Abstract
Diabetes mellitus, type 1 in particular, is a well-recognised risk factor for melioidosis, a disease caused by Burkholderia pseudomallei. Melioidosis is endemic in Southeast Asia and in northern Australia and has a variety of clinical presentation, isolated splenic abscess being one of them. B. pseudomallei, however, is an uncommon aetiology of splenic abscess. The diagnosis of melioidosis is often overlooked unless the clinician and the microbiologist are suspicious of the condition. Multiple splenic abscesses and perisplenic collection were noted in CT scan of the abdomen in a patient of type 1 diabetes, presenting with fever for preceding 4 weeks. B. pseudomallei was isolated from the splenic aspirate and the diagnosis was made based on gram stain and routine biochemical tests. He was successfully treated with antibiotics. We postulate that the likely route of infection was inoculation through skin, the integrity of which was compromised by multiple subcutaneous insulin injections.
Keywords: diabetes, tropical medicine (infectious disease)
Background
Splenic abscess, once considered an uncommon entity, is recently being diagnosed more frequently due to widespread use of various imaging modalities in routine practice, and increasing survival of immunocompromised patients. Abdominal trauma, intravenous drug abuse and haemoglobinopathies, like sickle cell anaemia, are the other important precipitating factors for splenic abscess. Splenic abscess may develop following local spread of infection from the surrounding structures (pancreas, colon, left kidney, pelvic organs), or from distant metastatic sources (like infective endocarditis).
Aerobic gram-positive (Staphylococci, Streptococci) and gram-negative (Salmonella, Escherichia coli and Klebsiella pneumoniae) organisms are the commonly encountered microbes in splenic abscess overall, with significant geographical variation. Splenic abscess often results from haematogenous spread of these pathogens. Mycobacteria, fungi and protozoa are also encountered, especially in immunocompromised individuals. Atypical microbes like Burkholderia, Brucella, Coxiella and Bartonella are rare, even in this group of patients.
Three medically important genera Pseudomonas, Burkholderia and Stenotrophomonas are termed pseudomonads, a group of gram-negative bacteria with an inability to ferment lactose. The genus Burkholderia comprises more than 40 species, of which B. pseudomallei perhaps is the most virulent one.1 Melioidosis, an infection caused by B. pseudomallei, is largely confined to the tropics and subtropical region including parts of Southeast Asia, Thailand, India and China in particular, and outback of Australia.2 3 The organism is found in soil and water, and human acquisition of infection most often occurs through inoculation, when skin integrity is compromised. The other modes of acquisitions are inhalation and less commonly, by ingestion. Though impaired host immunity is not essential for melioidosis, many such patients have diabetes mellitus (DM), chronic alcohol use, advanced kidney disease and some forms of underlying lung diseases.4 The clinical presentation of the disease is quite varied including pneumonia, ulcerative skin lesions with regional lymphadenopathy, bone and joint involvement, chronic suppurative lesions and disseminated disease.5 6 Pneumonia is the most common primary clinical presentation, and splenic abscess has variably been reported from one-quarter to as high as three-fourths of patients with melioidosis in different series.4 6 7 Melioidosis is an important aetiology of splenic abscess in endemic areas, contributing to about 19% of such cases in some series.8
Case presentation
An 18-year-old village dweller, a known patient of type 1 DM, presented with high-grade fever and ill-defined abdominal pain for the preceding 4 weeks. Type 1 DM had been diagnosed 1 year prior to his presentation with us, following hospitalisation with diabetic ketoacidosis. He had been advised multiple subcutaneous injections of insulin since diagnosis, however, his glycaemic control had never been satisfactory, with glycated haemoglobin (HbA1c) ranging between 9.5% and 12.5%. He had been injecting two shots of human regular insulin and one shot of premixed human insulin. He used to graze cattle and often shared the same pond for bathing along with his cattle. He had been treated with multiple oral antibiotics (amoxicillin, cefixime, cefpodoxime, azithromycin) for 5–7 days by the primary care physician before his contact with us.
Clinical examination revealed a fully oriented, thin built boy (weight: 45 kg, height: 164 cm, body mass index: 16.7 kg/m2) without cutaneous markers of insulin resistance. He had fever (102°F), pallor, splenomegaly and tender left upper abdomen with localised muscle guard. Insulin injection sites were overall healthy with few hyperpigmented spots and focal areas of lipohypertrophy. We could not find any ulcer or infection over skin or soft tissue. Rest of the systemic examination was unremarkable.
