A 49-year-old, female with a 5-year history of diabetes mellitus type 2, presents for an initial visit. She has no known complications of diabetes. She takes metformin, glyburide, and aspirin. On examination, you find a pleasant, obese female in no distress. Her blood pressure is 136/86 mm Hg. As you discuss monitoring her diabetes, you recommend screening for early kidney disease.
Question 5.1.1 Which of the following approaches is the recommended way to screen for diabetic kidney disease?
A) Obtain a 24-hour urine collection for albumin now and again in 3 years.
B) Obtain a spot urine albumin every year.
C) Obtain a spot urine albumin/creatinine ratio every year.
D) Obtain a urinalysis every year.
E) Obtain a serum creatinine every year.
Answer 5.1.1 The correct answer is "C." Moderately increased albuminuria (previously known by the misnomer "microalbuminuria") is a marker for increased risk of future kidney disease in diabetic patients. The best test to evaluate for moderately increased albuminuria is the urine albumin/creatinine ratio. Its advantages include ease of use, relatively low cost, and good correlation with 24-hour urine collections. Some of you may have chosen "B." A "spot microalbumin" (now a spot albumin) is a common but less accurate way to provide screening and may still be in use in some areas. As a practical matter, many physicians use urine albumin alone as a method of screening, but this method does not allow for corrections for variations in urine volume and dilution. A random spot urine albumin/creatinine ratio is normally less than 30 mg/g. Values above 30 mg/g are consistent with 24-hour measures showing abnormal amounts of albumin. Answers "D" and "E" offer measures of kidney function that simply are not sensitive enough to use for screening purposes.
Note on terminology. The term "microalbuminuria" has been replaced with "moderately increased albuminuria," but the definition remains the same (30–300 mg protein in the urine per 24 hours). Compare this to "severely increased albuminuria," defined as greater than 300 mg of protein in the urine per 24 hours.
Question 5.1.2 Her albumin/creatinine ratio is 42 mg/g. The next step to confirm moderately increased albuminuria is:
A) Repeat urine albumin/creatinine ratio.
B) Urine dipstick for protein.
C) 24-hour urine collection for total protein excretion.
E) Referral to a nephrologist.
Answer 5.1.2 The correct answer is "A." Verification by repeat urine albumin/creatinine ratio is sufficient for a diagnosis of moderately increased albuminuria, so 24-hour urine collections need not be performed for confirmation. Of note, the diagnosis of moderately increased albuminuria requires 2 of 3 urine specimens showing >30 mg/g albumin/creatinine over a 6-month period. Since protein excretion must exceed 300 to 500 mg/day for a urine dipstick to detect proteinuria, urinalysis ("B") is not sensitive enough to be diagnostic. Serum creatinine elevation may be a marker for diabetic kidney disease, but it would develop late in the process. Nephrology referral is premature—you went to medical school; you can do this!
Question 5.1.3 Which of the following can cause a false-negative albumin/creatinine ratio?
D) Poor glycemic control.
Answer 5.1.3 The correct answer is "E." Patients with a large muscle mass have a high rate of creatinine excretion, which may result in a falsely negative albumin/creatinine ratio (as the urine creatinine goes up, the ratio obviously goes down). Cachectic patients have the opposite problem, with low amounts of creatinine excretion, resulting in false-positive albumin/creatinine ratio. Fever, vigorous exercise, heart failure, and poor glycemic control can cause transient increases in albuminuria, potentially resulting in false-positive albumin/creatinine ratios.
Your patient's other laboratory studies reveal the following: hemoglobin A1c 6.4%, serum creatinine 1.4 mg/dL, and normal electrolytes. A month later, your patient returns. Her blood pressure is 138/84 mm Hg. Her urine albumin/creatinine remains elevated on a second measurement. According to an eye examination yesterday, she has nonproliferative diabetic retinopathy.
Question 5.1.4 Because your patient has type 2 diabetes mellitus and moderately increased albuminuria, you realize that her likelihood of progressing to overt nephropathy is:
B) About half that of a similar patient with type 1 diabetes.
C) Nearly equal to that of a similar patient with type 1 diabetes.
D) More than twice that of a similar patient with type 1 diabetes.
E) Absolutely certain (100% chance).
Answer 5.1.4 The correct answer is "C." Although earlier studies showed a greater progression to overt nephropathy in type 1 diabetics, more recent studies demonstrate a nearly equal rate of progression in types 1 and 2. About 20% to 40% of Caucasian patients with diabetes type 2 and moderately increased albuminuria will progress to diabetic nephropathy. The rate of progression to nephropathy in non-Caucasian populations is even higher.
