Gastrectomy osteopenia in the rat: the role of ... - Semantic Scholar

Clinical Science (1998) 95, 735–744 (Printed in Great Britain)

Gastrectomy osteopenia in the rat: the role of vitamin B12 deficiency and the type of reconstruction of the digestive tract A. WOJTYCZKA*1, B. BERGE; *, G. RU= MENAPF*†, P. O. SCHWILLE*, P. BALLANTI‡, M. SCHREIBER*†, W. FRIES* and W. HOHENBERGER† *Division of Experimental Surgery and Endocrine Research Laboratory, University of Erlangen, D-91023 Erlangen, Germany, †Department of Surgery, University of Erlangen, D-91023 Erlangen, Germany, and ‡Department of Experimental Medicine and Pathology, La Sapienza, University of Rome, Italy

A

B

S

T

R

A

C

T

1. The mechanisms underlying gastrectomy osteopenia are not yet clear. The gastrectomyassociated cobalamin (vitamin B12) deficiency may favour osteopenia and skeletal fractures. Also, the exclusion of the duodenum from the food passage may contribute to gastrectomy osteopenia. To investigate this, rats were gastrectomized and the passage of nutrients restored either with the duodenum excluded (Roux Y) or included (Longmire). Sham-operated rats served as controls. In half of the rats in each gastrectomy group the serum B12 levels were normalized by parenteral administration of B12. 2. Four months post operation, both gastrectomy groups showed a similar degree of osteopenia. There was normal bone mineralization ; serum levels of parathyroid hormone were normal, but decreased for 25-hydroxyvitamin D, and elevated for 1,25-dihydroxyvitamin D ; in urine there was decreased pH and excessive hyperphosphaturia. 3. B12 therapy had no influence on any of the essential bone and mineral metabolic parameters. 4. We conclude that osteopenia in the gastrectomized rat (i) is not due to B12 or folic acid deficiency, calcium deficiency or secondary hyperparathyroidism ; (ii) is independent of the type of anatomic reconstruction of the digestive tract ; (iii) appears to be related to disturbed vitamin D, phosphorus and acid–base metabolism.

INTRODUCTION In humans, total gastrectomy (GX) may lead to deficiency syndromes. Among these is osteopenia [1–4], the pathophysiology of which is poorly understood. The loss of gastric acid, gastric reservoir function, the exclusion of

the duodenum, pancreato–cibal asynchrony and malabsorption of calcium and vitamin D are currently being discussed as possible determining factors [5–7]. No attention has as yet been paid to the fact that GX patients or experimental animals are deprived of intrinsic factor, and consequently cannot utilize cobalamin (vitamin B )

"#

Key words : hyperphosphaturia, Longmire reconstruction, osteopenia, Roux-Y reconstruction, total gastrectomy, vitamin B "# deficit, vitamin B therapy. "# Abbreviations : GX, gastrectomy ; PTH, parathyroid hormone. Correspondence : Dr P. O. Schwille. " On leave from the Department of Gastrointestinal Surgery, Silesian University Medical School, Katowice, Poland.

# 1998 The Biochemical Society and the Medical Research Society

735

736

A. Wojtyczka and others

and folic acid [8] ; these individuals may present with atrophy of the intestinal mucosa and megaloblastic anaemia [9–13]. B deficiency decreases serum total [14] "# and bone [15] alkaline phosphatase levels, owing to impaired osteoblastic activity. Also, patients with pernicious anaemia have a high incidence of fractures [16] and osteoporosis [17]. The latter has been successfully treated with B and etidronate [18]. "# Currently, the Longmire type reconstruction after GX is considered the more physiological procedure, since the duodenum is preserved. On the other hand, the Roux-enY reconstruction is considered a technically simple procedure. In terms of the occurrence of osteopathy, the Longmire procedure appears to be superior [20], but more in-depth studies on the late complications of the two modes of surgery are still lacking. In the present study we used rats, since they have proved to be a suitable species for comparative studies of bone mineral [19,21,22], and because the post-GX complication rate is low [23–26]. The aims of the present work were to compare the effects of the Longmire and Roux reconstruction after removal of the stomach on the haematological status, the state of mineral metabolism, several hormones, and the physical state of bone in untreated and B -treated GX rats.

"#

MATERIALS AND METHODS Animals The experiments described in this paper comply with the ‘ Bundesdeutsches Tierschutzgesetz ’ (German Animal Protection Act) of 1986, and ‘ The Government Committee for Animal Experimentation ’. Male Sprague– Dawley rats of the CD strain (Charles River, Wiga, Sulzfeld, Germany) were housed in Makrolon cages at a room temperature of 23p1 mC and 60 % relative humidity, under a 12-h light\12-h dark cycle. Free access to tap water and a standard laboratory rat diet (code no. 1321 ; Altromin, Lage, Germany) was allowed. The diet contained (dry matter) : crude protein (19 %), crude fat (4 %), utilizable carbohydrates (38 %), calcium (0.9 %), phosphorus (0.8 %) and magnesium (0.2 %) ; the content of vitamin D was 1200 i.u.\kg, of vitamin B 0.041 mg\ $ "# kg, and of folic acid 10 mg\kg. Phosphorus was supplied in the form of calcium-dihydrogen phosphate, which supplied part of the calcium, the remaining calcium being in the form of the carbonate ; magnesium was supplied as the oxide. The drinking water contained 64 mg\l calcium and 22 mg\l magnesium, but no phosphorus. After a 2week acclimatization period food, but not water, was withheld for 18 h. Animals, with a body weight of approx. 280–300 g, were randomly allocated to the study groups defined below. Surgery was performed under diethyl ether anaesthesia. # 1998 The Biochemical Society and the Medical Research Society

Surgery Longmire procedure (GXL ; nl10) After a midline laparotomy, the stomach was excised together with a 1 mm cuff of oesophagus and a 4–5 mm cuff of duodenum. A free jejunal loop, supplied by two to three major jejunal vessels, was isolated by transection at about 2 cm and 5–6 cm distal to the ligament of Treitz, and the proximal end anastomosed end-to-end with the oesophagus in an antecolic and isoperistaltic position, using an invagination technique (A. Wojtyczka, unpublished work) guaranteeing a high survival rate. The operation was terminated by anastomosing the distal end of the jejunal loop to the duodenum, performing an endto-end jejuno–jejunostomy to restore the continuity, and then closing the abdomen. Using this technique resulted in minimal oesophageal inflammation and dilation cephalad to the anastomosis, and a less than 20 % mortality during the first month post operation.

