Two Systemic Lupus Erythematosus Patients With Severe Pleurisy: Similar Presentations, Different Causes


Address correspondence to Kyriakos A. Kirou, MD, Mary Kirkland Center for Lupus Research, Hospital for Special Surgery, 535 East 70th Street, New York, NY 10021. E-mail:


Case 1

History of the present illness

A 25-year-old woman was diagnosed with systemic lupus erythematosus (SLE) 2 years prior to admission, after she presented with fever, malaise, lymphadenopathy, arthritis, myalgias, malar rash, and positive serologies for antinuclear antibody (ANA) and anti–double-stranded DNA (anti-dsDNA). Additional serology tests were negative for anti-Ro/SSA, anti-La/SSB, RNP, and Sm antibodies, as well as antiphospholipid antibodies (aPL). She was treated with intermittent courses of glucocorticoids (GC) for disease flares. Eight months before admission, she had a severe flare that manifested with high fevers, weight loss, delirium, spastic bladder, and non-nephrotic proteinuria. She was ultimately diagnosed with gray matter transverse myelitis of the conus medullaris and lupus nephritis. She was treated aggressively with pulse GC and intravenous (IV) cyclophosphamide (CYC), which led to resolution of all symptoms. After completion of a 6-month course of IV CYC, she was maintained on 20 mg of prednisone. One month prior to admission, she was admitted to another hospital with left-sided pleurisy, dyspnea, and high fever. A chest computed tomography (CT) scan excluded pulmonary embolism, but showed bilateral lung lower lobe consolidations with pleural effusions (left larger than right) and a small fluid collection adjacent to her spleen. Her blood cultures grew Salmonella. Her symptoms improved with cefepime and high-dose GC, and she was discharged home 2 weeks later on 60 mg of daily prednisone and oral levofloxacin for a total of 3 weeks of antibiotics. In addition, she was discharged on trimethoprim/sulfamethoxazole (TMP/SMX) 160/800 mg 3 times weekly for Pneumocystis jiroveci pneumonia (PCP) prophylaxis. One week before her current admission, she was started on 1,000 mg of mycophenolate mofetil (MMF) daily. She then presented to our hospital with acute left-sided pleuritic chest pain, dyspnea, and arthralgias. She did not have fever or cough.

Medical history

She had a history of mild asthma and hypothyroidism.

Family and social history

Her family history was unremarkable except for a maternal aunt who also had SLE. There was no history of smoking, alcohol, or illicit drug use.


Her home medications included prednisone 40 mg daily, MMF 500 mg twice daily, hydroxychloroquine (HCQ) 200 mg daily, TMP/SMX 160/800 mg 3 times weekly, esomeprazole 40 mg, and levothyroxine 50 μg daily. She was not taking any supplements.

Review of systems

She denied fevers, chills, cough, dyspnea, leg swelling, joint pain, skin rash, or oral ulcers.

Physical examination

She appeared to have Cushingoid features, but was afebrile and not hypoxic. Her blood pressure was 140/90 mm Hg and her heart rate was 90 beats per minute. She had decreased breath sounds over her left lower lung field. She had a normal cardiac examination and no hepatosplenomegaly, lymphadenopathy, rashes, or ulcers. She had a normal neurologic examination and no joint swelling or tenderness.