Question 5.1.5 What is the most appropriate next step in the evaluation and management of this patient's moderately increased albuminuria?
A) Start an angiotensin-converting enzyme (ACE) inhibitor.
B) Start an angiotensin receptor blocker (ARB).
C) Order renal ultrasound with Doppler of the renal arteries.
E) Order a 24-hour urine collection for total protein.
Answer 5.1.5 The correct answer is "A." ACE inhibitors should be the first-choice drugs unless there is a contraindication to their use. "B," an ARB, should be second-line choice in the event that the patient cannot tolerate an ACE inhibitor. "C," a renal ultrasound, is not indicated at this point in time. "D" is also incorrect. Your patient already has good glucose control (HbA1c of 6.4%). As previously stated, a 24-hour urine collection is not necessary for diagnosis.
HELPFUL (AND VERY IMPORTANT) TIP:
ARBs are not just an ACE inhibitor without the cough. A large meta-analysis showed that ARBs, while lowering blood pressure, have no appreciable effect on MI or cardiovascular mortality versus placebo (BMJ. 2011;342: d2234; JAMA Intern Med. 2014;174:773). But that's heart disease in diabetics (which is kind of important). Back to renal disease … The only study (to date) to directly compare an ACE inhibitor (enalapril) to an ARB (telmisartan) for renal protection in diabetes did not demonstrate a statistically significant difference between the two, although there was a trend in favor of the ACE inhibitor (N Engl J Med. 2004;351(19):1952). However, this was a small study (250 individuals) with a high dropout rate. Another study of 4,000 subjects showed that while ARBs do reduce proteinuria, they have no benefit on mortality (N Engl J Med. 2011;364:907), which is what we really care about.
The patient has a full urinalysis to rule out renal inflammation (e.g., nephritis) and overt proteinuria (nephrotic syndrome). The urinalysis is entirely negative.
Question 5.1.6 What further investigations must your patient undergo to eliminate other potential causes of proteinuria?
B) Renal ultrasound with Doppler of the renal arteries.
Answer 5.1.6 The correct answer is "E." No further evaluation is necessary in this patient with moderately increased albuminuria. The combination of diabetic retinopathy (a marker for diabetic renal disease), hypertension (BP >140/90 mm Hg in a diabetic patient), and abnormal protein in the urine as measured by the urine albumin/creatinine ratio is sufficient to make the diagnosis of early diabetic nephropathy. Renal biopsy is quite invasive and unlikely to change management. ANA, ESR, CRP, and ultrasound are unlikely to offer new information. If things change (e.g., nephritic urine and gross proteinuria), further evaluation may be indicated.
You continue to follow this patient for several years. She ultimately is admitted for chest pain, rules out for myocardial infarction, but has a positive stress test. She will need to have a cardiac catheterization. Her creatinine is now 1.5 mg/dL, and her glomerular filtration rate is in the range of stage 3 chronic kidney disease.
Question 5.1.7 In addition to holding her metformin, which of the following interventions would be most likely to reduce her risk of developing contrast-induced nephropathy?
A) N-acetylcysteine and IV saline.
B) N-acetylcysteine and mannitol.
D) Sodium bicarbonate and mannitol.
E) Mannitol and IV saline.
Answer 5.1.7 The correct answer is "C." For patients at risk of contrast-induced nephropathy, contrast studies should be avoided if possible (see helpful tip below). If a contrast study must be done, stop aggravating medications like nonsteroidal anti-inflammatory drugs (NSAIDs-–this patient shouldn't be on these anyway since she has CKD!). Hydration, usually with IV saline, should be given if there are no contraindications (e.g., heart failure). Nonionic lower osmolality (or even iso-osmolar) contrast agents should be used, and we recommend discussing these cases individually with your radiologist, cardiologist or whomever is performing the procedure.
Sodium bicarbonate and N-acetylcysteine do not have good evidence for effectiveness, but they also do not appear to do harm. Sodium bicarbonate (isotonic sodium bicarbonate solution at 3 mL/kg for 1 hour before the procedure and continue at 1 mL/kg for 6 hours after the procedure) can be used as current evidence indicates that it has similar efficacy to IV saline. N-acetylcysteine needs to be started the day before the procedure and the evidence is conflicting. The diuretics (mannitol, furosemide, etc.) may be associated with an increased risk of nephropathy, and mannitol in particular has an undesirable side effect profile.
