Lupus Pleuritis

Lupus Pleuritis* Clinical Features and Pleural Fluid Characteristics with Special Reference to Pleural Fluid Antinuclear Antibodies James T. Good, Jr., M.D., F.C.C.P.;t Talmadge E. King, M.D. ;t Veena B. Antony, M.D.;t and Steven A Sahn, M.D., F.C.C.P.;

Eighteen patients with lupus erythematosus (LE) and pleural effusions were evaluated. Fourteen patients had lupus pleuritis and four had pleural effusions of other etiologies. All patients were symptomatic, and the presenting signs and symptoms did not help distinguish between lupus pleuritis and pleural effusions of other causes. The presence of LE cells con6rmed the diagnosis of lupus pleuritis in seven of eight patients. In 11 of 13 patients with lupus pleuritis, the pleuralftuid antinuclear antibody (ANA) titer was ;;J:1:160, and in nine of 13 patients with lupus pleuritis, the pleural ftuid to serum (PF/S) ANA ratio was

;;J: 1. In the four patients with LE and a pleural effusion of another etiology, the pleuralftuid ANA titer was negative in two and low titer in two (1:40, 1:80); the pleural Ouid to serum ANA titer was always less than one. Of 67 patients with pleural effusions of other etiologies, the pleural ftuid ANA was negative. The signs and symptoms of lupus pleuritis are nonspeciflc, however; the 6ndings ofLE cells in pleural ftuid con6rms the diagnosis and a high pleural ftuid ANA titer (;;J:l:l60) and a PF/S ANA ratio of ;;J:l strongly supports the diagnosis.

W,en patients present with acute pleurisy manifested by severe chest pain, dyspnea, fever, cough, and a friction rub, many diagnostic possibilities exist. While patients with collagen vascular disease, in particular systemic lupus erythematosus (SLE), may frequently have pleuro-pulmonary involvement, 1.a establishing the diagnosis of lupus pleuritis can be difficult. In this setting, those diagnoses which are most commonly confused with lupus include pleurisy secondary to (I) pulmonary embolism, (2) parapneumonic effusion, (3) viral infections, (4) tuberculosis, (5) other collagen vascular disorders including rheumatoid arthritis, and occasionally (6) congestive heart failure. This last group of patients may be receiving drugs such as procainamide or hydralazine, which are associated with the drug-induced lupus syndrome. As with most other pleural diseases, no one test has been shown to be pathognomic for lupus pleuritis. Because of the difficulty in differentiating this furm of pleurisy from other etiologies, we undertook a study to evaluate the clinical presentation and to analyze the pleural fluid characteristics of patients with lupus erythematosus and the acute development of pleural effusions.

Center and affiliated hospitals from January 1979 to June 1982. All patients had the clinical diagnosis of systemic lupus erythematosus based on having fuur or more of the American Rheumatism Association criteria fur the diagnosis oflupus.3 In 14 patients, lupus was thought to be responsible fur the pleural effusion and in fuur patients other etiologies were established. In ten of 14 patients with lupus pleuritis, the diagnosis of SLE had been established ten months to 25 years prior to the most recent episode of pleurisy. The other fuur patients had pleurisy as one of the initial manifestations oflupus, two had drug-induced disease, and two had systemic lupus erythematosus. One patient with drug-induced lupus had been taking 200 mg of hydralazine daily fur two years and had the findings of polyarthritis, positive lupus erythematosus (LE) cells in the peripheral circulation, and pleurisy. The second patient was treated with procainamide, and the diagnosis of drug-induced lupus was established by the presence of fever, arthritis, pleurisy, and a positive serum ANA. A clinical diagnosis of lupus pleuritis was established when the patient met the criteria fur either drug-induced or native lupus and no other etiology fur the pleural effusion could be established. All patients had posterior-anterior and lateral chest roentgenograms perfOrmed within 24 hours of admission to the hospital. Thoracocenteses were perfOrmed on all patients. Routine pleural fluid characteristics were determined which included protein concentration, lactate dehydrogenase (LDH) concentration, total leukocyte count with differential, glucose concentration, and pH determination. In eight patients, LE cells were looked fur in the pleural fluid by Wright stain technique. Seventeen patients had simultaneous pleural Huid and serum antinuclear antibody (ANA) determinations by standard techniques. • Briefly, the patient's serum is incubated with a section of mouse kidney. The slide is washed to remove excess serum and fluorescein-labelled goat anti-human IgG is added which binds to any human IgG present. The slide is again washed to remove excess Huorescein labelled IgG and is then examined under the Huorescence microscope. In some cases, when the pleural fluid ANA was positive at a 1:20 dilution, the samples were analyzed at progressive serial dilutions beginning at 1:8. In other laboratories, progressive dilutions from 1:20 were utilized. Eleven patients had total hemolytic complement levels determined. In addition to the 14 patients with lupus pleuritis, a second group