Roux-en-Y procedure (GXR ; nl20) Transection of the jejunum was carried out 2 cm distal to the ligament of Treitz, the proximal jejunal end anastomosed end-to-side to the main jejunum 5–6 cm downstream, and the distal jejunal end anastomosed to the oesophagus as described above.

Mock surgery (SHAM ; nl20) After laparotomy the stomach was exposed, and the proximal jejunum was transected and the cut ends anastomosed together. To imitate the non-specific stress encountered with the longer duration of gastrectomies (see above), the abdomen closure was delayed for 75 min. All operated animals had free access to water, but standard chow was not proffered until 48-h post operation. Half of the rats in each of the two GX groups received cobalamin injections (Cytobion 1000, Merck, Darmstadt, Germany), 2150 µg per rat, commencing in the second week after surgery and repeated every second week ; the remaining GXL and GXR as well as the SHAM rats received saline vehicle injections. The postoperative observation period was 116p3 days.

Other procedures Twenty and 10 days before killing, the animals were labelled by intramuscular injection of 925 kBq of %&Ca (Amersham, Braunschweig, Germany) in 0.4 ml of saline, to identify the movements of calcium. Nine and 4 days before the end of the trial, calcein (20 mg\kg body weight) was injected intraperitoneally. During the last week the intake of food and water was recorded, and faeces and paraffin-covered urine were collected over two consecutive 24-h periods from the animals housed in plastic metabolic cages. The volume and pH of the urine were immediately measured and the urine was then

Postgastrectomy osteopathy and vitamin B12

acidified to pH 2.0 by adding 6 M HCl (to prevent calcium precipitation) ; aliquots were stored at k20 mC until analysed. Wet and dried (100 mC for 24 h) faeces were weighed, the latter incinerated (800 mC for 12 h), reweighed and dissolved in 10 ml of 6 M HCl for analysis of minerals. On termination of the experiments the rats were fasted for 18 h, anaesthetized (see above), laparotomized and exsanguinated from the abdominal aorta. Blood gases and haematological parameters were immediately measured, and serum and plasma samples prepared for analysis of additional variables. Bones (right femur, right tibia) were excised, freed of adherent tissue and defatted in ethanol at 4 mC for 48 h. The dried (100 mC for 24 h) femur was weighed and incinerated (800 mC for 24 h), and the ash reweighed and dissolved in 10 ml of 6 M HCl for analysis of minerals. The tibia was further processed for histomorphometry and determination of several static parameters [19].

Analyses Haematocrit, haemoglobin and erythrocyte count were obtained using the Blood Analyser K 1000 (Sysmix, Hamburg, Germany), and blood gases using the pH\ Blood Gas Analyzer 1306 (Instrumentation Laboratory, Milan, Italy). Commercial kits were used for radioimmunoassay of serum cobalamin and folic acid (IBL, Hamburg, Germany), calcitonin, bioactive rat parathyroid hormone (PTH), 25-hydroxyvitamin D and 1,25dihydroxyvitamin D (all from Nichols Institute, Bad Nauheim, Germany). In serum we also measured glucose (enzymically, by Glucose Analyzer, Beckman, Palo Alto, U.S.A.), α-amino nitrogen (by colorimetry, using glycine as standard), free fatty acids (enzymically, by kit, Merck, Darmstadt, Germany), total protein (refractometry), phosphorus (also in urine and bone ash ; colorimetry) and alkaline phosphatase (colorimetry). Calcium in serum, faeces, urine and bone ash was determined by atomic absorption spectrophotometry (FL-6, Zeiss, Oberkochen, Germany), as was bone ash magnesium. Other analytes were urinary hydroxyproline (by colorimetry, after prior extraction), in a limited number of samples the bone collagen crosslinks pyridinium and deoxypyridinium (by HPLC, after extraction, using the reagents distributed by Immunodiagnostik, Bensheim, Germany), and urinary cAMP (by protein-binding in-house radioassay, intra- and inter-assay coefficients of variation 15 %). %&Ca was counted in a liquid scintillation spectrophotometer (Betaszint, Berthold, Munich, Germany). Several of the analytical techniques, including more details on sample processing, specificity and precision of determination, have been described previously [19,27,28]. The femur volume and mean specific density were determined according to Archimedes ’ principle. The mineral content of the femur was measured by computer-

ized X-ray tomography (for details see [28]) and expressed as equivalents of hydroxyapatite. The bone breaking force was expressed as the energy applied perpendicular to the femoral mid-diaphysis (Pendelschlagwerk 5102, Zwick, Ulm, Germany).

Calculations and statistics The ratio of the body weight gain post operation and the mean daily food intake during the period in the metabolism cage was taken as a measure of food efficiency. For the same period, intestinal calcium secretion was assessed as the mean daily %&Ca faecal excretion divided by the mean %&Ca specific activity in urine. Bone dry weight and minerals were factorized for bone volume, as were substances in urine for urinary creatinine. One-way analysis of variance was applied to the untreated (kB12) SHAM, GXL and GXR rats ; where the F-value was found to be significant, the experimental groups were compared with the SHAM group using the test for leastsquare differences, and P 0.05. Untreated (kB12) GXL and GXR rats were compared with their treated (jB12) counterparts using the t- or U-test for unpaired observations, as appropriate. Because post-GX urinary phosphorus excretion was markedly elevated, and as phosphorus deficiency has been shown to cause osteopathy [29,30], urinary phosphorus was correlated with a series of independent variables, using bi- and multivariate regression analysis (software STATISTICA, Statsoft, Tulsa, OK, U.S.A.).

RESULTS In each of the next three sections the results are given in the following order : effects of GX in untreated (kB ) "# GXL and GXR rats, i.e. in comparison with SHAM rats ; thereafter the effects of cobalamin treatment (jB ) are "# addressed.

