CPC: CASE RECORDS OF PGI CHANDIGARH

https://doi.org/10.5005/jp-journals-10028-1667
Journal of Postgraduate Medicine, Education and Research
Volume 58 | Issue 2 | Year 2024

Unraveling the Complexity: A Clinicopathological Odyssey of Neonatal Infective Endocarditis and Its Complications

Brijdeep Singh1, Uma N Saikia2, Karthi Nallasamy3, Anmol Bhatia4, Saniya Sharma5, Sanjay Jain6

1Department of Telemedicine, Postgraduate Institute of Medical Education and Research (PGIMER), Chandigarh, India

2Department of Histopathology, Postgraduate Institute of Medical Education and Research (PGIMER), Chandigarh, India

3Department of Pediatric Medicine, Postgraduate Institute of Medical Education and Research (PGIMER), Chandigarh, India

4Department of Radiodiagnosis and Imaging, Postgraduate Institute of Medical Education and Research (PGIMER), Chandigarh, India

5Department of Pediatric Clinical Immunology and Rheumatology, Postgraduate Institute of Medical Education and Research (PGIMER), Chandigarh, India

6Department of Internal Medicine, Postgraduate Institute of Medical Education and Research (PGIMER), Chandigarh, India

Corresponding Author: Uma N Saikia, Department of Histopathology, Postgraduate Institute of Medical Education and Research (PGIMER), Chandigarh, India, Phone: +91 9876288554, e-mail: umasaikia@gmail.com

How to cite this article: Singh B, Saikia UN, Nallasamy K, et al. Unraveling the Complexity: A Clinicopathological Odyssey of Neonatal Infective Endocarditis and Its Complications. J Postgrad Med Edu Res 2024;58(2):86–95.

Source of support: Nil

Conflict of interest: None

CLINICAL HISTORY (DR KARTHI NALLASAMY)

The baby presented with a history of multiple hospitalizations since birth, fever for 3 weeks, and lethargy for 2 weeks. The previous hospitalizations were in private hospitals; hence, details are not known.

Hospitalization 1: Born term by lower segment cesarean section (previously LSCS). Birth weight of 2.5 kg. The baby cried immediately after birth; however, he developed respiratory distress soon after. He was admitted to the neonatal intensive care unit (NICU) for 11 days, received intravenous (IV) antibiotics and fluids, and was discharged with a well status. The documents and records are not available.

Hospitalization 2: After 3 days, he developed lethargy and refused to feed; hence, he was hospitalized again for 10 days. He received IV antimicrobials (teicoplanin and amphotericin B) and was discharged home thereafter.

Hospitalization 3: After a reasonably well period of 10 days at home, the child developed high-grade fever (101–103°F, 4–5 spikes/day, unremitting) associated with excessive irritability. He was treated with oral medications at home for 1 week but became progressively lethargic with poor feeding; hence, he was admitted again to a different hospital. He received broad-spectrum antibiotics (ceftriaxone, tazobactam) and a blood component transfusion. Laboratory investigations revealed anemia, elevated C-reactive protein (CRP), and leukocytosis. After 7 days of treatment, he was referred to another private hospital due to nonresponse. At the second hospital, he remained admitted for 5 days, continued on intravenous antimicrobials, and was evaluated for fever without a focus. A two-dimensional (2D) echocardiogram (ECHO) performed showed a mobile, oscillating echogenic mass 7.9 × 6.8 mm in the tricuspid valve bulging into the right atrium (RA). A similar echogenic mass 16 × 9 mm was seen in the right ventricle, bulging into the proximal right ventricular outflow tract (RVOT). There was no history of cough, fast breathing, or cyanosis. There was no history of loose stools, vomiting, or abdominal distension. There was no history of seizures or umbilical/central venous catheters during hospitalization.

With this history, the patient was referred to the Postgraduate Institute of Medical Education and Research (PGIMER) with a clinical possibility of myxoma or bacterial endocarditis of the RA.

Family History

The family was of a low socioeconomic status. The father was a daily wage worker, and the mother was a housewife with no history of previous abortions. There was one elder sibling, a 2.5-year-old girl, who was alive and healthy.

Immunization

The patient received birth doses, but there are no records regarding the Bacillus Calmette–Guérin (BCG) scar or uptake.

General Examination

The baby presented to the pediatric emergency department with tachycardia, mild tachypnea, stable hemodynamics, and an oxygen (O2) saturation of 95%. He was administered 2L of O2via nasal cannula. Anthropometric parameters (weight, length, and head circumference) revealed a deviation of –3 standard deviations. There was mild pallor; however, no cyanosis, clubbing, or edema was identified.

Systemic Examination

  • Cardiovascular system (CVS): S1, S2 normal. No murmurs.

  • Respiratory system: Breath sounds equal and vesicular. Fine crepitations were heard.

  • Abdomen: Soft, enlarged liver up to 5 cm below the right costal margin, spanning 7 cm. No splenomegaly.

  • Central nervous system (CNS): Irritable, but no motor deficits have been documented.

Hematological, Biochemical, and Microbiological Investigations

Prior to admission to PGIMER and during the hospital stay, the patient had anemia and received one transfusion during the current hospitalization. There was leukocytosis; however, platelet counts were normal. A coagulogram revealed a very highly elevated D-dimer.

Both renal and liver functions were normal. Serum electrolyte levels revealed hyperkalemia initially and hypokalemia in the later stages of the hospital stay. CRP levels were markedly elevated on two occasions. Blood pH and other acid-base parameters were reasonably normal, considering that the measurements were based on capillary and venous blood gas.

