SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. References

Background:

Gastric acid secretion is important for absorption of dietary non-haem iron, and iron deficiency is common in gastric hyposecretory states such as after gastric resection. It is not known if prolonged, continuous treatment with potent acid suppressants such as omeprazole will lead to iron deficiency or lower body iron stores.

Aim:

To assess iron stores and the occurrence of iron deficiency anaemia in patients with Zollinger–Ellison syndrome (ZES) treated long-term with gastric antisecretory drugs.

Methods:

One hundred and nine patients with ZES but without previous gastric resections were studied. All patients underwent assessment of acid control on antisecretory agents, determination of tumour extent, evaluation of haematological parameters (Hct, haemoglobin, WBC, MCV, MCHC), and determination of serum iron parameters (iron, ferritin, transferrin, iron/transferrin ratio). Acid control values for the last 4 years were reviewed, the presence or absence of acid hyposecretion determined using three different criteria and this result correlated with haematological and iron parameters.

Results:

Eighty-nine patients were taking omeprazole, nine patients were taking histamine H2-antagonists and 11 patients were taking no drugs following curative resection. The mean duration of omeprazole treatment was 5.7 years (range 0.7–12.5 years) and total duration of any treatment was 10.1 years (range 0.7–21 years). Acid hyposecretion was present by at least one criterion in 45% of patients. There were no significant differences between patients with or without acid hyposecretion, taking or not taking omeprazole, having different durations of omeprazole treatment or different durations of total time receiving any antisecretory treatment, for any serum iron parameter, haematological parameter, or for the frequency of iron deficiency. Males and females did not differ in percentage with low ferritin levels or percentage with iron deficiency.

Conclusions:

Continuous treatment with omeprazole for 6 years or continuous treatment with any gastric antisecretory drug for 10 years does not cause decreased body iron stores or iron deficiency. These results suggest that, in contrast to recent results which show yearly monitoring of vitamin B12 in such patients is needed, such monitoring for iron parameters is not necessary.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. References

The chronic use of gastric acid suppressing agents has become more common due to the high prevalence of gastro-oesophageal reflux disease (GERD) in the general population and the recent approval of H+,K+-ATPase inhibitors for maintenance therapy. The possible need for long-term continuous treatment with potent gastric antisecretory agents such as H+,K+-ATPase inhibitors has led increasingly to considerations of their long-term safety.1, 2 H+,K+-ATPase inhibitors such as omeprazole can decrease gastric hydrogen ion concentration by 95–99% in doses of 30–40 mg in man.3, 4 In recent studies of patients treated long-term with omeprazole, long-term acid suppression is sufficiently profound that 80–100% of patients develop hypergastrinaemia.5, 6 The primary safety concerns that have received most attention are the possible long-term effects of the hypergastrinaemia produced secondary to the marked acid suppression. The ability of the hypergastrinaemia to cause proliferation of gastric ECL cells resulting in gastric carcinoid tumours has been extensively studied.7, 8 However, much less is known about the possible long-term metabolic effects of this degree of drug-induced hypochlorhydria itself.

The absorption of numerous nutrients, including vitamin B12, iron and calcium, are affected by gastric acidity, and the possibility that long-term, drug-induced hypochlorhydria could have nutritional side-effects has been raised.9[10][11][12][13][14]–15 Recent studies report that after 4 years of omeprazole treatment in patients with idiopathic peptic disease or oesophageal reflux disease, a decrease in serum vitamin B12 levels was seen,13 and in patients with Zollinger–Ellison syndrome (ZES) requiring continuous omeprazole treatment, decreases in serum vitamin B12 levels correlated closely with the extent of omeprazole-induced acid inhibition.16 In contrast to vitamin B12, little is known about the possible effects of chronic hypochlorhydria due to prolonged treatment with potent gastric acid antisecretory drugs on iron or calcium absorption.13 Dietary iron is present in food as either non-haem iron or haem iron (in meats largely as myoglobin or haemoglobin).17[18][19][20]–21 Dietary non-haem iron absorption is markedly improved in the presence of gastric acid.22[23][24]–25 A number of clinical associations suggest that gastric acid hyposecretion, especially over a prolonged period, may result in clinically significant malabsorption of iron. Decreased non-haem iron absorption has been reported in patients with achlorhydria,14, 26, 27 with gastric hyposecretion due to cimetidine treatment,12 due to omeprazole in rats on an iron-deficient diet,28 and is reported with various gastric acid-hyposecretory conditions (gastric resection, vagotomy, atrophic gastritis).13, 23, 25, 27, 29[30][31][32]–33

At present the information available on the possible long-term effect of omeprazole on iron absorption is very limited, because only recently has omeprazole been approved for long-term use in reflux oesophagitis. Therefore at present only small numbers of these patients have been treated long-term with omeprazole. In contrast, most patients with ZES (>70%) are not cured after gastrinoma resection so that daily, life-long treatment of the gastric acid hypersecretion is required.34 The cohort of patients with ZES followed long-term at the NIH is particularly well suited for assessment of the possible long-term effects of omeprazole on iron absorption for a number of reasons. First, none of these patients had undergone gastric resections, which have been shown (in patients without gastrinomas) to cause malabsorption of iron30, 31, 35, 36 and could confound the interpretation of the effect of omeprazole. Second, the mean length of treatment with omeprazole in these patients was long (mean 5.7 years) and the mean duration of total gastric antisecretory drug therapy with any antisecretory drug was even longer (mean 9.5 years). Therefore, if it takes a significant period of time to develop decreased iron body stores it should be possible to detect it. Third, yearly gastric acid secretory studies on antisecretory drugs were carried out on all patients so that the level of acid secretion on gastric antisecretory medications is known; those patients who were rendered achlorhydric can be identified, so that the effects on iron parameters in patients with different levels of acid suppression can be compared. Lastly, some patients are treated only with histamine H2-receptor antagonists, which cause less marked acid inhibition;34, 37 the results of less inhibition of acid suppression on iron parameters in these patients can be compared to those treated with omeprazole, which causes more profound acid inhibition.34

METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. References

One hundred and twenty-eight consecutive patients with ZES, who were admitted to the NIH from January 1995 to December 1995, were considered for the present study. The diagnosis of ZES was established as described previously.38 Each patient, prior to this study, underwent an evaluation to confirm the diagnosis; an evaluation of tumour location and extent; and an assessment to determine the presence or absence of Multiple Endocrine Neoplasia type I (MEN-I), as described previously.39 Tumour extent was determined by standard imaging studies.40[41]–42 The diagnostic criteria for the presence of MEN-I in a patient with ZES were as previously described.39

