Bone resorption and renal calcium reabsorption in renal cell carcinoma-bearing mice: the effects of bisphosphonate

Authors


Max G. Weissglas, Department of Urology, Kennemer Gasthuis locatie Zuid, PO Box 417, 2000 AK, Haarlem, the Netherlands.
e-mail: weissgla@kg.nl

Abstract

OBJECTIVE

To examine the contribution of the skeleton and the kidney to the development of humoral hypercalcaemia of malignancy (HHM) in a mouse model of HHM treated with a potent bisphosphonate.

MATERIALS AND METHODS

Mice bearing the human RCC cell line RC-9 were treated with bisphosphonate (subcutaneous, 0.25 mg/kg body weight olpadronate) or saline solution. Treatment was initiated at a tumour volume (TV) of ≈100 mm3 and 500 mm3, and the mice were monitored for ≈4 weeks. Serum calcium and phosphate concentrations and trabecular bone volume (TBV) were assessed during and/or after treatment.

RESULTS

Athymic mice implanted with the RCC RC-9, developed severe hypercalcaemia and bone resorption. During tumour growth the mean (sd) serum calcium concentration increased to 4.1 (0.3) mmol/L, and phosphate decreased to 1.6 (0.3) mmol/L, vs 2.3 (0.1) and 2.9 (0.4) mmol/L in controls, respectively. TBV decreased from 8.7 (1.8)% in mice with no tumour, to 5.3 (2.7)% in RC-9-bearing mice. Olpadronate initiated at a Tv of 100 mm3 prevented the loss of bone induced by RCC RC-9 cells, with a TBV of 12.8 (2.1)%, but the development of hypercalcaemia was unaffected. Olpadronate treatment at a TV of 500 mm3 did not influence the development of hypercalcaemia and did not protect against bone resorption. Kinetic monitoring showed an identical rate of tumour growth in the presence or absence of bisphosphonate, while under both conditions there was a tumour load-dependent increase in calcium concentration.

CONCLUSIONS

Bisphosphonate can prevent parathyroid hormone-related peptide (PTHrP)-mediated bone resorption when administered during the early phase of renal tumour growth, but has no effect on the tumour-induced development of hypercalcaemia, indicating a primary role for renal tubular reabsorption of calcium in the kidney by PTHrP in HHM.

Abbreviations
HHM

humoral hypercalcaemia of malignancy

PTH(rP)

parathyroid hormone(-related peptide)

IL

interleukin

TV

tumour volume

MMA

methyl methacrylate

TBV

trabecular bone volume.

INTRODUCTION

Cancer is an important cause of clinically significant hypercalcaemia, a well-known paraneoplastic syndrome in RCC [1–5]. The association of malignant hypercalcaemia with stimulation of osteoclastic bone resorption, either by local factors secreted by tumour cells in the bone marrow or by humoral factors secreted by the tumour at distance (humoral hypercalcaemia of malignancy, HHM) has been extensively addressed. Studies in vitro, in athymic mice transplanted with renal tumour and patients with RCC, showed that parathyroid hormone-related peptide (PTHrP) is the main stimulator of osteoclastic bone resorption [6–9]. PTHrP, in concert with other factors such as interleukin (IL)-1, IL-6 and IL-11, stimulates the synthesis of receptor activator of NF-κB ligand (RANKL) that binds to the RANK receptor of osteoclast progenitor cells, leading to differentiation into mature, bone resorption-mediating osteoclasts [10]. However, PTHrP, as PTH, also stimulates renal calcium reabsorption, thus contributing to the increase in calcium concentration [11]. By contrast, with existing and still expanding knowledge about the pathogenesis of tumour-induced bone resorption, the importance of the renal component in HHM has received little attention. This can be of considerable clinical significance, as patients with HHM appear to respond less well than patients with bone metastases to calcium-lowering treatment with bisphosphonates, which suppress exclusively the skeletal component of HHM. In the present study, we investigated the contribution of renal calcium handling to the development of hypercalcaemia induced by the RC-9 RCC line in nude mice.

MATERIALS AND METHODS

The human RCC cell line RC-9 was established from a patient with renal cancer at an advanced stage, and was maintained as a xenograft in the nude mice (nu/nu) as previously described [12,13]. During serial transplantation there were no changes in growth kinetics, histology and immunohistochemistry of the tumour.

Swiss nu/nu mice, 4–6 weeks old, were maintained in laminar flow-cage racks at a controlled temperature (25 °C) and humidity (60%). The mice had free access to acidified water and (irradiated) standard rodent food containing 0.84% calcium (RMH-GS, Hope Farms, Bodegraven, the Netherlands).

