| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Endocrinological Oncology |
University of Manchester Bone Disease Research Centre, Department of Medicine, Manchester Royal Infirmary (E.B.M., M.D.), Departments of Surgery and Medical Oncology (J.W., A.H., N.J.B.), University Hospital of South Manchester, Manchester, United Kingdom; and Wolfson Research Laboratories, Queen Elizabeth Medical Centre (W.A.R.), Birmingham, United Kingdom
Address all correspondence and requests for reprints to: Nigel J. Bundred, Department of Surgery, University Hospital of South Manchester, Nell Lane, Manchester, M20 8LR, United Kingdom.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
In a 6-month longitudinal study of normocalcemic patients with bone metastases undergoing hormonal therapy, serum 1,25-(OH)2D concentrations fell in patients whose disease progressed (P = 0.0056), but remained constant in those who were stable or responded to treatment. These changes in 1,25-(OH)2D preceded clinical signs of progression and predicted disease response. In the progressive group, five of whom died during the study, 1,25-(OH)2D decreased between the initial and final samples, PTH fell significantly from 24.8 to 13.5 ng/L (P = 0.025), serum calcium rose from 2.27 to 2.39 mmol/L (P = 0.017), and the urinary calcium/creatinine ratio rose from 0.37 to 0.68 (P = 0.046). PTH and 1,25-(OH)2D were significantly correlated in the final samples from this group, Spearman’s rank correlation = 0.80, P = 0.022. The results indicate that normocalcemia in these patients is maintained, at the expense of suppressing PTH and 1,25-(OH)2D, in the face of increased calcium released from lytic lesions in bone. Loss of the antiproliferative effects of 1,25-(OH)2D may then permit more rapid secondary growth of the tumor.
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
We investigated 129 women with breast cancer in a cross-sectional study; 88 had operable early breast cancers (EBC), 29 had bone metastases and normal serum calcium (BM), 12 had bone metastases with hypercalcemia (HC). In a separate, prospective study, 26 normocalcemic women with BM were followed at monthly intervals for 6 months to assess changes in hormone levels over the time course of the disease. Initially all patients were either on no treatment or adjuvant tamoxifen therapy and were subsequently changed to tamoxifen or megesterol acetate when they presented with advanced disease. Response in bone was defined according to Union Internationale Controle de Cancer (UICC) criteria. Evaluation of radiological response of measurable skeletal metastases was performed at 3-month intervals after presentation and at the time of progression. Patients were then classed as stable (nonprogressive disease) or progressive. Blood samples were taken as part of the routine investigation of the patients, and a fasting second void urine sample was collected. Consent for the study was given by the University Hospital of South Manchester Ethical Committee, and samples were taken with the informed consent of the patients after they had received a full explanation. Serum was analyzed for vitamin D metabolites, PTH, calcium, creatinine, albumin, and phosphate; plasma for PTHrP; and urine for calcium and creatinine. Hypercalcemia was defined as a serum calcium >2.6 mmol/L when adjusted for serum albumin. Urinary calcium excretion was expressed as the molar ratio of calcium to creatinine (normal = < 0.45).
PTHrP was measured as PTHrP 1–86 using an established two-site
immunoradiometric assay [normal = limit of detection 
0.23
pmol/L (12)]. PTH 1–84 was measured by immunoradiometric assay using
a Nichols Institute Allegro kit (Saffron Walden, U.K.), reference range
10–60 ng/L. Vitamin D metabolites were assayed by established in-house
methods after extraction and high performance liquid chromatography
separation; 25-hydroxyvitamin D, (25OHD) was quantitated by ultraviolet
absorbance during a second high performance liquid chromatography run
(reference range 10.8–58.5 nmol/L), inter- and intraassay coefficients
of variation 8.7% and 6.5%, respectively (13);
1,25-(OH)2D was measured by RIA using monoclonal antibody
5F2 (reference range 48–120 pmol/L), inter-and intraassay coefficients
of variation 10.7% and 7.8%, respectively, (14). As far as was
practicable, samples from any one patient were measured on the same
assay. Bone metastases were diagnosed by a positive bone scan or plain
x-rays.