Laboratory and radiologic evaluation with hospital course

Her initial blood tests are shown in Table 1. She had non-nephrotic proteinuria and microscopic hematuria without a rise in her baseline creatinine level. A kidney biopsy was not performed because her nephritis appeared mild and the results were unlikely to change her management. CT imaging of the chest, abdomen, and pelvis again excluded pulmonary embolism, but showed a left-sided pleural effusion with adjacent lung consolidation and a parasplenic fluid collection. Her pleural fluid was sampled and showed an exudative process with a neutrophilic inflammatory response (Table 1) without organisms on Gram stain or culture. A transthoracic echocardiogram was normal. Her GC dose was increased to 16 mg of IV methylprednisolone (IVMP) every 8 hours and she received an additional 500 mg dose of IVMP. MMF was continued at the same dose. IV oxacillin was initiated after 1 of 2 sets of blood cultures grew Staphylococcus warneri. Despite these interventions, her symptoms did not improve. Bronchoscopy was performed 9 days after admission and bronchoalveolar lavage (BAL) yielded nonbloody, cellular, frothy fluid. The fluid was negative for bacterial, fungal, and mycobacterial pathogens, but was positive for influenza A by polymerase chain reaction (PCR). Oseltamivir was prescribed for a 5-day course. Transbronchial biopsies (TBBX) were unable to be performed because of hypoxia during the procedure. Her abdominal fluid was sampled on day 14 by CT guidance to exclude an abscess and showed serosanguinous fluid, consistent with serositis. She had a repeat chest CT scan 16 days after admission that showed loculations within a moderate left pleural effusion and an unchanged left lower lobe consolidation despite high-dose GC, IV oxacillin, and oseltamivir (Figure 1A). The patient underwent video-assisted thoracoscopic surgery (VATS) 21 days after admission for a more definitive diagnosis.

Table 1. Laboratory values at initial presentationa
 Case 1Case 2Normal values
  1. aWBC = white blood cell; LDH = lactate dehydrogenase; ESR = erythrocyte sedimentation rate; anti-dsDNA = anti–double-stranded DNA; RBC = red blood cell; hpf = high-power field.
WBC count, × 103/μl11.59.83.5–10.7
Absolute lymphocyte count, × 103/μl0.30.20.9–5.2
Hemoglobin, gm/dl9.711.311.5–16
Platelet count, × 103/μl275160160–400
Creatinine, mg/dl1.31.620.44–1.03
Albumin, mg/dl2.73.33.5–4.8
Globulin, mg/dl2.91.81.8–3.3
LDH, IU/liter264209105–333
ESR, mm/hour86464–20
Complement C3, mg/dl7110979–152
Complement C4, mg/dl302916–38
Urine protein:creatinine ratio, gm1.780.76<0.2
Urine RBC count, per hpf17220–4
Urine WBC count, per hpf850–5
Pleural fluid   
LDH, IU/liter274280 
Total protein, mg/dl3.22.9 
Glucose, mg/dl107120 
RBC count, per μl30,0001,530 
WBC count, per μl21,0002,850 
Neutrophils, %8181 
CytologyReactive mesothelial cells  
Figure 1.

A, Consolidation with loculated effusion in case 1. B, Dense nodular consolidations within larger consolidation of the left lower lobe in case 2. Smaller nodules were seen near the hilum and subpleurally. Associated pleural and pericardial effusions are shown.

