- Email: [email protected]
Gastric Electrical Stimulation Improves Outcomes of Patients With Gastroparesis for up to 10 Years
CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2011;9:314 –319
ORIGINAL ARTICLES—ALIMENTARY TRACT Gastric Electrical Stimulation Improves Outcomes of Patients With Gastroparesis for up to 10 Years RICHARD W. MCCALLUM,* ZHIYUE LIN,‡ JAMESON FORSTER,§ KATHERINE ROESER,‡ QINGJIANG HOU,储 and IRENE SAROSIEK* *Department of Internal Medicine, Texas Tech University Health Sciences Center, El Paso, Texas; and Departments of ‡Medicine, §Surgery, 储Biostatistics, University of Kansas Medical Center, Kansas City, Kansas
BACKGROUND & AIMS: We assessed the long-term clinical outcomes of gastric electrical stimulation (GES) therapy with Enterra (Enterra Therapy System; Medtronic, Minneapolis, MN) in a large cohort of patients with severe gastroparesis. METHODS: Gastroparesis patients (n ⫽ 221; 142 diabetic, 48 idiopathic, and 31 postsurgical) treated with Enterra (Medtronic) for 1–11 years were retrospectively assessed; 188 had follow-up visits and data were collected for at least 1 year (mean 56 months, range 12–131 months). Total symptom scores (TSSs), gastric emptying, nutritional status, weight, hospitalizations, use of prokinetic and/or antiemetic medications, levels of HbA1c levels (in diabetic patients), and adverse events were evaluated at the beginning of the study (baseline) and during the follow-up period. RESULTS: TSS, hospitalization days, and use of medications were significantly reduced among all patients (P ⬍ .05). More patients with diabetic (58%) and postsurgical gastroparesis (53%) had a greater than 50% reduction in TSS than those with idiopathic disease (48%; P ⫽ .32). Weight significantly increased among all groups, and 89% of J-tubes could be removed. At end of the follow-up period, all etiological groups had similar, abnormal delays in mean gastric retention. Thirteen patients (7%) had their devices removed because of infection at the pulse generator site. CONCLUSIONS: GES therapy significantly improved subjective and objective parameters in patients with severe gastroparesis; efficacy was sustained for up to 10 years and was accompanied by good safety and tolerance profiles. Patients with diabetic or postsurgical gastroparesis benefited more than those with idiopathic disease. Keywords: Gastric Electrical Stimulation; Gastric Emptying; Gastroparesis; Nausea and Vomiting.
G
astroparesis (GP) is a gastric motility disorder characterized by delayed gastric emptying without evidence of mechanical obstruction,1 and clinically presents as nausea, vomiting, abdominal pain, with a compromised nutritional state. GP has multiple etiologies but the dominant ones are diabetes (DM) or idiopathic (ID) and postsurgical (PS). Frequent hospital admissions and severe symptoms make GP patients socially restricted. Standard medical therapy targets dietary modifications and the use of antiemetic and prokinetic agents. When medical therapy fails, nutritional support, typically via a feeding jejunostomy tube, is considered. Finally a total gastrectomy surgery may be necessary, particularly in the setting of GP following a partial gastric resection.2
The Enterra therapy was approved by the US Food and Drug Administration (FDA) as a Humanitarian Device Exemption in March 2000 (Enterra Therapy System; Medtronic, Minneapolis, MN) for symptomatic relief in patients with refractory DM and ID GP. This therapy requires surgery to implant an electrical stimulator with high frequency and low energy settings. The approval was largely based on promising positive data from a double-blind controlled trial (Worldwide Anti-Vomiting Electrical Stimulation Study).3 Subsequent open labeled trials with small patient numbers and short-term follow-up reported consistent evidence of reduced GP symptoms, improved nutritional status, and health-related quality of life (Schnelldorfer T, Abell TL, Limehouse VM, et al. Gastric electrical stimulation for gastroparesis: experience with higher setting. Abstract. Neurogastroenterol Motil 2006;18:481).3-16 Results from recent prospective, placebo controlled trials in DM and ID GP patients showed that GES significantly improved subjective and objective responses after 12 months of stimulation.17 The purpose of our study was to retrospectively review a large group of patients with refractory GP who underwent GES implantation at University of Kansas Medical Center (KUMC) from April 1998 to April 2008, including some patients reported in our previous publications which addressed up to 5 years of follow-up.4-9 The goals were to: (1) determine if clinical outcomes from Enterra (Medtronic) therapy can be sustained for as long as 10 years; (2) explore the data at the latest follow-up among DM, ID, and PS GP to investigate their different subjective and objective clinical outcomes; (3) evaluate GES parameters during sustained long-term therapy; (4) assess the natural progression of the disease and morbidity; and (5) review complications during GES therapy.
Methods Study Population The study population was comprised of 221 GP patients (median age of 38 years, range: 18 –70; 164 female, 74%) who were not responding to pharmacotherapy. After the WorldAbbreviations used in this paper: DM, diabetes; GE, gastric emptying; GES, gastric electrical stimulation; GI, gastrointestinal; GP, gastroparesis; ID, idiopathic; PS, postsurgical; TPN, total parenteral nutrition; TSS, total symptom score. © 2011 by the AGA Institute 1542-3565/$36.00 doi:10.1016/j.cgh.2010.12.013
April 2011
A 10-YEAR GES STUDY FOR GASTROPARESIS
315
Table 1. GP Patients Implanted Between 1998 and 2008 Demographic
All patients
Patients with diabetic gastroparesis
Patients with idiopathic gastroparesis
Patients with postsurgical gastroparesis
Patient numbers Age, median (range), y Gender (% female) Duration of gastroparesis, median (range), y
221 38 (18–70) 74 3.5 (1–33)
142 38 (20–65) 68 4 (1–20)
48 38 (18–63) 85 3 (1–27)
31 48 (20–70) 87 3 (1–33)
wide Anti-Vomiting Electrical Stimulation Study3 was completed, we broadened the inclusion criteria for GES beyond DM and ID etiologies to include patients with GP secondary to gastric surgery, specifically related to partial gastric resection, vagotomy, or vagal nerve damage.8 The other key entry criteria (indications and contraindications) remained the same as in already published trials.3,9 Of the 221 enrolled patients, 142 (64%) had DM for a median duration of 18 years (range: 1– 41), 48 were ID, and 31 had PS GP. Patient demographics are shown in Table 1. Twentyeight DM were receiving hemodialysis. The study protocol was approved by the Institutional Review Board (IRB) at University of Kansas Medical Center (KUMC), and consent forms were obtained from all subjects before the initiation of the GES therapy.
Study Protocol The patients were evaluated at baseline and at approximately 3 and 6 months, and 1 year intervals up to the time of their latest follow-up. Their baseline was defined at the 4-week run-in period before surgical placement of GES. The data collected were as follows: (1) upper gastrointestinal (GI) symptoms using a 5-point symptom interview scale; (2) nutritional status and weight; (3) days of hospitalization in the year prior to GES implantation and last year during therapy; (4) medications, specifically prokinetics and antiemetics; (5) gastric emptying utilizing the 4-hour scintigraphic technique with a standardized solid meal18 with glucose levels ⬎75 and ⬍275 mg/dL on the day of the test; (6) GES parameters; (7) HbA1c levels in DM; and (8) adverse events related to GES therapy. After initiating GES, patients were instructed to continue all medications they were taking before implantation. Some tapering of antiemetics was encouraged, as well as changing to smaller portions of low fat, low fiber meals as appropriate based on their clinical status. The detailed descriptions of the GES system, surgical technique, and stimulation parameters have been published previously.3,4,8
Assessment of Upper GI Symptoms Each patient completed a Symptoms Interview Form with a 5-point categorical scale (0 to 4). This form assessed the symptoms of GP occurring during the last 2 weeks and captured severity of 7 symptoms including: vomiting, nausea, early satiety, bloating, postprandial fullness, epigastric pain, and burning.6,7 The severity of each symptom was graded by the patients as 0 ⫽ absent, 1 ⫽ mild (not influencing the usual activities), 2 ⫽ moderate (diverting from, but not urging modifications, of usual activities), 3 ⫽ severe (influencing usual activities enough to urge modifications), and 4 ⫽ extremely severe (requiring bedrest). The sum of the severity ratings of the
7 subscores comprised the overall total symptom score (TSS) for severity.