General state of GXL and GXR animals, and blood levels of cobalamin (Table 1) GXL and GXR surgery was followed by a significant increase in food intake and wet weight of faeces, but a decrease in food efficiency and final body weight. The fasting serum levels of glucose, free fatty acids and αamino nitrogen were normal, indicating normal metabolism of the major calorie carriers (results not shown) ; the associated serum levels of gastrin, calcitonin and total protein were decreased, the degree of reduction being comparable for GXL and GXR. Serum creatinine remained normal in all groups. The blood pH was lowered in GXL but not GXR animals, while urinary pH was markedly decreased in both groups. There was a dramatic reduction in the mean serum B levels to 6 (GXL) and

"#


737

738


Table 1 General data of sham-operated (SHAM), gastrectomized (GXL ; GXR), vitamin B12-treated (jB12) and untreated (kB12) animals

Values are means (S.E.M.) ; n l 10 rats per group. Abbreviations : S, fasting serum ; b.w., body weight. *P 0.05 compared with kB12 (same protocol) ; †P 0.05 compared with the respective value in SHAM (by analysis of variance and least squares difference test) GXL

Initial body weight (g) Final body weight (g) Weight gain (g) Food intake (g:day−1:100 g−1 b.w.) Food efficiencya (g/g) Faeces wet weight (g:day−1:100 g−1 b.w.) Haematocrit (%) Erythrocytes (106 cells/µl) Haemoglobin (g/l) S-B12 (pmol/l) S-Folic acid (nmol/l) S-Creatinine (µmol/l) S-Total protein (g/l) S-Gastrin (pmol/l) S-Calcitonin (pmol/l) Blood pH Urinary pH a

GXR

SHAM kB12

kB12

jB12

kB12

jB12

294 (2) 552 (17) 258 (16) 3.6 (0.2) 13 (1) 2.2 (0.2) 47 (1) 8.9 (0.2) 160 (3) 464 (48) 54 (2) 50 (2) 58 (1) 28 (6) 8.0 (1.3) 7.44 (0.02) 7.60 (0.09)

284 (2)† 427 (8)† 143 (9)† 4.5 (0.2)† 8 (1)† 2.9 (0.2)† 41 (2) 9.5 (0.3) 150 (4) 27 (7)† 32 (2)† 48 (1) 53 (1)† 9 (1)† 4.7 (0.3)† 7.37 (0.03)† 6.73 (0.10)†

295 (3)* 488 (11)* 193 (11)* 4.1 (0.2) 10 (1)* 2.4 (0.1) 45 (1) 8.6 (0.4) 140 (7) 242 (21)* 45 (2)* 54 (2) 56 (1) 10 (1) 6.7 (1.0)* 7.39 (0.02) 7.05 (0.13)

282 (3)† 470 (18)† 188 (18)† 4.3 (0.3)† 10 (1)† 2.8 (0.2) 36 (2)† 7.9 (0.6) 100 (8)† 58 (22)† 39 (2)† 52 (2) 53 (1)† 10 (1)† 4.7 (0.3)† 7.47 (0.02) 6.76 (0.14)†

294 (3)* 471 (13) 177 (14) 4.2 (0.2) 9 (1) 2.7 (0.2) 42 (1)* 8.7 (0.2) 120 (1)* 246 (18)* 43 (2) 52 (2) 54 (1) 9 (0.5) 5.0 (0.3) 7.41 (0.01)* 6.80 (0.21)

Body weight gain (g) divided by food intake (g per animal and day) ; for details see Materials and methods section.

12 % (GXR) of control values. The levels of folic acid were also significantly reduced in both GX groups. The erythrocyte count remained statistically unchanged, while the low haemoglobin and haematocrit values signal the presence of microcytic anaemia, and in this respect the GXR procedure had a more pronounced negative effect. The effects of B treatment were as follows : in both "# GXL and GXR rats the blood levels of B and folic acid "# were substantially raised compared with the levels found in the untreated animals. Although within the normal range, the mean B levels in the B12-treated GX rats "# were sub-optimal relative to those of SHAM rats. In the GXL rats, B treatment significantly improved food "# efficiency, weight gain and calcitonin, but not in the GXR rats in whom, however, improvements in haemoglobin, erythrocyte count and haematocrit were seen. In this latter group the blood pH was decreased compared with the untreated animals.

State of minerals and hormones in faeces, serum and urine (Table 2) In GXL and GXR rats there was increased faecal calcium associated with high intestinal secretion of calcium (for comparison, intestinal calcium secretion in humans averages 7 % in infants and 14 % in women [31,32]). The # 1998 The Biochemical Society and the Medical Research Society

contribution of secreted calcium to the excess of calcium in faeces is approx. 50 % (GXL and GXR), the rest readily being accounted for by an increased dietary intake (8 mg in GXL, 5 mg in GXR). Also, in GXL, the mean total alkaline phosphatase serum level was higher than normal, but was significantly increased in GXR. In both GX groups the mean serum levels of 25-hydroxyvitamin D were low. Total serum calcium was unchanged, but showed a trend towards higher values in both GXL and GXR. Serum PTH was always within the normal range ; in GX rats, levels of the hormone tended to be low, not high. In contrast, in comparison with SHAM, the serum 1,25(OH) D levels were greatly increased # (GXL by factor 6, GXR by factor 5 ; for the associated serum and urinary phosphorus levels see below). The mean urinary calcium excretion was low (GXL and GXR), but the difference compared with SHAM rats did not reach significance. Urinary cAMP excretion, in general paralleling parathyroid gland activity, was only insignificantly higher than in the SHAM rats. In contrast, urinary phosphorus excretion was drastically increased (GXL, GXR), and the mean serum phosphorus concentrations were also higher than in SHAM rats. Urinary hydroxyproline was insignificantly lower after GX, and so were the bone collagen crosslinks. The mean values of the latter in the untreated rats were as follows :


Table 2

Data of the effect of gastrectomy, without and with vitamin B12 treatment, on minerals and other substances

