Tumour-induced osteomalacia (TIO), or oncogenic osteomalacia, is a paraneoplastic syndrome marked by hypophosphataemia, renal phosphate wasting, bone pain, weakness, and fractures. The syndrome has been reported with both benign and malignant tumours including parotid gland basal cell tumours, thyroid carcinomas, colon adenocarcinomas, and prostate cancer. Often, the syndrome is marked by an insidious course during which patients present with generalised bony pain and weakness, which do not resolve until the underlying tumour is identified and treated. We present a case of a patient with Parkinson’s disease whose subacute weakness, lower extremity paresis, and renal phosphate wasting led to the synchronous diagnosis of metastatic prostate adenocarcinoma and TIO.
- calcium and bone
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
Tumour-induced osteomalacia (TIO), or oncogenic osteomalacia, is a rare paraneoplastic syndrome featuring bone mineralisation imbalance and marked renal wasting of phosphate, the latter driven by increased levels of proteins such as fibroblast growth factor-23 (FGF-23), secreted frizzled related protein-4 (SFRP4) and matrix extracellular phosphoglycoprotein (MEPE), which are produced by an underlying tumour.1 The associated tumours are soft tissue or bone tumours of mesenchymal origin (phosphaturic mesenchymal tumour mixed connective tissue variant), which may be part of a known cancer diagnosis or may only be revealed after a thorough search for a tumour once the clinical features of the disease are recognised.2–7 The clinical features of this syndrome include decreased serum phosphorus levels, increased urinary excretion of phosphate, inappropriately low-to-normal 1,25-dihydroxyvitamin D levels and musculoskeletal weakness. These features resolve once the underlying tumour is removed. In this case report, we present a patient with long-standing Parkinson’s disease, a 2-month history of progressive weakness, difficulty walking and abdominal pain who was found to have a metastatic prostate adenocarcinoma-induced osteomalacia.
A 67-year-old man with a 20-year history of Parkinson’s disease was brought to the emergency department (ED) for progressive leg weakness. He had been in his usual state of health 2–3 months ago, walking up to 2 miles daily and performing activities of daily living (ADLs) independently. Subsequently, he became progressively weaker and experienced dizziness, dyspnoea on exertion, inability to walk unassisted, and difficulty with ADLs. Two weeks prior to presentation, due to progressive difficulty urinating and left lower abdominal quadrant pain, he had visited an ED at an outside hospital. There, a CT (computed tomography) abdomen and pelvis scan revealed significant constipation requiring disimpaction, and, incidentally, multiple sclerotic bone masses in the spine and pelvis suspicious for metastatic disease. He had lost 20 lbs over the last 2 months, noting decreased intake due to issues with weakness and manual dexterity. Colonoscopy 3 years prior had been negative for malignancy, and although he was a former smoker, imaging did not support a pulmonary malignancy. The prostate-specific antigen (PSA), however, had recently increased from the baseline level of <3 to >9 ng/mL in the prior 3 months (figure 1). He was discharged home from that ED with plans for follow-up with his primary physician but sought urgent medical attention at our facility as his weakness had progressed substantially.
On the day of presentation, his physical exam was notable for motor weakness in all extremities, specifically with an inability to raise his legs off the bed, suprapubic fullness and decreased rectal tone. A bladder scan indicated greater than 1 L of fluid and Foley catheter insertion drained 2 L of urine. A basic metabolic panel showed blood urea nitrogen (BUN) of 135 mg/dL and a creatinine of 24.11 mg/dL (up from normal values obtained at the previous ED 2 weeks ago). Urinalysis was notable only for monomorphic red blood cells, likely from a traumatic Foley insertion. An MRI (magnetic resonance imaging) of the lumbar spine was urgently obtained and showed no evidence of cauda equina syndrome, although it did demonstrate multiple hypointense lesions scattered throughout the lumbar spine, upper sacrum and medial aspects of both iliac bones. The Foley catheter was kept in place and he was admitted to the floor for postobstructive acute kidney injury (AKI), acute lower extremity paresis and osseous lesions of unknown aetiology. Within 24 hours of relief of the bladder outlet obstruction, his BUN and creatinine down-trended to normal values, in the setting of 7 L urine excretion (table 1). Interestingly, after relief of obstructive AKI, weakness persisted and recurrent hypophosphataemia was noted despite intensive phosphate repletion (figure 2). Further neurological evaluation confirmed that this presentation was not consistent with exacerbation of his underlying Parkinson’s disease.
Hypophosphataemia can be approached pathophysiologically from the perspective of acute phosphate shifts, decreased intake, decreased gastrointestinal absorption or increased renal wasting. Refeeding syndrome with acute phosphate shifts was considered initially given the patient’s history of poor oral intake, weight loss and the initially low electrolyte values (magnesium, phosphate and potassium). However, while the patient’s potassium and magnesium levels responded to supplementation, phosphorus did not, and refeeding syndrome was not felt to be a possible cause. In terms of evaluating other differential diagnostic possibilities, decreased intake alone is rarely a cause of hypophosphataemia, and intestinal malabsorption was not supported by the patient’s history. No offending medications were found in the patient’s history that would promote intestinal malabsorption. Thus, further investigations were initiated to identify the presence of renal phosphate wasting. Renal phosphate wasting secondary to tubular dysfunction and postobstructive polyuria were strongly considered as the main cause of hypophosphataemia initially. However, despite resolution of polyuria over time, intensive phosphate repletion was still required to maintain phosphate levels and additional aetiologies of renal wasting were explored.