Statistical Analysis All data were presented as median, interquartile range, and full range, or mean and standard deviation of the mean. Repeated measure analysis with the mixed model19 was used to account for the within subject correlation between baseline and follow-up measures in all patient etiologies separately. For all models with repeated measure analysis, we assumed compound symmetric covariance structure for the error term and used single degree contrast to differentiate the effects within a main factor. A P value of .05 was considered statistically significant. All analyses were performed using SAS statistical software (Version 9.1 of the SAS System for Windows, SAS Institute, Inc, Cary, North Carolina).
Results Patient Flow Twenty-six patients (12%), 23 DM and 3 ID, died of non-GES therapy-related causes (13 less than 1 year postimplant). Twenty out of the 26 (77%) died of cardiovascular or renal complications of DM; 2 patients committed suicide; 2 died of pulmonary embolism in the postoperative period; 1 died of malnutrition/anorexia, and 1 from complicated liver cirrhosis. Twenty-four patients (11%) had their GES devices removed: 13 (7%) caused by infection at the pulse generator or electrode sites, 5 due to lead dislodgement, and 6 related to lack of symptom improvement. Ten patients (7 DM and 3 PS) required a total gastrectomy for failure to improve symptoms, and an additional 10 patients were lost to long-term evaluations. All gastrectomy patients, subjects whose devices were removed due to the lack of efficacy and all patients who died after 6 months of GES (n ⫽ 197) were included in the response rate calculation. The clinical data for at least 1 year of Enterra (Medtronic)—the time that is regarded as adequate time for a response to be established—were analyzed for 188 (85%) of the 221 initially enrolled patients over a 10-year follow-up.
Upper GI Symptoms One hundred eighty-eight patients had follow-up data available for at least 1 year (mean: 56 months; range: 12–131 months). The average TSS severity decreased by 53% ⫾ 32%, improving significantly from 19.4 ⫾ 5.1 at baseline to 9.2 ⫾ 6.6 at the latest follow-up (P ⬍ .001). Our results showed that all 7 individual GP symptoms were significantly reduced for all patients (P ⬍ .0001) (Table 2) and also for all 3 groups of patients (P ⬍ .05). Patients with DM had greater symptom reduction than patients with PS and ID GP (55% vs 48% vs 47%, respec-
316
MCCALLUM ET AL
CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 9, No. 4
Table 2. Comparison of Individual Symptom Scores (Mean ⫾ SD) Between Baseline and Follow-Up Variables
Baseline
Follow-upa
Vomiting Nausea Early satiety Bloating Postprandial fullness Epigastric pain Epigastric burning
3.0 ⫾ 1.2 3.5 ⫾ 1.6 2.9 ⫾ 1.1 2.8 ⫾ 1.2 2.8 ⫾ 1.1 2.5 ⫾ 1.3 2.1 ⫾ 1.4
1.4 ⫾ 1.3 1.6 ⫾ 1.3 1.5 ⫾ 1.3 1.4 ⫾ 1.3 1.4 ⫾ 1.2 1.3 ⫾ 1.3 0.8 ⫾ 1.1
aP
⬍ .0001 compared with baseline (n ⫽ 188).
tively). Fifty-four percent of all patients had a more than 50% reduction in TSS severity, with a greater percent of DM than PS or ID GP (58% vs 53% vs 48%, P ⫽ .32). We used Kaplan–Meier survival analysis on TSS and found that there were no significant differences among the 3 groups of patients if we define 50% reduction from the baseline as the cut point. This was also true if we contrast DM against the 2 other groups. Also, 20% of the patients had a less than 20% improvement after GES therapy (Table 3).
Gastric Emptying Test Of 119 patients with GE follow-up data, 26% normalized their results after GES therapy. Overall, 2-hour gastric retention decreased from a median of 70% at baseline to 66% (P ⬍ .05) and 4-hour retention from 37% to 30% at time of last follow-up (P ⬍ .05; Table 4).
HbA1c Levels in Diabetes Mellitus Mean HbA1c levels in 37 DM patients with available results beyond 1 year were substantially reduced from 8.5% at baseline to 7.8% at last follow-up. Seven (19%) of those patients at baseline had HbA1c levels ⬎10% versus 4 (11%) at follow-up.
Nutritional Outcome and Weight Status All patients with total parenteral nutrition (TPN) (n ⫽ 6), TPN and J-tube (n ⫽ 23), or G-J tube combination (n ⫽ 13) were converted to jejunostomy. The 39 patients, who required a J-tube prior to GES, continued after surgery and 18 patients had a new placement of a J-tube. Thus, there were a total of 99 patients with J-tubes at the initiation of Enterra (Medtronic) Therapy. During follow-up only 11% continued supplemental nutritional support by J-tube (P ⫽ .032 by Fisher exact 2 test) and 2 required TPN. Compared with the results for DM and PS
groups, more patients with ID GP required nutritional support both at baseline and during the GES follow-up period (Table 5). Overall the weight for all patients (n ⫽ 124) significantly increased from 149 ⫾ 41 lbs at baseline to 162 ⫾ 43 lbs (P ⬍ .05) as well as within each group.
Hospitalizations The DM patients had significantly more hospitalization days than both the ID and PS patients at the baseline. The days of hospitalization decreased by 87% (52 ⫾ 4 –7 ⫾ 2 days) (P ⬍ .001) in the last year of follow-up. The major reasons for admissions prior to surgery were due to complications of GP. After surgery, they were related to the complications of DM (poor glucose control, Diabetic Ketoacidosis, and infection), some recurrence of GP symptoms, feeding tube complications, or infection at the pulse generator site, and need for repositioning of the dislodged electrode(s).
Other Interesting Observations All GES surgeries at our center were laparotomy cases. The average operating room time of the procedure was 2.5 hours (median 2, range 2.1–3.2 hours) including time for esophagogastroduodenoscopy, G-J tube revisions, original placement of J-tube, and cholecystectomy or lysis of adhesions, if necessary. In 3 cases repositioning of the electrodes was done based on esophagogastroduodenoscopy confirmation of penetration into the lumen. It was not necessary to reposition electrodes to achieve acceptable readings of impedance (200 – 800 ⍀) during surgery. Full thickness gastric biopsies for research purposes were performed in 87 cases. The average duration of post operation hospitalization was 5 days (median 6, range 2–9 days).
The Use of Medications Overall, the use of GP medications in all patient groups was reduced after 1 year of GES (74% vs 56% of patients given prokinetics, P ⫽ .05, and 65% vs 58% of patients given antiemetics, P ⫽ .025).
Pulse Generator Programming Settings The pulse generator was initially programmed postsurgically with pulse width of 330 sec, current of 5 mA, frequency of 14 Hz, cycle on for 0.1 sec and cycle off for 5.0 sec. At each follow-up visit, the standardized protocol was used to adjust the voltage according to the impedance value, to maintain the current as 5 mA. In the last 2 years of follow-up, incrementally increasing the stimulation current by 1 mA to help overcome
Table 3. Comparison of TSS Among GP Patients (n ⫽ 188) Over 1–10 Years
Baseline Follow-upa % improvement Follow-up (mo) ⬎50% reduction of TSS (n ⫽ 197; see “Upper GI Symptoms” section) aP
⬍ .001 compared with baseline.
