Pediatric and Developmental Pathology 2, 440–445, 1999
Pediatric and Developmental Pathology
r1999 Society for Pediatric Pathology
Vocal Cord Basement Membrane in Non-Sudden Infant Death Syndrome Cases EUMEˆ NIA C.C. CASTRO1
AND
L. CESAR PERES2*
1Disciplina
de Patologia, Departamento de Cieˆncias Biolo´gicas, Faculdade de Medicina do Triaˆngulo Mineiro, Prac¸a Manoel Terra s/no, Centro, 38015-050, Uberaba, Minas Gerais, Brazil 2Departamento de Patologia, Faculdade de Medicina de Ribeira ˜ o Preto, 14049-900, Universidade de Sa˜o Paulo, Ribeira˜o Preto, SP, Brazil Received June 19, 1998; accepted November 16, 1998.
ABSTRACT
INTRODUCTION
Vocal cord basement membrane thickening (VCBMT) has been observed in children with sudden infant death syndrome (SIDS). It has been proposed that this lesion could be used as a positive indicator of this syndrome in autopsies of children who have died unexpectedly. The present investigation aimed to analyze vocal cord basement membranes from autopsies of children 0 to 365 days old. A total of 134 larynges were analyzed. Histological sections of paraffin-embedded larynges stained with H&E and submitted to histochemical staining with periodic acid–Schiff (PAS), Masson’s trichrome, syrius red, and Carstairs were used for light microscopy analysis. Immunohistochemistry with monoclonal anti-collagen IV antibody was used to determine the nature of VCBMT. The study was completed with morphometry of H&E– and PAS-stained sections and revision of the clinical information contained in the hospital files. VCBMT was found in 25 cases (18.7%) and showed characteristics of normal basement membrane, including immunoreactivity to collagen IV. Our data support the conclusions that VCBMT is frequently seen in pediatric autopsies, is seen in children in all age-groups studied whose deaths were due to causes other than SIDS, and is commonly associated with infectious diseases. Like SIDS, VCBMT occurs in the first year of life.
In recent years, vocal cord membrane thickening has received the attention of authors who work with sudden infant death syndrome (SIDS). In their view, the thickening seen in children dying unexpectedly could be related to SIDS and so this lesion could be used as a positive criterion for its pathological diagnosis [1]. Vocal cord lesions have been found in children who died of other causes and were interpreted as fibrinoid necrosis [2,3] or fibrosis [4]. Basement membrane deposition and remodeling are not fully understood and the mechanisms involved in both physiological and pathological phenomena could bring clues for the understanding of its significance and relation to the underlying problems. The aim of this study was to identify vocal cord basement membrane thickening (VCBMT) in children from 0 to 365 days of age whose deaths were due to causes other than SIDS in order to understand its morphology, composition, and related factors.
Key words: vocal cord, basement membrane, non-SIDS, SIDS, sudden infant death syndrome, non-sudden infant death syndrome
The study consisted of 134 larynges from children less than 365 days of age at the time of death that were submitted to postmortem examination at Hospital das Clı´nicas and Servic¸o de Verificac¸a˜o de ´ bitos do Interior (SVOI), University of Sa˜o Paulo, O
*Corresponding author
METHODS
Table 1.
Morphometric analysis of vocal cords, age, sex, race, and place of autopsy in non-SIDS cases Thickened basement membrane
Vocal cord thickness (m)a Age
(days)b
Female Male White Non-white HC SVOI
Non-thickened basement membrane
5.1 ⫾ 2.1 m
n (%)
1.1 ⫾ 0.3 m
173.4 ⫾ 102.3
57.6 ⫾ 71.6
17c (26.2)
48 (73.8)
65 (100)
7 (10.3)
61 (89.7)
68 (100)
18d (16.1)
94 (83.9)
112 (100)
6 (46.2)
7 (53.8)
13 (100)
11 (11.3)
86 (88.7)
97 (100)
21 (60)
35 (100)
14e (40)
´ bitos do Interior. Numbers in parentheses are percents. HC, Hospital das Clı´nicas; SVOI, Servic¸o de Verificac¸a˜o de O aKruskall-Wallis, H ⫽ 79.16; P ⬍ 0.0001. bKruskall-Wallis, H ⫽ 47.3; P ⬍ 0.001. c2 Yates corrected ⫽ 4.63; P ⫽ 0.03. d2 Yates corrected ⫽ 4.99; P ⫽ 0.02. e2 Yates corrected ⫽ 11.96; P ⬍ 0.001.
