- Email: [email protected]
Massive pulmonary hemorrhage in the newborn
T h e Journal of P E D I A T R I C S
3
Massive pulmonary hemorrhage in the nevborn I. Pathologic considerations
The type and severity o[ massive pulmonary hemorrhage diagnosed at autopsy were reviewed for a 10 year period. The frequency (percentage of total autopsies) and incidence (comparison with all livebirths) were determined. Alveolar and interstitial hemorrhage were equally frequent in infants who died during the first day of life, but alveolar hemorrhage predominated in those surviving for longer periods. Pulmonary hemorrhage occurred more frequently than expected in infants of low birth weight for gestational age.
John R. Esterly, M.D., and Ella H. Oppenheimer, M.D. BALTIMORE9
PULMONARY
MD.
HEMORRHAGE
is o n e
of the most common postmortem findings in the newborn infant, but its pathogenesis and significance remain poorly defined. This study evaluates the variations in morphology of massive pulmonary hemorrhage and compares the findings with some of the perinatal factors with which hemorrhage has been associated. The clinical findings in these cases are the subject of a separate report (Part II).1
MATERIALS AND METHODS A review was made of the histologic material from 758 liveborn infants who survived two weeks or less during the period from November 1, 1954, to October 31, 1964. The gross findings in the lungs and other From the Department of Pathology,'The Johns Hopkins University School o[ Medicine and The Johns HopkinsHospital.
organs were noted from the autopsy protocols. This group contained more than 95 per cent of the liveborn deaths during the first month of life. Stillbirths were examined separately because of the difficulty of distinguishing superimposed maceration with extravasation of blood from antemortem hemorrhage. The findings in 216 consecutive stillbirths autopsied in a 30 month interval during the 10 year period have been summarized and are included for comparison. Autolysis prevented diagnosis in about half of the cases. "Massive" hemorrhage, present in 135 cases, was defined as involvement of at least two lobes by confluent loci of extravasated red cells. Some areas of normal lung were present on about half of the slides; the remainder showed nearly uniform hemorrhage. (A longitudinal section of the entire lung is routine in autopsies of newborns.)
4 Esterly and Oppenheimer
July 1966
N o a t t e m p t was m a d e to assess the significance of the p u l m o n a r y h e m o r r h a g e in the d e a t h of the infant, although it was clearly a m a j o r factor in the m a j o r i t y of instances. T h e cases were evaluated for both alveolar a n d interstitial hemorrhage. I n the former the red cells till alveoli and extend into the bronchioles a n d bronchi (Fig. 1). I n the latter h e m o r r h a g e is present in the potential connective tissue spaces of the lung, i.e., the pleuraI, interlobular septal,
Fig. 1. Pure alveolar hemorrhage seen in a 28 week, 1360 gram infant who survived for two days. The extravasated erythrocytes fill peripheral airways and often extend into major bronchi. Red cells are few or absent in the alveolar wall, perivascular space (upper right), and interlobar septa (lower left). (Hematoxylin and eosin. Original magnification x240.)
peribronchial, and perivascular areas a n d the dissected alveolar wall (Figs. 2 a n d 3). T h e severity (including extent of h e m o r r h a g e ) was c o m p a r a b l e for the two types. Capillary engorgement of varying severity was found in almost all cases. OBSERVATIONS
T h e observations have been g r o u p e d into those descriptive of: (1) the p u l m o n a r y hemorrhage, (2) the infant, (3) concomi-
Fig. 2. Interstitial hemorrhage is present in the interlobar and perivascular spaces of a 27 week, 1310 gram infant who died in the first hours of Iife. The area between aIveoli represents interstitial extravasation of red ceils in alveolar septa. (Hematoxylin and eosin. Original magnification •
T a b l e I. T h e type of p u l m o n a r y h e m o r r h a g e in liveborn infants
Birth weight (Gin,) 0-999 1000-1999 2000-2499 2500 and over All weights
I Pure I Predominantly alveolar alveolar 10 27 3 9 49
3 13 4 16 36
Equal alveolar Predominantly and interstitial interstitial 2 0 0 5 7
4 8 2 6 20
Pure interstitial 3 14 1 5 23
I
Total 22 62 10 4I 135
Volume 69 Number 1
Pulmonary hemorrhage in newborn. Part I
tant pulmonary findings, and (4) stillbirths. 1. Puhnonary hemorrhage Type and severity. Table I shows no association between birth weight and the type of hemorrhage, The degree of severity was estimated separately for each type and could not be correlated with birth weight. Although these were the least objective data in the study, the observations were consistent on repeated occasions. Frequency and incidence. The frequency was calculated by comparing the number of instances of massive pulmonary hemorrhage with the total number of autopsied liveborn infants who survived two weeks or less. The results (Table I I ) show an increase-from less than 10 per cent in infants under 1000 grams to over 40 per cent in autopsied term infants. The incidence was determined by comparing the totai number of liveborn infants (1955 to 1962) with those born and autopsied in this hospital during those years (Table I I I ) . Although the frequency of pulmonary hemorrhage as a postmortem finding increases with birth weight, the incidence falls abruptly in infants over 2000
5
grams (Fig. 4). The annual incidence was calculated and ranged from 2.4 per 1000 to 6.6 per 1000. No trend or pattern was apparent. 2. The infant Gestational age and low birth we@hr. The estimated gestational age was known in 110 cases. The type of hemorrhage, when related to gestational age (Table I V ) , showed a distribution similar to that for birth weight (Table I ) . No association was found be-
Table II. The frequency of massive pulmonary hemorrhage in autopsied livebirths
Birth weight I (Gin.) ] Total 0-999 252 1000-1999 354 2000-2499 62 2500 and over 90 All weights 758
Pulmonary hemorrhage 22 62 10 41 135
Per cent 8.7 17.5 16.1 45.6 17.8
Fig. 3. Severe involvement of the alveolar wall is present in this 25 week, 700 gram infant with pure interstitial hemorrhage. Septal hemorrhage is often less apparent in areas of atelectasis and congestion. (Hematoxylin and eosin. Original magnification x135.)
