These anti¬ bodies have been eluted and shown to be cross-reactive, i.e. the anti- glomerular basement membrane. (GBM) antibody reacts with lung basement ...
Bilateral nephrectomy for massive pulmonary hemorrhage in
Goodpasture's syndrome M. Silverman, m.d., f.r.c.p.[c],* D. Hawkins, m.d., f.r.c.p.[c]** and C. F. Douglas Ackman, m.d., f.r.c.s.[c],*** Toronto
several weeks' duration. On physical examination she appeared a thin, pale, young white woman. Blood pressure supine was 120/70; radial pulse, 116/ min. and regular; respiration, 18/ min.; oral temperature, 100°F.; weight, 47.8 kg. There was conjunc¬ tival and mucous membrane pallor, but the remainder of the findings were within normal limits. Results of laboratory investigation on admission were as follows: hemoglobin, 5.2 g./
100 ml.; hematocrit, 14%; erythro¬ cytes, 1.74 x 107c.mm.; reticulocytes, leukocytes, 7500/c. mm. with Summary: A case of Goodpasture's has been produced in non-human 3.8%; adult 62% neutrophils, 1% eosin¬ in is described which eluted from syndrome primates by antibody bilateral nephrectomy was the kidneys of a patient with Good¬ ophils, 34% lymphocytes and 3% platelets, 208,000/c.mm. undertaken because of massive pasture's syndrome.8 It is not known monocytes; showed a specific gravity Urinalysis whether the antibody is directed of 1.006, 2+ pulmonary hemorrhage. 0 sugar, pH 6; protein, Similar cases recorded in the primarily against lung or glomerular on microscopic examination there literature are reviewed. basement membrane, and the role were many erythrocytes, some leuko¬ of antibody in the production of cytes, occasional epithelial cells, oval Various hypotheses to explain the beneficial effects of renal pulmonary disease is not as clearly fat bodies, 2-3 erythrocyte casts and ablation on lung purpura are established as it is in the pathogene¬ occasional granular and blood casts considered. It is suggested that the sis of the renal lesion. per high power field. Chest x-ray a diffuse increase in lung of in The is pulmonary hemorrhage purpose this report to showed which was interstitial and al¬ density is document the of dramatic mediated Goodpasture's syndrome response veolar in location. The BUN was 14 in part by a non-antibody a patient with massive pulmonary serum creatinine, 1.6 ml., mg./lOO humoral factor with permeabilityhemorrhage to bilateral nephrec¬ ml.; sodium, 140 mEq./l.; increasing properties that is tomy and to review the clinical ex¬ mg./lOO 4.1 mEq./l., chloride 105 potassium released from the nephritic kidney. perience with similar cases in the mEq./l.; and biocarbonate 24 mEq./l. literature. The creatinine clearance was 33 l./day and the urine protein excretion was Goodpasture's syndrome is a dis¬ Case report 3.5. g./day. The ASO titre was within tinct clinical and pathological en¬ The an 18-year-old girl, was normal limits. patient, tity consisting of pulmonary hem¬ referred to the Montreal General Hos¬ An open biopsy was per¬ orrhage and glomerulonephritis.12 In pital on July 17, 1970 for investiga¬ formed whichrenal showed all glomeruli recent years immunologic phenom¬ tion of anemia and glomerulitis. There to be affected by disease (Fig. 1). In ena have been increasingly implicat- was no history of hemoptysis. The some there were segmental ed in the pathogenesis of this dis¬ main symptoms on admission were tive and exudative lesions, prolifera¬ while in order.3"5 In both kidney and lung, weakness, headaches and dizziness of others the process was diffuse. There immunoglobulins and often comple¬ ment are found localized along the basement membrane. These anti¬ bodies have been eluted and shown to be cross-reactive, i.e. the antiglomerular basement membrane (GBM) antibody reacts with lung basement membrane (LBM) in vitro, and the antibody eluted from lung binds in a linear pattern to the GBM.3 Furthermore, severe glom¬ erulonephritis of the anti-GBM type Presented in preliminary form at the Annual Meeting of the Royal College of Physicians and Surgeons of Canada in January 1971. ?Department of Medicine, Division of Nephrology, Toronto General Hospital. ??Department of Medicine, Division of Clinical Immunology, Montreal General Hospital. ???Department of Urology, Montreal General Hospital. Reprint requests to: Dr. M. Silverman, Division of Nephrology, Toronto General Hospital, Toronto, Ont.