Values are means (S.E.M.) ; n l 10 rats per group. For further information see Table 1, and Materials and methods section. *P 0.05 compared with kB12 (same protocol) ; †P 0.05 compared with the respective value in SHAM (by analysis of variance and least squares difference test). GXL SHAM Faeces Total calcium (mmol:g−1 dry matter:day−1) Endogenous calcium secretion (%) Serum 25-Hydroxyvitamin D (nmol/l) 1,25(OH)2D (pmol/l) Rat PTH (pmol/l) Calcium (mmol/l) Phosphorus (mmol/l) Alkaline phosphatase (i.u./l) Urine Calcium (µmol/mmol creatinine) Phosphorus (mmol/mmol creatinine) Cyclic AMP (nmol/mmol creatinine) Hydroxyproline (µmol/mmol creatinine) Bone Hydroxyapatite equivalent (Hounsfield units) Fracturing energy (mJ/100 g b.w) Dry weight (µg/100 g b.w.) Volume (µl/100 g b.w.) Density (g/ml) Ash weight (µg/100 g b.w.) Calciumb (mmol/ml) Phosphorusb (mmol/ml) Magnesiumb (mmol/ml) b

GXR

kB12

jB12

kB12

0.68 (0.03) 4.4 (0.2)

1.10 (0.03)† 21 (4)†

1.03 (0.03)* 13 (2)

57 (8) 174 (34) 1.8 (1.0) 2.35 (0.25) 1.87 (0.06) 45 (6)

31 (3)† 1049 (143)† 1.8 (0.7) 2.50 (0.05) 2.16 (0.05)† 52 (7)

34 (5) 834 (81) 1.5 (0.4) 2.43 (0.03) 1.53 (0.05)* 63 (13)

34 (5)† 951 (86)† 0.5 (0.1) 2.40 (0.05) 1.94 (0.05) 77 (11)†

283 (113) 2.66 (0.58) 1.70 (0.11) 46 (5)

150 (8) 7.47 (0.44)† 2.03 (0.23) 37 (4)

110 (14)* 7.58 (0.51) 2.15 (0.11) 33 (6)

181 (23) 8.38 (0.80)† 2.71 (0.45) 33 (6)

775 (48) 22 (2) 241 (6) 151 (4) 1.59 (0.01) 106 (4) 6.90 3.77 (0.19) 0.204 (0.08)

623 (29)† 608 (14) 19 (2) 18 (2) 262 (6)† 257 (3) 185 (6)† 178 (2) 1.42 (0.02)† 1.45 (0.01) 96 (3)† 99 (1) (0.23) 5.58 (0.23)† 5.40 (0.15) 3.10 (0.16)† 2.74 (0.16) 0.167 (0.008)† 0.167 (0.004)

0.98 (0.05)† 14 (2)

547 (21)† 19 (2) 259 (7)† 183 (6)† 1.42 (0.01)† 92 (2)† 5.40 (0.23)† 2.42 (0.26)† 0.154 (0.004)†

jB12 0.93 (0.03)*† 14 (1) 42 (5) 1187 (236) 1.0 (0.2)* 2.48 (0.05) 1.70 (0.10) 61 (11) 99 (8)* 6.93 (0.40) 2.15 (0.11) 28 (5) 565 (31) 20 (2) 260 (4) 183 (4) 1.42 (0.01) 98 (3) 5.18 (0.20) 2.94 0.29 0.163 (0.004)

Per unit bone volume.

pyridinium\creatinine ratios (nmol\mmol) in 4 to 5 rats per group – 56 (SHAM), 48 (GXL), 51 (GXR) ; deoxypyridinium\creatinine – 46 (SHAM), 37 (GXL), 43 (GXR). B treatment evoked a decrease in faecal calcium "# excretion (GXL, GXR) and serum phosphorus concentration (GXL), but some increase in serum PTH (GXR). B treatment also evoked a significant decrease "# in urinary calcium, and led to insignificantly lower hydroxyproline, but did not affect urinary crosslinks (results not shown). Except for a slight tendency to higher bone ash weight and ash magnesium, B treat"# ment also failed to change bone parameters.

State of bone (Tables 2 and 3) Compared with SHAM rats the GXL and GXR animals showed a decreased mean bone density, as a result of the underlying higher bone volume and lower bone dry weight (Table 2). These parameters were accompanied by decreased bone ash calcium, phosphorus and magnesium

content ; ash calcium corresponded well with the hydroxyapatite equivalent of the isolated femur (Table 2). The fracturing energy was statistically unchanged. The mean urinary hydroxyproline excretion, an established marker of bone resorption, was lower in GXL and GXR than in SHAM rats ; this finding precludes the possibility that increased bone resorption – as in secondary hyperparathyroidism – was a key change resulting in abnormal bone (see also below). In view of previous, detailed data [19], bone histomorphometry was obtained from only three animals each of the untreated and B -treated GXL, GXR and SHAM "# rats (Table 3). There was an impressive reduction in bone area relative to the whole tissue area, bone perimeter per tissue area, and of the parameters derived from the measurements of trabeculae. Obviously, in one sample each of the GXL and GXR groups the tissue section was obtained from an almost trabecular-free region, but in the remaining samples the ratio bone area\whole tissue area was far below that seen in the SHAM rats. # 1998 The Biochemical Society and the Medical Research Society

739

740


Table 3

Bone histomorphometry (static parameters)

Top panel : individual values of three SHAM rats and three rats each from the treatment groups. Bottom panel : mean values (S.E.M.) of the pooled (Σ ) observations. For further information see Table 1, and Materials and methods section. *Note that due to the small number of SHAM rats no statistical comparison was undertaken. GXL kB12

SHAM 2

Tissue area (mm ) Bone area/tissue area (%) Bone perimeter/tissue area (mm/mm2) Trabecular width (µm) Trabecular number (no./mm) Trabecular separation (mm)

Tissue area (mm2) Bone area/tissue area (%) Bone perimeter/tissue area (mm/mm2) Trabecular width (µm) Trabecular number (no./mm) Trabecular separation (mm)