Table III. The incidence of massive pulmonary hemorrhage in infants born at The Johns Hopkins Hospital, 1955-1962
Birth weight (Gin.) 0-999 1000-1999 2000-2499 2500 and over' Atlweights
I I
Total liveborn in[ants 227 730 1,683 22,180 24,820
Neonatal deaths 217 222 69 110 618
Massive pulmonary hemorrhage 15 46 9 25 95
Incidence per 1000 livebirths 66.1 63.0 5.3 1.1 3.8
6
Esterly and Oppenheimer
July 1966
Table IV. Gestational age of infants with massive p u l m o n a r y hemorrhage
Gestational age (weeks)
Predominantly alveolar
~30 30-33 34-37 38-41.5 41.5 All ages
23 13 7 22 4 69
FROUEC t
AT AUTOPSY
Equal alveolar and interstitial 1 0 0 4 0 5
ooA
Total
15 8 6 6 1 36
39 21 13 32 5 110
INC0EC0er I000 LIVEBIRTHS
75
50%
Predominantly interstitial
Loo -45
.15
1 2 3 BIRTH WEIGHT IN KILOGRAMS
Fig. 4. The frequency of massive pulmonary hemorrhage as a finding at autopsy is illustrated in birth weight groups by the shaded bars. For comparison the broken line demonstrates the incidence of this finding in all liveborn infants.
tween gestational age and the severity of hemorrhage. T h e birth weights of the infants of known gestational age were compared to the weight range for infants of the same length of gestationY, a A highly significant difference was observed between the numbers expected and observed with weights two standard deviations (S.D.) below the mean for their gestational age (Table V ) . N o particular gestational age was predominant, and low b i r t h weights were not associated with a specific type or severity of hemorrhage. Infants with erythroblastosis and multiple births were included in the 110 cases, but did not alter the distribution. Duration of survival. Forty per cent of
Table V. Gestational a g e - - b i r t h weight distribution of 110 infants
Birth weight 2 1 +- 1 1 2
S.D. S.D. S.D. S.D. S.D.
I No. observed [ No. expected 14 12 70 i0 4
2.5 15.0 75.0 15.0 2.5
the infants with massive pulmonary hemorrhage died on the first day of life, and twothirds survived tess than 48 hours. T h e changing proportion of alveolar hemorrhage is shown graphically in Fig. 5. Interstitial hemorrhage occurred in more than half the cases in the first day of life but was much less c o m m o n thereafter. T h e duration of survival was not related to birth weight or to the severity of hemorrhage. T h e duration of hemorrhage could not be estimated, and attempts to stain hemosiderin in instances of long survival, or in those with numerous alveolar macrophages or fibrin clots, were uniformly unsuccessful. Extrapulmonary hemorrhage. Table V I shows the frequency of bleeding in sites other than the lung. T h e thymus, skin, heart, and abdominal organs were the major sites of hemorrhagic lesions. T e n of the 14 infants weighing less than the second S.D. for gestational age had bleeding into the central nervous system; this was twice the frequency found in the group as a whole. Infection. Inflammation in extrapulmonary sites was present in eight infants, none of w h o m had pneumonia. Viral encephalitis and myocarditis were found in one infant, necrotizing esophagitis in another, and a
Volume 69 Number 1
Pulmonary hemorrhage in newborn. Part I
7
Table VI. E x t r a p u l m o n a r y hemorrhage Birth weight (Gin.) 0-999 1000-1999 2000-2499 2500 and over All weights
CNS 6 13 1 6 26
CNS and visceral 5 12 3 10 30
Visceral 3 15 3 7 28
Total 14 40 7 23 84
Series total 22 62 10 41 135
A 60
5O
.25
Number 40 of Cases 30
,20
20
I0
"15
10 0 . Duration of Lifein Days ]
.
.
.
.
.
.
.
.
; - - , r " 2 - I,
.
5
9
9
;
Stillbirths
13
Total
58 33
13 6
U
3
2
3
2
0
l
2
1
0
25
Alveolar
25 29
10 5
lO
2
1
0
1
0
1
1
0
0
12
Fig. 5. A, the number of infants who died on each day of the neonatal period is indicated by" the height of the top llne. The shaded areas represent cases of predominantly alveolar hemorrhage and illustrate the changing proportion of the type of hemorrhage after the first day of life. B, the proportion of predominantly interstitial hemorrhage in stillbirths is similar to that present in infants who survive less than 24 hours. third had congenital otitis. T h e remaining five infants had omphalitis. Pure alveolar hemorrhage was noted in seven of the eight cases. Erythroblastosis and kernieterus. During the 10 year period, etythroblastosis was diagnosed at autopsy in 23 infants. Massive pulmonary hemorrhage was found in six, four of w h o m had visceral and intracranial bleeding as well. Kernicterus was present in five infants, three of w h o m had pulmonary hemorrhage. Alveolar hemorrhage was present in the eight infants with erythroblastosis a n d / o r kernicterus. There was interstitial hemorrhage of equal or lesser severity in four. Malformations. Major congenital anomalies were present in 26 infants (19.3 per cent). Including multiple system defects, the largest categories were: cardiovascular, four-
teen; genitourinary, seven; and tracheoesophageal fistula, four. Corrective surgery was attempted in six of these infants, five of w h o m had severe alveolar hemorrhage at autopsy. Severe alveolar hemorrhage was present in the four infants with tracheoesophageal fistula. 3. Concomitant pulmonaryfindings Hyaline membranes. Peripheral atelectasis with focal air trapping and eosinophilic, laminated hyaline material (Fig. 6) was present in 28 infants (20.8 per cent) (Table V I I ) . Five of the sL,~ full-size infants with hyaline membranes weighed less than 2800 grams and at least one had a gestational age of only 34 weeks. T h e site of associated pulmonary hemorrhage and incidence of intracranial hemorrhage did not differ significantly from the series as a whole. Fibrin clots. Eosinophilic clotted fibrin was
8
Esterly and Oppenheimer
July 1966
Table VII. Additional pulmonary findings in infants with massive pulmonary hemorrhage Birth weight (Gm.) O-999 1000-1999 2000-2499 2500 and over All weights
Hyaline membranes 4 13 4 6 28
Fibrin clots 7 18 2 6 33
Fig. 6. Atelectasis, alveolar hemorrhage, and hyaline membranes shown in a 31 week, 1400 gram infant surviving 35 hours. Note the proximity of the membrane to the airway epithelium and its location in peripheral bronchioles. (Hematoxylin and eosin. Original magnification x250.)
