volume 42, Number 6, August 2013

934 www.pancreasjournal.com * 2013 Lippincott Williams & Wilkins


who present for evaluation with SST have their results inter-preted within the context of their gastric acid status. If ZESis strongly suspected in patients with sporadic disease who maybenefit from exploration and cure, retesting of PPI therapy isonly advised after a careful wean under controlled circum-stances33 and preferably together with gastric acid analysis incenters with experience.

REFERENCES

1. Zollinger RM, Ellison EH. Primary peptic ulcerations of the jejunum

associated with islet cell tumors of the pancreas. Ann Surg.

1955;142(4):709Y723; discussion, 724Y708.

2. Gregory RA, Tracy HJ, French JM, et al. Extraction of a gastrin-like

substance from a pancreatic tumour in a case of Zollinger-Ellison

syndrome. Lancet. 1960;1(7133):1045Y1048.

3. Jensen RT, Gardner JD, Raufman JP, et al. Zollinger-Ellison syndrome:

current concepts and management. Ann Intern Med. 1983;98(1):59Y75.

4. McGuigan JE, Trudeau WL. Immunochemical measurement of

elevated levels of gastrin in the serum of patients with pancreatic

tumors of the Zollinger-Ellison variety. N Engl J Med.

1968;278(24):1308Y1313.

5. Isenberg JI, Walsh JH, Grossman MI. Zollinger-Ellison syndrome.

Gastroenterology. 1973;65(1):140Y165.

6. Wolfe MM, Reel GM, McGuigan JE. Inhibition of gastrin release bysecretin is mediated by somatostatin in cultured rat antral mucosa.J Clin Invest. 1983;72(5):1586Y1593.

7. Metz DC, Starr JA. A retrospective study of the usefulness of acidsecretory testing. Aliment Pharmacol Ther. 2000;14(1):103Y111.

8. Maton P, Dayal Y. Clinical implications of hypergastrinemia. In: ZakimD, Dannenberg A, eds. Peptic Ulcer Disease and Other Acid-Related

Disorders. Armonk, NY: Academic Research Associates, Inc.;1991:213.

9. Klein K. Gastric secretory testing. In: Drossman D, ed. Manual of

Gastroenterologic Procedures. New York, NY: Raven Press; 1993:61.10. Frucht H, Howard JM, Slaff JI, et al. Secretin and calcium provocative

tests in the Zollinger-Ellison syndrome. A prospective study. Ann InternMed. 1989;111(9):713Y722.

11. Hansky J, Soveny C, Korman MG. Effect of secretin on serum gastrin asmeasured by immunoassay. Gastroenterology. 1971;61(1):62Y68.

12. Isenberg JI, Walsh JH, Passaro E Jr, et al. Unusual effect of secretin onserum gastrin, serum calcium, and gastric acid secretion in a patientwith suspected Zollinger-Ellison syndrome. Gastroenterology.1972;62(4):626Y631.

13. Kolts BE, Herbst CA, McGuigan JE. Calcium and secretin-stimulatedgastrin release in the Zollinger-Ellison syndrome. Ann Intern Med.1974;81(6):758Y762.

14. McGuigan JE, Wolfe MM. Secretin injection test in the diagnosis ofgastrinoma. Gastroenterology. 1980;79(6):1324Y1331.

FIGURE 2. Theorized response of serum gastrin level to secretin injection. A, Secretin injection leads to a serum gastrin level increasethat is normally attenuated by D-cell inhibition. This leads to a negative SST. B, Serum gastrin level rises in response to secretin injectiondue to exogenous release from a gastrinoma that is not subject to D-cell inhibition. C, A proposed mechanism for a rise in serumgastrin level in patients with achlorhydria after secretin injection. A relative G-cell hyperplasia from decreased acid secretion and lackof inhibitory signaling leads to an exaggerated response to secretin injection similar to that of a gastrinoma.

Pancreas & Volume 42, Number 6, August 2013 False-Positive SST in Achlorhydria

* 2013 Lippincott Williams & Wilkins www.pancreasjournal.com 935


15. Imamura M, Adachi H, Takahashi K, et al. Gastrin release fromgastrinoma cells stimulated with secretin. Dig Dis Sci.1982;27(12):1130Y1136.

