Mayo Clin Proc. 2001;76:1163-1166 1163 © 2001 Mayo Foundation for Medical Education and Research Residents' Clinic 40-Year-Old Man With Anemia, Chronic Low Back Pain, and Hypoalbuminemia CLAUDIA ORTIZ ZEIN, MD,* AND VANDANA NEHRA, MD† *Fellow in Gastroenterology, Mayo Graduate School of Medicine, Mayo Clinic, Rochester, Minn. †Adviser to fellow and Consultant in Gastroenterology and Hepatology, Mayo Clinic, Rochester, Minn. See end of article for correct answers to questions. Address reprint requests and correspondence to Vandana Nehra, MD, Division of Gastroenterology and Hepatology, Mayo Clinic, 200 First St SW, Rochester, MN 55905. A 40-year-old man was referred to our institution for evaluation of mild anemia found during routine blood testing. Aside from chronic low back pain, the pa- tient was completely asymptomatic. He had no gastrointes- tinal or systemic symptoms and specifically denied having abdominal pain, nausea, vomiting, changes in appetite, bloody stools, diarrhea, or weight loss. The patient’s medi- cal history was notable for paraplegia due to a traumatic spinal injury during a farm accident when he was 15 years old. Two years before presentation, he experienced pro- gressive low back pain. This symptom was evaluated at our institution and attributed to both lumbar facet arthritis and degenerative disk disease resulting from long-standing scoliosis. Because of the patient’s persistent low back pain, 1 year before presentation diclofenac sodium was initiated at a dose of 100 mg twice daily. He had no changes in bowel habit and had not changed his diet over the past year. Physical examination of the patient revealed right thora- columbar and left lumbar scoliosis, atrophy of the lower extremity muscles, and hyperreflexia and spasticity of the lower extremities (all related to his long-standing posttrau- matic neurologic condition). Findings on the rest of the examination were unremarkable. In particular, he had no peripheral edema or ascites. Laboratory tests yielded the following results (reference ranges shown parenthetically): hemoglobin, 10.7 g/dL (13.5-17.5 g/dL); leukocytes, 8.8 × 10 9 /L (3.5-10.5 × 10 9 / L); eosinophils, 0.17 × 10 9 /L; platelets, 409 × 10 9 /L (150- 450 × 10 9 /L); mean corpuscular volume, 82.8 fL (81.2-95.1 fL); iron, 17 µg/dL (50-150 µg/dL); total iron binding capacity, 250 µg/dL (250-400 µg/dL); transferrin satura- tion, 8% (14%-50%); ferritin, 14 µg/dL (20-300 µg/dL); and erythrocyte sedimentation rate, 14 mm/1 h (0-22 mm/1 h). Serum protein electrophoresis showed a total protein of 4.2 g/dL (6.3-7.9 g/dL), albumin of 2.0 g/dL (3.4-4.8 g/dL), and γ-globulin of 0.3 g/dL (0.7-1.7 g/dL). Results of other laboratory tests including liver enzymes, bilirubin, creati- nine, electrolytes, prothrombin time, serum lipids, and uri- nalysis were within the reference range. The patient’s iron deficiency anemia was presumed to be due to chronic blood loss secondary to possible nonsteroidal anti-inflammatory drug (NSAID)–induced gastropathy (al- though other causes of upper and lower gastrointestinal blood loss must be ruled out). However, his albumin level was found to be markedly and inexplicably low. 1. Which one of the following is the most likely diagnosis accounting for the low albumin level in this patient? a. Liver disease with decreased synthetic function b. Protein-losing enteropathy (PLE) c. Protein-energy malnutrition d. Hypoanabolic hypoalbuminemia e. Nephrotic syndrome The differential diagnosis of hypoalbuminemia includes liver disease with decreased synthetic function, PLE, mal- nutrition, hypoanabolic hypoalbuminemia, nephrotic syn- drome, etc. In patients with cirrhosis and ascites, hypoalbu- minemia commonly reflects diminished hepatic synthesis. Our patient had no clinical or laboratory evidence of liver disease. His liver enzymes and prothrombin time were within normal limits, making liver disease much less likely. Protein-losing enteropathy involves gastrointestinal loss of serum proteins in excessive amounts. Serum levels of pro- teins that normally have a low fractional catabolic rate (eg, albumin and immunoglobulins G, M, or A) are reduced substantially. Thus, PLE is a possible diagnosis in our patient. The most commonly used methods for determining pro- tein-energy malnutrition include a thorough history, a physical examination, and selected laboratory tests. The serum albumin concentration represents a net balance among synthesis, catabolism, losses, equilibrium between intravascular and extravascular compartments, and volume of distribution. Although albumin may be used as a marker of nutritional status, it is not the most sensitive marker of protein-energy malnutrition. Our patient had no clinical signs and symptoms of nutritional inadequacy associated with a poor protein intake, such as muscle wasting of the upper body (our patient had lower extremity muscle atro- phy but that was related to his long-standing posttraumatic For personal use. Mass reproduce only with permission from Mayo Clinic Proceedings.
40-Year-Old Man With Anemia, Chronic Low Back Pain, and Hypoalbuminemia
Feb 09, 2023
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
40-Year-Old Man With Anemia, Chronic Low Back Pain, and HypoalbuminemiaMayo Clin Proc, November 2001, Vol 76 Residents’ Clinic 1163
Mayo Clin Proc. 2001;76:1163-1166 1163 © 2001 Mayo Foundation for Medical Education and Research
Residents' Clinic
40-Year-Old Man With Anemia, Chronic Low Back Pain, and Hypoalbuminemia
CLAUDIA ORTIZ ZEIN, MD,* AND VANDANA NEHRA, MD†
*Fellow in Gastroenterology, Mayo Graduate School of Medicine, Mayo Clinic, Rochester, Minn.
†Adviser to fellow and Consultant in Gastroenterology and Hepatology, Mayo Clinic, Rochester, Minn.
See end of article for correct answers to questions.
Address reprint requests and correspondence to Vandana Nehra, MD, Division of Gastroenterology and Hepatology, Mayo Clinic, 200 First St SW, Rochester, MN 55905.
