Characterization of Stone Events in Patients with Type 3 Primary Hyperoxaluria

Abstract

Purpose:

Hallmarks of primary hyperoxaluria type 3 (PH3) are nephrolithiasis and hyperoxaluria. However, little is known about factors influencing stone formation in this disease. We characterized stone events and examined associations with urine parameters and kidney function in a PH3 population.

Materials and Methods:

We retrospectively analyzed clinical, and laboratory data of 70 PH3 patients enrolled in the Rare Kidney Stone Consortium PH Registry.

Results:

Kidney stones occurred in 65/70 PH3 patients (93%). Among the 49 patients with imaging available, the median [IQR] number of stones was 4 [2, 5], with largest stone 7 mm [4, 10] at 1^st imaging. Clinical stone events occurred in 62/70 (89%),) with median number of events per patient 3 [2,6], range 1-49. Age at 1^st stone event was 3 years [0.99, 8.7]. Lifetime stone event rate was 0.19 events/year [0.12,0.38] during follow-up of 10.7 [4.2, 26.3] years. Among 326 total clinical stone events, 139 (42.6%) required surgical intervention. High stone event rates persisted for most patients through the 6^th decade of life. Analysis was available for 55 stones: Pure calcium oxalate accounted for 69%, with mixed calcium oxalate and phosphate in 22%. Higher calcium oxalate supersaturation was associated with increased lifetime stone event rate after adjusting for age at 1^st event [IRR, 95%CI] [1.23 (1.16, 1.32), p<0.001]. By the 4^th decade, eGFR was lower in PH3 patients than the general population.

Conclusions:

Stones impose a lifelong burden on PH3 patients. Reducing urinary calcium oxalate supersaturation may reduce event frequency and surgical intervention.

Introduction:

The primary hyperoxalurias (PHs) are autosomal recessive disorders of glyoxylate metabolism characterized by endogenous hepatic overproduction of oxalate that results in elevated urinary oxalate excretion (>0.7 mmol/1.73 m²/day; normal < 0.46 mmol/1.73 m²/day) [1-3]. PH3 is the most recently defined subtype [4] (OMIM #613616), accounting for approximately 10% of currently diagnosed PH cases. Patients with PH3 typically present with recurrent nephrolithiasis, often in the first decade of life. However, kidney failure in PH3 patients appears rare, especially compared to PH1 and PH2 [4-8]. PH3 is likely underdiagnosed due to the relatively less severe clinical symptoms in comparison with the other types of PH, the recent identification of the responsible gene, and the lack of readily available clinical genetic testing until quite recently [5, 9].

Although childhood onset of kidney stones is often observed in PH3 patients, it has been proposed that stone formation decreases later in life [1, 3, 8, 10]. Since little is known about factors influencing stone formation in this disease, in the current study we characterized stone events and examined associations with urine parameters and kidney function in a PH3 population.

Patients and Methods:

A retrospective analysis was performed on patients with genetically confirmed PH3 who were enrolled in the Rare Kidney Stone Consortium (RKSC) PH Registry as of January 1, 2021. Informed consent for participation was obtained from each subject after Mayo Clinic Institutional Review Board approval. The registry was queried to obtain clinical and laboratory data throughout the patient’s lifetime before kidney failure [11]. The estimated glomerular filtration rate (eGFR) was calculated using Pottel’s Full Age Spectrum equation for serum creatinine values [12]. Twenty-four-hour urinary excretion rates of oxalate, calcium, and other determinants of supersaturation and all stone analyses were performed in clinical testing laboratories and extracted from medical records. Since the cohort included children and adults, excretion rates were corrected to 1.73 m² for analyses. Supersaturation was calculated using the EQUIL2 program [13]. All clinical stone events in a subject’s lifetime were ascertained from clinician medical record review and were classified as: symptomatic only (pain, hematuria, or other stone-related symptoms), spontaneous stone passage, urologic procedure for stone management, or unknown (symptoms noted but not specifically described). Urologic procedures included lithotripsy, percutaneous nephrolithotomy, ureteroscopy, ureteroscopic stone removal, or cystoscopy.