16. Long SH, Berna MJ, Thill M, et al. Secretin-receptor andsecretin-receptor-variant expression in gastrinomas: correlation withclinical and tumoral features and secretin and calcium provocative testresults. J Clin Endocrinol Metab. 2007;92(11):4394Y4402.

17. Berna MJ, Hoffmann KM, Long SH, et al. Serum gastrin inZollinger-Ellison syndrome: II. Prospective study of gastrin provocativetesting in 293 patients from the National Institutes of Health andcomparison with 537 cases from the literature. evaluation of diagnosticcriteria, proposal of new criteria, and correlations with clinical andtumoral features. Medicine (Baltimore). 2006;85(6):331Y364.

18. Corleto VD, Annibale B, Gibril F, et al. Does the widespread use ofproton pump inhibitors mask, complicate and/or delay the diagnosisof Zollinger-Ellison syndrome? Aliment Pharmacol Ther.2001;15(10):1555Y1561.

19. Goldman JA, Blanton WP, Hay DW, et al. False-positive secretinstimulation test for gastrinoma associated with the use of proton pumpinhibitor therapy. Clin Gastroenterol Hepatol. 2009;7(5):600Y602.

20. Poitras P, Gingras MH, Rehfeld JF. The Zollinger-Ellison syndrome:dangers and consequences of interrupting antisecretory treatment.Clin Gastroenterol Hepatol. 2012;10(2):199Y202.

21. Brady CE 3rd, Utts SJ, Dev J. False-positive gastrin rises after secretininjection. J Lab Clin Med. 1985;106(4):461Y462.

22. Feldman M, Schiller LR, Walsh JH, et al. Positive intravenous secretintest in patients with achlorhydria-related hypergastrinemia.Gastroenterology. 1987;93(1):59Y62.

23. Wollmuth RL, Wagonfeld JB. False-positive secretin test. Ann InternMed. 1978;88(5):718Y719.

24. Kuiper P, Biemond I, Masclee AA, et al. Diagnostic efficacy of thesecretin stimulation test for the Zollinger-Ellison syndrome: anintra-individual comparison using different dosages in patients andcontrols. Pancreatology. 2010;10(1):14Y18.

25. Metz DC, Buchanan M, Purich E, et al. A randomized controlledcrossover study comparing synthetic porcine and human secretins withbiologically derived porcine secretin to diagnose Zollinger-EllisonSyndrome. Aliment Pharmacol Ther. 2001;15(5):669Y676.

26. Solcia E, Vassallo G, Sampietro R. Endocrine cells in the antro-pyloricmucosa of the stomach. Z Zellforsch Mikrosk Anat. 1967;81(4):474Y486.

27. Walsh JH, Grossman MI. Gastrin (second of two parts). N Engl J Med.1975;292(26):1377Y1384.

28. Tatsuta M, Itoh T, Okuda S, et al. Effect of fundusectomy on serum andantral gastrin levels in rats. Gastroenterology. 1977;72(1):78Y81.

29. Hattori Y, Imamura M, Tobe T. Gastrin release from antral G cellsstimulated with secretin. Am J Gastroenterol. 1992;87(2):195Y200.

30. Brady CE 3rd, Hyatt JR, Utts SJ. Is the gastrin response to secretinprovocation a function of antral G-cell mass? Results in thehypergastrinemia of acid hyposecretion. J Clin Gastroenterol.1989;11(1):27Y32.

31. Raines D, Chester M, Diebold AE, et al. A prospective evaluation of theeffect of chronic proton pump inhibitor use on plasma biomarker levelsin humans. Pancreas. 2012;41(4):508Y511.

32. Ito T, Igarashi H, Jensen RT. Serum pancreastatin: the long soughtuniversal, sensitive, specific tumor marker for neuroendocrine tumors?Pancreas. 2012;41(4):505Y507.

33. Metz DC. Diagnosis of the Zollinger-Ellison syndrome.Clin Gastroenterol Hepatol. 2012;10(2):126Y130.

Shah et al Pancreas & Volume 42, Number 6, August 2013

936 www.pancreasjournal.com * 2013 Lippincott Williams & Wilkins


hypochlorhydria and achlorhydria are … and achlorhydria are associated with false-positive...

Documents