identified in 33 instances (24.4 per cent) of severe alveolar hemorrhage. In sections stained with hematoxylin and eosin, the material had a superficial resemblance to hyaline membranes, but formed in the midst of the hemorrhage or exudate rather than lined bronchiolar walls (Fig. 7). The material was less homogeneous and had more rounded edges than hyaline membranes, which axe usually thin and elongated in outline. No particular pattern of aeration was noted in these cases. Fibrin staining clearly
Squames 5 9 5 15 34
Pneumonia 3 5 3 8 19
Series total 22 62 10 41 135
distinguished the negative or weakly staining hyaline membranes from the strongly positive fibrin clots (Fig. 8) when both were present. The factors common to cases with fibrin clots were alveolar hemorrhage and longer than average survival. Some clot formation was present as early as six hours; it was extensive only after 36 to 48 hours. Squames. Multiple loci of squames were identified in 34 infants (25.2 per cent), including three of the four with a gestational age of 42 weeks or longer. Aspiration of maternal blood could not be excluded histologically, but the proportion of cases with interstitial hemorrhage (17 of 34) makes this source an unlikely major component. Pneumonia. Histologic evidence of widespread pneumonia was present in only six infants; in another eight it was limited to a single site; and in five others there were focal areas of amorphous debris with an acute inflammatory infiltrate suggestive of aspiration (Table V I I ) . In all instances the hemorrhage was more severe and extensive than the inflammation. 4. Stillbirths. The frequency (15.1 per cent), distribution of weight groups, and the type of hemorrhage in stillbirths were similar to the findings in the liveborn infants who survived less than 24 hours (Fig. 5). Six of the seven infants in the 1000 to 1999 gram group had predominantly interstitial hemorrhage. In the other groups alveolar hemorrhage was more severe, although both components were present in the majority of cases. COMMENTS
Spencer 4 in 1891 distinguished between subpleural bleeding and hemorrhage into
Volume 69 Number 1
Pulmonary hemorrhage in newborn. Part I
9
Fig. 7. Alveolar hemorrhage with fibrin clot formation found in a 41 week, 1120 gram infant who survived 46 hours. Fibrin is commonly intermixed with red cells and present in the center of airways. Clots may be identified in all airway divisions (alveoli, bronchioles, and bronchi). (Hematoxylin and eosin. Original magnification x135.)
Fig. 8. Differential staining of fibrin clots in a bronchus and hyaline membranes (arrow) is illustrated in this section from a 28 week, 1220 gram infant who survived only 10 hours. (Phosphotungstic acid-hematoxylin. Original magnification x170.)
the "substance of the lung." Later, Ylpp5 "~ differentiated interstitial hemorrhage in an infant who died shortly after birth from alveolar hemorrhage found in infants of longer survival. The first clinical pathological study of "acute hemorrhagic pneumonia" by Browne G-s led him to postulate several etiologies and to conclude that it was a distinct entity. The association of pulmonary hemorrhage with fetal anoxia has long been recognized,~, ~0 but MacGregor ~1 has also commented on the frequency of hemorrhage in neonatal pneumonia. She concluded that asphyxia produced congestion and bleeding, a situation often favorable for bacterial proliferation. The frequency of massive pulmonary hemorrhage in reviews of neonatal pulmonary pathology has varied from 1 to 10 per cent? 2-~s In studies of pulmonary hemor-
rhage, similar frequencies are cited. 19-~2 This contrasts with our incidence of 17.8 per cent. However, the previous studies are not strictly comparable to each other or to the present investigation. Several authors confined their observations to alveolar hemorrhage, and all listed single diagnoses. It is probable that the majority o5 infants with massive bleeding and concomitant pneumonia or hyaline membrane disease were tabulated as the latter. Likewise, when all organ systems were reviewed, many instances of pulmonary hemorrhage were omitted if intracranial hemorrhage was also present. The differences in age of the infants, variation in autopsy population, and the pathologic definition of "massive" hemorrhage are probably not as important in influencing the frequency of pulmonary hemorrhage as is the exclusion of cases with other diagnoses. Only the study of Butler and Bon-
10
EsterIy and Oppenheimer
ham is permits comparison of the incidence on the basis of livebirths: 0.7 per 1000 compared to 3.8 per 1000 in our series. The most obvious reason for an elevated frequency is inclusion of cases with less severe involvement. The criteria used in this study are similar to those used by other authors. Nevertheless, if only the most severe cases are included (i.e., 3+ and 4+) the frequencies in Table I I change by only 25 per cent and remain well above the 8 per cent found by others. Another reason for a higher frequency in our study is the higher (10.9 per cent) incidence of premature births, compared with 6.7 per cent in the British study. A point of interest in Butler and Bonham's report was the greater than average gestational age for birth weight among infants with pulmonary hemorrhage. Similarly, the