Fax: (619) 543-6723
Article first published online: 29 OCT 2004
Copyright © 2004 American Cancer Society
Volume 101, Issue 11, pages 2622–2628, 1 December 2004
How to Cite
Pardo, F. S., Lien, W. W., Fox, H. S., Efird, J. T., Aguilera, J. A., Burton, D. W. and Deftos, L. J. (2004), Parathyroid hormone–related protein expression is correlated with clinical course in patients with glial tumors. Cancer, 101: 2622–2628. doi: 10.1002/cncr.20689
Presented in part at the 45th Annual Meeting of the American Society for Therapeutic Radiology and Oncology, Salt Lake City, UT, October 19–23, 2003.
The opinions expressed herein do not necessarily reflect the views of the National Institutes of Health or the U.S. Government.
- Issue published online: 16 NOV 2004
- Article first published online: 29 OCT 2004
- Manuscript Revised: 24 AUG 2004
- Manuscript Accepted: 24 AUG 2004
- Manuscript Received: 29 APR 2004
- Moores University of California–San Diego Cancer Center (2000)
- National Institutes of Health. Grant Number: K22-CA95089
- California Department of Health and Human Services. Grant Number: 03-00169VRS-30017
- parathyroid hormone–related protein;
- glial tumors;
- central nervous system;
- immunopathologic criteria
Parathyroid hormone–related protein (PTHrP) expression modulates cell survival in a number of human solid tumors. Although PTHrP is expressed in normal developing and neoplastic central nervous system tissue, clinical data indicating the importance of this protein with respect to local control and/or survival in patients with glial tumors are scarce.
Using a standard immunoperoxidase technique, the authors examined PTHrP expression in a population of 51 patients with Daumas–Duport Grade II–IV astrocytomas over a 15-year period. Both local control and survival were calculated from the date of definitive irradiation to the last time of known follow-up examination using the actuarial method. PTHrP expression was scored on examination under 40× magnification, with the incidence of cellular staining averaged over 10 high-power fields. The intensity and extent of staining were characterized semiquantitatively using the standard World Health Organization classification criteria. The median follow-up duration was approximately 5.5 years. Multivariate analyses were performed to ascertain the statistical significance of several standard clinicohistopatholgic factors (Karnofsky functional status, age, gender, extent of surgical resection, radiotherapy dose, grade, and PTHrP expression) with respect to local control and survival. P < 0.05 was considered indicative of statistical significance.
Patients with high levels of PTHrP expression had significantly lower glial tumor local control rates and corresponding decreases in progression-free and overall actuarial survival after definitive irradiation (P < 0.01). In a Cox 3-variable model, the PTHrP staining score was independent of tumor grade or Karnofsky functional status. It is notable that the strongest predictor of survival was tumor grade (P < 0.001).
PTHrP may be an important adjunct to standard immunopathologic criteria in the determination of glial tumor responses. A number of mechanisms were explored to derive a more mechanistic understanding of these translational results. Subsequent prospective studies involving larger patient populations will be necessary before findings can be translated to clinical practice. Cancer 2004. © 2004 American Cancer Society.
Alterations in parathyroid hormone–related protein (PTHrP) expression are well documented in patients with hypercalcemia of malignancy.1–3 Soon after the cloning of PTHrP in the late 1980s, the homology between the first eight N-terminal amino acids of parathyroid hormone (PTH) and PTHrP was recognized.4, 5 Since then, a substantial amount of work has been conducted on the role of PTHrP in calcium metabolism and function.4 Although the name ‘PTHrP’ was chosen based on the homology between PTH and PTHrP, distinct N- and C-terminal amino acid sequences give each of these two proteins unique functions.4–9 For example, although both proteins share the function of binding to the G-protein-coupled receptor (the type 1 PTH/PTHrP receptor) while concomitantly stimulating cyclic adenosine monophosphate production and cytosolic calcium influx,8, 10 PTH also is a classic systemic peptide hormone that is responsible for regulating calcium and phosphorus homeostasis directly, as typified in bone growth and remodeling,11 whereas PTHrP function is limited to autocrine/paracrine mechanisms.8 Perhaps the most well defined function of PTHrP in adults is its effect on smooth muscle. PTHrP expression induces a decrease in calcium influx, resulting in the relaxation of muscle in organs.3, 4, 6, 12
Since the cloning of PTHrP, the expression of this protein has been observed in a variety of normal and malignant tissue specimens in the absence of hypercalcemia. PTHrP is expressed widely in various normal and malignant tissue specimens, where it serves as a local autocrine and/or paracrine factor involved in the regulation of cellular growth and differentiation.8, 12–14 Much of the initial work on PTHrP concentrated on its expression within the context of bone and cartilaginous abnormalities. Expression studies in various tumor types, including breast, prostate, colon, lung, renal, and ovarian carcinomas, indicate a more global role for PTHrP expression in the natural history of human malignancies.9, 15–24
In the central nervous system (CNS), PTHrP expression has been described in normal embryonic and human glial tumors, neurons, normal and neoplastic vasculature, as well as in several types of glial tumors.25–28 Although primary CNS tumors are not associated with the hypercalcemia of malignancy, other PTHrP functions, including signal transduction and angiogenesis, may be quite relevant to glial tumor cell survival both before and after irradiation.7, 29–35 There are, to our knowledge, no published clinical data on glial tumor local control and PTHrP expression, particularly with respect to clinical course after therapeutic irradiation.