2.25 15.4 1.4 63 2.4 0.35

GXR

2.55 8.0 2.9 50 1.6 0.57

2.25 8.4 2.9 51 1.6 0.56

2.55 1.90 0.82 42 0.46 2.1

jB12 2.28 0.10 0.10 18 0.06 16.6

2.55 3.00 1.36 40 0.76 1.3

ΣSHAM (n l 3)

ΣGXL* (n l 6)

ΣGXR* (n l 6)

2.55 (0.00) 10.6 (2.4) 2.4 (0.5) 55 (4.2) 1.9 (0.3) 0.49 (0.05)

2.53 (0.01) 1.7 (0.24) 0.73 (0.13) 39 (2) 0.40 (0.07) 4.8 (0.05)

2.55 (0.004) 1.9 (0.4) 0.48 (0.20) 43.0 (2.5) 0.27 (0.11) 6.7 (1.6)

kB12

2.54 2.02 0.86 42 0.42 2.9

2.21 1.40 0.57 43 0.32 3.1

2.55 1.78 0.64 50 0.36 2.7

2.25 1.63 0.68 4751 0.38 2.6

jB12 2.55 3.93 1.38 47 0.77 1.2

2.54 0.43 0.16 .238 0.09 11.1

2.25 2.02 0.19 37 0.11 9.4

2.53 1.40 0.22 39 0.12 9.2

2.25 1.78 0.27 0.15 6.6

Table 4 Bivariate (left) and multivariate (right) stepwise logistic regression analysis, using urinary phosphorus/creatinine excretion (see Table 2) as dependent variable, and independent variables considered of pathophysiological relevance in the development of GX osteopenia of untreated (kB12) GXL, GXR and sham-operated control rats

Total number of animals l 28. For further comments see text. β is the standardized regression coefficient ; R 2 for the model (adjusted for confounding variables) l 0.71 (P 0.0001).

r Cyclic AMP 1,25(OH)2D PTH Calcitonin Gastrin Cobalamin Folic acid Mean density* Osteo-CT* Calcium* Magnesium* Phosphorus*

0.62 0.53 k0.16 k0.18 k0.58 k0.66 k0.51 k0.66 k0.55 k0.57 k0.59 k0.40

P 0.0001 0.004 0.41 0.35 0.001 0.0001 0.005 0.0001 0.003 0.001 0.001 0.04

β Cyclic AMP Calcium* Gastrin PTH Calcitonin 1,25(OH)2D

0.474 k0.389 k0.386 k0.112 0.075 0.053

R 2 (%)++

P-value

22.5 15.1 14.9 1.3 0.6 0.3

0.0004 0.003 0.008 0.36 0.59 0.76

* Measured in femur.

B treatment had no effect on the physical and "# chemical quality (Table 2), or histomorphometry (Table 3) of bone.

Interrelationships of urinary phosphorus and independent variables (Table 4) The most striking abnormalities found in the GXL and GXR rats were the low B blood levels, high urinary "# phosphorus excretion, high serum 1,25(OH) D, and the # low bone mineral content, while normal serum calcium # 1998 The Biochemical Society and the Medical Research Society

and PTH values excluded the presence of secondary hyperparathyroidism [5,20,33–36]. The possibility of overlooking the latter was further minimized by regression analyses. There were highly significant positive correlations of urinary phosphorus with urinary cAMP and serum 1,25(OH) D, and significant negative corre# lations with serum B , folic acid and gastrin, bone "# minerals (chemically measured) and bone hydroxyapatite (X-ray measurement) ; serum PTH and calcitonin, inhibitors of renal tubular phosphorus reabsorption,


did not significantly correlate with urinary phosphorus (Table 4, left panel). On stepwise inclusion of selected variables in multivariate regression analysis, urinary cAMP, bone calcium and serum gastrin were identified as significant predictors of urinary phosphorus, while the contribution of PTH, 1,25(OH) D and calcitonin was # weak (Table 4, right panel). Together, the six variables account for approx. 55 % of the total variation of urinary phosphorus, indicating the existence of additional, as yet unknown, factors.

DISCUSSION Gastrectomy, B12 and general status The GX rats of the present study have decreased body weight and food efficiency, whether the duodenum is bypassed (GXR) or not (GXL). A number of causes for their reduced nutritional state have been discussed elsewhere [6,7,9,11,37]. Interestingly, the base metabolic rate after GX is normal [37], which is in line with the normal levels of fasting glucose, free fatty acids and αamino nitrogen in the present study. In earlier studies [38] using end-to-end oesophago–jejunostomy the postoperative weight gain of GXR rats was much lower than in GXL. Using invagination, this difference no longer exists. Thus the duodenal passage of food may be less essential for the nutritional state of GX than previously assumed [38,39]. Ongoing work indicates that the degree of surgery-induced inflammation of the oesophago– jejunostomy differs between GXL and GXR, suggesting that local changes in tissue morphology give rise to factors that may play a systemic role. Vitamin B serum levels of 200–900 pg\ml are normal "# in both humans and the rat ; values less than 140 pg\ml indicate significant B12 deficiency [40]. GX removes the intrinsic factor and causes ileal malabsorption of vitamin B ; in humans, the lag time for the onset of B deficit "# "# is up to 6 years, due to the large reserves of the body [41]. The GX rats of the present study developed serious deficiency of B and folate within 16 weeks post "# operation, but contrary to humans, megaloblastic anaemia was absent. GXR rats rather had microcytic anaemia, and B treatment increased both the number "# and haemoglobin content of erythrocytes. This obvious discrepancy between GXL and GXR may be explained by the exclusion of the duodenum from the food passage with subsequent malabsorption of iron. In this respect the Longmire reconstruction appears superior to RouxY. The positive effect of B on food efficiency and body "# weight of GXL rats remains unexplained.

GX osteopenia and calcium In both GXL and GXR rats a comparable reduction of bone density, ash weight and mineral content was found.

