Patients with primary hyperparathyroidism had a greater risk of renal stone disease even 10 years before the diagnosis was registered than did a population based control group matched for sex and age. This is in accordance with the higher risk of fractures observed in the same population 10 years before diagnosis.12
These findings suggest that the disease has started several years before diagnosis and emphasises the importance of early diagnosis and treatment. This view is further supported by the finding of a higher risk of stone events before diagnosis. Parathyroid adenomas were larger and preoperative serum calcium concentrations were higher in our patients than in other surgical series,14,15
supporting the theory of a delay in diagnosis.
However, other explanations of our findings regarding the delay in diagnosis of renal stone disease may exist. Because plasma calcium concentrations are routinely determined in Danish patients with renal stones, our findings may resurrect the possibility of symptomatic normocalcaemic hyperparathyroidism, an entity widely debated in the 1970s.7,16
Another possibility is that idiopathic renal hypercalciuria may, in some cases, be the cause of later primary hyperparathyroidism.5
This would explain why renal stone disease occurs many years before diagnosis, and it agrees with the persistent increase in the risk of renal stones several years after parathyroid surgery, especially among patients who previously had stones.. This interpretation presumes that the primary renal calcium leak continues after parathyroidectomy. Our findings of an increased risk of fractures up to 10 years before diagnosis12
does not contradict this theory because renal calcium loss may cause increased bone turnover and bone loss, leading to decreased skeletal integrity.
Relation between diseases
Our data do not unequivocally indicate a causal relation between primary hyperparathyroidism and renal stone disease. In evaluating a relation between two common diseases such as renal stone disease and primary hyperparathyroidism, the possibility of confounding by indication should be considered. The diagnosis of primary hyperparathyroidism depends on plasma calcium concentration. The risk of being diagnosed with a diagnosis of primary hyperparathyroidism would be more likely in patients who had a renal stone event than in controls, who are not submitted to systematic plasma calcium screening (Berkson's bias). The peak in hospital admissions for renal stones around the time of diagnosis corroborates this interpretation.
The epidemiological observation of an increased risk of renal stone disease many years before any clinical suspicion of primary hyperparathyroidism also indicates that biological associations exist between primary hyperparathyroidism and renal stone disease. The fact that a substantial proportion of patients never experience renal stones, and the observation that the risk of stone formation is decreasing with the growing awareness of the disease,2,17
suggest that there are factors that modulate the biological relation between primary hyperparathyroidism and renal stone disease.
Our study did not corroborate the previous finding that patients with nephrolithiasis had smaller tumour weight and lower serum calcium concentrations than patients with skeletal disorders.7,18
In the study by Lloyd, the specific hyperparathyroid disorder of osteitis fibrosa was used as an indicator of skeletal disease,18
whereas we used the occurrence of skeletal fractures as an indicator.
Surgery and risk
After surgical treatment of primary hyperparathyroidism, hospital admissions due to stone episodes were reduced. Without randomisation to surgery or no surgery, one cannot determine whether this reduction in stone episodes is caused by the treatment, by the natural course of the disease, or by the synchronisation effect induced by a possible Berkson's bias. In our study, the risk of new renal stone episodes before, as well as after, parathyroid surgery was lower with older age. However, the reduction of stone events after diagnosis and surgery was significantly higher than could be explained by the average increase in age during follow up, suggesting a treatment effect of around 8%. The risk of hospital admission because of stone disease returned to normal more than 10 years after surgery (fig ).
Patients also had a higher risk of strictures of the ureters and of hydronephrosis than the controls both before and after surgery. This could indicate that stones caused anatomical damage to the urinary tract. These structural changes could increase the risk of having stones subsequently even though the biochemical abnormalities had been normalised by surgery.
Our data accord with recent studies that have shown that patients with primary hyperparathyroidism and renal stones may continue to produce stones after parathyroid surgery even though normocalcaemia has been established.10,11
Patients with renal stones who have primary hyperparathyroidism have a higher renal calcium excretion than those without stones who have similar concentrations of plasma calcium.19
Finally, three years after parathyroidectomy, patients who continue to form stones have higher average renal calcium excretion than those without stones. This suggests again that idiopathic renal hypercalciuria could have a pathogenic role in patients with recurrent stone formation.20
Primary hyperparathyroidism is associated with an increased risk of renal stones more than 10 years before the diagnosis is established. Parathyroidectomy is associated with an 8.3% risk reduction in renal stone events, and more than 10 years after surgery the risk returns to that of controls. Male sex, younger age, and previous stone events are associated with a greater risk of stone disease. The study does not support the concept of dividing primary hyperparathyroidism patients into separate clinical entities of patients who form stones and patients with skeletal complications.