Longitudinal changes with age in bone mineral density of patients with primary biliary cirrhosis

Longitudinal changes with age in bone mineral density of patients with primary biliary cirrhosis

Hepatology Research 10 (1998) 49 – 57 Longitudinal changes with age in bone mineral density of patients with primary biliary cirrhosis Susumu Shiomi ...

114KB Sizes 0 Downloads 16 Views

Hepatology Research 10 (1998) 49 – 57

Longitudinal changes with age in bone mineral density of patients with primary biliary cirrhosis Susumu Shiomi a,*, Kyoko Masaki a, Tadashi Takeda a, Shuhei Nishiguchi a, Daiki Habu a, Tetsuo Kuroki a, Takashi Tanaka b, Hironobu Ochi c a

Third Department of Internal Medicine, Osaka City Uni6ersity Medical School, 1 -5 -7 Asahimachi, Abeno-ku Osaka 545, Japan b Department of Public Health, Osaka City Uni6ersity Medical School, 1 -5 -7 Asahimachi, Abeno-ku, Osaka 545, Japan c Di6ision of Nuclear Medicine, Osaka City Uni6ersity Medical School, 1 -5 -7 Asahimachi, Abeno-ku, Osaka 545, Japan Received 22 August 1997; received in revised form 20 October 1997; accepted 28 October 1997

Abstract Bone disorders are associated with primary biliary cirrhosis and cirrhosis of the liver. We used dual-energy X-ray absorptiometry to study bone involvement in primary biliary cirrhosis and in cirrhosis of the liver by measurement of the bone mineral density of lumbar vertebrae in Japanese patients. Some patients were monitored to identify how bone mineral density changed with age, as a first step in the establishment of suitable treatment. The mean bone mineral density of lumbar vertebrae L2 –L4 in 150 women (64 with primary biliary cirrhosis and 86 with cirrhosis) was calculated. The bone mineral density of these vertebrae of 61 patients (31 with primary biliary cirrhosis and 30 with cirrhosis) was measured more than once during a period of 8–72 months. In women with primary biliary cirrhosis or cirrhosis of the liver, the bone mineral density tended to decrease with age faster than the reference values published elsewhere. At 50 years or more, the patients with primary biliary cirrhosis had a mean significantly lower than the reference values. At 60 years or more, the cirrhotic patients had a mean significantly lower than the reference value. The mean annual change in the mean bone mineral density was − 3.5% in women with primary biliary cirrhosis and − 2.4% in * Corresponding author. Tel.: +81 6 6452116; fax: + 81 6 6452416; e-mail: [email protected] 1386-6346/98/$19.00 © 1998 Elsevier Science Ireland Ltd. All rights reserved. PII S 1 3 8 6 - 6 3 4 6 ( 9 7 ) 0 0 1 0 7 - 1

50

S. Shiomi et al. / Hepatology Research 10 (1998) 49–57

women with cirrhosis, significantly different from the reference value ( − 1.1%). This finding suggests that treatment for prevention of bone loss is needed for older women with primary biliary cirrhosis or cirrhosis of the liver. © 1998 Elsevier Science Ireland Ltd. Keywords: Bone disease; Primary biliary cirrhosis; Cirrhosis of the liver; Dual-energy X-ray absorptiometry

