Assessment of the validity of in vivo methods of measuring human macular pigment optical density.

Recently, Hammond et al. published a manuscript in Optometry and Vision Science entitled “Assessment of the Validity of in Vivo Methods of Measuring Human Macular Pigment Optical Density.” The authors claim to assess the validity of various optical methods used for the in vivo measurement of human macular pigment (MP). Specifically, they attempt to compare the MP measuring method of heterochromatic flicker photometry (HFP)—a psychophysical method that is their specific field of expertise—with several other emerging objective methods that are based on reflection, fluorescence, and resonance Raman spectroscopy. We feel obliged to point out a string of shortcomings, misrepresentations, and flawed conclusions of this manuscript, especially because a large portion of the manuscript provides a duplicative critique of our resonance Raman method for measurement of macular pigment in vivo. The authors recapitulate the points that they have already published as lengthy Letters to the Editor of the journals Investigative Ophthalmology and Visual Science 2,3 and Carotenoid News. Disappointingly, however, the authors fail to include any of our rebuttals to their arguments, as provided in detail in earlier published letters and in Carotenoid News. We highlight a few of the more important points of contention subsequently. 1. HFP is a subjective method because a subject uses his or her own photoreceptors to measure MP. Like with any psychophysical method, extreme care must be taken to ensure that the subject understands the task and performs it properly and reproducibly. Obviously, HFP cannot be externally validated because it is impossible for excised tissue to participate in an HFP measurement. In the absence of this possibility, it may provide comfort to the authors that HPF can generate spectral stimulation spectra that to some extent can follow the shape of absorption spectra measured for excised tissue or a model system with another method, i.e., absorption spectroscopy. A calibration to the absorption strength of the excised tissue, however, is fundamentally impossible for HFP. A spectral match of HFP stimuli to absorption shapes of external tissue alone is not sufficient. A very important aspect, of course, is the correlation of HFP perceived MP levels with levels existing in the tissue. In particular, one wonders about the linearity in this regard. Certainly, at some MP concentration level, the HFP measurements must start to saturate. However, because a calibration of concentration levels is fundamentally impossible with this technique, saturation effects cannot be taken into account. By contrast, objective optical spectroscopic techniques require no subject cooperation beyond fixation on a target (which can be monitored by a video camera), and validation can be performed on animal or model eyes. For the Raman method, we validated the technique by measurements in living monkey eyes followed by high-performance liquid chromatography measurements of MP levels after enucleation. 2. In their discussion of the influence of ocular media on MP Raman measurements, the authors fail to cite recently published work by Savage et al. and Zagers et al. questioning the validity of previously published crystalline lens density functions at wavelengths longer than 450 nm. If these more recent results had been used for their calculations, their putative reduction factors would have been much lower in magnitude. 3. Our finding that MP levels remain low in elderly pseudophakic eyes is not explainable by scattering at the intraocular lens interface with the posterior capsule as they assert. At the University of Utah Moran Eye Center, only modern silicone and acrylic lens are implanted during cataract surgery. These lenses are designed for high optical clarity above 450 nm, and they are shaped for a tight fit with the posterior capsule to prevent lens epithelial migration that can cause light scattering. In situations when posterior capsule opacification does eventually occur, the ophthalmologist routinely performs a YAG laser capsulotomy to clear the visual axis. All subjects enrolled in our studies had a dilated eye examination by a board-certified ophthalmologist to confirm that no media opacities were present that could attenuate the MP measurements. 4. The authors have never used our Raman instrumentation as evidenced by their erroneous statement that no chin rest is used during measurements, yet they assert that elderly subjects would be unable to hold fixation steady for the time period of a measurement (0.25–0.5 second). Clinical practice clearly contradicts their assertion. Elderly patients with impaired visual acuity routinely hold steady fixation under bright light conditions for much longer periods during therapeutic and diagnostic procedures such as ocular photodynamic therapy and video angiography. 5. HFP measures the perceived MP only at the edge of the test stimulus. Although the authors recognize correctly that autofluorescence and Raman methods are based on a signal that is integrated or averaged over the entire stimulus area, they state incorrectly that this is only an assumption that has not been tested. The authors completely ignore our published correlation results between in vivo Raman MP levels and direct high-performance liquid chromatography measurement levels in the same primate eyes. Similar correlation results could never be obtained for a psychophysical test such as HFP. 6. The authors state incorrectly that the Raman method does not provide spectral profiles. In fact, the intensity of resonance enhancement closely follows the spectral absorption shape. At any resonant excitation wavelength, the molecules yield very sharp, highly distinctive vibrational peaks, a spectral “fingerprint” of the molecules of interest. Contrary to the authors’ assertions, these Raman signals can be readily quantified, and unlike HFP, they can be calibrated against the concentration of external standards. 7. The authors fail to cite our published correlation results for Ramanand HFPderived MP levels. Similar correlation results have also been reported recently by another group. The correlations are statistically significant and would likely improve if HFP 1040-5488/06/8304-0254/0 VOL. 83, NO. 4, PP. 254–259 OPTOMETRY AND VISION SCIENCE Copyright © 2006 American Academy of Optometry