Dynamic bone histomorphometry in B -deficient GXL "# rats demonstrated high-turnover osteopenia, with increased mineral apposition rate and decreased osteoid maturation time, but no signs of secondary hyperparathyroidism [19]. The extra-osseous changes found in the B -deficient GXL and GXR rats in the present study "# confirm these findings : in the serum unchanged PTH and calcium, massively elevated 1,25(OH) D and decreased # 25-hydroxyvitamin D, as well as hyperphosphaturia. The histomorphometric data of the present study underscore that the decrease in bone mass is not due to general thinning of trabeculae, but rather to the complete loss of functional units [19]. Again, no difference is seen between GXL and GXR. Some PTH-independent increase of bone turnover may be present in GXL and GXR rats, based on higher serum total alkaline phosphatase. However, the contribution by the bone isoenzyme – a product of osteoblasts – to the high enzyme activity is questionable in view of the low urinary excretion of hydroxyproline and crosslinks, findings not commensurate with increased bone turnover. The difference of bone hydroxyapatite between GXL and GXR remains unexplained since there is no equivalent difference in bone calcium as determined by the direct and precise chemical method (Table 2). The bone hydroxyapatite may therefore not be sufficiently sensitive in the determination of post-GX bone mineral density in the rat. The energy required to break the bones of GXL and GXR rats is normal, indicating that ‘ bone strength ’ is not only determined by the integrity of bone mineral but also of non-mineralized tissue, which was not studied (see also below). The degree of bone loss is almost equal in GXL and GXR, although the two surgical procedures differ with respect to the duodenum, which has a high absorption capacity for calcium. Consequently, in the GXR rats, in whom the duodenum is bypassed, preservation of extracellular calcium homoeostasis may have been achieved via small intestinal adaptation. GX rats may therefore not suffer from calcium deficiency, and may not need to stress the intestinal absorptive reserves for calcium [42]. This assumption is supported by the failure of enteral and parenteral administration of calcium to prevent GX osteopenia ([42,43] ; T. Koch, M. Schreiber and P. O. Schwille, unpublished work), and by the fact that the increase in faecal calcium in GXL and GXR rats in the present study can be attributed equally to increased intestinal calcium secretion and to increased dietary calcium intake. Thus, the two reconstruction procedures appear similar with respect to intestinal calcium handling.

GX osteopenia and non-calcium factors B deficiency may be a risk factor for osteoporosis as "# well as skeletal fractures in patients with pernicious anaemia [16,17], and B therapy has been recommended "# as a stimulus for osteoblastic function [14,15]. However, # 1998 The Biochemical Society and the Medical Research Society

741

742


apart from one successfully treated patient [18], no clinical studies have been published on the effectiveness of B treatment, in particular in the context of post-GX "# osteopenia. In some studies in GX animals, B was "# supplemented, but the blood levels of B as well as "# effects on bone and mineral metabolism were not evaluated [42,43]. In the present GX models, there is no significant beneficial effect of B therapy on bone and mineral "# metabolism, in terms of serum alkaline phosphatase, urinary hydroxyproline and crosslinks, histomorphometry, density and the fracturing energy of bone. GX osteopathy as observed 4 months post operation is not therefore due to vitamin B deficiency. "# Low post-GX gastrinaemia results from removal of the major G-cell sites (antrum, proximal duodenum), while low serum calcitonin remains unexplained, since our own independent experiments failed to demonstrate a positive feedback of gastrin and calcitonin in the rat [44]. Low serum 25-hydroxyvitamin D was often attributed to intestinal malabsorption of vitamin D. However, recent work incriminates increased biliary excretion of 25hydroxyvitamin D as the underlying mechanism, resulting from interaction of high concentrations of circulating 1,25(OH) D with its hepatic receptors [35]. # Inorganic phosphate is an integral component of bone mineral, and it also affects the production of bone matrix [45]. The extreme hyperphosphaturia, common to both GXL and GXR and resistant to B treatment, may "# signal that post-GX osteopenia is a phosphorus rather than a calcium problem. Increase of urinary cAMP, due to changes in hormones capable of stimulating phosphorus excretion, can be ruled out as a causal factor [PTH is unchanged, calcitonin is decreased] ; therefore, cAMP filtered by glomeruli rather than renal adenylate cyclase stimulation should underlie the high urinary cAMP (see also below). Numerous factors can modify phosphaturia in the presence or absence of PTH [46]. Hence, subdiaphragmatic vagotomy in the rat is followed by altered renal handling of water and electrolytes [47], and renal denervation causes a decrease in tubular phosphate reabsorption ([48] ; A. Mayer, B. Berge! and P. O. Schwille, unpublished work)]. In our GX rats the subdiaphragmatic vagal trunks were cut, so that vagal afferent signals to the central nervous system and the relaying process to peripheral organs, including the kidney, are likely to have been altered. Ongoing studies of our own indicate that both injuries to the gastric vagus (Latarjet’s nerve) or sub-diaphragmatic truncal vagotomy elicit hyperphosphaturia ; in the latter case, there was also a rise in urinary cAMP. Chronic hyperphosphaturia may activate the 1α-hydroxylase and subsequently raise the blood levels of 1,25(OH) D. Phosphorus depletion# induced osteopathy has been reported [49], and high 1,25(OH) D itself can diminish bone calcium via stimu# lation of bone resorption [50], thereby attenuating # 1998 The Biochemical Society and the Medical Research Society

parathyroid gland activity. This would also explain why, in contrast to bone calcium, 1,25(OH) D is only insignif# icantly related to urinary phosphorus. Another possible cause of GX hyperphosphaturia may be that, in the absence of gastric acid production, protons from nutrients cannot be neutralized intra-intestinally, and need to be eliminated by alternative pathways. One of these may be proton buffering by bone, at the expense of bone mineral ; another may be the urine. In fact, urinary pH is low and phosphorus reabsorption declines with decreasing pH [51]. The correlations further support the idea of interrelations between GX, acid–base homoeostasis and bone minerals, because the latter, especially calcium, are inversely and significantly associated with urinary phosphorus. The hypocalciuria of GX rats was accentuated by B "# treatment. However, the role of urinary excretion in the calcium balance of the rat is minimal, and so this effect of B does not play a quantitative role. The negligible "# effect of B on extracellular calcium homoeostasis is "# reflected by its inability to alter the calcium-regulating hormones 1,25(OH) D, PTH and calcitonin.