METHODS

Eighteen patients with lupus erythematosus and pleural effusions were evaluated at the University of Colorado Health Sciences *From the Division of Pulmonary Sciences, Department of Medicine, University of Colorado Health Sciences Center, Denver Veterans Administration HospiW, and Denver General Hospital, Denver. t Assistant Professor of Medicine. *Professor of Medicine. Manuscript received February 28; revision accepted June 17. Reprint requeltB: Dr: Sahn, 4200 East Ninth Avenue, C-272, Denver

80262

714

with known SLE who presented with pleural effusions due to an etiology other than lupus were studied. While lupus was initially considered a possible diagnosis, the fOlioing diagnoses were established: (1) empyema (documented by positive pleural fluid culture); (2) congestive heart &ilure (clinical signs of biventricular &ilure, borderline exudative effusion which cleared with diureses); (3) parapneumonic effusion (exudative effusion associated with pneumococcal pneumonia); and (4) effusion secondary to a viral syndrome (confirmed by acute and convalescent serum titers) (see 'lllble 2). The purpose lOr evaluating this second group was to compare clinical presentations and pleural fluid characteristics to patients with lupus pleuritis. While all !Our patients had some manifestations of disease activity at the time of presentation (eg, arthritis, discoid lupus, proteinuria), none had overt activity requiring a change in therapy. A third group of 67 patients with diagnoses other than lupus had pleural fluid ANA determinations.

iJ

~

+

I

I

+

+ +

~

+

+ +

+

+ +

~+

++

+

I+

~+

++

+

I+

;::! +

+

I

+

I

::!+

++

+

++

+

I

+

+

+ +

I

+

+++I+~

+

+

+

I

+I

+I+~~

I

+I

+++

I

~

+

+

+

v

:r

rf .....

j

RESUIIS

'Thble 1 reviews the presenting signs and symptoms and pleural fluid characteristics on admission of the 14 patients with lupus pleuritis. In five of these patients, the diagnosis was initially suspected. The most common presenting symptoms in these patients were pleuritic pain and dyspnea. No patients had hemoptysis. Evidence of pleural effusion on physical examination was present in 12 of 14 patients. All patients with lupus pleuritis were symptomatic at the time of presentation and most had abnormal physical findings. While other manifestations of lupus erythematosus were present in all patients (see Table 1), the pleuropulmonary findings were the most striking, and in some cases, so prominent that other less obvious findings of SLE were initially overlooked. While the chest roentgenographic findings were not specific fur the diagnosis oflupus pleuritis, all patients had abnormal chest x-ray film findings at the time of admission. One-half of the patients had bilateral pleural effusions, while fOur patients had right-sided effusions, and three, left-sided effusions. Most effusions were of small to moderate size (400 ml to 1,000 ml). Other abnormalities were noted in 11 of 14 patients. These included alveolar infiltrates in fOur, atelectasis in three, and an enlarged cardiac silhouette in fOur. All patients with lupus pleuritis had exudative effusions by protein (PF/S >0.5) or LDH criteria (PF/S >0.6 or >200 IU/L). The LDH values never exceeded 550 IU/L. There was a wide range ofleukocyte counts from 230 to 15,000 cells/J.Ll. The differential leukocyte count also showed a wide range varying from 10 percent to 100 percent polymorphonuclear leukocytes. Fifty percent of the patients had polymorphonuclear predominate effusions. Four of the five patients with mononuclear effusions had their effusion present fur at least seven days. Eleven patients had normal glucose (>60 mVdl, PF/S >0.5) and normal pH (>7.30) effusions. Three patients had low glucose-low pH effusions, and these patients did not differ clinically

pi

l

f fI

f- H

- I

~

i i ~i i~i ~ !!:oil: J ~

~ ~

~ ~~

~ ~ ~ h

from those patients with normal pH, normal glucose effusions. Neither the total leukocyte count, differential count, size of effusion, nor duration of the effusion correlated with the pH or the glucose concentration of the effusion. Of the eight patients who had pleural fluid examined fur LE cells by the Wright stain, seven were round to have at least one typical LE cell. In 11 of 13 patients with lupus pleqritis, the pleural fluid ANA titer was ;a::l:l60, and nine of 13 patients with lupus pleuritis, the pleural fluid to serum ANA ratio was ;a::1 CHEST I 84 I 8 I DECEMBER, 1883

715

"'

t

1: 1280 1:640 Pleural Fluid

ANA

Titer

.....

z

•

1:2560

#

~

!ll

~

..,..,

3:

..,

'r

8

+ +

+

+

l- §.e:

!!

