Scand J Haematol(l975) 15,306-311

Mechanism of Anaemia in Experimental Bacterial Endocarditis ROBERT A. JOYCE,M.D.

&

MERLEA. SANDE, M.D.

Division of Haematology (Chief, D . R . Boggs), University of Pittsburgh School o f Medicine, Pittsburgh and Division o f Infectious Disease (Chief, G . Mandell), University of Virginia School of Medicine, Charlottesville, USA

Rabbits with Streptococcus viridans aortic valve endocarditis develop anaemia and reticulocytosis which increase with the duration of infection. Mean red cell counts decreased from 6.05 f 0.29 x 106 per p l before infection to 4.10 f 0.18 x 106 per p l after 11 t o 20 days of endocarditis and reticulocytes increased from 1.16 f 0.14 x 105 per p l t o 4.91 f 0.83 x 105 per p l after more than 20 days of endocarditis. The anaemia could not be explained by intravascular haemolysis. Anti-erythrocyte antibodies were not detected. Splenomegaly was a consistent finding and also increased with the duration of infection. Red cell half life (T%) was shortened to 4.7 f 0.3 days in rabbits with endocarditis compared with normal T % of 11.1 f 0.5 days. The T% of red cells from infected animals was prolonged when measured in noninfected rabbits and splenectomized animals had a mean red cell T % of 9.25 days after three weeks of infection. These studies suggest that splenic enlargement associated with infection results in red cell sequestration, a mechanism not well recognized as contributing to the anaemia of endocarditis. Key words: anaemia - endocarditis - extravascular haemolysis

Accepted for publication September 19, 1975 Correspondence to: Dr. Robert A. Joyce, Department of Medicine, School of Medicine, University of Pittsburgh, Pittsburgh, PA 15261, USA

Anaemia is common in patients with bacterial endocarditis (Parsons 1953), but the etiology has not been completely characterized. The anaemia which occurs with chronic infection is generally mild and is characterized by hypofunction of the erythroid

marrow (Bush et al 1956, Cartwright 1966). The anaemia in patients with endocarditis has also1 been associated with decreased serum iron and iron binding capacity, defective reutilization of iron, and impaired responsiveness of marrow heme synthesis to

A preliminary report was published in abstract form in CIin Res 22, 396 (abstr.).

ANAEMIA IN EXPERIMENTAL ENDOCARDITIS

erythropoietin (Haurani et a1 1965, Zucker et al 1974). In a rabbit model developed to study prolonged streptococcal endocarditis, anaemia and splenomegaly were detected, but were accompanied by reticulocytosis. The purpose of this study was to determine the nature of the anaemia and the role of the spleen in its development.

METHODS Endocarditis was induced in New Zealand white rabbits weighing 2 to 3 kg essentially as described by Perlman & Freedman (1971). Rabbits were anaesthetized with sodium pentobarbital, 40 to 60 mg i.v., and a polyethylene catheter (Intramedic, P E 90, Clay Adams, Parsippany, New Jersey) was inserted into the carotid artery and advanced into the left ventricle. The catheter was clamped off and tied to the artery with 00 silk suture and the skin incision was closed over the catheter. When haemotocrits and reticulocytes returned to preoperative levels, 106 Streptococcus viridans isolated from the blood of a patient with endocarditis were given t o the rabbits through a marginal ear vein. After 4 days the catheters were removed with negligible blood loss. Quantitative blood cultures and rectal temperatures were obtained daily. Positive blood cultures and temperatures greater than 39.6OC indicated bacterial endocarditis in all cases as described previously (Sande & Irvin 1974). To prevent early death from infection, 300,000 units of penicillin G (Wyeth Laboratories, Philadelphia, PA) were given i.m. daily or every other day when lo3 organisms or more per ml of blood were detected. This therapy suppressed the level of bacteraemia but was not sufficient t o cure the endocarditis. When the rabbits died, the aortic valvular area was examined t o verify the existence of endocarditis. Haematocrits and red cell and reticulocyte counts were obtained every two to five days during the infection. Red cell counts were determined using an electronic particle counter, Model S (Coulter Electronics, Inc., Hialeah, FL) and were done by micro technique t o minimize the volume of blood required. Blood smears were prepared at the same time and stained with Wright's stain to examine red cell morphology and with new

307

methylene blue for reticulocyte counts. Plasma haemoglobin was measured as described by Crosby & Furth (1956) when red cell survival studies were begun. The severity of haemolysis was measured by red cell T?h determination using W r (3.34 FCi per animal) as described by Sterling & Gray (1950). To evaluate the effect of splenic enlargement on the anaemia in infected rabbits, the T?h of blood from rabbits with endocarditis was measured in the donors and in noninfected rabbits. In further experiments, T % of red cells was measured after three weeks of endocarditis in rabbits splenectomized prior to infection. Less than 1 ml of blood was required on the days of study except during TYz determinations when not more than 2 ml were removed. In the longest surviving rabbits with endocarditis from which the most blood samples were obtained, this represented less than 15 ml blood during a four week period of study or a total of less than 8 % of the mean blood volume determined in seven noninfected rabbits weighing 2-3 kg.

