I do not know why anyone would diagnose distal RTA (dRTA) very often. As I will show you it has colorful and unusual characteristics as unmistakable as rare, so diagnosis is not difficult. But many more people think they have than have it. In my 50 years of kidney stone prevention I have perhaps a few dozen examples or so, out of many thousands of stone formers.
This is another of those long, elaborate articles only the most devoted read.
Even so, elaborate as it is, this article tells only part of the story. It simplifies or simply ignores the mechanism for low potassium in dRTA, and left for another time its genetic causes, and also the bone and mineral disorders and treatment outcomes. I forgive myself, as just this part has been most taxing to write and is equally so to read.
In a subsequent article I hope to expand on diagnosis and treatment, the bone and mineral disorders, genetic transporter disorders, and take up the novel modern issue of acid retention and its effects on kidneys. So consider the present article a part of my planned contribution.
The featured illustration of kidney tissue from a patient with dRTA shows many crystal deposits on a radiograph (panel a), that at surgery mostly are calcified deposits (panel b) that nearly replace papillary tissue (panels c and d).
What is distal RTA?
Kidneys make urine more acid than blood because most of us eat a diet that imposes an acid load on the body and kidneys need to remove that acid. But apart from balancing acid excretion to diet acid load, unless kidneys acidify urine calcium phosphate crystals may form in such profusion as to block kidney tubules and produce kidney stones.
This happens because as they conserve water kidneys concentrate urine calcium phosphate salts far above their levels in blood. If they simultaneously make urine more acid than blood the calcium and phosphate will stay in solution and make no crystals. That is the normal state of affairs.
But what would you predict might happen if by magic – bad magic – kidneys lost their normal ability to make urine acidic? Perhaps not all their ability, but some significant part of it? Yet withal, retained enough of their life sustaining glomerular filtration so that their other functions were close to or in fact normal.
You would predict acid would accumulate in the body and cause trouble, perhaps leach mineral from bone. Likewise you would predict calcium phosphate crystals will plug tubules and make stones. This because preservation of filtration assures delivery of ample calcium into the late nephron and final urine.
You would be right both times.
That is the very make and mark of distal renal tubular acidosis (dRTA).
What Does ‘Distal’ Mean?
The nephron begins at the glomerulus. The proximal tubule lies just beyond it. Prior articles do a reasonable job of showing where things are. Tubule cells acidify the filtrate in the proximal tubule, and then again later on, in the collecting ducts. When the latter is defective, urine itself is invariably too alkaline predisposing to calcium phosphate crystals and plugs. Such an alkaline urine is common among idiopathic calcium phosphate stone formers who certainly do not have dRTA. In those who do have such a defect, stones and crystals are usually far more massive and damaging to kidneys.
When the former is defective (proximal RTA), filtered bicarbonate cannot be fully reabsorbed as in normals, so more is delivered downstream to the collecting ducts. There, the cells are competent to secrete acid, but the amounts of bicarbonate can be so high as to use up all of it and escape into the urine. The result is too alkaline a urine. Stones, however, are not so likely. Serum and therefore filtrate bicarbonate falls, better matching what remains of proximal reabsorption so the distal delivery slows to a trickle.
But, sometimes a ‘proximal’ problem can indeed cause stones. For example drugs like topiramate reduce proximal tubule acid secretion and lead to calcium phosphate stones. In part this may reflect intermittent dosing, so serum bicarbonate rises between doses, and partly a more acid blood that lowers urine citrate, a powerful anti – stone molecule.
Our concern here is the distal form of RTA.
Clinical Laboratory Appearance
Should we not expect to see high urine pH – alkaline urine – with acid retention in the blood?
Some Explanation of Terms
Proton concentration varies over millions of fold from acids to alkali, and being a logarithm, pH spans that range gracefully. In fact, pH is the log to the base 10 of 1/[H+] where [H+] is the concentration of hydrogen ions – protons. For example, water has a pH of 7, meaning the concentration of protons is 10-7 (0.000,000,7) moles/liter. Blood is nearly the same, pH 7.4. But urine is normally pH 6, more than 10 times more acid than blood.
Blood is held at a pH just above that of water by a buffer system consisting of bicarbonate and carbon dioxide (CO2) gas. The latter is regulated by the brain using the lungs to clear CO2 from our metabolism at just the rate desired to maintain blood pH. Bicarbonate is a negative ion that can take up a proton – acid radical – or give one off. So it is a ‘buffer’ – meaning by its donation or uptake of protons it can stabilize the pH of blood.