1:320 1: 160

•

·-

------H·-----------------•

1:80 1:40 0

8.01

• •

f

g.

:c

Lupus Pleuritis

•

!I

e.

l

Lupus with Pleural Effusion of other Causes

1.0

0.5 0

+

+

+

8

+ +

+

+

!I

§

F.

z

<;I

+ +

l!J

I

I

"+

>

£:

!I

to

I

a"'

la.

f l "g

ai' fl)

c '§.

":! v


~

fl)

f

1

~~

~~

I

~

+

+

~[

+

r l l ,~

Q

l

ri.

:II

3.0

Fluid/ Serum Ratio 2·0 ANA Titer

I

ir

J Ple~al

<;I

I

•

~

::!!

>

>

8

+ +

+

•

!:!!

"l

E"

6:.

f;ifi!l ~J ~

•

•

Lupus Pleuritis

Lupus with Pleural Effusion of other Causes

1. Top, Only two patients with lupus pleuritis had pleural fluid ANA titers less than 1:160. No patient with lupus and pleural effusion secondary to other etiologies had a pleural fluid ANA titer greater than 1:80. Bottom, No patient with lupus and acute pleurisy &om other causes had a pleural fluid to serum ANA ratio of one or greater, while ten ofl3 patients with lupus pleuritis had ratios of one or greater.

(Fig 1). Of the nine patients in whom total hemolytic complement values were determined, five had low values (<20 units/ml), while fuur had normal values. The pleural6.uid findings in patients with SLE and acute pleural effusion of other etiologies are listed in Thble 2. Both symptoms and physical findings were similar to patients with lupus pleuritis. All patients had positive serum AN As which ranged from titers of 1:40 to 1:640. Two patients (one with CHF and one with a parapneumonic effusion) had positive pleural B.uid AN As (1:40 and 1:80). The pleural B.uid to serum ANA ratio was always less than one. In the one patient who had the pleural B.uid examined by Wright stain fur LE cells, none were fuund. Of 67 patients with pleural effusions of other etiologies who did not have lupus erythematosus, the pleural B.uid ANAs were negative. These diagnoses

~

r ~

~ > .~~l !!

---- - ~---------- - -------

FICURE

718

oo!:l

~

a.

t

••

~~

"'

n

:c ..,

f

s·

til

a

!;'

H fi X'

"' I+ 3:

1

e

i

I+

...iii

:--1:--1 !:!~

,.. ,.. ... :c ~

0

oo

I+ (l!

~!!!

e;

!!!

;; I+ ~

fi!:

$

~

... .,

"'~

~

[

j3

:]r

c ~

n

,.

~

~

!liN~

$

i

I

>

$~ z~

.a

!

~z

z

>

!

~

$

.a

z

>

z z

ii1l

i;l

~~

z

f l ri.

~

§ lll

I

r

f

~I+~

Cit I+

l

;!'

~

if

f

3

e. c.

~

[.

~

[-

:2.

':i

f

ii

ttl

~t"

~

=(ll

Q

:) ! n :c l!J Lupus Pleurltla (Good et-')

included the following: malignancy (n = 13), parapneumonic effusion (n =9), empyema (n =1), congestive heart failure (n =6), idiopathic {n =5), tuberculosis {n =5), cirrhosis (n =4), trauma {n =3), pulmonary embolism (n =3), subdiaphragmatic abscess (n =3), rheumatoid arthritis (n =3), atelectasis (n =3), and chylothorax, eosinophilic pleuritis and paragonimiasis (one each). DISCUSSION