Statistical analysis:

Results were expressed as median and interquartile range [IQR] for continuous variables due to the skewed distributions of the laboratory and stone event measures and as n (%) for categorical variables. Laboratory measures were reported as the median value across the patient’s lifetime unless otherwise specified. eGFR at last follow-up was defined as the last eGFR measurement occurring ≥15 years of age. Lifetime stone event rates were calculated by dividing the total number of stone events by age at the patient’s last follow-up. Spearman rank correlations were used to analyze associations between clinical features, imaging measures, and laboratory values. Multivariable Poisson regression models were fit to evaluate the association between number of lifetime stone events and laboratory measures after adjustment for age at first stone event, using the logarithm of age at last follow-up as an offset term in the models. Incidence rate ratios (IRR) and corresponding 95% confidence intervals (CI) are reported for these models. All analyses were performed using SAS version 9.4 (SAS Institute Inc., Cary, NC) and R version 4.0.3 (R Foundation for Statistical Computing, Vienna, Austria). All p-values were two-tailed and were considered statistically significant at the 0.05 alpha level.

Results:

As of January 1^st, 2021, 70 PH3 patients were enrolled in the RKSC registry (46 males and 24 females). Sixty-five of them (93%) were known to have stones, either seen on imaging or experiencing clinical events. Among patients with imaging available (70%, n=49), median [IQR] number of stones seen on imaging was 4 [2, 5] and maximum stone diameter was 7 mm [4, 10] at 1^st imaging. Most patients (n=62, 89%) experienced at least one clinical stone event in their lifetime. Individual patient trajectories are depicted in Figure 1 which includes age of diagnosis and individual stone events over time.

Figure 1:

Patient trajectories by age at each stone event, stratified by age at the 1^st stone event quartile, among patients with at least one stone event (N=62). Patients are ordered by age at their 1^st stone event. Stone event types are distinguished by shape and color, as well as age at PH diagnosis and age at the last follow-up. Median (IQR) ages at last followup, stratified by age at 1^st event, are shown in supplemental table 1.

Among patients with stone events, the median [IQR] number of events per patient was 3 [2, 6], with a range of 1-49 events (Supplementary Figure 1). All 8 patients (6 males and 2 females) who had not experienced any clinical stone events during follow-up had at least one kidney imaging report available, with 3/8 having at least 1 non-symptomatic stone first seen on imaging at ages 42 days, 3.7 years, and 7.9 years, respectively. Among these patients without stone symptoms, 7 were diagnosed via family screening and one was prenatally diagnosed; mean age at last follow-up was 9.4 years and ranged from 1 year to 29 years.

Among the 62 patients with clinical stone events, the first clinical stone event occurred at a median [IQR] age of 3 years [0.99, 8.7]. The total number of stone events for all patients was 326, among which 154 (47.2%) were spontaneous passage, 139 (42.6%) were surgical interventions, and 25 (7.7%) were characterized by symptoms only (e.g., pain, hematuria) without stone passage. Median [IQR] lifetime stone event rate for this cohort was 0.19 events per year [0.12, 0.38] with a median [IQR] age at follow-up of 12.9 years [6.3, 35.4] (Table 1).

Table 1.

Demographics, laboratory, and imaging characteristics of all PH3 patients (N=70)

Demographics and Stone Events	Median [Q1, Q3]	Number of patients (N=70)
Age at 1st stone event (years)	3.0 [0.99, 8.7]	62
Age at last stone event (years)	8.0 [2.2, 34.5]	62
Age at LFU (years)	12.9 [6.3, 35.4]	70
Time from 1st stone event to LFU (years)	10.7 [4.2, 26.3]	62
Lifetime Stone event rate (no. stones/age at LFU)	0.19 [0.12, 0.38]	62
Male	-	46
Laboratory Characteristics*
Urine Oxalate concentration (mmol/24hr)	0.82 [0.42, 1.07]	56
Urine Calcium concentration(mg/24hr)	83 [44, 146]	52
Urine Calcium*Urine oxalate concentration	67 [19, 152]	51
Urine Oxalate (mmol/1.73m2/24hr)	1.09 [0.87, 1.29]	58
Urine Citrate (mg/1.73m2/24hr)	672 [455, 944]	51
Urine Calcium (mg/1.73m2/24hr)	124 [80.8, 166]	50
Ca Oxalate Supersaturation (RSS)	2.12 [1.72, 2.36]	21
Urine Calcium*Urine Oxalate	119 [67.5, 178]	50
Urine Oxalate/Urine Citrate	0.0017 [0.0012, 0.0022]	53
Urine Calcium/ Urine Citrate	0.16 [0.11, 0.28]	51
24hr Urine Volume (ml/min/1.73m2/24hr <15yrs)	1744 [1331, 2735]	52
eGFR at last follow-up† (mL/min/1.73m2)	74.5 [60,0, 101.5]	28
Imaging Characteristics
First Image	-	-
No. stones	4 [2, 5]	48
Max stone diameter	7 [4, 10]	49
Last Image	-	-
No. stones	4 [3, 8]	49
Max stone diameter	5 [3, 9]	41