Fungal biomarkers—galactomannan index was normal or below the threshold, suggestive of Aspergillus; however, β-D-glucan was highly elevated.

Multiple blood cultures were sent during the hospital stay, suspecting infective endocarditis (IE). At least five cultures showed growth of yeast species, which was later identified to be Candida tropicalis. One of the cultures just before the terminal event grew staphylococcal species, which was identified to be coagulase-negative staphylococci. A detailed chronological overview of the laboratory investigations is provided in the figure.

Echocardiogram

An ECHO was performed on the first day of admission by a pediatric cardiologist, which showed a 12 × 10 mm vegetation on the tricuspid valve (anterior leaflet) and a 14 × 10 mm vegetation on the pulmonary valve. A good left ventricular function was obtained.

Imaging (Dr Anmol Bhatia)

Chest X-rays (10th and 13th July): The bilateral lungs were hyperinflated with heterogeneous airspace opacities suggestive of consolidation. Similar findings were noted in the preterminal X-ray performed prior to computed tomography (CT).

Computed tomography pulmonary angiography: In the caudal most sections, ill-defined hypodense soft tissue was noted in the region of the tricuspid valve, seen extending into the pulmonary valve with involvement of the main pulmonary artery, but it wasn’t seen to be extending beyond the proximal portion of the main or into the right or left pulmonary artery.

A lobulated ill-defined contrast-filled outpouching was seen arising from the left pulmonary artery at the hilum just proximal to its bifurcation, suggestive of a pseudoaneurysm. Further, in the lung windows, patchy areas of consolidation in both lungs with some cavitary lesions in the left lung, along with noncavitating nodules scattered in bilateral lung fields, were seen, suggestive of septic emboli. Overall, the possibility of disseminated bacterial infection like Staphylococcus was considered.

Contrast-enhanced computed tomography (CECT) of the brain was essentially normal, with no significant findings.

Immunology (Dr Saniya Sharma)

Lymphocyte subset analysis was performed in the pediatric immunology lab. The lymphocytes were gated on a cluster of differentiation 45 (CD45) vs low side scatter, and they constituted around 39%. The proportions of lymphocytes of all three subsets revealed that T and B lymphocytes were within normal limits and natural killer cells were slightly increased. T-cell immunophenotyping was performed, lymphocytes were gated, and proportions of CD4 T-cells were maintained. However, CD8-positive T-cells were slightly on the higher side, and the ratio was reversed. The naive and memory T-cell compartments also revealed some alterations. There was a slight reduction in CD4-positive naive T-cells, and CD8-positive naive T-cells were also reduced. The memory compartment revealed an increase in CD4 as well as CD8-positive memory T cells. Alterations in the naive and memory T cell compartment could be attributed to the ongoing infection or a possibility of an underlying T cell defect in this patient.

Additional immunological investigations were performed. The human immunodeficiency virus (HIV) enzyme-linked immunosorbent assay (ELISA) was nonreactive. Immunoglobulin levels were difficult to comment on at this age because of maternal antibodies and possibly immature antibody production in the baby at 2 months.

Other Investigations

Investigations to assess the extent of involvement and complications showed a heavy proteinuria from urine sample. Cerebrospinal fluid pleocytosis was encountered, possibly suggesting meningitis. Ocular fundus examination was normal.

Case Discussion

Course and management: Baby A was admitted with mild rapid breathing. He was given nasal O2 (L/minute) and breastfeeding on the first day. ECHO confirmed large vegetation on the right side valves. The first dose of antimicrobials (cefotaxime and gentamycin) was broadened to meropenem, vancomycin, and liposomal amphotericin to cover fungal infection. He had increasing respiratory distress on day 2 and required escalation of respiratory support, that is, high-flow O2 therapy (at 10 L/minute, FiO2—25%) and nasogastric feeding. On day 3 morning, acute worsening of respiratory distress was associated with hypoxemia and pulmonary hemorrhage (about 10 mm noted at intubation) that required invasive ventilation (positive end-expiratory pressure—7 cmH2O, FiO2—60%). A central venous catheter was placed, and a fluid bolus was administered. He underwent CT pulmonary angiography. Cardiothoracic and vascular surgery consultation was sought while continuing stabilization measures. Another episode of pulmonary bleeding (“torrential endotracheal tube bleed”) in the evening resulted in severe hypoxemia and circulatory shock. Around 10 mm/kg of packed red blood cells was transfused, and vasoactive infusions were added. With progressive deterioration, he developed ventricular tachycardia that required cardiopulmonary resuscitation and defibrillation. However, after a brief return of spontaneous circulation, he progressed to bradycardia and asystole.

Unit’s Final Clinical Diagnosis

Right-sided native valve IE with a mycotic aneurysm of the pulmonary artery and hemorrhagic shock

Clinical Discussion

The diagnosis initially seemed straightforward, considering the clinical history, available radiology, and microbiology. However, the absence of sufficient previous hospital records and a delayed presentation without the initial 2-month scores necessitate a detailed discussion regarding the case. This discussion aims to demystify any predispositions, the dissemination of infection to the involved organ system, and the events related to the patient’s demise.