RESULTS

Initially, 11 rabbits with streptococcal endocarditis of varying duration were examined to determine if anaemia occurred during infection. In 7 noninfected rabbits mean haematocrit was 40f0.8 % (SE); mean reticulocytes, 1.8k 0.2 % (SE); and mean spleen weight, 0.95f0.10 g (SE). In rabbits with endocarditis mean haematocrit was 30+ 1.4 % (SE). All infected rabbits had reticulocytosis with a mean of 1Ok 1.8 % (SE). Spleens from 4 rabbits, weighed when the animals died, were all enlarged. Mean weight was 3.12 k 1.27 g (SE). To determine when red cell changes occurred during endocarditis, 12 rabbits were studied sequentially during the infection (Table I). All rabbits became anaemic and developed reticulocytosis. The reticulocytosis and severity of anaemia increased with the duration of infection. Haematocrit and 20"

308

ROBERT A. JOYCE & MERLE A. SANDE TABLE I Effect of duration of endocarditis on anaemia and reticulocytosis

*

Duration of endocarditis (days)

Haematocrit

0 1- 5 6-10 11-20 > 20

3 8 . 2 f 0 . 5 (12)" 35.3 f:0.8 (12) 29.5 k 0.8 (10) 26.7 k 0.5 ( 9) 28.6 k 0.6 ( 5)

(%)

Reticulocytes

Red cell count (x 10Vpl)

(%)

6.05 f0.29 (10) 5.37k0.20 (10) 4.76k0.17 ( 8) 4.10k0.18 ( 7) 4.25 f0.23 ( 3)

1 . 9 f 0 . 2 (12) 3.1 k0.3 (11) 4.7 a 0.7 (10) 9.2 f 0.8 ( 9) 8 . 4 f 0 . 7 ( 5)

I

-

(x1Wlpl) 1.16 f 0.14 (10) 1.77 a 0.16 ( 9) 2.19k0.35 ( 8) 3.59 k 0.40 ( 7) 4.91 f 0 . 8 3 ( 3)

Mean k SE of determinations made at 2 to 5 day intervals from 12 rabbits with endocarditis; numbers in parentheses represent animals surviving and studied during the time interval noted.

red cell counts fell rapidly during the first 10 days and then tended to stabilize. Mean reticulocyte counts progressively increased, however, to more than 4 times baseline values after 20 days of infection. No gastrointestinal blood loss was detected from repeated stool analyses with Guaiac. Similar levels of anaemia and reticulocytosis were reached in animals with high fever and bacteraemia as in those with little or no fever and low levels of bacteraemia. Blood smears demonstrated polychromasia and variations in red cell size as well as basophilic stippling and occasional nucleated red blood cells when reticulocyte counts increased, but there was no significant variation in red cell indices compared with baseline values determined prior to infection. Blood smears did not demonstrate red cell fragmentation, plasma haemoglobin levels were not elevated above 4.1 mg per 100 ml and urine haemosiderin was not detected. As an additional means of excluding intravascular haemolysis, renal tissue from 5 rabbits was examined microscopically and only one animal had detectable tubular haemosiderin. Antiglobulin tests were done using sheep anti-rabbit antisera and serum anti-bodies against rabbit erythrocytes were not demonstrated. In separate experiments groups of rab-

bits with endocarditis were killed during the same intervals of endocarditis as above and spleen weights were recorded (Figure 1). Mean spleen weight also increased with the duration of endocarditis reaching 4.8& 0.72 g (SE) in the group with endocarditis longer than 20 days. Only congestion was evident in histologic examination of the enlarged spleens. The relationships between the level of anaemia and reticulocytosis and the splenic enlargement are illustrated in

5

0

1-5 6-10 11-20 > 2 0 DURATION OF INFECTION ( D A Y S )

Figure 1. Relationship of spleen weight to duration of endocarditis. Values represent the mean f SE of 7 (day O), 6 (days 1-5), 19 (days 6-10), 9 (days 11-20), and 6 (> 20 days) animals.

ANAEMIA IN EXPERIMENTAL ENDOCARDITIS

Figure 2. The calculated regression lines demonstrate an inverse correlation of red cell counts with spleen weight during endocarditis (r = - 0.964k0.046) and a positive correlatioa of reticulocytosis with spleen

+

309

weight (r = 0.941 0.059). Both correlations are statistically highly significant (P

Mechanism of anaemia in experimental bacterial endocarditis.

Rabits with Streptococcus viridans aortic valve endocarditis develop anaemia and reticulocytosis which increase with the duration of infection. Mean r...
412KB Sizes 0 Downloads 0 Views