data of Thorburn 22 show four of 21 infants below the second S.D. in weight fo~ gestational age. The same trend is evident in our study (Table V ) . We believe that these differences in frequency and survival are related, and result from our inclusion of infants with massive interstitial and mixed hemorrhage who died shortly after birth. Death in infants with such lesions has usually been attributed to asphyxia. The type and frequency of pulmonary hemorrhage in infants with short survival closely resembles that in stillborn infants, and differs from that of the infant who survives for several daysY s No difference was noted in the severity of hemorrhage in these clinical extremes. Comparison of infants who die shortly after birth with nonmacerated stillbirths is also important, because in either situation death is close to the events of labor and delivery. We suggest that the similar findings in the two groups reflect common etiologic factors. The significance of the present study lies in the detailed examination of the type of hemorrhage in relation to the other factors investigated. Two overlapping variants are suggested when type of hemorrhage and duration of survival are compared: (1) the infant who dies shortly after birth has hemor-
July 1966
rhage of either type with equal frequency; (2) the infant surviving the first day of life is found to have predominantly alveolar hemorrhage. It appears that the type of hemorrhage may be independent of etiologic factors and unrelated to most of the other findings examined, and that duration of life is the most important factor in determining the location of pulmonary hemorrhage. SUMMARY
AND
CONCLUSIONS
This study describes the pathology of massive pulmonary hemorrhage in a group of 135 newborn infants surviving two weeks or less. The frequency was highest in termweight infants, but the incidence among all liveborn infants decreased sharply with increasing birth weight. A disproportionate number of infants of low birth weight (less than the second standard deviation for gestational age) was present in the series. Death occurred during the first day in nearly half the infants, and only one-third survived longer than 48 hours. Alveolar and interstitial hemorrhage were equally frequent in the infants who died shortly after birth, but alveolar hemorrhage predominated when the infants survived over 24 hours. Both were found in nearly half the infants, and the distinction between the two types had only a limited correlation with other morphologic findings. The frequency and types of pulmonary hemorrhage were similar in stillborn and liveborn infants surviving less than 24 hours. REFERENCES
1. Rowe, D. S., and Avery, M. E.: Massive pulmonary hemorrhage. II. Clinical considerations, J. PEDIAT.69: 12, 1966. 2. Gruenwald, P., and Minh, H. N.: Evaluation of body and organ weights in perinatal pathology. II. Weight of body and placenta of surviving and autopsied infants, Am. J. Obst. & Gynec. 82: 312, 1961. 3. Gruenwald, P.: Growth of the human fetus. I. Normal growth and its variation, Am. J. Obst. & Gynec. In press. 4. Spencer, H. R.: On visceral hemorrhages in stillborn children, Tr. Obst. Soc. London 33: 203, 1891. 5. YlppS, A.: Pathologish-anatomische Studien bei Friihgeboren, Ztschr. Kinderh. 20: 212, 1919.
Volume 69 Number 1
6. Browne, F. J.: Stillbirth: Its causes, pathology, and prevention, Edinburgh M. J. 27: 153, 1921. 7. Browne, F. J.: Pneumonia neonatorum, Brit. M. J. 1: 469, 1922. 8. Browne, F. J.: Neonatal death, Brit. M. J. 2: 590, 1922. 9. Cruickshank, J. N.: The causes of neonatal death. Special Report Series #145, Medical Research Council, London, 1930, His Majesty's Stationery Office, p. 62. 10. Clifford, S. I-I.: The effects of asphyxia on the newborn infant, J. PEDIAT. 18: 567, 1941. 11. MacGregor, A. R.: Pneumonia in the newborn, Arch. Dis. Child. 14: 323, 1939. 12. Arey, J. B., and Dent, J.: Causes of fetal and neonatal death with special reference to pulmonary and inflammatory lesions, J. PEDIAT. 42: 1, 1953. 13. Bound, J. P., Butler, N. R., and Spector, W. G.: Classification and causes of perinatal mortality, Brit. M. J. 2: 1191, 1956. 14. Sivanesan, S.: Neonatal pulmonary pathology in Singapore, J. P~mAT. 59: 600, 1961. 15. Driscoll, S. G., and Smith, C. A.: Neonatal
Puhnonary hemorrhage in newborn. Part I
16. 17. 18. 19. 20. 21.
22. 23.
11
pulmonary disorders, Pediat. Clin. North America 9: 325, 1962. Sant, S. M., and Purandare, N. M . : Pulmonary lesions in neonatal deaths, J. Postgrad. Med. (Bombay) 8: 8, 1962. Potter, E. L.: Pathology of prematurity, Clin. Obst. & Gynec. 7" 733, 1964. Butler, N. R., and Bonham, D. G.: Perinatal mortality, London, 1963, E. & S. Livingstone, Ltd. Ahvenainen, E. K., and Call, J. D.: Pulmonary hemorrhage in infants, a descriptive study, Am. J. Path. 28: 1, 1952. Ahvenainen, E. K.: Massive pulmonary hemorrhage in newborns, Ann. paediat. Fenniae 2: 44, 1956. Polleri, J. O., Matteo, A. L., Petruccelli, A. L., and Toledo, N.: Pulmonary hemorrhage in the newborn infant, Ann. paediat. Fenniae 3: 727, 1957. Thorburn, M. J.: Neonatal death and massive pulmonary hemorrhage in Jaxnaica, Arch. Dis. Child. 38: 589, 1963. Briggs, J. N.: Some obsetwations on puhnonary hemorrhage in the newborn period (abstract). J. Dis. Child. 99z 591, 1955.