Case 2

History of the present illness

A 23-year-old woman was diagnosed with SLE 10 months before admission, after she presented with high fever, malaise, and alopecia. She was found to have pancytopenia, a pericardial effusion, and International Society of Nephrology/Renal Pathology Society class IV-G (A) lupus nephritis based on renal biopsy. Her serologic tests showed low complement levels, as well as antibodies to dsDNA, SSA, SSB, RNP, and Sm. aPL tests were negative. She improved after treatment with high-dose GC and one dose of IV CYC, and was started on MMF maintenance therapy. Four months before admission, while receiving prednisone 15 mg daily, she developed autoimmune hemolytic anemia. Her prednisone dosage was increased to 50 mg/day and HCQ was added to her regimen. Eight weeks before admission, she developed a new severe left-sided pleuritic chest pain that prompted admission at another hospital. She was noted to have a peripheral mass-like left lower lobe pulmonary infiltrate on a chest CT scan. She underwent a CT-guided needle biopsy that revealed “organizing pneumonia with diffuse alveolar damage.” She was treated with a short course of IV antibiotics, her prednisone dosage was increased from 35 mg to 50 mg/day, and MMF was continued. One week after discharge, her chest pain worsened and she was readmitted and treated with IV vancomycin and piperacillin tazobactam, followed by oral levofloxacin, for a total of 10 days of antibiotics. After initial improvement, her pleuritic chest pain returned and she developed a new dry cough, prompting her first admission to our hospital. Her chest CT scan showed a left lower lobe and inferior lingular lung consolidation with additional foci of more nodular consolidations (increased since 5 weeks earlier). In addition, there were right upper lobe scattered ground-glass densities and a small loculated left-sided pleural effusion that was not large enough to be sampled (Figure 1B). She was started empirically on IV vancomycin and piperacillin tazobactam, TMP/SMX was initiated for PCP prophylaxis, and her prednisone (50 mg/day), MMF, and HCQ doses were unchanged. Bronchoscopy with BAL was performed and yielded nonbloody, frothy fluid. Lavage studies revealed rhinovirus by PCR and 16,000 copies/ml of cytomegalovirus (CMV) DNA. Pathology from a fluoroscopically guided TBBX showed lymphocytic infiltration of the interstitium. There was also focal fibrin deposition within septal and alveolar spaces. On direct immunofluorescence (DIF), there was IgG, IgA, IgM, C3, C1q, fibrinogen, and C5b–9 granular staining within the septa and septal capillaries. There were no intracellular inclusions to suggest CMV infection. The pathology findings were interpreted as interstitial pneumonitis with significant septal capillary injury due to SLE. Blood and sputum cultures, as well as bacterial, fungal, Pneumocystis, and mycobacterial stains and cultures of the BAL fluid and biopsy material, were negative. In addition, Histoplasma and Legionella urinary antigens, galactomannan and cryptococcal serum antigens, and QuantiFeron-TB Gold testing were negative. Because of persistent pleurisy, she was treated with 1,000 mg of IVMP daily for 3 days, 750 mg of IV CYC for 1 day, and 1,000 mg of IV rituximab for 1 day, and was discharged the next day. She was also placed on valganciclovir for the CMV detected on BAL fluid. The day after discharge, she developed rapid deterioration of her left chest pain and dyspnea and was readmitted to our hospital.

Medical history

She had no significant medical history except for her SLE and complications of the disease, including chronic kidney disease with a baseline creatinine level of 1.6 mg/dl, autoimmune hemolytic anemia, lymphopenia, and arthralgias.

Family and social history

There was no family history of SLE or other autoimmune diseases. She had no history of smoking, alcohol, or illicit drug use.


Her home medications included prednisone 40 mg daily, HCQ 200 mg daily, TMP/SMX 80/400 mg daily, as well as carvedilol 12.5 mg twice daily, simvastatin 20 mg once daily, darbepoetin alfa, omeprazole 40 mg once daily, risedronate, calcium, and vitamin D.

Review of systems

She denied fevers, chills, or night sweats. Her weight was stable. She experienced throat pain and was found to have thrush. She had no rash, arthritis, or lymphadenopathy.

Physical examination

She appeared to have Cushingoid features and seemed acutely ill. Her blood pressure was 110/60 mm Hg and her heart rate was 115 beats per minute and regular. She was afebrile, but hypoxic to 85% on room air. She had decreased breath sounds and a rub at the left lung base. She had a normal cardiac and abdominal examination. She had no rashes except for some acne due to GC. She was neurologically intact and had no signs of synovitis on joint examination.

Laboratory and radiographic evaluation with hospital course

Laboratory tests upon her initial admission are shown in Table 1. She had anemia with mild renal insufficiency (at her baseline). She had mild proteinuria and microscopic hematuria. Her blood cultures were negative. A chest radiograph showed near complete opacification of the left hemithorax with a rightward mediastinal shift (Figure 2). A chest CT scan showed a large partially loculated left pleural effusion with near complete collapse of her left lung, a small right pleural effusion, and small to moderate pericardial effusion. She was started empirically on IV piperacillin tazobactam, her prednisone dosage was decreased to 20 mg/day, and TMP/SMX prophylaxis and HCQ were continued. A chest tube was placed to drain her enlarging pleural effusion. Fluid analysis showed an exudative process, normal complement levels, and a negative ANA. No lupus erythematosus cells were found. Pleural fluid Gram, potassium hydroxide, and acid-fast stains were negative and blood and urine cultures were negative. CMV from plasma showed 201 copies of DNA per ml. Given the size of the effusion, progressive hypoxia, and lack of a diagnosis, VATS was performed 6 days after admission for decortication and biopsies of the lung and pleura.

