Under physiologic conditions, the kidney, under the control of parathyroid hormone, is the sole source of circulating calcitriol [11]. Hypercalcemia inhibits parathyroid hormone production and thereby diminishes renal calcitriol synthesis [11, 12]. An elevated serum calcium level itself also inhibits renal calcitriol production [13], as does hyperphosphatemia [11]. Unless an autonomous source of parathyroid hormone is present, as in primary hyperparathyroidism [14, 15], serum calcitriol levels are suppressed during hypercalcemia; a failure to observe such suppression is evidence of disordered regulation of calcitriol synthesis and usually indicates extrarenal production. Such extrarenal calcitriol production causes hypercalcemia in sarcoidosis, in which activated macrophages within the granuloma synthesize calcitriol [16, 17].

The available reference ranges for serum calcitriol levels are derived from normocalcemic persons, and the appropriate serum level in the setting of hypercalcemia is unknown. Thus, recognizing abnormal calcitriol regulation in the absence of a frank elevation of the serum level is problematic. To better determine the incidence of abnormal calcitriol production during hypercalcemia that complicates non-Hodgkin lymphoma, we studied hypercalcemic patients with myeloma as a control group. Healthy persons rendered severely hypercalcemic by calcium infusion also would have been appropriate, but such study is not ethically possible. However, neither patients with primary hyperparathyroidism nor those with hypercalcemia that complicates solid tumors are suitable as controls. In both of these disorders, extraneous factors substantially influence calcitriol metabolism. In primary hyperparathyroidism, the elevated serum level of parathyroid hormone enhances calcitriol production [14, 15]. In patients with hypercalcemia that complicates solid tumors, the serum level of the novel hormone parathyroid hormone-related protein is often elevated [18-21]. Parathyroid hormone-related protein, like parathyroid hormone, can increase renal calcitriol production [22, 23]. Despite this capacity, serum calcitriol levels are rarely elevated in these patients [24], possibly because of a tumor-derived humoral factor that inhibits renal calcitriol production [25]. In patients with myeloma, however, it is known that calcitriol levels are suppressed during episodes of hypercalcemia and return to normal when hypercalcemia resolves after bisphosphonate treatment [26].

Our study had two aims: 1) to establish a reference range for serum calcitriol during hypercalcemia in the control group and, using this reference range, to determine the frequency of elevated serum levels of calcitriol in hypercalcemic patients with non-Hodgkin lymphoma and 2) to investigate the frequency of abnormal calcium metabolism in normocalcemic patients with non-Hodgkin lymphoma.

Methods

Top Methods Results Discussion Author & Article Info References

Patients

From July 1992 to May 1993, we determined the serum calcium levels in patients with lymphoma and myeloma to identify those who had non-Hodgkin lymphoma or myeloma and hypercalcemia. During this time, approximately 345 new patients were seen in the lymphoma clinic and 90 were seen in the myeloma clinic, although both newly diagnosed patients and patients with relapse were eligible for this study. Hypercalcemia was defined as a corrected serum calcium level of 11.0 mg/dL (2.75 mmol/L) or greater, in which corrected serum calcium (mg/dL) = total serum calcium (mg/dL) + 0.8(4.0 –serum albumin [g/dL]) [27]. Corrected serum calcium in mg/dL can be converted to mmol/L by multiplying by 0.25.

We identified 28 patients with non-Hodgkin lymphoma and hypercalcemia and excluded 6 from the study for the following reasons: One patient had a B-cell, diffuse large-cell lymphoma and a corrected serum calcium level of 2.94 mmol/L (11.8 mg/dL) and could not be studied; 3 patients had elevated serum levels of intact parathyroid hormone (144, 63, and 221 pg/mL, respectively), suggesting primary hyperparathyroidism; 1 patient had idiopathic absorptive hypercalciuria; and 1 patient had hypercalcemia caused by human T-cell lymphotrophic virus type 1-related T-cell leukemia-lymphoma, in which parathyroid hormone-related protein is known to mediate hypercalcemia [28]. The remaining 22 patients were studied within 48 hours of the documentation of hypercalcemia and before corticosteroids or cytotoxic drugs were administered. Patients usually received intravenous hydration during this interval, which may explain why 5 patients had corrected serum calcium values of less than 2.74 mmol/L (11.0 mg/dL) at the time of full evaluation. One patient had received bisphosphonate treatment one day before the study but was included because serum levels of parathyroid hormone, parathyroid hormone-related protein, and calcitriol are known not to be influenced by bisphosphonate therapy during this period [29-31]. We excluded this patient from the analysis of urinary values because these may change rapidly after therapy with bisphosphonates [32].

