A Young Man With Hemoptysis and Cavitary Lung Lesions

Case Presentation

A man in his 20s presented with 2 months of mild fatigue and intermittent hemoptysis of less than a tablespoon per episode. He was previously healthy and was on no medications. He denied fevers, night sweats, weight loss, wheezing, dyspnea, musculoskeletal symptoms, and rashes. He had emigrated from a South American country to the United States 3 years earlier. He worked as a groundskeeper but had no exposures to animals, mold, or dusts. He reported rare prior cigarette smoking with no history of alcohol or drug use. He was unsure whether he had received the Bacillus Calmette-Guérin vaccine.

Physical Examination Findings

The patient was well-appearing and lungs were clear to auscultation. The remainder of the physical examination was unremarkable.

Diagnostic Studies

Purified protein derivative (PPD) test was positive (15-mm induration), but QuantiFERON-TB Gold was negative. Three sputa were negative for acid-fast bacilli (AFB), and nucleic acid amplification testing for Mycobacterium tuberculosis was negative. CBC demonstrated normal WBC count with 7% eosinophils. Serum chemistry values and urinalysis were normal. Serology for coccidioides and strongyloides, histoplasma urine antigen, and HIV enzyme-linked immunosorbent assay were negative. Antinuclear, antineutrophilic cytoplasmic, and perinuclear antibodies were within normal limits. Sputum and stool were negative for ova and parasites. C-reactive protein was elevated at 6.1 mg/L (normal range, 0.1-3.0 mg/L), whereas erythrocyte sedimentation rate was normal. IgE level was elevated at 249 kU/L (normal range, 0-115 kU/L). CT scan of the chest demonstrated a 1-cm right upper lobe (RUL) cavitary lesion, bronchiectasis, and tree-in-bud nodular opacities (Fig 1). BAL of the RUL demonstrated normal flora, was AFB stain negative, and was unrevealing for actinomyces and nocardia.

Figure 1.

Right upper lobe cavitary lesion with associated bronchiectasis and tree-in-bud pattern.

Treatment for possible TB was initiated with rifampin, isoniazid, pyrazinamide, and ethambutol. Two months later, fatigue and hemoptysis had resolved, as had the cavitary lesion (Fig 2). AFB cultures remained negative. Given clinical and radiographic improvement, a diagnosis of culture-negative TB was inferred; he continued on isoniazid and rifampin.

Figure 2.

Resolution of right upper lobe cavity and improvement of bronchiectasis and tree-in-bud changes after 2 months of anti-TB therapy with rifampin, isoniazid, pyrazinamide, and ethambutol. No new pulmonary lesions were identified.

He presented 1 month later with recurrent fatigue and hemoptysis despite adherence to his TB regimen. CT scan demonstrated a new right lower lobe cavitary lesion (Fig 3A). Peripheral eosinophils rose to 12% and three sputa were again negative for AFB. CT-guided biopsy revealed eosinophilic infiltrates with a fragment of a white, smooth bordered structure (Fig 3B).

Figure 3.

A, Development of new right lower lobe thick-walled cavitary lesion with surrounding ground-glass opacities while on isoniazid and rifampin; after completing 3 months of anti-TB therapy. There were no new findings in the right upper lobe. B, Hematoxylin and eosin stain of CT-guided lung biopsy of the right lower lobe cavitary lung lesion demonstrating lung parenchyma with an eosinophilic infiltrate. A single structure with a refractile wall, roughly 30 microns in diameter was identified.

Image courtesy of Robert Homer, MD, PhD, professor of pathology, Department of Pathology, Yale School of Medicine.

What is the diagnosis?

Diagnosis: Pulmonary paragonimiasis

Discussion

Paragonimus is a fluke transmitted via consumption of undercooked crab or crayfish. After ingestion, the larvae enter the peritoneal cavity through the intestinal wall. They then migrate through the diaphragm into the lung, where the worms mature and produce eggs that are secreted in the sputa or swallowed and excreted in the feces. Although rare in the United States, Paragonimus infects approximately 20 million people worldwide. Its prevalence in South America is among the highest in the Americas. Paragonimus species are found in Asia, West Africa, and the Americas, although most individuals in North America diagnosed with paragonimiasis are immigrants from endemic regions. In the Unites States, case reports of local infection from Paragonimus kellicotti and from imported food exist. The early phase of infection, which occurs while the larvae travel from the peritoneal cavity into the chest wall, may consist of fevers, malaise, epigastric pain, and pleuritic chest pain. Peripheral eosinophilia is common during this time and chest imaging may demonstrate migratory pulmonary infiltrates, exudative pleural effusions with eosinophilic predominance, and pneumothoraces. The late phase of infection occurs after the worms settle in the lungs or extrapulmonary locations and induce inflammation and fibrosis with symptoms of recurrent hemoptysis and malaise. Imaging may demonstrate cavitary lesions, nodular infiltrates, and pleural effusions. Definitive proof of paragonimiasis is difficult to obtain because eggs are identified in sputum, stool, BAL, or tissue biopsy in only 50% of patients. Serologic testing has 92% sensitivity and 90% specificity and is available through the US Centers for Disease Control and Prevention. Paragonimiasis is easily treated with a short course of antihelminthic therapy; however, if untreated, the flukes can migrate to extrapulmonary sites including the brain, causing meningitis, encephalitis, and hemorrhage.