Investigations
Complete blood count revealed mild anaemia (haemoglobin: 90 g/L) and polymorphonuclear leucocytosis (total count: 12.9×109/L; polymorphonuclear cells: 90%; lymphocytes: 8%; monocytes: 2%). Erythrocyte sedimentation rate was 94 mm in 1st hour and C reactive protein was 1047.6 nmol/L (normal: <57 nmol/L). Random plasma glucose was 25.31 mmol/L with blood ketone of 0.4 mmol/L. HbA1c, measured by high-performance liquid chromatography, was 11.4%. Blood cultures did not grow any organism. Liver function tests, renal function tests and chest radiograph were all normal. Contrast-enhanced CT scan of his abdomen was suggestive of multiple splenic abscesses with perisplenic collection (figure 1). He was evaluated according to the institutional protocol for ‘febrile patients’, and work-up for malaria, enteric fever, dengue fever, leptospirosis and scrub typhus came out to be negative.
Figure 1.
Contrast-enhanced CT of the abdomen showing multiple hypodense lesions within the enlarged spleen suggestive of abscesses (right panel) and perisplenic extension (left and middle panels).
Aspirate of the splenic abscess was inoculated on blood agar, nutrient agar and MacConkey agar and was incubated aerobically at 37°C for 24 hours. Blood agar and nutrient agar produced 1–2 mm round, low convex, translucent, smooth surface colonies with an earthy odour without any pigmentation. MacConkey agar showed pale, non-lactose fermenting colonies with typical wrinkled appearance (figure 2). Gram stain, done from the colonies on nutrient agar, documented gram-negative rods with bipolar staining (figure 3). Following primary isolation of the organism on solid media, the isolate was put on Vitek-2 compact system which identified the organism as B. cepacia complex. However, the clinical presentation and the typical ‘safety pin’ appearance on gram stain prompted us to proceed for conventional modes of identification. Biochemical reactions showed oxidative changes in Hugh Leifson’s oxidation fermentation media (figure 4).
Figure 2.
Round, low convex, translucent non-lactose fermenting (pale coloured) colonies in MacConkey agar after overnight aerobic incubation at 37°C.
Figure 3.
Gram-negative irregularly stained bacilli with bipolar staining (safety pin appearance) arranged haphazardly.
Figure 4.

Hugh Leifson oxidation–fermentation (O/F) media showing oxidative changes (O+/F-) after 24 hours of incubation at 37°C (colour changes noticed in test tube without paraffin (left), but not in tube with paraffin (right)).
Identification of the isolate as B. pseudomallei was confirmed based on the biochemical reactions like positive oxidase and catalase reaction, nitrate reduction, arginine dihydrolase activity, oxidation of lactose, glucose and mannitol, and negative lysine and ornithine decarboxylase reactions. The antibiotic susceptibility testing was performed on Mueller Hinton agar according to Clinical Laboratory Standards Institute guidelines. The isolate was sensitive to ceftazidime, ciprofloxacin, meropenam and co-trimoxazole, and was resistant to azithromycin, gentamicin, amikacin and polymyxin B (300 μg/disc). This typical susceptibility pattern was suggestive of B. pseudomallei.9
Differential diagnosis
Yersinia pestis and Klebsiella granulomatis are the other two bacteria that exhibit bipolar staining on gram stain. However, they belong to the family of Enterobacteriaceae. Since the isolate had a typical morphology of ‘safety pin’ appearance, was a non-fermenter, showed positive oxidase reaction and produced ammonia from arginine, our provisional diagnosis was B. pseudomallei. Wrinkled colonies of B. pseudomallei may resemble that of Pseudomonas stutzeri, but the morphology is different on gram stain. B. pseudomallei needs to be differentiated from the other pseudomonads as listed in table 1.
Table 1.
Tests used to separate Burkholderia pseudomallei from related pseudomonads
| Morphology and biochemical reactions | Pseudomonas stutzeri | B. pseudomallei | B. mallei | B. cepacia complex |
| Gram stain morphology | Gram-negative rods, uniformly stained | Gram-negative rods, irregularly stained (bipolar staining) | Gram-negative rods, uniformly stained | Gram-negative rods, uniformly stained |
| Oxidase | Positive | Positive | Positive | Positive |
| Motility | Motile | Motile | Non-motile | Motile |
| Glucose oxidation | Positive | Positive | Positive | Positive |
| Lactose oxidation | Negative | Positive | Variable | Variable |
| Mannitol oxidation | Variable | Positive | Variable | Positive |
| Arginine dihydrolase | Negative | Positive | Positive | Positive |
| Lysine decarboxylase | Negative | Negative | Negative | Variable |
Treatment
Drainage of the abscess was performed under ultrasonographic guidance. Glycaemic control was achieved with basal–bolus insulin therapy (three prandial doses of human regular insulin and one dose of insulin glargine U-100 at bed time). Intravenous meropenem (3 g/day) was administered for 2 weeks followed by oral trimethoprim–sulfamethoxazole (1600/320 mg) for 3 months.