A 40-year-old man was referred to our institution for evaluation of mild anemia found during routine
blood testing. Aside from chronic low back pain, the pa- tient was completely asymptomatic. He had no gastrointes- tinal or systemic symptoms and specifically denied having abdominal pain, nausea, vomiting, changes in appetite, bloody stools, diarrhea, or weight loss. The patient’s medi- cal history was notable for paraplegia due to a traumatic spinal injury during a farm accident when he was 15 years old. Two years before presentation, he experienced pro- gressive low back pain. This symptom was evaluated at our institution and attributed to both lumbar facet arthritis and degenerative disk disease resulting from long-standing scoliosis. Because of the patient’s persistent low back pain, 1 year before presentation diclofenac sodium was initiated at a dose of 100 mg twice daily. He had no changes in bowel habit and had not changed his diet over the past year.
Physical examination of the patient revealed right thora- columbar and left lumbar scoliosis, atrophy of the lower extremity muscles, and hyperreflexia and spasticity of the lower extremities (all related to his long-standing posttrau- matic neurologic condition). Findings on the rest of the examination were unremarkable. In particular, he had no peripheral edema or ascites.
Laboratory tests yielded the following results (reference ranges shown parenthetically): hemoglobin, 10.7 g/dL (13.5-17.5 g/dL); leukocytes, 8.8 × 109/L (3.5-10.5 × 109/ L); eosinophils, 0.17 × 109/L; platelets, 409 × 109/L (150- 450 × 109/L); mean corpuscular volume, 82.8 fL (81.2-95.1 fL); iron, 17 µg/dL (50-150 µg/dL); total iron binding capacity, 250 µg/dL (250-400 µg/dL); transferrin satura- tion, 8% (14%-50%); ferritin, 14 µg/dL (20-300 µg/dL); and erythrocyte sedimentation rate, 14 mm/1 h (0-22 mm/1 h). Serum protein electrophoresis showed a total protein of 4.2 g/dL (6.3-7.9 g/dL), albumin of 2.0 g/dL (3.4-4.8 g/dL), and γ-globulin of 0.3 g/dL (0.7-1.7 g/dL). Results of other
laboratory tests including liver enzymes, bilirubin, creati- nine, electrolytes, prothrombin time, serum lipids, and uri- nalysis were within the reference range.
The patient’s iron deficiency anemia was presumed to be due to chronic blood loss secondary to possible nonsteroidal anti-inflammatory drug (NSAID)–induced gastropathy (al- though other causes of upper and lower gastrointestinal blood loss must be ruled out). However, his albumin level was found to be markedly and inexplicably low.
1. Which one of the following is the most likely diagnosis accounting for the low albumin level in this patient?
a. Liver disease with decreased synthetic function b. Protein-losing enteropathy (PLE) c. Protein-energy malnutrition d. Hypoanabolic hypoalbuminemia e. Nephrotic syndrome
The differential diagnosis of hypoalbuminemia includes liver disease with decreased synthetic function, PLE, mal- nutrition, hypoanabolic hypoalbuminemia, nephrotic syn- drome, etc. In patients with cirrhosis and ascites, hypoalbu- minemia commonly reflects diminished hepatic synthesis. Our patient had no clinical or laboratory evidence of liver disease. His liver enzymes and prothrombin time were within normal limits, making liver disease much less likely. Protein-losing enteropathy involves gastrointestinal loss of serum proteins in excessive amounts. Serum levels of pro- teins that normally have a low fractional catabolic rate (eg, albumin and immunoglobulins G, M, or A) are reduced substantially. Thus, PLE is a possible diagnosis in our patient.
The most commonly used methods for determining pro- tein-energy malnutrition include a thorough history, a physical examination, and selected laboratory tests. The serum albumin concentration represents a net balance among synthesis, catabolism, losses, equilibrium between intravascular and extravascular compartments, and volume of distribution. Although albumin may be used as a marker of nutritional status, it is not the most sensitive marker of protein-energy malnutrition. Our patient had no clinical signs and symptoms of nutritional inadequacy associated with a poor protein intake, such as muscle wasting of the upper body (our patient had lower extremity muscle atro- phy but that was related to his long-standing posttraumatic
For personal use. Mass reproduce only with permission from Mayo Clinic Proceedings.
Residents’ Clinic Mayo Clin Proc, November 2001, Vol 761164
paraplegia), sparse and thin hair, flaking dermatitis, or edema. In addition, his body mass index, a useful element in diagnosing protein-energy malnutrition, was calculated at 22 (normal nutrition is defined as a body mass index of 18 to 25). Hypoanabolic hypoalbuminemia is an inherited disorder present at birth and characterized by periodic pe- ripheral edema and serum albumin levels lower than 0.3 g/ dL. Both last-mentioned characteristics were absent in our patient. Hypoalbuminemia, edema, and hyperlipidemia characterize the nephrotic syndrome. Our patient had no edema, and his serum lipids were normal. In nephrotic syndrome, hypoalbuminemia is a consequence primarily of urinary protein loss. By definition, nephrotic range pro- teinuria entails albuminuria in amounts of more than 3 g/d.
A 24-hour urine collection demonstrated a total protein elimination of 39 mg/24 h (0-150 mg/24 h). In light of these findings, the possibility of PLE should be explored.
2. Which one of the following should be considered at this point to confirm the suspected diagnosis in our patient?
a. Fecal measurement of intravenously administered chro- mium Cr 51
b. Quantification of albumin in a 72-hour stool collection c. Random stool concentration of α
1 -antitrypsin
d. 24-Hour stool collection for α 1 -antitrypsin and calcula-
tion of clearance from plasma e. Esophagogastroduodenoscopy (EGD) and colonoscopy
In the past, the diagnosis of PLE was based on the measurement of fecal losses of intravenously administered radiolabeled macromolecules, such as chromium Cr 51. Although accurate, these tests involve radioactive expo- sure, are expensive and inconvenient, and thus are not ordinarily used. Albumin is digested extensively by intesti- nal proteases. Therefore, quantification of albumin in stool is not useful. α
1 -Antitrypsin is a protein with molecular
weight similar to that of albumin. Because of its anti- proteolytic activity, α
1 -antitrypsin is digested minimally by
intestinal proteases and is mainly excreted intact in stool. This can be used as an indirect measurement of albumin loss into the gastrointestinal tract. Random stool concentra- tion of α
1 -antitrypsin has been shown to have poor correla-
tion with plasma clearance measurements. The diagnostic test of choice for PLE is a 24-hour stool collection to measure the clearance of plasma α
1 -antitrypsin in stool.