Because of similar symptoms, radiographic features, and geographic distribution, 50% to 70% of patients with paragonimiasis are initially diagnosed and treated for TB, even in countries where it is endemic. Culture-negative TB is diagnosed in a host with a positive PPD or QuantiFERON TB-Gold test with presence of compatible radiographic findings, negative AFB cultures, and lack of alternative diagnosis. This diagnosis is further supported by clinical and radiographic response to empiric TB treatment. Although patients with culture-negative TB typically present with a lower burden of symptoms and less extensive radiographic abnormalities compared with culture-positive patients, it is important to note that hemoptysis is reported by 20% of patients, and 12% will have cavitary lesions.

Despite significant overlap in presentation between paragonimiasis and culture-negative TB, subtle differences can provide clues to the correct diagnosis. The presence of peripheral eosinophilia and predominance of cavities in the middle to lower lung zones should raise concern for paragonimiasis, whereas TB usually involves the upper lung zones. Additionally, the radiographic abnormalities seen in paragonimiasis are often migratory and can resolve spontaneously, in contrast to TB. On pathology, even if a specific organism is not identified, the background lung changes in parasitic infection commonly show prominent eosinophilia, while tuberculosis shows necrotizing granulomas. Improved awareness of epidemiologic risk may also aid with differential diagnosis. One tool to augment this process is offered by the Global Infectious Diseases and Epidemiology Online Network, which provides an online resource generating a differential for infectious diseases endemic to a given region.

Clinical Course

This patient’s clinical course highlights the importance of clinical reasoning in the construction and revision of a differential diagnosis as new information becomes available. When making a diagnosis, clinicians often use heuristics, which are simple and general rules that allow for efficient decision-making. Heuristics are mental shortcuts that work well in most scenarios; however, they can lead to errors that are cognitive biases. In this case, cognitive biases may have prevented an awareness of a knowledge gap in the epidemiologic risk and contributed to delay of the correct diagnosis. Four common types of heuristics were evident in this case: availability, anchoring, confirmation, and framing.

Availability bias is the notion that the perceived likelihood of a diagnosis is based on exposure to similar cases and is greatly influenced by where an individual physician has lived, trained, and practiced. When encountering a recent immigrant with hemoptysis and a cavitary lung lesion, most American clinicians may place TB at the top of the differential; although logical, this represents availability bias contributing to underrecognition of paragonimiasis. As this case progressed, the initial impression of culture-negative TB remained the anchoring narrative. Anchoring bias occurs when a clinician clings to an initial hypothesis despite new contradictory information. Culture-negative TB was a reasonable initial hypothesis with clinical and radiographic improvement after 2 months of rifampin, isoniazid, pyrazinamide, and ethambutol therapy, but became less plausible as the case evolved. Although initially anchored to TB, it was important to revisit the differential diagnosis once new information became available that was incongruent with the initial hypothesis. Confirmation bias is the tendency to seek evidence to support a diagnosis rather than to accept evidence to refute it. The elevated IgE and persistent eosinophilia were minimized because they did not fit the narrative of culture-negative TB. Even after the patient developed a second cavitary lesion, confirmation bias led to greater emphasis on his apparent clinical and radiographic response to the initial 2 months of anti-TB therapy than on the incongruent laboratory findings, the lack of AFB positivity despite persistent symptoms, and the negative QuantiFERON-TB Gold. The finding of an egg fragment on the lung biopsy helped reframe the case and uncover the correct diagnosis. Framing bias is inherent to the process of presenting and formulating an assessment for any case, but both presenters and listeners must be mindful of how the narrative guides subsequent clinical assessment and strive to remain open-minded to alternative diagnoses. Presentation of his case as “a South American man with hemoptysis, cavitary RUL lesion, positive PPD, and negative AFB cultures” excludes details incorrectly considered extraneous and builds a consensus toward an erroneous diagnosis of culture-negative TB. Alternatively, framing the presentation as “a South American man with recurrent hemoptysis, migratory cavitary lesions, elevated eosinophils, and IgE with negative AFB cultures” would have painted a more complex clinical picture and invited earlier expansion of the differential diagnosis. The four heuristics present in this case are common tools used by clinicians during a diagnostic workup; however, they can lead to cognitive errors and missed diagnoses, especially in more unusual cases. Acknowledging and confronting potential biases in our clinical decision-making may enhance clinical acumen.

In light of his lung biopsy, serology testing was sent to the Centers for Disease Control and Prevention. This was positive for Paragonimus infection. After a course of praziquantel 75 mg/kg/d for 3 days, the patient demonstrated complete clinical and radiographic response.

Clinical Pearls

• 1.Paragonimus and tuberculosis have overlapping symptoms, radiographic findings, and endemic geographic regions; however, peripheral eosinophilia and lower lung zone lesions should raise the suspicion for paragonimiasis.
• 2.Although rare in the United States, Paragonimus is a highly prevalent trematode infection in Asia, West Africa, and the Americas, acquired via consumption of raw or undercooked shellfish.
• 3.Although the presence of cavitary lesions and hemoptysis does not exclude culture-negative TB, these features are more common in culture-positive disease.
• 4.The presence of new or persistent cavitary lung lesions while on appropriate anti-TB therapy should prompt consideration of alternative etiologies.
• 5.Although heuristics assist in efficient clinical decision-making, cognitive biases create vulnerability to diagnostic errors. Improved awareness of cognitive biases may enhance critical thinking and improve patient care.