3
Massive pulmonary hemorrhage in the nevborn I. Pathologic considerations
The type and severity o[ massive pulmonary hemorrhage diagnosed at autopsy were reviewed for a 10 year period. The frequency (percentage of total autopsies) and incidence (comparison with all livebirths) were determined. Alveolar and interstitial hemorrhage were equally frequent in infants who died during the first day of life, but alveolar hemorrhage predominated in those surviving for longer periods. Pulmonary hemorrhage occurred more frequently than expected in infants of low birth weight for gestational age.
John R. Esterly, M.D., and Ella H. Oppenheimer, M.D. BALTIMORE9
PULMONARY
MD.
HEMORRHAGE
is o n e
of the most common postmortem findings in the newborn infant, but its pathogenesis and significance remain poorly defined. This study evaluates the variations in morphology of massive pulmonary hemorrhage and compares the findings with some of the perinatal factors with which hemorrhage has been associated. The clinical findings in these cases are the subject of a separate report (Part II).1
MATERIALS AND METHODS A review was made of the histologic material from 758 liveborn infants who survived two weeks or less during the period from November 1, 1954, to October 31, 1964. The gross findings in the lungs and other From the Department of Pathology,'The Johns Hopkins University School o[ Medicine and The Johns HopkinsHospital.
organs were noted from the autopsy protocols. This group contained more than 95 per cent of the liveborn deaths during the first month of life. Stillbirths were examined separately because of the difficulty of distinguishing superimposed maceration with extravasation of blood from antemortem hemorrhage. The findings in 216 consecutive stillbirths autopsied in a 30 month interval during the 10 year period have been summarized and are included for comparison. Autolysis prevented diagnosis in about half of the cases. "Massive" hemorrhage, present in 135 cases, was defined as involvement of at least two lobes by confluent loci of extravasated red cells. Some areas of normal lung were present on about half of the slides; the remainder showed nearly uniform hemorrhage. (A longitudinal section of the entire lung is routine in autopsies of newborns.)
4 Esterly and Oppenheimer
July 1966
N o a t t e m p t was m a d e to assess the significance of the p u l m o n a r y h e m o r r h a g e in the d e a t h of the infant, although it was clearly a m a j o r factor in the m a j o r i t y of instances. T h e cases were evaluated for both alveolar a n d interstitial hemorrhage. I n the former the red cells till alveoli and extend into the bronchioles a n d bronchi (Fig. 1). I n the latter h e m o r r h a g e is present in the potential connective tissue spaces of the lung, i.e., the pleuraI, interlobular septal,
Fig. 1. Pure alveolar hemorrhage seen in a 28 week, 1360 gram infant who survived for two days. The extravasated erythrocytes fill peripheral airways and often extend into major bronchi. Red cells are few or absent in the alveolar wall, perivascular space (upper right), and interlobar septa (lower left). (Hematoxylin and eosin. Original magnification x240.)
peribronchial, and perivascular areas a n d the dissected alveolar wall (Figs. 2 a n d 3). T h e severity (including extent of h e m o r r h a g e ) was c o m p a r a b l e for the two types. Capillary engorgement of varying severity was found in almost all cases. OBSERVATIONS
T h e observations have been g r o u p e d into those descriptive of: (1) the p u l m o n a r y hemorrhage, (2) the infant, (3) concomi-
Fig. 2. Interstitial hemorrhage is present in the interlobar and perivascular spaces of a 27 week, 1310 gram infant who died in the first hours of Iife. The area between aIveoli represents interstitial extravasation of red ceils in alveolar septa. (Hematoxylin and eosin. Original magnification •
T a b l e I. T h e type of p u l m o n a r y h e m o r r h a g e in liveborn infants
Birth weight (Gin,) 0-999 1000-1999 2000-2499 2500 and over All weights
I Pure I Predominantly alveolar alveolar 10 27 3 9 49
3 13 4 16 36
Equal alveolar Predominantly and interstitial interstitial 2 0 0 5 7
4 8 2 6 20
Pure interstitial 3 14 1 5 23
I
Total 22 62 10 4I 135
Volume 69 Number 1
Pulmonary hemorrhage in newborn. Part I
tant pulmonary findings, and (4) stillbirths. 1. Puhnonary hemorrhage Type and severity. Table I shows no association between birth weight and the type of hemorrhage, The degree of severity was estimated separately for each type and could not be correlated with birth weight. Although these were the least objective data in the study, the observations were consistent on repeated occasions. Frequency and incidence. The frequency was calculated by comparing the number of instances of massive pulmonary hemorrhage with the total number of autopsied liveborn infants who survived two weeks or less. The results (Table I I ) show an increase-from less than 10 per cent in infants under 1000 grams to over 40 per cent in autopsied term infants. The incidence was determined by comparing the totai number of liveborn infants (1955 to 1962) with those born and autopsied in this hospital during those years (Table I I I ) . Although the frequency of pulmonary hemorrhage as a postmortem finding increases with birth weight, the incidence falls abruptly in infants over 2000
5
grams (Fig. 4). The annual incidence was calculated and ranged from 2.4 per 1000 to 6.6 per 1000. No trend or pattern was apparent. 2. The infant Gestational age and low birth we@hr. The estimated gestational age was known in 110 cases. The type of hemorrhage, when related to gestational age (Table I V ) , showed a distribution similar to that for birth weight (Table I ) . No association was found be-
Table II. The frequency of massive pulmonary hemorrhage in autopsied livebirths
Birth weight I (Gin.) ] Total 0-999 252 1000-1999 354 2000-2499 62 2500 and over 90 All weights 758
Pulmonary hemorrhage 22 62 10 41 135
Per cent 8.7 17.5 16.1 45.6 17.8
Fig. 3. Severe involvement of the alveolar wall is present in this 25 week, 700 gram infant with pure interstitial hemorrhage. Septal hemorrhage is often less apparent in areas of atelectasis and congestion. (Hematoxylin and eosin. Original magnification x135.)