Selected, vital tumour pieces (≈2 × 2 × 2 mm) were inserted unilaterally and s.c. into the shoulder region of ether-anaesthetized mice. Tumour volume (TV) was monitored by measuring the two major diameters, d1 and d2, of the tumours with a graduated calliper, and estimated as π/6(d1 × d2)1.5. Mice with a TV of ≈100 mm3 were selected and divided into groups of seven. Mice were injected s.c. with either an appropriate volume of PBS or 0.25 mg/kg body weight olpadronate (dimethyl-APD), a potent nitrogen-containing bisphosphonate; this dose was three times the dose used in humans. In the first set of experiments treatment was started either at a TV of ≈100 mm3 or at ≈500 mm3, and was given three times per week (Fig. 1). At ≈4 weeks after xenograft transplantation the mice were killed, blood collected to measure serum calcium levels, and the femurs excised for bone histomorphometry. To compare the kinetics of the development of hypercalcaemia with and without bisphosphonate, we conducted a detailed study in which the administration of bisphosphonate or saline solution was started at a TV of ≈100 mm3 and serum calcium levels were assessed by blood sampling from each mouse once a week for 3 weeks.

Figure 1.

The treatment protocol. Groups of mice were injected with PBS or olpadronate at a TV of ≈100 mm3 and 500 mm3. sCa, calcium concentration; sPO4, phosphate concentration.

Serum calcium and phosphate concentrations were determined using an automated autoanalyser. For bone histomorphometry one femur of each mouse was cleaned of soft tissue, the bones fixed in 10% neutral buffered formalin, slightly trimmed, and dehydrated in an ascending series of ethanol, infiltrated in methyl methacrylate (MMA) and embedded in MMA. After polymerization, the MMA blocks were trimmed and cut using a heavy-duty microtome (HM 350, Microm, Heidelberg, Germany). Sections (4 µm) were stained according to von Kossa, which colours mineralized bone black. The trabecular bone volume (TBV) was measured using the Optimas histomorphometric package (Bioscan Inc. Edmonds, WA, USA) attached to a Nikon microphot FXA microscope (Melville, NY, USA) and an MX5 CCD camera (Adimex Image Systems BV, Eindhoven, the Netherlands) [14]. All measurements were made at a distance of 0.5–1.0 mm from the epiphyseal plate. These studies were approved by the ethical committee for animal experimentation of the Leiden University Medical Centre.

Statistical differences of serum calcium, phoshate and TBV between treated and untreated groups were analysed using Student’s t-test, with regression and correlation analysis used as appropriate.

RESULTS

All mice implanted with the RC-9 tumour developed hypercalcaemia and hypophosphataemia. The administration of bisphosphonate had no effect on the concentrations of serum calcium and phosphate at the end of the experiment, either in the mice treated at a TV of 100 mm3 or 500 mm3 (Table 1).

Table 1. 
The development of hypercalcaemia and bone resorption associated with the RC-9 tumour, and the effects of the three weekly administrations of the bisphosphonate olpadronate
GroupMean (sd) serum level of TBV, %
Ca, mmol/LPO4, mmol/L
Control, no tumour2.3 (0.1)2.9 (0.4) 8.7 (1.8)
Tumour ≈100 mm3
 no olpadrone4.1 (0.3)1.6 (0.30) 5.3 (2.7)
 olpadrone4.2 (0.5)1.7 (0.3)12.8 (2.1)
Tumour ≈500 mm3
 no olpadrone4.2 (0.5)2.2 (0.2) 5.9 (3.4)
 olpadrone4.0 (0.4)1.8 (0.1) 6.7 (2.2)

Bone histomorphometry showed that tumour-bearing mice treated with PBS had significantly (P = 0.021) lower TBV than age-matched controls with no tumour (Table 1). The TBV was similar at the end of the experiments in both groups of tumour-bearing mice, at 5.3 (2.7)% and 5.9 (3.4)%, vs 8.7 (1.8)% in controls. This result indicates a significant skeletal contribution to the hypercalcaemia. Treatment with olpadronate, started at a low TV, prevented the tumour-induced bone loss and even significantly increased TBV above control values. By contrast, when olpadronate was given to mice with a high tumour load, it could not prevent bone loss, the TBV being similar to that in control mice (Table 1).

These experiments showed that olpadronate given at an early stage of tumour growth could prevent bone resorption, but had no effect on the final calcium concentrations. To obtain further insight into the relationship between these apparently conflicting observations, we followed the changes in serum calcium concentration with time in a different group of mice transplanted with the RC-9 RCC. Blood samples from individual mice were taken when olpadronate or saline solution was started, at a TV of ≈ 100 mm3 (day 0) and the every week for 3 weeks. As shown in Fig. 2, the weekly administration of olpadronate affected neither the onset nor the rate of the increase in serum calcium concentration, compared to the control. Moreover, in both untreated and olpadronate-treated mice there was a significant and identical correlation between serum calcium levels and TV (Fig. 3).

Figure 2.

The development of hypercalcaemia (calcium concentration) without and with olpadronate. The dotted and solid lines represent the calcium concentration during three weekly administrations of PBS or olpadronate, respectively.

Figure 3.