Figure 2.

Two days before (A) and 3 days after (B) pulse glucocorticoid therapy in case 2.

Case summaries

Both cases describe patients with severe SLE requiring heavy immunosuppression who developed severe pleurisy while receiving high doses of GC and had concurrent loculated pleural effusions with lung consolidation. Pleural fluid analyses revealed exudative and neutrophilic inflammatory responses without evidence of empyema. Given their immunosuppression, there was concern for infection in both cases. Case 1 had just recovered from a systemic infection with Salmonella, had evidence of influenza infection in her BAL fluid, had one positive blood culture for coagulase-negative staphylococci, and did not respond to antiinfluenza and antistaphylococcal therapies. Case 2 had a more chronic course of pleurisy but an acute exacerbation after pulse GC therapy. In addition, case 2 had nodular consolidations on imaging that were of concern for infection and evidence of CMV reactivation in BAL fluid and plasma. Both cases had evidence for active SLE with pleurisy, as well as active lupus nephritis with proteinuria and active urinary sediment. In addition, case 2 had evidence for interstitial pneumonitis with septal capillary injury and immune deposits that were consistent with lupus pneumonitis.



Both patients had severe SLE and presented with painful pleural effusions and lung consolidation while being severely immunosuppressed. The foremost differential diagnosis in this clinical context is infection versus active lupus. Pulmonary infections are common in patients with SLE and should be a primary consideration in an SLE patient presenting with respiratory symptoms. Patients with SLE may be profoundly immunocompromised because of immunosuppressive medications and immunodysregulation associated with their underlying disease [1]. Both patients had been receiving prolonged courses of high-dose GC and also had received other immunosuppressive agents such as IV CYC and MMF. They were also severely lymphopenic (Table 1), likely related to GC and their underlying disease. Their immunocompromised states were highlighted by a recent history of salmonellosis in case 1 and CMV reactivation in case 2. Immunocompromised patients with SLE are at a high risk of pulmonary infections due to common bacteria, such as Staphylococcus aureus and Streptococcus pneumoniae, but also may be at risk for opportunistic infections such as Mycobacterium tuberculosis, Nocardia, fungi (including Cryptococcus neoformans and Aspergillus), and P jiroveci (2–7). It is also important to note that fever may be masked in patients receiving high doses of GC, and therefore the absence of fever does not exclude infection.

SLE pleurisy

Pleurisy is the most common pulmonary manifestation of SLE and occurs in 30–50% of patients [7]. The presentation varies and may consist of pleuritic chest pain alone without an effusion. Not infrequently, pleurisy is associated with pericarditis (pleuropericarditis), which can also present isolated. SLE pleurisy is thought to be caused by immune complex deposition in vessels with subsequent complement activation and direct binding of anti-dsDNA antibodies to the mesothelium [8-10]. Although rare, effusions can be unilateral [10, 11]. Pleural fluid can be lymphocytic or neutrophilic and typically has a pH >7.35, as in both of our cases [12]. Complement, ANA, and anti-dsDNA levels can be measured in pleural fluid and are helpful in assessing the role of SLE as a cause of pleural effusion [13, 14]. Pleural fluid ANA levels are nonspecific, since they may also be elevated in malignancy, but they are sensitive. Therefore, the absence of pleural fluid ANA makes it unlikely that SLE is the cause of a pleural effusion [15, 16]. On the other hand, the presence of either anti-dsDNA antibodies or lupus erythematosus cells (neutrophils with ingested nuclear debris appearing as a homogenous round inclusion) are diagnostic of SLE, but are not sensitive [17, 18]. In case 2, the pleural fluid was negative for ANA and lupus erythematosus cells, providing evidence against SLE as the primary etiology of the effusion. These tests were not performed in case 1. The differential diagnosis of pleurisy in SLE is shown in Table 2.

