Presenteof Endo

Accepte

Reprintgery, UnSciencesippel@

0039-60

� 2012

http://d

1232

Symptoms of gastroesophagealreflux disease improve afterparathyroidectomyAlexandra E. Reiher, MD,a,b Haggi Mazeh, MD,a Sarah Schaefer, NP,a Jon Gould, MD,c

Herbert Chen, MD,a and Rebecca S. Sippel, MD,a Madison and Milwaukee, WI, and Evanston, IL

Background. Primary hyperparathyroidism can be associated with symptoms related to GERD, but it isunclear which symptoms of GERD improve after parathyroidectomy. Our goal was to assess prospectivelyfor changes in specific GERD symptoms after parathyroidectomy using a validated questionnaire.Methods. Using the GERD health-related quality of life (GERD-HRQL) questionnaire, symptoms ofheartburn were prospectively assessed before and 6 months after treatment of hyperparathyroidism withparathyroidectomy. This validated questionnaire includes 10 items, with a Likert scale of 0–5. Scoresrange from 0 to 45, a lesser score indicates fewer/less severe symptoms.Results. Pre- and postoperative surveys were available for 51 patients. Parathyroidectomy improved theoverall questionnaire score (12.5 ± 1.3 vs 4.5 ± 0.9, P < .0001). Overall scores for each questionimproved after parathyroidectomy, including symptoms of dysphagia (P = .001) and overall satisfactionwith symptoms (P < .0001). However, the number of patients taking antireflux medication before andafter parathyroidectomy was not substantially different (34 vs 28 patients, P = .17).Conclusion. All symptoms of GERD improved after parathyroidectomy for hyperparathyroidism. Despitethe decrease in symptoms, there was not a change in the number of patients who remained on anti-refluxtherapy. For patients with symptoms of GERD, a trial off antireflux medications after parathyroidectomyshould be considered. (Surgery 2012;152:1232-7.)

From the Section of Endocrine Surgery, Department of Surgery,a University of Wisconsin, Madison, WI;Division of Endocrinology, Department of Internal Medicine,b Northshore University Healthcare System,Evanston, IL; and Department of General Surgery,c Medical College of Wisconsin, Milwaukee, WI

CLASSIC SYMPTOMS ASSOCIATED WITH PRIMARY HYPERPARA-

THYROIDISM (PHPT) include nephrolithiasis andbone disease. However, several symptoms can berelated to PHPT, including polydipsia, polyuria,depression, and decreased appetite. Gastric andduodenal ulceration, as well as heartburn or gas-troesophageal reflux disease (GERD), also havebeen described. The incidence of heartburn inpatients with PHPT is as great as 30%. Othergastrointestinal (GI) manifestations include consti-pation (33%), nausea (24%), and decreased appe-tite (15%). The underlying mechanism remainsunclear.1-3

d as an oral presentation at the American Associationcrine Surgeons, Iowa City, IA, May 1, 2012.

d for publication August 23, 2012.

requests: Rebecca S. Sippel, MD, Department of Sur-iversity of Wisconsin, 600 Highland Avenue, ClinicalCenter J4/703, Madison, WI 53792-3284. E-mail:

surgery.wisc.edu.

60/$ - see front matter

Mosby, Inc. All rights reserved.

x.doi.org/10.1016/j.surg.2012.08.051

SURGERY

Patients with symptomatic PHPT should un-dergo parathyroid surgery to achieve cure, andcertain asymptomatic patients should undergosurgical resection of their adenoma. Currently,symptoms of GERD are not considered an indica-tion for parathyroidectomy (PTX). Criteria devel-oped by the NIH Management of AsymptomaticPrimary Hyperparathyroidism for operative inter-vention include serum calcium concentration of1.0 mg/dL or more above the upper limit ofnormal; creatinine clearance <60 mL/min; bonedensity at the hip, lumbar spine, or distal radiusthat is more than 2.5 standard deviations less thanpeak bone mass (T score <�2.5) and/or previousfragility fracture; or age less than 50 years.4 PTXcures PHPT, improves bone mineral density, maydecrease the risk of fracture, decreases the riskof kidney stones, and improves some quality oflife (QOL) measurements5; however, its effect ongastrointestinal symptoms have been poorlydefined.