1. Introduction Complications such as osteoporosis are often seen in chronic liver diseases such as chronic active hepatitis [1], alcoholic cirrhosis [2] and primary biliary cirrhosis (PBC) [3]. Perhaps the bone changes seen in some patients with PBC and cirrhosis of the liver arise in part, from a disturbance in the absorption of vitamin D, Ca, etc., because of the steatorrhea that occurs as a result of long-term intra-hepatic cholestasis [4]. However, in one study of patients with cirrhosis other than alcoholic cirrhosis [5], there was a high rate of bone complications and the cirrhosis itself was suggested to give rise to the changes. By the usual X-ray imaging methods, evaluation of the lumbar vertebrae, composed mostly of spongy bone [6], is not possible. Another disadvantage of X-ray images is that they do not give information about the bone mineral density (BMD). Dual-photon absorptiometry made it possible to measure the BMD in any part of the body [7] and dual-energy X-ray absorptiometry (DXA) is a further improvement over this radionuclide-based method [8,9]. Bone metabolism markers such as serum osteocalcin [10] and urinary deoxypyridinoline [11] are said to be indices for the evaluation of pathological bone changes. Unfortunately, their high coefficients of variation make these markers unsuitable for monitoring changes in the BMD, which varies by only several percentage points per year. Therefore, the BMD itself should be measured. We used DXA to measure the BMD in women with PBC or cirrhosis, as part of a study of the bone changes in these diseases. Some patients with PBC or cirrhosis were examined more than once to identify how the BMD changed with age, as a step in improving treatment.

2. Materials and methods

2.1. Patients The subjects were 64 women aged 32–72 years with PBC and 86 women aged 37– 76 years with cirrhosis. All patients with cirrhosis of the liver had underlying viral infection (14 were infected with hepatitis B virus and 72 were infected with hepatitis C virus). All patients were admitted to the Third Department of Internal Medicine of our hospital. The diagnoses of PBC and cirrhosis were based on

S. Shiomi et al. / Hepatology Research 10 (1998) 49–57

51

histological examination of liver specimens obtained by laparoscopy or needle biopsy done under ultrasonic guidance. None of the 150 patients had been given steroids, supplements of calcium and vitamin D, or hormone replacement therapy. If such treatment was begun during the study, values obtained later by DXA were excluded; treatment began in this way for eight patients with PBC and seven patients with cirrhosis. The characteristics of both groups of patients, including results of the indocyanine green clearance test, are shown in Table 1.

2.2. Measurement of BMD For DXA, a Hologic QDR-1000 or -2000 apparatus was used. The X-ray tube operates at excitation voltages of 70 and 140 kV. The BMD of most bone can be calculated by computer processing of the difference in the decreases in energy of the X-rays after they pass through bone or soft tissue. The BMDs of lumbar vertebrae L2, L3 and L4 were measured and the mean was used as an index of bone density. Coefficients of variation for the QDR-1000 and -2000 were 2.2 and 2.0%, with L2— L4 of five healthy subjects being measured five times for each calculation. The BMD of these lumbar vertebrae of 61 patients (31 with PBC and 30 with cirrhosis) was measured at least twice over a period of 8–72 months. In these 31 patients with PBC, 18 patients were asymptomatic and 13 were symptomatic. No report on longitudinal changes in the BMD of healthy Japanese women measured by DXA has been published. Accordingly, we reviewed cross-sectional data obtained in 622 Japanese female volunteers of various ages [12]. Only means for different decades of life are available. With this grouping of the data, the BMD changed almost linearly in the 40-, 50- and 60-year-old groups and the mean annual change for that 30-year period was − 1.1%. That study was done with a Hologic QDR-1000 apparatus; the Table 1 Characteristics of patients with PBC or cirrhosis of the liver Characteristics

PBC

Cirrhosis

P-value

Number of patients Range of ages (years) Mean age 9 S.D. (years) Varices, with and without Ascites, with and without Serum Ca (mg/dl) Serum P (mg/dl) Albumin (g/dl) Total bilirubin (mg/dl) Alkaline phosphatase (KAU) Gamma globulin (%) 104 Platelets ICG15 (%)

64 32– 72 53 911 15/49 5/59 8.96 (8.50, 9.41) 3.51 (3.02, 3.96) 3.80 (3.41, 4.12) 1.02 (0.80, 2.12) 24.5 (9.4, 34.4) 20.1 (17.3, 23.6) 18.7 (11.4, 22.8) 13.5 (8.9, 21.3)

86 37 – 76 60 98 23/63 10/76 8.92 (8.58, 9.24) 3.55 (3.20, 3.92) 3.72 (3.36, 3.92) 0.92 (0.68, 1.41) 8.2 (6.3, 12.4) 25.2 (20.3, 28.9) 9.5 (7.6, 12.8) 22.4 (16.7, 36.4)

B0.001 0.645 0.442 0.751 0.874 0.043 0.076 B0.001 B0.001 B0.001 B0.001

Values from this point are medians (with 25th and 75th percentiles). ICG15, indocyanine green retention at 15 min; and PBC, primary biliary cirrhosis.

