Seminars in Arthritis and Rheumatism
Volume 35, Issue 3 , Pages 138-142, December 2005

Diffuse Alveolar Hemorrhage: A Nonthrombotic Antiphospholipid Lung Syndrome?

  • Ronald A. Asherson, MD (FRCP, FACP)

      Affiliations

    • Rheumatic Diseases Unit, University of Cape Town Health Science Center, Cape Town and the Rosebank Clinic, Johannesburg, South Africa
    • ,
  • Ricard Cervera, MD, PhD (FRCP)

      Affiliations

    • Department of Autoimmune Disease, Hospital Clinic, Barcelona, Catalonia, Spain
    • Corresponding Author InformationAddress reprint requests to: Ricard Cervera, MD, PhD, FRCP; Servei de Malalties Autoimmunes; Hospital Clinic; Villarroel, 170; 08036-Barcelona, Catalonia, Spain.
    • ,
  • Athol U. Wells, MD (FRCP)

      Affiliations

    • The Royal Brompton Hospital, London, England, United Kingdom

Article Outline

 

The authors of the article on diffuse alveolar hemorrhage (DAH) and pulmonary “capillaritis,” which appears in this issue of the journal (1), have drawn attention to the existence of specific pulmonary manifestations encountered in 4 patients with antiphospholipid antibody (aPL) positivity. Two of the patients had a prior history of primary antiphospholipid syndrome (APS). The first had seizures and recurrent pulmonary emboli and had been maintained on warfarin therapy. Their second patient developed a pulmonary embolus on a background of recurrent episodes of DAH and capillaritis over a 2-year period. The remaining 2 had no associated thromboembolic accompaniments either previous to, or simultaneous with, the occurrence of the DAH. The condition appeared to arise de novo in these aPL-positive patients.

DAH has been described in a number of systemic autoimmune diseases, including SLE, Behcet’s disease, microscopic polyarteritis, cryoglobulinemia vasculitis, Henoch–Schonlein purpura, Goodpasture’s syndrome, granulomatous vasculitis, and “overlap” situations as recently reviewed by Zandman-Goddard (3). To this list must now be added APS.

The potential clinical importance of this complication needs to be stressed. It is likely that the true frequency is significantly higher than the very low prevalence suggested by the paucity of reported cases. DAH syndromes are notoriously difficult to diagnose. In the majority of cases, there is little or no hemoptysis, even with large volume intra-alveolar bleeding. The radiological signs are sometimes florid but often highly nonspecific, consisting of amorphous ground-glass attenuation on chest radiography or high-resolution computed scans of the thorax (HRCT). Invasive or semi-invasive evaluation is generally required, specifically bronchoscopy, with or without a surgical lung biopsy. This, in turn, demands a higher index of diagnostic suspicion than often appears warranted, based on the prevailing perception that these disorders are rare. Newly recognized complications or disorders are generally underdiagnosed initially, with diagnosed cases tending to lie at the severe end of a much wider spectrum of disease (as in the current cases, characterized by overt hemoptysis and severe rapidly progressive disease). In time, increasing clinician awareness leads to increasing recognition of milder and milder disease. This widely recognized “tip-of-the-iceberg” phenomenon, historically seen in Wegener’s granulomatosis and systemic sclerosis alike, is especially likely to apply when clinical manifestations are nonspecific and invasive procedures are required to make the diagnosis.

A focus on diagnostic uncertainties is important because the very existence of this complication gives rise to major management dilemmas when the APS is complicated by apparent interstitial lung disease. In general, in treated autoimmune diseases, infiltrative lung disorders can, for practical therapeutic purposes, be divided broadly into opportunistic infection, which demands specific antimicrobial therapy and a reduction in immunosuppression, and a wide range of immunologically mediated processes which demand the opposite approach: viz, intensification of immunosuppressive therapy. Refinement of the differential diagnosis in the latter group is important, but less important than the exclusion of infection. Bronchoalveolar lavage (BAL) to exclude infection has been the pivotal investigation that has allowed an empirical immunosuppressive approach. The difficulty peculiar to the APS is the need to distinguish between a hemorrhagic and a thrombotic process, with crucial therapeutic consequences. Both continuation and cessation of anticoagulation may be disastrous, depending on the underlying diagnosis. A pragmatic approach of standard immunosuppressive therapy after the exclusion of infection is not, in itself, sufficient. Active steps must be taken to diagnose alveolar hemorrhage.