LFU=Last Follow-Up.

^*Laboratory measure is median value across the patient's lifetime unless otherwise specified.

^†eGFR labs were obtained when patients were 15 years of age or older.

Among the urologic procedures were 30 ESWL, 41 percutaneous nephrolithotomies, 45 ureteroscopic stone removals, 5 cystoscopic stone removals, 5 open lithotomies, and 13 not specified. Stone event rates in our cohort remained high, varying between 1.6 to 6.0 events per decade from birth to age 70 without any clear trend by age. Age at first stone event demonstrated an inverse association with lifetime stone event rate (r=−0.48, p<0.001) (Figure 2). Among patients with stone events, median age at last follow-up was 14.3 (7.0, 36.7) years, as shown in Supplemental Table 1. Age at first stone event, number of stones on first image, and lifetime stone event rates were not significantly associated with BSA corrected 24-hour urine excretion of oxalate (Uox), calcium (Table 2), or citrate; however, 24hr urine volume (1.73m2/24hr <15yrs) was positively correlated with age at first stone event (r=0.38, p=0.006) and negatively correlated with the number of stones in the first and last image (r= −0.48, p=0.002 and r= −0.61, p=0.001, respectively) (Supplementary Table 2).

Figure 2.

Scatterplots overlaid with boxplots of lifetime stone event rates, stratified by age at 1^st stone event quartile, among patients with at least 1 stone event (N=62). Different colors denote each quartile.

Table 2.

Spearman correlations between stone event measures and laboratory measures, among patients with at least 1 stone event in their lifetimes (N=62)

Laboratory Characteristic*		Lifetime stone event rate	Age at 1st stone event	No. stones on 1st image
Urine Oxalate Concentration (mmol/24hr)	N	52	52	40
	r	−0.44	0.53	−0.35
	P	<0.001	<0.001	0.025
Urine Calcium Concentration (mg/24hr)	N	49	49	39
	r	−0.12	0.31	−0.22
	P	0.43	0.028	0.18
Urine Calcium*Urine Oxalate Concentration	N	48	48	38
	r	−0.3	0.43	−0.28
	P	0.039	0.002	0.09
Urine Oxalate (mmol/1.73m2/24hr)	N	53	53	42
	r	−0.18	−0.11	−0.09
	P	0.21	0.44	0.57
Urine Citrate (mg/1.73m2/24hr)
Urine Calcium (mg/1.73m2/24hr)	N	47	47	38
	r	0.11	−0.05	−0.10
	P	0.45	0.75	0.56
Ca Oxalate Supersaturation (RSS)	N	20	20	15
	r	0.26	−0.19	0.29
	P	0.26	0.42	0.29
Urine Calcium*UOX concentration	N	47	47	38
	r	0.05	−0.02	−0.26
	P	0.72	0.87	0.12
eGFR at last follow-up †	N	28	28	19
	r	0.17	−0.49	−0.009
	P	0.39	0.008	0.97

The r statistics denotes spearman rank coefficients. P-values in bold indicate statistical significance at the 0.05 alpha level.

^*Laboratory measure is median value across the patient’s lifetime unless otherwise specified.

^†eGFR labs were obtained when patients were 15 years of age or older.