The linear correlation between serum calcium and RC-9 TV during three weekly administrations of PBS (control, dotted line: r = 0.72, P < 0.001) or olpadronate (solid line, r = 0.65, P = 0.002). The two regression lines were comparable with similar slopes, i.e. PBS 0.006 (0.001) vs olpadronate 0.007 (0.002) mmol/L per mm3) and intercepts, i.e. 2.1 (0.5) and 2.3 (0.6) mmol/L, respectively.

DISCUSSION

The present results show that nude mice transplanted with RC-9 RCC develop severe hypercalcaemia and hypophosphataemia, metabolic changes also apparent in patients with RCC and that have been attributed to increased production of PTHrP by the tumour cells. PTHrP, by binding to the PTH-1 receptor, evokes the same biological activity as PTH, resulting in increased bone resorption and enhanced renal tubular reabsorption, leading to hypercalcaemia.

Bisphosphonates can prevent osteoclastic bone resorption; currently they are the standard treatment in malignancy-associated hypercalcaemia. However, in patients with high circulating levels of PTHrP, their effect on serum calcium concentration is only partial and temporary. This might be explained by the action of PTHrP on renal calcium handling, leading to increased tubular reabsorption of calcium [15–17].

Previously we reported the induction of hypercalcaemia by two RCC cell lines, one of which was RC-9, in the nude mouse model [18]. There was a clear correlation between the concentration of serum calcium, level of PTHrP and amount of loss of TBV in tumour-bearing mice. These results suggested that the skeletal resorptive effect of PTHrP was the predominant cause for the hypercalcaemia. However, a contribution of renal calcium reabsorption, although less likely, could not be excluded. The present mouse model makes it possible to inhibit selectively bone resorption without affecting tumour-induced alterations of serum calcium, due to renal calcium reabsorption. To examine this possibility of differentiation we assessed the effects of the bisphosphonate olpadronate on the development of hypercalcaemia and bone destruction in athymic mice implanted with the RCC line RC-9. Treatment with olpadronate in the early phase of tumour growth prevented RC-9-associated bone resorption, but the development of hypercalcaemia (and tumour growth) was unaffected. Interestingly, despite the achieved elimination of bone resorption by olpadronate as a source of serum calcium, both the onset and rate of increase of serum calcium appeared to be equal to that in untreated tumour-bearing mice. Based on the assumption of a primary role for increased osteoclastic bone resorption, we expected, in the group of mice treated in the early phase of tumour growth, a later onset and a slower increase in serum calcium concentration. Other animal studies showed that PTHrP might stimulate renal tubular calcium resorption. The results of the present study seem to indicate that the renal effect of PTHrP on serum calcium concentrations is predominant to the osteoclastic effect, at least in this mouse model. The question remains as to why, in patients with HHM, bisphosphonates have a temporary effect. Possibly the answer is not a qualitative difference in the action of PTHrP on bone and kidney between men and mice, but more a quantitative difference. The nude mouse has a body weight of 25 g, with a relatively large s.c. implanted renal tumour, producing large amounts of PTHrP, capable of breaking down the skeleton of the mouse within 4 weeks. This differs clearly from the situation in the human patient, where a more gradual pathological process takes place. That PTHrP acts on the human kidney was shown in studies by Syed et al.[19] in healthy human volunteers, where PTHrP stimulated renal calcium reabsorption, suggesting a significant contribution of PTHrP-induced renal calcium reabsorption in HHM. Thus, the present findings of the importance of the renal effect of PTHrP might be applicable to the clinical situation, despite the possible different features of the rodent and human kidney.

The consequence of the important renal effect of PTHrP in tumours causing HHM is that therapy of the hypercalcaemia should be directed against PTHrP itself, rather than directed against the osteoclasts. However, that does not exclude the use of bisphosphonates in the treatment of these patients, as bisphosphonates are very effective in protecting the bone against osteoclastic bone resorption. Theoretically and under experimental conditions, the action of bisphosphonates will be more effective when given in an early phase of the tumour process, before significant bone destruction has occurred. In clinical situations this might be more difficult to realise, as tumour development and progression are less predictable. Clinical trials are needed to show the usefulness of treatment with bisphosphonates in an early phase in PTHrP-producing tumours to prevent bone destruction.

In conclusion, treatment with the bisphosphonate olpadronate in nude mice bearing the PTHrP-producing RC-9 RCC had no effect on the time of onset and rate of hypercalcaemia, although the skeletal resorptive effect of PTHrP could be blocked when olpadronate was administered in the early phase of tumour growth. These findings suggest an important, if not primary, contribution of tumour-induced renal calcium reabsorption in HHM. A renal effect of tumour-derived PTHrP might be relevant for the clinical situation, notwithstanding the possible different features between the physiology of rodent and human kidney. Bisphosphonate administration can have a positive effect in preventing bone resorption, but treatment of HHM by anti-osteoclastic bisphosphonate therapy alone will not suffice to correct the hypercalcaemia. Treatment to control hypercalcaemia in patients with RCC and other solid tumours, and with elevated levels of PTHrP, in whom removal of the primary tumour is not possible or indicated, should therefore be directed against PTHrP.

CONFLICT OF INTEREST

None declared.

Ancillary