Mycobacterial disease causing chylous effusions in two patients living with uncontrolled HIV
- 1 Critical care, University Hospital Aintree, Liverpool, UK
- 2 HIV department, Chelsea and Westminster Hospital NHS Foundation Trust, London, UK
- Correspondence to Dr Jonathan Falconer; johnathanfalconer@gmail.com
Abstract
Chylous effusions are a rare complication of disseminated non-tuberculous mycobacterial (NTM) infection. This is a case couplet reporting on the treatment challenge of chylous effusions secondary to NTM infection in two individuals living with advanced HIV. Their treatment was complicated by associated immune reconstitution inflammatory syndrome. They both required intermittent paracentesis, steroid treatment and transitioning on to fat-free diets alongside NTM treatment. Only after months of this treatment regimen was successful resolution of the associated chylous effusions achieved. Chylous effusions in immunosuppressed patients living with NTM infection are rarely reported and difficult to manage. This report discusses treatment approaches and highlights the difficulties faced by the treating team.
Background
Chylous effusions, including chylous ascites and chylothorax, have a wide range of non-traumatic causes including occurrence as a rare complication of disseminated non-tuberculous mycobacterial (NTM) infection.1 However, in contrast to a focal obstruction of the lymphatic system as seen in Mycobacterium tuberculosis infection,2 our cases demonstrate a diffuse lymphadenitis within the mesentery and the retroperitoneum. This has been described in previous case series as ‘sclerosing mesenteritis’ and is associated with a diffuse fibrotic reaction that is more commonly seen in infections with the NTM Mycobacterium avium and Mycobacterium genavense (MG),3 4 especially in patients living with HIV with advanced immunocompromise5 where immune reconstitution inflammatory syndrome (IRIS) plays a crucial role in the clinical progression. This case report is unique in summarising two cases which were preceded by IRIS and in highlighting the challenges presented by these two disease processes.
Case presentation
Case 1
A man in his 40s was diagnosed with HIV with a CD4 of 2 cells/µL and a viral load (VL) of 467 735 of copies/mL. He was given raltegravir and tenofovir disoprixil fumarate/emtricitabine. MG was cultured from a biopsy of a coalesced retroperitoneal lymph node mass, found on abdominal CT (figure 1). Antimycobacterial treatment with ethambutol, rifabutin and azithromycin was given. Azithromycin was subsequently switched to levofloxacin due to gastrointestinal intolerance. Eight months later, he presented with shortness of breath and abdominal swelling. A CT of the chest/abdomen/pelvis (CT CAP) revealed bilateral pleural effusions and large volume intra-abdominal ascites. The intra-abdominal lymph node mass had not changed in size. HIV VL was less than 20 copies/mL, and CD4 count was 50 cells/µL.
Coronal CT image of mesenteric nodal mass of case 1.

Pleural and ascitic drain fluid was chylous, confirmed by triglyceride level in the fluid of 130 mg/dL. Repeat biopsy of the abdominal lymph nodes demonstrated persistent mycobacteria and fibrosis. Mycobacterial culture of the tissue sample was negative after extended incubation. A lymphangiogram suggested the cause of the chylous effusion to be nodal involvement and lymphatic obstruction within the proximal small bowel mesentery. After extensive investigation, other non-traumatic causes of chylothorax were excluded, and a diagnosis of sclerosing mesenteritis was made.
He was given prednisolone 1.5 mg/kg to reduce inflammation driven by immune reconstitution and minimise the progression of fibrosis. Repeat ascitic drains were required due to persistent reaccumulation of chylous fluid. Octreotide 100 µg two times per day was given subcutaneously, and tamoxifen 20 mg once daily orally was added to the treatment regimen together with a low-fat diet. Over a 4-week admission, the chylous output reduced sufficiently for the patient to be discharged. He was weaned off steroids over the course of 4 months but required intermittent abdominal paracentesis.
Six weeks after weaning steroids completely, he was readmitted due to increasing reaccumulation of ascites. The octreotide dose was increased to 200 µg three times per day, thalidomide 100 mg once daily was added and prednisolone 60 mg was restarted. The patient was again discharged. Two weeks later, he was admitted a third time with increased chylous ascites (draining~1 L/day). Over the course of a 3-month admission, total parenteral nutrition (TPN) was started with fast and significant reduction of the chylous effusion. Ocreotide and thalidomide were stopped, and the patient was successfully discharged on TPN and fat-free only oral diet, without reaccumulation of chylous effusion for the past 6 months. Steroids were weaned slowly (over the course of 6 months). He continues to fail to immune-reconstitute (latest CD4+25, 7%; VL <20 copies/mL) and remains on antimycobacterials (22 months of treatment), aiming to complete 24 months of treatment.