Does this need to include thoracoscopic or open lung biopsy, potentially a very high-risk procedure in the compromised patient? Provided that the index of clinical suspicion is high, alveolar hemorrhage can generally be diagnosed without a surgical procedure. A striking fall in hemoglobin, in association with evidence of an alveolar filling process on chest radiography, is highly suggestive. Single-breath gas transfer measurement (DLCO) may be diagnostic as intra-alveolar blood causes a truly striking rise in gas transfer, often to 300 to 400% of normal in more severe disease; this finding is pathognomonic, although the test must be performed within 36 hours of hemorrhage and false-negative results can therefore occur. BAL is the pivotal test. In severe disease the diagnosis is usually obvious, with overt hemorrhage from multiple lung segments. In milder disease, the initial BAL yield may appear normal, with increasing blood staining strongly suggesting a diffuse hemorrhagic process.

Is it truly necessary to establish that capillaritis is present? In the article by Deane and West in the present issue (1), nonspecific neutrophilic infiltration was the overriding histologic abnormality in 2 of the 3 biopsied cases, and capillaritis was not seen. This did not alter the diagnosis and the therapeutic approach was vindicated in both cases by a response strongly indicative of an immunologically mediated process. Biopsy procedures may be prone to “sampling error,” with regional variation in histological findings and findings at biopsy may not be representative of the whole lung. Moreover, histological findings in pulmonary vasculitis may wax and wane, especially in more indolent disease. Even in active disease, there is ample precedence in Wegener’s granulomatosis for a histological finding of nonspecific alveolar hemorrhage without overt vasculitis, in association with a diagnostic renal biopsy. The writers question the view that a histological diagnosis is always required in high-risk situations. In this context, it is surely sufficient to diagnose alveolar hemorrhage noninvasively and then to ask oneself whether management would change depending on whether capillaritis is present. In the current cases, the answer was clearly “no.”

Vascular damage and rupture of small pulmonary vessels must be the basic pathogenesis in the group of patients with DAH and systemic autoimmune diseases. In other conditions associated with “bland” alveolar hemorrhage (eg, severe mitral stenosis, inhaled toxins, severe infections, and coagulopathies), the etiopathogenetic mechanisms and histopathological findings in the lungs are different and will not be reviewed in this Editorial.

The authors have constructed a most plausible hypothesis to explain these nonthrombotic associations of the aPL and have reviewed all similar cases in the literature including their own 4, numbering a total of 17 (1). Of these, a majority had suffered previous thrombotic complications and were examples of the APS. These included deep vein thromboses, in some accompanied by pulmonary embolization (9 patients), seizures (4), strokes (8), and other arterial occlusions, including myocardial infarctions (5). They do not comment on whether the patients have either SLE or “lupus-like” disease. They speculate as to whether the pulmonary damage is caused by endothelial cell activation induced by the aPL, along with complement activation and inhibition of activated protein C. This latter mechanism, however, is not operating in these patients as evidenced by the paucity of microthrombotic pathology in those patients who had undergone lung biopsies. Protein C deficiency is associated with large-vessel venous occlusions and not microvascular thrombosis (4). aPL-associated endothelial cell activation has been well demonstrated (5), and will result in the liberation of many cytokines as well as upregulation of adhesion molecules. Neutrophil binding to the endothelium will lead to “capillaritis” and vascular damage/disruption is then understandable. However, in the 2 patients who had biopsies performed, perivascular or vascular infiltrates indicative of a true vasculitis are not described; rather, septal and alveolar neutrophilic infiltrates were noted. The association of true vasculitic conditions (eg, polyarteritis nodosa, microscopic polyarteritis, etc) with the APS is very uncommon and may in fact be only coincidental. The role of complement activation in animal models with sera taken from patients with fetal loss and aPL has been most elegantly demonstrated by the group from the Hospital for Special Surgery in several articles (6), and its role in DAH remains to be adequately investigated, possibly with a similar model using sera from patients with DAH. The authors of these articles have speculated that the levels of complement activation may be too low for significant alteration in serum complement to be detected in patients. Staining for complement deposition was performed in 1 patient only (case 1) and was reported to be negative. They also speculate on the role of toll-like receptors and smoking; however, 3 of their patients were nonsmokers.