When food and metabolism add protons to blood, those protons are ‘buffered’ by blood bicarbonate. But that buffering has a strange property. As it takes up a proton, bicarbonate becomes carbonic acid (H2CO3) that almost immediately decomposes into dissolved CO2 (dCO2) and water:
H+ + HCO3– <—–>H2CO3 <—–> dCO2 + H2O
Dissolved CO2 (dCO2) is in equilibrium with CO2 gas whose pressure in the blood (pCO2) is fixed by the brain and lungs. The lungs carry away any excess, so the partial pressure of CO2 gas in our blood can be very constant. Given this, when a bicarbonate buffers a proton It disappears into thin air – CO2 gas, actually.
The arrows point both ways, because these equilibria can move back and forth. If kidneys remove a proton – they do this to make up for what we eat – things simply reverse. Just imagine a tiny tweezers removes the H+ at the left of the equation above. Immediately, the lost proton is replaced by a proton from H2CO3, that then becomes HCO3–and the ‘lost’ H2CO3 is replaced from dCO2 and H2O.
Does this mean that new bicarbonate appears in blood out of thin air?
Yes. As kidneys remove acid to make up for our diet, they make new bicarbonate. They make it in the capillaries around the renal tubules that secrete acid into the final urine, and the bicarbonate flows into the renal veins, and thence into the general circulation through the vena cava.
Does that mean if we anchored a tiny canoe in the vena cava just at the mouth of a renal vein, and dipped a pH meter into the blood around, us the pH would be higher at the vein’s mouth than in the central parts of that great vein?
What we measure in blood is the total of HCO3– + H2CO3 + dissolved CO2 so we call it total CO2 or just TCO2.Of these, HCO3– massively predominated. So when TCO2 goes down, acid has been retained. When it goes up, acid has been removed – or new alkali added.
Lets Use the Terms
The big graph shows serum TCO2 and urine pH. For simplicity I have left out the ‘T’.
Common Stone Formers
The tiny blue dots are my data from common calcium stone formers. Most of the points center between 23 and 30 for total CO2. Urine 24 hour pH averages about 5.9 but ranges high and low because these people have normal kidneys that are adjusting for the diet they have chosen to eat.
Normal Men and Women
Blue large circles are from normal people we have studied. They gave us 24 hour urines and let us draw a blood. Nothing more. Remarkably, their points overlap exactly with those from the common stone formers.
dRTA and Acid Loaded Normals
The red circles are clinical measurements from my patients with known dRTA. A few could not or would not discontinue their alkali treatments; their points show normal serum CO2 values and very high urine pH.
The red up triangles show dRTA cases given an acid load to force urine pH as low as it can get. I have plotted this low pH against their serum CO2 values before the acid load – the value one would see clinically. So low values mean the patients have abnormal blood acidity even without an acid load.
The blue up triangles, from our own and other published studies, show acid loaded normal people. Once again, when possible, I have plotted their lowest urine pH values – from acid loading – against their CO2 values before the acid load. The big blue square is an odd patient from one study – see below.
For those who want to see the data I have collected from published sources, here is a link to the references and the values plotted on this figure. My clinical data are not in the spreadsheet because never published explicitly in this form.
What Does the Graph Say?
Normals and Common Stone Formers
The blue circles and tiny dots show us the variable urine pH of normal people and common stone formers. Their serum CO2 values are almost all to the right – higher than – those for the red symbols (dRTA) on average.
Acid Loaded Normals
Their most acid urine pH values are all below 5.3. A few cases lacked pre-acid load serum CO2 so I plotted the values after acid loading. The one blue square shows a single patient labeled as ‘incomplete’ RTA in one of the references. Although urine pH fell normally, a more elaborate test failed – I view the person as having normal renal acidification.
When the kidneys are disabled by dRTA, either genetic or from special kinds of kidney diseases, the serum CO2 falls because acid accumulates. It accumulates not because we gave an acid load but because the kidneys cannot cope with our normal acid load from food. But the lower serum CO2 – meaning acid blood – cannot drive kidneys to lower urine pH. So most of the red up triangles have the unique signature of dRTA: high urine pH and low serum CO2.