Diabetic GP patients (n ⫽ 114)
Idiopathic GP patients (n ⫽ 43)
Postsurgical GP patients (n ⫽ 31)
19.8 ⫾ 5.0 8.7 ⫾ 6.04 55 54 58%
18.6 ⫾ 5.8 9.7 ⫾ 6.2 47 57 48%
19.1 ⫾ 3.4 10.9 ⫾ 7.6 48 63 53%
April 2011
A 10-YEAR GES STUDY FOR GASTROPARESIS
317
Table 4. Comparison of Gastric Retention Among GP Patients Gastric retention %
Diabetic gastroparetic patients (n ⫽ 75)
Idiopathic gastroparetic patients (n ⫽ 20)
Postsurgical gastroparesis (n ⫽ 24)
Baseline at 2 hours Follow-up at 2 hours Baseline at 4 hours Follow-up at 4 hours
70.5 (53.0–86.0)a 68.0 (45.0–84.0) 39.5 (21.0–68.0)a 30.0 (9.0–57.0)
63 (43.0–71.0) 60.5 (53.5–78.0) 30.5 (10.0–40.0) 20.5 (6.2–55.5)
80.5 (68.0–92.0)a 65.0 (35.5–86.0)b 48.0 (33.0–73.0)a 40.0 (4.5–73.0)b
NOTE. Gastric retention reported as median and interquartile range. aP ⬍ .05 compared with idiopathic patients at baseline. bP ⬍ .05 compared with baseline.
persisting GP symptoms was implemented. Voltage adjustment with subsequent change in current up to 10 mA took place in 10% of patients, and in 8 subjects we changed the on time to 1–2 sec on and 3– 4 sec off. The frequency and pulse width of stimulation was not changed in any of our implanted patients. Supplementary Table 1 shows a significant increase of median impedance and voltage values during the GES follow-up period for all 3 subgroups (P ⬍ .001).
Adverse Events Twenty-six patients (12%) died from non-GES therapyrelated causes (see Patient Flow section). The GES system was removed in 24 patients (16 DM, 5 ID, and 3 PS) from 1 to 43 months postimplant. Of these 24 patients, 13 (7%) were due to infection at the pulse generator or electrode sites (timing of infections: 1 patient developed infection 4 years after surgery; 5 patients 1–2 years after GES; 4 patients 6 –12 months postoperative, and 3 had infections in less than 2 months after receiving devices; 2 of these latter 3 patients started with postsurgical [laparotomy] wound infection, which migrated into the pocket housing of the pulse generator). Six of the 24 were removed based on lack of symptom improvement; 2 were related to the lead dislodgements, 1 from a small bowel obstruction caused by the wires, 1 associated with peptic ulcer disease, and 1 was due to penetration of the electrode into the lumen of the stomach. Another 10 patients (7 DM and 3 PS) underwent a total gastrectomy (3 were less than 1 year postimplant) to control their unimproved vomiting episodes and frequent hospitalizations. Ten patients had their GES system repositioned or replaced: 4 were due to depleted battery; 4 were related to lead dislodgement secondary to trauma or twisted wires; and 2 were due to a migration of the device.
No malfunctioning of the Enterra (Medtronic) GES System was reported.
Discussion A single surgeon (JF) using the same open laparotomy surgical approach and 1 gastroenterologist (RWM) and his team were involved in evaluation, decision making, as well as supervising the long term care of these patients receiving gastric neurostimulation therapy. In our current study, we evaluated 188 GP patients with a mean of 56 months of follow-up; the longest and largest single center GES therapy report in the world to date. Overall, the clinical outcomes we observed are consistent with previous shorter term studies reported from our center4,7-9 and other groups,3,5,10,13-16 indicating that sustained improvement and efficacy is able to be maintained with GES therapy for as long as 10 years. One of the major findings from our study is that DM and PS GP had superior outcomes to ID patients. In the DM group, the better symptom control was also reflected by reduction in HbA1c level which has implications for overall better morbidity and mortality in DM patients. The improvement in severity of vomiting makes these patients better candidates for renal and/or pancreas transplant because they could have reliable absorption of immunosuppressant medications. Also, we observed the normalization of GE in 23% of DM compared with 25% of ID and 25% of PS patients. Hence the symptom reduction in diabetics which exceeded that in idiopathic patients was achieved with similar GE changes. The majority of our patients (75%) continued to have similar degrees of delayed GE during the follow-up visits. Observations in symptom outcome are in agreement with previous small and short-term studies3,16 indicating that they
Table 5. Comparison of Use of Supplemental Nutrition at Baseline and During GES Follow-Up Diabetic and nutritional support (n ⫽ 48)
Idiopathic and nutritional support (n ⫽ 22)
Postsurgical and nutritional support (n ⫽ 11)
Nutrition type
Baseline
Follow-up
Baseline
Follow-up
Baseline
Follow-up
J-tube prior to GES TPN only TPN ⫹ J-tube G-J tubes
22 (46) 2 (4) 17 (35) 7 (15)
3 (6) 0 (0) 0 (0) 0
12 (54) 3 (14) 4 (18) 3 (14)
4 (18) 2 (9) 0 (0) 0
5 (46) 1 (9) 2 (18) 3 (27)
2 (18) 0 (0) 0 (0) 0
NOTE. Values are given as n (%). Eighteen gastroparetic patients received a new jejunostomy (J) tube at time of GES surgery. G-J tubes, gastrostomy-jejunal tubes.
318
MCCALLUM ET AL
were more consistent in the DM than the ID subgroup.3 In this current study, we provided a comparison of long-term symptom responses among 3 subgroups. Expectation of therapeutic response based on etiology is important for selecting patients to treat with GES. One explanation for the limited improvement in ID GP patients might be the fact that the idiopathic group consists of a relatively heterogeneous mix of patients and often with more concomitant abdominal pain complaints than seen in other subgroups with GP.3,20 It is well known that abdominal pain is the least likely to be improved by GES.20 In a long-term follow-up study, assessment of complications associated with GES system and overall safety of Enterra (Medtronic) therapy can be better appreciated. In this current study, 24 patients (11%) had their GES devices removed, which is similar to a 4-year follow-up study.10 Another major finding in this study is that symptom improvement with GES therapy is also associated with an increase in stimulation energy with a 6.0 mA median current. This is in agreement with the only 3 studies addressing any changes of stimulation parameters to maximize the symptom improvement with GES therapy.11-13 In 2 reports from Abell’s group11,12 the majority of patients treated with neurostimulation seemed to benefit from higher than baseline settings over time. In these studies, comparison of energy values in small GP subgroups with DM, ID, and PS showed that the postsurgical patients statistically required the most energy. In our current study we observed that the ID group had a greater voltage requirement both at baseline and at follow-up compared with the DM and PS groups, although the ID group had the worst symptom outcome. Hence, we conclude that energy parameters and increasing voltage do not predict or explain the likelihood of symptom reduction. Clinical studies with long-term follow-up often have the problem with missed visits or patients who become lost to follow-up, especially when contemplating the intention-totreat (ITT) analysis versus per protocol analysis in the final data set.21 In the current study, we used repeated measure analysis with the mixed model methodology.19 By fitting marginal models via generalized estimating equations, the mixed model gives an unbiased estimate for parameter estimate in the presence of missing values due to lost to followup. The model also adjusts for potential error deflation effect caused by positive intracluster correlation in response between baseline and follow-up.22 One limitation of this study is that it was not a placebocontrolled trial. Thus, the contribution of a placebo response during GES therapy could not be ruled out. However, the patients in our study had chronic symptoms that had not responded to numerous medications given over at least 1 and sometimes many years before receiving GES therapy and also requiring many admissions to address the complications of GP. Our patients over time had a significant reduction in hospitalization days as well as in use of medications, as we previously reported.23 In addition to increased weight there was essentially no reliance on J-tube and/or oral supplements. These outcomes are too profound to be considered a “placebo” effect; rather they pass the “test-of-time” challenge, where patients were able to be cared for by their local physicians, many in different states (42 states). Recently presented and published results of double-blind, randomized, multicenter, prospective trials of GES in DM and ID GP showed a significant improvement of TSS and quality of life
CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 9, No. 4
at 1 year of follow-up. However, when responses were compared during 3 months of on or off stimulation, there was no significant difference in GP symptom control. One explanation could be related to the study design involving 6 weeks of on stimulation which significantly decreased GP symptoms prior to the double-blind randomized phase. This raises the possibility of a sustained memory or carryover effect from these 6 “active” weeks and emphasizes the neuroplasticity of the central nervous system (CNS) nausea and vomiting control mechanisms.17 In summary, our data show that significant improvements in GP symptoms, nutritional support, days of hospitalization, and the use of medications with an accompanying good safety profile can be achieved and maintained for over 10 years through GES therapy in GP patients not responding to standard medical therapy. Enterra (Medtronic) GES therapy is more effective in treating diabetic and postsurgical gastroparesis than idiopathic GP patients.