A 2-month-old infant presented with fever without a focus for 3 weeks, with an ECHO revealing a 15 mm mass and multiple blood cultures growing Candida. Despite considerations like myxoma, IE emerges as the most probable diagnosis, meeting the latest Duke criteria with more than two positive blood cultures and typical findings such as vegetation and pseudoaneurysms. However, the presence of Staphylococcal (CoNS) species later in the hospital course introduces confusion. Coagulase-negative staphylococci are common in neonates, but a single organism fulfilling the major microbiological criteria should be considered the sole cause of endocarditis. In this case, a large vegetation and elevated β-D-glucan levels suggest with the culture that possibly we are dealing with only C. endocarditis. Furthermore, we are observing right-sided native valve endocarditis. The top three common organisms in the neonatal age-group include coagulase-negative staphylococci along with Candida. The microbiological, bacteremic, and fungemic profiles in our referred neonatal cohorts show CoNS as well as Candida to be equally dominant. However, a major microbiological criterion needs to be fulfilled to diagnose IE. Even if there are two organisms present, the one fulfilling the major microbiological criteria in the Duke’s criteria should be considered the sole organism causing endocarditis. Additionally, there were no clinical symptoms to suggest heart failure or congenital heart disease early in the first 2 months of life, and a reasonably good ECHO by two expert operators showed no major defects. CT angiogram also revealed no major defects or signs suggestive of paradoxical embolism. So probably we are not dealing with a structural heart disease here. It looks like a native valve endocarditis. The question arises—why right-sided endocarditis? A look at the profile of fungal endocarditis in neonatal and infant age-groups indicates that right side is the most commonly involved. Although both valvular and mural vegetations are equally common, the images suggest valvular endocarditis. A mural involvement would probably be revealed on pathology. Right ventricular involvement could also be present in this child.

The most common predisposition seen in infants or neonates with IE is a central venous catheter, which can induce trauma and serve as a contributing factor in about two-thirds of cases of fungal endocarditis. In our case, although it is mentioned twice in the file that there were no umbilical venous catheter or central venous catheter lines, this information is solely based on parental history, and the documents from previous hospitalizations are unavailable. Also, it’s important to note that about 30% of neonates can still develop IE without a central venous catheter in place. Additionally, prematurity is a significant risk factor for candidemia, typically associated with long hospitalizations, broad-spectrum antibiotics, and parenteral nutrition. However, in our index child, who was a term infant, had multiple hospitalizations, and received broad-spectrum antibiotics, the status of hyperglycemia or steroid intake is unknown, and parenteral nutrition was possibly not administered.

Considering the possibility of immune dysfunction, HIV ELISA was negative, and the child was nonneutropenic. While primary immunodeficiency diseases (PID), such as severe combined immunodeficiency (SCID) and chronic granulomatous disease (CGD), can manifest with candidiasis, their presentation differs from the observed course. SCID may be considered due to factors such as disseminated infection and unexplained initial hospitalization, but flow cytometry results were inconclusive, and there were no other organisms cultured. Also, a small thymus was observed. For CGD, candidiasis can be a manifestation, but Candida is less common compared to Aspergillus, and the age of presentation is unusual. The presentation of Candida seeding into the endocardium and dissemination does not align with the typical PID presentation. Therefore, PID seems unlikely.

Speculating on the course evolution, it’s plausible that Candida was acquired early, leading to progressive candidemia and endocarditis with subsequent embolic complications. Candida infections in neonates can occur within the 1st week, with dissemination being a risk, especially with virulent strains like C. tropicalis. Clinical evidence of dissemination in the lung is evident, and it will possibly be demonstrated in the autopsy findings as well.

Available investigations suggest a possible meningitis, likely glomerular injury, which may be immune-mediated. We are unlikely to encounter a fungal ball, but micronodules and microabscesses in the solid organs will be a possibility. As for the terminal event, pulmonary thromboembolism is a possibility, as there was progressive hypoxemia and hemorrhage. Large vegetations can embolize, but at the same time, the presentation looks like a torrential bleed, which was seen at the time of death, or just hours before death, points toward rupture of myocardial aneurysm in this baby.

In conclusion, a single organism disseminated and caused multiple features in this child. Adhering to the law of parsimony, where the simplest explanation is often the best, helps guide clinical reasoning and decision-making.

Dr Sanjay Jain: What we have seen is a very clear-cut case of neonatal IE. Now we have to discuss whether there is real immunodeficiency or there’s no immunodeficiency and what is the extent of involvement by Candida. With this, can we have the comments from the treating unit?

Dr Rakesh Pilania: We have a 2-month-old term neonate who lacks obvious external risk factors other than broad-spectrum antimicrobials with disseminated candidiasis, possibly originating from IE and then spreading to other organs. There is an absence of lymphopenia in the records, but the two primary immunodeficiencies that can present at this age are SCID and phagocytic defects. Regarding SCID, investigations show some abnormalities in CD8 T cells, which are reduced, but it’s not a full-blown picture of severe combined immunodeficiency. CGD couldn’t be ruled out, but there is a significant possibility in the presence of invasive candidiasis and significant leukocytosis—that possibility would be there. Additionally, CARD9 defects can present with invasive candidiasis.

Dr Sanjay Jain: What about the extent of involvement?

Dr Parag: Before the pathologist provides the real answers, I would like to share a few points. I’m hesitant to consider immunodeficiency because typically, both sides are involved. In this case, in the ECHO I performed, significant vegetation was observed on both the tricuspid and pulmonary valves. With multiple lung emboli causing mycotic aneurysm and a pseudoaneurysm in the left pulmonary artery, this suggests bloodstream dissemination likely due to prolonged hospitalization. The patient had multiple admissions outside PGIMER, was likely exposed to broad-spectrum antibiotics. Additionally, the lack of clarity regarding antisepsis precautions during NICU stays raises concerns. As pointed out correctly, the terminal event likely involved rupture of mycotic aneurysm, the predominant involvement appears to be on the right side. Therefore, I’m cautious about attributing the condition solely to immunodeficiency, as it seems predominantly related to the long-term hospital stay.

