Localised intestinal Kaposi sarcoma in a patient with non-coeliac seronegative villous atrophy

Abstract

Seronegative villous atrophy (SNVA) is a diagnostic challenge for gastroenterologists, which is defined by villous atrophy and negative coeliac serology. Non-coeliac forms of SNVA, such as autoimmune enteropathy, can be life-threatening leading to intractable diarrhoea and severe malabsorption that require systemic immunosuppression. When all known causes have been excluded, it is termed idiopathic villous atrophy (IVA). We present a case of non-coeliac SNVA complicated by Kaposi sarcoma (KS). A previously well HIV-negative man in his 30s presented with a 4-month history of watery diarrhoea and 25 kg weight loss. After prolonged investigation, he was diagnosed with non-coeliac SNVA without an identified aetiology that would be consistent with IVA. Clinical recovery was achieved with parenteral nutrition for type II intestinal failure and immunosuppression using high-dose corticosteroids. On subsequent gastroscopy, he was diagnosed with localised intestinal KS prompting cessation of all immunosuppression but remained in clinical remission.

Background

Kaposi sarcoma (KS) is an angioproliferative disorder requiring infection with human herpes virus-8 (HHV-8) to develop.¹ KS is classically divided into four types known as ‘classic’, ‘endemic’, ‘iatrogenic’ and ‘epidemic’.² The epidemic form of KS is associated with HIV and AIDS. It is the most common tumour associated with AIDS, typically presenting with multiple mucocutaneous lesions but visceral involvement is common, particularly in the gastrointestinal (GI) tract.³ GI involvement in iatrogenic KS is poorly characterised but commonly seen in post-transplant cohorts.⁴ However, it is increasingly recognised in chronic inflammatory conditions that require immunosuppressive therapies such as inflammatory bowel disease (IBD).⁵

Seronegative villous atrophy (SNVA) is a term used to describe the presence of villous atrophy in the absence of positive coeliac serology, and the most common cause is seronegative coeliac disease (SNCD).⁶ SNVA poses a diagnostic challenge, as outlined by the Paris consensus, because there are a myriad of causes with similar histopathological and clinical features.⁷ One of these conditions is autoimmune enteropathy (AIE), which is a rare malabsorption syndrome first described in children but increasingly recognised in adults. It is characterised by intractable diarrhoea and severe malabsorption usually requiring systemic immunosuppression and nutritional support.^{8 9} The disease is heterogeneous and associated with circulating anti-enterocyte (AEA) and anti-goblet cell autoantibodies, but these are not required for the diagnosis and found in other severe enteropathies.¹⁰ When a definitive diagnosis of SNVA is not identified, which is seen in up to 14% of adults, the term idiopathic villous atrophy (IVA) is used.⁶ IVA is a relatively new and poorly defined term of which three subtypes have been proposed.¹¹ Like AIE, patients with IVA can develop severe complications and may require immunosuppression. Here, we present a case of non-coeliac SNVA secondary to IVA with intestinal failure that was complicated by localised intestinal KS.

Case presentation

A previously fit and well man in his 30s presented with a prolonged 7-month history of nausea and watery diarrhoea in association with 25 kg weight loss. His body mass index (BMI) fell from 24.9 to 16.6 kg/m². There was no significant family history, he was a non-smoker and worked in construction. Clinical examination was unremarkable except for low body weight.

During this 7-month period, he had an initial 2-month admission in Poland undergoing a series of investigations including gastroscopy, colonoscopy, coeliac serology and cross-sectional imaging. This confirmed duodenal villous atrophy with chronic inflammatory infiltrates in association with shortened villi in the distal ileum. He was negative for both IgA tissue transglutaminase (TTG) and IgA endomysial antibodies (EM), and CT of the abdomen and pelvis showed non-specific mesenteric lymphadenopathy. There was no evidence of IgA deficiency. He was diagnosed with suspected SNCD in the absence of human leucocyte antigen (HLA) testing and started on a gluten-free diet (GFD) with a weaning course of oral prednisolone for refractory symptoms.

