Acquired Cytomegalovirus Infection
in
Preterm Infants Roberta A. Ballard, MD; W. Lawrence Drew, MD, PhD; Karen G.
\s=b\Small sick preterm infants requiring
in a neonatal intensive care unit for more than 28 days were cultured for cytomegalovirus in urine and/or nasopharynx during two periods lasting a total of 13 months. Sixteen of 51 such infants began excreting the virus at 28 to 148 days of age (mean, 55 days). In 14 of the 16, a recognizable, self-limited symptom complex developed that consisted of care
respiratory deterioration, hepatosplenomegaly, a remarkable gray pallor, and both an atypical and absolute lymphocy-
tosis. All of the infants with the clinical symptom complex had underlying chronic lung disease and all had received multiple blood transfusions during their hospitalization. Acquired cytomegalovirus may be relatively common in sick preterm infants and should be distinguished from other causes of rapid deterioration. (Am J Dis Child 133:482-485, 1979)
clinical of cy¬ infection have
incidence, features, The sequelae congenital tomegalovirus (CMV) and
been studied
extensively and re¬ viewed by numerous authors.1-7 In contrast, the impact of acquired CMV infection in the newborn is less clear.
From the Department of Pediatrics (Drs Ballard, Hufnagle, and Riedel) and Pathology and Laboratory Medicine (Dr Drew), Mount Zion Hospital and Medical Center, San Francisco. Dr Hufnagle is now at the William Beaumont Hospital, Royal Oak, Mich. Reprint requests to Department of Pediatrics, Mount Zion Hospital and Medical Center, PO Box 7921, San Francisco, CA 94120 (Dr Ballard).
Hufnagle, MD; Philip
A. Riedel, MD
Granstrom et alR has reported that newborns who acquired the disease perinatally (^ 5 months of age) have few symptoms and develop normally; however, protracted inter¬ stitial pneumonitis has been reported in two term infants9 who apparently acquired the virus from the maternal genital tract. In the immunologically compromised host, the disease may be severe, with hepatitis, hemolytic anemia, pneumonitis, myocarditis and pericarditis, and possibly involvement of the CNS.1 Because sick preterm infants are immunologically com¬ promised, they might be considered candidates for the more serious disease. Since 1968, there have been occa¬ sional reports of serious CMV infec¬ tion in preterm infants in whom the illness was presumed to be acquired rather than congenital.101- In July 1975, we noted a distinct syndrome associated with CMV infection in three sick, preterm infants. After this observation, we initiated a systematic survey for the incidence and clinical manifestations of CMV infection in preterm infants and found 16 infants to be excreting the virus.13 This report describes 14 of these infants in whom a recognizable symptom complex due to CMV infection developed that was thought to be acquired in the inten¬ sive care nursery.
healthy
Downloaded From: http://archpedi.jamanetwork.com/ by a Oakland University User on 06/08/2015
METHODS Patient Population The Mount Zion Hospital intensive care nursery is a 16-bed unit with 300 admis¬ sions a year of which 50% are inborn and 50% are transported from outlying hospi¬ tals. The population is composed primarily of preterm infants with hyaline membrane disease. After identification of three
cytomegalovirus disease, two prevalence of CMV coloniza¬ done. During the first survey,
infants with surveys for
tion were 100 consecutive sick infants admitted to the unit for at least four days during a seven-month period underwent weekly cul¬ tures of urine and nasopharynx for virus. Since none of these infants excreted virus prior to 28 days of age, in the second phase of the survey, weekly cultures were taken from all infants who had been in the unit for more than 28 days. Twenty-one addi¬ tional infants met the criteria for screen¬ ing during a six-month period.
Viral Studies Urine specimens were transported to the laboratory within four hours of being obtained; 10 mL of urine were centrifuged at 1,000 g for 15 to 20 minutes. The sedi¬ ment was resuspended in 0.5 mL of super¬ natant fluid and 0.2 mL of this was added to the media of each of two tubes of WI-38
cells. Cultures of WI-38 were observed for 38 days to detect CMV cytopathogenic effect. Throat cultures were obtained with ster¬ ile cotton-tipped swabs that were immedi¬ ately immersed in a tube of WI-38 tissue culture. The tube was capped and sent to
virology laboratory. After approxi¬ mately one hour for adsorption of the virus to the tissue culture, the swab was removed and 0.2-mL aliquots of fluid were removed for inoculation into tubes of primary monkey kidney, primary human embryonic kidney, and Hep-2 tissue cultures. Only the
WI-38 cultures were inoculated if the infant was not acutely ill; the additional cultures were inoculated if the infant was sick. Autopsy cultures were prepared by mincing centimeter-sized portions of vari¬ ous organs in Hanks' balanced salt solution containing gentamicin sulfate and amphotericin B. The specimen was then sonically disrupted in a sonic oscillator at 10,000 Hz for 40 s. It was then centrifuged at 1,000 g for five minutes, and 0.2 mL of superna¬ tant fluid was inoculated per tube of tissue culture. Cultures of pooled breast milk were performed exactly as for urine.