52

S. Shiomi et al. / Hepatology Research 10 (1998) 49–57

manufacturer reports a coefficient of variation of 1.0% or less. An internalized automatic calibration system calibrates the apparatus each time it is turned on.

2.3. Statistical analysis Only means9S.D. were available for the cross-sectional results for healthy women (reference values); therefore, in comparisons of BMD in subjects with PBC or cirrhosis at different ages with that in the healthy controls, Student’s t-test (two-tailed) was used. Annual changes in BMD are expressed as medians with 25th and 75th percentiles. The significance of differences between medians was evaluated by the Mann-Whitney U-test (two-tailed). The significance of differences in rates of change was examined by the Wilcoxon rank-sum test (two-tailed). For all three tests, differences with P B 0.05 were considered to be statistically significant.

3. Results

3.1. BMD of lumbar 6ertebrae For healthy Japanese women, the mean BMD for L2 –L4 in different decades of life is highest at 30 – 49 years and is lower thereafter [12]. The BMDs for our 150 patients are compared in Fig. 1 with the means 9 S.D. for the healthy women classified by decade of life. In patients with PBC, the BMD tended to decrease with

Fig. 1. Bone mineral density (BMD) of lumbar vertebrae L2, L3 and L4 in healthy women (continuous lines; mean9 S.D., with a darker line for the mean), 64 women with primary biliary cirrhosis (closed circles) and 86 women with cirrhosis of the liver (open circles). Values for healthy women are quoted from [12].

S. Shiomi et al. / Hepatology Research 10 (1998) 49–57

53

age at more than the normal rate of decline. Up to the end of the 40’s group, significant differences in the means 9 S.D. were not found for the healthy subjects and our patients with PBC, but for subjects in their 50’s, the mean BMD in the patients with PBC (0.792 90.138 g/cm2) was significantly lower than that for the healthy subjects (0.86990.101; P B0.01). In the 60’s group as well, the mean BMD in the patients with PBC (0.652 90.167) was significantly lower than that for the healthy subjects (0.7499 0.101; P B0.001). In patients with cirrhosis of the liver, the BMD tended to decrease with age at more than the reference rate of decline. Up to the end of the 50’s group, significant differences in the means 9 S.D. were not found for the healthy subjects and our patients with cirrhosis of the liver, but for subjects in their 60’s, the mean BMD in the patients with cirrhosis (0.6929 0.100) was significantly lower than that for the healthy subjects (PB 0.01).

3.2. Changes in BMD The BMD was measured from two to five times over a period of 8–70 months for patients with PBC (Fig. 2), and from two to five times over a period of 10–72 months for patients with cirrhosis (Fig. 3). The BMD decreased in 30 of the 31 patients with PBC and it increased in the other patient (Fig. 4). The BMD decreased more in 27 of the 31 patients with PBC than the mean annual change of −1.1% in the cross-sectional study of healthy women [12] and less in the four other patients. In the large cross-sectional study, the mean annual decrease was statistically significant (P B0.001). Therefore, in our comparison of longitudinal values in

Fig. 2. Changes in the bone mineral density (BMD) of women with primary biliary cirrhosis.