After returning to the UK for work and completing 2 months of a GFD, he was admitted to our institute with persistent symptoms of weight loss and diarrhoea. His admission weight was 48.7 kg and BMI 16.3 kg/m². We initially reassessed the diagnosis of SNCD by completing a series of basic investigations which are shown in table 1. Importantly, testing confirmed a negative IgA anti-TTG and anti-EM with no evidence of IgA deficiency. HLA-DQ2/DQ8 typing was negative, which essentially excluded SNCD. We opted to perform imaging with CT of the abdomen and pelvis and MRI of the small bowel, which again showed multiple non-specific small mesenteric lymph nodes but no other pathology. A repeat gastroscopy with duodenal biopsies confirmed evidence of duodenitis with villous atrophy, cryptitis and crypt abscesses without significant intraepithelial lymphocytosis (figure 1A). Congo red, Ziehl-Neelsen and periodic acid-Schiff-diastase stains were all negative. Repeat biopsies confirmed villous atrophy with an absence of significant intraepithelial lymphocytosis and paucity of goblet cells, which was suggestive of AIE.

Table 1.

Summary of basic initial investigations

Abnormal results	Abnormal value	Normal range
Haemoglobin	127	135–170 (g/L)
Iron	5.5	11–36 (μmol/L)
Vitamin B12	1649	160–925 (ng/L)
Sodium	134	135–145 (mmol/L)
Potassium	3.1	3.5–5.1 (mmol/L)
Urea	9.5	2.1–7.1 (mmol/L)
Bilirubin	30	0–21 (μmol/L)
Alanine transaminase	56	10–50 (U/L)
Calcium (adjusted)	2.16	2.20–2.60 (mmol/L)
Calprotectin	409	0–51 (μg/g)
Prothrombin time	15	9–12 (s)
SARS-CoV-2 antibody	Positive
Varicella zoster virus IgG qualitative	Positive
Epstein-Barr virus IgM screen	Positive
Epstein-Barr virus IgG screen	Positive
Cytomegalovirus IgG screen	Positive
Hepatitis B surface antibody	Positive
Hepatitis B core antibody	Positive
Hepatitis A total antibody	Positive
Normal results
APTT, albumin, alkaline phosphatase, aspartate transaminase, autoiummune screen, coeliac screen, creatinine, ferritin, folate, Helicobacter pylori stool antigen, hepatitis B surface antigen, hepatitis C antibody, HIV antibody, immunoglobulins, magnesium, phosphate, platelets, stool viral, bacterial, and Clostridioides difficile PCR, stool ova, cysts and parasites, syphilis screen, thyroid function, white cell count

APTT, activated partial thromboplastin time.

Figure 1.

Histopathology of duodenal biopsies. (A) Duodenal mucosa showing villous atrophy and crypt hyperplasia with a little intraepithelial lymphocytosis and active chronic inflammation with crypt abscess. (B) Duodenal mucosa showing proliferation of spindle cells with poorly formed vascular spaces and red blood cells consistent with Kaposi sarcoma.

A full dietetic assessment was completed due to significant weight loss and malnutrition. A decision was made to start on total parenteral nutrition (TPN) for intestinal failure. We opted to trial a series of empirical treatments for SNVA while managing his nutrition and conducting further investigations (summarised in figure 2). This included a trial of budesonide (9 mg daily, open capsule) for 2–3 weeks that was followed by high-dose hydrocortisone (100 mg intravenous four times a day) due to a lack of clinical response. This was completed for 7 days followed by a weaning course of prednisolone (starting dose 40 mg) over 4 weeks. Despite these treatments, he continued to be symptomatic, but weight did improve with TPN. We completed autoantibodies for AIE including anti-enterocyte and anti-goblet cell, which were both negative implying a diagnosis of IVA. Due to the severity and rarity of the case, it was reviewed by several local histopathologists and discussed at a national enteropathy unit. Based on all information, he was treated with the working diagnosis of seronegative AIE with a plan to initiate anti-tumour necrosis factor alpha (anti-TNF-alpha) monoclonal antibody therapy as second-line treatment. This decision was delayed due to complications with suspected line sepsis and hospital-acquired pneumonia, likely precipitated by a state of immunosuppression from protein-losing enteropathy.