Hypophosphataemia was approached by looking first at the renal response in the face of hypophosphataemia. Urine studies were striking for 24-hour phosphate excretion of >1.5 g and a calculated fractional excretion of phosphate (FePi) of 34.97% (normal<5%) (table 2). This confirmed renal wasting of phosphorus. Serum calcium values were concurrently measured and when corrected for albumin, were normal between 8.6 and 9.4 mg/dL (table 2). Intact parathyroid hormorone (iPTH) was 84 pg/mL and confirmed no evidence of primary hyperparathyroidism as an aetiology of renal phosphate wasting. Urinalysis as noted was bland without proteinuria or glucosuria to suggest a primary tubular defect. Curiously, 1,25 dihydroxy (OH) vitamin D levels were measured and showed a normal value of 38.9 pg/mL in the face of concomitant hypophosphataemia (table 2). Based on this inappropriately low 1,25-dihydroxy vitamin D level, FGF-23 (fibroblast growth factor - 23) levels were drawn and found to be elevated at 392 RU/mL (reference range 44–215), confirming the diagnosis of FGF-23 excess, potentially due to tumour-induced osteomalacia.
To investigate the extent and metabolic activity of the bony lesions, a 2-deoxy-2-[fluorine-18] fluoro-d-glucose integrated with CT whole body scan was performed. This revealed several mediastinal and bilateral hilar foci of increased uptake, suspicious for reactive lymph nodes, as well as multiple sclerotic lesions in the thoracolumbar spine, pelvis, right iliac and right scapula, all with mild to no increased uptake. Evaluation for potential multiple myeloma revealed no evidence of a monoclonal gamma spike and a normal ratio of kappa and lambda (ratio 1.25) despite elevated absolute levels of kappa and lambda light chains (table 1).
At this point, tumour-induced osteomalacia (TIO) with unconfirmed primary malignancy was suspected as the aetiology of the hypophosphataemia, and the patient underwent biopsy of the lesion in the right iliac crest to confirm the diagnosis. The bone marrow biopsy results returned 2 weeks later revealing areas with trilineage haematopoiesis as well as areas of fibrosis containing crushed cells which stained positive for PSA, prostate specific acid phosphatase, prostate specific membrane antigen, Nk3 homeobox 1, cytokeratin AE1/AE3 and cytokeratin CK8/18, all consistent with prostate cancer. A final diagnosis of metastatic carcinoma compatible with prostate primary was made based on this histology and the clinical presentation.
The patient was started on leuprolide and daily bicalutamide for treatment of metastatic prostate cancer. A repeat PSA obtained 1 week after initiation of leuprolide, revealed lowered PSA levels to 5.881 ng/mL, with no discernible change in bladder outlet obstruction, therefore the patient continued to rely on the use of an indwelling Foley catheter. Standing repletion of sodium phosphate was doubled to 2 packets of K-Phos Neutral three times a day (78 mEq sodium, 1.5 g phosphorus and 6.6 mEq potassium daily). Twice daily dark soda was added to his diet for additional phosphorus (94 mg for two 12 oz. dark sodas daily)8 and vitamin D supplementation was provided.
Outcome and follow-up
A few days after the biopsy, the patient had achieved stable serum phosphate levels for three consecutive days on his aforementioned electrolyte-repletion regimen and was discharged to a skilled nursing facility. Upon discharge, patient discontinued the daily sodium-phosphate packets and used only ad lib dark soda and active vitamin D supplementation to maintain normal serum phosphate levels. With physical therapy, the patient was able to start taking a few independent steps out of his wheelchair 1 month after discharge from the hospital. More than a month after discharge, he continued to maintain a normal serum phosphate level of 3.9 mg/dL.
The case illustrates how a very morbid diagnosis (tumour-induced osteomalacia) can masquerade under the guise of a relatively non-specific symptom complex of weakness and the relatively ubiquitous electrolyte abnormality of hypophosphataemia. Failure to diagnose hypophosphataemia as the source of weakness and the subsequent aetiology of hypophosphataemia (tumour-induced osteomalacia) in a systematic manner would have resulted in further delay in diagnosis of prostate cancer and treatment.
