M RUOSS¹, C SEGER², C TIMM², M RISCH<sup>2

1</sup>Kantonspital Glarus AG, Glarus, Switzerland, ²Risch Laboratory Group, Schaan, Principality of Liechtenstein

The formation of solid material of organic or inorganic origin in the urinary system (urolithiasis) is a rather common disease caused by various environmental or genetic factors (1). Once such crystal aggregates are isolated from patients detailed analysis of their composition is, beside conventional urinary analysis including microscopy and sediment characterization, the most important step towards differential diagnosis of the principle of causality.

State of the art technologies for “stone analysis” (2) are infrared spectroscopy (IRS) and powder x-ray diffraction (XRD). If using IRS the attenuated total reflectance (ATR) technology allows an accelerated workflow with less material needed since embedment of analyte material into a KBr matrix is not necessary. Both in conventional IRS and ATR-IRS spectral readouts are analyzed either manually or automatically by comparison with reference spectra since characteristic absorption line groups allow distinguishing the different crystal types. If using automated analysis systems the user relies on both the quality of the comparison algorithm and the reference collection – both provided by the manufacturer of both instrument and software.

In the course of a platform design and evaluation study we tested two ATR-IRS systems from independent manufacturers by measuring diagnosed patient samples (n = 51), proficiency testing samples (PTS, n = 18), and mixtures of pure crystalline materials (n = 64) including interference testing with water, urine and whole blood (n = 30). Whereas most of the PTS materials were identified at least partially by both platforms, patient sample analysis revealed the data base dependency of the analysis setup – both between the manufacturers platforms and within a vendor platform since the primary diagnosis of these materials was done with an instrument of one of the vendors. The analysis of mixtures showed the limitations of semi-quantitative (percentage) stone analysis with ATR-IRS – a pure qualitative composition description with the information which component is predominant seems to be more feasible.

Taken together it can be concluded, that software and database evaluation is a critical step in introducing IRS based stone analysis in a routine laboratory setting. In critical cases (e.g. inconsistent “hit list” entries) a second opinion – preferably with a different method (e.g. XDR) seems necessary to minimize diagnostic errors.

(1) Romero V, Akpinar H, Assimos DG. Rev Urol. 2010;12:e86-e96.
(2) Schubert G. Urol Res 2006;34:146-150.