Outcome and follow-up
Fever subsided completely after about 6 days of antibiotic therapy. Repeat CT scan after 3 months documented complete radiological resolution. No evidence of relapse has been reported after 1 year of follow-up.
Discussion
B. pseudomallei, though not an opportunistic pathogen, commonly infects people with DM. Percentage of patients with DM varies from 40% to 60% in different series of confirmed melioidosis.2 10 An association of melioidosis with endogenous insulin deficiency has been suggested long time ago. Insulin markedly inhibits the growth of this organism both in vitro and in vivo, and insulin levels, thus, may play a significant role in modulating the pathogenesis of B. pseudomallei, particularly in disseminated infections.11 This is also supported by the fact that melioidosis is encountered more frequently in type 1 DM. However, insulin deficiency was seen in 5%–10% of all patients with DM in some cohorts of melioidosis.12 13 The role of insulin deficiency as a risk factor of melioidosis, hence, has been challenged and a number of alternate explanations have been put forward. Patients with DM, irrespective of type, demonstrate altered innate immunity due to impairment of neutrophil functions (impaired chemotaxis, phagocytosis, oxidative burst and killing activity). Patients with DM, thus, are predisposed to all infections including melioidosis.14 It is important to note that, m-cresol (3-methylphenol), a preservative used in all injectable insulin preparations, was found to inhibit the growth of B. pseudomallei and might have confounded the results of the earlier studies.15 16 In addition, direct dermal inoculation through insulin injection sites or insulin pump tube insertion sites might also explain the higher risk in type 1 DM.17 Bathing in muddy pool and breached cutaneous integrity due to subcutaneous insulin injections probably led to acquisition of infection in this patient.
Visceral abscesses in melioidosis are often multiple, small and discrete with ‘target-like’ appearance. However, larger multiloculated splenic abscess and splenic abscess with perisplenic extension are also common.7 18
Laboratory diagnosis of melioidosis is often encountered with difficulties; the bacterium may not be readily isolated from clinical specimens or may not be properly identified even if isolated. It is often considered a contaminant or misidentified as Pseudomonas species.19 Newer advancements in diagnosis like genotypic identification, metabolomics are often beyond the reach of most laboratories in the developing world. Improved isolation of B. pseudomallei can be achieved by using the Ashdown agar medium, especially for specimens obtained from non-sterile sites. Most of the commercial systems API 20NE (bioMerieux, Marcy l’Etoile, France), Microbact 24E (Oxoid, Hampshire, UK), Vitek-1 or Vitek-2 (bioMerieux) and MicroScan WalkAway (Dade Behring, Sacramento, California, USA) can identify B. pseudomallei. The earlier version of Vitek-2 could correctly identify only 19% of confirmed cases with B. pseudomallei infection and the remaining isolates were misidentified as other non-fermenters.20 However, Vitek-1 performs much better with definite identification of the organism in almost 99% cases. Such issues have also been encountered with another automated system (Phoenix NMIC-ID4, Becton Dickinson, Sparks, Maryland, USA).21 We came across similar problems using the Vitek-2 compact system. The introduction of colorimetric identification cards to Vitek-2 instead of the fluorometric cards used originally, has improved the identification performance for non-fermenting gram-negative bacilli. Isolation of B. pseudomallei remains the ‘gold standard’ for diagnosis; and in resource-restricted settings, this perhaps is the only way to reach the diagnosis.
Splenic abscess, if left untreated is universally fatal, whereas in treated patients, the mortality varies from 12.4% to 27.6%.22 23 Patients with multiple splenic abscesses or underlying immunodeficiency are likely to have a poor prognosis and high mortality.