This is done by multiplying the 24-hour stool volume times the α
1 -antitrypsin concentration in the stool and dividing
this product by the α 1 -antitrypsin concentration in the se-
rum (stool volume/24 h) × (stool concentration α 1 -anti-
trypsin)/(serum concentration α 1 -antitrypsin). Colonoscopy
and EGD are useful tests in the work-up of a patient with a diagnosis of PLE; however, they are not useful per se in diagnosing PLE.
A 24-hour stool collection for α 1 -antitrypsin was per-
formed. α 1 -Antitrypsin clearance was calculated and found
to be markedly elevated at 607 mL/24 h (≤27 mL/24 h). Quantification of fat in a 72-hour stool specimen was only minimally above normal at 8.9 g/24 h (2-7 g/24 h). The results of these stool studies are consistent with PLE. The slightly elevated results of stool fat are unimportant in the setting of such a high α
1 -antitrypsin clearance. Antigliadin
antibody and endomysial antibodies were negative. Caro- tene levels were 76 µg/dL (48-200 µg/dL).
3. At this stage, which one of the following investigations would be the least helpful in establishing the cause of PLE in this patient?
a. Small bowel contrast study b. EGD and colonoscopy with biopsies c. Small bowel biopsy d. Pedal lymphangiography e. Abdominal computed tomography
Diverse disorders are associated with PLE. A small bowel contrast study is occasionally useful in identifying radiologic abnormalities of the intestinal tract associated with several disorders that could cause PLE. However, these are typically nonspecific findings, and further tests are almost always necessary to identify the specific cause of PLE in an individual patient. An EGD with biopsies would prove useful in diagnosing mucosal ulcerative (eg, erosive gastroenteritis, Crohn disease) and nonulcerative (Ménétrier disease, eosinophilic gastroenteritis, celiac sprue, Whipple disease, parasitic diseases, bacterial over- growth) causes of PLE. Colonoscopy with biopsies would also be useful in diagnosing similar causes of PLE affecting the colon or terminal ileum (eg, pseudomembranous entero- colitis, Crohn disease, microscopic or collagenous colitis, neoplasm). Small bowel biopsies are important in diagnos- ing many of the potential PLE etiologies mentioned previ- ously. Several biopsies may be necessary to establish the exact diagnosis because many disorders associated with PLE are patchy in nature. Pedal lymphangiography may show abnormalities consistent with congenital lymphangiectasia; however, normal findings on a small bowel biopsy exclude this diagnosis. It is unlikely that a pedal lymphangiogram would yield any additional information in this patient. Ab- dominal computed tomography is helpful to rule out lym- phoma and retroperitoneal processes that could cause PLE.
Findings on upper gastrointestinal x-ray film and small- bowel follow through were negative. Small erosions in the body and antrum of the stomach were evident on EGD. Biopsy specimens were consistent with reactive gas- tropathy suggestive of NSAID-type injury. No evidence of Helicobacter pylori infection was found on biopsy. The proximal small bowel appeared normal on endoscopy. A
For personal use. Mass reproduce only with permission from Mayo Clinic Proceedings.
Mayo Clin Proc, November 2001, Vol 76 Residents’ Clinic 1165
postbulbar duodenal biopsy showed normal mucosa with no diagnostic abnormality. Specifically, no Whipple dis- ease, celiac sprue, or Giardia was evident. Stains for amy- loid (Congo red, sulfated alcian blue) were negative. No lymphangiectasia changes were evident on biopsy. Colo- noscopy, performed as part of the work-up and to rule out any malignant lesions that could be associated with iron deficiency anemia, showed a shallow ulceration at the ileo- cecal valve, consistent with an NSAID-type injury. Biop- sies of the area showed nonspecific focal inflammation. The terminal ileum appeared normal, the colonic mucosa was unremarkable, and findings on random biopsies were normal. Lymphoscintigraphy showed no evidence of lym- phatic obstruction from the lower limbs into the blood pool. Computed tomography of the abdomen and pelvis re- vealed mild mesenteric adenopathy without retroperitoneal adenopathy. To obtain further diagnostic information, a technetium Tc 99m hexamethylpropyleneamine oxime (HMPAO) white blood cell scan was performed to deter- mine whether inflammation was present in the bowel.
4. In the context of all the evaluations performed to date and based on the results of the HMPAO white blood cell scan, which one of the following is the most likely etiology of this patient’s condition?
a. Long-term use of diclofenac b. Crohn disease of the small bowel c. Chronic ulcerative colitis d. Microscopic and collagenous colitis e. Eosinophilic gastroenteritis
The HMPAO white blood cell scan showed increased activity in the small bowel (left upper quadrant) and colon (ascending, transverse, and descending) at 30 minutes and more pronounced findings after 2 to 3 hours. These find- ings suggest an inflammatory process involving the small bowel (likely proximal) and the colon. None of the classic causes of severe PLE were found. Our patient’s history of long-term use of NSAID therapy, along with the HMPAO white blood cell scan results and the findings of NSAID- type injury on EGD and colonoscopy, points to a diagnosis of NSAID-induced PLE. Although HMPAO leukocyte im- ages have been proved useful and accurate in the assess- ment of patients with active Crohn disease of the small bowel,1,2 our patient had no historical, laboratory, or endo- scopic findings typically associated with this condition. Crohn disease is an unlikely diagnosis in light of the ab- sence of endoscopic or histological findings consistent with Crohn disease on colonoscopy including terminal ileal endoscopy and EGD. The absence of gastrointestinal symptoms and the normal colonic mucosa on endoscopy and biopsies refute a diagnosis of chronic ulcerative co- litis. Both microscopic and collagenous colitis cause co-
lonic inflammation, and an epidemiological relationship to long-term use of NSAIDs has been reported; however, by definition, inflammation is seen on colonic biopsies. Eo- sinophilic gastroenteritis may also present with a clinical picture of PLE. However, histology shows infiltration of various layers of the gastrointestinal tract with eosino- phils, and peripheral eosinophilia is present in 75% of such patients.
