Severe hypophosphataemia following oral bisphosphonate treatment in a patient with osteoporosis

Background

Bisphosphonates are commonly used in pharmacological treatment of osteoporosis. They have an antiresorptive effect on bone metabolism which, potentially, affects both calcium and phosphate homeostasis. Mild and transient decreases in p-calcium and p-phosphate are thus frequent side effects.1 Few cases of bisphosphonate-induced severe hypophosphataemia have been reported but only following intravenous administration of the potent bisphosphonate, zoledronic acid.2–4

Severe hypophosphataemia is defined as p-phosphate <0.3 mmol/L. It is a potentially fatal condition where symptoms include bone pain, muscle weakness, rhabdomyolysis, arrhythmias, convulsions, confusion and loss of conscious.5

We report a clinically relevant case of two consecutive episodes of severe hypophosphataemia following oral administration of alendronate in a challenge–dechallenge–rechallenge scenario. This is to our knowledge the first case of severe hypophosphataemia following oral bisphosphonate treatment in a patient with osteoporosis.

Case presentation

A 76-year-old woman was referred to the osteoporosis outpatient clinic due to a random finding of a non-traumatic vertebral compression fracture on a CT scan (figure 1). She had earlier been diagnosed with a microcytic (82 mmol/L), hypochromic (19.1 mmol/L) anaemia (haemoglobin 108 g/L) due to iron deficiency (ferritin 21 mg/L) and undergone investigations, including the CT scan, for malignancy and malabsorption without any findings. The patient was previously supplemented with oral iron supplements (ferrous sulfate 200 mg daily).

Furthermore, the patient had a history of hypertension, hypercholesterolemia, type 2 diabetes and ischaemic heart disease for which she was treated with amlodipine, enalapril, atorvastatin, metformin, vildagliptin and acetylsalicylic acid.

In the outpatient clinic, a dual-energy x-ray absorptiometry (DXA) showed osteopenia (T-score lumbar spine (L2–L4) −2.1, femoral neck left −2.4 and right −2.2, total hip left −2.2 and right –2.3) but due to the significant compression fracture of the 12th thoracic vertebra the patient was diagnosed with manifest osteoporosis, yet without clinical symptoms. The aetiology was evaluated to be postmenopausal osteoporosis as secondary causes were excluded, such as primary hyperparathyroidism, thyrotoxicosis, chronic kidney disease, Chron’s disease or treatment with glucocorticoids or other potential medications. The patient was prescribed with oral alendronate 70 mg once weekly as well as magnesium hydroxide 360 mg, calcium carbonate 800 mg and cholecalciferol 20 μg daily.

Follow-up after 4 months showed good patient compliance which was confirmed with the finding of suppressed bone markers (p-P1NP and p-CTx) and the patient experienced no subjective side effects. Pharmacy records confirmed collection of alendronate and did not suggest overdosage (figure 1). However, blood tests surprisingly revealed severe hypophosphataemia (p-phosphate 0.20 mmol/L, reference range 0.76–1.41 mmol/L, table 1).

The patient was hospitalised for acute treatment. Her only problem was pain in both feet. There was no prior history of diarrhoea, nutritional issues or alcohol misuse. The patient did not use antacids, proton-pump inhibitors or diuretics. Physical examination and vital parameters were all normal. ECG and telemetry (18 hours) were without signs of arrhythmias. Blood tests reconfirmed severe hypophosphataemia (table 1). Phosphate levels were restored within 3 days during hospitalisation and alendronate treatment, as well as calcium and vitamin D supplements, were paused. The patient was discharged to outpatient follow-up (figure 1).

Short follow-up after discharge only showed slightly elevated phosphate levels (table 1). Four months after the episode of severe hypophosphataemia, plasma phosphate levels were normal and alendronate, calcium and vitamin D supplements were resumed as treatment for osteoporosis (figure 1). One month after reintroduction of alendronate, the patient once again presented unaffected with severe hypophosphataemia (p-phosphate<0.16 mmol/L), but reported pain and muscular weakness of left hip and leg and difficulty walking. The hypophosphataemia was corrected during hospitalisation and alendronate treatment discontinued (figure 1).

Investigations

Hypophosphataemia was never observed in this patient prior to alendronate treatment. Due to her history of iron deficiency, the patient was previously investigated for malignancy and malabsorption without any findings including normal gastroscopy and colonoscopy, normal thoracic/abdominal CT-scan, negative M-component, negative screening for coeliac disease and normal levels of faecal calprotectin and elastase.

Baseline blood tests prior to alendronate treatment were all within the normal range despite a random finding of a mild hypomagnesaemia (table 1).

During both episodes of severe hypophosphataemia and in subsequent follow-up, a complete set of blood tests were obtained in order to evaluate phosphate metabolism, including p-phosphate, p-ionised calcium, p-parathyroid hormone (p-PTH), p-25-OH-D, p-magnesium and estimated glomerular filtration rate (table 1). Laboratory values for alkaline phosphatase, potassium and indicators of haemolysis, such as lactate dehydrogenase, bilirubin and haptoglobin, were all within the normal range.