FIG. 1. Hematoxylin and eosin-stained section of renal and sclerotic glomerular lesions. xl52.
336 C.M.A. JOURNAL/FEBRUARY 3,
1973/VOL. 108
biopsy showing proliferative
were foci of fibrinoid necrosis, adhe¬ sions of tuft to capsule and epithelial crescent formation. Some glomerular tufts were completely sclerosed while in others the sclerosis was segmental. Lymphocytes and plasma cells surrounded many of the tufts but there was no granuloma formation. There was some atrophy of tubules associ¬ ated with interstitial fibrosis but no significant vessel abnormality was de¬ tected. The diagnostic impression was
in
hospital there was no hemoptysis; evidence of fluctuating interstitial and alveolar infiltrates (Fig. 3). On the evening of the 29th hospital day the patient experienced the sud¬ den onset of severe hemoptysis for the first time. At that time the leukocyte count was 8300/c. mm with a nor¬ mal differential; the platelet count was also normal. The bleeding was accompanied by a fall in hematocrit of generalized and segmental proli¬ from 40 to 25%. A chest film taken ferative and necrotizing glomerulone¬ shortly after the onset of hemoptysis shows the rather extensive nature of phritis. Immunofluorescent staining of renal the intrapulmonary hemorrhage (Fig. tissue demonstrated deposition of host 4). During the period of massive IgG and complement (C3) in a con¬ bleeding and in the next five days the tinuous, fine, linear fashion along the leukocyte count varied from 4700 to GBM. The pattern was typical of the 6400/c. mm. Over the next two days the pa¬ presence of anti-glomerular basement membrane antibody. tient's condition deteriorated rapidly. Marked clinical improvement fol¬ By the 31st hospital day severe hypo¬ lowed transfusions of packed red xemia necessitated intubation and con¬ blood cells. A course of prednisone, tinuous automatic ventilation. 40 mg. daily, and azathioprine, 100 Because of the life-threatening na¬ mg. daily, was initiated in an attempt ture of the intrapulmonary hemorr¬ to treat the active glomerulitis. As hage, and despite the fact that the shown in Fig. 2, this therapy had no however, there was definite radiologic
observable effect on either the 24hour protein excretion or the creatin¬ ine clearance. During the first 29 days
O Q0
15
20
25
Days jn Hospital FIG. 2.Creatinine clearance and 24 hours* protein excretion during
prior to nephrectomy; of hemoptysis.
H
=
onset
platelet
was 86,000/c. operation.
count
the time of
mm.
at
In order to assess and document the response to bilateral nephrectomy we undertook a retrospective analysis of some of the clinical parameters which were monitored during the
perioperative period. Physical examination On the basis of multiple bedside ex¬ aminations the following clinical ob¬ servations were made: (1) Preoperatively there was con¬ tinuous pulmonary bleeding as evidenced by suctioning of
fresh blood from the endo¬ tracheal tube. (2) From the time the patient was returned from the operating room there was a dramatic stabilization in vital signs, and fresh blood could no longer be recovered from the endotra¬ cheal tube. Hematocrit variation Fig. 5 is a graphical presentation of the patient's hematocrit. The level on admission of 14% was increased to 40% by transfusion. It remained stable until the occurrence of hemop¬ tysis when it fell to 25%. Over the next 40 hours it was maintained at about this level by multiple packed cell transfusions. The immediate post¬ operative hematocrit showed a sub¬ stantial rise to 35% without further administration of blood. We interpret this rise as evidence that the pul¬ monary hemorrhage stopped imme¬ diately after nephrectomy.
Oxygenation period FIG. 4.Chest radiograph, Aug. 15 (29th hospital day), several hours after onset of hemoptysis.