What about the red triangles with normal (above 23 mmol/l) serum CO2? For them I show their lowest urine pH after an acid challenge. It does not reach below the limits for normal people. These cases are labeled ‘Incomplete RTA’ and I accept them as they cannot lower their urine pH as much as normal people would given an acid load (blue up triangles). Unlike the dRTA patients with reduced serum CO2 those with incomplete dRTA can stave off ‘acidosis’, an acid blood as gauged by reduced serum CO2. But their kidneys respond to an acid load challenge less well than normal thus disclosing some defect in lowering urine pH.
Do Patients with dRTA Make Stones?
Yes. The cases, my few and those in the papers I selected, mostly came for care because of stones. Some had bone disease, some children did not grow properly. These are thought to be consequences of the acid retention.
They not only make stones, but generally the stones we expect: Calcium phosphate, the stones that form in alkaline urine. Like calcium phosphate stone formers, patients with dRTA can form mixed stones that contain some calcium oxalate, and occasionally produce stones predominantly of calcium oxalate. But in general calcium phosphate predominates. Likewise, patients with dRTA and idiopathic calcium phosphate stones plug terminal portions of their nephrons with masses of calcium phosphate crystals.
But unlike the not uncommon idiopathic calcium phosphate stone formers, those with dRTA have unrelentingly high urine pH and acid retention. Perhaps because the nephrons have lost their ability to lower pH, so tubule fluid is perpetually alkaline, they form much larger masses of crystals, in the urinary system and as terminal collecting duct plugs. Radiologists often encounter massive calcium deposits in the kidneys and label the condition ‘nephrocalcinosis‘,
We were not the first group to study the kidney tissue in dRTA, but those before us sampled the kidney cortex, where glomeruli and proximal tubules predominate. We expect crystals to form not there but in the terminal portions of the nephrons, where urine is most concentrated, and it is precisely from there we obtained our tissue samples, and within that narrow precinct where we made our observations.
Radiographs Can Mislead
Among our few cases (5) we described two with multiple masses of calcifications in the kidneys – nephrocalcinosis. As in phosphate stone formers, these calcifications can be stones or masses of tissue calcium deposits.
Here is one kidney. Many calcified objects more or less fill it. Arrows point to some obvious clusters. This is a perfect image of what radiologists correctly label ‘nephrocalcinosis’. It is the name I would use. The large clusters marked by asterisks were in an obstructed upper pole calyx and certainly masses of crystal, but whether in tissue or the collecting system we would not be able to find out.
Inside the kidney, at surgery, no stones. Instead, the papillae were filled with crystals inside the tubules.
Here is a biopsy from a papilla. We studied it using high resolution micro-CT. Inside the tissue are masses of crystal deposits filling much of the space. In fact, when cut by a microtome to make microscopic sections, it shredded as the knife could not well cut the crystal deposits and pulled the tissue apart. The arrow points to a huge agglomeration of crystals in tubules.
Kidneys in dRTA
You might say that in plugging their terminal nephrons with crystals, and forming mainly calcium phosphate stones, patients with dRTA are just like calcium phosphate stone formers, who do the very same thing. But saying this is to miss the main point. When compared quantitatively in terms of numbers of deposits and deposit size dRTA vastly outweighs other stone diseases.
Elsewhere we have compared deposit size and numbers in multiple stone diseases, but not exactly as I propose to do here.
Idiopathic Calcium Phosphate Stone Formers
Brushite stone formers – one of the two types we encounter – plug occasional tubules with large masses of crystals, so that one tubule is utterly destroyed, but most tubules are fine. See its point on the graph? About 3 deposits per mm2 but over 1.5 mm in size.
Apatite stone formers plug a higher fraction of tubules but with much smaller deposits. Their point on the graph shows about 12 deposits per mm2 but only about 0.6 mm in size.
By contrast, dRTA is a more diffuse disease, affecting more or less all of the tubules. So mineral deposits are much more dense – about 45 deposits per mm2. They are also larger than apatite deposits, about 0.8 mm, so the total volume of tissue occupied is much greater.
Other Kinds of Stone formers
Notably, primary hyperparathyroidism (PHPT) is about like apatite stones and ileostomy – the latter a state of constant high risk from dehydration.
The other stone diseases cluster to the far left: small bowel resection, obesity bypass, cystinuria, uric acid stones. They cause few and small deposits. The most common calcium stone disease of all, idiopathic calcium oxalate stones, causes very few and small deposits. We found almost none; the Mayo Clinic group described tiny ones.
None match dRTA for its sheer diffuseness, its contamination of so large a fraction of tubules with crystals. Take another look at the micro-CT. You can see very little uninvolved tissue.