5. Which one of the following options is most appropriate for our patient at this point?
a. Discontinuation of diclofenac b. Substitution of diclofenac with rofecoxib c. Low-fat diet d. Use of medium-chain triglycerides e. Observation since he is asymptomatic
More than one third of patients receiving long-term NSAID therapy develop intestinal inflammation, which may lead to bleeding and protein loss. Recognition of the cause of the enteropathy and discontinuation of the offend- ing agent (diclofenac in this case) are the mainstays of therapy. Short-term studies assessing intestinal permeabil- ity show that rofecoxib does not increase intestinal perme- ability. However, long-term studies are needed before we could confidently initiate this medication in our patient. In patients with congenital intestinal lymphangiectasia, en- teric protein loss is generally diminished with a low-fat diet. Medium-chain triglycerides, which do not require intestinal lymphatic transport and therefore do not stimu- late intestinal lymph flow, can also be used in such patients to reduce enteric protein loss. The appropriate medical intervention should be adapted in patients with severe PLE. Observation only is inappropriate.
Diclofenac was discontinued, and tramadol was initi- ated for the patient’s low back pain. Determination of an albumin level 7 weeks after discontinuation of NSAID therapy showed a substantial increase from 2.0 to 3.2 g/dL. Ten weeks after discontinuation of NSAID therapy, a re- peated stool collection for α
1 -antitrypsin clearance showed
a dramatic improvement (from 607 to 176 mL/24 h).
DISCUSSION Our case illustrates the occurrence of severe PLE second- ary to long-term NSAID use. Although most cases of intes- tinal protein loss secondary to NSAID are mild, severe loss can occur. This outcome in elderly patients or in those with chronic rheumatologic disorders can lead to clinically im- portant hypoalbuminemia.
Long-term use of NSAIDs is associated with a high risk of gastrointestinal injury, and the effects of NSAIDs on the gastroduodenal mucosa remain the primary clinical prob- lem. Studies have shown that 10% to 30% of patients
For personal use. Mass reproduce only with permission from Mayo Clinic Proceedings.
Residents’ Clinic Mayo Clin Proc, November 2001, Vol 761166
REFERENCES 1. Martin-Comin J, Prats E. Clinical applications of radiolabeled
blood elements in inflammatory bowel disease. Q J Nucl Med. 1999;43:74-82.
2. Jaakkola K, Knuuti J, Soderlund K, Saraste A, Jalkanen S, Voipio- Pulkki LM. Labelling lymphocytes with technetium99m- hexamethyl propyleneamine oxime for scintigraphy: an improved labelling procedure. J Immunol Methods. 1998;214:187-197.
3. Lanza FL, Graham DY, Davis RE, Rack MF. Endoscopic compari- son of cimetidine and sucralfate for prevention of naproxen-in- duced acute gastroduodenal injury: effect of scoring method. Dig Dis Sci. 1990;35:1494-1499.
4. Langman MJ. Epidemiologic evidence on the association between peptic ulceration and antiinflammatory drug use. Gastroenterol- ogy. 1989;96(2, pt 2, suppl):640-646.
5. Bjarnason I, Zanelli G, Prouse P, et al. Blood and protein loss via small-intestinal inflammation induced by non-steroidal anti-in- flammatory drugs. Lancet. 1987;2:711-714.
6. Tibble JA, Sigthorsson G, Foster R, et al. High prevalence of NSAID enteropathy as shown by a simple faecal test. Gut. 1999;45:362-366.
7. Somasundaram S, Sigthorsson G, Simpson RJ, et al. Uncoupling of intestinal mitochondrial oxidative phosphorylation and inhibition of cyclooxygenase are required for the development of NSAID- enteropathy in the rat. Aliment Pharmacol Ther. 2000;14:639-650.
8. Reuter BK, Davies NM, Wallace JL. Nonsteroidal anti-inflamma- tory drug enteropathy in rats: role of permeability, bacteria, and enterohepatic circulation. Gastroenterology. 1997;112:109-117.
9. Sigthorsson G, Tibble J, Hayllar J, et al. Intestinal permeability and inflammation in patients on NSAIDs. Gut. 1998;43:506-511.
10. Tibble JA, Sigthorsson G, Foster R, Bjarnason I. Comparison of the intestinal toxicity of celecoxib, a selective COX-2 inhibitor, and indomethacin in the experimental rat. Scand J Gastroenterol. 2000; 35:802-807.
11. Davies NM, Saleh JY, Skjodt NM. Detection and prevention of NSAID-induced enteropathy. J Pharm Pharm Sci. 2000;3:137-155.
12. Bjarnason I. Forthcoming non-steroidal anti-inflammatory drugs: are they really devoid of side effects? Ital J Gastroenterol Hepatol. 1999;31(suppl 1):S27-S36.
13. Hayllar J, Smith T, Macpherson A, Price AB, Gumpel M, Bjarnason I. Nonsteroidal antiinflammatory drug-induced small intestinal inflammation and blood loss: effects of sulfasalazine and other disease-modifying antirheumatic drugs. Arthritis Rheum. 1994;37:1146-1150.
14. Leite AZA, Sipahi AM, Damião AOMC, et al. Protective effect of metronidazole on uncoupling mitochondrial oxidative phosphoryla- tion induced by NSAID: a new mechanism. Gut. 2001;48:163-167.
Correct answers: 1. b, 2. d, 3. d, 4. a, 5. a
taking NSAIDs have gastroduodenal ulceration and that 20% to 40% have milder lesions, such as petechiae or erosions.3,4 Notably, 40% to 70% of patients taking long- term NSAIDs develop enteropathy.5,6 Enteropathy consists of increased intestinal permeability and inflammatory changes with low-grade bleeding and protein loss. Despite the high prevalence of enteropathy in patients receiving long-term NSAID therapy, most have mild bleeding and protein loss and are asymptomatic. Other patients present with bile acid malabsorption, small intestinal ulcers, and perforations or strictures.