Table III. The incidence of massive pulmonary hemorrhage in infants born at The Johns Hopkins Hospital, 1955-1962
Birth weight (Gin.) 0-999 1000-1999 2000-2499 2500 and over' Atlweights
I I
Total liveborn in[ants 227 730 1,683 22,180 24,820
Neonatal deaths 217 222 69 110 618
Massive pulmonary hemorrhage 15 46 9 25 95
Incidence per 1000 livebirths 66.1 63.0 5.3 1.1 3.8
6
Esterly and Oppenheimer
July 1966
Table IV. Gestational age of infants with massive p u l m o n a r y hemorrhage
Gestational age (weeks)
Predominantly alveolar
~30 30-33 34-37 38-41.5 41.5 All ages
23 13 7 22 4 69
FROUEC t
AT AUTOPSY
Equal alveolar and interstitial 1 0 0 4 0 5
ooA
Total
15 8 6 6 1 36
39 21 13 32 5 110
INC0EC0er I000 LIVEBIRTHS
75
50%
Predominantly interstitial
Loo -45
.15
1 2 3 BIRTH WEIGHT IN KILOGRAMS
Fig. 4. The frequency of massive pulmonary hemorrhage as a finding at autopsy is illustrated in birth weight groups by the shaded bars. For comparison the broken line demonstrates the incidence of this finding in all liveborn infants.
tween gestational age and the severity of hemorrhage. T h e birth weights of the infants of known gestational age were compared to the weight range for infants of the same length of gestationY, a A highly significant difference was observed between the numbers expected and observed with weights two standard deviations (S.D.) below the mean for their gestational age (Table V ) . N o particular gestational age was predominant, and low b i r t h weights were not associated with a specific type or severity of hemorrhage. Infants with erythroblastosis and multiple births were included in the 110 cases, but did not alter the distribution. Duration of survival. Forty per cent of
Table V. Gestational a g e - - b i r t h weight distribution of 110 infants
Birth weight 2 1 +- 1 1 2
S.D. S.D. S.D. S.D. S.D.
I No. observed [ No. expected 14 12 70 i0 4
2.5 15.0 75.0 15.0 2.5
the infants with massive pulmonary hemorrhage died on the first day of life, and twothirds survived tess than 48 hours. T h e changing proportion of alveolar hemorrhage is shown graphically in Fig. 5. Interstitial hemorrhage occurred in more than half the cases in the first day of life but was much less c o m m o n thereafter. T h e duration of survival was not related to birth weight or to the severity of hemorrhage. T h e duration of hemorrhage could not be estimated, and attempts to stain hemosiderin in instances of long survival, or in those with numerous alveolar macrophages or fibrin clots, were uniformly unsuccessful. Extrapulmonary hemorrhage. Table V I shows the frequency of bleeding in sites other than the lung. T h e thymus, skin, heart, and abdominal organs were the major sites of hemorrhagic lesions. T e n of the 14 infants weighing less than the second S.D. for gestational age had bleeding into the central nervous system; this was twice the frequency found in the group as a whole. Infection. Inflammation in extrapulmonary sites was present in eight infants, none of w h o m had pneumonia. Viral encephalitis and myocarditis were found in one infant, necrotizing esophagitis in another, and a
Volume 69 Number 1
Pulmonary hemorrhage in newborn. Part I
7
Table VI. E x t r a p u l m o n a r y hemorrhage Birth weight (Gin.) 0-999 1000-1999 2000-2499 2500 and over All weights
CNS 6 13 1 6 26
CNS and visceral 5 12 3 10 30
Visceral 3 15 3 7 28
Total 14 40 7 23 84
Series total 22 62 10 41 135
A 60
5O
.25
Number 40 of Cases 30
,20
20
I0
"15
10 0 . Duration of Lifein Days ]
.
.
.
.
.
.
.
.
; - - , r " 2 - I,
.
5
9
9
;
Stillbirths
13
Total
58 33
13 6
U
3
2
3
2
0
l
2
1
0
25
Alveolar
25 29
10 5
lO
2
1
0
1
0
1
1
0
0
12
Fig. 5. A, the number of infants who died on each day of the neonatal period is indicated by" the height of the top llne. The shaded areas represent cases of predominantly alveolar hemorrhage and illustrate the changing proportion of the type of hemorrhage after the first day of life. B, the proportion of predominantly interstitial hemorrhage in stillbirths is similar to that present in infants who survive less than 24 hours. third had congenital otitis. T h e remaining five infants had omphalitis. Pure alveolar hemorrhage was noted in seven of the eight cases. Erythroblastosis and kernieterus. During the 10 year period, etythroblastosis was diagnosed at autopsy in 23 infants. Massive pulmonary hemorrhage was found in six, four of w h o m had visceral and intracranial bleeding as well. Kernicterus was present in five infants, three of w h o m had pulmonary hemorrhage. Alveolar hemorrhage was present in the eight infants with erythroblastosis a n d / o r kernicterus. There was interstitial hemorrhage of equal or lesser severity in four. Malformations. Major congenital anomalies were present in 26 infants (19.3 per cent). Including multiple system defects, the largest categories were: cardiovascular, four-
teen; genitourinary, seven; and tracheoesophageal fistula, four. Corrective surgery was attempted in six of these infants, five of w h o m had severe alveolar hemorrhage at autopsy. Severe alveolar hemorrhage was present in the four infants with tracheoesophageal fistula. 3. Concomitant pulmonaryfindings Hyaline membranes. Peripheral atelectasis with focal air trapping and eosinophilic, laminated hyaline material (Fig. 6) was present in 28 infants (20.8 per cent) (Table V I I ) . Five of the sL,~ full-size infants with hyaline membranes weighed less than 2800 grams and at least one had a gestational age of only 34 weeks. T h e site of associated pulmonary hemorrhage and incidence of intracranial hemorrhage did not differ significantly from the series as a whole. Fibrin clots. Eosinophilic clotted fibrin was
8
Esterly and Oppenheimer
July 1966
Table VII. Additional pulmonary findings in infants with massive pulmonary hemorrhage Birth weight (Gm.) O-999 1000-1999 2000-2499 2500 and over All weights
Hyaline membranes 4 13 4 6 28
Fibrin clots 7 18 2 6 33
Fig. 6. Atelectasis, alveolar hemorrhage, and hyaline membranes shown in a 31 week, 1400 gram infant surviving 35 hours. Note the proximity of the membrane to the airway epithelium and its location in peripheral bronchioles. (Hematoxylin and eosin. Original magnification x250.)