We applied identical inclusion criteria for the control patients with myeloma. We identified 26 hypercalcemic patients and evaluated 17. One patient was excluded because of an elevated serum intact parathyroid hormone level (56 pg/mL); 16 patients thus remained for further study. All patients provided informed consent for investigational studies in accordance with institutional guidelines, no patients had preexisting chronic renal impairment (with a serum creatinine level of 132.6 mmol/L [1.5 mg/dL]), and none reported receiving vitamin D supplements, lithium, or thiazide diuretics before the study.

Laboratory Evaluation

We determined serum electrolyte, albumin, and creatinine levels using an automated analyzer (Kodak, Rochester, New York). Urine was collected as a random fasting specimen, and urinary calcium excretion was calculated as the product of urinary calcium and serum creatinine levels divided by the urinary creatinine level, where all values are in mg/dL with a normal range of less than 0.14 mg/dL (<0.035 mmol/L) of glomerular filtrate [33]. We calculated the renal tubular phosphate threshold (TmPO₄^3-), an index of phosphate resorption, using the method of Walton and Bijvoet [34]. The reference range for TmPO₄^3- is 2.5 to 4.2 mg/dL (0.80 to 1.35 mmol/L) of glomerular filtrate; a lower value indicates a greater degree of phosphaturia [34]. We measured serum levels of intact parathyroid hormone using a two-site chemiluminometric immunoassay (CIBA-Corning Diagnostics, Medfield, Massachusetts), with a normal range of 11 to 54 pg/mL and a lower limit of detection of 3 pg/mL. We extracted calcitriol from serum, purified it as described [35], and measured it using the chicken intestine cytosolic receptor competitive binding assay with tritium-labelled calcitriol standards of known concentration (SmithKline Beecham, Houston, Texas). The reference range for serum calcitriol during normocalcemia is 20 to 76 pg/mL; the assay used has a lower limit of detection of 5 pg/mL and a coefficient of variation of 15% at concentrations of 26 to 90 pg/mL. We measured serum 25-hydroxyvitamin D levels using a competitive protein-binding assay (SmithKline Beecham); the reference range is 10 to 55 ng/mL (25 to 138 nmol/L). Parathyroid hormone-related protein in serum was measured using an immunoradiometric assay (Nichols Institute, San Juan, California). The assay uses a solid-phase antiparathyroid hormone-related protein^37-74 antibody and Iodine-125-labeled antiparathyroid hormone-related protein^1-36 as tracer and has no cross-reactivity with parathyroid hormone; it has a reference range of less than 1.5 pmol/L and a lower limit of detection of 0.2 pmol/L [21].

Statistical Analysis

We compared categorical data using the chi-square test or the Fisher Exact test and ordinal data using the Wilcoxon two-sample test or the Kruskal-Wallis test as appropriate. Sequential values were compared using the paired student t-test. We evaluated correlations using the Spearman rank correlation coefficient and compared them on the basis of the slope determined from the linear regression analysis using the least-squares method. All P values are two-sided.

Results

Top Methods Results Discussion Author & Article Info References

Control Group

The median age of the hypercalcemic patients with myeloma was 60 years (range, 37 to 79 years); nine (56%) were male, and all had lytic bone lesions. Their mean serum calcitriol level was 11.6 ±10.0 pg/mL, and the highest level observed was 32 pg/mL (Figure 1). We defined the upper limit of expected serum calcitriol in hypercalcemia as the mean plus 3 standard deviations of the serum calcitriol levels measured in the control patients (<42 pg/mL). No patient in the control group had an elevated serum level of parathyroid hormone-related protein (median, 0.2 pmol/L; range, <0.2 to 1.5 pmol/L). Other biochemical features are given in Figure 1 and Table 1.