Figure 2.

A comprehensive summary of the treatments and nutritional support given during the inpatient admission. Loperamide 2 mg three times a day; metronidazole (METRO) 400 mg three times a day; doxycycline (DOXY) 100 mg two times per day; budesonide 9 mg once daily; mesalazine 2 g two times per day; bile acid sequestrant: colesevelam 625 mg two times per day, increased to 1875 mg two times per day then combined with colestipol 5 g two times per day; hydrocortisone (HYDRO) 100 mg four times a day; prednisolone 40 mg weaning by 5 mg every 5 days; ciprofloxacin (CIPRO) 400 mg two times per day; Creon 40 000–50 000 units three times a day; octreotide (OCT) 50–100 µg three times a day.

Once improved, before administration of anti-TNF-alpha, we reassessed for histological improvement following steroids and identified a focal area of intestinal KS that was positive for HHV-8 on duodenal biopsy (figure 1B). A repeat HIV-1/2 serum antibody, HIV viral load and HHV-8 viral load were negative. A full dermatological assessment excluded any mucocutaneous lesions. Following discussion with a specialist KS unit, the lesion was felt to be reactivation from previously acquired HHV-8 in the context of a relative immunodeficiency state from the underlying enteropathy. Due to the risk of progression with systemic immunosuppression, we proceeded with a trial off TPN as diarrhoea was starting to improve. Given stability off TPN, a decision was reached to take a watch-and-wait approach rather than instigating an immunomodulator as a potential long-term option to prevent relapse. He was discharged with a weight of 63.8 kg, BMI 21.3 kg/m², off any form of supplemental nutrition. He remains in clinical remission 6 months on from discharge but repeat biopsies were not completed to confirm remission. There has been a self-reported steady increase in weight, and he is planned for 6-month follow-up.

Discussion

We present a severe case of non-coeliac SNVA with intestinal failure that was complicated by KS following treatment with high-dose corticosteroids. This represents a novel finding of KS in the context of a severe non-coeliac enteropathy possibly secondary to seronegative AIE or a type 2 IVA given the absence of AEA. We hypothesise that a prolonged state of protein-losing enteropathy in combination with high-dose corticosteroids resulted in reactivation of HHV-8 leading to focal intestinal KS. Immunosuppression is crucial to the development of intestinal KS by enabling reactivation and proliferation of previously acquired HHV-8.⁴ This is represented in several cases of KS in patients treated with long-term immunosuppression for IBD.⁵ In these cases, corticosteroids are heavily represented, often combined with other immunosuppressive agents.^{5 8} In transplant cohorts, the mainstay of treatment is attenuation of immunosuppression while preserving graft function.⁴ However, KS has often been observed in refractory inflammatory diseases, so this strategy is not viable. A recently reported case of KS in association with ulcerative colitis treated with vedolizumab was managed with surgical intervention (colectomy).⁵ This is a familiar strategy in IBD-associated colorectal KS because colectomy enables both definitive treatment of KS and cessation of immunosuppression. Nevertheless, when surgery is not optional, a difficult balance needs to be drawn between treating the inflammatory condition and not perpetuating this rare complication.

