Summary: In this comparison of patients with FGF-23 associated hypophosphatemia and control subjects, PTH concentration was higher in patients, 25% of whom had PTH above the upper limit of normal. PTH was positively associated with serum calcium in patients but not in controls, and was negatively associated with serum phosphate. 7 subjects had evidence of parathyroid autonomy, which was effectively treated surgically in all cases.

This study in context: Hyperparathyroidism is a significant and clinically challenging complication of XLH and other forms of FGF23 mediated disease. This cohort with systematic data collection is a useful resource to study the prevalence and natural history of this complication, particularly in light of novel therapies recently becoming available.

Major comments:
o I am unclear of the value of doing a PCA analysis in this dataset – what is it adding? It is unclear how this analysis adds to our understanding of hyperparathyroidism in FGF23-mediated hypophosphatemia.
o It would be interesting to look at the relationship between PTH and other variables including calcium and phosphate after excluding the patients with parathyroid autonomy – I wonder if the same relationships would be observed (possibly suggestive of evolving autonomy) or whether the patients without clinical parathyroid autonomy would look like the controls. As it stands, I worry that the relationships described in the patient population as a whole are being driven by those with tertiary hyperparathyroidism.
o The authors propose in the discussion that impaired 1,25(OH)2D may be involved in the pathogenesis of secondary hyperparathyroidism in this patient population – would it be possible to test this hypothesis by looking at this relationship or looking at the history of therapeutic calcitriol/active vitamin D treatment with the development of hyperparathyroidism?

Minor comments:
o I find “genetic hypophosphatemia” to be somewhat confusing and vague as it could include other entities such as HHRH – it may be more clear if it is specified that this is FGF23-mediated hypophosphatemia specifically.
o Intro p4 line 47: this is called a “prospective” study in the introduction, but it is not really prospective in that greater than half of the patients with HHPT were diagnosed prior to data collection. Most data was collected systematically, which might be a more appropriate term.
o Methods p5 line 11: similarly, this is called a “case-control” study, but that term has a specific epidemiologic meaning in which exposures are retrospectively assessed in order to determine whether an exposure increases risk of an outcome – this is not that kind of study but rather a cross-sectional comparison of a disease and control cohort.
o Methods p 7 line 12 – was there any cross-calibration of the 2 different 1,25(OH)2D assays?
o Results p 8 line 59 – do the authors have any speculation as to the genetic disorder in the patients without evidence of PHEX or FGF23 mutations?
o Fig 1 – is redundant with table 1; could be eliminated if space is a concern.
o Fig 3 – it looks like x-axis is cut-off in fig 3B – where are most of the controls? And consider using open and closed circles rather than black and blue as these are hard to distinguish in a black and white copy.
o Table 2, fig 4, and p 11 line 8, : what is meant by “maximal median at time of diagnosis”.
o Table 2 – it is notable that a substantial number of patients without HHPT have evidence of renal calcification – what do the authors think about this group?
o Figure 4 – consider making boxes black/white/gray for ease of reading in black and white copies.
o P. 11 line 15 – the statement that 25OHD was >30ng/mL in all patients (presumably at time of hyperparathyroidism as this is what this paragraph is discussing) is discordant with Fig 4 in which prior to surgery the median 25OHD appears to be about 25.
o P. 13 line 24 – the authors appear to state quite definitively that using CKD-EPI is inappropriate in this patient population – so why present it in the results?
o P15 line 3 – the methods state that serum and plasma were frozen for FGF23 measurement, but the discussion alludes to the fact that the serum was used rather than the plasma. Please clarify in methods.
o P. 16 lines 28-33 – the statement “careful management of standard therapy with oral phosphate salts and active vitamin D metabolites remains essential to prevent hyperparathyroidism” does not seem entirely consistent with the findings of this paper which do not show systematic management differences between those with and without hyperparathyroidism, and in fact presents a patient who developed hyperparathyroidism absent treatment. Is there information available about the dosing of medications which might help clarify this?

Thank you for addressing all the points in the first review.

I continue to disagree with the characterization of this study as “prospective”. “Prospective” has a specific meaning in which subjects are followed longitudinally. If I am reading correctly, neither XLH patients nor the healthy controls are followed longitudinally. Rather, each subject is assessed at a single time point. This is however a semantic point, so I leave it to the Editor to decide whether or not this is acceptable.

P14 L47:
The finding, currently mentioned in the discussion, that there was no significant relationship between PTH and calcium after excluding the patients with hypercalcemic hyperparathyroidism should be moved to the Results section. In addition, details as to the relationship should be presented: the beta estimate, the R (or R2) and the p value.

P14 L52:
The phrase “…but the non-autonomous patients still remained above the regression slope of the controls” is vague. Looking at the scatter plots in Figure 2, several of the individual patients have PTH values below the control regression slope, so this doesn’t seem correct. Would it be correct to say that, after excluding the patients with HHPT, mean (or median) PTH is significantly higher than the controls?

Figure 2:
It is unusual and thus confusing to have 2 regression lines but individual points for only one group. I would prefer to see this as overlaid scatter plots or 2 separate panels. In panel B, it is not appropriate to cut off the regression line for the controls – please extend the X-axis so that the entire regression line can be shown.

Minor comment:
I apologize for not noticing this on the first review, but in Table 2, in the “ongoing treatment” section, is it possible that the authors mean “vitamin D” and not 25OHD?