MATERIALS AND METHODS
We examined the clinical records of 51 patients with glial tumors treated with definitive radiotherapy over an approximate 15-year period and evaluated at the University of California–San Diego (San Diego, CA). These patients were chosen at random, and each tumor specimen was analyzed immunohistochemically using the PTHrP monoclonal antibody and a standard avidin-biotin complex immunoperoxidase technique. Two pathologists (E.T.H.W. and H.S.F.) interpreted the staining results and assigned tumor grades. An antibody dilution of 5 μg/mL was used on 5 μm thick paraffin sections to localize PTHrP expression in the cytoplasm of glial tumor cells. Positive cells were stained with NovaRed (Vector Laboratories, Burlingame, CA), which, when used at a 1:50 dilution, yielded red staining of positive cells. Samples were then counterstained with hematoxylin (Sigma, St. Louis, MO). Tumor specimens were characterized according to the Daumas–Duport system (Grade I–IV).36 Tumor specimens with oligodendroglial components were not used in the current analysis. Tissue specimens of normal cortex, midbrain, and cerebellum were used as negative controls, whereas a prostate carcinoma cell line designed to overexpress PTHrP was used as the positive control.
All specimens were classified according to the World Health Organization histologic classification criteria.37 The intensity of staining was graded on a scale of 0 to 3—a score of 0 indicated no staining in comparison to the control, a score of 1 indicated weak staining, a score of 2 indicated moderate staining, and a score of 3 indicated strong staining. The extent of staining also was graded on a scale of 0 to 3—a score of 0 indicated no immunostaining, a score of 1 indicated faint and patchy immunostaining (1–30% positive cells), a score of 2 indicated moderate immunostaining (31–60% positive cells), and a score of 3 indicated prominent immunostaining (61–100% positive cells). The extent and intensity of PTHrP expression were scored by averaging the findings made in 10 high-power (40×) fields (Fig. 1).
Progression-free survival (PFS) and overall actuarial survival (OS) were calculated from the date of commencement of radiotherapy to the date of the last assessment of neurologic Karnofsky functional status (i.e., the date of last follow-up). Kaplan–Meier statistics were used to analyze potential correlations between a number of independent variables (age, grade, extent of surgical resection, Karnofsky functional status, radotherapy dose, and PTHrP expression) and PFS and OS.
Kaplan–Meier analysis revealed that patients whose glial tumor specimens overexpressed PTHrP as measured by both the intensity of staining and the percentage of positively stained cells had reduced PFS and OS (P < 0.01).38 This effect was independent of tumor grade and Karnofsky performance status. Negative correlations between survival and staining intensity and between survival and percentage of positively stained cells were documented (Figs. 2–4). Analysis of covariance demonstrated highly statistically significant differences across staining groups (P < 0.001). Age, extent of surgical resection, and radiotherapy dose were not predictive of survival.