The old teaching about renal toxicity and iodinated contrast has recently been called into question (Radiology. 2014;273(3):714–725; Radiology. 2014;271(1):65–73). These well done articles have been game changers. Do not start using contrast with wild abandon in renal disease. However, if it is really needed, it seems that the risk of renal injury is low. One caveat; few patients in these studies had a creatinine of >3 mg/dL.
Hemofiltration and hemodialysis have been studied in the prevention of contrast-induced nephropathy for patients at risk. There does not appear to be a benefit over more conservative approaches. Also, there is a move to dialyze patients who are already on dialysis within 24 hours of a contrast study. There is no good data, however, and studies show potential harm. Their kidneys are already shot … they are on dialysis. What additional harm can you do to their kidneys? (There is some argument for patients who continue to produce urine despite having end-stage renal disease and being on hemodialysis).
HELPFUL (AND VERY IMPORTANT) TIP:
The use of MRI contrast (gadolinium) in patients with renal disease has been associated with a scleroderma-like syndrome (nephrogenic systemic fibrosis). Gadolinium should be avoided in those with a CrCl of <30 mL/min. The data are less compelling in those with a CrCl of 30 to 60 mL/min. Since nephrogenic systemic fibrosis can affect more than just the kidneys, early dialysis may help prevent this disease by reducing the half-life of the gadolinium.
Further, it has recently been discovered that repeated use of MRI contrast may remain in the CNS and form deposits. What affect this has on cognition, etc. is unknown, but is being investigated by the FDA. FDA MedWatch suggests limiting the use of MRI contrast (http://www.fda.gov/Safety/MedWatch/SafetyInformation/SafetyAlertsforHumanMedicalProducts/ucm456012.htm).
Upon cardiac catheterization, she is found to have several lesions. She undergoes coronary artery bypass grafting and is discharged. Her creatinine remains stable at 1.5 mg/dL.
A few months later, she presents with gradually increasing dyspnea and cough. Her urine output is reported to be normal. Her vitals show temperature 37°C, pulse 76 bpm, respiratory rate 24, and blood pressure 92/46 mm Hg. You note crackles at both lung bases. Her heart rhythm is regular and an S4 is audible. She has JVD of 9 cm and 2+ pitting pretibial edema.
The ECG shows sinus tachycardia but no evidence of potassium toxicity (peaked T-waves). Laboratory results: troponin-T and CK normal, BUN 70 mg/dL, Cr 2.0 mg/dL (her baseline was 1.5 mg/dL), Na 128 mEq/L, K 5.5 mEq/L (reference range 3.5–5.0), HCO3 19 mEq/L, WBC 14,500 per mm3, remainder of the CBC is normal. Urinalysis shows protein and glucose and a specific gravity >1.030, but there are few cells and no casts. Cultures and a chest x-ray are pending.
Question 5.1.8 You suspect that her elevated creatinine is primarily due to which of the following processes?
A) Adverse toxic effects of drugs on the kidney.
B) Heart failure or other prerenal cause of renal failure such as dehydration.
C) Sudden progression of diabetic nephropathy.
E) Urinary tract infection.
Answer 5.1.8 The correct answer is "B." Your patient has clinical evidence of heart failure (HF), including edema and rales. With a BUN/Cr ratio greater than 20 and an elevated urine-specific gravity, she appears to have a prerenal azotemia, likely secondary to insufficient cardiac output. Drugs ("A") may play a role in her dehydration, but the BUN/Cr ratio of >20 argues against a renal cause of her elevated creatinine. Therefore, the most likely culprit is HF. Diabetes should not cause a sudden worsening of renal disease. Although it has not been eliminated as a cause, urinary obstruction ("D") is not likely since she has good urine output. Answer "E" is incorrect since the urinalysis does not support diagnosis of infection, and CHF explains the overall clinical picture much better.
Remember that a BUN/Cr ratio >20 generally (but not always!) indicates a prerenal cause of azotemia. These causes include dehydration and poor renal perfusion (shock such as sepsis, HF, and hypotension). A BUN/Cr ratio <20 is suggestive of intrinsic renal disease or urinary outlet obstruction. These broad generalizations apply to adults only—use clinical judgment.
Question 5.1.9 Given her diabetes and renal disease, which of the following is the most likely cause of her hyperkalemia?
A) Renal tubular acidosis (RTA) type 1.
Answer 5.1.9 The correct answer is "D." RTA type 4 is due to aldosterone deficiency or resistance to the activity of aldosterone. The most common cause of RTA type 4 is hyporeninemic hypoaldosteronism, which is often seen in diabetic nephropathy. The disorder is recognized by hyperkalemia and mild acidosis. RTA types 1 and 2 usually are hypokalemic and these forms of RTA are not associated with diabetes. RTA type 3 is a rare autosomal-recessive disorder. RTA type 5 does not exist.