GERD is defined by the American College ofGastroenterology as symptoms or mucosal damageproduced by the abnormal reflux of gastric

SurgeryVolume 152, Number 6

Reiher et al 1233

contents into the esophagus.6 The most commonsymptoms associated with GERD are heartburn,dysphagia, and regurgitation. Diagnostic testingfor GERD includes upper endoscopy, ambulatoryesophageal pH monitoring, barium swallow stud-ies, and assessing response to treatment with anti-secretory therapy such as proton pump inhibitorsor H2 antagonists.6,7

Treatment of GERD includes antireflux medi-cations or operative intervention for refractorycases. Standard medical treatment with antirefluxmedications is safe overall but there are side-effectsand risks associated with taking these medications.The side-effect profile for proton-pump inhibitors(PPIs) includes headaches (1–2%), diarrhea (1.9–4.1%), nausea (0.02–0.9%), and rash (0.4–1.1%).Chronic use of PPIs has been associated withdecreased calcium absorption due to decreasedgastric acid, as well as community-acquired pneu-monia.8 There have also been conflicting studiesregarding a possible increased risk of fractures inpatients using proton pump inhibitors for longerthan 6 months.9 Removing such medicationsfrom a patient’s medication list would, therefore,be of benefit to the patient.

Health-related QOL questionnaires are consid-ered a standardized instrument to evaluate theeffects of treatment of GERD in clinical trials.10

The GERD-HRQL questionnaire was designed tomeasure symptomatic outcomes of GERD withthe use of a simple instrument---a questionnaire.The questionnaire was designed and validated byVelanovich et al11 in 1996 and includes 10 items,with a Likert scale of 0–5. The GERD-HRQL totalscore is the sum of the individual item scores,with the worse possible score reaching 45 points.After medical or operative treatment, with an inter-val of 1 to 6 months, patients were again asked tocomplete the GERD-HRQL questionnaire. Thequestionnaire was able to identify the most com-mon and bothersome symptoms of GERD, as wellas differentiate between satisfied and unsatisfiedGERD patients. Velanovich et al11 evaluated 72 pa-tients who were asked to complete the GERD-HRQL at initial assessment.

Although the authors of previous studies havedescribed symptoms of heartburn improving afterPTX,1 no published studies have investigated spe-cific symptoms of GERD before and after PTX.Our goal was to prospectively assess the changesin specific GERD symptoms before and after PTXby using a validated questionnaire, the GERDhealth-related quality of life (GERD-HRQL)questionnaire.

METHODS

This prospective study was performed at a singleacademic institution from March 2010 throughJune 2011 and approved by its institutional reviewboard. Patients undergoing PTX were asked tocomplete a survey as part of their preoperativeevaluation. PTX was performed by 1 of 2 endo-crine surgeons at our institution. All patients wereasked whether they had symptoms of heartburn. Ifpatients answered yes to having heartburn, theywere asked to complete an additional survey, theGERD-HRQL questionnaire. The questionnairewas then provided to patients to complete at theirfollow-up appointment 6 months after PTX.

The GERD-HRQL validated questionnaire in-cludes 10 items, with a Likert scale of 0–5. Six itemsmeasure satisfaction with the degree of heartburnsymptoms, 2 with dysphagia/odynophagia, and1 with the impact of medication on daily life.One item measures overall satisfaction with thepresent condition. Scores range from 0 to 45, witha lower score indicating fewer/less severe symp-toms (Supplementary File 1).10,11

Paired t tests were used to compare individualdata pre and op post-operatively, and unpaired ttests and the Fisher exact were used to comparethe responders and nonresponders. Statistical anal-ysis was performed with Microsoft Excel (2003 forMicrosoft Windows; Microsoft, Redmond, WA). Re-sults are expressed as the mean ± SEM.

RESULTS

Pre- and postoperative surveys were available for51 patients. Mean age at time of PTX was 59 ± 2years, and 78% of patients were female. Mean bodymass index (BMI) was 30.7 ± 6.3 kg/m2 (range,20.1–52.0 kg/m2). Mean preoperative calciumwas 10.6 ± 0.1 mg/dL, and mean preoperativePTH was 142 ± 24 pg/mL. Forty-eight patientshad PHPT, 2 patients had secondary hyperparathy-roidism, and 1 patient had tertiary hyperparathy-roidism. Thirty-two patients had 1 parathyroidadenoma (63%), 11 patients had 2 parathyroid ad-enomas (22%), and 8 patients underwent subtotalPTX (16%). Follow-up surveys were completed 6months postoperatively and were available on100% of patients. Average 6-month postoperativecalcium was 9.4 ± 0.1 mg/dL and average postoper-ative PTH was 50.7 ± 6.7 pg/mL. All patients wereeucalcemic at 6 months’ follow-up.