One more point I’d like to raise is the possibility of a combined infection with Staphylococcus. With such prolonged hospitalization, it’s unlikely that only one organism is involved, especially considering the size of the vegetation. Staphylococcus lugdunensis, rather than Staphylococcus aureus, is known for causing large vegetation and could also be a factor, even in adults.

Dr Sarang: Not only the study from PGIMER, but very large studies from developing countries and one big study from Delhi—the Delhi Neonatal Infections Study have shown that one of the biggest risk factors for Candida is actually being born in a small center and getting referred to a big hospital. Nursing care practices are appalling in many of these hospitals, where stock solutions are repeatedly used on multiple babies, and flush saline is taken out at the beginning of the nursing shift and repeatedly used from open syringes. And these have been shown to be, so that’s why anywhere between 5 and about 20% of sepsis in outborn referred babies to tertiary care centers is because of Candida. Many of these cases actually don’t have any typical immunodeficiency.

In the second hospitalization, the NICU gave amphotericin B, which is a little unusual for a small hospital. So I wonder whether it had already been detected in some previous hospitalization and was treated because within a 10-day span, amphotericin B had been started. And although, as you’re saying, there was no record of an umbilical line or a central line insertion, I wouldn’t be surprised if somebody had put it in and not kept any records of the same. So having Candida is not surprising. We see a lot of it in outbound babies. But it is this extent of right-sided involvement which is unusual. I think there must have been a central line; somebody may have attempted to go outside.

Dr Sanjay Jain: Certainly, on pathology, we’ll get a clearer picture of the extent of the disease. So, apart from the lungs and heart, we’ve observed massive proteinuria. Is it indicative of proliferative glomerulonephritis, neonatal nephrotic syndrome, or could Candida be embolizing? Although textbooks highlight IE as a significant cause, it’s important to note that right-sided endocarditis differs from left-sided cases, as Dr Parag mentioned. They tend to be indolent, challenging to diagnose, with usually negative cultures, often leading to immune-complex diseases. This might also explain the development of mycotic aneurysms, as they all require time to evolve.

Dr Karthi: Probably a comment about CARD9 deficiency. Yes, systemic candidiasis is reported, but from what I’ve seen, it’s mostly in older age-groups, not in neonates. In the second hospital, amphotericin was administered. Again, we don’t have records of culture or any other investigation, but possibly due to suspicion or a positive culture for Candida. However, it was inadequately treated, for sure, as he stayed for 10 days but likely received <10 days of treatment.

AUTOPSY FINDINGS (DR UMA NAHAR SAIKIA)

Thank you, sir, and good morning to all. Well, indeed, this patient had a candidal infection, but what it did to the baby and how it came is what I am going to demonstrate at the autopsy. I chose this case for two reasons. One, although candidal infection is very common in neonates and the pediatric age-group, the way it has disseminated to various organs, and the terminal effects that led to the demise of this patient.

This was a partial autopsy. The brain was not examined, so I do not know whether there were microabscesses in the brain. And the prosector had noted little pericardial effusion. The pleural and the pericardial cavities did not yield any excess of fluid.

Heart: The heart weighed approximately 140 gm, slightly overweight. Anteriorly, the bulk of the heart is mainly formed by the right side. Normally, the heart is two-thirds formed by the left ventricle and one-third by the right ventricle, but here, the ratio is reversed, with almost two-thirds formed by the right side and one-third by the left side. The pulmonary artery, especially at the root, is dilated. Posteriorly, the right side is much more prominent, indicating significant right ventricular hypertrophy. Upon examining the right inflow tract, the rough part of the right atrium appears normal. However, there is a vegetation on both the posterior and anterior tricuspid valve leaflets, measuring 1 × 0.5 × 0.5 cm. The vegetation is friable, soft, and necrotic with some areas of hemorrhage. Importantly, there is a large opening forming an atrial septal defect (ASD) measuring around 5 mm in diameter and is of the ostium secundum type (Fig. 1A). In the outflow tract of the heart, the pulmonary artery is narrowed at its infundibulum. One of the cusps, specifically the posterior cusp, appears rudimentary, with a linear lesion measuring 15 × 5 mm. Upon cutting open, there is appreciable narrowing of the infundibulum of the pulmonary artery with dilatation of pulmonary artery, along with a linear lesion that is a vegetation measuring 15 × 5 mm, firmly attached to the pulmonary artery. However, the vegetation on the tricuspid valve was friable and soft. There was no rupture of the pulmonary artery, hematoma, or endothelial reaction noted. The pulmonary valve essentially appears normal, with no damage to the cusp, and the root of the pulmonary artery is dilated, and the vegetation was jutting out from the tricuspid valve (Fig. 1B).