Table 2. Differential diagnosis of pleuritic chest pain with or without pleural effusion in patients with SLEa
  1. aEffusions without pain may be due to nephrotic syndrome, hypoalbuminemia, heart failure, atelectasis, trapped lung (due to previous severe SLE pleurisy), etc. SLE = systemic lupus erythematosus.
1. SLE related
Lupus serositis (pleurisy or pericarditis)
Acute lupus pneumonitis
Shrinking lung syndrome
2. Musculoskeletal pain in the rib cage, i.e., costochondritis
3. Pulmonary embolism/infarction
4. Infection related
Bacterial pneumonia (empyema or uncomplicated parapneumonic effusion)
Atypical pneumonia (viral, Mycoplasma pneumoniae, Legionella species, Chlamydophila pneumoniae)
Pleural tuberculosis
Fungal disease (aspergillosis, histoplasmosis, cryptococcosis, pneumocystis)
Subphrenic abscess
5. Drug induced, including drug-induced lupus, i.e., due to hydralazine, procainamide, etc.
6. Malignant or paramalignant effusions (non-Hodgkin's lymphoma, primary lung cancer, etc.)
7. Selected miscellaneous causes
Uremic pleurisy
Post–cardiac injury syndrome
Ascites (it may represent serositis due to SLE)
Pancreatitis (acute or chronic)

Acute lupus pneumonitis (ALP)

Other acute pulmonary manifestations of SLE include ALP and diffuse alveolar hemorrhage (DAH). ALP is relatively uncommon, occurring in 1–12% of patients with SLE [19-23]. The clinical presentation includes cough, pleurisy, dyspnea, hypoxia, and patchy infiltrates, often with bilateral consolidations that resemble “bronchopneumonia” [20, 21, 23]. Given the indistinguishable clinical and radiographic presentation from infectious etiologies of pneumonia, establishing the diagnosis requires ruling out infectious processes, often necessitating bronchoscopy. The histopathology of ALP often shows nonspecific findings, such as diffuse alveolar damage with hyaline membranes and interstitial edema [4, 19]. The presence of vasculitis and hematoxylin bodies with pleuritis is thought to be more specific for SLE-induced disease [4, 19]. DAH is the most severe acute pulmonary manifestation of SLE. These patients typically present acutely, as in ALP, and often present with hemoptysis and a precipitous drop in hemoglobin. Imaging shows bilateral alveolar infiltrates and sequential BAL fluid reveals progressively more hemorrhagic lavage aliquots. Cytology is notable for hemosiderin-laden macrophages. Our patients' presentations and BAL fluid did not suggest DAH, but they could have been consistent with ALP. Nevertheless, the largely unilateral lung involvement and lack of significant hypoxia provided evidence against ALP.

Chronic interstitial disease

In addition to these acute pulmonary manifestations, SLE is also associated with chronic interstitial lung disease, which is seen in 3–13% of patients and typically presents with dry cough and dyspnea on exertion [7]. Radiologic findings include ground-glass and reticular infiltrates as well as honeycombing. Associated histologic patterns include nonspecific interstitial pneumonia, usual interstitial pneumonia and, less often, lymphocytic interstitial pneumonia. In SLE, this entity is associated with anti-Ro antibodies [24-26]. Magro et al have found evidence for microvascular injury and IgG, IgA, and/or complement deposition in septal capillaries in patients with chronic interstitial lung disease in the setting of various collagen vascular diseases, including anti-Ro–positive SLE [25, 26]. The same authors also found reactivity in sera from these patients against endothelial cells and suggested a possible pathogenic link between this antibody-mediated (Gell and Coombs type II immune reaction) microvasculopathy and pulmonary fibrosis [26]. In addition to anti-Ro, other antibodies such as aPL have also been implicated. Similar microvascular injury and DIF findings were noted on the TBBX of case 2.