Even in the absence of KS, SNVA poses a challenge for the general gastroenterologist. IVA and AIE are both extremely rare and seen in those unresponsive to a GFD. These conditions may mimic SNCD and often referred to tertiary centres as refractory coeliac.^{7 12} We point the reader to review articles, clinical practice guidelines and data from the national centre for coeliac disease in the UK for more information on the differential diagnosis and challenges in SNVA.^{6 13 14} In our case, the absence of an underlying cause of enteropathy including negative AEA was suggestive of IVA. However, after extensive case discussion, it was felt it could represent seronegative AIE given the prolonged history, presentation with intestinal failure and typical histopathological features including minimal intraepithelial lymphocytosis, paucity of goblet cells and crypt abscesses. The presence of AEA antibodies is not required for the diagnosis of AIE due to poor specificity.⁹ In one European cohort of patients with adult-onset AIE, AEA antibodies were identified in 77% (10 of 13) of patients, but also 7.7% (4 of 5) of coeliac disease, 16.7% (3 of 18) of type II refractory coeliac disease and 20% (2 of 10) of enteropathy-associated T cell lymphoma.¹⁵ Therefore, these antibodies may not be useful at differentiating between severe causes of enteropathy because they are not pathognomonic. In fact, they have been considered a secondary epiphenomenon reflecting severe enteropathy with reduced titres seen on resolution of mucosal inflammation.¹⁰ Therefore, treatment with corticosteroids in our case may have caused a fall in autoantibodies below detectable levels. In this setting, it could represent a true seronegative form of AIE without generation of this humeral immune response, but the diagnosis most resembles IVA. The latter may be the best distinction given that the presence of AEA antibodies is part of the diagnostic criterion in the Paris consensus on non-coeliac enteropathies, and seronegative AIE is poorly defined in adults.^{7 11}

Whether this represented IVA or a true seronegative AIE, first-line treatment remains the same with corticosteroids such as open-capsule budesonide as for other enteropathies of unknown origin.¹² We know from cases of AIE that this is associated with a response rate of 85% but longer follow-up is needed regarding disease relapse.^{12 16} When steroids fail, patients require escalation of therapy with a variety of drugs being used including thiopurines,⁹ infliximab,¹⁷ ustekinumab¹⁸ and even autologous stem cell transplantation.¹⁵ Maintenance therapy is usually offered due to the risk of relapse or refractory symptoms with low-dose budesonide or immunomodulators.^{9 15 16} Early case series suggest over two-thirds of patients require additional maintenance therapy following remission with steroids.⁹ However, more recent experience suggests one-third of patients can maintain long-term drug-free remission.¹⁵ In IVA, there is a paucity of data on treatment options. The largest retrospective series of IVA reported second-line treatment with azathioprine, adalimumab and vedolizumab.¹¹ In group 2 IVA, that is most akin to our case, clinical response was seen in 64% (9 of 13).

Unfortunately, there is no clear consensus on the most appropriate second-line treatment in enteropathy of unknown origin. We observed a clinical response within 4–6 weeks, but due to the severity of his illness and TPN dependence, we had discussions about early escalation of therapy that could have exacerbated the KS. This novel complication of non-coeliac SNVA reflects the severity of intestinal failure and loss of mucosal immunity that may have been attenuated by earlier disease control. Additionally, the use of open-capsule budesonide followed by monoclonal antibody therapy in favour of systemic corticosteroids may have altered the disease course. Long-term, how we identify patients at risk of HHV-8 reactivation across all chronic inflammatory GI disease needs to be defined.

In summary, we report a novel case of intestinal KS in a patient with non-coeliac SNVA secondary to IVA that was treated with high-dose corticosteroids. Clinicians should be aware that severe forms of non-coeliac SNVA can include AIE and IVA that represent potentially fatal malabsorption syndromes unresponsive to dietary manipulation. There is an urgent need for a treatment strategy among this cohort of patients, particularly for second-line treatments and how to address maintenance therapy to reduce risk of relapse. Given the current lack of guidance, treatment should involve a multidisciplinary team discussion, and clinicians should be vigilant to the potentially serious complication of GI KS with immunosuppressive therapies in patients with severe GI disease.

Learning points.

• Non-coeliac seronegative villous atrophy poses a diagnostic challenge for gastroenterologists.

• Treatment of non-coeliac seronegative villous atrophy should involve a multidisciplinary team.

• Severe forms of idiopathic villous atrophy and autoimmune enteropathy can lead to intestinal failure and secondary complications including Kaposi sarcoma.

• Patients with suspected autoimmune enteropathy should have anti-enterocyte antibodies but these are not positive in all cases.

• Treatment of non-coeliac seronegative villous atrophy of unknown cause should be with open-capsule budesonide, but there is urgent need for a treatment strategy when first-line therapy fails.

Ethics statements

Patient consent for publication

Obtained.