While the pleuropulmonary manifestations of systemic lupus occur commonly, lupus pleuritis initially may be overlooked as the cause of the pleural effusion. First, SLE may not have been diagnosed previously. Second, the presentation is nonspecific and compatible with other diagnoses. In the 14 patients with lupus pleuritis, only five had this admitting diagnosis, while the remainder had initial diagnoses of pulmonary embolism, empyema, tuberculous pleurisy, congestive heart failure, and nephrotic syndrome. It was only after pleural fluid analysis in conjunction with the clinical presentation that the diagnosis of lupus pleuritis was established. In patients with lupus pleuritis, all had either LE cells present, high pleural fluid ANA titers or a pleural fluid to serum ANA ratio greater than or equal to one. In two patients with lupus and pleural effusions of other etiologies, the pleural fluid ANA was fuund to be positive, but at low titers (S1:80). There was marked variation in the appearance of pleural fluid in the 14 patients with lupus pleuritis. Approximately one-half of the patients had clear, serous fluid, while the other one-half had sanguineous or serosanguineous fluid. Thus, the gross appearance of the fluid does not help distinguish lupus pleuritis from pleuritis associated with pulmonary embolism or tuberculosis. In contrast to reports of others, a low glucose, low pH effusion can occur in patients with lupus pleuritis. 5 •6 The findings of a normal pleural fluid glucose has been suggested as a point of differentiation between lupus and rheumatoid pleuritis; however, three patients in this series had low pH, low glucose effusions. Lupus effusions tend to have higher glucose concentrations (<;!:32 mgldl) and pH values compared with rheumatoid effusions. Patients with rheumatoid pleurisy tend to have marked pleural fibrosis which can lead to a selective influx block to glucose and an effiux block to hydrogen ion. 7' 10 Neither the total leukocyte count nor the differential count on the pleural fluid helped to establish a diagnosis oflupus pleuritis. Most effusions were polymorphonuclear predominant; however, this probably represents the timing of the thoracocentesis. It is well recognized that with acute pleural injury, polymorphonuclear predominant effusions are the rule.

Recently, there has been a great interest in pleural fluid immune complexes and complement values as a means of establishing the diagnosis of lupus pleuritis. Shocket et al 11 reported a patient with immune complex vasculitis causing ascites and pleural effusions in SLE. 11 Chandrasekhar et al12 reported that antibody deposition in the pleural fluid was a finding associated with drug-induced lupus and did not occur in effusions of malignant, tuberculous or rheumatoid etiologies. The issue of pleural fluid complement activity in patients with lupus remains intriguing. Reduced complement values have been reported in both SLE and rheumatoid arthritis. 13 Hunder et al14 suggested that pleural fluid complement depletion may be secondary to immunologic activation and that immune complexes probably contribute to the development of pleuritis in both lupus and rheumatoid arthritis. Later reports by Glovsky et al15 demonstrated that marked diminution of pleural fluid total hemolytic complement activity was fuund in both SLE and rheumatoid arthritis, but not in patients with nonrheumatic disorders. They concluded that the complement profiles in sero-positive rheumatoid arthritis and SLE suggest primary complement pathway activation, and functional hemolytic assays were more sensitive indicators of active disease than were the usual quantitative protein determinations of individual complement components. Of nine patients with lupus pleuritis, four had normal hemolytic complements and five had depressed values. When one considers the mechanism responsible fur low total hemolytic complement, it is not surprising that certain patients' total hemolytic complement may not be depressed. To adequately evaluate the complement system, both C 3 and C 4 conversion products should be measured. In H under and McDuffies's article, evaluation of pleural fluid complement conversion and immune complexes in immunologic and nonimmunologic diseases, the highest concentration of immune complexes in pleural fluid were found in patients with rheumatoid arthritis. 14 Pleural fluid immune complex concentration correlated positively with conversions of C 3 • These findings suggested that the reduced levels of pleural fluid complement in rheumatoid arthritis and systemic lupus were secondary to complement conversion by these complexes. Similar findings have been reported by Halla et al16 where low complement levels and immune complexes were fuund in the pleural fluids of patients with rheumatoid arthritis and SLE. However, in rheumatoid arthritis, there appeared to be local pleural space production of immune complexes as contrasted with SLE where the presence of immune complexes reflected more closely the serum concentration. While the complement levels differentiated rheumatoid from control effusions, they did not separate lupus pleuritis CHEST I &4 I 8 I DECEMBER, 1983