Supplementary Material Note: To access the supplementary material accompanying this article, visit the online version of Clinical Gastroenterology and Hepatology at www.cghjournal.org, and at doi:10.1016/ j.cgh.2010.12.013. References 1. Parkman HP, Hasler WL, Fisher RS, et al. Technical review on the diagnosis and treatment of gastroparesis. Gastroenterology 2004; 124:1592–1622. 2. Jones MP, Maganti K. A systematic review of surgical therapy for gastroparesis. Am J Gastroenterol 2003;98:2122–2129. 3. Abell T, McCallum RW, Hocking M, et al. Gastric electrical stimulation for medically refractory gastroparesis. Gastroenterology 2003;125:421– 428. 4. Forster J, Sarosiek I, Delcore R, et al. Gastric pacing is a new surgical treatment for gastroparesis. Am J Surg 2001;182: 676 – 681. 5. Abell TL, Cutsem EV, Abrahamsson H, et al. Gastric electrical stimulation in intractable symptomatic gastroparesis. Digestion 2002;66:204 –212. 6. Familoni BO, Abell TL, Nemoto D, et al. Electrical stimulation at a frequency higher than basal rate in human stomach. Dig Dis Sci 1997;42:885– 891. 7. Lin ZY, Forster J, Sarosiek I, et al. Treatment of diabetic gastroparesis by high-frequency gastric electrical stimulation. Diabetes Care 2004;27:1071–1076. 8. McCallum RW, Lin ZY, Sarosiek I, et al. Clinical response to gastric electrical stimulation in patients with postsurgical gastroparesis. Clin Gastroenterol Hepatol 2005;3:49 –54. 9. Lin ZY, Sarosiek I, Forster J, et al. Symptom responses, long-term outcomes and adverse events beyond 3 years of high-frequency gastric electrical stimulation for gastroparesis. Neurogastroenterol Motil 2006;18:18 –27. 10. Anand C, Al-Juburi A, Familoni B, et al. Gastric electrical stimulation is safe and effective: a long-term study in patients with drug-refractory gastroparesis in three regional centers. Digestion 2007;75:83– 89. 11. Cutts TF, Luo J, Starkebaum W, et al. Is gastric electrical stimulation superior to standard pharmacologic therapy in improving GI symptoms, healthcare resources, and long-term health care benefits? Neurogastroenterol Motil 2005;17:35– 43. 12. Abidi N, Starkebaum WL, Abell TL. An energy algorithm improves symptoms in some patients with gastroparesis and treated with
April 2011
13.
14.
15. 16.
17.
18.
19. 20.
21.
22. 23.
gastric electrical stimulation. Neurogastroenterol Motil 2006; 18:334 –338. Mason RJ, Lipham J, Eckerling G, et al. Gastric electrical stimulation: an alternative surgical therapy for patients with gastroparesis. Arch Surg 2005;140:841– 848. McKenna D, Beverstein G, Reichelderfer M, et al. Gastric electrical stimulation is an effective and safe treatment for medically refractory gastroparesis. Surgery 2008;144:566 –574. Brody F, Vaziri K, Saddier A, et al. Gastric electrical stimulation for gastroparesis. J Am Coll Surg 2008;207:533–538. Maranki JL, Lytes V, Meilahn JE, et al. Predictive factors for clinical improvement with Enterra gastric electric stimulation treatment for refractory gastroparesis. Dig Dis Sci 2008;53: 2072–2078. McCallum R, Snape W, Brody F, et al. Gastric electrical stimulation with Enterra Therapy improves symptoms from diabetic gastroparesis in prospective study. Clin Gastroenterol Hepatol 2010;8:847–954, e1. Tougas G, Eaker EY, Abell TL, et al. Assessment of gastric emptying using a low fat meal: establishment of international control values. Am J Gastroenterol 2000;95:1456 –1462. Little RC, Milliken GA, Stroup WW, et al. SAS system for mixed models. Cary, NC: SAS Institute, 1996. McCallum RW, Dusing RW, Sarosiek I, et al. Mechanisms of symptomatic improvement after gastric electrical stimulation in gastroparetic patients. Neurogastroenterol Motil 2010;22:161– 167. Schoenfeld PS. Evidence-based medicine in practice: applying intention-to-treat analysis and per-protocol analysis. Am J Gastroenterol 2005;100:3– 4. Finney D. Repeated measurements: what is measured and what repeats? Stat Med 1990;9:639 – 644. Lin Z, McElhinney C, Sarosiek I, et al. Chronic gastric electrical stimulation for gastroparesis reduces the use of prokinetic and/or antiemetic medications and the need for hospitalizations. Dig Dis Sci 2005;50:1328 –1334.
A 10-YEAR GES STUDY FOR GASTROPARESIS
319
Reprint requests Address requests for reprints to: Richard W. McCallum, MD, Chairman, Department of Internal Medicine, Texas Tech University Health Sciences Center, Paul L. Foster School of Medicine, 4800 Alberta Avenue, El Paso, Texas 79905. e-mail: [email protected]; fax: (915) 545-6634. Acknowledgments Richard W. McCallum and Irene Sarosiek contributed equally to the work. Medtronic research and technical personnel were not involved in data collection, interpretation of the results, writing, or statistical analysis of this manuscript. The authors thank the Medtronic GI Group, specifically representatives Bjorn Barner and Junius Hill, for their technical support, and Warren Starkebaum for scientific advice over the years. The authors acknowledge the following individuals for their contributions: Pernilla Foran, LPN; Teri Lavenbarg, RN; and Irene King in the Motility Center; and also faculty, fellows, and nursing staff (particularly Sandra Sostarich) in the GI Division and the OR staff, as well as surgical residents for their patient-care contributions. The authors acknowledge Dr Robert Twillman in Psychology and Pain Management, and Drs Reginald Dusing, James Traylor, and Christine McMillin in the Nuclear Medicine Department at KUMC. Conflicts of interest The authors disclose the following: Over the past 10 years, Richard W. McCallum and Irene Sarosiek have occasionally participated in teaching and consulting activities sponsored by Medtronic, Inc. The remaining authors disclose no conflicts. Funding This work was partially financially supported by Medtronic, Inc, Minneapolis, MN, as some of our patients participated initially in 2 double-blind trials.
April 2011
A 10-YEAR GES STUDY FOR GASTROPARESIS
319.e1
Supplementary Table 1. Comparison of GES Parameters Among the 3 Groups of GP Patients
Impedance (⍀) Baseline Follow-up Voltage (V) Baseline Follow-up Current (mA) Baseline Follow-up
Diabetic gastroparetic patients (n ⫽ 118)
Idiopathic gastroparetic patients (n ⫽ 39)
Postsurgical gastroparesis (n ⫽ 24)
480 (390–534) 579 (494–655)a,b
524 (429–579) 617 (511–724)a
465 (379–531) 608 (521–716)a
2.10 (2.00–2.60)c 3.75 (2.70–5.00)a
2.60 (2.00–2.90) 4.15 (3.10–5.00)a
2.20 (2.00–2.70) 4.00 (2.90–5.00)a
5.0 6.0 (3.0–16.0)
5.0 6.0 (3.0–13.0)
5.0 6.0 (3.0–18.0)
NOTE. Impedance and voltage are reported as median and interquartile range. aP ⬍ .001 compared with respective baseline. bP ⬍ .05 compared with idiopathic group at follow-up. cP ⬍ .05 compared with PS ID patients at respective baseline.