TIO was first described unwittingly in 1947 by McCance who reported the case of a young woman with an 8-year history of progressive muscular weakness, pseudofractures on imaging and chronic hypophosphatemia of unclear aetiology, who presented to the physician barely able to walk. He would go on to initiate an extremely high dose of oral vitamin D therapy while simultaneously locating and removing a tumour from the femur. With removal of the tumour, which consisted of degenerated osteoid tissue, the woman experienced recovery of her strength and serum phosphate levels. While McCance attributed his case report to acquired resistance to vitamin D,4 a few years later, Prader et al, 9 were credited with publishing the first evidence linking an unknown phosphaturic substance released from a tumour to the development of rickets in a young child. Since then, it has been discovered that FGF-23, SFRP4 and MEPE are all so-called ‘phosphatonins’10 that can cause phosphaturia associated with tumour-induced osteomalacia.1 11–14 Of these phosphatonins, FGF-23 is thought to be the most common and well-studied.
FGF-23 levels are elevated in tumour-induced osteomalacia relative to several other causes of hypophosphataemia including Fanconi’s syndrome, isolated transporter defects or vitamin D deficiency.15 While the role of FGF-23 in the metabolism of phosphate and Vitamin D remains an active area of investigation,16 17 there is now improved understanding of FGF-23 physiology. Homozygous FGF-23-mutant mice have demonstrated that there is increased tubular maximum transport of phosphate/GFR (TmP/GFR), hyperphosphataemia, increased serum 1,25-dihydroxy vitamin D and calcium levels, stunted growth and abnormal bone development in the total absence of FGF-23.17 Not surprisingly, the converse is true: excess FGF-23 protein causes a hypophosphataemic syndrome in the mice. Specifically, transgenic mice that overexpress FGF-23 protein develop hypophosphataemic rickets, decreased 1,25-dihydroxy vitamin D levels and renal phosphate wasting (increased FePi), similar to the syndrome seen in patients. Furthermore, the FGF-23 transgenic animals additionally show stunted growth, abnormal cranio-facial development, and, in females, infertility.18 Importantly, the FGF-23-mediated renal wasting in the murine model is associated with reduction in the expression of sodium phosphate cotransporter type IIa (NaPi-2a), a transporter in the proximal tubule that is required for reabsorption of phosphate in the kidneys, without significant changes in PTH (parathyroid hormone) or calcium levels.18 Further work has revealed that FGF-23 binds two receptors, FGFR1 and FGFR4, in the proximal convoluted tubule, to cause downregulation of NaPi-2a.19 Thus, FGF-23 regulation of phosphate wasting via the kidneys may depend on the NaPi-2a transporter. Humans with markedly enhanced FGF-23 activity (seen in autosomal dominant hypophosphataemic rickets, X-linked hypophosphataemic rickets and autosomal recessive hypophosphataemic rickets) develop rickets, osteomalacia, hypophosphataemia, renal phosphate wasting, bone pain and deformed lower extremities.16 20 21
In tumour-induced osteomalacia, the source of the FGF-23 is the tumour. While initially described in association with benign tumours of mesenchymal origin, some malignant mesenchymal tumours have also been associated with the syndrome. It has been proposed that all the mesenchymal tumours, whether benign or malignant, that give rise to tumour-induced osteomalacia have common histological features that may aid in their identification and allow grouping under the single category of phosphaturic mesenchymal tumour-mixed connective tissue variant.12 22 Interestingly, tumour-induced osteomalacia has also been reported in association with malignant neoplasms such as prostate adenocarcinoma,23–31 anaplastic thyroid carcinoma32 and colon adenocarcinoma,33 all of which are non-mesenchymal tumours but undergo epithelial to mesenchymal transitions during metastases. How these carcinomas and mesenchymal tumours are linked and converge in tumour-induced osteomalacia is not fully understood but as shown in this case report, can cause substantial pleotropic effects beyond the burden of tumour.
Clinically, these tumours should be suspected in a patient with low or inappropriately normal 1,25-dihydroxy vitamin D level in the setting persistent hypophosphataemia of renal origin. Unlike other disorders associated with hypophosphataemia, 1,25-dihydroxy vitamin D production is uniquely not increased in patients with FGF-23 overproduction as FGF-23 downregulates 1,25-dihydroxy production.33 Recognition of this key point is often the major branch point in allowing the physician to suspect overproduction on of FGF-23 and to begin the search for an underlying tumour.
In summary, we present a case of tumour-induced osteomalacia where a systematic approach to underlying hypophosphataemia revealed malignant prostate cancer as a cause of FGF-23 production and ultimately led to timely treatment.
Severe hypophosphataemia that is difficult to control should prompt a clinical workup in a systematic manner measuring renal excretion of phosphate to determine if renal phosphate wasting is present.
Low or even normal 1,25-dihydroxy vitamin D levels in the face of hypophosphataemia and renal phosphate wasting should prompt clinical suspicion of fibroblast growth factor-23 overproduction and tumour-induced osteomalacia.
Tumour-induced osteomalacia is not limited to benign mesenchymal tumours and may be associated with malignancy.
Treatment of tumours associated with tumour-induced osteomalacia may lead to resolution of symptoms and improvement of hypophosphataemia, as in this case report.
The authors would like to thank their patient and his family for consent to use his experience as a teaching tool for others.
Contributors AAKL wrote the article with revisions from SJ and SS. All three authors were involved in providing care for the patient.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.
Patient consent for publication Next of kin consent.