The pathogenesis of NSAID-induced intestinal damage is believed to involve non–prostaglandin-dependent effects and prostaglandin-dependent effects. A theory is that un- coupling of enterocyte mitochondrial oxidative phosphory- lation leads to increased intestinal permeability and low- grade inflammation. Concurrently, suppression of prosta- glandin synthesis seems to be important in the development of intestinal ulcerations.7 Enterohepatic recirculation of drug has been shown to be of crucial importance in the pathogenesis of enteropathy caused by NSAIDs.8 NSAIDs that do not undergo enterohepatic recirculation (aspirin and nabumetone) do not cause intestinal inflammation.8,9 The role of luminal aggressive factors (eg, bile, pancreatic en- zymes, bacteria) also seems important in the pathogenesis of NSAID-induced enteropathy.
The diagnosis of NSAID-induced enteropathy is com- plex. Radiolabeled laboratory tests that use indium In– labeled leukocytes, chromium Cr 51–labeled red blood cells, or ethylenediaminetetraacetic acid are expensive and cumbersome. Our diagnostic approach was to use an HMPAO white blood cell scan to detect the presence of intestinal inflammation. This test has been evaluated exten- sively and has been shown to be accurate in the assessment of patients with intestinal inflammation caused by active Crohn disease of the small bowel. However, its perfor- mance has not been formally assessed in diagnosing NSAID-induced enteropathy. Recently, an assay detecting fecal calprotectin was shown to correlate with results of a 4-day fecal excretion of indium In–labeled white blood cells in diagnosing NSAID-induced enteropathy.6 Small bowel enteroscopy has been used to investigate blood loss in patients taking NSAIDs. However, the procedure is lengthy and often nondiagnostic.
The recommended therapy for NSAID-induced enter- opathy is still undefined. After the cause of the enteropathy has been determined, discontinuation of use of the offend- ing agent is the ideal strategy. If this strategy is not pos- sible, the NSAID dosage should be decreased. Cyclooxy- genase-2 selective inhibitors (celecoxib, rofecoxib) have fewer toxic effects on the small intestine vs traditional nonselective NSAIDs,10-12 therefore becoming a more suit-
able choice for such patients. Sulfasalazine has been shown to reduce intestinal inflammation and blood loss in patients taking NSAIDs.13 Synthetic prostaglandins are known to limit NSAID-induced gastroduodenal damage and may also help prevent development of NSAID-induced intesti- nal ulcerations. Proton pump inhibitors have been shown to allow healing of NSAID-induced ulcers in patients receiv-…
Mayo Clin Proc. 2001;76:1163-1166 1163 © 2001 Mayo Foundation for Medical Education and Research
Residents' Clinic
40-Year-Old Man With Anemia, Chronic Low Back Pain, and Hypoalbuminemia
CLAUDIA ORTIZ ZEIN, MD,* AND VANDANA NEHRA, MD†
*Fellow in Gastroenterology, Mayo Graduate School of Medicine, Mayo Clinic, Rochester, Minn.
†Adviser to fellow and Consultant in Gastroenterology and Hepatology, Mayo Clinic, Rochester, Minn.
See end of article for correct answers to questions.
Address reprint requests and correspondence to Vandana Nehra, MD, Division of Gastroenterology and Hepatology, Mayo Clinic, 200 First St SW, Rochester, MN 55905.
A 40-year-old man was referred to our institution for evaluation of mild anemia found during routine
blood testing. Aside from chronic low back pain, the pa- tient was completely asymptomatic. He had no gastrointes- tinal or systemic symptoms and specifically denied having abdominal pain, nausea, vomiting, changes in appetite, bloody stools, diarrhea, or weight loss. The patient’s medi- cal history was notable for paraplegia due to a traumatic spinal injury during a farm accident when he was 15 years old. Two years before presentation, he experienced pro- gressive low back pain. This symptom was evaluated at our institution and attributed to both lumbar facet arthritis and degenerative disk disease resulting from long-standing scoliosis. Because of the patient’s persistent low back pain, 1 year before presentation diclofenac sodium was initiated at a dose of 100 mg twice daily. He had no changes in bowel habit and had not changed his diet over the past year.
Physical examination of the patient revealed right thora- columbar and left lumbar scoliosis, atrophy of the lower extremity muscles, and hyperreflexia and spasticity of the lower extremities (all related to his long-standing posttrau- matic neurologic condition). Findings on the rest of the examination were unremarkable. In particular, he had no peripheral edema or ascites.
Laboratory tests yielded the following results (reference ranges shown parenthetically): hemoglobin, 10.7 g/dL (13.5-17.5 g/dL); leukocytes, 8.8 × 109/L (3.5-10.5 × 109/ L); eosinophils, 0.17 × 109/L; platelets, 409 × 109/L (150- 450 × 109/L); mean corpuscular volume, 82.8 fL (81.2-95.1 fL); iron, 17 µg/dL (50-150 µg/dL); total iron binding capacity, 250 µg/dL (250-400 µg/dL); transferrin satura- tion, 8% (14%-50%); ferritin, 14 µg/dL (20-300 µg/dL); and erythrocyte sedimentation rate, 14 mm/1 h (0-22 mm/1 h). Serum protein electrophoresis showed a total protein of 4.2 g/dL (6.3-7.9 g/dL), albumin of 2.0 g/dL (3.4-4.8 g/dL), and γ-globulin of 0.3 g/dL (0.7-1.7 g/dL). Results of other
laboratory tests including liver enzymes, bilirubin, creati- nine, electrolytes, prothrombin time, serum lipids, and uri- nalysis were within the reference range.
The patient’s iron deficiency anemia was presumed to be due to chronic blood loss secondary to possible nonsteroidal anti-inflammatory drug (NSAID)–induced gastropathy (al- though other causes of upper and lower gastrointestinal blood loss must be ruled out). However, his albumin level was found to be markedly and inexplicably low.