ORIGINAL ARTICLES—ALIMENTARY TRACT Gastric Electrical Stimulation Improves Outcomes of Patients With Gastroparesis for up to 10 Years RICHARD W. MCCALLUM,* ZHIYUE LIN,‡ JAMESON FORSTER,§ KATHERINE ROESER,‡ QINGJIANG HOU,储 and IRENE SAROSIEK* *Department of Internal Medicine, Texas Tech University Health Sciences Center, El Paso, Texas; and Departments of ‡Medicine, §Surgery, 储Biostatistics, University of Kansas Medical Center, Kansas City, Kansas
BACKGROUND & AIMS: We assessed the long-term clinical outcomes of gastric electrical stimulation (GES) therapy with Enterra (Enterra Therapy System; Medtronic, Minneapolis, MN) in a large cohort of patients with severe gastroparesis. METHODS: Gastroparesis patients (n ⫽ 221; 142 diabetic, 48 idiopathic, and 31 postsurgical) treated with Enterra (Medtronic) for 1–11 years were retrospectively assessed; 188 had follow-up visits and data were collected for at least 1 year (mean 56 months, range 12–131 months). Total symptom scores (TSSs), gastric emptying, nutritional status, weight, hospitalizations, use of prokinetic and/or antiemetic medications, levels of HbA1c levels (in diabetic patients), and adverse events were evaluated at the beginning of the study (baseline) and during the follow-up period. RESULTS: TSS, hospitalization days, and use of medications were significantly reduced among all patients (P ⬍ .05). More patients with diabetic (58%) and postsurgical gastroparesis (53%) had a greater than 50% reduction in TSS than those with idiopathic disease (48%; P ⫽ .32). Weight significantly increased among all groups, and 89% of J-tubes could be removed. At end of the follow-up period, all etiological groups had similar, abnormal delays in mean gastric retention. Thirteen patients (7%) had their devices removed because of infection at the pulse generator site. CONCLUSIONS: GES therapy significantly improved subjective and objective parameters in patients with severe gastroparesis; efficacy was sustained for up to 10 years and was accompanied by good safety and tolerance profiles. Patients with diabetic or postsurgical gastroparesis benefited more than those with idiopathic disease. Keywords: Gastric Electrical Stimulation; Gastric Emptying; Gastroparesis; Nausea and Vomiting.
G
astroparesis (GP) is a gastric motility disorder characterized by delayed gastric emptying without evidence of mechanical obstruction,1 and clinically presents as nausea, vomiting, abdominal pain, with a compromised nutritional state. GP has multiple etiologies but the dominant ones are diabetes (DM) or idiopathic (ID) and postsurgical (PS). Frequent hospital admissions and severe symptoms make GP patients socially restricted. Standard medical therapy targets dietary modifications and the use of antiemetic and prokinetic agents. When medical therapy fails, nutritional support, typically via a feeding jejunostomy tube, is considered. Finally a total gastrectomy surgery may be necessary, particularly in the setting of GP following a partial gastric resection.2
The Enterra therapy was approved by the US Food and Drug Administration (FDA) as a Humanitarian Device Exemption in March 2000 (Enterra Therapy System; Medtronic, Minneapolis, MN) for symptomatic relief in patients with refractory DM and ID GP. This therapy requires surgery to implant an electrical stimulator with high frequency and low energy settings. The approval was largely based on promising positive data from a double-blind controlled trial (Worldwide Anti-Vomiting Electrical Stimulation Study).3 Subsequent open labeled trials with small patient numbers and short-term follow-up reported consistent evidence of reduced GP symptoms, improved nutritional status, and health-related quality of life (Schnelldorfer T, Abell TL, Limehouse VM, et al. Gastric electrical stimulation for gastroparesis: experience with higher setting. Abstract. Neurogastroenterol Motil 2006;18:481).3-16 Results from recent prospective, placebo controlled trials in DM and ID GP patients showed that GES significantly improved subjective and objective responses after 12 months of stimulation.17 The purpose of our study was to retrospectively review a large group of patients with refractory GP who underwent GES implantation at University of Kansas Medical Center (KUMC) from April 1998 to April 2008, including some patients reported in our previous publications which addressed up to 5 years of follow-up.4-9 The goals were to: (1) determine if clinical outcomes from Enterra (Medtronic) therapy can be sustained for as long as 10 years; (2) explore the data at the latest follow-up among DM, ID, and PS GP to investigate their different subjective and objective clinical outcomes; (3) evaluate GES parameters during sustained long-term therapy; (4) assess the natural progression of the disease and morbidity; and (5) review complications during GES therapy.
Methods Study Population The study population was comprised of 221 GP patients (median age of 38 years, range: 18 –70; 164 female, 74%) who were not responding to pharmacotherapy. After the WorldAbbreviations used in this paper: DM, diabetes; GE, gastric emptying; GES, gastric electrical stimulation; GI, gastrointestinal; GP, gastroparesis; ID, idiopathic; PS, postsurgical; TPN, total parenteral nutrition; TSS, total symptom score. © 2011 by the AGA Institute 1542-3565/$36.00 doi:10.1016/j.cgh.2010.12.013
April 2011
A 10-YEAR GES STUDY FOR GASTROPARESIS
315
Table 1. GP Patients Implanted Between 1998 and 2008 Demographic
All patients
Patients with diabetic gastroparesis
Patients with idiopathic gastroparesis
Patients with postsurgical gastroparesis
Patient numbers Age, median (range), y Gender (% female) Duration of gastroparesis, median (range), y
221 38 (18–70) 74 3.5 (1–33)
142 38 (20–65) 68 4 (1–20)
48 38 (18–63) 85 3 (1–27)
31 48 (20–70) 87 3 (1–33)
wide Anti-Vomiting Electrical Stimulation Study3 was completed, we broadened the inclusion criteria for GES beyond DM and ID etiologies to include patients with GP secondary to gastric surgery, specifically related to partial gastric resection, vagotomy, or vagal nerve damage.8 The other key entry criteria (indications and contraindications) remained the same as in already published trials.3,9 Of the 221 enrolled patients, 142 (64%) had DM for a median duration of 18 years (range: 1– 41), 48 were ID, and 31 had PS GP. Patient demographics are shown in Table 1. Twentyeight DM were receiving hemodialysis. The study protocol was approved by the Institutional Review Board (IRB) at University of Kansas Medical Center (KUMC), and consent forms were obtained from all subjects before the initiation of the GES therapy.
Study Protocol The patients were evaluated at baseline and at approximately 3 and 6 months, and 1 year intervals up to the time of their latest follow-up. Their baseline was defined at the 4-week run-in period before surgical placement of GES. The data collected were as follows: (1) upper gastrointestinal (GI) symptoms using a 5-point symptom interview scale; (2) nutritional status and weight; (3) days of hospitalization in the year prior to GES implantation and last year during therapy; (4) medications, specifically prokinetics and antiemetics; (5) gastric emptying utilizing the 4-hour scintigraphic technique with a standardized solid meal18 with glucose levels ⬎75 and ⬍275 mg/dL on the day of the test; (6) GES parameters; (7) HbA1c levels in DM; and (8) adverse events related to GES therapy. After initiating GES, patients were instructed to continue all medications they were taking before implantation. Some tapering of antiemetics was encouraged, as well as changing to smaller portions of low fat, low fiber meals as appropriate based on their clinical status. The detailed descriptions of the GES system, surgical technique, and stimulation parameters have been published previously.3,4,8
Assessment of Upper GI Symptoms Each patient completed a Symptoms Interview Form with a 5-point categorical scale (0 to 4). This form assessed the symptoms of GP occurring during the last 2 weeks and captured severity of 7 symptoms including: vomiting, nausea, early satiety, bloating, postprandial fullness, epigastric pain, and burning.6,7 The severity of each symptom was graded by the patients as 0 ⫽ absent, 1 ⫽ mild (not influencing the usual activities), 2 ⫽ moderate (diverting from, but not urging modifications, of usual activities), 3 ⫽ severe (influencing usual activities enough to urge modifications), and 4 ⫽ extremely severe (requiring bedrest). The sum of the severity ratings of the
7 subscores comprised the overall total symptom score (TSS) for severity.