REPORT OF A CASE This infant of 1,100-g birthweight and 29 weeks' gestation was born to a 25-year-old, gravida 4, para 2 mother with fever and foul-smelling amniotic fluid and had Apgar scores of 4 and 5 at one and five minutes, respectively. The infant had respiratory distress at birth and required ventilation with 80% oxygen. At 5 hours of age, the infant was transported to Mount Zion Hospital with a respirator setting of 0.8 FIO.,, a pressure of 22/7 cm H,0, and a ventilation rate of 40/min. The infant had a
complicated course requiring a peak pres¬
H,0 and FIO, of 0.7. Chest at 10 days of age showed early development of bronchopulmonary dysplasia. The infant's condition improved and then stabilized with a ventilator setting of FIO, 0.35 and pressure of 24/6 cm H ,0 for the next four weeks. Because of multiple blood studies, the infant received frequent blood transfusions of 10 to 15 mL of packed RBCs obtained from 19 different donors. At 42 days of age, respiratory status worsened, requiring an increase in FIO, from 0.35 to 0.85 and an increase in inspiratory pressure from 24 to 30 cm H,0 over a period of nine days. At this time, the infant had hepatosplenomegaly, a remark¬ able gray pallor, and looked "septic"; feed¬ ing was poor. Laboratory studies showed a lymphocytosis of 76%, with 11% atypical lymphocytes, and viral cultures of urine were positive for CMV. The infant's condi¬ tion then improved, and 19 days after the initial deterioration, respiratory setting returned to 0.35 FIO, and pressure to 24/6 cm H,0. The infant's color returned to normal, the hepatosplenomegaly gradually disappeared, and the lymphocytosis re¬ solved. He was treated with high-dose sure
of 33
cm
roentgenograms
furosemide for chronic
lung disease, which gradually improved, so that FIO, and pres¬ sures were gradually reduced during the next three weeks until he
was extubated. Urine cultures were still positive for CMV six weeks after the initial positive culture.
RESULTS Viral Cultures In addition to the three original eight of 100 infants screened on a weekly basis during a seven-month period began excreting CMV in urine and /or nasopharyngeal culture speci¬ mens some time after 28 days of age. Cytomegalovirus was recovered only from infants who had been in the nursery for four weeks or more; of 30 such infants, 11 (37%) were culturepositive. Since CMV acquired at or shortly after birth does not become detectable in urine until 28 days of age, routine clinical surveying of infants was resumed, but only among those 1 month of age or older. Of 21 such infants tested over the next sixmonth period, five (24%) were identi¬ fied as having acquired CMV infec¬ tion. Altogether, a total of 16 CMVinfected babies were identified, of whom eight had urine cultures for CMV within ten days of birth, all of which were negative. No other viruses were recovered in any of the infants, despite frequent use of tissue culture
systems that would have permitted
their recovery. In the three infants who died, CMV inclusions were found in the cells of liver, pancreas, kidney, testes, and lung (Fig 1). The virus was cultured from all of these organs, as well as from cerebral cortex, cerebellum, ileum, adrenal gland, esophagus, and
spleen.
cases,
Clinical Features
Cytomegalovirus was documented in 16 infants during the 18-month period of this study, and 14 of these demonstrated clinical features that were part of a recognizable symptom complex. The most frequent clinical feature was hepatosplenomegaly, which was found in 13 (93%) of the 14 infants; 12 (88%) infants appeared to be "septic"; ten (71%) experienced deterioration in respiratory status at the time of infection; and five (36%) had fevers above 38 °C. Ten (71%) of the infants had a peculiar gray pallor at the time they first began excreting virus. Three of the sickest infants died. An
Manifestations
atypical
lymphocytosis
of
greater than 8% was found in 13 (93%) of the 14 infants, a distinctly abnor¬ mal
Lung of infant dying with cytomegalovirus infection alovirus intranuclear inclusion (x320).
Downloaded From: http://archpedi.jamanetwork.com/ by a Oakland University User on 06/08/2015
Laboratory
finding
showing
in infants 28
cells with
days of age
typical cytomeg¬
or
older. An absolute
greater than 75%
lymphocytosis of
also present in nine (64%) infants. In four (29%), the clinical impression of bacterial sepsis was strengthened by an absolute increase in the nonsegmented polywas
morphonuclear leukocyte
greater than 1,400/cu
mm.