Figs 1A to H: (A) Gross photo of the right inflow tract showing a friable, soft, necrotic, focally hemorrhagic vegetation on both posterior and anterior tricuspid valve leaflets. Probe passing through atrial septal defect measuring 5 mm; (B) Outflow of the right heart showing narrowing of the pulmonary artery at the infundibulum along with linear vegetation measuring 15 by 5 mm, firmly attached; (C) Left side of the heart—opening of septal defect measuring 5–6 mm in diameter with a pin, confirming ostium secundum type of ASD. Right ventricular wall measuring 5 mm; (D) Microphotograph from vegetation on tricuspid valve showing fibrin, neutrophil-rich infiltrate, capillary proliferation with numerous colonies of Candida: hematoxylin and eosin, 400×; (E) Fungal colonies with budding pseudohyphae and spores, highlighted on PAS stain: 400×; (F) Section from the pulmonary artery showing vegetation causing ulceration of the endothelium with maintained underlying elastic tissue: hematoxylin and eosin, 400×; (G) Photomicrograph from pulmonary valve showing multinucleated giant cells, engulfing candidal spores: hematoxylin and eosin, 400×; (H) Photomicrograph from the pulmonary artery showing thickened endothelium and multiple layers of cellular proliferation: hematoxylin and eosin, 400×

Now, moving to the left side, the opening of the septal defect is also observed from the left side, measuring 5–6 mm in diameter. Therefore, this confirms the presence of an ostium secundum type of ASD in this case. The left ventricular wall measured 5 mm, similar to the measurement of the right ventricular wall, which is markedly increased for this age (Fig. 1C). The aorta and the aortic valves appear normal.

Microscopic examination: Sections examined from the vegetation on the tricuspid valve showed ulceration of the endothelium, predominantly fibrin, and neutrophil-rich infiltrate with a lot of capillary proliferation. In addition, relatively paler-looking areas were noted in between the fibrin strands. Elastin stain performed did not reveal any elastic fibers at the base of the valve. On higher magnification, colonies of Candida could be appreciated accompanied by fibrin, nuclear debris, and neutrophilic infiltrate (Fig. 1D). The fungal colonies were strongly highlighted on periodic acid–Schiff (PAS) stain, forming budding colonies of Candida. Spores and pseudohyphae of Candida could be appreciated on further higher magnification (Fig. 1E). Some of the fungal profiles focally gave the appearance of branching, raising the suspicion of an additional fungus, but the polymerase chain reaction performed did not yield anything other than the Candida. The valve in the lower half showed exuberant vascular proliferation with mixed inflammation and edema. Occasionally multinucleated giant cells and thrombi within the arterioles were noted. The PAS-Alcian blue stain was performed to appreciate the normal morphology of the valve. The spongiosa and the fibrosa were maintained. So there was no thickening of the valve tissue. There was no increase in fibrosis or any elastic tissue, as highlighted on the elastin stain though increased cellularity and vascularity were present.

Sections from the pulmonary artery and pulmonary valve revealed similar vegetation with large areas of necrosis, fibrin, and candidal colonies. On PAS stains, the fungal colonies were highlighted as bright pink. The pulmonary artery base and part of the attachment of the pulmonary valve appeared essentially normal. Exuberant fibrin deposition was noted, but a lesser inflammatory infiltrate was encountered compared to the tricuspid valve. A Masson’s trichrome-stained section revealed no increase in fibrosis in the pulmonary valve or in the pulmonary artery. A vegetation was noted causing disruption and ulceration of the normal pulmonary artery endothelium, with normally maintained underlying elastic tissue (Fig. 1F). In addition to the pseudohyphae of Candida, many spores were identified, which were not very brightly PAS-positive, but the capsule showed bright positivity. A mucicarmine stain was performed to exclude the possibility of Cryptococcus. This appearance could be attributed to the florid proliferation and rapid turnover of the candidal colonies. Many multinucleated giant cells were seen in the sections from the pulmonary valve, some of which were engulfing the Candida colonies, indicating some chronicity in the pulmonary artery, whereas the tricuspid vegetation was much fresher, rich in fibrin and neutrophilic infiltrate (Fig. 1G). The pulmonary artery showed thickened endothelium and multiple layers of cellular proliferation (Fig. 1H).

Additionally, in the normal area and the right-sided heart sections, significant inflammatory infiltrate and fibroblastic proliferation were observed, indicating that the lesion had been of some duration. No abnormality, fibrosis, or mycotic abscesses were observed in the myocardium.

Lungs: Coming to the lungs, which bore the brunt of the disease—both lungs weighed 190 gm and were heavy, diffusely consolidated, with right more affected than the left. Hemorrhagic areas interspersed with pale areas were observed, particularly affecting the upper lobes more than the lower lobes. Pleural tags were also noted. Posteriorly, on the cut surface, cavitatory lesions were visible on both the right and left sides, predominantly localized to the pulmonary arterial branches. These lesions appeared hemorrhagic in color but firm to touch. Additionally, there was diffuse hemorrhagic consolidation with multiple infarcts, along with areas of brownish discoloration in some regions. The tracheobronchial tree, however, appeared normal (Fig. 2A). Another cut surface of the lung revealed multiple whitish lesions, with thrombi observed in the pulmonary artery and its branches. The surrounding parenchyma appeared hemorrhagic and consolidated on both sides. Additionally, a relatively older appearing cavitary lesion with dilatation was also noted (Fig. 2B). A mycotic dilatation of the pulmonary artery at the hilum was noted both on the left and the right side.