Statistical Analysis All data were presented as median, interquartile range, and full range, or mean and standard deviation of the mean. Repeated measure analysis with the mixed model19 was used to account for the within subject correlation between baseline and follow-up measures in all patient etiologies separately. For all models with repeated measure analysis, we assumed compound symmetric covariance structure for the error term and used single degree contrast to differentiate the effects within a main factor. A P value of .05 was considered statistically significant. All analyses were performed using SAS statistical software (Version 9.1 of the SAS System for Windows, SAS Institute, Inc, Cary, North Carolina).
Results Patient Flow Twenty-six patients (12%), 23 DM and 3 ID, died of non-GES therapy-related causes (13 less than 1 year postimplant). Twenty out of the 26 (77%) died of cardiovascular or renal complications of DM; 2 patients committed suicide; 2 died of pulmonary embolism in the postoperative period; 1 died of malnutrition/anorexia, and 1 from complicated liver cirrhosis. Twenty-four patients (11%) had their GES devices removed: 13 (7%) caused by infection at the pulse generator or electrode sites, 5 due to lead dislodgement, and 6 related to lack of symptom improvement. Ten patients (7 DM and 3 PS) required a total gastrectomy for failure to improve symptoms, and an additional 10 patients were lost to long-term evaluations. All gastrectomy patients, subjects whose devices were removed due to the lack of efficacy and all patients who died after 6 months of GES (n ⫽ 197) were included in the response rate calculation. The clinical data for at least 1 year of Enterra (Medtronic)—the time that is regarded as adequate time for a response to be established—were analyzed for 188 (85%) of the 221 initially enrolled patients over a 10-year follow-up.
Upper GI Symptoms One hundred eighty-eight patients had follow-up data available for at least 1 year (mean: 56 months; range: 12–131 months). The average TSS severity decreased by 53% ⫾ 32%, improving significantly from 19.4 ⫾ 5.1 at baseline to 9.2 ⫾ 6.6 at the latest follow-up (P ⬍ .001). Our results showed that all 7 individual GP symptoms were significantly reduced for all patients (P ⬍ .0001) (Table 2) and also for all 3 groups of patients (P ⬍ .05). Patients with DM had greater symptom reduction than patients with PS and ID GP (55% vs 48% vs 47%, respec-
316
MCCALLUM ET AL
CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 9, No. 4
Table 2. Comparison of Individual Symptom Scores (Mean ⫾ SD) Between Baseline and Follow-Up Variables
Baseline
Follow-upa
Vomiting Nausea Early satiety Bloating Postprandial fullness Epigastric pain Epigastric burning
3.0 ⫾ 1.2 3.5 ⫾ 1.6 2.9 ⫾ 1.1 2.8 ⫾ 1.2 2.8 ⫾ 1.1 2.5 ⫾ 1.3 2.1 ⫾ 1.4
1.4 ⫾ 1.3 1.6 ⫾ 1.3 1.5 ⫾ 1.3 1.4 ⫾ 1.3 1.4 ⫾ 1.2 1.3 ⫾ 1.3 0.8 ⫾ 1.1
aP
⬍ .0001 compared with baseline (n ⫽ 188).
tively). Fifty-four percent of all patients had a more than 50% reduction in TSS severity, with a greater percent of DM than PS or ID GP (58% vs 53% vs 48%, P ⫽ .32). We used Kaplan–Meier survival analysis on TSS and found that there were no significant differences among the 3 groups of patients if we define 50% reduction from the baseline as the cut point. This was also true if we contrast DM against the 2 other groups. Also, 20% of the patients had a less than 20% improvement after GES therapy (Table 3).
Gastric Emptying Test Of 119 patients with GE follow-up data, 26% normalized their results after GES therapy. Overall, 2-hour gastric retention decreased from a median of 70% at baseline to 66% (P ⬍ .05) and 4-hour retention from 37% to 30% at time of last follow-up (P ⬍ .05; Table 4).
HbA1c Levels in Diabetes Mellitus Mean HbA1c levels in 37 DM patients with available results beyond 1 year were substantially reduced from 8.5% at baseline to 7.8% at last follow-up. Seven (19%) of those patients at baseline had HbA1c levels ⬎10% versus 4 (11%) at follow-up.
Nutritional Outcome and Weight Status All patients with total parenteral nutrition (TPN) (n ⫽ 6), TPN and J-tube (n ⫽ 23), or G-J tube combination (n ⫽ 13) were converted to jejunostomy. The 39 patients, who required a J-tube prior to GES, continued after surgery and 18 patients had a new placement of a J-tube. Thus, there were a total of 99 patients with J-tubes at the initiation of Enterra (Medtronic) Therapy. During follow-up only 11% continued supplemental nutritional support by J-tube (P ⫽ .032 by Fisher exact 2 test) and 2 required TPN. Compared with the results for DM and PS
groups, more patients with ID GP required nutritional support both at baseline and during the GES follow-up period (Table 5). Overall the weight for all patients (n ⫽ 124) significantly increased from 149 ⫾ 41 lbs at baseline to 162 ⫾ 43 lbs (P ⬍ .05) as well as within each group.
Hospitalizations The DM patients had significantly more hospitalization days than both the ID and PS patients at the baseline. The days of hospitalization decreased by 87% (52 ⫾ 4 –7 ⫾ 2 days) (P ⬍ .001) in the last year of follow-up. The major reasons for admissions prior to surgery were due to complications of GP. After surgery, they were related to the complications of DM (poor glucose control, Diabetic Ketoacidosis, and infection), some recurrence of GP symptoms, feeding tube complications, or infection at the pulse generator site, and need for repositioning of the dislodged electrode(s).
Other Interesting Observations All GES surgeries at our center were laparotomy cases. The average operating room time of the procedure was 2.5 hours (median 2, range 2.1–3.2 hours) including time for esophagogastroduodenoscopy, G-J tube revisions, original placement of J-tube, and cholecystectomy or lysis of adhesions, if necessary. In 3 cases repositioning of the electrodes was done based on esophagogastroduodenoscopy confirmation of penetration into the lumen. It was not necessary to reposition electrodes to achieve acceptable readings of impedance (200 – 800 ⍀) during surgery. Full thickness gastric biopsies for research purposes were performed in 87 cases. The average duration of post operation hospitalization was 5 days (median 6, range 2–9 days).
The Use of Medications Overall, the use of GP medications in all patient groups was reduced after 1 year of GES (74% vs 56% of patients given prokinetics, P ⫽ .05, and 65% vs 58% of patients given antiemetics, P ⫽ .025).
Pulse Generator Programming Settings The pulse generator was initially programmed postsurgically with pulse width of 330 sec, current of 5 mA, frequency of 14 Hz, cycle on for 0.1 sec and cycle off for 5.0 sec. At each follow-up visit, the standardized protocol was used to adjust the voltage according to the impedance value, to maintain the current as 5 mA. In the last 2 years of follow-up, incrementally increasing the stimulation current by 1 mA to help overcome
Table 3. Comparison of TSS Among GP Patients (n ⫽ 188) Over 1–10 Years
Baseline Follow-upa % improvement Follow-up (mo) ⬎50% reduction of TSS (n ⫽ 197; see “Upper GI Symptoms” section) aP
⬍ .001 compared with baseline.