When the patient presented with severe hypophosphataemia the first time, blood tests aside from confirmation of severe hypophosphataemia showed ionised calcium levels high within the reference range or slightly elevated, suppressed p-PTH levels, and the pre-existing hypomagnesaemia. Phosphate levels were quickly normalised and after phosphate replacement therapy, a transient hyperphosphatemia was observed. The same tendency in the blood tests was seen during the second episode of hypophosphataemia.

Urine-phosphate and fibroblast growth factor-23 levels were not evaluated (see the Discussion section).

Due to the patient history of iron deficiency and a persisting mild hypomagnesaemia, malabsorption was considered as a cause for hypophosphataemia. Therefore, indicators of malabsorption, such as albumin levels (35–40 g/L), fecal-calprotectin (41 mg/kg), fecal-elastase (398 µg/g) and tissue transglutaminase antibodies (negative), were re-evaluated. Patient body mass index (BMI) was calculated to 22.2.

Differential diagnosis

In a case of severe hypophosphataemia, several mechanisms and hence underlying differential diagnoses must be considered. In general, there are three major mechanisms by which hypophosphataemia can occur: redistribution of phosphate into cells from extracellular fluids, decreased intestinal absorption of phosphate and increased urinary phosphate loss.5 6

Redistribution of phosphate can be observed in refeeding syndrome, ketoacidosis, hyperosmolar hyperglycaemic, sepsis and following insulin treatment. Our patient did not, at any point suffer from sepsis, critically illness, or starvation. Even though the patient was diabetic, blood glucose levels remained within the normal range during the episodes of hypophosphataemia and the patient did not receive insulin at any time. Consequently, redistribution of phosphate as a mechanism for the severe hypophosphataemia was ruled out.

Decreased intestinal absorption of phosphate is linked to inadequate intake, alcohol misuse, chronic diarrhoea, vitamin D deficiency, use of phosphorous-binding antacids and malabsorption.

However, the patient confirmed a varied diet containing dairy products, had a normal BMI, had no history of alcohol misuse and had a normal stool pattern. Vitamin D deficiency was ruled out in the blood tests and she did not use phosphorous-binding antacids or proton-pump inhibitors.

Malabsorption was considered in the case as the patient also had a history of iron deficiency and hypomagnesaemia. However, previous and repeated investigations for malabsorption, including coeliac disease, were all negative. Albumin levels were within the normal range. Furthermore, both episodes of severe hypophosphataemia were predominantly corrected by oral phosphate replenishment. Consequently, malabsorption did not seem likely as the only cause for hypophosphataemia, but could not be ruled out entirely.

Increased urinary loss of phosphate is seen following primary hyperparathyroidism, vitamin D deficiency, diuretics and some rare genetic disorders. Hyperparathyroidism and vitamin D deficiency were not present and the patient denied any family history of calcium or phosphate disturbances. As the patient had never experienced hypophosphataemia prior to the alendronate treatment, genetic disorders were considered very unlikely.

Drug-induced hypophosphataemia was strongly considered. The patient did not receive any antacids, proton-pump inhibitors, chemotherapy, diuretics or steroids that could cause drug-induced hypophosphataemia. Moreover, she was not using iron supplementations during the times hypophosphataemia did occur. However, the fact that hypophosphataemia first developed following alendronate treatment (challenge), no episodes of hypophosphataemia was observed during the period where alendronate was paused (dechallenge) and hypophosphataemia reoccurred when alendronate was resumed (rechallenge), posed a strong indication that bisphosphonate-induced hypophosphataemia was the most likely diagnosis.

Treatment

During the first hospitalisation, the patient received oral phosphate replacement therapy (20 mmol three times a day with an oral solution containing a combination of potassium dihydrogen phosphate and disodium phosphate dihydrate). p-Phosphate was restored after 2 days and the patient was discharged on day 3. The second episode of severe hypophosphataemia was initially treated with intravenous phosphate (20 mmol once with Addiphos, a combination of potassium dihydrogen phosphate and disodium phosphate dihydrate) the first day and followed by oral phosphate replacement (20 mmol two times per day). p-Phosphate was normalised within 2 days.

Outcome and follow-up

The patient survived both episodes of severe hypophosphataemia and ongoing follow-up in the outpatient clinic was established. After the second episode of hypophosphataemia, alendronate was registered as a medical CAVE to indicate that the patient did not tolerate bisphosphonate treatment. As a future treatment for osteoporosis, denosumab (Prolia) was chosen. Interestingly, follow-up blood tests after the first dose of denosumab, showed a mild and transient hypophosphataemia. This episode of hypophosphataemia was untreated and spontaneously reversed after 2 weeks.

The patient later on represented with iron deficiency and moderate asymptomatic hypophosphataemia and was referred to gastroenterological re-evaluation for malabsorption. Once again, faecal parameters and coeliac screening were negative, however, malabsorption could not be ruled out at this point.

After correction of hypophosphataemia, the patient’s symptoms of lower extremity pain, muscle weakness and difficulty walking persisted and the cause was re-evaluated. During hospitalisation, an ultrasound of the hip, a CT scan of the pelvis and MRI of the lumbar spine were performed (figure 1). The ultrasound and CT scan were both normal but the MRI of the lumbar spine revealed severe lumbar spinal stenosis for which the patient underwent surgery relieving the symptoms completely.