After adjustment for age at 1^st stone event, increased calcium oxalate supersaturation was positively associated with lifetime stone event rates (IRR (95% CI): 1.23 (1.16 to 1.32), p<0.001), while higher levels of Uox concentration, Uox excretion, and Ucalcium*Uox concentration were negatively associated with lifetime stone event rates (Table 3).

Table 3.

Associations between lifetime stone event rate and laboratory measurements after adjusting for age at 1^st stone event, among patients with at least 1 stone event.

Laboratory Characteristic*	N	Lifetime Stone Event rate
		IRR (95% CI)	P-value
Urine Oxalate concentration (per 1 unit)	52	0.38 (0.25 to 0.57)	<0.001
Urine Calcium concentration (per 100 units)	49	1.02 (0.86 to 1.20)	0.84
Urine Calcium* Urine Oxalate concentration (per 100 units)	48	0.87 (0.76 to 0.99)	0.041
Urine Oxalate (per 1.73m2/24hr)	53	0.53 (0.34 to 0.82)	0.005
Urine Citrate (per 10,000 1.73m2/24hr)	48	1.02 (0.86 to 1.20)	0.84
Urine Calcium (per 100 1.73m2/24hr)	47	1.13 (0.94 to 1.36)	0.18
Ca OX Supersaturation (per 1 unit)	20	1.23 (1.16 to 1.32)	<0.001
Urine Calcium* Urine Oxalate (per 100 units)	47	0.96 (0.81 to 1.14)	0.64
Urine Oxalate/ Urine Citrate (per 0.001 units)	50	0.95 (0.90 to 1.002)	0.06
Urine Calcium/ Urine Citrate (per 1 unit)	48	0.81 (0.33 to 2.01)	0.65
24hr Urine Volume (per 1,000 1.73m2/24hr <15yrs)	49	1.13 (0.98 to 1.30)	0.09
eGFR at last follow-up (per 10 mL/min/1/.73m2) †	28	1.02 (0.97 to 1.08)	0.45

IRR=Incidence rate ratio.

Incidence rate ratios and P-values were derived using Poisson regression models modeling the number of lifetime stones with log (age at last follow-up) considered an offset term.

^*Laboratory measure is median value across the patient's lifetime unless otherwise specified.

^†eGFR labs were obtained when patients were 15 years of age or older.

Stone analysis was available for 55 stones in 39 patients. Stones were comprised entirely of calcium oxalate in 38/55 (69%), with calcium oxalate monohydrate (COM) alone in 12 stones, calcium oxalate dihydrate (COD) alone in 2 stones, and both COM/COD in 24 stones. Mixed composition was found in 17 stones, 12 (22%) of which were mixed calcium oxalate/calcium phosphate (CaOx/CaP), and 5 (9%) of which were of other composition, including 2 stones with struvite. Stone composition did not differ for stones analyzed before versus after age 18 (Supplementary Table 5). For comparison, we also examined the stone composition that was available for 25 stones from 19 PH2 patients enrolled in the RKSC PH Registry. Stones were comprised solely of calcium oxalate in 16 (64%) of PH2 patients, with 12/16 (75%) having COM alone, none with COD alone, and 4/16 (25%) having both COM/COD. Among the remaining 9 PH2 stones, 7 (28%) were composed of CaOx/CaP, and 2 (8%) were of other compositions (Table 4).

Table 4.

Stone composition analysis

Stone Type	Arnous	Higueras 2021 [21]	Arnous	Jacob 2013 [26]	Daudon 1998 [29]
	PH3 n=55	PH3 n=30	PH2 n=25	PH1 n=12	PH1 n=87
CaOx	38 (69%)	24 (80%)	16 (64%)	10 (83%)	84 (97%)
COM	12/38	7/24	12	9	81
COD	2/38	4/24	0	0	0
COM/COD	24/38	13/24	4	1	3
CaOx/CaP	12 (22%)	6 (20%)	7 (28%)	2 (17%)	0
Other	5 (9%)	0	2 (8%)	0	3 (3.4%)

CaOx=Calcium Oxalate; COM=calcium oxalate monohydrate; COD=calcium oxalate dihydrate; CaP=calcium phosphate; PH=primary hyperoxaluria.