Diabetic GP patients (n ⫽ 114)
Idiopathic GP patients (n ⫽ 43)
Postsurgical GP patients (n ⫽ 31)
19.8 ⫾ 5.0 8.7 ⫾ 6.04 55 54 58%
18.6 ⫾ 5.8 9.7 ⫾ 6.2 47 57 48%
19.1 ⫾ 3.4 10.9 ⫾ 7.6 48 63 53%
April 2011
A 10-YEAR GES STUDY FOR GASTROPARESIS
317
Table 4. Comparison of Gastric Retention Among GP Patients Gastric retention %
Diabetic gastroparetic patients (n ⫽ 75)
Idiopathic gastroparetic patients (n ⫽ 20)
Postsurgical gastroparesis (n ⫽ 24)
Baseline at 2 hours Follow-up at 2 hours Baseline at 4 hours Follow-up at 4 hours
70.5 (53.0–86.0)a 68.0 (45.0–84.0) 39.5 (21.0–68.0)a 30.0 (9.0–57.0)
63 (43.0–71.0) 60.5 (53.5–78.0) 30.5 (10.0–40.0) 20.5 (6.2–55.5)
80.5 (68.0–92.0)a 65.0 (35.5–86.0)b 48.0 (33.0–73.0)a 40.0 (4.5–73.0)b
NOTE. Gastric retention reported as median and interquartile range. aP ⬍ .05 compared with idiopathic patients at baseline. bP ⬍ .05 compared with baseline.
persisting GP symptoms was implemented. Voltage adjustment with subsequent change in current up to 10 mA took place in 10% of patients, and in 8 subjects we changed the on time to 1–2 sec on and 3– 4 sec off. The frequency and pulse width of stimulation was not changed in any of our implanted patients. Supplementary Table 1 shows a significant increase of median impedance and voltage values during the GES follow-up period for all 3 subgroups (P ⬍ .001).
Adverse Events Twenty-six patients (12%) died from non-GES therapyrelated causes (see Patient Flow section). The GES system was removed in 24 patients (16 DM, 5 ID, and 3 PS) from 1 to 43 months postimplant. Of these 24 patients, 13 (7%) were due to infection at the pulse generator or electrode sites (timing of infections: 1 patient developed infection 4 years after surgery; 5 patients 1–2 years after GES; 4 patients 6 –12 months postoperative, and 3 had infections in less than 2 months after receiving devices; 2 of these latter 3 patients started with postsurgical [laparotomy] wound infection, which migrated into the pocket housing of the pulse generator). Six of the 24 were removed based on lack of symptom improvement; 2 were related to the lead dislodgements, 1 from a small bowel obstruction caused by the wires, 1 associated with peptic ulcer disease, and 1 was due to penetration of the electrode into the lumen of the stomach. Another 10 patients (7 DM and 3 PS) underwent a total gastrectomy (3 were less than 1 year postimplant) to control their unimproved vomiting episodes and frequent hospitalizations. Ten patients had their GES system repositioned or replaced: 4 were due to depleted battery; 4 were related to lead dislodgement secondary to trauma or twisted wires; and 2 were due to a migration of the device.
No malfunctioning of the Enterra (Medtronic) GES System was reported.
Discussion A single surgeon (JF) using the same open laparotomy surgical approach and 1 gastroenterologist (RWM) and his team were involved in evaluation, decision making, as well as supervising the long term care of these patients receiving gastric neurostimulation therapy. In our current study, we evaluated 188 GP patients with a mean of 56 months of follow-up; the longest and largest single center GES therapy report in the world to date. Overall, the clinical outcomes we observed are consistent with previous shorter term studies reported from our center4,7-9 and other groups,3,5,10,13-16 indicating that sustained improvement and efficacy is able to be maintained with GES therapy for as long as 10 years. One of the major findings from our study is that DM and PS GP had superior outcomes to ID patients. In the DM group, the better symptom control was also reflected by reduction in HbA1c level which has implications for overall better morbidity and mortality in DM patients. The improvement in severity of vomiting makes these patients better candidates for renal and/or pancreas transplant because they could have reliable absorption of immunosuppressant medications. Also, we observed the normalization of GE in 23% of DM compared with 25% of ID and 25% of PS patients. Hence the symptom reduction in diabetics which exceeded that in idiopathic patients was achieved with similar GE changes. The majority of our patients (75%) continued to have similar degrees of delayed GE during the follow-up visits. Observations in symptom outcome are in agreement with previous small and short-term studies3,16 indicating that they
Table 5. Comparison of Use of Supplemental Nutrition at Baseline and During GES Follow-Up Diabetic and nutritional support (n ⫽ 48)
Idiopathic and nutritional support (n ⫽ 22)
Postsurgical and nutritional support (n ⫽ 11)
Nutrition type
Baseline
Follow-up
Baseline
Follow-up
Baseline
Follow-up
J-tube prior to GES TPN only TPN ⫹ J-tube G-J tubes
22 (46) 2 (4) 17 (35) 7 (15)
3 (6) 0 (0) 0 (0) 0
12 (54) 3 (14) 4 (18) 3 (14)
4 (18) 2 (9) 0 (0) 0
5 (46) 1 (9) 2 (18) 3 (27)
2 (18) 0 (0) 0 (0) 0
NOTE. Values are given as n (%). Eighteen gastroparetic patients received a new jejunostomy (J) tube at time of GES surgery. G-J tubes, gastrostomy-jejunal tubes.
318
MCCALLUM ET AL
were more consistent in the DM than the ID subgroup.3 In this current study, we provided a comparison of long-term symptom responses among 3 subgroups. Expectation of therapeutic response based on etiology is important for selecting patients to treat with GES. One explanation for the limited improvement in ID GP patients might be the fact that the idiopathic group consists of a relatively heterogeneous mix of patients and often with more concomitant abdominal pain complaints than seen in other subgroups with GP.3,20 It is well known that abdominal pain is the least likely to be improved by GES.20 In a long-term follow-up study, assessment of complications associated with GES system and overall safety of Enterra (Medtronic) therapy can be better appreciated. In this current study, 24 patients (11%) had their GES devices removed, which is similar to a 4-year follow-up study.10 Another major finding in this study is that symptom improvement with GES therapy is also associated with an increase in stimulation energy with a 6.0 mA median current. This is in agreement with the only 3 studies addressing any changes of stimulation parameters to maximize the symptom improvement with GES therapy.11-13 In 2 reports from Abell’s group11,12 the majority of patients treated with neurostimulation seemed to benefit from higher than baseline settings over time. In these studies, comparison of energy values in small GP subgroups with DM, ID, and PS showed that the postsurgical patients statistically required the most energy. In our current study we observed that the ID group had a greater voltage requirement both at baseline and at follow-up compared with the DM and PS groups, although the ID group had the worst symptom outcome. Hence, we conclude that energy parameters and increasing voltage do not predict or explain the likelihood of symptom reduction. Clinical studies with long-term follow-up often have the problem with missed visits or patients who become lost to follow-up, especially when contemplating the intention-totreat (ITT) analysis versus per protocol analysis in the final data set.21 In the current study, we used repeated measure analysis with the mixed model methodology.19 By fitting marginal models via generalized estimating equations, the mixed model gives an unbiased estimate for parameter estimate in the presence of missing values due to lost to followup. The model also adjusts for potential error deflation effect caused by positive intracluster correlation in response between baseline and follow-up.22 One limitation of this study is that it was not a placebocontrolled trial. Thus, the contribution of a placebo response during GES therapy could not be ruled out. However, the patients in our study had chronic symptoms that had not responded to numerous medications given over at least 1 and sometimes many years before receiving GES therapy and also requiring many admissions to address the complications of GP. Our patients over time had a significant reduction in hospitalization days as well as in use of medications, as we previously reported.23 In addition to increased weight there was essentially no reliance on J-tube and/or oral supplements. These outcomes are too profound to be considered a “placebo” effect; rather they pass the “test-of-time” challenge, where patients were able to be cared for by their local physicians, many in different states (42 states). Recently presented and published results of double-blind, randomized, multicenter, prospective trials of GES in DM and ID GP showed a significant improvement of TSS and quality of life
CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 9, No. 4
at 1 year of follow-up. However, when responses were compared during 3 months of on or off stimulation, there was no significant difference in GP symptom control. One explanation could be related to the study design involving 6 weeks of on stimulation which significantly decreased GP symptoms prior to the double-blind randomized phase. This raises the possibility of a sustained memory or carryover effect from these 6 “active” weeks and emphasizes the neuroplasticity of the central nervous system (CNS) nausea and vomiting control mechanisms.17 In summary, our data show that significant improvements in GP symptoms, nutritional support, days of hospitalization, and the use of medications with an accompanying good safety profile can be achieved and maintained for over 10 years through GES therapy in GP patients not responding to standard medical therapy. Enterra (Medtronic) GES therapy is more effective in treating diabetic and postsurgical gastroparesis than idiopathic GP patients.