Ribeira˜o Preto, Brazil. All deaths were due to defined causes other than SIDS and none was sudden. Larynges were fixed in formalin, cut along a median sagittal plane and then along a coronal plane perpendicular to the vocal cords, and routinely processed. Twenty 5-m-thick histological serial sections were taken from the paraffinembedded blocks, stained with hematoxylin and eosin (H&E), and submitted to histochemical staining with periodic acid–Schiff (PAS), Masson’s trichrome (MT), syrius red (SR), and Carstairs for light microscopy analysis and immunohistochemistry, using the streptavidin method with monoclonal anti–collagen IV antibody (Dako, Carpinteria, CA) to determine the nature of basement membrane thickening. Computer-assisted morphometry (MOPVideoplan, Kontron Elektronic, Munich, Germany) of H&E– and PAS–stained sections was used to determine the thickness of basement membranes. Measurements were made at five different points with a 40⫻ objective and the arithmetic mean was used for analysis. The basement membrane was considered thickened when the mean was above the control plus 2 times the standard deviation (1.13 ⫾ 2 ⫻ 0.26 m). The study was completed with revision of the clinical information contained in the hospital files. EPI-INFO 6.0 software was used to analyse the data, with the level of significance being at 5%.
RESULTS Basement membrane thickening was found in 25 (18.7%) of the 134 cases studied. In the remaining 109 cases (81.3%), the basement membrane was not thickened. The mean thickness of the basement membrane was 1.1 ⫾ 0.3 m in the normal cases and 5.1 ⫾ 2.1 m in the thickened ones. The measurements ranged from 5.2 m to 11.0 m in the thickened cases whereas in controls there was less variation (1.1 ⫾ 0.3 m). There was no statistically significant difference in basement membrane thickening between histological sections stained with H&E and those stained with PAS. The mean age of VCBMT cases was 5.2 months, and the condition was more frequent among females and non-white children. With respect to the origin of the cases, VCBMT was observed in 40% of SVOI’s cases and only in 11.3% of hospital cases (Table 1). The postmortem diagnoses in the VCBMT cases are listed in Table 2. Infection was responsible for 12 deaths, corresponding to 35.3% (P ⬍ 0.016). Congenital heart disease was seen in only five cases, three of which exhibited VCBMT, corresponding to 60% (P ⬍ 0.04). Histologic and histochemical analysis of the vocal cord basement membrane showed that it reacted in the same way in the two groups (Fig 1A,B,D,E), indicating that it is made up of intrinsic VOCAL CORD
IN
NON-SIDS CASES
441
Table 2.
Vocal cord basement membrane thickness and postmortem diagnosis Thickened
Postmortem diagnosis
n
Infectious diseasesa
Non-thickened (%)
n
Sum (%)
n
(%)
12
(35.3)
22
(63.7)
34
(100)
Heart diseases
3
(60)
2
(40)
5
(100)
Diarrhea
2
(100)
0
2
(100)
Prematurity
2
(5.9)
32
(94.1)
34
(100)
Perinatal anoxia
1
(3.8)
25
(96.2)
26
(100)
Congenital defects
1
(8.4)
11
(91.6)
12
(100)
Cerebral edema
1
(25)
3
(75)
4
(100)
Aspiration
1
(33.3)
2
(66.6)
3
(100)
Extrahepatic biliary atresia
1
(100)
0
1
(100)
Malnutrition
1
(100)
0
1
(100)
Pulmonary edema
1
(100)
0
1
(100)
25
(18.6)
101
134
(100)
Total
(75.4)
2 ⫽ 56, P ⬍ 0.0001. a2 ⫽ 5.7; P ⫽ 0.016.
components. This is confirmed by the immunoreactivity to anti–collagen IV antibody (Fig 1C,F).