54

S. Shiomi et al. / Hepatology Research 10 (1998) 49–57

Fig. 3. Changes in the bone mineral density (BMD) of women with cirrhosis of the liver.

patients with PBC, in which we wanted to know if the decrease caused by the disease was statistically significant, we corrected the PBC values by adding 1.1%. The remaining change was that due to disease. The Wilcoxon rank-sum test showed a greater decrease in the patients with PBC (P B 0.001). The mean annual change in the mean BMD of women with PBC was − 3.5%. The BMD decreased in 25 of the 30 patients with cirrhosis and increased in the five other patients (Fig. 4). Compared with the mean change in healthy women, the BMD decreased more in nine patients with cirrhosis and less in the 21 other patients. With correction of BMD values of patients with cirrhosis by the addition of 1.1%, the Wilcoxon rank-sum test showed a significantly greater decrease in the patients with cirrhosis (PB 0.05). The mean annual change in the mean BMD of these women with cirrhosis was − 2.4%. The median (25th and 75th percentiles) of annual changes in BMD was − 2.9% (−4.7 and −1.8%) in PBC and − 1.9% (− 3.9 and − 0.9%) in cirrhosis. This difference between the two groups was not significant. In patients examined longitudinally, the mean annual change in the mean BMD of women with asymptomatic PBC was − 2.6% and that of women with symptomatic PBC was −4.6%. The median (25th and 75th percentiles) of annual changes in BMD was −2.4% ( −3.3 and −1.4%) with asymptomatic PBC and −3.5% (− 6.4 and −2.1%) with symptomatic PBC. This difference between the two medians was significant (P =0.045).

S. Shiomi et al. / Hepatology Research 10 (1998) 49–57

55

Fig. 4. Annual changes in the bone mineral density of women with primary biliary cirrhosis (PBC) or cirrhosis of the liver. Horizontal lines on the bars show the 25th percentile, median and 75th percentile, from bottom to top. Dotted line, mean annual change for healthy women, quoted from [12]; closed circles, asymptomatic PBC; open circles, symptomatic PBC.

4. Discussion The BMD of patients with PBC in their 50’s was significantly lower than the reference value for healthy subjects in the same decade of life. However, changes in BMD must be studied not only cross-sectionally but also longitudinally. The mean annual bone loss was significantly greater in women with cirrhosis than in healthy women and the absolute value of this bone loss was still greater in the patients with PBC; the conclusion based on the longitudinal data was the same as that based on the cross-sectional data. The bone loss associated with PBC seemed not to be simply the result of the cirrhosis that develops in some patients with this disease. Our results taken together suggested that the bone loss in patients with PBC was slight in their 20’s (when these patients and healthy subjects had their peak BMD), but increased later. That the mean annual change in the BMD in patients with PBC or cirrhosis was not significantly different, although there were significant differences in the cross-sectional part of the study even in women not yet in middle-age, suggests that PBC causes bone loss early. Many studies of the relationship of PBC and metabolites of vitamin D have been made. In this disease, the serum concentration of 25-hydroxyvitamin D3 [25(OH)D3] decreases [13,14]. In an earlier study [15] of 21 Japanese subjects with PBC, nine of whom are reported here as well, correlation between the BMD and the serum level of 25(OH)D3 was significant. Of those 21 patients with PBC, nine had abnormally low levels of 25(OH)D3, so the bone changes in PBC may be

56

S. Shiomi et al. / Hepatology Research 10 (1998) 49–57

related to vitamin levels. There are reports that 25(OH)D3 is not much decreased in PBC [16,17] and another report that it does decrease, but that the administration of vitamin D, which restores the level of 25(OH)D3 to normal, does not stop the progress of the bone disease [13]. Hodgson et al. [18] found that PBC is associated with low serum osteocalcin concentrations and with a significantly lower BMD than normal. They did not detect significant changes in the number of osteoblasts, which were histologically normal, as well and they concluded that osteoblast function is reduced, accounting for the decrease in bone formation that they detected [18]. Other authors did a histological study in which they found accelerated bone resorption in patients with PBC [19]. We referred to our results showing the mean annual change in PBC in decisionmaking about when to treat these patients with calcium or vitamin D products. Older women with PBC or cirrhosis of the liver were treated to help prevent bone loss. Current treatments for bone loss are still controversial, but in the evaluation of new treatments, this information about changes in the natural course of the bone disorder in PBC and cirrhosis of the liver may be of use.