Overall scores for each question improved afterPTX, including severity of symptoms (P < .0001),the presence of heartburn lying down (P = .001)or standing up (P = .004), symptoms after meals

Table I. Questionnaire scores before and 6months after PTX

Preoperative Postoperative P value

Overall score 12.5 ± 1.3 4.5 ± 0.9 <.0001Severity 2.01 ± 0.20 0.84 ± 0.14 <.0001Present lying down 1.53 ± 0.21 0.76 ± 0.15 .001Present standing up 1.04 ± 0.18 0.49 ± 0.13 .004After meals? 1.16 ± 0.18 0.55 ± 0.13 .002Affects diet 0.71 ± 0.17 0.37 ± 0.11 .03Wakes you from sleep 1.04 ± 0.18 0.49 ± 0.13 .004Difficulty swallowing 1.03 ± 0.21 0.37 ± 0.12 .002Pain with swallowing 0.48 ± 0.13 0.18 ± 0.09 .03Affects daily life 0.55 ± 0.16 0.14 ± 0.07 .02Satisfaction 2.16 ± 0.25 0.55 ± 0.14 <.0001No. patients onheartburnmedications (%)

34 (67%) 28 (55%) .11

PTX, Parathyroidectomy.

SurgeryDecember 2012

1234 Reiher et al

(P = .002), impact of symptoms on diet (P = .03),difficulty swallowing (P = .002), pain with swallow-ing (P = .03), and overall satisfaction with symp-toms (P < .0001). Also, the impact of symptomson daily life improved (P = .02). Interestingly, theoverall number of patients on antireflux medica-tion before and after PTX was not significantly dif-ferent (34 vs 28 patients, P = .11; Table I) Patientswith a preoperative score <20 were older than pa-tients with a pre-operative score$20 (60 ± 2 vs 54 ±2 years, respectively; P = .02.

Surgery significantly improved the overall ques-tionnaire score (12.5 ± 1.3 vs 4.5 ± 0.9; P < .0001).BMI did not explain the changes observed. Therewas not a significant difference in percent changein score between patients with an average BMI <25(�41 ± 30%) and patients with an average BMI$25 (�63 ± 8%), P = .49. Greater than 50%improvement in scores was recorded in 67% of pa-tients after PTX. Only 7 patients (14%) reportedno change in symptoms or worsening of symptoms.There was no difference in preoperativePTH levels (P = 1.0) or calcium levels (P = .57)between patients who had >50% improvement (re-sponders) versus those with <50% improvement(nonresponders) in symptoms of GERD after sur-gery (Table II). Interestingly, the nonrespondershad lesser baseline scores before surgery than theresponders, but this difference was not significant.In addition, there were fewer responders on antire-flux medications 6 months after surgery than non-responders (P = .005).

Patients with a preoperative score of >15 (n =15) had a 68 ± 9% improvement in their scorepost-operatively, while patients with a preoperativescore <15 (n = 36) had a 54 ± 12% improvement intheir score postoperatively. The differences inscore improvement were not substantially different(P = .35).

DISCUSSION

This is the first study to investigate specific GERDsymptoms in patients with PHPT before and 6months after parathyroidectomy. Patients withsymptoms of GERD before parathyroidectomy hadan average score on theGERD-HRQLquestionnaireof 12.5 ± 1.3, which significantly improved to 4.5 ±0.9 after PTX (P < .0001). Overall, each symptomscore improved after PTX. Although symptoms ofGERD have been linked to hyperparathyroidismand hypercalcemia, the pathophysiology of GERDsymptoms and peptic ulcer disease in these patientsremains unclear.12 In a study published in 1982,Mowschenson et al13 demonstrated changes inlower esophageal sphincter pressures before and

after PTX. Patients with heartburn had a loweresophageal sphincter pressure below normal beforeoperation, and 4 of the 5 patients had an increase inlower esophageal pressure.13 Studies have demon-strated increased gastric acid secretion and gastrinlevels in the presence of hypercalcemia, but otherstudies have found conflicting results.14-18

GERD can impact substantially a patient’s QOLand work productivity. In a study that used surveys,authors demonstrated that greater 30% of patientsexperiencing heartburn reported decreased workproductivity.19 The impact of GERD symptoms ona patient’s QOL is most important when evaluatingfor treatment response.