Figs 2A to H: (A) Gross photograph showing diffusely consolidated lungs with hemorrhagic consolidation and cavitatory lesions on both sides, along pulmonary arterial branches with multiple infarcts; (B) Gross photograph with a large cavitatory lesion in the upper lobe; (C) Microscopy of the pulmonary artery shows fibrin, endothelial damage, and neutrophil-rich infiltrate in the lumen, admixed with nuclear debris, neutrophils, histiocytes, and occasional giant cells: hematoxilin and eosin, 100×; (D) Photomicrograph of the right hilum section with exuberant neutrophilic infiltrate and fibrin in pulmonary artery wall, with secondary fungal vasculitis having fungal colonies within the lumen: Masson’s trichrome, 100×; (E) Higher power view exhibiting endothelial damage, fungal vasculitis, and colonies of Candida within the lumen of the pulmonary artery, spilling over into the adjoining lung parenchyma: hematoxilin and eosin, 400×; (F) Photomicrograph exhibiting damage to pulmonary artery elastic lamina, suggestive of vasculitis. Candidal spores are evident on Gram’s stain, 100×; (G) Photomicrograph showing prominent intra-acinar arteries, demonstrating endothelial proliferation, medial hypertrophy, and luminal narrowing: hematoxilin and eosin, 100×; (H) Plexiform lesion having a boomerang appearance: hematoxylin and eosin, 100×. Inset showing the formation of new vessels within a major artery: EVG stain, 400×

On microscopy, sections from the pulmonary artery show extensive fibrin deposition in its wall. There is damage to the endothelium in the form of ulceration, and the lumen is filled with a neutrophil-rich infiltrate, mixed with fibrin. Under high power, it is evident that the infiltrate is predominantly composed of nuclear debris and neutrophils, with some histiocytes and very occasional giant cells (Fig. 2C). A section taken from the hilum of the right side exhibited exuberant damage with a neutrophilic infiltrate and fibrin deposition in the wall of the pulmonary artery, indicating angioinvasion by the fungus. Numerous fungal colonies were visualized within the lumen of the pulmonary artery (Fig. 2D). Upon higher power examination, endothelial damage, fungal vasculitis, and colonies of Candida within the lumen of the pulmonary artery, spilling over into the adjoining lung parenchyma, could be appreciated (Fig. 2E). PAS stain was used to highlight these fungal spores. Damage to the elastic lamina of the pulmonary artery, indicative of fungal vasculitis, was demonstrated on the Gram stain, compensating for the Elastica van Gieson (EVG) stain, which did not work due to technical reasons. Candidal spores were also highlighted by the Gram stain (Fig. 2F). Additionally, other vessels showed recanalizing thrombi, but no fungus was seen within them. Furthermore, other pulmonary artery branches exhibited damage to the elastic lamina and the presence of fresh fibrin in the lumen. The veins were also affected by this process, showing fibrin deposition in their walls indicative of endothelitis. This was highlighted in red by Martius Scarlet Blue stain. The areas that grossly appeared yellowish and paler showed microabscesses in the form of neutrophilic infiltrate, cellular debris, and collections of candidal colonies in the center.

Given the prominent ASD in this patient, the lung vasculature bore the brunt. Multiple sections exhibited prominent intra-acinar arteries, demonstrating endothelial proliferation, medial hypertrophy, and fibrosis around the vessels (Fig. 2G). Additionally, the interalveolar septae were broadened, showing diffuse pulmonary capillary hemangiomatosis attributable to the back pressure due to the ASD and left-to-right shunt. Many prominent intra-acinar arteries, highlighted by the EVG stain, were seen, which are usually the first sign of pulmonary arterial hypertension. Reticulin stain highlighted capillary proliferation within the alveolar septae. Formation of plexiform lesions having a boomerang appearance (Fig. 2H) and aneurysmal outpouchings, some containing fibrin thrombi, in the vicinity of the bronchial tree were also noted, consolidating the evidence of pulmonary arterial hypertension. One of the photomicrographs exhibited a plexiform lesion, indicating long-standing pulmonary arterial hypertension. This was evidenced by the formation of new vessels within a major artery, highlighted by the EVG (Fig. 2H). These vessels bore an antler horn appearance, and they also contained fibrin thrombi within the lumen. In addition, hyaline membrane formation was highlighted on the PAS stain.

The liver weighed 190 gm and was slightly enlarged, felt slightly greasy, and appeared pale, with a hemorrhagic lesion identified at the center of both the right and left lobes (Fig. 3A). Microscopically, the architecture of the portal tracts and the lobular architecture appeared normal, but the sinusoids were extremely dilated and congested. In comparison to a control from the liver of a healthy 2-month-old, the portal tracts appeared expanded but regular (Fig. 3B). The central vein exhibited perivenular sclerosis with outpouchings and multiple portal venous radicals at its periphery. Reticulin stain highlighted sinusoidal dilatation and sclerosis around the central vein, suggesting congestive hepatopathy. No elastic tissue was visualized in the sinusoids even at high power. Focally, a raquet-like appearance formed by aneurysmal outpouching around the central veins within the liver was noted, but no fibrin deposition was seen in the sinusoids (Fig. 3C). Additionally, there was focal increase in iron pigment within the hepatocytes and Kupffer cells. The portal tracts exhibited proliferation and hyperplastic hepatic arteries, along with opening up of the portal venous radicals, highlighted on Masson’s trichrome stain. Elastic stain highlighted the prominent hepatic arteries. Normally, the caliber of the hepatic artery should be almost the same as that of the bile duct. However, in this case, the hepatic artery appeared significantly more hyperplastic than expected (Fig. 3D). The immunohistochemistry (IHC) for CK7 immunostain, which usually stains bile ducts, showed arterioles much larger than the CK7-stained bile ducts. Sections from the hemorrhagic area showed extensive sinusoidal dilatation and congestion, resembling peliosis hepatis (Figs 3E to H). The hepatocytes did not reveal any increase in hemosiderin. Extramedullary hematopoiesis was evident in the sinusoids and portal tracts, highlighted by a myeloperoxidase (MPO) stain of a brown color.