What Makes Crystals So Massive?
For this analysis, I will use only dRTA cases in my clinical series because their lab values were obtained exactly as were values for the other stone forming groups I need as a contrast.
Certainly the high urine pH is part of the story.
This quantile plot (qplot) shows the distribution of values for uric acid stone formers (red circles, to the left), CaOx stone formers (blue circles), normal people (black circles), CaP stone formers (green triangles), and our dRTA patients in red large squares.
It is true that the CaP stone formers make a more alkaline urine than do the CaOx stone formers, and that is a perfectly reasonable mechanism for their stones. The higher pH removes a proton from the second site on phosphate so a higher fraction of urine phosphate has two negative sites to mate better with calcium that has 2 positive sites – I know this is a bit oversimplified, but it is not wrong.
Likewise, the much lower urine pH of uric acid stone formers is indeed exactly why they produce those self same stones.
But RTA is a powerful exaggeration of high pH. The whole group of patients is shifted about 1/2 pH urine up, and pH is a logarithm. This will free up even a higher fraction of phosphate to mate with calcium, thereby raising supersaturation for any urine calcium, volume and citrate.
Urine Calcium and Citrate
dRTA causes not only remarkably high urine pH, but also an extreme deficit of urine citrate, and the latter certainly contributes to the exuberance of crystallizations.
The Calcium to Citrate Proportion
Other articles explore the remarkable properties of urine citrate. It binds urine calcium into a highly soluble complex so it is not free to combine with oxalate or phosphate. If crystals do form, citrate can attack their lattice and abridge growth, even in some cases destroy them.
Essentially, the citrate in urine competes against oxalate and phosphate for the attentions of calcium. Especially when pH is high, and a large fraction of phosphate has two negative charges to offer as calcium binding sites, the relative concentrations of calcium, phosphate and citrate are critical in determining how much calcium will combine with each. Moreover, to poison new formed crystals and abridge their growth, citrate must be free – not bound to calcium. So howsoever wonderful citrate may be as a calcium binder and crystal inhibitor, when it is outweighed by excess calcium its magical properties do not so much wane as become inadequate to the tasks.
Calcium to Citrate Plot
This means that the relationship between urine citrate and calcium cannot be but critical in the assessment of stone forming conditions.
The best way to visualize this relationship is simply to graph one against the other in comparable units. On the graph below, urine citrate is plotted against urine calcium, both in mmol/d. The long blue dashed diagonal line of identity bisects the graph. Points above the line mean an excess of citrate over calcium, below the opposite.
Uric Acid Stones
Uric acid stone formers (red circles) scatter above and below the line of identity. The best fit regression line – red line – runs below identity. Note the regression lines stretch no further to the right or left than the highest to lowest urine calcium values.
Normals and CaOx Stone Formers
Like uric acid stone formers, values from normals scatter above and below the line of identity (black triangles). The black regression line is not different from uric acid.
CaOx Stone Formers
Being so numerous, I plotted calcium oxalate stone formers using tiny blue dots. Otherwise they would obliterate the other symbols.
Their regression line (solid blue line) stretches to the right border of the figure because of hypercalciuria – that much raises urine calcium. Although the regression lines for uric acid stone formers, normals, and CaOx stone formers are more or less the same, one can see the last of the three is below the other two. The lower proportion of citrate to calcium would favor calcium crystallization. But the lower pH would make that crystal CaOx, not CaP. And in general, the result is stones, with little or no tubule plugging.
CaP Stone Formers
These nearest relatives to RTA have, like them, a higher than average urine pH and, interestingly, a lower level of citrate in relation to urine calcium than the other aforementioned groups. See where the CaP stone former points (green) have a regression line that moves in parallel with the blue line for CaOx stone formers but lies lower down. For any amount of calcium they have less citrate to put against it. And, they have a higher urine pH. So they would be facile producers of calcium phosphate crystals, and plug some tubules, like dRTA patients do. Their regression line runs to the right hand margin of the graph because CaP stone formers, like CaOx stone formers, are often very hypercalciuric.
At the very bottom of the graph, in large red squares, they have more or less no citrate at all to balance their calcium. Their regression line is flat and also short, the latter meaning they have much less urine calcium losses than either CaOx or CaP stone formers. But despite that limitation, their exteme citrate deficit and extreme pH elevation permits formation of masses of crystals to fill most collecting ducts and create numerous calcium phosphate stones.