Other potential causes

Another important consideration when dealing with acute pleurisy in SLE is pulmonary embolism, especially when patients are positive for aPL or have other risk factors for thromboembolic disease. Both cases were negative for aPL and did not have other pulmonary embolism risk factors, such as intake of oral contraceptives or prolonged immobilization. In addition, pulmonary embolism was excluded by CT angiography in case 1, and the chronic presentation of case 2 with marked lung consolidation was not consistent with pulmonary embolism. Another consideration in the differential diagnosis is malignancy, including bronchial carcinoma or lymphoma. However, both of our patients had disease courses that were marked by acute and subacute flares of their pleurisy, providing evidence against this diagnosis and in favor of either active SLE or infection. The absence of lymphadenopathy on chest CT scan and a negative history for smoking also provided evidence against malignancy.

Comparison of cases

In case 1, influenza infection was identified. However, there was concern for an alternate disease process, particularly given the dense consolidation and pleural effusion (both unusual in influenza), the lack of improvement with oseltamivir, and the broad infectious differential diagnosis. The positive blood culture for S warneri was thought to represent skin contamination because it grew in only one set of blood cultures and coagulase-negative staphylococci do not typically cause pneumonia. The 2 most likely etiologies of the presentation of case 1 were thought to be severe lupus pleurisy with compression atelectasis or pneumonia due to an unidentified pathogen with associated parapneumonic effusion/empyema. VATS was therefore undertaken to differentiate between these 2 possibilities.

In contrast to case 1, the infiltrates in case 2 appeared to be true consolidations with distinct nodules. This radiographic appearance, as well as their chronicity and progression over time, provided evidence for an infection instead of ALP. The acute clinical deterioration and worsening of the effusion after receiving pulse GC also supported an infectious etiology. The chronic course of disease was not typical of bacterial pneumonia, and nodular infiltrates were more suggestive of fungal, mycobacterial, and Nocardia infection than PCP or viral processes. The TBBX in case 2 revealed histologic and DIF features of chronic interstitial lung disease [26]. Her clinical and radiologic presentation, however, was not consistent with this diagnosis. Therefore, it was unclear if the pathologic findings represented a concomitant chronic interstitial lung disease in a preclinical stage or were false-positive in the setting of inflammation/immune response due to concomitant infection.


Case 1

The patient underwent VATS with decortication and biopsies were taken of the lung and pleura. Grossly, a moderate amount of serous fluid was identified in the pleural space and a thick pleural rind and pleural adhesions were also found. The histology of the lung and pleura showed mild chronic interstitial lung disease and marked reactive fibrinous pleuritis, respectively (Figure 3). DIF showed IgG and IgM deposition and complement activation with capillary injury in interalveolar septa. The patient markedly improved after surgery. She was treated with 1 gm of pulse IVMP and rituximab (1,000 mg every 2 weeks for 2 doses) and was discharged on 2,000 mg per day of MMF and 40 mg of daily prednisone. Her infiltrate and effusion resolved completely within a few months and she was able to taper her prednisone to 10 mg daily at the last followup, 7 months later.

Figure 3.

Fibrinoid pleuritis with pleural edema and extensive deposition of fibrin on the pleural surface in case 1.

Case 2

The patient also underwent a VATS with decortication and biopsies were taken of the lung and pleura. The lung was strongly adherent to the diaphragm and chest well, requiring a lengthy blunt dissection. On gross inspection, the pleural surfaces, pericardium, and diaphragm were studded with fibrous nodules (Figure 4). Biopsy samples of both the pleura and lung revealed a neutrophilic, suppurative, organizing granulomatous inflammatory response with necrotizing abscesses in the setting of a background of capillary injury and fibrin deposition, as seen previously on TBBX (Figures 5A and B). Gram stain of the pleura and pleural fluid revealed branching, beaded gram-positive rods (Figure 5C), and modified acid-fast Kinyoun stain revealed partially acid-fast bacilli. Bacterial cultures subsequently grew Nocardia nova that eventually tested susceptible to TMP/SMX and carbapenems. At the time of her second admission to our hospital, it was noted that a sputum culture that had been sent during her first admission to our laboratory to detect Legionella had grown beaded gram-positive rods. Unfortunately, the culture had been discarded and was not available for further analysis. After the diagnosis of nocardiosis was made, she was started on 15 mg/kg of daily TMP/SMX, divided into 3 doses, and 500 mg of IV meropenem every 8 hours. She continued to receive dual therapy for 6 weeks and then was transitioned to TMP/SMX oral monotherapy for 1 year. A magnetic resonance image of her brain was normal. The patient improved rapidly with complete resolution of infiltrate and effusion within a few months.

