
80 Mandel and Mandel
two or more components. The admixture of calcium oxalate or apatite accounts for 45%
(15) to 9.2% (5,17) and the incidence of pure whewellite stones varies from 24.9% (11)
to 4% (25). Although the actual percentages vary among the different studies, the most
frequent stone compositions are whewellite/weddellite, pure whewellite, apatite/
whewellite/weddellite, and apatite/struvite. Calcium oxalate was present in 73 ± 7% of
all stones analyzed in these studies.
Because most stones are multicomponent, the method employed in the analysis of
stone material should be capable of resolving all components of the stone, especially all
the crystalline components. The literature on stone analysis methods clearly supports the
use of XRD or FTIR as the prime choices. The use of both methods to obtain supportive
and the most accurate compositional data is ideal. For XRD and FTIR, the accuracy of
the analysis is very strongly dependent on the quality of standard spectra. Most labora-
tories conducting stone analyses prepare their own standards libraries. Unfortunately,
many analysis laboratories use patient stone material to create their standard spectra. As
their stones are analyzed by the same method as they are using to analyze other stone
samples, their unknowns become their standard. As virtually no stone is composed of
only one pure crystalline component, such spectral libraries are very inaccurate and the
potential for skewed and inaccurate stone analysis is highly probable. Preparation of
synthetic stone components for the generation of standards and verification of compo-
sition by alternative methods is the only correct way to prepare a standards library for
either XRD or FTIR, especially for FTIR. Commercial libraries should only be used for
supplemental data in those rare instances when experimental data cannot be correlated
with defined stone component standards, especially for identification of nonbiologic or
false stones.
One issue not yet resolved in the literature is the level of accuracy one should accept
in analysis reports. Specifically, rank order of compositional analysis in multicompo-
nent stones analyzed by XRD is probably the best one can expect. In FTIR analyses,
although it is possible to generate computer aided admixed spectral standards at any
percent admixture, many years of experience and cross-correlating analysis results with
accurately weighed admixtures of pure standards suggests that percent compositional
analysis should be quoted at 10% or more for accuracy.
The issue of stone composition analysis accuracy is certainly clouded by the frag-
mented samples now seen by most analysis labs with the common utilization of litho-
tripsy. Virtually no laboratories can be assured that they have received the entire voided
sample or if the sample they have received for analysis accurately represents overall
stone composition. Accuracy of analysis by other methods such as microscopic evalu-
ation that is quoted as approaching or exceeding the XRD or FTIR levels should be
questioned for scientific basis.
In conclusion, high resolution XRD and FTIR spectroscopic methods of analysis are
currently the only two methods that should be considered for the accurate and complete
characterization of crystals in urine and for the compositional analysis of kidney stones.
Frequently, the use of both XRD and FTIR is beneficial and occasionally necessary for
an accurate analysis of multicomponent stones. The accuracy of these methods of analy-
sis is very much dependent on the quality and completeness of a standards library as well
as on the experience of the instrument operator, especially in the use of computer-based
match routines supporting both XRD and FTIR instruments.