Patients had their acid hypersecretion controlled medically, up to 1984 with histamine H2-receptor antagonists and from 1984 to the present with either H+,K+-ATPase inhibitors (omeprazole or lansoprazole) or histamine H2-receptor antagonists (ranitidine). Acid control was established by reducing acid secretion to ≤ 10 mEq/h for the hour before the next dose of antisecretory drug.34 For patients with moderate to severe oesophageal disease and patients who had previous gastric acid-reducing surgery, acid secretion was reduced to < 5 mEq/h prior to the next dose of antisecretory drug.34 Basal acid output and maximal acid output were determined for each patient.34 All patients who did not have evidence of metastatic liver disease or MEN-I underwent exploratory laparotomy,41 whether or not imaging studies disclosed a tumour. At laparotomy a duodenotomy was carried out and, in addition, intraoperative oesophagogastroduodenoscopy and ultrasound were performed.41 Each patient was reassessed after surgery for possible cure.38

Specific protocol

One hundred and twenty-eight consecutive patients with ZES were considered for the study. Patients were eligible for the study if they had not undergone total or partial gastrectomy or if they were not new admissions without any follow-up data. Nineteen patients were excluded either because this was their initial admission to the NIH (n=11) or because they had previously undergone total gastrectomy (n=1) or partial gastrectomy (n=7).

Each patient was evaluated at least within the last year. Evaluation included an assessment of gastric acid secretion (control of acid hypersecretion by current antisecretory drug). Tumour extent and size was assessed with standard imaging studies and somatostatin receptor scintigraphy using (111In-DTPA,DPhe1)octreotide. Each patient had a history taken prior to the admission for evaluation and was asked to stop taking all multivitamin preparations or oral iron supplements 1 week prior to admission. A history of other medications was taken, of any recent bleeding, infections, arthritic conditions and, in premenopausal females, a history of menstrual bleeding was taken. Each patient underwent determinations of fasting serum gastrin level; serum iron studies including a serum iron determination (normal 50–150 μg/dL), serum ferritin level (normal 10–125 μg/L for females and 10–300 μg/L for males), serum transferrin levels (normal 230–390 mg/dL), serum iron/transferrin ratio (normal 15–62%); peripheral blood haematological studies including haematocrit (normal 31.8–43.3% for females and 36.7–48.3% for males), haemoglobin (normal 11.1–15.0 g/dL for females and 12.7–16.7 g/dL for males), white blood cell count, mean corpuscular volume (normal 77–99 fL for females and 79–98 fL for males) and mean haemoglobin concentration. Iron deficiency was considered to be present if the serum ferritin level was < 10 μg/L, because a number of studies show that a low serum ferritin (< 10 μg/L) alone is a sensitive indicator of depleted body iron stores.43[44][45][46][47]–48 In other studies < 12 μg/L or < 15 μg/L are used as the criteria and they were also used for analysis.43[44]–45, 48, 49 The acid secretory control value data from each of the previous four annual admissions (n=97) was reviewed, as well as the gastric antisecretory drugs and dosages. For the 12 patients followed for < 4 years the acid control values for their previous admissions were used to determine the acid secretory control category. Gastric acid hyposecretion was defined in three different ways, based on three different criteria of the acid control value obtained 1 h prior to the administration of the next dose of oral drug. The three criteria of gastric acid hyposecretion while taking antisecretory drugs were: acid output < 0.2 mEq/h for 2 out of the last 4 years; acid output < 0.2 mEq/h or < 1 mEq/h on the most recent acid control study. The duration of total gastric acid antisecretory treatment with any drug was defined as the total time of continuous treatment with either an H+,K+-ATPase inhibitor or a histamine H2-receptor antagonist.

The serum iron studies were performed on peripheral blood. The serum iron was determined using the Reagen Agent from Abbott Laboratories (Abbott Park, IL). It was processed by a Roche Corbas Fara centripetal analyser (Roche Diagnostics, Belleville, NJ (catalogue #2910–01)). The coefficient of variation was 1.4% intra-assay and 2.8% inter-assay. Serum transferrin was assayed using the Incstar SPQ transferrin test system (Incstar Corporation, Stillwell Water, MN (catalogue #86057)). The processing was done by a Roche Corbas Fara centripetal analyser (Roche Diagnostics, Belleville, NJ (catalogue #2910–01)). The coefficient of variation was 2.7% intra-assay and 4.0% inter-assay. Serum ferritin values before October 1995 were analysed using the Abbott Axsym system (Abbott Park, IL) with the Ferritin reagent (catalogue #7A5820). The method of analysis was the microparticle enzyme immunoassay. The coefficient of variation was 3.1% intra-assay and 3.9% inter-assay. From October 1995 to the end of the study, serum ferritin analysis was performed with the Access Immunoassay system using the Ferritin reagent (Sanofi, Chaska, MN (catalogue #33020)). The method of analysis was chemiluminesence. The coefficient of variation for inter-assay was 5.8%.

Statistical analysis

For statistical analysis the chi-squared test, Fisher’s exact test, analysis of variance, and the Mann–Whitney U-test were performed to compare variables in men and women, in patients with or without acid secretion < 0.2 mEq/h for 2 of the last 4 years or to compare results for patients treated and not treated with omeprazole. The Kruskal–Wallis test was used for comparing more than two groups of serum ferritin values. Differences were considered significant when P was < 0.05.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. References

One hundred and nine patients with ZES were included in this study. The majority (57%) were male and the mean age was in the sixth decade ( Table 1). The disease was long-standing in most patients, with a mean duration of disease from onset of 12.8 years and a mean duration of diagnosis of 8.9 years. The mean fasting serum gastrin levels, percentage with MEN-I syndrome, basal acid output and maximal acid output were similar to those in other large series of patients with ZES.37 Twenty-four per cent of patients had histologically proven metastatic disease to the liver and the remainder had no liver metastases proven at surgery ( Table 1). Most patients had a duodenal gastrinoma (40%) or pancreatic tumour (31%), with a lesser number (21%) having a primary tumour in other sites (primarily lymph nodes), and in 20 patients (18%) the primary site was unknown. In terms of tumour extent and location, this population of patients resembles those in other large series of patients with gastrinomas.37