Supplementary Material Note: To access the supplementary material accompanying this article, visit the online version of Clinical Gastroenterology and Hepatology at www.cghjournal.org, and at doi:10.1016/ j.cgh.2010.12.013. References 1. Parkman HP, Hasler WL, Fisher RS, et al. Technical review on the diagnosis and treatment of gastroparesis. Gastroenterology 2004; 124:1592–1622. 2. Jones MP, Maganti K. A systematic review of surgical therapy for gastroparesis. Am J Gastroenterol 2003;98:2122–2129. 3. Abell T, McCallum RW, Hocking M, et al. Gastric electrical stimulation for medically refractory gastroparesis. Gastroenterology 2003;125:421– 428. 4. Forster J, Sarosiek I, Delcore R, et al. Gastric pacing is a new surgical treatment for gastroparesis. Am J Surg 2001;182: 676 – 681. 5. Abell TL, Cutsem EV, Abrahamsson H, et al. Gastric electrical stimulation in intractable symptomatic gastroparesis. Digestion 2002;66:204 –212. 6. Familoni BO, Abell TL, Nemoto D, et al. Electrical stimulation at a frequency higher than basal rate in human stomach. Dig Dis Sci 1997;42:885– 891. 7. Lin ZY, Forster J, Sarosiek I, et al. Treatment of diabetic gastroparesis by high-frequency gastric electrical stimulation. Diabetes Care 2004;27:1071–1076. 8. McCallum RW, Lin ZY, Sarosiek I, et al. Clinical response to gastric electrical stimulation in patients with postsurgical gastroparesis. Clin Gastroenterol Hepatol 2005;3:49 –54. 9. Lin ZY, Sarosiek I, Forster J, et al. Symptom responses, long-term outcomes and adverse events beyond 3 years of high-frequency gastric electrical stimulation for gastroparesis. Neurogastroenterol Motil 2006;18:18 –27. 10. Anand C, Al-Juburi A, Familoni B, et al. Gastric electrical stimulation is safe and effective: a long-term study in patients with drug-refractory gastroparesis in three regional centers. Digestion 2007;75:83– 89. 11. Cutts TF, Luo J, Starkebaum W, et al. Is gastric electrical stimulation superior to standard pharmacologic therapy in improving GI symptoms, healthcare resources, and long-term health care benefits? Neurogastroenterol Motil 2005;17:35– 43. 12. Abidi N, Starkebaum WL, Abell TL. An energy algorithm improves symptoms in some patients with gastroparesis and treated with
April 2011
13.
14.
15. 16.
17.
18.
19. 20.
21.
22. 23.
gastric electrical stimulation. Neurogastroenterol Motil 2006; 18:334 –338. Mason RJ, Lipham J, Eckerling G, et al. Gastric electrical stimulation: an alternative surgical therapy for patients with gastroparesis. Arch Surg 2005;140:841– 848. McKenna D, Beverstein G, Reichelderfer M, et al. Gastric electrical stimulation is an effective and safe treatment for medically refractory gastroparesis. Surgery 2008;144:566 –574. Brody F, Vaziri K, Saddier A, et al. Gastric electrical stimulation for gastroparesis. J Am Coll Surg 2008;207:533–538. Maranki JL, Lytes V, Meilahn JE, et al. Predictive factors for clinical improvement with Enterra gastric electric stimulation treatment for refractory gastroparesis. Dig Dis Sci 2008;53: 2072–2078. McCallum R, Snape W, Brody F, et al. Gastric electrical stimulation with Enterra Therapy improves symptoms from diabetic gastroparesis in prospective study. Clin Gastroenterol Hepatol 2010;8:847–954, e1. Tougas G, Eaker EY, Abell TL, et al. Assessment of gastric emptying using a low fat meal: establishment of international control values. Am J Gastroenterol 2000;95:1456 –1462. Little RC, Milliken GA, Stroup WW, et al. SAS system for mixed models. Cary, NC: SAS Institute, 1996. McCallum RW, Dusing RW, Sarosiek I, et al. Mechanisms of symptomatic improvement after gastric electrical stimulation in gastroparetic patients. Neurogastroenterol Motil 2010;22:161– 167. Schoenfeld PS. Evidence-based medicine in practice: applying intention-to-treat analysis and per-protocol analysis. Am J Gastroenterol 2005;100:3– 4. Finney D. Repeated measurements: what is measured and what repeats? Stat Med 1990;9:639 – 644. Lin Z, McElhinney C, Sarosiek I, et al. Chronic gastric electrical stimulation for gastroparesis reduces the use of prokinetic and/or antiemetic medications and the need for hospitalizations. Dig Dis Sci 2005;50:1328 –1334.
A 10-YEAR GES STUDY FOR GASTROPARESIS
319
Reprint requests Address requests for reprints to: Richard W. McCallum, MD, Chairman, Department of Internal Medicine, Texas Tech University Health Sciences Center, Paul L. Foster School of Medicine, 4800 Alberta Avenue, El Paso, Texas 79905. e-mail: [email protected]; fax: (915) 545-6634. Acknowledgments Richard W. McCallum and Irene Sarosiek contributed equally to the work. Medtronic research and technical personnel were not involved in data collection, interpretation of the results, writing, or statistical analysis of this manuscript. The authors thank the Medtronic GI Group, specifically representatives Bjorn Barner and Junius Hill, for their technical support, and Warren Starkebaum for scientific advice over the years. The authors acknowledge the following individuals for their contributions: Pernilla Foran, LPN; Teri Lavenbarg, RN; and Irene King in the Motility Center; and also faculty, fellows, and nursing staff (particularly Sandra Sostarich) in the GI Division and the OR staff, as well as surgical residents for their patient-care contributions. The authors acknowledge Dr Robert Twillman in Psychology and Pain Management, and Drs Reginald Dusing, James Traylor, and Christine McMillin in the Nuclear Medicine Department at KUMC. Conflicts of interest The authors disclose the following: Over the past 10 years, Richard W. McCallum and Irene Sarosiek have occasionally participated in teaching and consulting activities sponsored by Medtronic, Inc. The remaining authors disclose no conflicts. Funding This work was partially financially supported by Medtronic, Inc, Minneapolis, MN, as some of our patients participated initially in 2 double-blind trials.
April 2011
A 10-YEAR GES STUDY FOR GASTROPARESIS
319.e1
Supplementary Table 1. Comparison of GES Parameters Among the 3 Groups of GP Patients
Impedance (⍀) Baseline Follow-up Voltage (V) Baseline Follow-up Current (mA) Baseline Follow-up
Diabetic gastroparetic patients (n ⫽ 118)
Idiopathic gastroparetic patients (n ⫽ 39)
Postsurgical gastroparesis (n ⫽ 24)
480 (390–534) 579 (494–655)a,b
524 (429–579) 617 (511–724)a
465 (379–531) 608 (521–716)a
2.10 (2.00–2.60)c 3.75 (2.70–5.00)a
2.60 (2.00–2.90) 4.15 (3.10–5.00)a
2.20 (2.00–2.70) 4.00 (2.90–5.00)a
5.0 6.0 (3.0–16.0)
5.0 6.0 (3.0–13.0)
5.0 6.0 (3.0–18.0)
NOTE. Impedance and voltage are reported as median and interquartile range. aP ⬍ .001 compared with respective baseline. bP ⬍ .05 compared with idiopathic group at follow-up. cP ⬍ .05 compared with PS ID patients at respective baseline.