We found only weak correlations between pre- and postoperative biochemical variables and adenoma weight. Although adenoma weight was moderate correlated to preoperative PTH levels, it was not strong enough to be used in the estimation of adenoma weight. Adenoma weight correlates strongly with its maximum diameter, i.e. with its size. This makes possible to use either variable for determining correlation with biochemical markers [8].
The normal parathyroid gland has a mean maximum diameter of 6 mm and a mean weight of 60 mg [8]. In our sample, the mean maximum diameter was 1.8 cm and the mean weight 1000 mg (1 g), values consistent with all western studies. A study from Iran reports much heavier adenomas, attributing the mean weight of 3.8 g to the effect of vitamin D deficiency and low calcium intake on parathyroid glands. The former seems to be the reason why patients in Iran are diagnosed at much younger age compared to the western countries (42 years vs. 60 years in our study) [7].
Several studies could not demonstrate any relationship between serum calcium and PTH and adenoma weight, although these findings may be influenced from patients with hyperplasia, double adenomas and renal hyperparathyroidism, who were not excluded [8]. Kandil et al., in their impressive series of 447 cases, found significant mean weight differences (410 vs. 910 mg) in patients with low and high baseline PTH values (i.e. PTH < or > 150 pg/ml, respectively) [6]. However, they made a comparative study and did not explore correlations in the whole cohort. Strickland et al. report no difference between hypocalcemic and normocalcemic patients with respect to preoperative serum calcium, PTH levels or adenoma weight. They also found higher mean preoperative calcium levels but comparable preoperative PTH and postoperative calcium levels in patients with adenomas > 2 g [9]. However, their method of categorizing a continuous variable may have led to biased parameter estimates and loss of efficiency in predictions [10]. Randhawa et al. dichotomised at several thresholds from 1 to 2.5 g but still failed to identify any adenoma weight predictors.
It is reported that larger parathyroid adenomas secrete PTH at a lower rate than lighter adenomas. This may be explained from the fact that larger adenoma may be filled with inactive zones, i.e. fibrosis, calcification, cystic spaces or hemorrhage into the gland [7, 11]. Another explanation could be that the excess autonomous PTH secretion, triggers negative feedback mechanisms, suppressing PTH secretion by the normal glands. On the other hand, it is not clear if and at which stage of the disease the normal glands cease secretion [8]. Stern et al. found a significant correlation between chief-cell percentage and adenoma weight but there was no correlation with preoperative calcium and PTH levels [3].
Serum PTH levels reflect adenoma and normal parathyroid gland PTH output. It is unclear how much PTH contributes the adenoma but there is evidence suggesting stronger correlation between PTH and adenoma weight in extreme levels of PTH. Therefore, several authors recommend that very high levels of PTH should alarm the surgeon of the possible existence of a too large adenoma [7]. When the PTH level is below 6 pmol/l (57 pg/ml) the adenoma is likely to weigh < 400 mg, whereas PTH levels > 170 pg/ml usually indicate an adenoma weight > 800 mg [2]. The relationship between adenoma weight and calcium is much more complex, as serum calcium values are the result of multiple endocrine calciferol-mediated mechanisms [8].
We found a negative moderate though statistically significant correlation between phosphate and adenoma weight. This is consistent with the studies of Bindlish et al. [5] and Mozes et al. [2] and is not surprising, since hypophosphatemia is observed in about 40% of HPT [2]. Vitamin D deficiency may be associated with heavier parathyroid adenomas, therefore in some departments it is common practice to correct vitamin D deficiency before establishing a HPT diagnosis [3].
According to a large cohort study with > 340 patients, adenoma weight correlates with the percentage decrease in calcium levels from before to after surgery but not with the PTH decrease [3]. Other authors report a significant correlation between adenoma weight and PTH-decrease at 10 min postexcision [12].
In our study, patients who required bilateral neck exploration, had also significantly (20–25%) lighter adenomas. For this reason, bilateral exploration should always be considered in cases of inadequate PTH decrease and adenoma size that can not explain high PTH values, even if imaging studies are strongly suggestive of single-gland disease. Interestingly, additional abnormal glands are present in some patients despite appropriate reduction in IOPTH. Unilateral exploration has been associated with recurrence/persistense rates of 5%, while recurrence rates after bilateral exploration recurrences are < 1% [13].
The controversial literature results can also be associated with other methodological divergences or limitations, except for the above mentioned approach of categorizing a continuous variable. Sadly, the majority of the studies do not report how correlation was tested. Many studies used the Pearson correlation without mentioning if its assumptions were met (normal distribution, homoscedasticity, linearity) [14]. We performed the more conservative Spearman correlation, as our data was not normally distributed [15]. Outliers can also bias the results in small samples, if they are not properly analyzed. Astonishingly, the removal of two heavy adenomas in a series of 44 patients, reduced the correlation coefficient from 0.83 (very strong correlation) to 0.28 (moderate to weak) [11].
Our study has limitations inherent to retrospective studies. Firstly, it is based on medical records, which are captured for clinical purposes. Moreover, the specimens were analyzed from different pathologists during the long study period, with the potential for interobserver bias. Last but not least, the weight of small adenomas may have been overestimated, as their size does not always allow precise removal of the surrounding fat.