View larger version (28K): [in this window] [in a new window]   Figure 1. Distribution of serum levels of parathyroid hormone-related protein (PTHrP), intact parathyroid hormone (PTH), and calcitriol and corrected serum calcium of control patients with multiple myeloma (right column) and of pa-tients with non-Hodgkin lymphoma and either elevated (left column) or normal (center column) serum calcitriol levels. The reference range for each variable is indicated by the shaded area, and the median values are indicated by the horizontal bars. The reference range for serum calcitriol during hypercalcemia was defined as the mean of the serum calcitriol levels in the control patients ±3 SD. To convert serum calcium values to mmol/L, multiply by 0.2495.

Hypercalcemic Patients with Non-Hodgkin Lymphoma

We next applied the derived reference range for serum calcitriol levels during hypercalcemia to the 22 hypercalcemic patients with non-Hodgkin lymphoma and used the upper limit to divide the patients into two groups: those with elevated serum calcitriol levels ( 42 pg/mL) and those with normal serum calcitriol levels (<42 pg/mL). Twelve patients (55%) had elevated serum calcitriol levels (median, 83 pg/mL; range, 51 to 170 pg/mL), and 10 patients (45%) had normal serum levels (median, 15 pg/mL; range, <5 to 31 pg/mL).

The median age of the 22 hypercalcemic patients with non-Hodgkin lymphoma was 55 years (range, 34 to 77 years); 14 patients (64%) were male. One patient was seropositive for antibodies to the human immunodeficiency virus (HIV). As described above, human T-cell lymphotrophic virus type 1 infection can cause T-cell leukemia-lymphoma, which is often associated with hypercalcemia. Only two patients studied (9%) had T-cell disease, and both were seronegative for antibodies to human T-cell lymphotrophic virus type 1. We also tested 12 of the 20 patients with B-cell disease; all were seronegative. In 19 of 22 cases (86%), the lymphoma was classified histologically as the diffuse large-cell type according to the International Working Formulation [36]. Two patients had follicular small-cleaved cell lymphoma (both had elevated serum calcitriol levels) and one had follicular mixed-cell lymphoma and a normal serum calcitriol level. The two patients with T-cell disease had normal serum levels of calcitriol.

Most patients had advanced-stage disease during the study: Two patients (9%) were classified as Ann Arbor stage II, 2 (9%) were classified as stage III, and 18 (82%) were classified as stage IV. Eight patients (36%) had systemic symptoms, which were defined as the presence of one or more of the following: drenching night sweats, unexplained temperature of greater than 38 °C, or unintentional weight loss of more than 10% of body weight in the preceding 6 months. Five patients (23%) were studied at the time of diagnosis and 17 (77%) at the time of relapse. Patients were evaluated for osseous disease at the discretion of the treating physician; we documented bone lesions in 9 patients (41%). We observed no significant differences (all P values > 0.1) in the distribution of these features between the patients with elevated or normal serum calcitriol levels (data not shown).

When we compared the entire group of 22 hypercalcemic patients who had non-Hodgkin lymphoma with the control group, the serum calcitriol levels were significantly higher in the patients with non-Hodgkin lymphoma (P < 0.001): Median serum levels were 51 pg/mL and 5 pg/mL, respectively. This difference was seen despite similar serum levels of the major factors influencing calcitriol production (all P values > 0.1). The following levels are expressed as medians for control and hypercalcemic patients with non-Hodgkin lymphoma, respectively: intact parathyroid hormone (9 pg/mL and 11 pg/mL), corrected serum calcium (2.92 mmol/L and 2.92 mmol/L [11.7 mg/dL and 11.7 mg/dL]), and serum phosphate (1.29 mmol/L and 1.16 mmol/L [4.0 and 3.6 mg/dL]). Serum levels of intact parathyroid hormone within the normal range were found in 9 of 22 (41%) patients with hypercalcemic non-Hodgkin lymphoma and in 7 of 16 (44%) control patients. The serum levels of the inactive precursor of calcitriol, 25-hydroxyvitamin D, were also similar (P > 0.1): Median levels were 17 ng/mL and 16 ng/mL, respectively.