Therapeutic options for splenic abscesses are conservative management (appropriate antimicrobial therapy, guided percutaneous aspiration) and splenectomy, though, the optimal treatment remains unclear and treatment needs to be customised. Splenectomy, once considered the gold standard, has recently been put into question as current evidences have reported comparable treatment efficacy and mortality between conservative management and splenectomy.24 In young patients and in children, preservation of splenic function by conservative therapy has a major immunological advantage. The survival rate is high in patients treated with antimicrobial therapy alone and appropriate antibiotics should be offered first.25 Percutaneous drainage can be used as an alternative for patients with severe comorbidities or patients who are critically ill.23 Splenectomy should be reserved for patients with abscess size more than 10 cm and for those not responding to conservative treatment.24 In one series of splenic abscess that reported B. pseudomallei as the most common aetiological agent and DM as an important risk factor, majority of the cases were successfully managed with conservative treatment and splenectomy was only occasionally required.26 The most recent guideline has acknowledged the importance of debridement and drainage in managing deep-seated abscesses (abscess anywhere other than skin, lungs, bones, central nervous system or vasculature) in melioidosis to prevent recrudescence and relapse. The antibiotic protocol in such cases includes a minimum 4 weeks of intravenous treatment (intensive phase) followed by 3 months of oral therapy (eradication phase).27
Learning points.
Burkholderia pseudomallei is an unusual cause of splenic abscess. However, the diagnosis needs to be considered in an endemic area, particularly in any insulin-treated patients with diabetes having multiple splenic abscesses.
The condition is often either overlooked or misdiagnosed as infection due to other pseudomonads. The typical ‘safety pin’ appearance of the bacterium along with supportive biochemical tests and typical antimicrobial sensitivity in appropriate clinical background clinches the diagnosis.
Splenic abscess due to B. pseudomallei should be treated with appropriate antibiotics with/without splenic aspiration. Antibiotics are to be given for long duration to avoid relapse. Splenectomy needs to be reserved for those patients not responding to conservative management and having large-sized abscesses.
Acknowledgments
We sincerely thank Professor Partha Sarathi Satpathi, head of the Department of Microbiology, Midnapore Medical College and Hospital, for his help and intellectual inputs to diagnose the condition.
Footnotes
Contributors: APG, RH and PPC were involved in patient care. MC helped in laboratory diagnosis. All four authors contributed to literature search and writing the manuscript.
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.
Competing interests: None declared.
Patient consent for publication: Obtained.
Provenance and peer review: Not commissioned; externally peer reviewed.
References
- 1.Ramphal R Infections Due to Pseudomonas Species and Related Organisms : Kasper DL, Fauci AS, Hauser SL, et al., Harrison’s Principles of Internal Medicine. 19th edn New York: McGraw Hill, 2015: 1042–9. [Google Scholar]
- 2.Suputtamongkol Y, Chaowagul W, Chetchotisakd P, et al. Risk factors for melioidosis and bacteremic melioidosis. Clin Infect Dis 1999;29:408–13. 10.1086/520223 [DOI] [PubMed] [Google Scholar]
- 3.Currie BJ, Fisher DA, Howard DM, et al. Endemic melioidosis in tropical northern Australia: a 10-year prospective study and review of the literature. Clin Infect Dis 2000;31:981–6. 10.1086/318116 [DOI] [PubMed] [Google Scholar]
- 4.Kingsley PV, Leader M, Nagodawithana NS, et al. Melioidosis in Malaysia: a review of case reports. PLoS Negl Trop Dis 2016;10:e0005182. 10.1371/journal.pntd.0005182 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Meumann EM, Cheng AC, Ward L, et al. Clinical features and epidemiology of melioidosis pneumonia: results from a 21-year study and review of the literature. Clin Infect Dis 2012;54:362–9. 10.1093/cid/cir808 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Saravu K, Mukhopadhyay C, Vishwanath S, et al. Melioidosis in southern India: epidemiological and clinical profile. Southeast Asian J Trop Med Public Health 2010;41:401–9. [PubMed] [Google Scholar]