1. Which one of the following is the most likely diagnosis accounting for the low albumin level in this patient?
a. Liver disease with decreased synthetic function b. Protein-losing enteropathy (PLE) c. Protein-energy malnutrition d. Hypoanabolic hypoalbuminemia e. Nephrotic syndrome
The differential diagnosis of hypoalbuminemia includes liver disease with decreased synthetic function, PLE, mal- nutrition, hypoanabolic hypoalbuminemia, nephrotic syn- drome, etc. In patients with cirrhosis and ascites, hypoalbu- minemia commonly reflects diminished hepatic synthesis. Our patient had no clinical or laboratory evidence of liver disease. His liver enzymes and prothrombin time were within normal limits, making liver disease much less likely. Protein-losing enteropathy involves gastrointestinal loss of serum proteins in excessive amounts. Serum levels of pro- teins that normally have a low fractional catabolic rate (eg, albumin and immunoglobulins G, M, or A) are reduced substantially. Thus, PLE is a possible diagnosis in our patient.
The most commonly used methods for determining pro- tein-energy malnutrition include a thorough history, a physical examination, and selected laboratory tests. The serum albumin concentration represents a net balance among synthesis, catabolism, losses, equilibrium between intravascular and extravascular compartments, and volume of distribution. Although albumin may be used as a marker of nutritional status, it is not the most sensitive marker of protein-energy malnutrition. Our patient had no clinical signs and symptoms of nutritional inadequacy associated with a poor protein intake, such as muscle wasting of the upper body (our patient had lower extremity muscle atro- phy but that was related to his long-standing posttraumatic
For personal use. Mass reproduce only with permission from Mayo Clinic Proceedings.
Residents’ Clinic Mayo Clin Proc, November 2001, Vol 761164
paraplegia), sparse and thin hair, flaking dermatitis, or edema. In addition, his body mass index, a useful element in diagnosing protein-energy malnutrition, was calculated at 22 (normal nutrition is defined as a body mass index of 18 to 25). Hypoanabolic hypoalbuminemia is an inherited disorder present at birth and characterized by periodic pe- ripheral edema and serum albumin levels lower than 0.3 g/ dL. Both last-mentioned characteristics were absent in our patient. Hypoalbuminemia, edema, and hyperlipidemia characterize the nephrotic syndrome. Our patient had no edema, and his serum lipids were normal. In nephrotic syndrome, hypoalbuminemia is a consequence primarily of urinary protein loss. By definition, nephrotic range pro- teinuria entails albuminuria in amounts of more than 3 g/d.
A 24-hour urine collection demonstrated a total protein elimination of 39 mg/24 h (0-150 mg/24 h). In light of these findings, the possibility of PLE should be explored.
2. Which one of the following should be considered at this point to confirm the suspected diagnosis in our patient?
a. Fecal measurement of intravenously administered chro- mium Cr 51
b. Quantification of albumin in a 72-hour stool collection c. Random stool concentration of α
1 -antitrypsin
d. 24-Hour stool collection for α 1 -antitrypsin and calcula-
tion of clearance from plasma e. Esophagogastroduodenoscopy (EGD) and colonoscopy
In the past, the diagnosis of PLE was based on the measurement of fecal losses of intravenously administered radiolabeled macromolecules, such as chromium Cr 51. Although accurate, these tests involve radioactive expo- sure, are expensive and inconvenient, and thus are not ordinarily used. Albumin is digested extensively by intesti- nal proteases. Therefore, quantification of albumin in stool is not useful. α
1 -Antitrypsin is a protein with molecular
weight similar to that of albumin. Because of its anti- proteolytic activity, α
1 -antitrypsin is digested minimally by
intestinal proteases and is mainly excreted intact in stool. This can be used as an indirect measurement of albumin loss into the gastrointestinal tract. Random stool concentra- tion of α
1 -antitrypsin has been shown to have poor correla-
tion with plasma clearance measurements. The diagnostic test of choice for PLE is a 24-hour stool collection to measure the clearance of plasma α
1 -antitrypsin in stool.
This is done by multiplying the 24-hour stool volume times the α
1 -antitrypsin concentration in the stool and dividing
this product by the α 1 -antitrypsin concentration in the se-
rum (stool volume/24 h) × (stool concentration α 1 -anti-
trypsin)/(serum concentration α 1 -antitrypsin). Colonoscopy
and EGD are useful tests in the work-up of a patient with a diagnosis of PLE; however, they are not useful per se in diagnosing PLE.
A 24-hour stool collection for α 1 -antitrypsin was per-
formed. α 1 -Antitrypsin clearance was calculated and found
to be markedly elevated at 607 mL/24 h (≤27 mL/24 h). Quantification of fat in a 72-hour stool specimen was only minimally above normal at 8.9 g/24 h (2-7 g/24 h). The results of these stool studies are consistent with PLE. The slightly elevated results of stool fat are unimportant in the setting of such a high α
1 -antitrypsin clearance. Antigliadin
antibody and endomysial antibodies were negative. Caro- tene levels were 76 µg/dL (48-200 µg/dL).
3. At this stage, which one of the following investigations would be the least helpful in establishing the cause of PLE in this patient?
a. Small bowel contrast study b. EGD and colonoscopy with biopsies c. Small bowel biopsy d. Pedal lymphangiography e. Abdominal computed tomography
Diverse disorders are associated with PLE. A small bowel contrast study is occasionally useful in identifying radiologic abnormalities of the intestinal tract associated with several disorders that could cause PLE. However, these are typically nonspecific findings, and further tests are almost always necessary to identify the specific cause of PLE in an individual patient. An EGD with biopsies would prove useful in diagnosing mucosal ulcerative (eg, erosive gastroenteritis, Crohn disease) and nonulcerative (Ménétrier disease, eosinophilic gastroenteritis, celiac sprue, Whipple disease, parasitic diseases, bacterial over- growth) causes of PLE. Colonoscopy with biopsies would also be useful in diagnosing similar causes of PLE affecting the colon or terminal ileum (eg, pseudomembranous entero- colitis, Crohn disease, microscopic or collagenous colitis, neoplasm). Small bowel biopsies are important in diagnos- ing many of the potential PLE etiologies mentioned previ- ously. Several biopsies may be necessary to establish the exact diagnosis because many disorders associated with PLE are patchy in nature. Pedal lymphangiography may show abnormalities consistent with congenital lymphangiectasia; however, normal findings on a small bowel biopsy exclude this diagnosis. It is unlikely that a pedal lymphangiogram would yield any additional information in this patient. Ab- dominal computed tomography is helpful to rule out lym- phoma and retroperitoneal processes that could cause PLE.