Statistical analysis
Results were expressed as the mean ± SE or as median (range) for data that were not normally distributed. Statistical analysis was undertaken using Instat, instant statistics software (Graphpad, San Diego, CA). The statistical significance of differences between groups were assessed as appropriate by Student’s paired or unpaired t test or by the Mann-Whitney or Wilcoxon tests. Differences in the concentration-time curves in the longitudinal study were analyzed by the Kruskal-Wallis nonparametric ANOVA test. Associations between variables were examined using Pearson’s correlation for parametric data and Spearman’s rank correlation (rs) for nonparametric data.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Mean 1,25-(OH)2D levels fell with increasing severity
of disease, being 102 ± 3.7 pmol/L for EBC, 52 ± 5.3 pmol/L
for BM, and 33 ± 5.6 pmol/L for HC (Fig. 1
). Of
the 88 patients with operable breast cancer, 21 (24%) had raised
1,25-(OH)2D levels. Median 25OHD levels were not
significantly different between the various groups, measuring 37.5
nmol/L (range, 5–118) in EBC, 48 nmol/L (range, 20–115) in BM, and 35
nmol/L (range, 8–100) in HC. Plasma PTHrP was detectable, but only at
very low levels in 8 of the 88 women (9%) with EBC (median 0.5 pmol/L;
range; 0.37–0.78) and in 1 of the 29 with BM (1.20 pmol/L). PTHrP was
present in 11 of the 12 patients (92%) with HC (median 2.14 pmol/L;
range, 0.46–20.74). Neither PTHrP nor 1,25-(OH)2D levels
correlated with any routine prognostic markers in patients with EBC,
and there was no relationship between 1,25-(OH)2D levels
and serum calcium or urinary calcium excretion. Serum calcium values
were normal in those EBC patients with raised 1,25-(OH)2D.
PTH concentrations did not differ between EBC patients with raised
1,25-(OH)2D (25.2; range, 17–47 ng/L) and those with
normal 1,25-(OH)2D (28.6; range, 6–56 ng/L).
|
UICC assessment demonstrated disease progression in 13 of the
women, 5 of whom died between 3 and 4 months; the remaining 8 patients
had a change of hormonal treatment at 3 months, and 2 of these
subsequently responded. In this progressive group there was a
significant fall from the initial 1,25-(OH)2D levels
(P = 0.0056) (Fig. 2). The apparent
rebound in the mean seen after 3 months (Fig. 2) reflects the deaths of
patients with the lowest final 1,25-(OH)2D levels (0, 17,
22, 26 and 41 pmol/L) plus a contribution from the 2 patients in this
group who responded to treatment and whose 1,25-(OH)2D
levels increased (Fig. 2). Serum PTHrP (1.20 pmol/L) was detected in
the final sample from 1 woman. In contrast, 13 women with BM whose
disease responded or remained stable showed no significant changes in
serum 1,25-(OH)2D (Fig. 2), and PTHrP was undetectable.
Values for serum and urine biochemistry at 0 and 2 months of treatment
are shown in Table 1. The fall in serum
1,25-(OH)2D seen within the progressive group from 108
± 14 to 56 ± 8 pmol/L was highly significant at 2 months
(P = 0.002), and preceded any clinical evidence of
disease progression by UICC criteria. Concentrations of serum
1,25-(OH)2D at 2 months in this group were significantly
lower than those for the stable group (P = 0.031)
(Table 1). The fall was independent of changes in serum 25OHD levels,
and there were no significant differences in 25OHD levels between or
within groups; stable group, initial: 40.1 nmol/L (range, 16.0–90.8),
2 months: 40.0 nmol/L (range, 7.7–78.0); progressive, initial: 40.8
nmol/L (range, 6.0–93.3), 2 months: 43.0 nmol/L (range, 14.0–90.3).