Adjusted Mean Citrate Excretions
One cannot leave this without some quantification of differences. In a general linear model with urine citrate as dependent, urine calcium as independent and stone type as categorical variable, the mean urine citrate excretions (means in mmol/d) adjusted for urine calcium differ.
The highest citrate is among normals, the very lowest among dRTA. In a simple comparison between all groups taken two at a time, differences are highly significant – p<0.0001 in all cases except CaOx vs. None (p=0.0073) and CaP vs dRTA (p=0.0003). These p values are adjusted for multiple comparisons. The high significance for dRTA is remarkable given I use only my small group of cases. I chose against merging data from published papers with mine, because their conditions of lab measurement and patient condition might not match.
Slope Dependence of Urine Citrate on Urine Calcium
Although the slopes look different on the graph – citrate rises less with calcium in dRTA, for example, this difference is not significant – p=0.26. The four regression slopes are: CaOx SF = 0.197, CaP SF = 0.171, Normals = 0.156, and dRTA = 0.027 mmol/d citrate per mmol/d calcium. Given the huge spread, I suspect with more data the low slope for dRTA would be highly significant.
You might ask why urine citrate indeed varies so remarkably with urine calcium. In our regression, the F value for urine calcium as a covariate was 17, that for stone former type was 6.28. I do not know why.
What Makes Urine Citrate So Low?
Kidneys can let filtered citrate out into the urine, or reabsorb it into its proximal tubule cells where it becomes fuel for energy production. This reabsorption is regulated by the acid – base status of the blood and the kidney tubule cells.
The ‘acidosis’ of dRTA is reflected in low serum CO2. Kidney cells can sense that condition and reabsorb citrate.
One might at this point like some hard statistics about the actual values of serum CO2 even though the large pH graph has made a visual impression.
Serum CO2 is low meaning blood is acidic. In an ANOVA without any adjustments (values for least square (LS) means are mmol/l):
|Factor||Level||LS Mean||Standard Error||N|
Serum CO2 is obviously low in dRTA, the very hallmark of the disease. Pairwise comparisons show dRTA is below each of the other groups with very high significance (p<0.0001 all three comparisons with full adjustment for multiple comparisons).
Blood Potassium Is Often Low
Not only does dRTA cause acidosis, it causes potassium depletion. In turn, potassium depletion lowers the pH inside kidney cells, and thereby and independently of acid retention, raises citrate reabsorption so urine citrate falls.
Normal subjects and a mass of calcium stone formers are shown as large blue circles and a cloud of tiny blue points. The normal points overlay those from stone formers because common stone formers, before treatments that might alter it, have normal values for serum potassium and CO2. The large blue square is an ‘Incomplete’ RTA patient, the large blue triangle a normal mean, both from published studies.
Values from dRTA patients are in red up triangles. The larger red triangles are from my practice, the smaller ones are published data. The spreadsheet of all published data is available for review.
Not always, but very often, dRTA produces a low serum potassium. Because most of the potassium in the body is inside cells, low serum values almost always mean low potassium levels inside cells.
Although serum potassium and CO2 appear correlated, they are not (Using only my own clinical data, p for correlation in a general linear model = 0.166). But in a pairwise comparison of mean values adjusted for serum CO2 the mean value of serum potassium from dRTA differed from those of normals, and CaP and CaOx stone formers: dRTA=3.73 vs. normals = 4.18, CaP SF = 4.24, and CaOx SF = 4.29 mEq/l, respectively, p< 0.0001 for all comparisons but dRTA vs. none (p=0.003).
Although I hesitate to make direct comparisons or results from our clinic and the published sources in my spreadsheet, the latter give results much like ours. The mean of 41 published serum potassium values from dRTA cases is 3.64 (95% CI for mean 3.48 – 3.81). Therefore the mean of our cases – 3.73 lies within the 95% CI for the published cases.
The Anion Gap
As though I have trapped you in an inexhaustible maze, yet another detail of dRTA demands our attention. So esoteric a point yet one we must encounter and wrestle with.
Diet Acid Load
Protons do not come into the blood unattended. They come with a negative partner, the so called proton donor. Carbonic acid is a proton donor – it can give off a proton. But the one I care about – and you need to care about – is sulfate. The diet acid load is mainly – for the most part – sulfuric acid derived from the oxidation of sulfur on the two amino acids cysteine and methionine. Both occur naturally in food.