Findings on upper gastrointestinal x-ray film and small- bowel follow through were negative. Small erosions in the body and antrum of the stomach were evident on EGD. Biopsy specimens were consistent with reactive gas- tropathy suggestive of NSAID-type injury. No evidence of Helicobacter pylori infection was found on biopsy. The proximal small bowel appeared normal on endoscopy. A
For personal use. Mass reproduce only with permission from Mayo Clinic Proceedings.
Mayo Clin Proc, November 2001, Vol 76 Residents’ Clinic 1165
postbulbar duodenal biopsy showed normal mucosa with no diagnostic abnormality. Specifically, no Whipple dis- ease, celiac sprue, or Giardia was evident. Stains for amy- loid (Congo red, sulfated alcian blue) were negative. No lymphangiectasia changes were evident on biopsy. Colo- noscopy, performed as part of the work-up and to rule out any malignant lesions that could be associated with iron deficiency anemia, showed a shallow ulceration at the ileo- cecal valve, consistent with an NSAID-type injury. Biop- sies of the area showed nonspecific focal inflammation. The terminal ileum appeared normal, the colonic mucosa was unremarkable, and findings on random biopsies were normal. Lymphoscintigraphy showed no evidence of lym- phatic obstruction from the lower limbs into the blood pool. Computed tomography of the abdomen and pelvis re- vealed mild mesenteric adenopathy without retroperitoneal adenopathy. To obtain further diagnostic information, a technetium Tc 99m hexamethylpropyleneamine oxime (HMPAO) white blood cell scan was performed to deter- mine whether inflammation was present in the bowel.
4. In the context of all the evaluations performed to date and based on the results of the HMPAO white blood cell scan, which one of the following is the most likely etiology of this patient’s condition?
a. Long-term use of diclofenac b. Crohn disease of the small bowel c. Chronic ulcerative colitis d. Microscopic and collagenous colitis e. Eosinophilic gastroenteritis
The HMPAO white blood cell scan showed increased activity in the small bowel (left upper quadrant) and colon (ascending, transverse, and descending) at 30 minutes and more pronounced findings after 2 to 3 hours. These find- ings suggest an inflammatory process involving the small bowel (likely proximal) and the colon. None of the classic causes of severe PLE were found. Our patient’s history of long-term use of NSAID therapy, along with the HMPAO white blood cell scan results and the findings of NSAID- type injury on EGD and colonoscopy, points to a diagnosis of NSAID-induced PLE. Although HMPAO leukocyte im- ages have been proved useful and accurate in the assess- ment of patients with active Crohn disease of the small bowel,1,2 our patient had no historical, laboratory, or endo- scopic findings typically associated with this condition. Crohn disease is an unlikely diagnosis in light of the ab- sence of endoscopic or histological findings consistent with Crohn disease on colonoscopy including terminal ileal endoscopy and EGD. The absence of gastrointestinal symptoms and the normal colonic mucosa on endoscopy and biopsies refute a diagnosis of chronic ulcerative co- litis. Both microscopic and collagenous colitis cause co-
lonic inflammation, and an epidemiological relationship to long-term use of NSAIDs has been reported; however, by definition, inflammation is seen on colonic biopsies. Eo- sinophilic gastroenteritis may also present with a clinical picture of PLE. However, histology shows infiltration of various layers of the gastrointestinal tract with eosino- phils, and peripheral eosinophilia is present in 75% of such patients.
5. Which one of the following options is most appropriate for our patient at this point?
a. Discontinuation of diclofenac b. Substitution of diclofenac with rofecoxib c. Low-fat diet d. Use of medium-chain triglycerides e. Observation since he is asymptomatic
More than one third of patients receiving long-term NSAID therapy develop intestinal inflammation, which may lead to bleeding and protein loss. Recognition of the cause of the enteropathy and discontinuation of the offend- ing agent (diclofenac in this case) are the mainstays of therapy. Short-term studies assessing intestinal permeabil- ity show that rofecoxib does not increase intestinal perme- ability. However, long-term studies are needed before we could confidently initiate this medication in our patient. In patients with congenital intestinal lymphangiectasia, en- teric protein loss is generally diminished with a low-fat diet. Medium-chain triglycerides, which do not require intestinal lymphatic transport and therefore do not stimu- late intestinal lymph flow, can also be used in such patients to reduce enteric protein loss. The appropriate medical intervention should be adapted in patients with severe PLE. Observation only is inappropriate.
Diclofenac was discontinued, and tramadol was initi- ated for the patient’s low back pain. Determination of an albumin level 7 weeks after discontinuation of NSAID therapy showed a substantial increase from 2.0 to 3.2 g/dL. Ten weeks after discontinuation of NSAID therapy, a re- peated stool collection for α
1 -antitrypsin clearance showed
a dramatic improvement (from 607 to 176 mL/24 h).
DISCUSSION Our case illustrates the occurrence of severe PLE second- ary to long-term NSAID use. Although most cases of intes- tinal protein loss secondary to NSAID are mild, severe loss can occur. This outcome in elderly patients or in those with chronic rheumatologic disorders can lead to clinically im- portant hypoalbuminemia.
Long-term use of NSAIDs is associated with a high risk of gastrointestinal injury, and the effects of NSAIDs on the gastroduodenal mucosa remain the primary clinical prob- lem. Studies have shown that 10% to 30% of patients
For personal use. Mass reproduce only with permission from Mayo Clinic Proceedings.
Residents’ Clinic Mayo Clin Proc, November 2001, Vol 761166
REFERENCES 1. Martin-Comin J, Prats E. Clinical applications of radiolabeled
blood elements in inflammatory bowel disease. Q J Nucl Med. 1999;43:74-82.
