How bad is kidney stone analysis?
But does this not raise the obvious question: How good are stone analysis labs? At first one might say why ask? We use labs all the time and trust them. As things turn out, stone analysis varies more in quality than serum electrolytes, or blood hemoglobin. Moreover, some stone crystals pose greater problems than others.
There is a gold standard
X Ray Diffraction
Basiri et al recently reviewed all available papers concerning analysis of kidney stone crystals. Like prior investigators, some of whom they reference, X-ray diffraction does indeed reveal stone crystal structures and is a gold standard.
You need to sample several regions of a stone because stones can contain multiple crystals. If you do, you can rely on x ray diffraction as the ideal method on which to base research or against which to compare other methods. But commercial kidney stone analysis cannot rely on x ray diffraction – too time consuming and expensive.
High Resolution CT Scanning of Stones
My own co-workers regularly use high resolution CT scanning of individual stones. Although unlike x ray diffraction, high resolution CT scanning of a stone does not reflect crystal structure, it does offer highly calibrated density measurements. These measurements identify stone crystals with considerable accuracy. Moreover, the CT technique shows not only some selected part of a stone but can show the whole stone and reveal its separate crystal components.
Even so, commercial kidney stone analysis cannot use high resolution CT – too time consuming and expensive, like x ray diffraction.
How bad is kidney stone analysis?
Struvite is a problem
Krambeck et al collected stone fragments whose composition they determined with considerable precision and accuracy. They then sent multiple samples of the fragments to 5 commercial laboratories.
The laboratories failed to detect struvite about half of the time. The featured pictures note this as a negative sign (-). On the other hand, the laboratories reported struvite in stone fragments that had none (+ signs).
Being a result of infection with bacteria that possess urease, struvite stones pose complex surgical and medical treatment problems. The bacteria can produce large stones that grow rapidly and damage kidney tissue. They may super-infect more common calcium stones more easily identified by stone analysis laboratories. Thinking only of those common stones clinicians may miss opportunities to prevent enlargement of struvite fragments not passed or removed at surgery.
Atazanavir, a common antiviral medication was frequently missed
Because of their chemical structures the antiviral drugs often possess spare solubility in urine. This agent is insoluble enough to produce drug stones. All five laboratories failed to detect this drug crystal.
Hydroxyapatite (HA) was missed too frequently
The HA content of stones gives clinicians important clues to renal papillary histopathology, clinical course of stone disease, and even urine chemistries. So the fraction of stones HA comprises matters.
For example, in patients whose stones were >50% calcium oxalate on average, stones were found growing outside the kidney on the papillary surface over deposits of interstitial apatite plaque. Among patients whose stones are predominantly calcium phosphate crystals terminal collecting ducts are plugged with hydroxyapatite deposits. Their phosphate stones are rarely found on plaque but rather seem either to form in free solution or as overgrowths on the open surface of ductal plugs – the end exposed to the urine.
Most remarkably, even the distinction between stone formers whose stones did or did not contain brushite seemed to matter. Those whose stones were entirely hydroxyapatite in their phosphate component formed numerous small tubule plugs whereas those whose stones contained brushite formed few but very large ductal plugs.
Is there a problem?
Struvite and HA failures suggest one exists
I want the highest accuracy in stone analysis, but realize $25 or so per test buys only so much. This study makes one concerned that laboratories frequently miss struvite. Likewise, but less often, they under report calcium phosphate crystals. Even worse, they report struvite when not present, needlessly raising the specter of infection.
Our own research suggests circumscribed failure
On the other hand, commercial stone analyses must possess considerable accuracy if one leaves struvite to one side. Our own research group has pioneered in tissue biopsy of renal papillae of stone forming patients during the course of surgical stone removal. In the course of the work we found a remarkably stable relationship between tissue changes and average stone calcium phosphate vs. calcium oxalate composition.
The analyses for these studies did sometimes arise from the sophisticated methods I have already mentioned. But often, perhaps even usually, they arose from a variety of commercial sources over a period of many years. If random commercial laboratory reports were adequate to support distinctions of such subtlety, how bad can they be?
Likewise, we have published that the percent of phosphate in kidney stones correlates very well with the urine calcium phosphate supersaturation of the patients who produced these stones.
However, these observations concern mainly the calcium oxalate and calcium phosphate crystals, leaving aside the problem with struvite identification.
Do we need more research on this issue?
I would think perhaps we might.
My questions concern the clinical consequences of errors that laboratories make. For example, missing minor apatite components in a stone would make no difference to me. On the other hand, missing apatite when abundant could change diagnosis, prognosis, and even treatment. Likewise, failure to report struvite might delay treatment of infected stones. Failure to report uric acid would obviate proper treatment with potassium alkali.
If I were doing research in this area I might seek designs that capture these kinds of clinical consequences from errors.
Judging from the several references I have shown, we do not yet know the frequency of such consequential errors.