DISCUSSION Histological examination of the larynges with their vocal cords is part of the routine neonatal postmortem examination [5], is highly recommended in the investigation of unexpected deaths [6], and is facultative in older children with no suspicion of SIDS. Many lesions can be found, reflecting local or systemic problems important for the interpretation of the other findings. Among them, VCBMT has recently received more attention because of its identification in SIDS cases. The supposed specificity of this lesion in such cases was determinant in the development of this study because there is no reasonable explanation for this fact and also because this lesion can be found in children dying of other causes. SIDS is virtually unknown in Brazil, as there are only two studies reporting identification of such cases [7,8]. Because we have many autopsies of the same age-group, we conducted this prospective study to assess the morphological characteristics of VCBMT as well as the possible mechanisms involved in or associated with it, to try to correlate them with the cause of death. Our cases were distributed in the first year of life and the total number (134) was adequate for the study. In previous studies conducted by Cullity 442
E.C.C. CASTRO
AND
L.C. PERES
and Emery [4], Valdes-Dapena et al. [2], and Risse et al. [3], the number of cases was 102, 100, and 56, respectively. The first study included only cases of known causes of death, the second was based on cases of unexpected deaths, among which some may have been SIDS, and the third included proven SIDS and non-SIDS cases. VCBMT occurred frequently in our cases (18.7%), a finding that differs from that of Shatz et al. for control cases, whose deaths were due to known causes [1]. Our results, however, are supported by those presented by Cullity and Emery [4], ValdezDapena [2], and Rise et al. [3], who have shown that VCBMT and other lesions are seen in non-SIDS cases. VCBMT was seen more frequently in female babies. Although this difference was statistically significant in the present study, there is no plausible explanation for this fact because sex distribution was similar and the events related to the deaths did not show any preponderance between sexes. On the other hand, there are no data in the pertinent literature to allow a comparison. The age in our cases was distributed along the first year of life with a mean age of 2.2 month (⫾3.3) but in the cases with VCBMT, the mean age was 5.8 months (⫾4.3). This finding may indicate that the lesion is progressive and is present during the same period of incidence of SIDS, a result that
Figure 1. A–C: Paraffin-embedded, 5-µm sections of control vocal cords. H&E staining shows squamous epithelium over a thin, regular basement membrane (A), better seen with PAS staining (B). The basement membrane reacted as a thin line just beneath the basal cells of the epithelium with immunostaining for collagen-IV (C) (original magnification, ⫻40). D–F: Paraffin-embedded,
5-µm sections of thickened vocal cord basement membrane. H&E staining shows squamous epithelium over a thick, regular basement membrane (D), better seen with PAS staining (E). The basement membrane reacted as a thick, somewhat irregular line just beneath the basal cells of the epithelium with immunostaining for collagen-IV (F) (original magnification, ⫻40).
VOCAL CORD
IN
NON-SIDS CASES
443
may cause some misinterpretation when dealing with SIDS without a proper control group. With respect to the origin of the cases, VCBMT occurred more frequently among those postmortem from SVOI, which is responsible for autopsies of cases that come from the community. These deaths are usually unexpected and without medical supervision. This means that intubation or other manipulations were unlikely and the lesion probably was due to other mechanisms. If SIDS cases were to be identified in our city it is expected that they would be found among the SVOI cases instead of those from the hospital. Nevertheless, all cases analyzed in the present study had a defined cause of death. The association of VCBMT with infectious diseases, among which meningoencephalitis and pneumonia were more frequent, was highly significant. Local trauma due to coughing could be expected in the latter but not in the former. Longer intubation time, which is expected to occur in such cases, could be a reasonable explanation. However, intubation was not associated with VCBMT in our cases. There are few studies reporting VCBMT with which to compare our data. Our results are similar to those of Shatz et al. [1,9] and Rise et al. [3] with respect to the number of cases with thickened membranes. However, there is a significant difference in relation to the control cases, because in Shatz et al.’s report [9] the authors presented a very small thickness. This explains why these authors found the cases with thickened membranes to have VCBMT that was 30 to 100 times greater than that of the controls. This difference between the two studies is too large and is apparently due to a misinterpretation by Shatz et al. [9], since the measure they presented for controls was close to that obtained by electron microscopy, based on basal lamina thickness. The basement membrane examined by light microscopy includes other constituents, resulting in a larger measure, which is similar to our results. In addition, the mean thickness of the control basement membrane in their cases was 0.015 m, far less than the optical resolution of the best light microscope, which is around 0.2 m [10]. The intrinsic mechanisms involved in the genesis of the VCBMT secreted by epithelial cells is 444
E.C.C. CASTRO
AND
L.C. PERES
not fully understood. Basement membrane production is very slow, being estimated at 0.25 m per month in rat crystalline [11]. Epithelial regeneration is followed by normal basement membrane deposition in most cases and the basement membrane is destroyed and remodeled under many different physiological and pathological conditions [12]. In cases of basement membrane thickening, an abnormal, intrinsic modulation of this process could be present. Another possible explanation for basement membrane thickening in these cases could be a particularity of basement membrane production and remodeling of the vocal cord. The basement membrane of amniotic epithelium is frequently thickened in areas of squamous metaplasia, although it is also seen in apparently normal areas, as well as in placentas of diabetic women [13]. Duplication of blood vessel basement membrane is observed in cases of porphyria cutanea tarda, possibly indicating the existence of diverse mechanisms in its genesis. Thus, biochemical, inflammatory, metabolic, and irritant factors, among others, may play a role in basement membrane epithelium thickening, some of which could be involved in VCBMT. The age distribution of the cases with VCBMT during the first year of life is not regular. Sixty-eight percent of the cases are seen between the third and eighth months, and 40% of them are between the third and fourth months. There is a temporal superimposition with SIDS cases, although there is no evidence of a common mechanism. A difficulty is the definition of a good or true control for SIDS. Probably SIDS victims are not completely normal and so could have disorders, whatever they are, that would facilitate VCBMT as in our cases. There are some indications that at least some SIDS victims show a particular cry [14] that could induce phonotrauma [3], and mild upper respiratory inflammation is also quite common in such children and could induce vocal cord lesion due to coughing [15,16]. The reason why some children have VCBMT will be better understood when we know how the basement membrane of vocal cord is produced and remodeled. It is only speculative to relate VCBMT to any disease or condition before we answer this question. In summary, the data presented in this study show that VCBMT is frequent in pediatric autop-
sies, is seen in all age-groups studied, and is made up of intrinsic components of the basement membrane. VCBMT is observed in children whose deaths were due to causes other than SIDS, usually infectious diseases, and like SIDS, VCBMT occurrs during the first year of life.