Case 2
A man in his early 30s was diagnosed with HIV and disseminated M. avium complex (DMAC) via laparoscopic retroperitoneal lymph-node biopsy. His CD4 count at diagnosis was 15 cells/µL, and his HIV VL 10 million copies/mL. He was given antiviral therapy with raltegravir+tenofovir DF/emtricitabine together with rifabutin, ethambutol and azithromycin for DMAC. Two months into treatment, he suffered virological failure with development of M184V in the reverse transcriptase gene and 155NH in the integrase gene.
Resistance sequencing on his original VL had not demonstrated any previous resistance, so non-adherence to his antiretroviral treatment (ART) was suspected. The importance of adherence was emphasised, and a direct observation of treatment strategy (DOTS) was instituted to improve adherence to the patient’s DMAC medications. His antiretrovirals were switched to doravirine 100 mg two times per day, dolutegravir 50 mg two times per day and tenofovir DF/emtricitabine, with subsequent VL suppression. The VL suppression and good engagement with DOTS suggested improved adherence.
Over the following 12 months, the patient had multiple admissions with abdominal pain and enlargement and necrosis of mesenteric and retroperitoneal lymph nodes, thought to be secondary to IRIS, managed with high-dose steroids.
Five months after starting DMAC treatment, further mycobacterial blood cultures were positive and resistant to clarithromycin. His antimycobacterial regimen was optimised to rifabutin, linezolid and ethambutol.
Twelve months after treatment optimisation, he was admitted with shortness of breath and abdominal swelling. CT CAP showed bilateral pleural effusions, significant ascites and an unchanged appearance of the mesenteric and retroperitoneal lymph-nodal mass. Drains in the peritoneum and chest drained high volumes of chylous fluid daily (303.8 mg/dL), with an output of up to 5 L/day from the chest. A lymphangiogram showed no lymphatic focal obstruction but evidenced extremely slow progression of the dye above the diaphragm. Repeat mycobacterial blood cultures were negative, and extensive investigation for non-traumatic causes of chylous effusions was negative (see differential section).
On the basis of radiological features and clinical picture, he was diagnosed with IRIS and sclerosing mesenteritis. High dose oral steroids (1.25 mg/kg of prednisone) were given, and octreotide 100 mg two times per day and tamoxifen 100 mg once daily were added into his treatment regimen. A medium-chain triglyceride diet was given.
In spite of this, he continued to output high volumes of chyle from his chest drain. TPN was trialled. Chylous output reduced drastically to a few millilitres a day, and the chest drain was removed. A low-fat oral diet was gradually introduced with minimal reaccumulation, and the patient was subsequently able to be discharged (CD4 22 cells/µL, 7%; VL <20 copies/mL). Ocreotide was stopped, and steroids were gradually weaned over 3 months. Four months later, he was readmitted for recurrence of pleural and ascitic chylous effusions. A repeat CT of the abdomen showed increase in size in the known intra-abdominal mass; this was biopsied with evidence of granulomatous inflammation and acid fast bacilli seen. Mycobacterial cultures failed to grow. Chest and abdominal drains were inserted, and steroid treatment was restarted (1.5 mg/kg prednisolone). High chylous ascites production continued (1–2 L/day) despite a fat-free diet, so an intra-abdominal pleurex drain was inserted to facilitate discharge from hospital. Ascitic production reduced gradually over the following 3 months and the drain was removed.
Steroids were weaned gradually over the course of 5 months. Investigations at 12 months show evidence of some immune-restoration (CD4 143 cells/µL, 12%; VL <20) and reduction in the size of intra-abdominal lymph nodal mass, with no new ascites nor pleural effusion.