Table 1.  . Clinical and laboratory characteristics of patients studied Thumbnail image of

All patients had been treated with a gastric antisecretory agent during the course of their disease. The mean duration of treatment with any gastric antisecretory drug was 10.1 years ( Table 2). At the time of this study 89 patients were being treated with omeprazole, nine patients with histamine H2-receptor antagonists and 11 patients were not being treated with any gastric acid antisecretory agent because they were cured after resection of a gastrinoma.38 The mean duration of treatment for patients who were receiving omeprazole and histamine H2-receptor antagonists were 5.7 and 5.0 years, respectively, with the longest duration of treatment with omeprazole being 12.5 years and with histamine H2-receptor antagonists 18.9 years. Forty-five per cent of the patients (49/109) met at least one of the three criteria for gastric acid hyposecretion while taking gastric antisecretory drugs ( Table 2). Twenty-three per cent of patients (25/109) met the most stringent criterion, which was acid secretion while taking antisecretory agents of < 0.2 mEq/h for 2 out of the last 4 years ( Figure 1, Table 3) and 43% (47/109) met the least stringent criterion of acid secretion < 1.0 mEq/h during the last year. There were no significant differences for patients with or without acid hyposecretion using the most stringent level of < 0.2 mEq/h for 2 of the last 4 years in the serum ferritin levels (geometric mean, 43.4 and 40, or arithmetic mean, 68 ± 11 and 73 ± 9.3 μg/L, respectively), serum iron levels (84 ± 8 and 86 ± 5 μg/L, respectively), serum transferrin levels (292 ± 11 and 289 ± 6 mg/dL, respectively), and the iron/transferrin ratio (24.5 ± 2 and 25 ± 1.5%, respectively) ( Figure 1). When a similar analysis was performed using the acid hyposecretory criteria of < 0.2 mEq/h in the last year, which was seen in 33/109 patients (30%) or < 1 mEq/h during the last year, which occurred in 47/109 patients (43%), there were also no significant differences in the mean iron parameters in patients with or without hyposecretion by these criteria ( Figure 2). Furthermore, there was no difference in the percentage of patients with or without each criterion of acid hyposecretion who had iron deficiency ( Figure 2, Table 3). Specifically, in patients meeting the most stringent criterion of acid hyposecretion of < 0.2 mEq/h for 2 of the last 4 years, three of the 25 patients (12%) were iron deficient, which was not significantly different from the 13% seen in patients without acid hyposecreti on by this criteria. Similarly, using serum ferritin values of < 12 μg/L or < 15 μg/L for determining iron deficiency, there was no significant difference between patients with and without acid hyposecretion. The percentage of patients with and without acid hyposecretion did not differ significantly (P=0.41) in patients with serum ferritin levels>  100 μg/L, in whom iron deficiency is rare if no inflammatory disorder is present50 ( Table 3). In patients with or without hyposecretion by the < 0.2 mEq/h or < 1 mEq/h for the last year criterion, the percentage of patients with or without iron deficiency did not differ significantly (18% vs. 11%, P=0.29 and 14% vs. 12%, respectively) ( Figure 2, Table 3). Furthermore, the haematocrit and mean corpuscular volume (MCV) of patients with gastric acid hyposecretion by the most stringent criterion ( Table 3), or by the other criteria (data not shown), did not differ from patients who did not have gastric acid hyposecretion by these criteria ( Table 3).

Table 2.  . Type and duration of gastric antisecretory drug treatment and the percentage of patients with gastric acid hyposecretion Thumbnail image of
image

Figure 1. Effect of gastric acid hyposecretion on serum iron parameters. Gastric acid hyposecretion (acid output < 0.2 mEq/h for 2 out of the last 4 years) was present in 25/109 of the patients (23%) and absent in 84/109 (77%) of the patients. Each dot represents the results obtained from each patient. The line at the bottom indicates the lower limit of normal for each laboratory value. The dark horizontal line represents the mean and the thin vertical line represents the S.E.M. The serum ferritin data are plotted on a logarithmic scale because a previous study46 demonstrates its distribution in a normal population is geometric. The mean for serum ferritin is a geometric mean. All other parameters are plotted arithmetically with arithmetic means. Fourteen patients had low levels of serum ferritin (< 10 μg/L), 24 patients had low levels of serum iron (< 50 μg/dL), 11 patients had low transferrin levels (< 230 mg/dL) and 22 patients had a low iron/transferrin ratio (< 1. 5%).

Download figure to PowerPoint

Table 3.  . The effect of gastric acid hyposecretion and treatment with omeprazole on haematologic parameters, serum iron values and the presence or absence of iron deficiency Thumbnail image of
image

Figure 2. Effect of gastric acid hyposecretion, defined by different criteria, on serum iron parameters and the development of iron deficiency. The three different criteria for gastric acid hyposecretion were: acid output < 0.2 mEq/h for 2 of the last 4 years, < 0.2 mEq/h in the last year and less than 1 mEq/h in the last year. The dotted line represents the lower limit of normal of all values. Shown are the geometric means for the serum ferritin data and the arithmetic mean ± S.E.M. for the other parameters. For the criteria of < 0.2 mEq/h for 2 of the last 4 years, < 0.2 mEq/h or < 1.0 mEq/h during the last year, 25/109 (23%), 33/109 (30%) and 47/109 (43%) of the patients, respectively, had gastric acid hyposecretion, and 84/109 (77%), 76/109 (70%), 62. /109 (57%) of the patients, respectively, did not have the criterion present. Iron deficiency was defined as serum ferritin < 10 μg/L. There were no significant differences in either the iron values or in the percentage of patients who developed iron deficiency among the patients with or without gastric acid hyposecretion in the three categories.