Among PH3 patients with eGFR measures taken after age 40 (n=15) eGFR levels were well below that of the general population, both in men and women [14] (Supplementary Figure 2). However, eGFR at last follow-up was not associated with lifetime stone events after adjusting for age at the 1^st stone event (p=0.45). There were no significant correlations between Uox, Uox concentration and eGFR at last follow up among patients with at least one stone event (Table 5). Two patients experienced kidney failure in this cohort, one at 8 years of age [5] and the other at 33 years [9]. There were no deaths.

Table 5.

Spearman correlations between Urine oxalate measures and eGFR at last follow-up, among patients with at least 1 stone event in their lifetimes (N=62)

Laboratory Characteristic*		eGFR at last follow-up†
Urine Oxalate (1.73m2/24hr)	N	27
	r	0.02
	P	0.91
Urine Oxalate concentration	N	27
	r	−0.22
	P	0.26

The r statistics denotes spearman rank coefficients. P-values in bold indicate statistical significance at the 0.05 alpha level.

^*Laboratory measure is median value across the patient's lifetime unless otherwise specified.

^†eGFR labs were obtained when patients were 15 years of age or older.

Discussion

Marked overproduction of oxalate in all types of PH results in stone formation and impairment of kidney function. Increased plasma oxalate levels and consequent Uox excretion have been associated with kidney function deterioration in PH patients with eGFR > 40 mL/min/1.73 m²[15]. In a large cohort of PH patients with types 1, 2, and 3 who had preserved kidney function at diagnosis, Uox excretion at diagnosis stratified by quartile was strongly associated with incident kidney failure[16] Moreover, increased 24h Uox excretion was an independent risk factor that has been associated with CKD progression in a large non-PH cohort [17]

Kidney stones in infancy and early childhood are a predominant feature of PH3 [18], even though Uox excretion rates tend to be lower than in PH1 and PH2 [5, 6, 8]. In this study, we evaluated associations between clinical stone events, urine biochemical parameters, and kidney function in patients with PH3. Among our cohort of 70 PH3 patients, 62 (89%) experienced at least one clinical stone event in their lifetimes, with the first event often occurring in early childhood. In addition, patients with their first stone event at younger ages were found to have higher lifetime stone event rates. Moreover, PH3 patients experienced high stone rates recurring throughout the 6^th decade, with more than 40% of stone events requiring surgical intervention. Overall stone event rates varied from 1.6 in the first decade to 6.0 in the 6ht decade of life. Although we cannot be definitive about comparisons between decades since we do not have lifetime data for all patients form birth, the data do suggest that PH3 patients experience frequent stones through their lifetime. Taken together, our results suggest that stone events impose a substantial clinical burden on PH3 patients, adding to our knowledge regarding the lifetime burden of this disease. Patients who experienced a first stone event at a younger age tended to have higher lifetime stone event rates. However, it is important to note that since these patients were also younger at last follow-up, accrual of additional follow-up data for the youngest patients will be important to confirm that this trend persists out to older ages. PH3 patients also exhibited lower eGFR levels than the general population after the 4^th decade of life [14], both in men and women. Yet, no significant association was found between eGFR at last follow up and stone event rates in PH3 patients.

Increased Uox excretion (>62 mg/24 hrs) is pathognomonic for PH, and supersaturation of the urine with calcium oxalate is widely presumed the cause of stone formation [19, 20]. Indeed, in our cohort calculated urinary calcium oxalate supersaturation was significantly associated with lifetime stone event rate. As a group, all PH3 patients have persistently high urinary oxalate excretion and calcium oxalate supersaturation, with the mean for this group (2.12 DG) well above the reference population mean (1.79 DG). However, somewhat counterintuitively, Uox and UCalcium*Uox were negatively associated with lifetime stone rate. The unexpected lack of an association between Uox excretion and stone events was recently noted in another PH3 cohort [21]. This observation could partly be due to the narrow range of Uox in PH3 patients which is moderately high in all and differing from other diseases like enteric hyperoxaluria where a broader range of Uox occurs and correlations of Uox and stone events were observed [22]. Thus, our data do not imply that strategies to reduce Uox towards or into the normal range might not reduce stone event rates in PH3 patients.