|Staining intensity||OSa (mos)||OS rate at 12 mos (%)||OS rate at 24 mos (%)|
|None||No disease recurrences||100||100|
|Weak||17.4 ± 1.4||81.20||43.80|
|Moderate||9.6 ± 0.8||15.40||0|
|Strong||10.3 ± 1.8||33.30||0|
|Percentage of positive cells|
|None||No disease recurrences||100||100|
|1–30||13.4 ± 1.0||53.80||37.70|
|31–60||14.5 ± 1.5||51.30||0|
|61–100||7.3 ± 0.5||0||0|
|PFSa (mos)||PFS rate at 24 mos (%)||PFS rate at 12 mos (%)|
|None||56.0 ± 5.6||100||91.70|
|Weak||28.2 ± 5.4||81.30||37.50|
|Moderate||9.4 ± 0.8||13.90||0|
|Strong||10.3 ± 1.8||33.30||0|
|Percentage of positive cells|
|None||56 ± 5.6||100||91.70|
|1–30||24.2 ± 5.0||53.90||32.30|
|31–60||12.0 ± 1.4||46.20||0|
|61–100||7.3 ± 0.5||0||0|
An increasing body of evidence suggests a role for PTHrP in the modulation of tumor progression in a wide variety of tumors, including neoplasms of the breast, thyroid, colon, pancreas, testis, liver, glial cells, and melanocytes.9, 15–24 PTHrP is differentially regulated not only in neoplastic cell types but also in phenotypically normal cells and during normal embryonic development.8, 11, 35 PTHrP and its receptor are widely expressed in the mammalian CNS. PTHrP is expressed in primary human astrocytes, meningeal cells, and in some astrocytomas.15 Recent studies have demonstrated PTHrP expression in both immature and malignant adult astrocytes, but not in normal astrocytes, a finding that suggests that PTHrP expression in the brain may arise from the supporting glial cells.39 In cerebellar granule cells, as in smooth muscle, PTHrP appears to negatively regulate voltage-sensitive calcium channels.40 Because PTHrP is expressed in phenotypically normal glial cells and neurons, it is intriguing to speculate that this peptide is a mediator of crosstalk between glial cells and neurons.15 PTHrP knockout mice exhibit a systemic chondrodysplasia that is lethal at birth. Transgenic replacement of PTHrP rescues these mice from death.31 These mice are six times more sensitive than controls to kainic acid, and the addition of PTHrP was protective against kainite toxicity. PTHrP-mediated calcium influx into the granule cells protects these cells against the excitotoxic effects induced by kainic acid.27 Thus, expression of PTHrP may prevent neuronal cell death via calcium channel regulation.41 PTHrP expression is also induced during CNS inflammation or injury in reactive astrocytes.42 These reactive astrocytes have an immature phenotype, as evidenced by their reexpression of fetal markers.43 Together with tumor necrosis factor-alpha, PTHrP induces the expression of interleukin-6, a cytokine with documented neuroprotective effects.43 Because astrocytes play such a major role in CNS inflammation and injury, it is possible that PTHrP expression may be altered in glial tumor cells after irradiation.
Despite numerous studies demonstrating the overexpression of PTHrP in tumor cells, studies addressing the potential effects of PTHrP expression on human solid tumor progression and therapeutic response are scarce.21 Most published reports emphasize the role of PTHrP in the regulation of calcium-mediated osteogenic signal transduction pathways, with this role often being clinically manifested in the form of serum hypercalcemia.11, 44 However, only relatively recently has the importance of differential expression of PTHrP in tumor progression been postulated.9, 16, 21, 28 Although calcium and other ions are involved in signal transduction in both developmental and tumorigenic cellular proliferation, the role of calcium in the regulation of cell cycle kinetics and genomic instability are less apparent. Studies of ras/mutant p53–related primary rat embryo cell transformation suggest that PTHrP expression may be transcriptionally regulated by ras and/or p53.45 Similarly, inhibition of retinoblastoma function also modulates levels of both PTH and PTHrP.46 Unlike dominant or tumor suppressor function, however, PTHrP function appears to hold a downstream position in the signal transduction cascade.