FIG. 3.Left: admission chest film.
(28th hospital day).
creatinine was 2.1 mg./lOO ml., elected to carry out an emergency bilateral nephrectomy 42 hours after the first episode of hemoptysis. The serum
we
Right: chest radiograph, Aug. 14
The degree of arterial hypoxemia in the pre- and postoperative periods is shown in Fig. 6. Also indicated are estimates of the Po2 of the inspired air which were determined from the concentration of oxygen being admin¬ istered at the time the blood gases were drawn. It is evident that pre¬ operatively the arterial Po2 was main¬ tained at 25 to 50 mm. Hg only by progressively increasing the Po2 of the inspired air. Almost immediately post¬ operatively this trend was reversed. Although the Po2 of the inspired air was gradually lowered, the arterial Po2 remained above 60 mm. Hg. In other words, the Po2 gradient across the alveolar capillary barrier that was required to maintain an arterial Po2 of 50 mm. Hg was rapidly reduced postnephrectomy. We interpret this as an¬ other indication of cessation of pul¬ monary hemorrhage.
C.M.A. JOURNAL/FEBRUARY 3, 1973/VOL. 108 337
Radiologic examination kidney. Postoperatively there was cases there is strong evidence that in pulmonary function renal ablation alleviates the pulmon¬ Fig. 7 shows two portable chest films. improvement as evidenced by an increase in ar¬ ary hemorrhage in Goodpasture's The one on the left was taken imme¬
diately preoperatively while that on right was taken about 36 hours postoperatively. The radiologic clearing that is apparent in the second film is further evidence that the pul¬ monary hemorrhage stopped after re¬ moval of the kidneys. In summary, each of these clinical parameters points to the fact that the massive pulmonary hemorrhage in this case of Goodpasture's syndrome ceased rather dramatically after bi¬ lateral nephrectomy. Our observa¬ tions also suggest that the response was rapid within minutes to hours of the removal of the second kidney. Despite this dramatic clinical im¬ provement postnephrectomy the pa¬ tient succumbed to gram-negative septicemia on the sixth postoperative day. the
.
Discussion To our knowledge, since 1967 there have been six other reported cases of Goodpasture's syndrome in which bilateral nephrectomy was
terial Po* and
a
reduction in the
intrapulmonary sequestration of erythrocytes that had been docu¬ mented prior to nephrectomy. These studies were carried out in the sec¬ ond and third weeks postnephrec¬ tomy and there is no way of know¬ ing how soon after kidney removal the pulmonary hemorrhage ceased. Of importance is the fact that there has been no recurrence of pulmon¬ ary bleeding three and one-half years after transplantation. Cleveland et aP° report the trans¬ plant of a cadaver kidney in a pa¬ tient with Goodpasture's syndrome who had terminal uremia. A bilat¬ eral nephrectomy was performed prior to transplantation. There is no mention as to whether this pa¬ tient showed any evidence of pul¬ monary hemorrhage preoperatively. At the time of publication (Novem¬ ber 1964) the patient had survived 16 months and a second transplant was being contemplated because of chronic rejection. It is perhaps sig¬ nificant that the authors do not re¬ port any pulmonary hemorrhage complicating the postoperative course. On the basis of these reported
syndrome. This suggests that there is a definite pathophysiologic inter¬ action between the lung and kidney in this disease and not just a demon¬ strable antigenic similarity of base¬ ment membranes in the two organs. Whether this interaction develops at some time during the course of the illness or is an integral part of the disease from the time of onset cannot be stated at the present time. On the basis of their experience Maddock and associates9 have sug¬ gested that the pulmonary mani¬ festations may actually be a conse¬ quence of the nephritis. In a similar vein, clinical observations surround¬ ing a case of Goodpasture's syn¬ drome associated with influenza As virus infection led to the suggestion that pneumonitis might initiate anti¬ body response by causing an anti¬ genic alteration of GBM.12 The in¬ fection could also perpetuate the pulmonary component of the dis¬ ease by exposing normally hidden lung basement membrane to circu¬ lating antibody or immune com¬
