Diarrhea in a Long-Term Kidney-Pancreas Recipient

On the Basis of the Clinical and Laboratory Data, Which Is the Most Likely Explanation for Her Diarrhea?

1. Diabetic enteropathy

2. Mycophenolate toxicity

3. CMV colitis

4. Lymphoma

Discussion

The correct answer is “B,” mycophenolate toxicity. Diarrhea in long-term recipients of kidney or kidney–pancreas transplants can be because of infectious causes, drug causes, malignancy, and metabolic causes, such as diabetic enteropathy (Table 1). Common infectious causes include CMV colitis, Clostridium difficile colitis, and norovirus infections, as well as bacterial, parasitic, and fungal organisms (1,2). Drug causes are also common, with mycophenolic acid (MPA) being the most common. Malignancy and metabolic causes must also be considered. It is critical to have a reliable laboratory to identify not only common pathogens but also viruses, especially norovirus, which is common and may be severe and not self-limited in organ recipients (3).

In this patient, tests for viral, bacterial, and parasitic pathogens and C. difficile toxin were negative. Although CMV colitis can occur at any time after transplant, it is more common in the early (1–2 years) post-transplant period and in patients who are seronegative receiving a seropositive CMV kidney (unlike this patient), and it would be unusual to persist for 2 months without fever and leukopenia (4). Lymphoma is always a possibility but the absence of abdominal pain or findings on computed tomography scan make this diagnosis less likely. Diabetic enteropathy is not usually associated with significant weight loss and would be unusual in a recipient with no previous enteropathy symptoms, who is now 7 years post-transplant.

Subsequent blood tests (CMV and Epstein–Barr virus PCR) and lactate dehydrogenase provided more evidence against CMV and lymphoma as etiologies. A colonoscopy is performed and the results are negative.

Which Is the Best Treatment for This Patient?

1. Start loperamide

2. Lower mycophenolate dose

3. Discontinue mycophenolate and maintain on tacrolimus and prednisone

4. Switch mycophenolate to azathioprine

Discussion

The correct answer is “D,” switch from mycophenolate to azathioprine. Starting loperamide would not get at the proximal cause. Further reduction in MMF dosage (to 250 mg twice daily) or stopping it without adding azathioprine could leave her underimmunosuppressed, especially as a recipient of a pancreas transplant. Discontinuing MMF when a 50% reduction in drug dose does not yield resolution, and substituting azathioprine is the approach recommended in a recent review of severe diarrhea in recipients of kidney transplant (1). This was done and within weeks her diarrhea had subsided. Within a month, she gained back much of the weight she had lost.

MPA is an antimetabolite that reversibly inhibits the inosine monophosphate dehydrogenase pathway needed for purine synthesis. Activated B and T lymphocytes depend on this pathway for de novo synthesis of purine nucleotides because they lack a salvage pathway for this process. Thus, MPA is ideal for inhibiting lymphocyte proliferation. Because it is poorly absorbed, MPA is available as two prodrugs, MMF (Cellcept) and mycophenolate sodium (Myfortic). Both drugs are converted to MPA to exert their biologic effect. For the remainder of this discussion, we will use the term MPA instead of MMF or mycophenolate sodium. MPA has largely replaced azathioprine in the prevention of rejection in most United States transplant centers and is recommended as the antimetabolite of choice in Kidney Disease Improving Global Outcomes Guidelines (5). It is also used to treat lupus nephritis, membranous glomerulopathy, psoriasis, and autoimmune disorders. Although it has been demonstrated to reduce rejections more than azathioprine in cyclosporine-treated transplant recipients, it causes more gastrointestinal symptoms, including diarrhea (6).

Most diarrhea with MPA occurs early after transplant (7) and may respond to a reduction in dose. Tacrolimus can also cause diarrhea but MPA is a more common culprit. Caution must be exercised when reducing or discontinuing MPA dose, as this has been associated with rejection and decreased allograft survival (8).

The incidence of severe diarrhea occurring later, after the initial period of MPA introduction, and associated with significant weight loss and malabsorption may be a distinct entity. Patients have been reported as presenting with diarrhea, weight loss, and malabsorption and subsequently, villous atrophy of the small bowel is demonstrated with duodenal biopsy or capsule endoscopy (9,10). Other patients have been reported with colitis found on colonoscopy and pathology demonstrating lesions similar to Crohn disease, graft versus host disease, or ischemic lesions (11–13). Cases have been most commonly described in recipients of kidney transplant but have also been reported in other solid organ transplants (13), as well as patients with autoimmune disease. Although the mechanism of these lesions is not completely understood, the presence of crypt atrophy, abscesses, and apoptosis in the lesions has led to the hypothesis that apoptotic cell death may be a pathogenic factor in the mucosal injury (12). Whether this is because of direct drug toxicity or an immune effect is not clear. It is also not clear whether this mechanism of intestinal epithelial cell injury plays any role in diarrhea observed when MPA is initially introduced, and which often responds to a dose reduction.

Many patients with reported cases of severe diarrhea and malabsorption have described the need for drug discontinuation to relieve symptoms (9,14). This is why it is important to recognize the entity of MPA colitis or small intestinal villous atrophy as an entity. It has been speculated that that some of the previously described cases of this entity could be undiagnosed norovirus infection (1), but further study is needed to determine whether this is true. This was not the case with our patient, as norovirus was not detected. There does not seem to be a correlation between MPA blood level and diarrhea (15). Although a few case reports have described improvement by switching to enteric-coated MPA (10), one would not expect this result to be common as the effects of MMF and enteric-coated mycophenolate in the small intestine and colon are similar.

In summary, when severe diarrhea occurs in a recipient of a kidney or kidney–pancreas transplant, the most common infectious causes must first be excluded (1–3,13). Viruses, especially norovirus, need to be looked for, as this infection is becoming increasingly appreciated as a cause of severe, chronic diarrhea, requiring a reduction or change in immunosuppression to achieve resolution (1,3). Colonoscopy with biopsies should always be performed when an initial evaluation is negative, to look for colitis and to exclude CMV colitis and lymphoma. CMV gastrointestinal involvement may occur in the absence of CMV viremia. When infectious causes are excluded with reliable testing, one needs to consider the entity of MPA-induced colitis or small intestinal villous atrophy. The diarrhea and associated malabsorption in such cases often requires complete discontinuation of the drug, with the substitution of azathioprine in most patients to reduce the risk of rejection.