Parathyroid hormone related protein

Parathyroid hormone related protein (PTHrP), isolated and cloned in 1987, is now thought to be the aetiological factor in many cases of hypercalcaemia associated with malignancy. This is particularly true for solid tumours (Ratcliffe et al., 1992). PTHrP has all the known biological effects of PTH and produces them by acting on a common PTH/PTHrP receptor (Martin, 1990). These actions of PTHrP can be produced by synthetic amino-terminal fragments of PTHrP with full activity residing in the first 34 amino acids. Since the identification of the full sequence of PTHrP, it has seemed likely that such a large molecule containing a number of proteolytic cleavage sites might well have other actions associated with different parts of the molecule. Provisional evidence for this has now been presented. PTHrP seems to be involved in placental calcium transport and the regulation of fetal calcium homeostasis. PTH and 1-34 PTHrP cannot produce these effects whereas 1-141 PTHrP can. Care and his colleagues (1990) have suggested that this activity resides in the 75-84 part of the peptide sequence. The Melbourne group have now produced evidence that the carboxy terminal sequence inhibits osteoclastic bone resorption (Fenton et al., 1991). PTHrP is produced by many normal cells in culture as well as by malignant cell lines (Docherty et al., 1989). This suggests that PTHrP is produced by normal cells, especially epithelial cells, and that it has local autocrine or paracrine actions which are currently unknown. Under normal physiological circumstances, local production of PTHrP would not lead to significant amounts gaining access to the circulation and no convincing evidence has been presented to indicate that PTHrP normally circulates. Under pathological conditions PTHrP produced in excess by malignant or benign processes may lead to high circulating PTHrP concentrations which may then cause hypercalcaemia. Following the isolation of PTHrP, it has been possible to develop immunoassays and, as might be expected, two-site immunoradiometric assays have proved to be the most sensitive and specific (Ratcliffe et al., 1991). The availability of antibodies has also allowed the cellular distribution of PTHrP to be studied using immunocytochemistry. The combination of the two techniques has recently permitted interesting new insights into metastatic breast cancer (Bundred et al., 1991). It has always been known that hypercalcaemia associated with breast cancer is almost always associated with metastatic disease, often involving bone. For many years, the hypercalcaemia was thought to be due to local osteolysis and not to have a humoral basis. The study by Bundred et al. showed that 92% of hypercalcaemic patients with breast cancer and metastases to bone had elevated PTHrP concentrations. This is further evidence that hypercalcaemia in breast cancer is likely to have a humoral mechanism. This complements other indirect evidence of a humoral cause (Gallacher et al., 1990). Patients with bone metastases who were normocalcaemic had elevated PTHrP values in 36% of cases while it was elevated in only 9% of women without obvious metastases. The obvious explanation of these results is that the greater the tumour burden, the more PTHrP is produced and hence the greater the chance of hypercalcaemia. However, clinical observations show a poor correlation between the extent of metastases and hypercalcaemia. If we now accept that hypercalcaemia in breast cancer is usually due to the excessive production of PTHrP, does the presence or absence of PTHrP in the primary tumour tell us anything more about the likely behaviour of that tumour? Obviously the PTHrP + ve tumours might be predicted to be more likely to cause hypercalcaemia if the disease becomes disseminated. In addition to this, the Melbourne group are now suggesting that such tumours are more likely to metastasize to bone. This suggestion is based on a study in which metastases to bone were immunohistochemically positive for PTHrP in 92% of cases while metastases to nonosseous sites were positive in only 17% of cases (Powell et al., 1991). Unfortunately, in most cases it was not possible to study the original primary breast tumour. Would it be important to know in advance which tumours are likely to metastasize to bone? The answer to this is likely to be, ‘yes’. Recent studies suggest that the prophylactic use of bisphosphonate drugs reduces the chances of hypercalcaemia developing and may reduce the chances of metastases to bone. In such studies many patients who do not develop bone metastases are treated. The presence or absence of PTHrP in the primary tumour might identify an at risk group who warrant special treatment. It is therefore important to verify as quickly as possible whether the Australian studies are correct. Primary breast tumours have twice been studied for the presence of PTHrP. The results were similar with 52 and 60% of the tumours being positive (Southby et al., 1990; Bundred et al., 1991). Of women presenting with breast cancer, approximately 60% are cured by initial surgery, though this varies consider-