One value for 25OHD in each group was subnormal, but these did not
relate to the low 1,25-(OH)2D levels. The fall in
1,25-(OH)2D remained significant (P =
0.009) even when data from the women who subsequently died were
excluded, indicating that the group was homogeneous. Serum calcium rose
significantly between 0 and 2 months from 2.27 ± 0.02 to
2.36 ± 0.02 mmol/L (P = 0.005), in the
progressive group, and the urinary calcium/creatinine ratio increased
but nonsignificantly, from 0.37 ± 0.09 to 0.54 ± 0.18
(Table 1). The rise in serum creatinine was not significant in either
group.
|
|
shows mean data for the initial and terminal
samples from the two groups. The time of the terminal sample was 6
months for the stable group, but varied from 3–6 months for the
progressive group. In the latter group, serum calcium had risen further
(compared with 2-month data in Table 1
) to 2.39 ± 0.03 mmol/L
(P = 0.017), and urinary calcium excretion continued to
rise to 0.68, the difference now becoming significant
(P = 0.046). Serum PTH fell in the progressive group
from 24.8 ± 2.9 to 13.5 ± 4.5 ng/L (P =
0.025) (Table 2). Values for serum calcium and inorganic phosphate were
all within the normal ranges, and there were no significant differences
initially between the groups, however the final value for inorganic
phosphate in the progressive group, was higher than in the stable group
(P = 0.031) (Table 2). No significant correlations
could be shown between the different parameters in Table 1 at either of
the time points studied with the exception of urinary calcium excretion
and serum 1,25-(OH)2D in the initial samples from the
stable group, r = -0.608 (P = 0.027). The same
correlation in the progressive group at 2 months just failed to reach
significance, r = -0.56 (P = 0.059). Serum
1,25-(OH)2D could not be related to serum calcium in either
group; however the final data (Table 2) for serum
1,25-(OH)2D and PTH in the progressive group were
significantly related, rs = 0.80 (P =
0.022).
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The cross-sectional study showed that serum 1,25-(OH)2D levels fell with increasing severity of the disease. The reason for the small number of EBC patients with elevated 1,25-(OH)2D levels is not understood, because PTH was not elevated. Several nonrenal cells can synthesize 1,25(OH)2D, including cancer cells (15, 16); it is possible that some breast cancer cells may also have this ability, and production of 1,25-(OH)2D has been reported in the rat Walker 256 mammary carcinoma (17). In hypercalcemic patients in the cross-sectional study, PTHrP was detectable, but mean serum 1,25-(OH)2D was decreased, with the majority of values being subnormal. Normal or increased 1,25-(OH)2D was associated with undetectable PTHrP in patients with EBC or responsive metastases, so we could find no evidence in this study that PTHrP stimulates 1,25-(OH)2D synthesis. Indeed, the appearance of PTHrP in serum was associated with a lowering of the 1,25-(OH)2D.
The longitudinal study showed that the group of normocalcemic women with bone metastases could be divided into two subgroups. In one there was a sequential fall in 1,25-(OH)2D as bone metastases progressed, with subnormal values being recorded in several patients. In the other, consisting of patients with stable or responding metastases, 1,25-(OH)2D remained normal or increased.