Offsetting the sulfate acid load, foods contain considerable amounts of mostly potassium anion salts that are metabolized to alkali – bicarbonate – in the body. The difference between the sulfate and alkali anions measures the acid load from a given food, and the average of foods the acid load from a diet.
Cheeses and egg yolks are highly acid loading, raisins alkali loading, as examples. One can find charts depicting a majority of foods. Here in the US and most of the West, the diet is overbalanced toward net acid load. If it were biased toward alkali load, dRTA would not appear as it does. Tubules would still be unable to lower urine pH but serum CO2 need be no lower than among normals. Bone mineral could remain stable, and urine calcium lower. It would be as if dRTA were treated with extra potassium alkali.
Reduced Kidney Function
As kidneys succumb to disease, they lose their ability to remove diet acid efficiently, and serum CO2 falls, just like in dRTA. But, the anions, sulfate mainly, from which the extra protons arose, are retained along with the protons. This is because kidneys have lost their normal glomerular filtration rate, and cannot clear sulfate from blood normally. The disease may or may not have impaired the ability of tubule cells to lower urine pH, so urine pH may be low or high. If it is high, one might imagine a false diagnosis of dRTA.
This points out in practical terms how the very essence of dRTA is a loss of tubule acidification out of proportion to loss of glomerular filtration. The very name of the disease is renal tubular acidosis, and so pronounced, to overweight and emphasize the specific loss of a tubule vs. a glomerular function. So when filtration is low, one might think everyone would avoid the diagnosis as essentially moot. The problem is that dRTA can lead to kidney disease from all the crystal deposits, and perhaps the many stones, too. Moreover, along their progress to a final death of function, kidneys may pass through transient phases where loss of acidification exceeds loss of filtration, giving rise to periods of disproportionate acid retention.
The Anion Gap Discloses
Blood has many components, but few with both charge and high abundance. Those few are sodium, potassium, bicarbonate, sulfate, and chloride. If you add up the charges on the positive ones – sodium + potassium – and subtract those of the negative ones – sulfate and chloride, you would get the residual anion gap.
But we almost never measure serum sulfate. Instead we calculate the – much larger – anion gap that excludes sulfate from the calculation leaving it – so to speak – as part of the gap.
In dRTA, most of the time, the loss of acid excretion far exceeds the loss of glomerular filtration, so the anion gap – mainly sulfate – remains near normal while the serum bicarbonate – gauged by serum CO2 falls. Essentially chloride ion replaces the lost bicarbonate.
A Way To Picture It
What all this means is that as serum CO2 falls in dRTA, serum chloride should more or less rise in proportion. But when glomerular filtration falls, chloride will not rise as briskly because kidneys no longer excrete sulfate efficiently.
On this plot, tiny blue points are CaOx stone formers. I have omitted other groups for visual clarity. Red triangles and circles are dRTA with creatinine clearances above 60 ml/minute, reasonable filtration. Green triangles are patients with clearances below 60 ml/minute.
For the most part, serum chloride does rise smoothly with fall in serum CO2 when clearance is above 60, and most of the failures – lower serum chloride for a given serum CO2, are from patients with clearance below 60 (green). Serum sodium must play an obvious role, and in fact accounts for the few scattered red points to the lower left.
After all this, this labyrinthine odyssey, this thicket of numbers and graphs, what are patients to understand, and what am I telling physicians who are not themselves as particularly interested in dRTA as I am?
Detection and Diagnosis
This disease is detected from fasting serum CO2 and chloride. The former is below normal, the latter runs high so the anion gap is not much above 12 or so. If glomerular filtration is reduced, below 60 ml/min, for example, the gap may have increased.
Diagnosis is the combination of reduced serum CO2 with a urine pH that is not maximally reduced – below 5.3. But in fact, most dRTA produce a urine far higher in pH, usually above 6. The graph I presented earlier of all dRTA cases shows how few have urine pH values below 6 even during acid loading.
Low serum potassium otherwise unexplained – diuretics, laxatives, vomiting – adds weight to the diagnosis. So does a near absence of urine citrate in the 24 hour urine.
Family history of RTA is so obvious a clue I hesitate to mention it here.
Expected Values for Key Measurements
In fairness, reliable values for serum chloride, CO2, anion gap, and sodium – the last part of the gap calculation – are not readily available for stone forming populations. So I calculated them from my own files, and show mean values here for dRTA, struvite, CaP, uric acid (UA) and CaOx stone formers, along with normals (N). Values are with standard deviations, not standard errors, to show the variability of the values within the populations.