performed.7"11 In addition, we are aware of three cases encountered plexes. In a recent paper Siegel8 has dis¬ by Dr. D. L. Shires (personal com¬ cussed two hypotheses that might munication). In seven of these nine cases the bilateral nephrectomy was undertaken because of life-threatening pulmonary hemorrhage, and in two, including our own, the renal function was only mildly impaired at the time of nephrectomy. One patient succumbed within 12 hours of nephrectomy but in this instance there was complete consolidation of both lungs preoperatively. In every FIG. 6.Variation of arterial Po2 preother case there was a dramatic cesand postnephrectomy. Numbers on sation of lung purpura within 24 to graph refer to estimated Poa of inspired 48 hours of operation, and the pa¬ FIG. 5.Variation of hematocrit preair at the time of blood gas tients were subsequently treated and postnephrectomy. determinations. with chronic dialysis or renal trans¬ plantation. Recurrence of pulmon¬ ary bleeding postnephrectomy has been documented in only one in¬ stance. In this case intermittent minor hemorrhages recurred for a six-month period; however, there has been no evidence of lung pur¬ pura in the last year. It is noteworthy that this patient (presently maintained on dialysis) has demon¬ strable serum levels of anti-GBM antibodies despite complete healing of the lung lesions. In the report of Siegel8 bilateral nephrectomy was performed after FIG. 7.Left: portable chest radiograph taken immediately preoperatively. homograft rejection of a cadaver Right: portable chest radiograph taken about 36 hours postnephrectomy. 338 C.M.A. JOURNAL/FEBRUARY 3, 1973/VOL. 108
explain the interaction between lung and kidney in Goodpasture's syn¬ drome. He first rejects the explana¬ tion that antibody with anti-LBM activity could alone be responsible for the intrapulmonary bleeding, since anti-basement membrane anti¬ body levels are higher following nephrectomy (i.e. after removal of the large antigen mass of GBM).6 We agree with this conclusion. As an alternative he proposes the fol¬ lowing explanation. Active nephritis releases GBM antigen into the renal vein where it is transported to the pulmonary capillary bed. In the presence of circulating antibody with anti-LBM specificity, immune complexes are formed and deposited along the LBM, thus giving rise to the observed lung pathology. This hypothesis has been extended to include the possibility that anti-lung (rather than anti-glomerular) base¬ ment membrane antibodies may be injurious to the lung after complexing with circulating GBM antigen released from necrotic glomeruli.11 There are at least two major criticisms of Siegel's hypothesis. First, although significant amounts of GBM antigen may reach the lung, the levels of circulating anti-base¬ ment membrane antibodies tend to be so low as to be usually undetectable and therefore large numbers of immune complexes are not like¬ ly to be formed. Furthermore, some complexes should also reach the GBM and be filtered out there. However, no such phenomenon has been observed. The second major objection to the proposal by Siegel is that, despite the focal distribution of immunofluorescence in the lung, the pattern of deposition of immune reactants is linear. It is identical to that which is found in the glomeruli and is virtually pathognomonic of Ag Lung B.M. =
anti-basement
membrane
anti¬ and the adenine nucleotides.13 Ex¬ amples of known agents with vasoactive properties that might be implicated are the kinins and prosta¬
body.8'4 The focal distribution along the LBM may be related to a vari¬ able and inhomogeneous pulmonary capillary blood flow or to peculiarities of alveolar capillary barrier per¬ meability. Whatever the explanation for the focal distribution it cannot be equated with the "lumpy-bumpy" or granular pattern that is char¬ acteristic of the presence of im¬ mune complexes. Therefore, there does not appear to be justification for ascribing the link between neph¬ ritis and the development of pul¬ monary hemorrhage to the deposi¬ tion of immune complexes in the lung. One other fact must be taken into consideration in any proposal aimed at explaining the pulmonary hemorrhage in Goodpasture's syn¬ drome. There is no morphologic evi¬ dence of destruction of any com¬ ponent of the alveolar capillary bar¬ rier endothelial lining cells, basement membrane or alveolar lin¬ ing. Bearing in mind the foregoing remarks, we would like to propose as a working hypothesis the mechan¬ ism illustrated schematically in Fig. 8 to explain the interaction between the lung and the kidney in Good¬ pasture's syndrome. We postulate that the antigenantibody interaction which takes in a place in the glomeruli results nonrelease into the renal vein of a antibody humoral agent with permeability-increasing properties. This humoral factor reaches the pulmon¬ ary circulation in high concentra¬ tion where it increases the perme¬ ability of the alveolar capillary bar¬ rier. The presence of antibody with LBM specificity is probably neces¬ sary in order for the humoral factor to exert its full effects. A corollary of this hypothesis is that the nephri¬ tis in Goodpasture's syndrome is the cause of the pulmonary hemorrhage. In order to satisfy the clinical ob¬ servation that the response of pul¬ monary hemorrhage to bilateral within nephrectomy is rapid the postulated minutes to hours .