Download figure to PowerPoint

For all patients, treatment with omeprazole did not have a significant effect on serum ferritin levels (geometric mean 37.7 with, and 58 μg/L without omeprazole, P=0.08) ( Figure 3, Table 3). When the duration of omeprazole treatment was considered, increasing duration of omeprazole treatment did not correlate with greater decreases in serum ferritin levels ( Figure 3). In fact, only the patients treated for < 5 years with omeprazole had serum ferritin levels significantly lower than patients not treated with omeprazole (<  0.03, Figure 3). In contrast, the serum ferritin levels in patients treated with omeprazole for more than 5 years were not significantly different from patients without omeprazole treatment ( Figure 3). The presence or absence of omeprazole therapy and the duration of omeprazole treatment (i.e. ≤ 5 years or>  5 years) had no significant effect on serum iron levels ( Table 3, Figure 4, top right panel), serum transferrin levels ( Table 3, Figure 4, bottom left panel), the iron-transferrin ratio ( Table 3), the MCV ( Table 3) or the haematocrit ( Table 3). Furthermore, the use of omeprazole or its duration had no significant effect on the rate of occurrence of iron deficiency ( Table 3, Figure 4, bottom right panel). Specifically, iron deficiency was present in 10% of patients not taking omeprazole, in 13% taking omeprazole for any duration, in 12% taking omeprazole for < 5 years and in 15% taking omeprazole for>  5 years, all of which were not significantly different ( Figure 4). Similarly, using serum ferritin criteria of < 12 μg/L or < 15 μg/L as a measure of iron deficiency did not change the lack of effect of omeprazole on the occurrence of iron deficiency ( Table 3). Furthermore, there was no significant difference (P=0.29) in the proportion of patients without iron deficiency with serum ferritin>  100 μg/L50 taking or not taking omeprazole ( Table 3).

image

Figure 3. Effect of the use of omeprazole and the duration of treatment with omeprazole on serum ferritin levels. Patients were divided into one of four groups depending on whether omeprazole was taken and its duration of use. Twenty patients did not receive omeprazole and 89 patients were treated with omeprazole. Thirty-nine patients were treated with omeprazole for 5 years or less and 50 patients were treated for more than 5 years. The serum ferritin value of each patient is shown by a dot. The horizontal line is the geometric mean for the patients in the indicated groups. The geometric mean values for the no omeprazole group, any omeprazole treatment, omeprazole treatment ≤ 5 years or> 5 years were 59, 38, 28 and 50. Only the ferritin value for patients treated < 5 years was significantly different from patients not treated with omeprazole (P < 0.03. ).

Download figure to PowerPoint

image

Figure 4. . Effect of treatment with omeprazole and the duration of omeprazole treatment on serum iron parameters and the development of iron deficiency. The number of patients with no omeprazole treatment, any omeprazole treatment, and the various durations of treatment were the same as in the Figure 3 legend. The values shown are the geometric mean for serum ferritin and the arithmetic mean ± S.E.M. for the other parameters. Iron deficiency was defined as serum ferritin < 10 μg/L. Iron deficiency was present in 10% of patients with no omeprazole therapy, 13% with any omeprazole therapy, 12% taking omeprazole ≤ 5 years, and 15% taking omeprazole>  5 years. The dotted line represents the lower limit of normal for the parameter listed.

Download figure to PowerPoint

Eighty-seven per cent of the patients (87/100) taking omeprazole or currently taking no antisecretory drug had their acid hypersecretion controlled with histamine H2-receptor antagonists prior to starting omeprazole or undergoing curative resection. The mean duration of such treatment was 4.5 ± 0.4 years. During this time sufficient histamine H2-antagonist was used such that acid secretion was reduced in all patients to < 10 mEq/h34 if no previous gastric surgery and < 5 mEq/h or greater in patients with reflux disease or previous gastric acid-reducing surgery. The total duration of treatment with either histamine H2-receptor antagonist or omeprazole was 10.1 ± 0.4 years (mean ± 1 S.E.M.), with some patients being treated for up to 21 years ( Table 3). To determine whether this prolonged period of antisecretory treatment affected iron parameters, a similar analysis was performed ( Figures 5 and 6 ) to that with omeprazole alone. The duration of total antisecretory treatment had no effect on serum ferritin levels ( Figure 5). Specifically, the geometric means of the serum ferritin levels were 34.4 μg/L for treatment < 5 years, 33.3 μg/L for 5–9.9 years, 47.9 μg/L for 10–14.5 years, and 51.6 μg/L for>  15 years, all of which were not significantly different (P=0.34). Similarly, when the serum iron levels ( Figure 6, top right), serum transferrin levels ( Figure 6, bottom left), or the iron/transferrin ratios (data not shown) were compared for these four different durations of total treatment, there were no significant differences between any of the time periods for any serum iron parameters. There was also no difference in the percentage of patients having iron deficiency with any of the different total treatment periods ( Figure 6, bottom right). Specifically, 11% of the 18 patients treated for less than 5 years of total treatment time had iron deficiency which was not significantly different from the 15% of the 32 patients treated for 5–9.9 years, the 10% of the 39 patients treated for 10–14.9 years and the 17% of the 18 patients treated for longer than 15 years ( Figure 6).

image

Figure 5. Effect of total duration of treatment with any gastric secretory drug on serum ferritin levels in individual patients. Each dot represents the result from a single patient. The horizontal line at the bottom is the lower limit of normal. The small horizontal line is the geometric mean for each group of patients. Eighteen patients had taken any gastric antisecretory drug for < 5 years, 34 patients for 5–9.9 years, 39 patients for 10–14 years, and 18 patients for> 18 years. The geometric means of the serum ferritin values for the different groups were: < 5 years, 34.4 μg/L; 5–9.9 years, 33.3 μg/L; 10–14.9 years, 47.9 μg/L; and ≥ 15 years, 5. 1.6 μg/L, none of which were significantly different (P=0.34, Kruskal–Wallis test).

Download figure to PowerPoint

image

Figure 6. . Effect of total duration of treatment with any gastric antisecretory drug on serum iron parameters and the development of iron deficiency. The duration of treatment with a gastric antisecretory drug was the total duration of treatment with either a histamine H2-receptor antagonist or with a H+,K+-ATPase inhibitor (omeprazole or lansoprazole) and was the same as outlined in the Figure 5 legend. The < 5 year group’s ferritin, serum iron/transferrin ratios, or percentage with iron deficiency did not differ significantly from the other time periods. Two of 18 patients (11%) treated < 5 years had iron deficiency, 5/34 patients (15%) treated for 5–9.9 years, 4/39 patients (10%) treated 10–14.9 years, and 3/18 patients (17%) treated for>  15 years. The geometric mean of the serum ferritin data and the arithmetic mean ± S.E.M. for the other parameters is shown. Iron deficiency was defined as a serum ferritin < 10 μg/L.