Despite significant improvement in symptomsof GERD, there was not a substantial change in thenumber of patients receiving antireflux medica-tion before and after PTX in our study (P = .11).Only 6 patients (12%) had been taken off antire-flux medication after PTX. A study from the Uni-versity of Pittsburgh demonstrated that 11% ofpatients had medications discontinued or de-creased after PTX. However, only 3 patients werenoted to have proton pump inhibitors discontin-ued after surgery.3 A Swedish study demonstratedthat 27% of 97 long-term PPI users were able todiscontinue PPIs after PTX without recurrence ofsymptoms.20 Given the financial and psychologicburden for patients to take medications on a dailybasis, patients and their primary care providersneed to be educated that discontinuation of antire-flux medications after PTX should be considered,especially in patients with a significant improve-ment or resolution of their GERD symptoms.

Classic symptoms of PHPT---nephrolithiasis,bone disease, and neuropsychiatric symptoms---are

Table II. Comparison between patients with >50% improvement in scores versus patients with <50%improvement in scores after PTX

>50% improvement inscores (n = 34, 67%)

<50% improvement inscores (n = 17, 33%) P value

Age, years 60 ± 2 56 ± 3 .34BMI, kg/m2 31.2 ± 1.5 29.5 ± 1.5 .42Preoperative calcium, mg/dL 10.6 ± 0.1 10.7 ± 0.3 .57Preoperative PTH, pg/mL 142 ± 32 142 ± 34 1.025-hydroxy vitamin D, ng/mL 36 ± 3 36 ± 4 .91Patients on antireflux medications

preoperatively, %n = 20 (59%) n = 14 (82%) .09

Patients on antireflux medications6 months postoperatively, %

n = 12 (34%) n = 14 (82%) .005

Average total preoperative score 12.1 ± 1.5 11.9 ± 2.3 .96Average total postoperative score 1.1 ± 0.3 11.4 ± 1.1 <.0001

Data are presented as mean ± SEM.BMI, Body mass index; PTH, parathyroid hormone; PTX, parathyroidectomy.

SurgeryVolume 152, Number 6

Reiher et al 1235

an indication for curative PTX.1-3 For patients with-out these symptoms, the indications for PTX are de-batable.4 Our study showed substantialimprovement in all symptoms of GERD after PTXfor PHPT. Therefore, despite the lack of a clear un-derstanding behind the mechanism of improvedsymptoms, patients with PHPTwho have symptomsof GERD should be considered for curative PTX be-cause it may decrease symptoms and the need forlong-term medication. The use of a preoperativesurvey can help identify such patients.

This is the first study to evaluate specific symp-toms of GERD attributable to hyperparathyroidismand assess for changes in symptoms of GERD aftercurative PTX using a validated questionnaire. TheGERD-HRQL questionnaire was used in this studybecause it is a validated questionnaire, it is easy forpatients to use, and it is easy for physicians tointerpret the results. Despite our study relativelysmall sample size, our results show a fairly impor-tant improvement in the majority of patients. Weselected patients for this study based upon self-reporting of heartburn symptoms. We did not doformal pH monitoring or radiographic imaging inour patients; therefore, we do not know if thesymptoms they reported were truly related togastroesophageal reflux or if the symptoms wererelated to another condition. Therefore, we do notknow if the improvement that was seen was trulydue to changes in gastric pH or decreasing refluxtime or if they are potentially attributable to otherreasons. The questionnaire that we used has beenwell validated and does correlate highly with sen-sitivity to the effects of GERD treatment.

Although all the patients in this study didcomplain of GERD symptoms, this was not the

primary indication for surgery and the degree ofsymptomatology varied among our cohort. There-fore, our results are reflective of the generalpopulation of patients undergoing parathyroidsurgery and may not be representative of thesubset of patients with more severe or refractoryGERD symptoms. Our findings will need to bevalidated with additional studies in this popula-tion. Interestingly, we did find that those patientswith the greatest symptoms pre-operatively werethe most likely to show improvement after PTX.