Question 5.1.10 In an effort to reduce her serum potassium level (not temporize), you should do which of the following?
A) Temporarily hold her furosemide and ACE inhibitor.
B) Increase her furosemide and temporarily hold her ACE inhibitor.
C) Administer IV calcium gluconate.
D) Bolus IV normal saline 1 to 2 L.
E) Give albuterol by nebulizer.
Answer 5.1.10 The correct answer is "B." Ah, test-taking—clearly not real life! The best answer here is to increase renal clearance of potassium with furosemide while holding the ACEI. Because ACE inhibitors can increase serum potassium, you should stop hers. "But what about her heart failure?" you ask. Well, good question. Given that her potassium is only slightly elevated, the ACEI could be continued while monitoring her serum potassium. "A" is incorrect; stopping the furosemide may lead to worsening HF symptoms and worsening hyperkalemia. In hyperkalemia, calcium gluconate ("C") is used to protect cardiac conduction, but it is not effective for reducing potassium concentrations. A bolus of normal saline ("D") is contraindicated in a patient with an acute HF exacerbation. Albuterol ("E") is a temporizing measure, reducing serum potassium by forcing potassium into the intracellular space.
In a patient with hyperkalemia and ECG changes (or a potassium of 7.5 mg/dL), it would be time for a full court press: calcium, insulin, glucose, sodium polystyrene sulfonate (Kayexalate), nebulized albuterol, etc. Note the absence of bicarbonate. It likely doesn't work and its use has fallen out of favor. Finally, don't use Kayexalate with sorbitol and be careful using even plain Kayexalate in individuals with GI dysfunction. Kayexalate has been associated with, and is likely the cause of, bowel perforation (J Am Soc Nephrol. 2010;21:733).
Question 5.1.11 Because her creatinine clearance is low (and she is in HF) you should also discontinue:
Answer 5.1.11 The correct answer is "C." Your patient's creatinine increased from 1.5 to 2.0 mg/dL, corresponding to a 33% reduction in creatinine clearance—a rough measurement of glomerular filtration rate (GFR). When creatinine clearance decreases, patients on metformin are at higher risk of developing the rare adverse effect of lactic acidosis. There is some evidence that patients with mild HF or slight elevations in creatinine can safely take the drug, but stopping metformin in renal failure continues to be the standard of care.
Metformin is often discontinued in diabetic patients with mild-to-moderate renal disease. As per the manufacturer, metformin is contraindicated in men with serum creatinine ≥1.5 mg/dL and in women with serum creatinine to ≥1.4 mg/dL. This is due to the risk of lactic acidosis. However, the risk of lactic acidosis is extremely small and does not even seem to be related directly to the degree of renal dysfunction. United Kingdom guidelines (2008) and American Diabetes Association guidelines (2011) recommend continuing metformin in patients with mild-to-moderate chronic kidney disease and monitoring renal function every 3 to 6 months.
A rough estimate of the creatinine clearance can be calculated using the Cockcroft–Gault formula:
Estimated creatinine clearance = (140 – age [year])(body weight [kg])/(72 × (serum creatinine [mg/dL]))
For women, multiply this figure by 0.85. One caveat is that this formula may not reflect early renal injury because of compensatory hypertrophy of the remaining glomeruli. Normal for healthy adult is 94 to 140 mL/min for men and 72 to 110 mL/min for women.
There is no perfect equation for calculating the GFR. The abbreviated modified diet in renal disease (MDRD) calculation is increasingly used because it seems to be an accurate representation of kidney function in adults with renal disease. Unlike the Cockcroft–Gault equation, the MDRD makes an adjustment for African-American race but does not take weight into account. Both equations adjust for gender. Although the equation is a little complicated, many labs now calculate the MDRD GFR. If you want to calculate it yourself, use the National Kidney Foundation website at http://www.kidney.org/professionals/KDOQI/gfr_calculator.
From the authors' perspective, any of these are OK. We don't need 2 mL/min accuracy in the CrCl. Just make sure you use one of them.
You administer intravenous furosemide. The night after admission she starts vomiting, and your partner inserts a nasogastric tube, which is inadvertently left to continuous suction overnight (don't do this … there is almost no indication for an NG tube—especially on continuous suction). The next morning, your patient's laboratory results are as follows: BUN 49 mg/dL, Cr 2.0 mg/dL, Na 132 mEq/L, K 3.5 mEq/L, and HCO3 35 mEq/L (normal 23–29 mEq/L).