Several important papers recently published have also stressed a major role for complement. Pierangeli et al (7) studying thrombus dynamics and adhesion of leukocytes in mice deficient in complement components C3 and C5, found that these mice were resistant to the enhanced thrombosis and endothelial cell activation induced by aPL. Furthermore inhibition of C5 activation using anti-C5 monoclonal antibodies prevented thrombophilia induced by aPL. Hart et al (8) and Fleming et al (9) both from the same group in Boston also recently published 2 important papers, potentially of major significance in understanding the unusual non-thrombotic pulmonary complications of the APS. In their first paper they studied the role of the classical and lectin complement pathways in gastrointestinal ischemia/reperfusion injury (GI/R injury) in C1q deficient mannan-binding lectin (MBL)-A/C deficient (MBI-null) and other genetically deficient mice models. Gastrointestinal ischemia followed by 3-h reperfusion induced both intestinal and lung injury. They elegantly demonstrated that C2 and MBL, but not C1q, were necessary for gut injury after GI/R. The lung injury was MBL and C1q independent, but C2 dependent, suggesting a role for ficolins in this model. C1q KO and MBL null mice demonstrated secondary lung injury and pulmonary neutrophil infiltration (as demonstrated by Deane and West in their biopsied cases) MBL-null mice reconstituted with MBL, after GI/R. In addition to MBL, H-ficolin and L-ficolin, can activate the lectin pathway, which is mainly synthesized in the liver. However L-ficolin alveolar hemorrhage in this group lends great credence to this theory; it is also produced in the lung by alveolar type-11 cells and unciliated bronchial epithelial cells. L-ficolin binds to Escherichia coli and lipoteichoic acid, a cell wall constituent of Gram-positive bacteria. Gut barrier dysfunction may lead to bacterial translocation to the lung resulting in increased complement neutrophil infiltration as a result of lectin complement pathway activation via ficolins. MBI activates the lectin complement in the intestines; ficolins may be activating complement in the lungs. This hypothesis explains why with neutrophilic infiltration initially, secondary disruption of alveolar blood vessels might take place resulting in DAH. The high frequency of abdominal syptomatology in catastrophic APS patients and the much higher frequency of pulmonary complications such as alveolar hemorrhage in this group supports this link. The ficolin hypothesis also ties in well with the high frequency of infections as “triggering” mechanisms for CAPS.

The second paper from this group (9) reported that both murine and human monoclonal and polyclonal antibodies against negatively charged phospholipids could reconstitute mesenteric I/R-induced intestinal and lung tissue damage in complement receptor-2-deficient mice. Antibodies against beta 2 glycoprotein 1 restored local and remote tissue damage in these mice. Unlike the complement receptor-2-deficient mice, reconstitution of I/R tissue damage in the injury resistant Rag-1 mouse required the infusion of both anti-B2 GP1 and aPL. They concluded that aPL could bind to tissues subjected to I/R insult and mediated tissue damage, just as they mediate fetal growth retardation and loss when injected into pregnant mice.

The rarity of DAH, and the fact that in 2 of the patients in Deane and West series (1), clinical APS had not been documented, suggests that differing pathogenic “idiotypes” of the aPL may exist and that the end-organ damage/thrombosis which occurs depends on the specificity of the idiotypes. While most pathogenic aPL result in large-vessel thrombosis, others may only cause fetal loss and/or thrombocytopenia, while others may result in essentially “nonthrombotic” presentations (eg, chorea, many cases of recurrent fetal loss where no thromboses have been detected on placental histopathological examination or DAH). The size of the vessels involved also may be affected by the differing specificity of the aPL, eg, predominance of small-vessel occlusions in patients with the catastrophic APS (Asherson’s Syndrome) (10). This concept is not new in the field of autoimmune diseases and has been postulated to explain the diversity of manifestations which characterize SLE or even rheumatoid arthritis (RA). How else can we explain the patient who presents only with renal SLE or chorea or the RA patient who does not have involvement of the metacarpophalangeal joints or erosive disease in the hands but instead demonstrates bilateral hip or shoulder arthritis only? Of course, multiple pathogenetic factors may also play important roles. These may involve the HLA Class II genes, gene polymorphisms, which we are now only beginning to unravel, and many other variables. Of more than passing interest is that none of these 4 patients developed an acute (adult) respiratory distress syndrome (ARDS), so prevalent in patients with catastrophic APS (11).