The reason for the low levels of the vitamin D hormone in patients with
progressive skeletal disease is most likely to be decreased stimulation
of the renal 1
-hydroxylase by the lowered concentration of PTH,
which is the trophic hormone for the reaction. Alternatively, it is
possible that, at least in some cases, cancer cells may produce an
inhibitor of the 1-hydroxylase. From our data, we conclude that
PTHrP is unlikely to function as an inhibitor, and that it is possible
that some other hormonally secreted factor may be produced by the tumor
cells that inhibits the renal 1-hydroxylase, as has been reported in
a nude rat model of human uterine cancer (18). An analogous situation
in which tumors are believed to secrete a factor that inhibits the
1-hydroxylase is oncogenous osteomalacia. In this condition grossly
suppressed serum levels of 1,25-OH)2D are found in
conjunction with normal serum calcium (19). The most likely
explanation, though, can be deduced from the data in Tables 1 and 2. In
those tables, we show that three of the parameters known to regulate
renal synthesis of 1,25-(OH)2D change in directions that
would tend to suppress that reaction, namely increases in serum calcium
and phosphate and a decrease in serum PTH. Although we were unable to
demonstrate significant correlations between serum calcium or inorganic
phosphate and 1,25-(OH)2D levels, the trend of these
results suggests that the primary event is release of calcium and
phosphate from lytic lesions arising from bone metastases. This calcium
increases the renal filtered load and serum calcium rises slowly within
the normal range. Serum PTH secretion is decreased and
1,25-(OH)2D synthesis down-regulated by lack of PTH and
possibly also directly by rising calcium and phosphate. This hypothesis
is supported by the significant relationship observed between PTH and
1,25-(OH)2D in the progressive group.
There are, however, individual patients within the progressive group
(e.g. patients 1, 11, and 12; Table 1) for whom it is
difficult to explain the results on this basis, leaving open the
possibility that a different mechanism may operate in some cases.
Assessment of patients with malignancy-associated hypercalcemia treated
with bisphosphonates has shown variation in the 1,25-(OH)2D
and PTH response to lowering the serum calcium (20). In about one third
of patients, 1,25-(OH)2D remained low despite the fall in
calcium. This suggests that factors other than calcium may be
regulating 1,25-(OH)2D levels.
The absence of circulating PTHrP in the normocalcemic patients reported in this study suggests that bone resorption is more likely to have arisen from local invasion with lysis. A role for PTHrP cannot be wholly excluded because a recent report has described an animal model of metastatic breast cancer in which PTHrP is produced locally in bone metastases without increasing circulating PTHrP or producing hypercalcemia (21). It is clear though that PTHrP could not be exerting effects on renal handling of calcium in our normocalcemic patients.
There are various ways in which lowered levels of 1,25-(OH)2D could influence the progress of the disease. 1,25-(OH)2D might itself control PTHrP levels. It has been reported that the hormone decreases steady state PTHrP messenger RNA levels in a neuroendocrine cell line (22) and in normal breast epithelial cells (23); thus, maintenance of normal concentrations of 1,25-(OH)2D may inhibit PTHrP synthesis.
Maintenance of adequate levels of vitamin D metabolites through sunlight exposure is associated epidemiologically with reduced incidence and morbidity of breast cancer (24). This may be effected by the fundamental effects of 1,25-(OH)2D on cell growth. 1,25-(OH)2D inhibits proliferation of breast cancer cells in vitro and in animal models in vivo (2, 25, 26). The findings in our study of decreasing levels of serum 1,25-(OH)2D in women with progressing breast cancer support the hypothesis that human breast cancer growth in vivo may be inhibited by high 1,25-(OH)2D levels. In addition to its effects on tumor cell proliferation and differentiation, the hormone also has potent inhibitory effects on angiogenesis in embryos and in transgenic murine retinoblastoma (27, 28), and therefore low serum levels of 1,25-(OH)2D may lead to increased, uncontrolled angiogenesis with the progression of breast cancer.