7.
8.
ACKNOW LEDG M E N T S We are grateful to Prof. Dr. Floreˆncio Figueiredo for the immunohistochemical study and to Mrs. Deise Lucia Chesca Simo˜es for her technical assistance. This work was conducted at the Departamento de Patologia, Faculdade de Medicina de Ribeira˜o Preto, Universidade de Sa˜o Paulo, Ribeira˜o Preto, Sa˜o Paulo, Brazil.
9.
10.
11.
12.
REFERENCES 1.
2. 3.
4.
5.
6.
Shatz A, Hiss J, Hammel I, Arensburg B, Variend S. Age-related basement membrane thickening of the vocal cords in sudden infant death syndrome (SIDS). Laryngoscope 1994;104:865–868. Valdes-Dapena M. Crib deaths and focal fibrinoid necrosis of the infant larynx. J Forensic Sci 1958;3:503–509. Risse M, Weiler G, Glanz H, Dreyer Th. Histomorphologic studies of the glottic region to ellucidate the pathogenesis and functional significance of fibrinoid vocal cord lesions in infants and young children. Int J Ped Otorhinolaryngol 1997;42:125–134. Cullity GJ, Emery JL. Ulceration and necrosis of vocal cords in hospital and unexpected child deaths. J Pathol 1975;115:27–31. Keeling JW. The perinatal necropsy. In: Keeling JW, ed. Fetal and Neonatal Pathology, 2nd ed. New York: SpringerVerlag, 1993;1–32. Valdes-Dapena M, McFeeley PA, Hoffman HJ, et al. Death investigation and post-mortem examination. In: Valde´r-
13.
14.
15.
16.
Dapina M, McFeeley PA, Hoffman HJ, Damus KH, Franciosi RR, Allison DJ, Jones M, Hunter JC, eds. Histopathology Atlas for the Sudden Infant Death Syndrome. Washington, DC: Armed Forces Institute of Pathology, 1993;23–37. Victora CG, Nobre LC, Lombardi C, et al. Quadro epidemiolo´gico das mortes su´bitas na infaˆncia em cidades gau´chas. Rev Saude Publica 1987;21:490–496. Peres LC. Sudden infant death syndrome in Ribeira˜o Preto, Brazil. Rev. Paul. Med 1998;116:1803–1807. Shatz A, Hiss Y, Arensburg B. Basement-membrane thickening of the vocal cords in sudden infant death syndrome. Laryngoscope 1991;101:484–486. Junqueira LC, Carneiro J. Me´todos de estudo. In: Junqueira LC, Carreiro J, eds. Histologia Ba´sica, 7th ed. Rio de Janeiro: Guanabara-Koogan, 1990;1–20. Haddad A, Bennett G. Synthesis of lens capsule and plasma membrane glycoproteins by lens epithelial cells and fibers in the rat. Am J Anat 1988;183:212–225. Alberts B, Bray D, Lewis J, Raff M, Roberts K, Watson JD. Cell junctions, cell adhesion, and the extracellular matrix. In: Alberts B, Bray D, Lewis J, Raff M, Roberts K, Watson JD, eds. Molecular Biology of the Cell, 3rd ed. New York: Garland Publishing, 1994;97:971–995. Benirschke K, Kauffman P. Maternal diseases complicating pregnancy: diabetes, tumors, preeclampsia, lupus anticoagulant. In: Benirschke K, Kaufmann P, eds. Pathology of the Human Placenta, 3rd ed. New York: Springer-Verlag, 1995;476–515. Corwin MJ, Lester BM, Sepkoski C, Peucker M, Kayne H, Golub H. Newborn acoustic cry characteristic of infants subsequently dying of sudden infant death syndrome. Pediatrics 1995;96:73–77. Lindgren C, Lin J, Graham BS, et al. Respiratory syncytial virus infection enhances the response to laryngeal chemostimulation and inhibits arousal from sleep in young lambs. Acta Pediatr 1996;85:789–797. Alessandri LM, Read AW, Burton PR, Stanley FL. Sudden infant death syndrome in Australian aboriginal and nonaboriginal infants: an analytical comparison. Paediatr Perinat Epidemiol 1996;10:309–318.
VOCAL CORD
IN
NON-SIDS CASES
445