#

ACKNOWLEDGMENTS We are grateful to Professor R. Eckstein, Division of Transfusion Medicine, University Hospital, Erlangen, who provided facilities for assessing the haematological status, to Professor E. Bonu) cci, Department of Experimental Medicine and Pathology, La Sapienza University of Rome, Italy for advice and cooperation with his institution, and to our co-workers B. Schreiber, K. Schwille and I. Goldberg for technical and secretarial work. Financial support was provided by the Catholic Academic Exchange Service, Wuppertal, Germany, to one author (A. W.) ; and the University of Erlangen Hospital Research Funds, Erlangen, Germany.

REFERENCES 1 Bisballe, S., Eriksen, E. F., Melsen, F., Mosekilde, L., Sorensen, O. H. and Hessov, I. (1991) Osteopenia and osteomalacia after gastrectomy : interrelations between biochemical markers of bone remodelling, vitamin D metabolites, and bone histomorphometry. Gut 32, 1303–1307 2 Nishimura, O., Furumoto, T., Nosaka, K. et al. (1986) Bone disorder following partial and total gastrectomy with reference to bone mineral content. Jpn. J. Surg. 16, 98–105 3 Koga, S., Tsuchiya, S. and Mutou, T. (1990) Bone disease after gastric surgery : cooperative multihospital study in Japan. Dig. Surg. 7, 19–25 4 Gertner, J. M., Lilburn, M. and Domenech, M. (1977) 25Hydroxycholecalciferol absorption in steatorrhoea and postgastrectomy osteomalacia. Br. Med. J. 1, 1310–1312 5 Imawari, M., Kozawa, K., Akanuma, Y., Koizumi, S., Itakura, H. and Kosaka, K. (1980) Serum 25-


6 7 8

9

10 11 12 13

14

15 16

17 18

19

20 21 22 23 24

25

26 27

hydroxyvitamin D and vitamin D-binding protein levels and mineral metabolism after partial and total gastrectomy. Gastroenterology 79, 255–258 Heptner, G., Domschke, S. and Domschke, W. (1989) Exocrine pancreatic function after gastrectomy. Gastroenterology 97, 147–153 Olbe, L. and Lundell, L. (1987) Intestinal function after total gastrectomy and possible consequences of gastric replacement. World J. Surg. 11, 713–719 Seetharam, B. (1994) Gastrointestinal absorption and transport of cobalamin (Vitamin B12). In Physiology of the Gastrointestinal Tract, 3rd edn. (Johnson, L. R., ed.), pp. 1997–2026, Raven Press, New York Goebell, H. (1979) Pathophysiological consequences of total gastrectomy. In Gastric Cancer (Herfarth, C. and Schlag, P., eds.), pp. 228–236, Springer Verlag, Berlin, Heidelberg, New York Scott, J. M. (1992) Folate–vitamin B12 interrelationships in the central nervous system. Proc. Nutr. Soc. 51, 219–224 Meyer, J. H. (1994) Nutritional outcomes of gastric operations. Gastroenterol. Clin. N. Am. 22, 227–260 Paulson, M. and Harvey, J. C. (1954) Hematological alterations after total gastrectomy. Evolutionary sequences over a decade. J. Am. Med. Assoc. 156, 1556–1560 Festen, H. P. M. (1991) Intrinsic factor secretion and cobalamin absorption – physiology and pathophysiology in the gastrointestinal tract. Scand. J. Gastroenterol. (Suppl.) 188, 1–7 Van Dommelen, C. K. V. and Klaassen, C. H. L. (1964) Cyanocobalamin-dependent depression of the serum alkaline phosphatase level in patients with pernicious anemia. N. Engl. J. Med. 271, 541–544 Carmel, R., Lau, K. H. W., Bayhuk, D. J., Saxena, S. and Singer, F. R. (1988) Cobalamin and osteoblast-specific proteins. N. Engl. J. Med. 319, 70–75 Goerss, J. B., Kim, C. H., Atkinson, E. J., Eastel, R., O’Fallon, W. M. and Melton, L. J. (1992) Risk of fractures in patients with pernicious anemia. J. Bone Miner. Res. 7, 73–79 Eastell, R., Vieira, N. E., Yergey, A. L. et al. (1992) Pernicious anaemia as a risk factor for osteoporosis. Clin. Sci. 82, 81–85 Melton, M. E. and Kochman, M. L. (1994) Reversal of severe osteoporosis with vitamin B12 and etidronate therapy in a patient with pernicious anemia. Metabolism 43, 68–69 Ru$ menapf, G., Schwille, P. O., Erben, R. G., Schreiber, M., Fries, W. and Hohenberger, W. (1997) Osteopenia following total gastrectomy in the rat – state of mineral metabolism and bone histomorphometry. Eur. Surg. Res. 29, 209–221 Kozawa, K., Imiwara, M., Shimazu, H., Kobori, O., Osuga, T. and Morioka, Y. (1984) Vitamin D status after total gastrectomy. Dig. Dis. Sci. 29, 411–416 Frost, H. M. and Jee, W. S. S. (1992) On the rat model of human osteopenias and osteoporoses. Bone Mineral 18, 227–236 Fries, W., Ru$ menapf, G. and Schwille, P. O. (1992) Disturbances of mineral and bone metabolism following gastric antrectomy in the rat. Bone Mineral 19, 245–256 Johnston, I. D. A., Nevin, H. O. and Welbourn, R. B. (1960) The syndrome of nutritional deficiency after gastrectomy in the rat. Br. J. Surg. 48, 81–86 Li Destri, G., Trombatore, G., La Greca, G. et al. (1992) Total gastrectomy : nutritional status after different reconstruction techniques. An experimental study. J. Surg. Oncol. 49, 98–102 Morsiani, E., Tassinati, E., Rescazzi, R. and Benea, G. (1985) Rapid intestinal transit and postgastrectomy malnutrition. An experimental study in growing rats. Ital. J. Surg. Sci. 15, 139–140 Helsingen, Jr., N. (1959) Oesophageal lesions following total gastrectomy in rats. Acta Chir. Scand. 118, 202–216 Blaurock, P., Schwille, P. O., Manoharan, M., Scheele, J., Fries, W. and Ru$ menapf, G. (1992) Effects of jejuno–ileal bypass on oxalate and mineral metabolism in rats. Eur. J. Surg. 158, 595–602