Changes in serum 1,25-(OH)2D have not previously been reported in detail during the prehypercalcemic stage of breast cancer with skeletal involvement. The fall in 1,25-(OH)2D may identify patients who require a change in therapy at an early stage and may be a useful marker of disease response. The present study indicates that the decline in 1,25-(OH)2D levels precedes clinical deterioration as judged by UICC criteria by up to 3 months. If serum 1,25-(OH)2D does not return to physiological levels following a change in hormonal therapy or chemotherapy, disease stabilization does not occur. Long-term maintenance of high levels of serum 1,25-(OH)2D by treatment with the hormone is not an option because of the potential toxicity of this approach. However studies in vitro with analogs of 1,25-(OH)2D, which have little calcemic effect, are producing promising results (29), and the rationale for developing such compounds as therapeutic agents is supported by the findings of our study.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received September 21, 1995.
Revised May 10, 1996.
Accepted August 27, 1996.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  D. FISCHER, S. BECKER, T. CORDES, B. BUCKER, K. DIEDRICH, M. FRIEDRICH, D. SALEHIN, and M. THILL Vitamin D-24-Hydroxylase in Benign and Malignant Breast Tissue and Cell Lines Anticancer Res, September 1, 2009; 29(9): 3641 - 3645. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. SIMMONS, E. AMIR, G. DRANITSARIS, M. CLEMONS, B. WONG, R. VEITH, and D. E.C. COLE Altered Calcium Metabolism in Patients on Long-term Bisphosphonate Therapy for Metastatic Breast Cancer Anticancer Res, July 1, 2009; 29(7): 2707 - 2711. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Wong, A. Hayen, J. R. Chapman, A. C. Webster, J. J. Wang, P. Mitchell, and J. C. Craig Association of CKD and Cancer Risk in Older People J. Am. Soc. Nephrol., June 1, 2009; 20(6): 1341 - 1350. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Dhawan, R. Weider, and S. Christakos CCAAT Enhancer-binding Protein {alpha} Is a Molecular Target of 1,25-Dihydroxyvitamin D3 in MCF-7 Breast Cancer Cells J. Biol. Chem., January 30, 2009; 284(5): 3086 - 3095. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. R. Perez-Lopez, P. Chedraui, and J. Haya Review Article: Vitamin D Acquisition and Breast Cancer Risk Reproductive Sciences, January 1, 2009; 16(1): 7 - 19. [Abstract] [PDF]
|
|
|
|
 |
|
 T. M. Chlon, D. A. Taffany, J. Welsh, and M. J. Rowling Retinoids Modulate Expression of the Endocytic Partners Megalin, Cubilin, and Disabled-2 and Uptake of Vitamin D-Binding Protein in Human Mammary Cells J. Nutr., July 1, 2008; 138(7): 1323 - 1328. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. E. Valrance, A. H. Brunet, and J. Welsh Vitamin D Receptor-Dependent Inhibition of Mammary Tumor Growth by EB1089 and Ultraviolet Radiation in Vivo Endocrinology, October 1, 2007; 148(10): 4887 - 4894. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C Palmieri, T MacGregor, S Girgis, and D Vigushin Serum 25-hydroxyvitamin D levels in early and advanced breast cancer J. Clin. Pathol., December 1, 2006; 59(12): 1334 - 1336. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Cui and T. E. Rohan Vitamin d, calcium, and breast cancer risk: a review. Cancer Epidemiol. Biomarkers Prev., August 1, 2006; 15(8): 1427 - 1437. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. F. Garland, F. C. Garland, E. D. Gorham, M. Lipkin, H. Newmark, S. B. Mohr, and M. F. Holick The Role of Vitamin D in Cancer Prevention Am J Public Health, February 1, 2006; 96(2): 252 - 261. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Nagpal, S. Na, and R. Rathnachalam Noncalcemic Actions of Vitamin D Receptor Ligands Endocr. Rev., August 1, 2005; 26(5): 662 - 687. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Townsend, C. M. Banwell, M. Guy, K. W. Colston, J. L. Mansi, P. M. Stewart, M. J. Campbell, and M. Hewison Autocrine Metabolism of Vitamin D in Normal and Malignant Breast Tissue Clin. Cancer Res., May 1, 2005; 11(9): 3579 - 3586. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Welsh Vitamin D and breast cancer: insights from animal models Am. J. Clinical Nutrition, December 1, 2004; 80(6): 1721S - 1724S. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. M. Zinser and J. Welsh Vitamin D receptor status alters mammary gland morphology and tumorigenesis in MMTV-neu mice Carcinogenesis, December 1, 2004; 25(12): 2361 - 2372. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Welsh, J. A. Wietzke, G. M. Zinser, B. Byrne, K. Smith, and C. J. Narvaez Vitamin D-3 Receptor as a Target for Breast Cancer Prevention J. Nutr., July 1, 2003; 133(7): 2425S - 2433. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C S Haworth, P L Selby, A W Horrocks, E B Mawer, J E Adams, and A K Webb A prospective study of change in bone mineral density over one year in adults with cystic fibrosis Thorax, August 1, 2002; 57(8): 719 - 723. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. A. Moffatt, W. U. Johannes, T. E. Hedlund, and G. J. Miller Growth Inhibitory Effects of 1{alpha}, 25-Dihydroxyvitamin D3 Are Mediated by Increased Levels of p21 in the Prostatic Carcinoma Cell Line ALVA-31 Cancer Res., October 1, 2001; 61(19): 7122 - 7129. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. T. WARD, S. K. YAU, A. P. MEE, E. B. MAWER, C. A. MILLER, H. O. GARLAND, and D. RICCARDI Functional, Molecular, and Biochemical Characterization of Streptozotocin-Induced Diabetes J. Am. Soc. Nephrol., April 1, 2001; 12(4): 779 - 790. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. J. Mantell, P. E. Owens, N. J. Bundred, E. B. Mawer, and A. E. Canfield 1{alpha},25-Dihydroxyvitamin D3 Inhibits Angiogenesis In Vitro and In Vivo Circ. Res., August 4, 2000; 87(3): 214 - 220. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C S Haworth, P L Selby, A K Webb, M E Dodd, H Musson, R M. Niven, G Economou, A W Horrocks, A J Freemont, E B Mawer, et al. Low bone mineral density in adults with cystic fibrosis Thorax, November 1, 1999; 54(11): 961 - 967. [Abstract] [Full Text]
|
|
|
|
|
|
E. M. John, G. G. Schwartz, D. M. Dreon, and J. Koo Vitamin D and Breast Cancer Risk: The NHANES I Epidemiologic Follow-up Study, 1971-1975 to 1992 Cancer Epidemiol. Biomarkers Prev., May 1, 1999; 8(5): 399 - 406. [Abstract] [Full Text]
|
|
|
|
 |
|
 C. J. Haug, P. Aukrust, E. Haug, L. Mørkrid, F. Müller, and S. S. Frøland Severe Deficiency of 1,25-Dihydroxyvitamin D3 in Human Immunodeficiency Virus Infection: Association with Immunological Hyperactivity and Only Minor Changes in Calcium Homeostasis J. Clin. Endocrinol. Metab., November 1, 1998; 83(11): 3832 - 3838. [Abstract] [Full Text]
|
|
|
|
|
|
E. B. Mawer 1,25 Dihydroxyvitamin D and Cancer--Author's Response J. Clin. Endocrinol. Metab., October 1, 1997; 82(10): 3517 - 3517. [Full Text] [PDF]
|
|
|
|
|
|
G. M. A. Palmieri 1,25 Dihydroxyvitamin D and Cancer J. Clin. Endocrinol. Metab., October 1, 1997; 82(10): 3516a - 3517. [Full Text]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
(highly significant fall with time,
P = 0.0056). The rebound in the nonresponsive group
after 3 months reflects the deaths of 5 women (shown as
+) who had the lowest 1,25-(OH)2D levels.
Horizontal lines show upper and lower limits of
reference range (mean ± 2 SD). Serum
1,25-(OH)2D levels did not differ between the groups, and
no patient was hypercalcemic.