- 7.Wibulpolprasert B, Dhiensiri T. Visceral organ abscesses in melioidosis: sonographic findings. J Clin Ultrasound 1999;27:29–34. 10.1002/(SICI)1097-0096(199901)27:1<29::AID-JCU5>3.0.CO;2-C [DOI] [PubMed] [Google Scholar]
- 8.Mohan A, Manan K, Tan L-S, et al. Detection of spleen abscesses facilitates diagnosis of melioidosis in Malaysian children. Int J Infect Dis 2020;98:59–66. 10.1016/j.ijid.2020.06.025 [DOI] [PubMed] [Google Scholar]
- 9.Khosravi Y, Vellasamy KM, Mariappan V, et al. Antimicrobial susceptibility and genetic characterisation of Burkholderia pseudomallei isolated from Malaysian patients. ScientificWorldJournal 2014;2014:1–9. 10.1155/2014/132971 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Currie BJ, Ward L, Cheng AC. The epidemiology and clinical spectrum of melioidosis: 540 cases from the 20 year Darwin prospective study. PLoS Negl Trop Dis 2010;4:e900. 10.1371/journal.pntd.0000900 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Woods DE, Jones AL, Hill PJ. Interaction of insulin with Pseudomonas pseudomallei. Infect Immun 1993;61:4045–50. 10.1128/IAI.61.10.4045-4050.1993 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Simpson AJ, Newton PN, Chierakul W, et al. Insulin, and melioidosis in Thailand. Clin Infect Dis 2003;36:e71–2. [DOI] [PubMed] [Google Scholar]
- 13.Currie B Pseudomonas pseudomallei-insulin interaction. Infect Immun 1995;63:3745. 10.1128/IAI.63.9.3745-3745.1995 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Geerlings SE, Hoepelman AI. Immune dysfunction in patients with diabetes mellitus (DM). FEMS Immunol Med Microbiol 1999;26:259–65. 10.1111/j.1574-695X.1999.tb01397.x [DOI] [PubMed] [Google Scholar]
- 15.Simpson AJH, Wuthiekanun V. Interaction of insulin with Burkholderiapseudomallei may be due to a preservative. J ClinPathol 2000;53:159–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Modi KD, Gadge PV, Jain P, et al. Clinical challenges with excipients in insulin formulations and role of concentrated insulin. Int J Basic Clin Pharmacol 2019;8:821–6. 10.18203/2319-2003.ijbcp20191125 [DOI] [Google Scholar]
- 17.Katz M, Smith S, Conway L, et al. Melioidosis in a patient with type 1 diabetes mellitus on an insulin pump. Endocrinol Diabetes Metab Case Rep 2018;2018:18–62. 10.1530/EDM-18-0062 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Kunnathuparambil SG, Sathar SA, Tank DC, et al. Splenic abscess due to chronic melioidosis in a patient previously misdiagnosed as tuberculosis. Ann Gastroenterol 2013;26:77–9. [PMC free article] [PubMed] [Google Scholar]
- 19.Lau SKP, Sridhar S, Ho C-C, et al. Laboratory diagnosis of melioidosis: past, present and future. Exp Biol Med 2015;240:742–51. 10.1177/1535370215583801 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Lowe P, Engler C, Norton R. Comparison of automated and nonautomated systems for identification of Burkholderia pseudomallei. J Clin Microbiol 2002;40:4625–7. 10.1128/JCM.40.12.4625-4627.2002 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Koh TH, Yong Ng LS, Foon Ho JL, et al. Automated identification systems and Burkholderia pseudomallei. J Clin Microbiol 2003;41:41. 10.1128/JCM.41.4.1809.2003 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Lee W-S, Choi ST, Kim KK. Splenic abscess: a single institution study and review of the literature. Yonsei Med J 2011;52:288–92. 10.3349/ymj.2011.52.2.288 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Liu Y-H, Liu C-P, Lee C-M. Splenic abscesses at a tertiary medical center in Northern Taiwan. J Microbiol Immunol Infect 2014;47:104–8. 10.1016/j.jmii.2012.08.027 [DOI] [PubMed] [Google Scholar]
- 24.Alvi AR, Kulsoom S, Shamsi G. Splenic abscess: outcome and prognostic factors. J Coll Physicians Surg Pak 2008;18:740–3. doi:12.2008/JCPSP.740743 [PubMed] [Google Scholar]
- 25.Tang Y, Deng J, Zhang J, et al. Epidemiological and clinical features of melioidosis: a report of seven cases from southern inland China. Am J Trop Med Hyg 2018;98:1296–9. 10.4269/ajtmh.17-0128 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Ng CY, Leong EC, Chng HC. Ten-Year series of splenic abscesses in a general Hospital in Singapore. Ann Acad Med Singap 2008;37:749–52. [PubMed] [Google Scholar]
- 27.Sullivan RP, Marshall CS, Anstey NM, et al. 2020 review and revision of the 2015 Darwin melioidosis treatment guideline; paradigm drift not shift. PLoS Negl Trop Dis 2020;14;14:e0008659 10.1371/journal.pntd.0008659 [DOI] [PMC free article] [PubMed] [Google Scholar]