RTA patients are the only ones with low average serum CO2 values, although struvite patients – now that I look at them – are also a bit low. Anion gaps all range between 11 and 12 on average. Serum sodium varies little, between 140 and 141 as a mean.
Since the anion gap and serum sodium in dRTA are like that of all the other conditions, yet the serum CO2 is low, serum chloride must be high – and is.
So it is the balance between serum chloride and CO2 that really swings in dRTA compared to – I guess I can make this sweeping a statement here – all other groups.
Sources of Confusion
If serum CO2 is clearly low – below 23 and preferably even lower, the anion gap is below 12, and 24 hour urine pH is clearly high – above 6, serum potassium is low, and urine citrate low as well, one need not scruple much about the diagnosis except for a few exceptions.
One such is infection with bacteria that hydrolyse urea to ammonia and thereby produce ammonia. Mostly that kind of infection is obvious because urine pH is very high, often above 8, and urine ammonia as well. Another confusion, rather esoteric, is over-ventilation that has reduced CO2 gas pressure and therefore lowered total serum CO2, but raised blood pH. This occurs with normal pregnancy, and can be an artefact during blood drawing – anxiety stimulating hyperventilation. If one confidently excludes these confounders, the combine of low serum CO2 and urine pH above 5.3 – usually higher! – will usually do.
Of course, chronic kidney disease will lower serum CO2, but everyone measures eGFR and will know. Typically, CKD below stage 3b (3a and better) rarely lowers serum CO2 below 23, and when it does urine pH is usually acid and the anion gap may be high. Moreover, contemporary treatment of CKD is to treat low serum CO2 with alkali, so if RTA is present or not one will do the same thing. I have already remarked that along the path to end stage kidney disease acidification may fail to keep up with filtration rate so transient non gap acidosis occurs, and will not explore that topic further.
Sometimes some physicians will want to know for sure. One alternative is to administer an oral acid load, typically ammonium chloride, and determine the minimum urine pH achieved. Another is to administer a sodium retaining steroid and a dose of furosemide and determine the lowest pH attained. This reference discusses both. When a young man I did these things, and sometimes published my results, but stopped years ago because almost no patient posed a real problem. Persistently low serum CO2 with an alkaline urine, normal anion gap and low urine citrate is rather obvious. Just the low serum CO2 itself is, after all, something needing remedy. So, I no longer put my patients to the bother of such testing.
My Own Prejudices
Here is a bafflement for me. Normal blood findings, but if challenged with an acid load these patients cannot acidify their urine as normal people do. Urine citrate may be low, but simple CaP stone formers have lower urine citrate than CaOx stone formers, and I doubt all of them have acidification problems. Excellent scientists have found such cases, and I show points them on my graph as red symbols – meaning failure to lower urine pH fully – with normal serum values. My question is simply what to do about such patients as I nor any other physician can find them sans a provocative test that challenges urine acidification.
I do nothing but treat them as idiopathic calcium stone formers. If low urine citrate is raising stone risk, I attempt to correct it. Likewise for all other stone risks for which we have some confidence: urine volume, calcium, and oxalate.
Burden of Heterozygosity. The brilliant work of Orson Moe showing some people heterozygous for dRTA have impaired acidification tells us how significant the research yield from acid loading tests might be if we all did them. In other words, inability to lower urine pH maximally may point to carrier states for hereditary dRTA. But this offers no specific therapy beyond what we always offer and therefore one cannot justify such testing as patient care but only as research.
Prospective Study of Minimum Urine pH. A recent and important prospective study of maximal acidification using both acid loading and a more technical method of lowering urine pH suggests incomplete dRTA may be a diagnosis without meaning. This snip from figure 2 of the paper shows one of their main points. The minimum urine pH is a distribution, and 5.3 merely one signpost along the way. In fact the central position of the pH distribution lies at about pH 4.7. No break in the distribution marks pH 5.3 as special.
Although urine citrate is indeed lower in those with a minimum urine pH below 5.3, a glance at their graph (not reproduced here) of urine citrate excretion and minimum urine pH shows – to my eye – no significant correlation.
Urine calcium excretion is unrelated to minimum urine pH.