Multiple prior reports suggested an increased urinary calcium excretion in PH3 patients [5, 8, 18]. However, no significant association was found between urinary calcium and stone event rates, either before or after adjusting for age at first stone. Citrate is a natural urinary inhibitor of calcium oxalate stone formation [23]. In studies of non-PH patients, increased urinary citrate excretion was associated with reduced calcium stone recurrence [24, 25]. However, the current study did not demonstrate an association between urinary citrate excretion and lifetime stone event rates in is PH3. Ultimately, we can only speculate that PH3 patients with relatively greater stone burden have superimposed risk factors in addition to hyperoxaluria that increase urinary calcium oxalate supersaturation. Since calculated calcium oxalate supersaturation did correlate with stone event rates, our findings also suggest that calculated calcium oxalate supersaturation is worthy of further study as a target for therapeutic intervention in PH3. Data from 3 large prospective cohorts also support the predictive value of calculated urinary calcium oxalate supersaturation for kidney stone events in non-PH populations[20].

Similar to prior PH literature, most analyzed stones (69%) were comprised entirely of calcium oxalate [1, 3, 5, 21, 26-28]. While we observed a higher proportion of stones containing calcium phosphate and mixed composition than previously reported in PH1 [3], rates were similar to PH2 from both the RKSC registry and the report of Daudon (Table 4) [29]. When Daudon and colleagues analyzed stone composition using stereo microscopy and infrared spectroscopy, COM was the sole component in 81/84 (96%) stones from 72 PH1 patients [29]. Interestingly, only 22% of stones in our PH3 cohort were comprised solely of COM, while 48% contained COD, either alone or mixed with COM. COM stones are characteristically associated with hyperoxaluria, while COD stones have been described in association with hypercalciuria [30] Among idiopathic calcium oxalate stone formers, Guerra et al. found that stone COD/COM ratio was correlated with higher urine calcium excretion[31] Urinary calcium excretion rates observed in PH3 are typically within the normal range and differ from low urine calcium excretion typically seen in PH1. Thus, the relative excretion rates of oxalate versus calcium could be a factor that influences the subtly different composition of stones between PH type. In addition, 22% of stones analyzed in our PH3 cohort were mixed calcium oxalate and calcium phosphate, consistent with several patients reported from the OxalEurope population [21]. Thus, stone composition in PH3 patients is more heterogeneous compared to PH1, and similar to PH2.

Our study has limitations. The small number of patients in our cohort is due to the rarity of the disease, its relatively recent identification, and the lack of universal and ready access to affordable genetic testing [32]. Thus, analyses in this study were exploratory in nature given the multiple comparisons without correction and lack of an a priori hypothesis to be tested. The identification of HOGA1 as the cause of PH3 occurred just a decade ago. Thus, many patients in our cohort that were diagnosed at a younger age have relatively shorter follow up. In contrast, many older patients were genetically confirmed later in life and had a more extensive medical history. This heterogeneity in observation periods may have influenced the results. Additionally, based on population data, the carrier frequency of PH3 is 1:185, suggesting the presence of many other individuals with PH3 who are undiagnosed or do not have clinical manifestations. Is unclear whether the number of symptomatic PH3 patients will increase in coming years as awareness of this disorder and wider availability of genetic testing occurs, or whether many patients with underlying HOGA1 genetic changes are instead asymptomatic. Lastly, data available from retrospective medical record review is inherently incomplete, and certain stone events may have been missed.

Despite these limitations shared by most retrospective cohort studies, this study of a large PH3 cohort provides a detailed analysis of stone events over a patient’s lifetime which pose a substantial clinical burden. Clinical stone events typically begin in early childhood, recur throughout life, cause symptoms, require surgical interventions, and therefore often significantly interfere with school attendance, employment, and quality of life.

Conclusions:

Stone events impose a considerable life-long burden in PH3 patients. There is a compelling need for more effective therapeutic agents. Calcium oxalate supersaturation of the urine positively correlates with lifetime stone events and may be a target for future treatment efforts.

Abbreviations and Acronyms:

HOGA1

4-hydroxy-2-oxoglutarate aldolase

PH3

primary hyperoxaluria type 3

IQR

interquartile range

IRR

Incidence rate ratio

COM/COD

calcium oxalate monophosphate/ calcium oxalate diphosphate