Figs 3A to H: (A) Gross of the liver showing slight enlargement, with a hemorrhagic lesion in the center of the right and left lobes; (B) Photomicrograph showing preserved lobular architecture with dilatation and congestion of sinusoids in comparison to a control liver matched for age: hematoxylin and eosin, 100×; (C) Photomicrograph displaying perivenular sclerosis with outpouchings, forming a racquet-like appearance. Sinusoidal dilatation and sclerosis are also evident: reticulin, 100×; (D) Photomicrograph showing a portal tract exhibiting hyperplastic hepatic arteries: Verhoeff’s elastic stain, 400×; (E) Photomicrograph from the hemorrhagic area showing extensive sinusoidal dilatation and congestion, resembling peliosis hepatis: hematoxylin and eosin, 100×; (F) Gross of the spleen exhibiting follicular prominence; (G) Photomicrograph from the spleen showing intact white pulp and a cellular infiltrate within the red pulp: hematoxylin and eosin, 100×; (H) High power from the spleen showing extramedullary hematopoiesis: hematoxylin and eosin, 400×. The presence of megakaryocytes highlighted on the MPO stain (inset)

The spleen weighed 30 gm and exhibited follicular prominence on gross examination. Microscopically, the white pulp appeared intact and the red pulp showed a cellular infiltrate, identified as extramedullary hematopoiesis. This was confirmed by the presence of megakaryocytes, highlighted by MPO stain.

The gastrointestinal tract was essentially unremarkable. No focal lesions or ulcerations were observed upon gross examination (Fig. 4A). Microscopy revealed only lymphoid hyperplasia in the small intestine, (Fig. 4B) while the large intestine appeared normal. Sections from the pancreas showed prominence of Islet cells, in regions that show the least number of Islet cells as scattered small collections (Fig. 4C). These cells were highlighted by the Insulin immunostain, confirming it to be Nesidioblastosis, likely reactive in this neonate (Fig. 4D).

Figs 4A to H: (A) Gross image of the small intestine which is largely unremarkable; (B) Photomicrograph from the small intestine showing lymphoid hyperplasia: hematoxylin and eosin, 100X; (C) Photomicrograph from the pancreas showing an increase in islet cells: hematoxylin and eosin, 400×; (D) Highlighted on: insulin stain, 400×; (E) Gross image swollen kidneys and medullary congestion; (F) Photomicrograph showing neoglomerulogenesis, with tiny glomeruli: hematoxylin and eosin, 100×; (G) Photomicrograph showing arterioles in the kidney exhibiting vascular changes: medial hypertrophy, intimal proliferation, and the presence of interstitial infiltrate, indicating extramedullary hematopoiesis: hematoxylin and eosin, 100×; (H) Photomicrograph from the adrenals showing hemorrhage in the adrenocortical and medullary junction: hematoxylin and eosin, 400×

Both kidneys were swollen but did not exhibit any focal lesions. The renal arteries were patent. On the cut surface, only medullary congestion was observed, with no focal lesions (Fig. 4E). Microscopic analysis revealed neoglomerulogenesis, characterized by the presence of tiny glomeruli. Other glomeruli appeared normal (Fig. 4F). At higher magnification, prominent podocyte proliferation was seen. There was no evidence of immune complex glomerulonephritis. Additionally, the arterioles in the kidney exhibited vascular changes such as medial hypertrophy and intimal proliferation. Furthermore, there was some infiltrate within the interstitium, attributed to extramedullary hematopoiesis (Fig. 4G). Occasionally, tubules exhibited casts, suggesting probable terminal acute kidney injury.

The adrenals also exhibited hemorrhage, particularly evident in the adrenocortical and medullary junction, as highlighted on high power (Fig. 4H).

The thymus appeared small in size, especially when compared to control thymus at this age. However, it contained adequate representation of all lymphoid cells. IHC staining on thymus encountered technical issues, but there was still a clear distinction between the cortex and medulla, with identifiable Hassall’s corpuscles.

The bone marrow was also normal for the age, with representation of all three hematopoietic elements.

FINAL AUTOPSY DIAGNOSIS (PM 31522)

During the clinical discussion, we focused on the prevalent issue of candidemia or candidal IE. However, we overlooked the fact that the patient had an ASD, which predisposed him to this condition. Additionally, multiple infusions and extended hospital stays outside may have further increased the child’s susceptibility to fungal endocarditis. This 2-month-old infant presented with IE caused by C. tropicalis, as confirmed by various blood cultures. The infection affected the normal tricuspid valve and pulmonary artery, with an uncomplicated ostium secundum type of ASD.

In literature, fungal endocarditis, specifically candidal IE, accounts for only 1.5–4 cases per 10 million children. Even among children, candidal IE comprises only 1.1% of all fungal endocarditis. These findings align with the established risk factors discussed during clinical discussion (Fig. 5).

Fig. 5: Detailed overview of laboratory investigations performed during hospital stay

FINAL DISCUSSION

Dr Parag Barwad: My take on ASD and PAH will be a little reversed from what we are thinking. We are dealing with a neonate and not a grown-up child, so PAH at 2.5 months of age is not possible. So it is actually a nonregression of a PAH in a neonate. The two-thirds to one-third physiology of the left-to-right ventricle starts when the baby reaches the age of a few months. The intrauterine circulation is right ventricle-dependent, and the right ventricle is hypertrophied. As soon as the baby starts getting his first breath, the PAH regresses slowly, and the regression takes a few weeks to a few months. So the right ventricle is still dominant, and the plexiform lesions in the lungs are actually a feature of a prominent PAH that’s already existent in the baby in the neonatal period. So this PAH isn’t attributed to ASD but to the inherent condition of the baby’s lungs. The plexiform lesions slowly regress over a period of time as the right ventricular pressure goes down as the baby’s age increases. Regarding the ASD, if the margins are clearly defined, it can be labeled as an ASD; however, a 5–6 mm defect in the interatrial septum, which is a bit tangential and doesn’t allow for a direct probe passage, is likely a persistent foramen ovale (PFO) which is seen in up to 50% of the babies at birth. The other important thing is that the baby immediately after birth had an infectious endocarditis impacting the right side of the heart leading to tricuspid and pulmonary regurgitation. Consequently, the right ventricle increases in size over a period of time due to this effect.