Download figure to PowerPoint

In some gastric acid hyposecretory states, such as after partial gastric resection or gastroenterostomy, the probability of developing iron deficiency is higher in women than men.30, 32, 33, 51 To determine whether a similar trend might be found in our study we compared the results from men and women ( Table 4). There were 62 men and 47 women in the study. The women differed from the men in being significantly younger (<  0.008), there was a higher proportion of women ≤ 45 years of age (<  0.001), women had a significantly shorter total antisecretory drug treatment duration (9.3 vs. 10.7 years, P=0.03), and the proportion treated with omeprazole for longer than 5 years was significantly higher (P=0.03) ( Table 4). In contrast, men and women did not differ in BAO, MAO, percentage treated with omeprazole, histamine H2-receptor antagonist or receiving no antisecretory treatment or the total duration of omeprazole treatment ( Table 4). Women had significantly lower mean serum ferritin, serum iron levels, iron/transferrin ratios, and a higher significant percentage with low serum iron or iron/transferrin ratios, but not serum ferritin levels ( Table 4). The percentage with iron deficiency did not differ between men and women (6%, 95% CI: 3–17% vs. 17%, 95% CI: 8–31%, > 0.27) and the percentage meeting any of the three criteria of acid hyposecretion did not differ between men and women ( Table 4). As expected, men had a significantly higher mean haematocrit, however, the mean MCVs did not differ.

Table 4.  . Comparison of clinical characteristics, serum iron study results and haematologic parameters in males and females Thumbnail image of

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. References

The purpose of this study was to determine whether marked suppression of gastric acid secretion with oral gastric acid antisecretory drugs (omeprazole, histamine H2-receptor antagonists) for a prolonged period of time caused effects on serum iron parameters, body iron stores, or caused iron deficiency anaemia. This study was undertaken because at present it is unclear whether prolonged treatment with such potent gastric antisecretory drugs will cause a clinically significant malabsorption of iron.10, 13

Numerous studies provide strong support that gastric acid secretion is important for optimal absorption of non-haem iron. Approximately two-thirds of dietary iron is food-bound, primarily as ferric salts (non-haem iron), and the remaining one-third is derived from meat as muscle myoglobin or blood haemoglobin.17 Gastric acid secretory rate does not affect the absorption of haem iron.17 The mechanism of the enhancing effect on non-haem iron absorption has been attributed to the ability of gastric acid to dissociate the iron salts in the food, and to help solubilize the iron salts which allows them to be reduced to the ferrous form and which allows the formation of complexes with ascorbate, sugars and amines.15, 17, 22[23][24]–25, 27, 52, 53 Ferric ions are insoluble in aqueous solutions above pH 3, whereas the ferrous form is 100 times more soluble at intestinal pHs of 5–7.17 Clinical studies as early as 1932 suggested that gastric acidity might have a role in the absorption of iron, because the response to oral iron supplements in patients with iron deficiency was greater in patients with normal gastric acidity than those with achlorhydria.54 Subsequent studies showed that decreased non-haem iron absorption can occur in achlorhydric patients and that the addition of acid can increase the absorption.14, 23, 25[26]–27, 55, 56 A number of clinical conditions that cause chronic hypochlorhydria or achlorhydria have been shown to cause anaemia, decreased iron absorption, or result in iron deficiency including vagotomy, atrophic gastritis and gastric resections.13, 23, 29[30]–31, 35, 55[56][57][58][59][60]–61 Patients after gastric resection or who are achlorhydric have been shown to have a reduced ability to absorb iron salts and non-haem iron mixed with food at neutral pH levels, yet they have a high capacity to absorb iron salts when it is ingested in an acidic solution.15, 27, 31, 35 However, very limited data are available on gastric antisecretory drugs. Cimetidine caused an inhibition of iron absorption following single doses of 300–900 mg,12 but this appears to be without clinical significance because iron deficiency is not a well-recognized complication of long-term cimetidine therapy.13, 28 Omeprazole, in a study in rats,28 decreased iron absorption in animals on an iron-deficient diet but had no effect on absorption in ani132;mals on a normal diet. In one recent study of 34 patients with peptic diseases (primarily reflux oesophagitis) treated for up to 48 months, two patients had low serum iron levels and three had low ferritin levels; however, repeated measurements did not show statistically significant changes.10 Whereas the above studies establish that acid hyposecretion can affect iron absorption as well as various gastric operations and diseases such as atrophic gastritis, these results may not be generalizable to long-term gastric acid antisecretory therapy. These conditions may not be analogous to treatment with potent acid suppressants such as H+,K+-ATPase inhibitors because the degree of acid suppression may differ, may be more intermittent, or other abnormalities of gastric secretion or absorption may occur in these conditions that do not occur with long-term H+,K+-ATPase inhibitor treatment. This question is becoming increasingly important to address with the recent approval by the FDA of long-term maintenance therapy with omeprazole for oesophageal reflux disease. At present few patients with chronic oesophageal disease have been treated long-term with H+,K+-ATPase inhibitors, so it is difficult to perform the study in sufficient numbers. In contrast, a number of groups of patients with ZES have been treated on a daily basis with omeprazole for more than 5–10 years.34, 37 The group of patients followed at the NIH is particularly useful for such a study because not only have they not routinely undergone gastric resections (which can alter iron absorption25, 55, 56), but these patients have undergone yearly routine assessment of gastric acid secretory rates on antisecretory drugs, allowing correlations between the degree of acid suppression and the results on iron parameters.

Using this population of patients, we determined the effect of prolonged acid suppression with either histamine H2-receptor antagonists, omeprazole, or both in 109 consecutive patients with ZES. Our results showed no increased rate of iron deficiency in any group of patients examined. Specifically, no evidence of alterations in clinically significant iron absorption could be detected in patients treated or not treated with omeprazole, even when stratified into patients who had the absence or presence of gastric acid hyposecretion defined by three different criteria. Furthermore, no evidence of an increased rate of iron deficiency for women compared to men was found, which has been reported in patients followed after gastric resection.30, 32, 33, 51 A number of possibilities could be proposed that might account for these results in light of the numerous studies which suggest acid secretion can affect iron absorption. These would include the possibility that the patients were not treated with gastric antisecretory drugs for a sufficient time; that the inhibition of gastric acid secretion was not sufficiently profound, perhaps because the patients had ZES; that the patients’ dietary intake of haem-iron might compensate for the non-haem iron malabsorption, or that hypochlorhydria alone does not cause iron deficiency. It is unlikely that a failure to see an effect was due to the fact that the patients were not treated with gastric antisecretory agents for long enough. The mean duration of omeprazole treatment in the present study was 6 years, the duration of total antisecretory treatment 10 years, and some patients were treated for up to 12 years with omeprazole and 21 years of total treatment with either omeprazole or a histamine H2-receptor antagonist. Furthermore, stratification of the patients for time of treatment either with omeprazole or any antisecretory drug showed no correlation with the results of the iron studies. On average, after gastric resection, 30% of patients develop iron deficiency;35, 36 however, the percentage varies widely in different series from 9 to 60% with a mean follow-up time of 7 years, varying from 3 to 15 years in various studies.29, 35, 51, 57[58][59][60]–61. In two studies evaluating patients at different times after resection, the highest rate of iron deficiency was seen at 4 years in one study57 and in a second study33 the maximal rate of iron deficiency was reached in < 5 years. In some studies the percentage increased with the time after the resection,35, 51 with up to 90% of females less than 50 years of age 4–5 years partial gastrectomy developing iron deficiency anaemia.51 These results would suggest that if one of the primary mechanisms of iron deficiency after gastric resection is hypochlorhydria, then the follow-up time in the present study was sufficiently long to have seen iron deficiency in a significant percentage of patients, especially high-risk groups such as females ≤ 45 years of age.35

