Windows to the Soul…and the Vasculature

477 4
Continuing the theme of useful Nephrology bedside clinical signs, today we’ll focus on band keratopathy (BK). Often overlooked, this is a band of calcium deposition across the central cornea commonly seen in patients with advanced CKD. It reflects chronic positive calcium balance or hypercalcemia, and is seen in a wide variety of calcium overload states, including myeloma, sarcoid, hyperparathyroidism and renal tubular acidosis. Interestingly, it can regress with treatment of hypercalcemia, (occasionally dramatically, as in this patient) and will often regress following renal transplant too. Calcium is deposited on the corneal surface (directly under the epithelium) as a horizontal band that begins at the periphery and moves centrally as BK becomes more severe. The BK severity grading system reflects this process, with the mildest grades most peripheral (see figure). Deposits begin as a gray haze, progressing to dense white with a pebbly surface. They can lead to pain, a foreign body sensation and recurrent erosions, but are not a cause of visual loss.
The pathophysiology is not fully understood, but passive calcium precipitation is believed to be responsible. Tears and aqueous humor contain calcium and phosphate at concentrations approaching their solubility product. As tears evaporate from the intrapalpebral area, the concentration increases and precipitation occurs. The most severely affected area is the junction of the middle and lower thirds of the cornea, which is the area of maximum atmospheric exposure. Elevated serum calcium increases the likelihood of precipitation occurring. It is worth remembering that a passive precipitation model used to be proposed as the mechanism of vascular calcification in calcium overload states. This has since been debunked; vascular calcification is an active process involving a phenotypic switch in the vascular smooth muscle cell, which comes to resemble an osteoblast. We may yet learn that calcification occurring on the eye is also more complex than just passive deposition.
This last point is relevant as BK and vascular calcification are closely linked: the presence of BK is associated with a ten-fold increased prevalence of vascular calcification, and over half of dialysis patients with BK have radiologic evidence of VC. Furthermore, BK has recently been shown to independently predict 1-year mortality in dialysis patients, with each increment in the BK severity index being associated with a 25% increased mortality risk in multivariate analyses. It would be interesting to study whether BK also predicts calciphyllaxis. When I encounter BK clinically, I tend to treat it as I would extraskeletal calcification at any other site. As it implies chronic positive calcium balance, it prompts me to consider measures to reduce calcium balance in the patient, such as reducing the dialysate calcium bath, switching to a non-calcium containing phosphate binder and/or a non-calcemic vitamin D analog.


  1. You're right on both counts Marta! A lower pH is also thought to contribute to Ca precipitation on the cornea; it is the VSMC and not the endothelial cell that undergoes a phenotypic switch – I've updated the piece. Many thank

  2. This is very interesting… as for the pathophysiology, I remember hearing that part of it is that the pH is different (?lower) in that part of the eye… Did you come across anything like this?

    Minor point: vascular calcification I think is felt to involve changes in the vascular smooth muscle cell rather than the endothelial cell, though I would not be surprised if there is a (signaling) contribution by the endothelial cell as well to the smooth muscle cell. Marta

  3. Thanks for the feedback Gopi. I agree that the images look similar to pingueculae, however you can differentiate the 2 by a sharply demarcated peripheral edge in BK, which is not present in a pinguecula.
    These images illustrate the Porter and Crombie classification of BK, and are taken from here:

  4. The accompanying series of corneal/ conjuctival images do not show band keratopathy. The arrows seem to be pointing to pingeucula.

Leave a Reply