identified in 33 instances (24.4 per cent) of severe alveolar hemorrhage. In sections stained with hematoxylin and eosin, the material had a superficial resemblance to hyaline membranes, but formed in the midst of the hemorrhage or exudate rather than lined bronchiolar walls (Fig. 7). The material was less homogeneous and had more rounded edges than hyaline membranes, which axe usually thin and elongated in outline. No particular pattern of aeration was noted in these cases. Fibrin staining clearly
Squames 5 9 5 15 34
Pneumonia 3 5 3 8 19
Series total 22 62 10 41 135
distinguished the negative or weakly staining hyaline membranes from the strongly positive fibrin clots (Fig. 8) when both were present. The factors common to cases with fibrin clots were alveolar hemorrhage and longer than average survival. Some clot formation was present as early as six hours; it was extensive only after 36 to 48 hours. Squames. Multiple loci of squames were identified in 34 infants (25.2 per cent), including three of the four with a gestational age of 42 weeks or longer. Aspiration of maternal blood could not be excluded histologically, but the proportion of cases with interstitial hemorrhage (17 of 34) makes this source an unlikely major component. Pneumonia. Histologic evidence of widespread pneumonia was present in only six infants; in another eight it was limited to a single site; and in five others there were focal areas of amorphous debris with an acute inflammatory infiltrate suggestive of aspiration (Table V I I ) . In all instances the hemorrhage was more severe and extensive than the inflammation. 4. Stillbirths. The frequency (15.1 per cent), distribution of weight groups, and the type of hemorrhage in stillbirths were similar to the findings in the liveborn infants who survived less than 24 hours (Fig. 5). Six of the seven infants in the 1000 to 1999 gram group had predominantly interstitial hemorrhage. In the other groups alveolar hemorrhage was more severe, although both components were present in the majority of cases. COMMENTS
Spencer 4 in 1891 distinguished between subpleural bleeding and hemorrhage into
Volume 69 Number 1
Pulmonary hemorrhage in newborn. Part I
9
Fig. 7. Alveolar hemorrhage with fibrin clot formation found in a 41 week, 1120 gram infant who survived 46 hours. Fibrin is commonly intermixed with red cells and present in the center of airways. Clots may be identified in all airway divisions (alveoli, bronchioles, and bronchi). (Hematoxylin and eosin. Original magnification x135.)
Fig. 8. Differential staining of fibrin clots in a bronchus and hyaline membranes (arrow) is illustrated in this section from a 28 week, 1220 gram infant who survived only 10 hours. (Phosphotungstic acid-hematoxylin. Original magnification x170.)
the "substance of the lung." Later, Ylpp5 "~ differentiated interstitial hemorrhage in an infant who died shortly after birth from alveolar hemorrhage found in infants of longer survival. The first clinical pathological study of "acute hemorrhagic pneumonia" by Browne G-s led him to postulate several etiologies and to conclude that it was a distinct entity. The association of pulmonary hemorrhage with fetal anoxia has long been recognized,~, ~0 but MacGregor ~1 has also commented on the frequency of hemorrhage in neonatal pneumonia. She concluded that asphyxia produced congestion and bleeding, a situation often favorable for bacterial proliferation. The frequency of massive pulmonary hemorrhage in reviews of neonatal pulmonary pathology has varied from 1 to 10 per cent? 2-~s In studies of pulmonary hemor-
rhage, similar frequencies are cited. 19-~2 This contrasts with our incidence of 17.8 per cent. However, the previous studies are not strictly comparable to each other or to the present investigation. Several authors confined their observations to alveolar hemorrhage, and all listed single diagnoses. It is probable that the majority o5 infants with massive bleeding and concomitant pneumonia or hyaline membrane disease were tabulated as the latter. Likewise, when all organ systems were reviewed, many instances of pulmonary hemorrhage were omitted if intracranial hemorrhage was also present. The differences in age of the infants, variation in autopsy population, and the pathologic definition of "massive" hemorrhage are probably not as important in influencing the frequency of pulmonary hemorrhage as is the exclusion of cases with other diagnoses. Only the study of Butler and Bon-
10
EsterIy and Oppenheimer
ham is permits comparison of the incidence on the basis of livebirths: 0.7 per 1000 compared to 3.8 per 1000 in our series. The most obvious reason for an elevated frequency is inclusion of cases with less severe involvement. The criteria used in this study are similar to those used by other authors. Nevertheless, if only the most severe cases are included (i.e., 3+ and 4+) the frequencies in Table I I change by only 25 per cent and remain well above the 8 per cent found by others. Another reason for a higher frequency in our study is the higher (10.9 per cent) incidence of premature births, compared with 6.7 per cent in the British study. A point of interest in Butler and Bonham's report was the greater than average gestational age for birth weight among infants with pulmonary hemorrhage. Similarly, the