Serum ferritin has been shown to be a sensitive indicator of body iron stores;43[44][45][46][47][48][49]–50 however, other factors could obscure the assessment of the effect of drug-induced hypochlorhydria on iron stores. These include increasing the ferritin because it is an acute phase reactant (infection, inflammatory process, chronic renal disease, malignancies) or due to taking oral iron prior to the study, or decreasing it secondary to blood loss (recent GI bleeding, menstrual bleeding, or bleeding from other sites). No patients in this study had an active infection, another malignancy except for the gastrinoma, or other inflammatory process. None were taking oral iron prior to the study and all patients had stopped taking multivitamin preparations at least 1 week prior to admission. No patient had a history, within 1 year, of bleeding, donating blood, and in the 38% of premenopausal females there was no history of increased menstrual bleeding within 6 months of the study.

The possibility could be raised that acid secretion was not sufficiently inhibited to see effects on iron absorption in these patients because they had ZES. It is unlikely that the presence of the ZES was a factor. In the present study 45% of the patients met at least one criterion of acid hyposecretion and 25% met the most stringent criterion of acid secretion < 0.2 mEq/h for 2 of the last 4 years. In a recent study this level of acid hyposecretion is sufficient to result in decreased vitamin B12 levels in patients with ZES.16 In the present study, because the acid secretory rates are taken immediately before the next dose of antisecretory drug, they represent the highest level of acid secretion during the day in the ZES patients whose parietal cells are continually stimulated by autonomous gastrin release from the tumour.37 In patients after gastric resection with a vagotomy the maximal acid output is reduced by 87 ± 1% in a number of studies.62[63][64]–65 The mean maximal acid output after gastric resection varied from 4.7 to 9.4 ± 6.4 mEq/h for men with a range of 0.6–14.5, and in females a mean ± S.E.M. of 8.6 ± 7.2 mEq/h with a range up to 24 mEq/h is reported.35, 62, 64, 65 Each of the three acid hyposecretory criteria used in the present study is more than 8-fold less than the peak secretory values usually seen in a patient after gastric resection without ZES. The mean basal acid secretory rate after a Billroth II or Billroth I operation in patients without ZES in one study was 1.3 ± 1.0 mEq/h and 0.9 ± 0.3 mEq/h,66 respectively, and in a second study after a partial gastrectomy were 0.83 and 1.9 mEq/h in patients with or without a post-operative marginal ulcer.67 These values are also significantly higher than the value of 0.2 mEq/h used as the most stringent criterion of hyposecretion in the present study. The basal values after resection in the above studies represent the minimum value of acid secretion of these patients’ gastric resection, whereas in our patients the 0.2 mEq/h value represents the maximal level because it is determined when the antisecretory drug action is at its nadir. Therefore, in comparison with the acid secretory results after gastric resection for idiopathic peptic ulcer disease, which is associated with a high percentage of patients developing iron deficiency, the depression of acid secretion in our patients is at least, and probably more, profound. This result would suggest that if the presence of the acid hyposecretion is one of the most important factors in causing iron deficiency after gastric resection, then the acid inhibition in the present study with antisecretory drugs was sufficient to expect similar changes in iron parameters.

Haem iron is absorbed independent of acid secretion whereas non-haem iron absorption depends on acid secretion as reviewed above.17, 19[20]–21 Therefore, it could be argued that a high intake of haem iron could compensate for any iron loss due to hypochlorhydria causing non-haem iron malabsorption. There is no evidence to suggest the patients in the present study ate an unusual diet that was particularly high in haem iron. However, in the US 20–30 mg of iron per day is frequently ingested as haem iron in the form of myoglobin or haemoglobin with an average daily iron requirement of 1 mg/day of iron.18 It is therefore a very likely possibility that this could be a factor in failing to see an effect of this extent of acid suppression on iron body stores in the present study. An additional possibility is that gastric hypochlorhydria or achlorhydria alone may not cause iron deficiency. In fact, Wintrobe in 195625, 68 stated that ‘alone achlorhydria does not lead to iron deficiency, as is attested by many observed instances of normal haemoglobin in the face of presumably long-standing achlorhydria.’ Our results are also consistent with this possibility.

The results of the present study suggest that, unlike vitamin B12, iron malabsorption with long-term treatment with potent gastric antisecretory agents such as omeprazole is unlikely to be sufficiently frequent that it requires regular monitoring for it. A recent study10 reported no increased occurrence of iron deficiency in a small number of patients with idiopathic peptic disease treated for up to 4 years and concluded that the risk of iron malabsorption within 3–4 years of starting treatment is probably low. Our results establish that this proposal is correct and extend the observations to the 5–10-year time period. The results with iron absorption are in marked contrast to those found recently by us16 in a similar group of patients for vitamin B12 absorption. We found 7% of patients develop low levels of vitamin B12 with prolonged omeprazole treatment and there was an excellent correlation between decreasing serum vitamin B12 levels and the length of omeprazole treatment. The results of these two studies suggest yearly monitoring of vitamin B12 levels, but not of iron parameters, is required in patients with ZES maintained on long-term omeprazole treatment.