Figure 4.

Intraoperative granulomas (arrows) noted on the pericardium and diaphragm in case 2.

Figure 5.

A, Areas of neutrophilic and granulomatous inflammation consistent with an infectious-based etiology in case 2. Adjacent areas of morphologic pattern that could be interpreted as changes of lupus pneumonitis based on the pattern of paucicellular septal and intraalveolar fibrin deposition are shown. B, Septal and intraalveolar fibrin deposition accompanied by red cell extravasation. The process is pauci-inflammatory. Overall, the pattern of parenchymal injury is characteristic for a Gell and Coombs type II immune reaction targeting the endothelium. Immunohistochemistry showed prominent granular deposition of IgG within the septa of the lungs, largely localized to the septal capillaries compatible with an anti–endothelial cell antibody reaction pattern. There was also prominent granular deposition of C5b–9 within the septa. C, Gram-positive branching rods are seen in pleural fluid, consistent with Nocardia.


These cases describe young women with SLE who presented with pleuritic chest pain, pleural effusion, and pulmonary infiltrate. This is a relatively common clinical scenario encountered by physicians who care for patients with SLE, and our cases offer several instructive points for their management. Differentiating active lupus from infection is the main challenge in such cases. This can be very difficult, since typically many of these patients have active lupus and at the same time are markedly immunosuppressed. A meticulous microbiologic evaluation for infection, including opportunistic infection, should be performed in all cases. This often requires thoracentesis to obtain pleural fluid and bronchoscopy to obtain BAL fluid and TBBX samples. If these tests are negative but there is still a strong suspicion for infection as in our cases, VATS may provide a more definitive diagnosis (SLE-associated fibrinous pleuritis in case 1 and Nocardia infection in case 2).

Adding to the diagnostic challenge, GC, especially at relatively high doses as in our cases, both alter the presentation of infections and increase the spectra of possible infectious etiologies. GC impair both innate and acquired immunity and impair the generation of inflammatory exudates by decreasing the production of various proinflammatory mediators [27]. Clinically, these effects are manifested by a decreased febrile response and decreased lung infiltrates and effusions. The fact that the pleural fluid in case 2 did not have clear features of empyema was possibly related to an antiinflammatory GC effect. Nevertheless, despite high doses of GC, there was still substantial inflammation with development of adhesions and loculations of pleural fluid not only in case 2, but also in case 1, in which the effusion was not due to infection, but rather to SLE itself.