References [1] Kato Y, Epstein O, Dick R, Sherlock S. Radiological patterns of cortical bone modeling in women with chronic liver disease. Clin Radiol 1982;33:313 – 7. [2] Long RG, Meinhard E, Skinner RK, Varghese Z, Wills MR, Sherlock S. Clinical, biochemical, and histological studies of osteomalacia, osteoporosis, and parathyroid function in chronic liver disease. Gut 1978;19:85–90. [3] Atkinson M, Nordin BEC, Sherlock S. Malabsorption and bone disease in prolonged obstructive jaundice. Q J Med 1956;25:299–312. [4] Kehayoglou AK, Holdsworth CD, Agnew JE, Whelton MJ, Sherlock S. Bone disease and calcium absorption in primary biliary cirrhosis with special reference to vitamin-D therapy. Lancet 1968;i:715–9. [5] Nakano A, Kanda T, Miyamoto T, Ishigami Y, Sato T, Shimizu Y. A study of osteopenia in liver cirrhosis by dual energy X-ray absorptiometry (DXA). Jpn J Gastroenterol 1993;90:1689 – 94. [6] Genant HK, Faulkner KG, Glher C-C. Measurement of bone mineral density: current status. Am J Med 1991;91:49S–53. [7] LeBlanc AD, Evans HJ, Marsh C, Schneider V, Johnson PC, Jhingran GJ. Precision of dual photon absorptiometry measurements. J Nucl Med 1986;27:1362 – 5. [8] Cullum ID, Ell PJ, Ryder JP. X-ray dual-photon absorptiometry: a new method for the measurement of bone density. Br J Radiol 1989;62:587 – 92. [9] Mazess RB, Collick B, Trempe J, Barden H, Hanson J. Performance evaluation of a dual- energy X-ray bone densitometer. Calcif Tissue Int 1989;44:228 – 32. [10] Brown JP, Delmas PD, Malaval L, Edouard C, Chapuy MC, Meunier PJ. Serum bone Gla-protein: a specific marker for bone formation in postmenopausal osteoporosis. Lancet 1984;i:1091 – 3. [11] Uebelhart D, Gineyts E, Chapuy M-C, Delmas PD. Urinary excretion of pyridinium crosslinks: a new marker of bone resorption in metabolic bone disease. Bone Miner 1990;8:87 – 96. [12] Kushida K, Inoue T, Sumi Y, et al. Osteoporosis: bone mineral measurement using DEXA. Jpn J Clin Med 1990;48:2845–50. [13] Herlong HF, Recker RR, Maddrey WC. Bone disease in primary biliary cirrhosis: histologic features and response to 25-hydroxyvitamin D. Gastroenterology 1982;83:103 – 8. [14] Compston JE, Crowe JP, Wells IP, et al. Vitamin D prophylaxis and osteomalacia in chronic cholestatic liver disease. Dig Dis Sci 1980;25:28 – 32.

S. Shiomi et al. / Hepatology Research 10 (1998) 49–57

57

[15] Shiomi S, Kuroki T, Masaki K, et al. Osteopenia in primary biliary cirrhosis and cirrhosis of the liver in women, evaluated by dual-energy X-ray absorptiometry. J Gastroenterol 1994;26:605 – 9. [16] Danielsson A, Lorentzon R, Larsson S-E. Intestinal absorption and 25-hydroxylation of vitamin D in patients with primary biliary cirrhosis. Scand J Gastroenterol 1982;17:349 – 55. [17] Cuthbert JA, Park CYC, Zerwekh JE, Glass KD, Combes B. Bone disease in primary biliary cirrhosis: increased bone resorption and turnover in the absence of osteoporosis or osteomalacia. Hepatology 1984;4:1–8. [18] Hodgson SF, Dickson ER, Wahner HW, Johnson KA, Mann KG, Riggs BL. Bone loss and reduced osteoblast function in primary biliary cirrhosis. Ann Intern Med 1985;103:855 – 60. [19] Mitchison HC, Malcolm AJ, Bassendine MF, James OFW. Metabolic bone disease in primary biliary cirrhosis at presentation. Gastroenterology 1988;94:463 – 70.

.