REFERENCES

1. Chan AK, Duh QY, Katz MH, Siperstein AE, Clark OH. Clin-ical manifestations of primary hyperparathyroidism beforeand after parathyroidectomy. A case-control study. AnnSurg 1995;222:402-12; discussion 412–4.

2. Fraser WD. Hyperparathyroidism. Lancet 2009;374:145-58.3. Melck AL, Armstrong MJ, Stang MT, Carty SE, Yip L. Med-

ication discontinuation after curative surgery for sporadicprimary hyperparathyroidism. Surgery 2010;148:1113-8; dis-cussion 1118–9.

4. Bilezikian JP, Khan AA, Potts JT Jr. Guidelines for the man-agement of asymptomatic primary hyperparathyroidism:summary statement from the third international workshop.J Clin Endocrinol Metab 2009;94:335-9.

5. Heath H 3rd, Hodgson SF, Kennedy MA. Primary hyperpar-athyroidism. Incidence, morbidity, and potential economicimpact in a community. N Engl J Med 1980;302:189-93.

6. DeVault KR, Castell DO. Updated guidelines for the diagno-sis and treatment of gastroesophageal reflux disease. Am JGastroenterol 2005;100:190-200.

7. Giannini EG, Zentilin P, Dulbecco P, Vigneri S, Scarlata P,Savarino V. Management strategy for patients with gastroe-sophageal reflux disease: a comparison between empiricaltreatment with esomeprazole and endoscopy-oriented treat-ment. Am J Gastroenterol 2008;103:267-75.

8. Thomson AB, Sauve MD, Kassam N, Kamitakahara H. Safetyof the long-term use of proton pump inhibitors. World JGastroenterol 2010;16:2323-30.

SurgeryDecember 2012

1236 Reiher et al

9. Ito T, Jensen RT. Association of long-term proton pump in-hibitor therapy with bone fractures and effects on absorp-tion of calcium, vitamin B12, iron, and magnesium. CurrGastroenterol Rep 2010;12:448-57.

10. Chassany O, Holtmann G, Malagelada J, Gebauer U, Doer-fler H, Devault K. Systematic review: health-related qualityof life (HRQOL) questionnaires in gastro-oesophageal re-flux disease. Aliment Pharmacol Ther 2008;27:1053-70.

11. Velanovich V, Vallance SR, Gusz JR, Tapia FV, Harkabus MA.Quality of life scale for gastroesophageal reflux disease.J Am Coll Surg 1996;183:217-24.

12. Abboud B, Daher R, Boujaoude J. Digestive manifestationsof parathyroid disorders. World J Gastroenterol 2011;17:4063-6.

13. Mowschenson PM, Rosenberg S, Pallotta J, Silen W. Effectof hyperparathyroidism and hypercalcemia on lower esoph-ageal sphincter pressure. Am J Surg 1982;143:36-9.

14. Reeder DD, Jackson BM, Ban J, Clendinnen BG, DavidsonWD, Thompson JC. Influence of hypercalcemia on gastricsecretion and serum gastrin concentrations in man. AnnSurg 1970;172:540-6.

15. Patterson M, Wolma F, Drake A, Ong H. Gastric secre-tion and chronic hyperparathyroidism. Arch Surg 1969;99:9-14.

16. Wise SR, Quigley M, Saxe AW, Zdon MJ. Hyperparathyroid-ism and cellular mechanisms of gastric acid secretion. Sur-gery 1990;108:1058-63; discussion 1063–4.

17. Barreras RF. Calcium and gastric secretion. Gastroenterol-ogy 1973;64:1168-84.

18. Stremple JF, Watson CG. Serum calcium, serum gastrin, andgastric acid secretion before and after parathyroidectomyfor hyperparathyroidism. Surgery 1974;75:841-52.

19. Dean BB, Crawley JA, Schmitt CM, Wong J, Ofman JJ. Theburden of illness of gastro-oesophageal reflux disease: im-pact on work productivity. Aliment Pharmacol Ther 2003;17:1309-17.