data of Thorburn 22 show four of 21 infants below the second S.D. in weight fo~ gestational age. The same trend is evident in our study (Table V ) . We believe that these differences in frequency and survival are related, and result from our inclusion of infants with massive interstitial and mixed hemorrhage who died shortly after birth. Death in infants with such lesions has usually been attributed to asphyxia. The type and frequency of pulmonary hemorrhage in infants with short survival closely resembles that in stillborn infants, and differs from that of the infant who survives for several daysY s No difference was noted in the severity of hemorrhage in these clinical extremes. Comparison of infants who die shortly after birth with nonmacerated stillbirths is also important, because in either situation death is close to the events of labor and delivery. We suggest that the similar findings in the two groups reflect common etiologic factors. The significance of the present study lies in the detailed examination of the type of hemorrhage in relation to the other factors investigated. Two overlapping variants are suggested when type of hemorrhage and duration of survival are compared: (1) the infant who dies shortly after birth has hemor-
July 1966
rhage of either type with equal frequency; (2) the infant surviving the first day of life is found to have predominantly alveolar hemorrhage. It appears that the type of hemorrhage may be independent of etiologic factors and unrelated to most of the other findings examined, and that duration of life is the most important factor in determining the location of pulmonary hemorrhage. SUMMARY
AND
CONCLUSIONS
This study describes the pathology of massive pulmonary hemorrhage in a group of 135 newborn infants surviving two weeks or less. The frequency was highest in termweight infants, but the incidence among all liveborn infants decreased sharply with increasing birth weight. A disproportionate number of infants of low birth weight (less than the second standard deviation for gestational age) was present in the series. Death occurred during the first day in nearly half the infants, and only one-third survived longer than 48 hours. Alveolar and interstitial hemorrhage were equally frequent in the infants who died shortly after birth, but alveolar hemorrhage predominated when the infants survived over 24 hours. Both were found in nearly half the infants, and the distinction between the two types had only a limited correlation with other morphologic findings. The frequency and types of pulmonary hemorrhage were similar in stillborn and liveborn infants surviving less than 24 hours. REFERENCES
1. Rowe, D. S., and Avery, M. E.: Massive pulmonary hemorrhage. II. Clinical considerations, J. PEDIAT.69: 12, 1966. 2. Gruenwald, P., and Minh, H. N.: Evaluation of body and organ weights in perinatal pathology. II. Weight of body and placenta of surviving and autopsied infants, Am. J. Obst. & Gynec. 82: 312, 1961. 3. Gruenwald, P.: Growth of the human fetus. I. Normal growth and its variation, Am. J. Obst. & Gynec. In press. 4. Spencer, H. R.: On visceral hemorrhages in stillborn children, Tr. Obst. Soc. London 33: 203, 1891. 5. YlppS, A.: Pathologish-anatomische Studien bei Friihgeboren, Ztschr. Kinderh. 20: 212, 1919.
Volume 69 Number 1
6. Browne, F. J.: Stillbirth: Its causes, pathology, and prevention, Edinburgh M. J. 27: 153, 1921. 7. Browne, F. J.: Pneumonia neonatorum, Brit. M. J. 1: 469, 1922. 8. Browne, F. J.: Neonatal death, Brit. M. J. 2: 590, 1922. 9. Cruickshank, J. N.: The causes of neonatal death. Special Report Series #145, Medical Research Council, London, 1930, His Majesty's Stationery Office, p. 62. 10. Clifford, S. I-I.: The effects of asphyxia on the newborn infant, J. PEDIAT. 18: 567, 1941. 11. MacGregor, A. R.: Pneumonia in the newborn, Arch. Dis. Child. 14: 323, 1939. 12. Arey, J. B., and Dent, J.: Causes of fetal and neonatal death with special reference to pulmonary and inflammatory lesions, J. PEDIAT. 42: 1, 1953. 13. Bound, J. P., Butler, N. R., and Spector, W. G.: Classification and causes of perinatal mortality, Brit. M. J. 2: 1191, 1956. 14. Sivanesan, S.: Neonatal pulmonary pathology in Singapore, J. P~mAT. 59: 600, 1961. 15. Driscoll, S. G., and Smith, C. A.: Neonatal
Puhnonary hemorrhage in newborn. Part I
16. 17. 18. 19. 20. 21.
22. 23.
11
pulmonary disorders, Pediat. Clin. North America 9: 325, 1962. Sant, S. M., and Purandare, N. M . : Pulmonary lesions in neonatal deaths, J. Postgrad. Med. (Bombay) 8: 8, 1962. Potter, E. L.: Pathology of prematurity, Clin. Obst. & Gynec. 7" 733, 1964. Butler, N. R., and Bonham, D. G.: Perinatal mortality, London, 1963, E. & S. Livingstone, Ltd. Ahvenainen, E. K., and Call, J. D.: Pulmonary hemorrhage in infants, a descriptive study, Am. J. Path. 28: 1, 1952. Ahvenainen, E. K.: Massive pulmonary hemorrhage in newborns, Ann. paediat. Fenniae 2: 44, 1956. Polleri, J. O., Matteo, A. L., Petruccelli, A. L., and Toledo, N.: Pulmonary hemorrhage in the newborn infant, Ann. paediat. Fenniae 3: 727, 1957. Thorburn, M. J.: Neonatal death and massive pulmonary hemorrhage in Jaxnaica, Arch. Dis. Child. 38: 589, 1963. Briggs, J. N.: Some obsetwations on puhnonary hemorrhage in the newborn period (abstract). J. Dis. Child. 99z 591, 1955.