Our data support a role for PTHrP expression in the modulation of human astrocytic tumor progression. Several studies describe overexpression of PTHrP in human primary glial cells, glial tumors, and reactive glial cells.25, 26, 39, 43 In one study, neurons, normal glial cells, endothelial cells, reactive astrocytes, and glial tumors all exhibited varying degrees of expression of PTHrP ligand and PTH/PTHrP receptor.26 There was a trend toward PTHrP overexpression in astrocytic glial tumors compared with oligodendroglial neoplasms. The current study reveals a correlation between PTHrP expression and clinical course after irradiation in patients with astrocytoma. Our data suggest that autocrine and/or paracrine regulation of PTHrP metabolism possesses clinical relevance in terms of glial tumor progression after the administration of ionizing radiation in vivo. PTHrP was an independent prognostic factor with respect to local control and OS in patients with astrocytic tumors. In three-variable survival models, PTHrP expression retained its statistical significance when compared with tumor grade and Karnofsky functional status. It is not surprising that tumor grade was by far the strongest predictor of astrocytic tumor local control and survival. In fact, throughout a number of studies on astrocytic tumor progression, grade has remained the most straightforward and reproducible independent variable in most statistical models.47–49
The mechanism by which PTHrP overexpression may lead to increases in recurrence-free survival in patients with astrocytic tumors is unclear. Data regarding the effects of irradiation on PTHrP expression are scarce. To our knowledge, there are no published reports characterizing PTHrP expression after irradiation in any of the cell types present in the CNS. In chondrocytes, prolonged arrests and delays in cell growth after irradiation result in increases in PTHrP expression.35 Indeed, it may be the case that alterations in PTHrP expression and the regulation of calcium metabolism are not a cell type– or tissue-specific biologic process, but rather a more global mediator of cell metabolism, proliferation, and apoptotic capacity.
PTHrP expression may be a downstream mediator of signal transduction via the Ras/Raf/MEK/extracellular signal–related kinase pathway.2, 50 After radiotherapy-induced DNA damage in astrocytes, PTHrP function may be altered, not only with respect to autocrine and/or paracrine mechanisms, but also with respect to transcriptional and translational control of radiotherapy-induced signal transduction. On this basis, alterations in PTHrP signal transduction were mediated by protein kinase C, via activation of the Ras/mitogen-activated protein kinase signaling pathway.2, 3, 46, 51 Similarly, alterations in PTHrP signaling can have effects on transforming growth factor-beta and can lead to functional changes in activator protein-1, nuclear factor-κB, E26 transformation-specific sequence-1, and specificity protein-1 signaling.32, 34 PTHrP also can inhibit the apoptotic effects of functional mutations in fibroblast growth factor, resulting in decreases in Bcl-2 expression and apoptosis.52 Finally, PTHrP can be a potent tumor angiogenic factor.7, 29 Although the underlying mechanism was not readily apparent, PTHrP expression was found to modulate VEGF expression.7 It is noteworthy that vascularity is a hallmark of glial tumor grade, the independent factor most strongly correlated with survival in the current series.36 Thus, PTHrP-mediated calcium signal transduction may lead to tumor progression via cell cycle proliferative changes, cell survival, inhibition of apoptosis, and angiogenic mechanisms.
PTHrP staining may be an important adjunct to standard immunopathologic criteria in the determination of how glial tumors respond to treatment. In the current study, increased staining intensity and a higher proportion of positively stained cells both were correlated with decreased PFS and OS. Patients who had tumors with moderate-to-strong staining for PTHrP had significantly poorer outcomes than did patients who had tumors that stained weakly or did not stain at all. Furthermore, as the percentage of positive cells increased, there were corresponding decreases in OS and PFS. The overexpression of PTHrP in glial tumors after irradiation appears to be in agreement with previous studies on PTHrP. Expression of PTHrP has been documented in immature/fetal and malignant astrocytes, but not in normal adult astrocytes. The overexpression of PTHrP in patients with glial tumors may be an indication of the loss of differentiation, leading to a more aggressive phenotype. The finding of inverse correlations between increased PTHrP expression and local control and survival in patients with astrocytoma may have therapeutic implications. Guanosine nucleotides inhibit a variety of syndromes caused by increased PTHrP expression in human tumors.10 In addition, inhibitors of ras function have exhibited activity against human malignancy–associated hypercalcemia.50 The roles of both ras-farnesylation inhibitors and guanosine nucleotides in inhibiting PTHrP-mediated astrocytic cell proliferation are intriguing. In spite of these novel results, tumor grade remained the most important predictor of OS and PFS in patients who had glial tumors that were definitively treated with radiotherapy. Additional prospective studies involving an expanded patient population will be necessary before clinical applications are initiated.
The authors thank Dr. Matthew Spear and Dr. Stephen Seagren for their expertise and for their assistance with the current study.
- 8Expression patterns of parathyroid hormone-related peptide (PTHrP) and parathyroid hormone receptor type 1 (PTHR1) during human development are suggestive of roles specific for each gene that are not mediated through the PTHrP/PTHR1 paracrine signaling pathway. Gene Expr Patterns. 2003; 3: 59–63., , .