Two hypercalcemic patients with non-Hodgkin lymphoma, both with B-cell diffuse large-cell lymphoma, had elevated serum levels of parathyroid hormone-related protein (2.7 pmol/L and 8.4 pmol/L). The calcitriol levels of these two patients were 87 pg/mL and 27 pg/mL, respectively, and both had normal serum phosphate levels. Neither patient had previous or concurrent solid tumors.

We then examined the biochemical features in the hypercalcemic patients with non-Hodgkin lymphoma according to whether the serum calcitriol level was elevated or normal (Figure 1). Because we used calcitriol levels as the basis for classification, it is not meaningful to compare these levels between the two groups. We observed no significant differences among any of the other biochemical measurements (all P values 0.3) (Table 1).

Urinary Electrolytes

Many study patients were acutely ill. It was not always possible to obtain the desired urine specimens before initiating necessary treatment measures. Therapy was never delayed to obtain appropriate specimens. Thus, urinary results are available for 11 of the 22 (50%) hypercalcemic patients with non-Hodgkin lymphoma and 11 of the 16 (69%) control patients. We observed no differences (P > 0.1) between patients with urine samples obtained for each disease category and those without. No significant differences in urinary calcium excretion or TmPO₄^3-were found between the control group and hypercalcemic patients with non-Hodgkin lymphoma and either elevated or normal serum calcitriol levels (P 0.4) (Table 1).

Relation between Serum Calcitriol and Calcium

In patients with elevated serum calcitriol levels, the serum levels of corrected calcium and calcitriol were strongly correlated (r = 0.60; P = 0.04). There was no such correlation in the patients with non-Hodgkin lymphoma and normal serum calcitriol levels (r = –0.20;P > 0.2), whereas in the control patients, these levels were negatively correlated (r = –0.66;P = 0.006). The results in the patients with elevated serum calcitriol levels differed significantly from the results in control patients (P < 0.001) and those with normal serum calcitriol levels (P = 0.037).

Sequential Studies

Serum calcitriol levels were sequentially determined in five patients with non-Hodgkin lymphoma. One patient had asymptomatic hypercalcemia and a serum calcitriol level of 75 pg/mL. After a 5-week period without intervention, the serum calcitriol level remained elevated at 99 pg/mL. In four patients, repeated determinations were done a median of 13 days (range, 6 to 50 days) after the initiation of chemotherapy (Table 2). In each case, a reduced serum calcitriol level accompanied a decrease in corrected serum calcium levels, and these differences were highly significant (P = 0.015 and P < 0.001, respectively). Four months after the initial studies, patient 4 developed progressive lymphoma accompanied by recurrent hypercalcemia (corrected serum calcium level, 2.89 mmol/L [11.6 mg/dL]). Again, the serum calcitriol level was elevated at 190 pg/mL, despite suppression of the serum level of intact parathyroid hormone (6 pg/mL).

Normocalcemic Patients with Non-Hodgkin Lymphoma

To determine the frequency of abnormal calcium metabolism in patients with non-Hodgkin lymphoma in the absence of hypercalcemia, we evaluated 25 consecutive new-ly diagnosed normocalcemic patients with non-Hodgkin lymphoma before they received therapy. Because diffuse large-cell lymphoma was the most common histologic type among the hypercalcemic patients, we limited our analysis to patients with this type of lymphoma. We excluded 3 patients with elevated serum levels of intact parathyroid hormone that were consistent with primary hyperparathyroidism. The median age of the 22 evaluable patients was 57 years (range, 17 to 71 years); 15 (68%) were male. Three patients were seropositive for HIV. Advanced-stage disease was less common in these patients than in the hypercalcemic patients: Eleven patients (50%) were classified as Ann Arbor stage III or IV. Only 1 patient had documented bone involvement. Six (27%) patients had systemic symptoms.

All patients had normal corrected serum calcium levels (median, 2.32 mmol/L [9.3 mg/dL]; range, 2.09 to 2.49 mmol/L [8.4 to 10.0 mg/dL]) and normal serum levels of both phosphate and creatinine. The median level of 25-hydroxyvitamin D, the inactive precursor of calcitriol, was 24 ng/mL (normal range, 10 to 55 ng/mL).