Dr Sarang: I am not sure what Dr Parag meant by “we don’t get to see PAH.” PAH is a well-known entity. It cannot be solely attributed to the ASD, but there could be other reasons for having persistent pulmonary arterial hypertension. The ASD doesn’t typically generate such high-pressure jets or velocities that it could be attributed as the reason for having endocarditis. While the ASD could have been an incidental finding, we still cannot directly link the endocarditis to the ASD.

Dr Harsimran: We have also seen patients who have ASD, but candidemia occurs postsurgical exploration and not as such having ASD as a risk factor for candidemia.

Dr Karthi: Infective endocarditis is relatively rare in infants even with congenital heart disease with turbulent circulation, and ASD is the least turbulent one. The most common reason for IE in a young infant is usually postprocedural in congenital heart disease rather than congenital heart disease itself.

Dr Uma: First of all, taking up Dr Parag’s question, I totally agree with you, but ASD has definitely contributed. When I compared the lungs of a 2-month-old baby with this case, we did not find these kinds of plexiform lesions in the normal 2-month-old baby. I studied the control to compare as I had anticipated these questions. Besides, I showed the control of the liver, showing changes in the hepatic arteries, veins, and thymus. So there was an element of ASD which contributed to the regressive PAH. These plexiform lesions are not typically observed in an otherwise normal 2.5-month baby. We do get hypertrophied arteries in preterm neonates as well as in infants up to 2–4 months of age. The ASD is not small; it measures 5–6 mm and may have contributed to the nonregression of PAH. The patient had multiple admissions outside which may have predisposed him to candidemia. This is not PFO but a clear-cut ASD as it is visible from the left side, through which a probe could be passed and have clear margins. In PFO, one has to stretch the septum to appreciate the opening; however, here, the opening is visible to the naked eye from both sides. So it is a clear-cut ASD of the ostium secondum type. I am not saying that ASD is the cause of IE, but I am reiterating the fact that the baby presented with respiratory distress in the first go. Why was the patient having respiratory distress? The ASD complicated the situation and candidemia came much later.

Dr Sanjay Jain: This will continue to haunt us, but the ECHO has been performed by one of the certified fetal medicine experts, and his report states very clearly that the interatrial and interventricular septum is intact, so I don’t know why there is a discrepancy between the two findings.1-5

COMMENTARY

This clinicopathological case presents a 2-month-old neonate with a history of recurrent hospitalizations, fever, and lethargy. Complicating the clinical picture was an ostium secundum ASD which was not detected on ECHO and predisposed the patient to septic embolization to the lungs and multi-organ involvement secondary to congestion. Despite challenges posed by incomplete medical records and delayed presentation, a comprehensive analysis unraveled C. tropicalis endocarditis affecting the native tricuspid valve and pulmonary artery, alongside pulmonary arterial hypertension. This case underscores the intricate interplay between predisposing factors such as prolonged hospitalization and broad-spectrum antibiotics and the dissemination of infection to the lungs. Notably, ECHO failed to detect the ASD, emphasizing the importance of clinical acumen and advanced imaging modalities. Autopsy findings revealed extensive cardiovascular involvement, shedding light on the complications of neonatal IE in a predisposed setting. This clinical journey highlights the significance of considering common explanations while remaining vigilant for atypical presentations and complications, thereby guiding clinical decision-making and management strategies.

REFERENCES

1. Kumar J, Soni PK, Angrup A, et al. Antimicrobial resistance patterns among neonates referred to pediatric emergency in North India: a prospective cohort study. Pediatr Infect Dis J 2023;42(11):1007–1011. DOI: 10.1097/INF.0000000000004056

2. Fowler VG, Durack DT, Selton-Suty C, et al. The 2023 Duke-International Society for cardiovascular infectious diseases criteria for infective endocarditis: updating the modified Duke criteria. Clin Infect Dis 2023;77(8):1222. 2023; 77 (4): 518– 526. DOI: 10.1093/cid/ciad271

3. Murphy CR, Teoh Z, Whitehurst D, et al. Disseminated disease after candidemia in children and young adults: epidemiology, diagnostic evaluation and risk factors. Pediatr Infect Dis J 2024;43(4):328–332. DOI: 10.1097/INF.0000000000004212

4. Kara A, Devrin L, Mese T, et al. The frequency of infective endocarditis in Candida bloodstream infections: a retrospective study in a child hospital. Braz J Cardiovasc Surg 2018;33(1):54–58. DOI: 10.21470/1678-9741-2017-0049

5. Puig-Asensio M, Padilla B, Garnacho-Montero J, et al. Epidemiology and predictive factors for early and late mortality in Candida bloodstream infections: a population-based surveillance in Spain. ClinMicrobiol Infect 2014;20(4):O245–O254. DOI: 10.1111/1469-0691.12380

________________________
© The Author(s). 2024 Open Access. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (https://creativecommons.org/licenses/by-nc/4.0/), which permits unrestricted use, distribution, and non-commercial reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.