glandins. It may be possible to test this hypothesis by taking pre- and post¬ operative blood samples from the renal vein of patients with Good¬ pasture's syndrome who are under¬ going bilateral nephrectomy. It might then be determined if a fac¬ tor or factors capable of increasing vascular permeability were present. Measurement of kinins, prostaglan¬ dins and other vasoactive substances could also be performed. As of March 1969 about 100 cases of Goodpasture's syndrome
have been recorded in the literature, 56 of these since 1965.14 To our knowledge the present case is the eighth instance of Goodpasture's syndrome in which bilateral neph¬ rectomy was undertaken because of massive lung purpura and the sec¬ ond case where the renal function was relatively normal at the time of the bilateral nephrectomy. In seven of these cases a definite, dramatic, beneficial response of pulmonary hemorrhage has been recorded fol¬ lowing removal of both kidneys. The majority of such patients have then received renal transplants. Re¬ sults available from the transplant registry (in 1971) show a one-year kidney survival of 52, 77 and 79% in grafts from cadaver, parental and sibling donors, respectively.15 Pa¬ tient survival statistics are not avail¬ able; however, they must be at least as good as the transplant survival rates. In contrast, for 43 patients with Goodpasture's syndrome re¬ ported by Proskey et al14 the mean survival was 10.2 months and just under 50% died of pulmonary hemorrhage. Proskey reported that corticosteroids prolonged life from 5.7 to 16.8 months when compared to a control group. In cases of life-
threatening pulmonary hemorrhage, however, there is no evidence to Permeability suggest that steroids will prolong survival.9 Bilateral nephrectomy is a drasAg Glomerular B.M. humoral agent must have a short tic procedure under any circum¬ half-life in the circulation. It may, stances; however, our experience in fact, be inactivated by enzymes and that of others reviewed to date FIG. 8.Diagrammatic representation located on the luminal surface of indicate that lung purpura in Good¬ of proposed hypothesis to explain the endothelial lining of the lung; pasture's syndrome can be checked interaction of kidney and lung in mechanism has been sug¬ by removal of both kidneys. Al¬ Goodpasture's syndrome. Ag antigen, a similar to gested explain the metabolism of though we do not advocate bilateral Ab antibody, BM basement membrane. circulating angiotensin I, bradykinin nephrectomy in every case, such inIncreased
Pulmonary Capillary
.
.
=
=
=
=
C.M.A. JOURNAL/FEBRUARY 3, 1973/VOL. 108 339
tervention might be contemplated References EW: The significance earlier in those instances in which 1. GOoDPASTURE lesions in relaof certain pulmonary pulmonary hemorrhage becomes tion to the etiology of influenza. Am manifest. J Med Sci 158: 863, 1919 Resumen
Nephrectomie bilaterale pratiquee pour hemorragie pulmonaire massive dans un syndronm de Goodpasture L'auteur presente un cas de syndrome de Goodpasture qui a impose une nephrectomie bilaterale en presence d'une hemorragie pulmonaire massive. II passe en revue des cas similaires publies dans la litterature medicale. II 'tudie les diverses hypotheses permettant d'expliquer les effets benefiques de l'ablation renale sur le purpura pulmonaire. II estime que l'hemorragie pulmonaire dans le syndrome de Goodpasture peut etre declenchee partiellement par un facteur humoral (qui n'est pas un anticorps) dote de la propriete d'augmenter la permeabilite et qui est libere par le rein nephretique.