Four patients (18%) had serum calcitriol levels greater than the normocalcemic reference range (20 to 76 pg/mL). The median serum calcitriol level was 63 pg/mL (Figure 2). Seventeen patients provided a fasting urine specimen, and 12 (71%) had an elevated urinary calcium excretion ( 0.14 mg/dL of glomerular filtrate) (Figure 2). The one patient with lymphomatous bone involvement had an elevated calcitriol level of 90 pg/mL. We noted no correlation between serum calcitriol levels and either corrected serum calcium (r = –0.16;P > 0.2) or urinary calcium excretion (r = 0.27; P > 0.2).

View larger version (36K): [in this window] [in a new window]   Figure 2. Distribution of urinary calcium excretion and serum levels of intact parathyroid hormone (PTH), calcitriol, and corrected calcium among normocalcemic patients with non-Hodgkinlymphoma. The normal range is indicated by the shaded area, and the median value is indicated by the horizontal bars. To convert serum calcium values to mmol/L, multiply by 0.2495. GF = glomerular filtration.

It is not possible to directly compare the relation of the serum levels of calcitriol and corrected calcium between the normocalcemic and hypercalcemic patients with non-Hodgkin lymphoma. Such an analysis is compounded by the strong inhibitory influence of hypercalcemia on calcitriol production and the different reference ranges for serum calcitriol in these settings as discussed above. However, serum levels of calcitriol greater than the appropriate reference range were more common in the hypercalcemic patients (55%) than in the normocalcemic patients (18%) (P = 0.015).

Discussion

Top Methods Results Discussion Author & Article Info References

We found that serum calcitriol levels in hypercalcemic patients with non-Hodgkin lymphoma are significantly elevated compared with those of controls, showing that calcitriol production in patients with non-Hodgkin lymphoma and hypercalcemia is frequently dysregulated. We selected hypercalcemic patients with myeloma as a control group because it has previously been shown that in these patients, the production of calcitriol is appropriately and profoundly suppressed during hypercalcemic episodes and returns to normal after bisphosphonate therapy in the absence of cytotoxic treatment [26]. Further, serum levels of parathyroid hormone-related protein, which may stimulate renal calcitriol production [22, 23], were not elevated in the patients with myeloma in our study.

Fewer than 20 previous reports have implicated calcitriol as a mediator of hypercalcemia in non-Hodgkin lymphoma [4-9, 37]. These studies have not been of consecutive patients, and the inability to define a reference range for serum calcitriol during hypercalcemia has limited these reports to patients with serum calcitriol levels frankly elevated above the normocalcemic reference range. Using the mean serum calcitriol level of the control patients plus 3 standard deviations, we defined the upper limit of expected serum calcitriol during hypercalcemia as 42 pg/mL, approximately half of the upper limit of the normocalcemic reference range, 76 pg/mL.

In the largest series of this type, 10 hypercalcemic patients with lymphoma were studied [6], excluding those patients with lymphoma associated with human T-cell lymphotrophic virus type 1, in which parathyroid hormone-related protein-mediated hypercalcemia is common [28]. Five of these patients had elevated calcitriol levels, although parathyroid hormone-related protein levels were not available. In our study, 12 (55%) of the 22 hypercalcemic patients with non-Hodgkin lymphoma had elevated serum calcitriol levels; 7 of these patients had levels that were greater than the upper limit of the normocalcemic reference range. The finding of an elevated serum parathyroid hormone-related protein level in only 2 (9%) of the 22 patients, only 1 of whom had an elevated serum calcitriol level, excludes stimulation of renal calcitriol production by parathyroid hormone-related protein as the cause of the elevated serum calcitriol. Further, it is unlikely that parathyroid hormone-related protein acted in a local paracrine manner to stimulate calcitriol production by macrophages because macrophage 1 -hydroxylase, in contrast to the renal enzyme, is not regulated by parathyroid hormone-like substances [38].

In our study, 42% of the hypercalcemic patients had intact parathyroid hormone serum levels within the normal range. On the basis of various assays, most previous studies have found only 5% to 20% of patients with hypercalcemia of malignancy to have serum intact parathyroid hormone levels within the normal range [21, 39-41]. Using the same CIBA-Corning assay that was used in our study, other groups found serum parathyroid hormone levels within the normal range in 9 of 10 [42] and 9 of 36 patients with humoral hypercalcemia of malignancy (information on file at CIBA-Corning). The slightly higher serum parathyroid hormone levels seen in our study may be due to the inclusion of patients with mild hypercalcemia or to some recovery from suppression of parathyroid hormone secretion during hydration before the study. Regardless, the serum parathyroid hormone levels did not differ between the two groups. Thus, the marked difference observed in calcitriol levels cannot be explained solely by any residual activity of parathyroid hormone.

