Residual stone volume outperforms stone-free rate as risk marker

A secondary analysis of the ASPIRE randomized trial presented at the 41st Annual Congress of the European Association of Urology (EAU) in London, UK, found that residual stone volume (RSV) after endoscopic stone surgery is a significantly stronger predictor of downstream health care consumption events than conventional stone-free status classifications—suggesting that what surgeons leave behind matters more than whether any stone remains.¹

Among 101 evaluable patients followed for 2 years across 11 US centers, each 100-mm³ increase in RSV was associated with more than double the odds of a health care consumption event (adjusted OR 2.08; 95% CI 1.29–3.34), with a model discrimination area under the curve (AUC) of 0.776. By contrast, baseline stone volume, stone density, and categorical stone-free definitions were not statistically significant predictors.

In practical terms, 100 mm³ of residual stone corresponds roughly to a 6-mm spherical fragment—a size commonly encountered post-ureteroscopy. The AUC of 0.776, as lead investigator Brett Johnson, MD, explained, means that in a head-to-head comparison of 2 patients, the model correctly identifies the higher-RSV patient as the one more likely to experience a clinical event roughly 78% of the time.

Johnson, an associate professor of urology at UT Southwestern Medical Center, said the field's long reliance on binary stone-free status has obscured meaningful differences in surgical outcomes—noting, for example, that a patient with a single 3-mm fragment and a patient with 20 fragments are currently classified the same way under many definitions. He attributed the slow shift toward volumetric measurement largely to the limited clinical availability of automated CT-based segmentation tools, although he expects that to change as AI-driven software becomes more accessible and widely validated.

Please provide an overview of this study and its notable findings.

This is a secondary analysis of the ASPIRE 2-year outcome study—the same data set I mentioned previously—a randomized, very high-quality study where we very carefully watched these patients. What we noted was a significant difference with the SURE procedure vs traditional ureteroscopy. But this secondary analysis asked: Why is that the case? Because stone-free status wasn't that different between the 2 groups.

For context, historically, stone-free rate has been the standard for how we determine success of kidney stone surgeries, which makes sense, right? Are you stone free or are you not? There's a little bit of granularity there—there's grade A stone-free status, which is essentially nothing on CT, and then grade B, where you can have up to a 2-mm fragment. So, there's some granularity, but it's largely a binary thing. And although historically, that has been the primary outcome measure, it basically did not capture any meaningful differences in clinical outcome for this study. When we went back and looked at the data, residual stone volume was the independent predictor of downstream health care use. And it wasn't where you started—it wasn't baseline stone volume or density. It was how much stone did you have when the surgery was over. Not whether you had stone, but how much.

It makes sense logically. If you look at a person with 1 3-mm fragment in the lower pole and say, “You're not stone free,” and another patient who has 20 fragments ranging from 1 to 5 mm and say, “You guys are the same”—you wouldn't say that. That doesn't make sense. So, what this analysis showed was that residual stone volume was the strongest predictor of whether the patient would have one of these clinical encounters—an ER visit, surgery, that kind of thing.

The field has used stone-free status as its primary outcome measure for decades. Your data suggest RSV is a significantly better predictor of downstream health care consumption than categorical stone-free definitions. Why do you think the field has been slow to move toward volumetric measurement, and what has that inertia cost us in terms of what we do and don't understand about ureteroscopy outcomes?

You mentioned the word right in the question—inertia is a big thing. When we design clinical trials, we often model our outcomes from previously accepted clinical trials. And sometimes, it's hard to change that status quo. The people who review all these studies are used to seeing outcomes a certain way, and they say, “This study used this outcome, so why are you trying to change it?” So, in general, the scientific community does have some inertia in how we define outcomes, and it takes a long time for those to change.

But probably the biggest barrier has just been the clinical availability of volumetric measurement. That's getting better now that we have CT software packages that can do this automatically, and there are procedural or even AI-driven models that can do it for you, but they're not widely available. And some of them are not FDA approved or have some disclaimer that they shouldn't be used clinically. So, there's often a technical barrier to measuring stone volume. It's a lot easier when you're just looking at a scan to take the little measuring tool and drag it across the stone and say, “That's a 3-mm fragment,” than to segment the stone and then either do the calculation by hand—measuring all the dimensions—or register the stone on the scan and have a software package do it. But when you think about it, it's kind of silly to describe a 3-dimensional object in 1 linear dimension. If I said, “You drive a Ford Explorer, that's a big car—that car is 6 feet”—it doesn't make sense volumetrically. So, I do think it's changing. More studies are reporting volume, both preoperatively and postoperatively. And I think it will eventually become the standard of care once it's widely accessible and it just gets reported on the CT. I think we're going there, but it just takes time.

The adjusted odds ratio for RSV as a predictor of healthcare consumption events was 2.08 per 100 mm³, with an AUC of 0.776 after adjusting for age and prior urological surgery. Can you put that discrimination value in practical terms—what does an AUC of 0.776 actually mean for a urologist trying to identify which post-ureteroscopy patients are at meaningful risk for retreatment, ED visits, or hospitalization?

We did the receiver operating characteristic—the ROC curve—so basically we're quantifying how well the model will separate patients into, "Yes, you will have an event" or, “No, you will not." An AUC of 0.5 basically says we're not predicting anything—it's a coin flip. Our AUC is basically 78%, call it 80%. The idea is, if you had 1 patient with 10 mm³ of RSV vs 1 with 150 mm³, the model is implying that 80% of the time in a head-to-head comparison, the larger RSV will be the one with the clinical encounter. That’s pretty strong as far as ROC analysis goes. What we're saying is that RSV is very strongly predicting those who will have the event—we're classifying low-concern vs higher-concern patients. This basically indicates that residual stone volume is strongly predictive of that. And then the odds ratio—for every 100 mm³, we're doubling the risk of an event. For context, 100 mm³ of stone is basically a 6-mm residual stone if you look at a spherical shape. So, for every additional 6-mm stone, you double your risk of an event.

Baseline stone volume and density were not predictive of healthcare consumption events—only residual stone volume after the procedure mattered. What does it mean clinically that where you start doesn't predict downstream burden, but where you finish does?

I think historically a lot of weight was put on how bad your stone was initially—if you're kind of a "stone factory"—and saying, well, you treat a stone and 18 months later, it's probably just a new stone. But I think we're starting to understand now that when you have surgery for stones, residual stone is really the driving factor of events, because that stone is not only a nidus for stone growth, but a larger stone can obstruct in and of itself. And it's kind of almost a matter of time before that stone would be an issue. So, it kind of makes sense: If you have a significant stone burden but you treat all of it and you're clean as a whistle, you're probably better off than someone who has less stone to start with but has a bunch left over. That makes sense empirically. But I think historically, not enough emphasis has been made on residual stone volume—that even small particles and dust can serve as a nidus for more stone growth. We're seeing CT becoming a standard for following these patients, where historically we'd just get an ultrasound and maybe it would show something, maybe it wouldn't, and you wouldn't really worry about it because it should all pass. I think more emphasis is being put on trying to get out as much stone as possible, and less on where you started.

REFERENCE

1. Johnson B, Stern K, Chi T, et al. Residual stone volume, rather than stone-free status, predicts downstream healthcare utilization after ureteroscopy: Secondary analysis of the ASPIRE trial. Presented at: 41st Annual Congress of the European Association of Urology. London, UK. March 13-16, 2026. Abstract P0417