2. BENOIT FL, RULON DB, THIL GB, et al: Goodpasture's syndrome: a clinicopathologic entity. Am J Med 37: 424, 1964 3. KOFFLER D, SANDSON J, CARR R, et al: Immunologic studies concerning the pulmonary lesions in Goodpasture's syndrome. Am J Pathol 54: 293, 1970 4. BEIRNE GJ, OCTAVIANO GN, KoPP WL, et al: Immunohistology of the lung in Goodpasture's syndrome. Ann Intern Med 69: 1207, 1968 5. Posiurr TR: Immunologic and electron microscopic studies in Goodpasture's syndrome. Am J Med 49: 250, 1970 6. LERNER RA, GLASSOCK RJ, DIXON FJ: The role of antiglomerular basement membrane antibody in pathogenesis of human glomerulonephritis J Exp Med 126: 989, 1967 7. SHRES DL, PFAFF WW, DEQUESADA A, et al: Pulmonary hemorrhage and glomerulonephritis; treatment of two cases by bilateral nephrectomy and renal transplantation. Arch Surg 97: 699, 1968 8. SIEGEL RR: The basis of pulmonary disease resolution after nephrectomy
9.
10.
11.
12.
13.
14.
15.
in Goodpasture's syndrome. Am J Med Sci 259: 201, 1970 MADOCK RK JR, STEVENS LE, REETSMA K, et al: Goodpasture's syndrome: cessation of pulmonary hemorrhage after bilateral nephrectomy. Ann Intern Med 67: 1258, 1967 CLEVELAND RJ, LEE HM, PRouT GR, et al: Preservation of the cadaver kidney for renal homotransplantation in man. Surg Gynecol Obstet 119: 991, 1964 NowAKowsKI A, GROVE RB, LEROY HK JR, et al: Goodpasture's syndrome: recovery from severe pulmonary hemorrhage after bilateral nephrectomy. Ann Intern Med 75: 243, 1971 WILsoN CB, SMITH RC: Goodpasture's syndrome associated with influenza A. virus infection. Ann Intern Med 76: 91, 1972 RYAN JW, SMITH U, NIEMEYER RS: Angiotensin I: metabolism by plasma membrane of lung. Science 176: 64, 1972 PRosKEY AJ, WEATHERBEE L, EASTERLING RE, et al: Goodpasture's syndrome. A report of five cases and review of the literature. Am J Med 48: 162, 1970 MuRuRY JE, BARNES BA, ATKiNSON JC: Eighth report of the human kidney transplant registry. Transplantation HI: 328, 1971
Before you select aRegistered Retirement Savings Plan fund, you owe it to yourself o get all these facts. Performance record of Canadian Equity and Income funds currently qualified for Registered Retirement Savings Plans: 1. Ranked by performance (up to 10 years) 2. Growth of investment 3. Rate of return Pirbeck, well known to major institutional investors, is now offering a new measurement facility to the individual investor. Pirbeck does not sell mutual funds, solicit investments, or act as an advisory service. This is the latest factual performance record, issued quarterly.
I
"These two lessons experience has taught me that getting the facts of a situation before acting is of crucial importance, and that getting these facts is a continual job which requires eternal vigilance." Bernard Baruch 340 C.M.A. JOURNAL/FEBRUARY 3, 1973/VOL. 108
m m U Send your Canadian R.R.S.P. Funds Performance Survey: January, 1973 - results to December 1972 $10.00 Z1 All four 1973 Quarterly Issues $30.00 DOI My cheque or money order is enclosed.
NAME
ADDRESS CITY
PROV.
Pirbeck Investment Measurement Ltd.
806 Sun Life Building
Montreal 110, Quebec, Canada.