717

from rheumatoid pleurisy. While the finding of low pleural 8uid complement levels and the presence of immune complexes supports the diagnosis of lupus pleuritis, it is not pathognomic. In our series, the most specific pleural fluid characteristic to differentiate lupus pleuritis from other etiologies was the presence of the LE cell. With improved techniques of cytocentrifugation, the in vivo demonstration of the LE phenomenon in body fluids is being reported with increasing frequency. 17 In addition to native lupus, the finding of LE cells in pleural 8uid also has been established in drug-induced lupus pleuritis. 18•19 Currently, as with our data, there are no reports of false positive LE cells from pleural fluids from patients who do not have lupus. Thus, this finding appears highly specific for pleural involvement with lupus. Leechawengwong et al10 performed ANA tests on 100 consecutive pleural8uids of multiple etiologies. In seven of the effusions, the ANA was positive and all seven patients had either drug-induced or native lupus. Their observations support our findings in which all67 patients who had pleural effusions of other etiologies had negative pleural 8uid ANAs. In general, the serum ANA reSects disease activity and would be expected to be high in lupus pleuritis. However, our data have shown that lupus pleuritis can occur with low serum titers, and conversely, may not be present when serum titers are high. We conclude that in patients with acute pleurisy, a pleural8uid ANA titer o£1:160 or greater or a pleural fluid to serum ANA ratio of one or greater suggests that the pleurisy is secondary to active lupus. REFERENCES 1 Osler W. On the visceral manifestations of erythema group of skin diseases. Am J Med 1904; 127:1-29 2 Alarcon-Segovia D, Alarcon DG. Pleuro-pulmonary manifestations of systemic lupus erythematosus. Chest 1961; 39:7-17 3 Rodnan GB (ed). Primer on the rheumatic diseases. JAMA 1973; 225:139.

718

4 GreenwaldCA, PeeblesCL, NakamuraRM . Laboratorytestsfor antinuclear antibody in rheumatic disease. Lab Med 1978; 9:19-27 5 Carr DT, Lillington GA, Mayne JG. Pleural fluid glucose in systemic lupus erythematosus. Mayo Clinic Proc 1970; 45:409-12 6 Good JT Jr, Thryle DA, Maulitz RM , Kaplan RL, Sahn SA. The diagnostic value of pleural fluid pH. Chest 1980; 78:55-59 7 Good JT Jr, King TE, Sahn SA. Native and drug-induced lupus pleuritis: a cause of low glucose-low pH pleural effusion. Chest 1980; 78:518 8 Sahn SA, Kaplan RL, Maulitz RM, Good JT Jr. Rheumatoid pleurisy: observations on the development of low pleural fluid pH and glucose level. Arch Intern Med 1980; 140:1237-38 9 Dodson WH, Hollingsworth WJ. Pleural effusion in rheumatoid arthritis. New Eng) J Med 1966; 275:1337-42 10 'llu-yle DA, Good JT Jr, Sahn SA. Acid generation by pleural fluid: possible role in the determination of pleural fluid pH. J Lab Clin Med 1979; 93:1041-46 11 ShocketAL, Lain D, Kohler PF, Steigerwald]. Immune complex vasculitis as a cause of ascites and pleural effusions in systemic lupus erythematosus. J Rheumatol1978; 5:33-38 12 Chandrasekhar AJ, Robinson J, Barr LB. Antibody deposition in the pleura: a finding in drug-induced lupus. J Allerg Clin lmmunol1978; 61:399-402 13 Hunder GG, McDuffie FC, Hepper NGG. Pleural fluid completement in systemic lupus erythematosus and rheumatoid arthritis. Ann Intern Med 1972; 76:357-63 14 Hunder GG, McDuffie FC, Huston KA, Elveback LR, Hepper NG. Pleural fluid complement, complement conversion and immune complexes in immunologic and nonimmunologic disease. J Lab Clin Med 1977; 90:971-80 15 Glovsky MM, Louis JS, Pitts WH Jr, Alenty A. Reduction of pleural fluid complement ability in patients with systemic lupus erythematosus and rheumatoid arthritis. Clin Immunol Immunopathol1976; 6:31-41 16 Halla JT, Schrohenloher RE, Volanakis JE. Immune complexes and other laboratory features of pleural effusions. Ann Intern Med 1980; 92:748-52 17 Gamet RF Jr, Atkinson BF, Bonner H, Wurzel HA. Rapid screening for lupus erythematosus cells using cytocentrifuge prepared huffy coat monolayers. Am J Clin Pathol 1977; 67:537-39 18 Keshgegian AA. Lupus erythematosus cells in pleural fluid (letter to editor). Am J Clin Pathol 1978; 69:570-71 19 Carel RS, Shapiro MS, Shoham D, Gutman A. Lupus erythematosus cells in pleural effusion. Chest 1977; 72:670-72 20 Leechawengwong M, Berger HW, Sukumaran M. Diagnostic significance of antinuclear antibodies in pleural effusion. Mt Sinai J Med 1979; 46:137-39

Lupue Pleurttla (Good et a/)