The View from the Trenches
I wrote to Dr. Daniel Fuster who has guided this work and he offered his personal view:
‘I think only longitudinal studies that reveal that this subset of patients is indeed unique as it deserves special treatment can rescue provocative loading / idRTA unmasking. Elevated urinary pH, hypocitraturia and low bone mass are also not specific to idRTA but common findings in stone formers and we adapt our treatment to the individual phenotype of the stone former. There is much to be learned about stone formers with high urinary pH. But I am afraid, idRTA as distinct entity in stone formers is likely gone.’
Requiescat in pace.
Causes of dRTA
Being uncommon, dRTA patients invariably attract considerable clinical attention, and the cause of the RTA rarely escapes detection. Given this, I make only primitive remarks about the subject. Textbook chapters and reviews carry long lists of possibilities not appropriate for this kind of overview. A brief and recent one appeals to me for its restraint and focus. But a brief search on PubMed will yield many others as well.
Autoimmune diseases (sjogren’s syndrome, lupus, rheumatoid arthritis, and both primary biliary cirrhosis and autoimmune hepatitis) are not rare causes. Renal sarcoidosis and amyloidosis are causes. I have found most of my acquired dRTA among those with Sjogren’s syndrome and recommend this excellent review.
In a search for a good review of genetic causes of dRTA, this somewhat older paper deserves attention. Likewise, the remarkable Moe paper on heterozygotes has what seems a complete listing of genetic papers. Family history of dRTA is common in patients with dRTA and a common clue to diagnosis. Conversely, a diagnosis of dRTA should alert families that may harbor more people who have the condition but may not know.
Amphotericin B, lithium, and trimethoprim are commonly used and can cause dRTA. Drugs like Topiramate and other carbonic anhydrase inhibitors interfere with acid secretion in the proximal tubule. They produce normal anion gap acidosis, and urine pH can be high, citrate low just as in dRTA. So evaluation must be very sensitive to all medication use.
Treatment of dRTA
So massive an exposition and so narrow a range of treatments!
If any systemic disease is the cause of dRTA, sjogren’s syndrome for example, one treats it as a primary aim. Oftimes, the dRTA can improve, but may not. If a drug is culprit we stop it.
Lacking a systemic disease or drug, which usually means the trait is genetic, or in the event that treatment of systemic disease leaves residual systemic acidosis, one treats with potassium alkali in a dose sufficient to restore serum CO2 to within the normal range. We use potassium, not sodium alkali to restore systemic potassium deficits and avoid raising urine calcium by sodium loading.
If urine calcium is high and does not fall to with treatment of acidosis, I lower diet sodium and, sometimes, urine thiazide diuretics as in idiopathic calcium stone formers. High urine oxalate can be treated, as always, with reduced oxalate intake and increased diet calcium. In other words they are treated as every idiopathic calcium stone former with the exception of a requirement for alkali.
Treatment of dRTA in children is perhaps more urgent than in adults. This form of dRTA is almost always genetic. A single center retrospective analysis of treatment for all patients at a referring center in India. Failure to grow and rickets, were both the rule. With treatment, growth resumed and rickets resolved, in most cases without added vitamin D treatment. In a multivariate analysis only initial growth deficit correlated with eventual height gain – greater deficit, greater gain. Nephrocalcinosis was common and associated only with higher urine calcium excretion. They found that acidosis, not malnutrition was seemingly the main factor affecting growth.
In another article – this one is already too long and tiring – I will review more of the scanty long term outcome data for this disease. But it will add little. We must always correct acidosis, and then use fluids, diet, and our few other means to control stone risk.
A Final Word
Almost no stone formers have dRTA, it is very uncommon. Those told they have it should inquire as to the criteria employed and – I hesitate here – harbor a sceptical sensibility best soothed by a second opinion. Incomplete dRTA seems ridiculous to me and Dr Fuster. It is a name without clinical value in that treatment proceeds as if it were not present. More than ridiculous, it distracts physicians and patients, and can imply to a common calcium phosphate stone former much worse pathology than they have. With him, I recommend we stop using the term outside of research protocols.
Thanks to Joan Parks
The large and colorful graphs of data from our practice all come from a table that Joan made years ago and left me as a legacy.
One might think it a small thing to compile its thousands of rows of data, given we have computers. But that is naive and foolish. One can accumulate considerable volumes of trash from almost any collecting source – like our electronic medical records – but well curated data sets from patient care sources take a long time to clear of error.
In her long career, Joan did this many times.
Although she retired some years ago and now writes excellent novels, we benefit, still, from her legacy of remarkable and accurate research work.