Question 5.1.12 On the basis of history and laboratory data provided, you strongly suspect:
D) Respiratory alkalosis.
Answer 5.1.12 The correct answer is "C." Loss of gastric acid, through emesis or gastric suction, can result in a metabolic alkalosis. Although no pH is available to confirm alkalosis, you are able to infer the diagnosis based on the elevation in serum HCO3.
Diuretics will often cause a hypochloremic, "contraction," alkalosis from volume contraction. Watch for this in your patients on a diuretic.
Your patient recovers surprisingly well from her heart failure. Her creatinine returns to baseline (1.5 mg/dL). You try to re-challenge the patient with an ACE inhibitor. However, her hyperkalemia recurs and she is unable to tolerate either ACEI or ARB. With beta-blocker and loop diuretic therapy, her average blood pressure is 130/80 mm Hg. At a follow-up visit, you find no signs or symptoms of HF.
Question 5.1.13 You want to lower your patient's risk of progressing to end-stage renal disease. To reduce proteinuria, which of the following strategies is best?
A) Add isosorbide dinitrate.
B) Add a nondihydropyridine calcium channel blocker (e.g., diltiazem, verapamil).
C) Add a dihydropyridine calcium channel blocker (e.g., amlodipine, nifedipine).
D) Restrict protein in the diet.
E) Aim for a higher A1C target, around 8%.
Answer 5.1.13 The correct answer is "B." Nondihydropyridine calcium channel blockers reduce protein excretion in diabetic patients with nephropathy and slow down the progression of renal disease—but not to the degree of ACE inhibitors. "A," a long-acting nitrate, would be effective for angina and will lower blood pressure but does not demonstrate an effect on diabetic nephropathy. The dihydropyridine calcium channel blockers ("C") do not slow down progression to nephropathy to the same degree as some other antihypertensive drugs. The effect of protein restriction ("D") on nephropathy is controversial and not clearly beneficial; therefore, more effective therapies should be initiated first. Finally, "E" is incorrect since better glycemic control is associated with renal protection.
For diabetic nephropathy, always start with an ACE inhibitor and try to maximize the dose. Full dose ACE inhibitor treatment is associated with improved survival in diabetic nephropathy compared with low-dose ACE inhibitor. The dose of ACE inhibitor may be limited by serum potassium level or the patient's blood pressure. Sodium restriction added to an ACE inhibitor is actually more effective than adding an ARB for proteinuria and blood pressure control. (BMJ. 2011;343:d4366). In this study, the actual sodium in the diet was 2,500 mg (although the goal was 1,200 mg). Despite initial enthusiasm (or drug company marketing), an ARB plus an ACE is not useful in essentially any patient and results in increased side effects; this combination should be avoided.
Over the next year, your patient experiences increasing difficulties with glycemic control. Despite your efforts, her proteinuria and serum creatinine increase. As you discuss referral to a nephrologist, she asks about dialysis.
Question 5.1.14 Which of the following is NOT an indication for dialysis?
A) Severe hyperkalemia due to renal failure.
B) Accelerated hypertension.
D) Persistent nausea and vomiting despite treatment.
E) Bleeding secondary to uremia.
Answer 5.1.14 The correct answer is "C." In chronic renal failure due to any cause, absolute clinical indications for initiating dialysis include: persistent nausea and vomiting, pericarditis, fluid overload, uremic encephalopathy, accelerated hypertension, bleeding due to uremia, serum creatinine greater than 12 mg/dL, and severe electrolyte abnormalities that cannot be otherwise handled. These are potentially life-threatening situations that must be dealt with acutely. Malnutrition is a relative indication that occurs more indolently.
HELPFUL (AND VERY IMPORTANT) TIP:
Counter to what you might expect, early dialysis increases mortality (or at least is of no benefit). Do not initiate dialysis until the GFR is 5 to 7 mL/min/ 1.73 m2 or the patient is having clinical problems that cannot be otherwise managed (N Engl J Med. 2010;363(7): 609–619). Patients on early dialysis die from infection, complications from dialysis (hypotension), problems with their grafts (again leading to infection), etc.
Objectives: Did you learn to…
Screen for moderately increased albuminuria in diabetic patients?
Evaluate and treat moderately increased albuminuria and overt proteinuria in diabetic patients?
Identify prerenal acute kidney injury?
Describe the features of RTA type 4?
Identify gastric suctioning as a cause of metabolic alkalosis in the hospital?
Discuss when to initiate dialysis in a patient with chronic renal failure?