The risk of Nocardia infection is significantly increased in immunocompromised patients, and GC are an independent risk factor for nocardiosis [28, 29]. Lung infections caused by Nocardia can manifest as single or multiple lung nodules, reticulonodular infiltrates, lobar consolidations with or without cavitation, and pleural effusions [30, 31]. Therefore, nocardiosis should be considered in immunosuppressed patients with SLE, particularly those receiving high doses of GC who have unexplained pulmonary infiltrates. The lack of growth of Nocardia on BAL fluid and TBBX cultures in case 2 demonstrates the fastidious nature of Nocardia and the difficulty in cultivating these organisms by conventional cultures. These properties explain why the time from onset of symptoms until diagnosis of pulmonary nocardiosis is typically more than 1 month [32]. In case 2, the initiation of prophylactic TMP/SMX 2 days prior to the bronchoscopy may also have decreased the culture yield. If infection with Nocardia is suspected, clinicians should ask the microbiology laboratory to perform a modified acid-fast stain on respiratory specimens that uses a weaker acid as a decolorizer rather than a full acid-fast stain. This modification to the conventional acid-fast stain allows the moderately acid-fast Nocardia to retain the initial stain and to be visualized [33]. Although routine aerobic culture media can support Nocardia, clinicians can also ask for selective media, such as buffered charcoal yeast extract that is used to isolate Legionella. These media are beneficial to support Nocardia because they decrease the overgrowth of other organisms in specimens obtained from nonsterile sites [33]. In case 2, a sputum specimen obtained before initiation of TMP/SMX prophylaxis during her initial hospitalization in our institution that was sent for Legionella culture grew gram-positive rods, which was the first clue that suggested nocardiosis. This case also highlights that Nocardia are not a component of the normal respiratory flora and should be considered a pathogen when grown from any respiratory specimen [33]. Finally, if Nocardia are suspected, the microbiology laboratory should be instructed to hold respiratory specimens for an extended period of time. Respiratory cultures are often discarded at 48–72 hours, and Nocardia often require a longer incubation period to grow [33].

Although there are no randomized trials to determine the optimal treatment of nocardiosis, sulfonamides, such as TMP/SMX, are considered the standard of therapy. However, most authorities recommend initial combination therapy, particularly given recent reports of emerging resistance to TMP/SMX [34, 35]. Although clinical improvement is typically noted within 2 weeks of initiating therapy, prolonged antibiotic courses of up to 12 months are recommended because of high rates of relapse. Nocardia often invade blood vessels and have a propensity to cause brain infections. In fact, the central nervous system (CNS) is the most common extrapulmonary location for nocardiosis [28]. Therefore, any symptoms of CNS disease warrant brain imaging to exclude brain abscess, and many authorities recommend brain imaging in all immunocompromised hosts with Nocardia (34). Fortunately, case 2 did not have CNS lesions on imaging. Case 2 had a dramatic response to antibiotics and is expected to complete a full year of therapy with TMP/SMX.

Mild pleurisy due to active SLE is treated first with nonsteroidal antiinflammatory drugs (NSAIDs) [7]. However, if the presentation is severe (as in our cases), or if the patient does not respond to NSAIDs in a few days, moderate to high doses of GC may be used [7]. The most severe cases may also be treated with IV pulse GC. In rare cases where GC cannot be tapered because of severe (such as in our case 1) or recurrent inflammation, other immunosuppressive agents, such as azathioprine and MMF, may be needed. Tetracycline or talc pleurodesis have been used for recurrent large effusions [7]. Of note, uncontrolled pleural inflammation may lead to fibrothorax and trapped lung [7, 36]. Our case 1 responded dramatically to decortication, supporting its therapeutic role in severe forms of this disease.

The discrepant diagnoses from these 2 similar case presentations provide evidence for early consideration of VATS to distinguish between infection and SLE-associated pleurisy. Despite extensive evaluations with imaging, bronchoscopy, and pathology (using 2 different sampling techniques in case 2, specifically, percutaneous CT-guided needle biopsy and TBBX), the correct diagnoses were not identified and therefore, appropriate care was delayed. Not only does VATS offer diagnostic value, it may have a therapeutic benefit [7, 36]. Case 1 improved rapidly after VATS, before one would expect a benefit from rituximab or an increased dose of MMF. In case 2, the diagnosis was not made until VATS was performed. These cases highlight the important role of VATS in patients with SLE and pulmonary pathology of unclear etiology despite less invasive diagnostic procedures.


Case 1

Marked reactive fibrinous pleuritis due to active SLE with compression atelectasis of the adjacent lung. Mild chronic interstitial lung disease due to SLE.

Case 2

Pneumonia and empyema due to Nocardia nova infection. Possible interstitial lung disease due to SLE.


All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Kirou had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Gulati, Magro, Kirou.

Acquisition of data. Gulati, Satlin, Magro, Kirou.

Analysis and interpretation of data. Gulati, Magro, Kirou.