2. Jaakkola K, Knuuti J, Soderlund K, Saraste A, Jalkanen S, Voipio- Pulkki LM. Labelling lymphocytes with technetium99m- hexamethyl propyleneamine oxime for scintigraphy: an improved labelling procedure. J Immunol Methods. 1998;214:187-197.
3. Lanza FL, Graham DY, Davis RE, Rack MF. Endoscopic compari- son of cimetidine and sucralfate for prevention of naproxen-in- duced acute gastroduodenal injury: effect of scoring method. Dig Dis Sci. 1990;35:1494-1499.
4. Langman MJ. Epidemiologic evidence on the association between peptic ulceration and antiinflammatory drug use. Gastroenterol- ogy. 1989;96(2, pt 2, suppl):640-646.
5. Bjarnason I, Zanelli G, Prouse P, et al. Blood and protein loss via small-intestinal inflammation induced by non-steroidal anti-in- flammatory drugs. Lancet. 1987;2:711-714.
6. Tibble JA, Sigthorsson G, Foster R, et al. High prevalence of NSAID enteropathy as shown by a simple faecal test. Gut. 1999;45:362-366.
7. Somasundaram S, Sigthorsson G, Simpson RJ, et al. Uncoupling of intestinal mitochondrial oxidative phosphorylation and inhibition of cyclooxygenase are required for the development of NSAID- enteropathy in the rat. Aliment Pharmacol Ther. 2000;14:639-650.
8. Reuter BK, Davies NM, Wallace JL. Nonsteroidal anti-inflamma- tory drug enteropathy in rats: role of permeability, bacteria, and enterohepatic circulation. Gastroenterology. 1997;112:109-117.
9. Sigthorsson G, Tibble J, Hayllar J, et al. Intestinal permeability and inflammation in patients on NSAIDs. Gut. 1998;43:506-511.
10. Tibble JA, Sigthorsson G, Foster R, Bjarnason I. Comparison of the intestinal toxicity of celecoxib, a selective COX-2 inhibitor, and indomethacin in the experimental rat. Scand J Gastroenterol. 2000; 35:802-807.
11. Davies NM, Saleh JY, Skjodt NM. Detection and prevention of NSAID-induced enteropathy. J Pharm Pharm Sci. 2000;3:137-155.
12. Bjarnason I. Forthcoming non-steroidal anti-inflammatory drugs: are they really devoid of side effects? Ital J Gastroenterol Hepatol. 1999;31(suppl 1):S27-S36.
13. Hayllar J, Smith T, Macpherson A, Price AB, Gumpel M, Bjarnason I. Nonsteroidal antiinflammatory drug-induced small intestinal inflammation and blood loss: effects of sulfasalazine and other disease-modifying antirheumatic drugs. Arthritis Rheum. 1994;37:1146-1150.
14. Leite AZA, Sipahi AM, Damião AOMC, et al. Protective effect of metronidazole on uncoupling mitochondrial oxidative phosphoryla- tion induced by NSAID: a new mechanism. Gut. 2001;48:163-167.
Correct answers: 1. b, 2. d, 3. d, 4. a, 5. a
taking NSAIDs have gastroduodenal ulceration and that 20% to 40% have milder lesions, such as petechiae or erosions.3,4 Notably, 40% to 70% of patients taking long- term NSAIDs develop enteropathy.5,6 Enteropathy consists of increased intestinal permeability and inflammatory changes with low-grade bleeding and protein loss. Despite the high prevalence of enteropathy in patients receiving long-term NSAID therapy, most have mild bleeding and protein loss and are asymptomatic. Other patients present with bile acid malabsorption, small intestinal ulcers, and perforations or strictures.
The pathogenesis of NSAID-induced intestinal damage is believed to involve non–prostaglandin-dependent effects and prostaglandin-dependent effects. A theory is that un- coupling of enterocyte mitochondrial oxidative phosphory- lation leads to increased intestinal permeability and low- grade inflammation. Concurrently, suppression of prosta- glandin synthesis seems to be important in the development of intestinal ulcerations.7 Enterohepatic recirculation of drug has been shown to be of crucial importance in the pathogenesis of enteropathy caused by NSAIDs.8 NSAIDs that do not undergo enterohepatic recirculation (aspirin and nabumetone) do not cause intestinal inflammation.8,9 The role of luminal aggressive factors (eg, bile, pancreatic en- zymes, bacteria) also seems important in the pathogenesis of NSAID-induced enteropathy.
The diagnosis of NSAID-induced enteropathy is com- plex. Radiolabeled laboratory tests that use indium In– labeled leukocytes, chromium Cr 51–labeled red blood cells, or ethylenediaminetetraacetic acid are expensive and cumbersome. Our diagnostic approach was to use an HMPAO white blood cell scan to detect the presence of intestinal inflammation. This test has been evaluated exten- sively and has been shown to be accurate in the assessment of patients with intestinal inflammation caused by active Crohn disease of the small bowel. However, its perfor- mance has not been formally assessed in diagnosing NSAID-induced enteropathy. Recently, an assay detecting fecal calprotectin was shown to correlate with results of a 4-day fecal excretion of indium In–labeled white blood cells in diagnosing NSAID-induced enteropathy.6 Small bowel enteroscopy has been used to investigate blood loss in patients taking NSAIDs. However, the procedure is lengthy and often nondiagnostic.
The recommended therapy for NSAID-induced enter- opathy is still undefined. After the cause of the enteropathy has been determined, discontinuation of use of the offend- ing agent is the ideal strategy. If this strategy is not pos- sible, the NSAID dosage should be decreased. Cyclooxy- genase-2 selective inhibitors (celecoxib, rofecoxib) have fewer toxic effects on the small intestine vs traditional nonselective NSAIDs,10-12 therefore becoming a more suit-
able choice for such patients. Sulfasalazine has been shown to reduce intestinal inflammation and blood loss in patients taking NSAIDs.13 Synthetic prostaglandins are known to limit NSAID-induced gastroduodenal damage and may also help prevent development of NSAID-induced intesti- nal ulcerations. Proton pump inhibitors have been shown to allow healing of NSAID-induced ulcers in patients receiv-…
Related Documents