In our study, most patients with hypercalcemia had intermediate-grade lymphoma (91%), predominantly of the diffuse large-cell type, and 20 of 22 patients had B-cell disease. Although our study did not evaluate the relative incidence of hypercalcemia according to histologic grade, our findings are consistent with those of previous reports [1, 2, 43]. We could not determine any features that could identify those patients subsequently shown to have elevated calcitriol levels.

In the hypercalcemia that complicates non-Hodgkin lymphoma, it is likely that the calcitriol is of extrarenal origin. Lymphomatous tissue has been shown to convert 25-hydroxyvitamin D to calcitriol in vitro [8]. Further, cytokine-stimulated macrophages can produce significant calcitriol [37] and macrophage infiltration of tumor tissue, recognized histologically as a "starry-sky" appearance [36], is prominent only in intermediate- and high-grade lymphomas, in which hypercalcemia is also most common [1, 2]. It is probable but not yet proven that infiltrating macrophages rather than lymphoma cells per se are the source of calcitriol.

Although the pathogenic role of calcitriol in the hypercalcemia of non-Hodgkin lymphoma has been questioned [44], many data support such a role. Previous studies have shown enhanced intestinal calcium absorption [6], decreases in calcitriol level with therapy, and recurrent elevations with relapsing hypercalcemia [4-7, 9]. We also found the serum levels of corrected calcium and calcitriol to be strongly correlated (r = 0.60; P = 0.04) among the hypercalcemic patients with non-Hodgkin lymphoma and elevated serum calcitriol levels. The previously reported reduction in the serum calcitriol levels in concert with normalization of the corrected serum calcium after the initiation of cytotoxic therapy [6, 7] is also consistent with a pathogenic role for calcitriol. Calcitriol is certainly not the sole factor responsible for hypercalcemia in patients with non-Hodgkin lymphoma. As stated in a recent review [37], other additional factors may contribute to the hypercalcemia. Renal function, dietary intake of calcium and fluids, mobility, and the interacting influences of other active mediators such as interleukin-1, tumor necrosis factor-, and transforming growth factor- all influence the degree of hypercalcemia in response to a given level of calcitriol. Further, 45% of hypercalcemic patients in our study had appropriately suppressed serum calcitriol levels. A few patients with lymphoma, 9% in our study, will have elevated parathyroid hormone-related protein levels that contribute to their hypercalcemia. The pathogenesis of the hypercalcemia in patients without elevations of calcitriol or parathyroid hormone-related protein levels remains uncertain. The possible role of the additional calciotropic factors described above requires further study.

The study of normocalcemic patients at diagnosis establishes that dysregulated production of calcitriol is common in patients with diffuse large-cell lymphoma. Although no patients were hypercalcemic, 71% were hypercalciuric, and 18% had elevated serum calcitriol levels. It is uncertain why none of these patients were hypercalcemic despite significant elevations of the calcitriol levels, although serum creatinine levels were normal in all the normocalcemic patients. However, we observed elevations in the serum creatinine levels above 132.6 µmol/L (1.5 mg/dL) in 45% of the hypercalcemic patients with non-Hodgkin lymphoma. Previous reports have also suggested that hypercalciuria may be present in some normocalcemic patients with non-Hodgkin lymphoma [5, 45].

Our results confirm that abnormalities of calcitriol production and regulation are common during hypercalcemia in intermediate-grade non-Hodgkin lymphoma and suggest that dysregulated calcium metabolism, hypercalciuria, and significant extrarenal calcitriol production are common events in these diseases, even in the absence of hypercalcemia. More detailed longitudinal evaluation of vitamin D and calcium metabolism in both normocalcemic and hypercalcemic patients with non-Hodgkin lymphoma are required. It is likely that the abnormalities described are caused by the activities of cytokines involved in the host-tumor interaction, and clarification of this process will increase the understanding of the pathogenesis of malignant lymphoproliferative diseases.