Deane and West (1) were only able to list 17 patients with DAH and aPL. This is a small harvest considering that worldwide there may be thousands of patients with APS. In the catastrophic APS series of 200 patients (“CAPS Registry”), all documented on the web site (http://www.med.ub.es/MIMMUN/FORUM/CAPS.HTM), we have already registered 18 patients with DAH to date (12). Thus, there is an enormous difference in prevalence between its occurrence in the simple APS and in the catastrophic APS. This differing prevalence is also evident with ARDS. Few cases have been reported with the simple or classic APS (13), whereas 47 patients of 250 with catastrophic APS had this condition (11). It has been hypothesized that DAH is dependent on the systemic inflammatory response syndrome (SIRS), is cytokine-mediated, and is caused by the extensive tissue necrosis, which is the hallmark of this condition.

The frequency of coexisting pulmonary microthrombosis also is small (2 patients only in each series) However, it is relatively common in catastrophic APS (27/250 patients). One patient only in the series of patients with CAPS and DAH had pulmonary hypertension (PHT).

PHT and its relationship to the APS/aPL has been the topic of many articles over the past few years. There is no difficulty in understanding the relationship between thromboembolic PHT and the aPL. The prevalence of aPL positivity in patients with chronic thromboembolic PHT varies from 10 to 20% (14, 15, 16). The association of the aPL with PHT in SLE patients was first reported in 1983 (17) and has been documented both in patients with SLE as well as with the “Primary” APS (2, 18).

Several cases of elevated aPL were studied in an early series of “primary” nonthromboembolic PHT patients studied by Asherson and coworkers in 1983, confirming the later clinical associations (19). The pathogenesis of PHT is not understood in these nonthrombotic cases but there may be a relationship to an interaction of the aPL with endothelin-1, a highly speculative hypothesis.

Therefore, the pathogenesis of DAH, ARDS, microthrombosis, and PHT differs. The occurrence of DAH (and/or “capillaritis”) is 1 of the more frequent manifestations of the catastrophic APS as opposed to its infrequent occurrence in patients with simple/classic APS, and it does not appear to coexist with any of the above-mentioned conditions.

Treatment of the severe pulmonary manifestations associated with aPL probably requires a similar approach to that required in the catastrophic APS. When several combination regimens were analyzed in the “CAPS Registry,” the highest survival rate (70%) was achieved with the combination of anticoagulation, corticosteroids, and plasma exchange or intravenous gamma-globulins, which has been highlighted in the proposed guidelines for the management of patients with catastrophic APS (20). The authors have reviewed the therapy of DAH admirably (1). Pulmonary involvement in APS has recently been extensively reviewed by Espinosa and coworkers (21).

Should DAH, ARDS, primary nonthromboembolic PHT, and, very uncommonly, fibrosing alveolitis (3 patients documented with the APS) be grouped together under the all-embracing term “Nonthrombotic Antiphospholipid Lung Syndrome” (Fig. 1)? The thrombotic manifestations would of course result from microthrombotic occlusions or pulmonary emboli with resultant thromboembolic PHT.

The existence of the “nonthrombotic” manifestations, increasingly reported, demonstrates the need for a reclassification of the APS into “thrombotic” and “nonthrombotic” manifestations, which hopefully will take place during the next biannual International APS Symposium to be held in Italy in 2006.