28 Berge! , B. (1995) Orchiektomie-induzierte Osteopathie bei der Ratte. Untersuchungen zur Pathophysiologie und Prophylaxe mittels Calciumsalzen organischer Sa$ uren, unter besonderer Beachtung von bildgebenden Verfahren und biomechanischer Pru$ fung des Knochens. Medical Dissertation, University of Erlangen 29 Cuisinier-Gleizes, P., George, A., Giuliano, C. and Mathieu, H. (1971) Stimulation of bone resorption by phosphorus deprivation in the rat. Israel J. Med. Sci. 7, 355 (Abstract) 30 Baylink, D., Wergedal, J. and Stauffer, M. (1971) Formation, mineralization, and resorption of bone in hypophosphatemia rats. J. Clin. Invest. 50, 2519–2550 31 Bronner, F. and Abrams, S. A. (1993) Urinary and fecal endogenous calcium excretion in the age range of 5–15 Y. Am. J. Clin. Nutr. 57, 944 (Technical note) 32 Bronner, F. (1994) Calcium and osteoporosis. Am. J. Clin. Nutr. 60, 831–836 33 Maier, G. W., Kreis, M. E., Zittel, T. T. and Becker, H. D. (1997) Calcium regulation and bone mass loss after total gastrectomy in pigs. Ann. Surg. 225, 181–192 34 Imamura, M., Yamauchi, H., Fukushima, K., Sasaki, I. and Ouchi, A. (1988) Bone metabolism following gastric surgery : microdensitometry and single-photon absorptiometry. Tohoku J. Exp. Med. 156, 237–249 35 Davies, M., Heys, S. E., Selby, P. L., Berry, J. L. and Mawer, E. B. (1997) Increased catabolism of 25hydroxyvitamin D in patients with partial gastrectomy and elevated 1,25-dihydroxyvitamin D levels. Implications for metabolic bone disease. J. Clin. Endocrinol. Metab. 82, 209–212 36 Tovey, F. I., Hall, M. L., Ell, P. J. and Hobsley, M. (1991) Postgastrectomy osteoporosis. Br. J. Surg. 78, 1335–1337 37 Johnston, I. D. A., Nevin, H. O. and Welbourn, R. B. (1960) The syndrome of nutritional deficiency after gastrectomy in the rat. Br. J. Surg. 48, 81–86 38 Ru$ menapf, G., Schwille, P. O., Steiner, G., Schreiber, M., Kissler, H. and Gall, F. P. (1994) Einfluss von Duodenalpassage und Vaguserhaltung nach totaler Gastrektomie (GX) auf den Knochenstoffwechsel der Ratte – Untersuchungen zur Pathogenese der Postgastrektomie-Knochenerkrankung. Langenbecks Arch. Chir. Suppl. Forum, 83–86 39 Miholic, J., Meyer, H. J., Mu$ ller, M. J., Weimann, A. and Pichlmayr, R. (1990) Nutritional consequences of total gastrectomy : the relationship between mode of reconstruction, postprandial symptoms, and body composition. Surgery 108, 488–494 40 Jeejeebhoy, K. N. (1993) Nutrient requirements and nutrient deficiencies in gastrointestinal diseases. In Gastrointestinal Disease, 5th edn (Sleisenger, M. H. and Fordtran, J. S., eds.), pp. 2017–2047, Saunders, Philadelphia 41 Babior, B. M. and Bunn, H. F. (1983) Megaloblastic anemias. In Harrison’s Principles of Internal Medicine, 10th edn (Petersdorf, R. G. et al., eds.), pp. 1852–1860, McGraw-Hill, New York 42 Persson, P., Gagnemo-Persson, R., Chen, D. et al. (1993) Gastrectomy causes bone loss in the rat : is lack of gastric acid responsible ? Scand. J. Gastroenterol. 28, 301–306 43 Klinge, B., Lehto-Axtelius, D., Akerman, M. and Hakanson, R. (1995) Structure of calvaria after gastrectomy. An experimental study in the rat. Scand. J. Gastroenterol. 30, 952–957 44 Ru$ menapf, G., Schwille, P. O., Erben, R. G. et al. (1998) Gastric fundectomy in the rat – Effects on mineral and bone metabolism, with emphasis on the gastrin–calcitonin–parathyroid hormone–vitamin D axis. Calcif. Tissue Int. (1998), in the press 45 Caverzasio, J., Montessuit, C. and Bonjour, J. P. (1996) Functional role of Pi transport in osteopenic cells. News Physiol. Sci. 11, 119–125 46 Murer, H. and Malmstro$ m, K. (1987) How renal phosphate transport is regulated. News Physiol. Sci. 2, 45–49 47 Reyes, A. A., Robertson, G., Jenden, D. J., Cryer, P. E., Chasalow, F. I. and Klahr, S. (1992) Subdiaphragmatic vagotomy in rats induces systemic hypertension and sodium retention. Miner. Electrol. Metab. 18, 375–381


743

744


48 Szalay, L., Bencsath, P. and Takacs, L. (1977) Effect of splanchnicotomy on the renal excretion of inorganic phosphate after renal denervation in the anaesthetized dog. Pflu$ gers Arch. 367, 283–286 49 Day, H. G. and McColum, E. V. (1939) Mineral metabolism, growth, and symptomatology of rats on a diet extremely deficient in phosphorus. J. Biol. Chem. 130, 269–274

50 Henry, H. L. and Norman, A. W. (1992) Metabolism of vitamin D. In Disorders of Bone and Mineral Metabolism (Coe, F. L. and Favus, M. J., eds.), pp. 149–162, Raven Press, New York 51 Quamme, G. A., Mizgala, C. L. and Wong, N. L. M. (1985) Effects of intraluminal pH and dietary phosphate on phosphate transport in the proximal convoluted tubule. Am. J. Physiol. 249, F759–F768 Received 8 April 1998/7 July 1998; accepted 9 July 1998