References

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. References
  • 1
    Arnold R. Safety of proton pump inhibitors—an overview. Aliment Pharmacol Ther 1994; 8: 65 70.
  • 2
    Maton P. Risks of hypergastrinemia. Dig Dis Sci 1995; 13: 1 2.
  • 3
    Walt RP, Gomes MD, Wood EC, Logan LH, Pounder RE. Effect of daily oral omeprazole on 24 hour intragastric acidity. Br Med J 1983; 287: 12 14.
  • 4
    Prichard PJ, Yeomans ND, Mihaly GW, et al. Omeprazole: a study of its inhibition of gastric pH and oral pharmacokinetics after morning or evening dosage. Gastroenterology 1985; 88: 64 9.
  • 5
    Jansen JB, Klinkenberg-Knol EC, Meuwissen SG, et al. Effect of long-term treatment with omeprazole on serum gastrin and serum group A and C pepsinogens in patients with reflux esophagitis. Gastroenterology 1990; 99: 621 8.
  • 6
    Lamberts R, Creutzfeldt W, Struber HG, Brunner G, Solcia E. Long-term omeprazole therapy in peptic ulcer disease: gastrin, endocrine cell growth, and gastritis. Gastroenterology 1993; 104: 1356 70.
  • 7
    Freston JW. Clinical significance of hypergastrinaemia: relevance to gastrin monitoring during omeprazole therapy. Digestion 1992; 51: 102 14.
  • 8
    Ekman L, Hansson E, Havu N, Carlsson E, Lundberg C. Toxicological studies on omeprazole. Scand J Gastroenterol 1985; 20(Suppl. 108); 53 69.
  • 9
    Dutta SK. Editorial: vitamin B12 malabsorption and omeprazole therapy. J Am Coll Nutr 1994; 13: 544 5.
  • 10
    Koop H & Bachem MG. Serum iron, ferritin, and vitamin B12 during prolonged omeprazole therapy. J Clin Gastroenterol 1992; 14: 288 92.
  • 11
    Saltzman JR, Kemp JA, Golner BB, Pedrosa MC, Dallal GE, Russell RM. Effect of hypochlorhydria due to omeprazole treatment or atrophic gastritis on protein-bound vitamin B12 absorption. J Am Coll Nutr 1994; 13: 584 91.
  • 12
    Skikne BS, Lynch SR, Cook JD. Role of gastric acid in food iron absorption. Gastroenterology 1981; 81: 1068 71.
  • 13
    Koop H. Review article: metabolic consequences of long-term inhibition of acid secretion by omeprazole. Aliment Pharmacol Ther 1992; 6: 399 406.
  • 14
    Jacobs A, Rhodes J, Peters DK, Campbell H, Eakins JD. Gastric acidity and iron absorption. Br J Haematol 1966; 12: 728 36.
  • 15
    Jacobs P, Bothwell T, Charlton RW. Role of hydrochloric acid in iron absorption. J Appl Physiol 1964; 19: 187 8.
  • 16
    Stewart CA, Termanini B, Gibril F, Weber HC, Jensen RT. Prospective study of the effect of long-term gastric acid antisecretory treatment on serum vitamin B12 levels in patients with Zollinger–Ellison syndrome . Gastroenterology 1995; 108: A226 (Abstract).
  • 17
    Conrad ME. Iron absorption. In: Johnson LR, ed. Physiology of the Gastrointestinal Tract. 2nd edn. New York: Raven Press, 1987: 1437 53.
  • 18
    Turnberg LA & Riley SA. Digestion and absorption of nutrients and vitamins. In: Sleisenger MH, Fordtran JS, eds. Gastrointestinal Disease. 5th edn. Philadelphia: W. B. Saunders, 1993: 977 1008.
  • 19
    Rucker RB, Lonnerdal B, Keen CL. Intestinal absorption of nutritionally important trace elements. In: Johnson LR, ed. Physiology of the Gastrointestinal Tract. 3rd edn. New York: Raven Press, 1994: 2183 202.
  • 20
    Charlton RW & Bothwell TH. Iron absorption. Ann Rev Med 1983; 34: 55 68.
  • 21
    Wheby MS & Crosby WH. The gastrointestinal tract and iron absorption. Blood 1963; 22: 416 28.
  • 22
    Conrad ME & Schade SG. Ascorbic acid chelates in iron absorption: a role for hydrochloric acid and bile. Gastroenterology 1968; 55: 35 45.
  • 23
    Schade SG, Cohen RJ, Conrad ME. The effect of hydrochloric acid on iron absorption. N Engl J Med 1968; 279: 672 4.
  • 24
    Bezwoda W, Charlton R, Bothwell T, Torrance J, Mayet F. The importance of gastric hydrochloric acid in the absorption of nonheme food iron. J Lab Clin Med 1978; 92: 108 16.
  • 25
    Williams J. The effect of ascorbic acid on iron absorption in post-gastrectomy anaemia and achlorhydria. Clin Sci 1959; 18: 521 5.
  • 26
    Goldberg A, Lochhead AC, Dagg JH. Histamine-fast achlorhydria and iron absorption. Lancet 1963; i: 848 50.
  • 27
    Cook JD, Brown GM, Valberg LS. The effect of achlyia gastrica on iron absroption. J Clin Invest 1964; 43: 1185 91.
  • 28
    Golubov J, Flanagan P, Adams P. Inhibition of iron absorption by omeprazole in rat model. Dig Dis Sci 1991; 36: 405 8.
  • 29
    Hines JD, Hoffbrand AV, Mollin DL. The hematologic complications following partial gastrectomy. A study of 292 patients. Am J Med 1967; 43: 555 69.
  • 30
    Wheldon EJ, Venables CW, Johnston ID. Late metabolic sequelae of vagotomy and gastroenterostomy. Lancet 1970; i: 437 40.
  • 31
    Turnberg LA. The absorption of iron after partial gastrectomy. Q J Med 1966; 35: 107 18.
  • 32
    Hobbs JR. Iron deficiency after partial gastrectomy. Gut 1961; 2: 141 9.
  • 33
    Wallensten S. Results of the surgical treatment of peptic ulcer by partial gastrectomy according to billroth I and II methods; Supplementum 191. Acta Chir Scand, 1954; 1 157.
  • 34
    Metz DC & Jensen RT. Advances in gastric antisecretory therapy in Zollinger–Ellison syndrome. In: Mignon M, Jensen RT, eds. Endocrine Tumors of the Pancreas: Recent Advances in Research and Management. Series: Frontiers of Gastrointestinal Research. Basel, Switzerland: S. Karger, 1995: 240 57.