20. Bjornsson E, Abrahamsson H, Simren M, Mattsson N, Jen-sen C, Agerforz P, et al. Discontinuation of proton pump in-hibitors in patients on long-term therapy: a double-blind,placebo-controlled trial. Aliment Pharmacol Ther 2006;24:945-54.

DISCUSSION

Dr Keith Heller (New York, NY): Did you do any-thing to control for the placebo effect of surgery?Could you have used as a control a comparablegroup of patients with GERD who had a thyroid lo-bectomy, perhaps?

Dr Alexandra Reiher: We did not use thyroid pa-tients as a control group. We did perform the post-operative questionnaire 6 months after PTX toeliminate any placebo effect from the procedure.

Dr Wen Shen (San Francisco, CA): With Tumsbeing, obviously, a double agent, working both inour realm of calcium supplementation, but alsofor the symptoms of GERD, did you control forpeople taking Tums as calcium supplementationafterwards?

Dr Alexandra Reiher: You bring up an excellentpoint. Again, having it be 6 months postopera-tively, any treatment for transient hypocalcemiaor as prophylaxis had been resolved. With patients

who were taking more than 1,000 milligrams perday, that was considered more than just calciumsupplementation.

Dr Wen Shen (San Francisco, CA): Did thatcount within your preoperative GERD medicationlist, Tums as well?

Dr Alexandra Reiher: Yes. Patients were asked ifthey were on medications for heartburn and thento describe it, and we looked at their doses as well.

Dr Marco Raffaelli (Rome, Italy): I would like toknow if you have any data about the diagnosis ofgastroesophageal reflux disease, if it was clinicalor if the patient had undergone an examinationbefore taking drugs. The second question, doyou have any data about the gastrin levels in thisgroup of patients that could be affected by theircalcium levels?

Dr Alexandra Reiher: If I understand your firstquestion regarding medications, the patientswere asked at the time of preoperative evaluationif they had symptoms of heartburn. And if theydid, they completed the questionnaire. And thenthose patients were followed postoperatively. Wedid not perform any other measure of diagnosingGERD, pre- or postoperatively. Regarding calciumlevels, postoperatively, the average calcium at 6months was 9.4 in this group. Gastrin levels werenot measured.

Dr Ashok R. Shaha (New York, NY): One of thebest ways to evaluate the GERD effect is to performfiberoptic laryngoscopy and observe the laryngo-pharyngeal changes. Another way is to measurethe pH levels. I know you did not do that, butthat may be something you may want to pursue ifyou really feel that the PTX does have an effecton GERD.

PTX does have an effect on general symptomsof the patient, and I’m not sure whether thepatients are interpreting the general symptoms ofwellbeing or true GERD. So, in future, you maywant to look more critically at the pH value andlaryngoscopic findings.

Dr Alexandra Reiher: You bring up an excellentpoint. And I agree that the next step would be touse more diagnostic studies, such as pH monitor-ing or endoscopy. But this study did look at specificsymptoms of GERD rather than just generalizedsymptoms. So I think this is the next step towardsa study such as that.

Dr Quan-Yang Duh (San Francisco, CA): Didany of these patients undergo manometry? Thegold standard of diagnosis would be esophagealmanometry.

Dr Alexandra Reiher: No, we did not use spe-cific diagnostic tests. Although I think it’s

SurgeryVolume 152, Number 6

Reiher et al 1237

important to keep in the mind that GERD is re-lated to how symptomatic it is for the patient,and that guides management of GERD.

Dr Stuart Wilson (Milwaukee, WI): Quan Duhstole some of my thunder, but I wanted to makeone comment and ask one question. I think it’sbeen 31 years ago, but right after our first annualmeeting, Bill Silen, who is one of our foundingmembers, was very interested in parathyroid dis-ease and made really good observations. And hehad a group of about a dozen patients, and I thinkhe had talked 5 or 6 of them into esophageal

manometry. This was 31 years ago. And I think 4of the 5 showed clearly that lower esophagealsphincter (LES) pressure that was decreased be-forehand had increased afterwards.

Now, of course, you can have a defect in esoph-ageal hernia so they are not all going to respond,but I think from a historical point of view that wasinteresting. So I was going to ask one question ifyou could separate out your responders and non-responders by measuring LES pressure, becausethe effect of calcium on motility of the esophagusand LES, I think, is clear.