Differential diagnosis
There is a wide differential for non-congenital, non-traumatic causes of chylous effusions in these patients. This includes lymphoma, lymphangiolieomyomatosis, Kaposi sarcoma, oesophageal cancer, mulitcentric castlemann’s disease, histoplasmosis and filariasis. Extensive investigation including biopsy excluded these. Differentiating between treatment failure, IRIS and what seems to be the healing end-result of fibrosis can prove a diagnostic challenge. Delayed onset of symptoms, repeated culture being negative for mycobacterial infection (with the caveat of MG being difficult to grow) and effective virological suppression were suggestive in these cases of ongoing inflammation (IRIS) leading to scarring, with fibrosis of the involved lymphatic sites and their surrounding structures.
Treatment
Management of chylous effusions secondary to NTM infection in HIV hosts starts with viral control and treatment of the mycobacterial infection, with steroids when indicated for IRIS.
A high mycobacterial burden requires prolonged treatment with three drugs (ie, rifamycins, a macrolide and ethambutol) to be efficacious and avoid onset of resistance.6 Persistence of high antigenic stimuli and immune reconstitution most likely drive the inflammatory process, even after antimicrobial success; hence, the frequent prolonged need for steroids.
Agents with potential antifibrotic activity like thalidomide and tamoxifen have been reported to improve or even resolve chylous effusions;7 8 however, it is hard to discern efficacy in this complex clinical picture.
Nutrition plays a crucial role in reducing and slowing the reaccumulation of chylous fluid through the implementation of a diet restricted to medium chain triglycerides only or, in severe cases, by administration of TPN,1 as it has been the case for our patients.
The effusions, however, are often refractory and further medical treatments, although with limited evidence, include the use of octreotide.9 This somatostatin analogue is known to decrease the intestinal absorption of fats, thereby reducing lymph flow in the major vessels.10 Lymphangiography with high-density lipiodol dye has been suggested to offer benefit therapeutically as well as being a diagnostic tool, most likely through inflammation or embolization of the lymphatic vessels,11 although this applies mainly for traumatic chylous leaks and did not prove beneficial in our patients.
Symptomatic effusions required multiple hospital admissions in both cases. The use of intermittent pleural and abdominal drainage gave temporary symptomatic benefit and facilitated discharge from hospital even if it did not address the underlying aetiology. In light of the chylothorax and chylous ascites, a pleuroperitoneal shunt was suspected. The subsequent use of indwelling drainage catheters, specifically an indwelling abdominal drain for 3 months in case 2, was helpful in avoiding hospital admissions while the medical therapies took effect.
Pleurodesis was considered in case 2 for refractory chylothorax but was not necessary in the end. Surgical pleurovenous or pleuroperitoneal shunting has been used in previously reported cases to manage high-output chylous effusions, although usually for traumatic chylous effusions postoperatively.12 13
Outcome and follow-up
Both patients remain well and continue to improve in the outpatient setting. They are followed up on a 3 monthly basis by their parent HIV team.
Discussion
There is no standard of care for sclerosing mesenteritis with chylous effusions, and the management of these patients remains challenging. In spite of advances in the treatment and control of HIV and NTM infection, the prognosis for this group of patients remains poor.14 As with the majority of cases in the literature, both our cases required prolonged treatment with a combination of antibiotics, steroids and the therapies mentioned above. Optimal dose and duration of steroids were determined on a case-by-case basis, balancing the associated side-effects and complications of steroids with the symptomatic and nutritional relief they brought.
Learning points
-
In a patient with advanced HIV and non-tuberculous mycobacterial infection, recurrence of symptoms, even months after establishing treatment, can be due to a delayed immune reconstitution inflammatory syndrome (IRIS) complicating their recovery.
-
Differentiating between treatment failure or IRIS in these difficult-to-culture organisms can be complex.
-
There remains a limited evidence base for these chylous complications of mycobacterial infections. A multi-disciplinary team (MDT) approach is therefore crucial to managing these cases.
Ethics statements
Patient consent for publication
Footnotes
-
Contributors The following authors were responsible for the drafting of the text, the sourcing and editing of clinical images, investigation results, and critical revision for important intellectual content:KED, JF, MB and AP. The following authors gave final approval of the manuscript: AP, MB, KED and JF.
-
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
-
Case reports provide a valuable learning resource for the scientific community and can indicate areas of interest for future research. They should not be used in isolation to guide treatment choices or public health policy.
-
Competing interests None declared.
-
Provenance and peer review Not commissioned; externally peer reviewed.
- © BMJ Publishing Group Limited 2023. No commercial re-use. See rights and permissions. Published by BMJ.
References
Use of this content is subject to our disclaimer