  • 35
    Magnusson BEO. Iron absorption after antrectomy with gastroduodenostomy. Scand J Haematology 1976; (Suppl. 26): 1 107.
  • 36
    Deller DJ, Richards WCD, Witts LJ. Changes in the blood after partial gastrectomy with special reference to vitamin B12. Q J Med 1962; 31: 89 102.
  • 37
    Jensen RT & Gardner JD, Gastrinoma. In: Go VLW, DiMagno EP, Gardner JD, Lebenthal E, Reber HA, Scheele GA, eds. The Pancreas: Biology, Pathobiology and Disease. 2nd edn. New York: Raven Press, 1993: 931 78.
  • 38
    Fishbeyn VA, Norton JA, Benya RV, et al. Assessment and prediction of long-term cure in patients with Zollinger–Ellison syndrome: the best approach. Ann Intern Med 1993; 119: 199 206.
  • 39
    Benya RV, Metz DC, Venzon DJ, et al. Zollinger–Ellison syndrome can be the initial endocrine manifestation in patients with multiple endocrine neoplasia-type 1. Am J Med 1994; 97: 436 44.
  • 40
    Orbuch M, Doppman JL, Strader DB, et al. Imaging for pancreatic endocrine tumor localization: recent advances. In: Mignon M, Jensen RT, eds. Endocrine Tumors of the Pancreas: Recent Advances in Research and Management. Series: Frontiers of Gastrointestinal Research. Basel, Switzerland: S. Karger, 1995: 268 81.
  • 41
    Sugg SL, Norton JA, Fraker DL, et al. A prospective study of intraoperative methods to diagnose and resect duodenal gastrinomas. Ann Surg 1993; 218: 138 44.
  • 42
    Gibril F, Reynolds JC, Doppman JL, et al. Somatostatin receptor scintigraphy: its sensitivity compared with that of other imaging methods in detecting primary and metastatic gastrinomas: a prospective study. Ann Intern Med 1996; 125: 26 34.
  • 43
    Cook JD. Iron-deficiency anaemia. Bailliere’s Clin Haematol 1994; 7: 787 804.
  • 44
    Guyatt GH, Oxman AD, Ali M, Willan A, McIlroy W, Patterson C. Laboratory diagnosis of iron-deficiency anemia: An overview. J Gen Intern Med 1992; 7: 145 53.
  • 45
    Hastka J, Lasserre JJ, Schwarzbeck A, Reitter A, Hehlmann R. Laboratory tests of iron status: correlation or common sense? Clin Chem 1996; 42: 718 24.
  • 46
    Cook JD, Lipschitz DA, Miles LEM, Finch CA. Serum ferritin as a measure of iron stores in normal subjects. Am J Clin Nutr 1974; 27: 681 7.
  • 47
    Walters GO, Miller FM, Worwood M. Serum ferritin concentration and iron stores in normal subjects. J Clin Path 1973; 26: 770 2.
  • 48
    Lipschitz DA, Cook JD, Finch CA. A clinical evaluation of serum ferritin as an index of iron stores. N Engl J Med 1974; 290: 1213 16.
  • 49
    Skoog DP. Laboratory evaluation of iron deficiency. In: Bick RL, ed. Hematology. St. Louis: Mosby Yearbook, 1993: 273 9.
  • 50
    Cook JD.Clinical evaluation of iron deficiency. Semin Hematol 1982; 19: 6 18.
  • 51
    Baird IM, Blackburn EK, Wilson GM. The pathogenesis of anaemia after partial gastrectomy. I. Development of anaemia in relation to time after operation, blood loss, and diet. Q J Med 1959; 109: 21 34.
  • 52
    Jacobs A. Effect of gastric juice and pH on inorganic iron in solution. Nature 1967; 216: 707 8.
  • 53
    Jacobs A & Miles PM. Role of gastric secretion in iron absorption. Gut 1969; 10: 226 9.
  • 54
    Minot GR & Heath GW. The response of reticulocytes to iron. Am J Med Sci 1932; 183: 110 18.
  • 55
    Baird IM & Wilson GM. The pathogenesis of anaemia after partial gastrectomy. II. Iron absorption after partial gastrectomy. Q J Med 1959; 109: 35 41.
  • 56
    Choudhury MR & Williams J. Iron absorption and gastric operations. Clin Sci 1959; 18: 527 32.
  • 57
    Shafer RB, Ripley D, Swaim WR, Mahmud K, Doscherholmen A. Hematologic alterations following partial gastrectomy. Am J Med Sci 1973; 266: 240 8.
  • 58
    Brookes VS, Meynell MJ, Bold AM, Kingston RD. A review of symptoms, haematology and clinical chemistry following a partial gastrectomy. Br J Surg 1974; 61: 9 15.
  • 59
    Svensson A. Vagotomy with antrum resection. Acta Chir Scand 1974; 140: 50 6.
  • 60
    Wastell C. Long-term clinical and metabolic effects of vagotomy with either gastrojejunostomy or pyloroplasty. Ann R Coll Surg Engl 1969; 45: 193 211.
  • 61
    Wastell C. Metabolic effects of vagotomy and gastric drainage. Postgrad Med J 1967; 43: 481 6.
  • 62
    Baron JH. Selective surgery for duodenal ulcer. In: Clinical Tests of Gastric Secretion History, Methodology and Interpretation. New York: Oxford University Press, 1979: 152 263.
  • 63
    Rovelstad RA, Maher FT, Adson MA. Gastric analysis. Surg Clin North Am 1971; 51: 969 78.
  • 64
    Talstad I, Fretheim B, Myren J, Gjone E, Holter O. Graded gastrectomy for duodenal ulcer—a five-year prospective study. Scand J Gastroenterol 1975; 33(Suppl.): 1 27.
  • 65
    Myren J, Gjeruldsen S, Fretheim B. Gastric secretion before and after graded partial gastrectomy for duodenal ulcer. Scand J Gastroenterol 1966; 1: 132 7.
  • 66
    Thompson JC, Fender HR, Watson LC, Villar HV. The effects on gastrin and gastric secretion of five current operations for duodenal ulcer. Ann Surg 1976; 183: 599 608.
  • 67
    Aoyagi T & Summerskill WH. Gastric secretion with ulcerogenic islet cell tumor. Importance of basal acid output. Arch Intern Med 1966; 117: 667 72.
  • 68
    Wintrobe MM. Clinical Haematology. 4th edn. London: Kimpton, 1956.