Back to Article Outline

References 

  1. Deane KD , West SG . Antiphospholipid antibodies as a cause of pulmonary capillaritis and diffuse alveolar hemorrhage (a series and literature review) . Semin Arthritis Rheum . 2005;35:154–165
  2. Asherson RA , Khamashta MA , Ordi-Ros J , Derksen RHWM , Machin SJ , Barquinero J , et al.   The “primary” antiphospholipid syndrome (major clinical and serological features) . Medicine (Baltimore) . 1989;68:366–374
  3. Zandman-Goddard G . Diffuse alveolar hemorrhage in autoimmune diseases . J Isr Med Assoc J . 2002;4:461–462
  4. Esmon CT , Gu J-M , Xu J , Qu D , Stearns-Kurosawa DJ , Kurosawa S . Regulation and functions of the protein C anticoagulant pathway . Haematologica . 1999;84:363–368
  5. Pierangeli SS , Colden-Stanfield M , Liu X , Barker JH , Anderson GL , Harris EN . Antiphospholipid antibodies from antiphospholipid antibody syndrome patients activate endothelial cells in vitro and in vivo . Circulation . 1999;99:245–250
  6. Salmon JE , Girardi G , Holers VM . Activation of complement mediates antiphospholipid antibody-induced pregnancy loss . Lupus . 2003;12:535–538
  7. Pierangeli SS , Girardi G , Vega-Ostertag M , Liu X , Espinola RG , Salmon J . Requirement of activation of complement C3 and C5 for antiphospholipid antibody-mediated thrombophilia . Arthritis Rheum . 2005;52:2120–2124
  8. Hart ML , Ceonzo KA , Shaffer LA , Takahashi K , Rother RP , Reenstra WR , et al.   Gastrointestinal ischemia-reperfusion injury is lectin complement pathway dependent without involving C1q . J Immunol . 2005;174:6373–6380
  9. Fleming SD , Egan RP , Chai C , Girardi G , Holers VM , Salmon J , et al.   Anti-phospholipid antibodies restore mesenteric ischemia/reperfusion-induced injury in complement receptor 2/complement receptor 1-deficient mice . J Immunol . 2004;173:7055–7061
  10. Piette J-C , Cervera R , Levy RA , Nasonov EL , Triplett DA , Shoenfeld Y . The catastrophic antiphospholipid syndrome—Asherson’s syndrome . Ann Med Interne (Paris) . 2003;154:195–196
  11. Bucciarelli S , Espinosa G , Asherson RA , Cervera R , Claver G , Gómez-Puerta JA , et al.   The acute respiratory distress syndrome in catastrophic antiphospholipid syndrome. Analysis of a series of 47 cases. . Ann Rheum Dis . 2005; (in press)
  12. Cervera R , Gómez-Puerta JA , Espinosa G , Font J , De la Red G , Gil V , et al.   “CAPS Registry” (a review of 200 cases from the International Registry of patients with catastrophic antiphospholipid syndrome (CAPS)) . Ann Rheum Dis . 2003;62(suppl 1):88
  13. Ghosh S , Walters HD , Joist JH , Osborne TG , Moore TL . Adult Respiratory Distress Syndrome associated with antiphospholipid—antibody syndrome . J Rheumatol . 1993;20:1406–1408
  14. Moser KM , Auger WR , Fedullo PF . Chronic major-vessel thromboembolic pulmonary hypertension . Circulation . 1990;81:1735–1743
  15. Jamieson SW , Auger WR , Fedullo PF , Channick RN , Kriett JM , Tarazzi RY , et al.   Experience and results with 150 pulmonary thromboendarterectomy operations over a 29-month period . J Thorac Cardiovasc Surg . 1993;106:116–127
  16. Sandoval J , Amigo MC , Barragon R , Izaguirre R , Reyes PA , Martinez Guerrero ML , et al.   Treatment with thromboendarterectomy . J Rheumatol . 1996;23:772–775
  17. Asherson RA , Mackworth-Young CG , Boey ML , Hull RG , Saunders A , Gharavi AE , et al.   Pulmonary hypertension in systemic lupus erythematosus . Br Med J . 1983;287:1024–1025
  18. Nagai H , Yasuma K , Katsuki T , Shimakura A , Usuda K , Nakamura Y , et al.   Primary antiphospholipid syndrome and pulmonary hypertension with prolonged survival . A case report. Angiology . 1997;48:183–187
  19. Asherson RA , Harris EN , Bernstein RM , Mackworth-Young CG , Hughes GRV . Immunological studies in “Primary” idiopathic pulmonary hypertension . Eur J. Rheumatol Inflamm . 1985;7:105
  20. Asherson RA , Cervera R , de Groot PG , Erkan D , Boffa MC , Piette JC , et al.   Catastrophic antiphosphospholipid syndrome (international consensus statement on classification criteria and treatment guidelines) . Lupus . 2003;12:530–534
  21. Espinosa G , Cervera R , Garcia-Carrasco M , Font J , Asherson RA . Pulmonary involvement in the antiphospholipid syndromes . In:  Asherson RA ,  Wells AU ,  Denton CP editor. In Handbook of Systemic Autoimmune Diseases; Pulmonary Involvement in the Systemic Autoimmune Diseases . Amsterdam: Elsevier Press; 2004;p. 137–143 (Series Editor)

PII: S0049-0172(05)00171-X

doi:10.1016/j.semarthrit.2005.08.006

Refers to article:

  • Antiphospholipid Antibodies as a Cause of Pulmonary Capillaritis and Diffuse Alveolar Hemorrhage: A Case Series and Literature Review

    Kevin D. Deane, Sterling G. West
    Seminars in Arthritis and Rheumatism December 2005 (Vol. 35, Issue 3, Pages 154-165)

Seminars in Arthritis and Rheumatism
Volume 35, Issue 3 , Pages 138-142, December 2005

Article Tools

* Image Usage & Resolution