count to Four (29%)
infants also had thrombocytopenia of less than 75,000/cu mm. Repeated cultures of blood, urine, and CSF for bacteria and fungi were negative. General Characteristics of Infected Infants
The Table compares the general characteristics of the infected infants with those of noninfected infants who were cultured during the initial screening period. The mean gestation¬ al age was 28.5 weeks, and the mean birth weight was 1,058 g compared with a mean gestational age of 33.6 weeks and mean birth weight of 1,976 g in the noninfected infants. The infants were 28 to 148 days of age when the first positive CMV culture was obtained. All the infected infants had received multiple blood transfu¬ sions prior to the first positive CMV culture (mean, 21 separate units). Associated conditions or procedures in the infected infants were respiratory distress syndrome, 12; ventilatory support, 13; chronic lung disease, 14; multiple blood transfusions, 14; patent ductus arteriosus ligation, five; and pneumothoraces, five. COMMENT We first became aware of acquired CMV infection in our intensive care nursery in July 1975, during an inves¬ tigation for persistent conjugated hyperbilirubinemia in a 6-week-old preterm infant who had experienced an extremely stormy neonatal course. Two additional cases were identified within two weeks in similar, very sick, preterm infants. At that time, the
program of all infants admitted to the nursery was under¬ taken. The majority of CMV-infected infants manifested hepatosplenomeg¬ aly, a remarkable gray pallor, a selflimited respiratory deterioration, and a lymphocytosis, with an increase in atypical lymphocytes. The infants affected could all be considered as compromised hosts since they were all very small, preterm infants suffering from severe respiratory distress and its complications, including chronic lung disease. All had survived their initial illness, usually severe respirato¬ ry distress, and were progressing satisfactorily until approximately 4 to 6 weeks of age when they either became ill again or suddenly suffered abrupt deterioration in respiratory status. Prior to this study, infants manifesting this symptom complex would have been thought by our staff to be "septic," or to be suffering from severe respiratory deterioration sec¬ ondary to their chronic pulmonary disease. Recognition of this self-
screening
limited, predictable symptom complex a course of ten to 14 days, followed by return to the baseline
with
respiratory status has greatly aided in
the management of these infants, as well as in working with their fami¬ lies. The three preterm infants who died early in the study had the most severe underlying respiratory disease, and at autopsy had multiple organ cultures positive for the virus. Heretofore, there have been only occasional reports of clinical sympto¬ matology in infants with acquired CMV infection. In 1969, McCracken et al1" described two infants (one pre¬ term) who were thought to have symptomatic CMV infection 39 and 58
days postnatally, possibly secondary to multiple blood transfusions in the immediate neonatal period. At autop¬ sy of the infant who died, hydrocepha-
General Characteristics of Infected and Noninfected Infants Infected
Noninfected
(Mean & Range) 1,976 (830-3,969) (26-32)_33.6 (26-42)
(Mean & Range) Birth weight, g_1,058 (720-1,600) Gestational age, wk_28.5 at first positive cytomegalovirus culture,
Age
days_55 (28-148)_None_
No. of separate blood units transfused
21
(8-36)
4
(0-30)
Downloaded From: http://archpedi.jamanetwork.com/ by a Oakland University User on 06/08/2015
lus and hepatosplenomegaly were not¬ ed, and CMV inclusions were present in the pancreas, parotid, and kidney. McEnery and Stern1- described another CMV-infected, premature in¬ fant who received intrauterine blood transfusions and died seven weeks after birth with pneumonia and hepa¬
tosplenomegaly bilaterally. Yeager et
al11 described two preterm infants who apparently acquired CMV infec¬ tion in the intensive care nursery and clinically manifested hepatospleno¬ megaly and lymphocytosis. In a later study, Yeager11 described six infants (three preterm) who probably ac¬ quired CMV infection in the nursery. In this group, a febrile illness devel¬ oped in one, two had hepatospleno¬ megaly, and two had splenomegaly. Pneumonia was noted in one and thrombocytopenia in one. We found 16 of 51 (31%) infants hospitalized in our intensive care nursery for more than 28 days to be excreting CMV. The reasons for such a high prevalence of CMV is not entirely clear. It has been reported that 1% to 2% of all newborn infants and 3% to 4% of premature infants excrete CMV at birth,1 i; suggesting congenital infection with the virus. This does not seem to be the major route of infection in our infants, since, of the 16 infected infants, cultures were made in eight within a few days of birth and all were negative. In a study of the effect of maternal cyto¬ megalovirus excretion on early peri¬ natal infection, Reynolds et al1"' found that 37% of infants born to thirdtrimester cervical CMV excreters be¬ came infected by 3 months of age compared with 4% of infants born to nonexcreters. Viral excretion in the infected infants was found to begin between 30 and 60 days of age. Kumar et alie also found that 20% of CMVnegative infants born to mothers with cytomegaloviruria during pregnancy will begin excreting virus before 14 weeks of age. Although we do not have data on CMV excretion in the mothers in this report, this route of infection is possible for at least some of our infants. Another possible source of infection is breast milk. Cytomegalovirus has been recovered from the breast milk
of 27% of
seropositive mothers,17 and it, therefore, might well be expected that virus would be present in pooled specimens. However, 25 samples from pools of such milk from our breast milk bank were cultured for CMV, and all were negative. Nosocomial spread might also have contributed to a high prevalence of CMV infection, although there was no
evidence for increased occurrence in cohorts or other suggestion of patientto-patient spread. In addition, during this time, there were periods of one to two months when there were no infected infants in the unit. Finally, all of these infants received multiple transfusions with either packed RBCs or whole, heparinized blood, and it is possible that the source
of infection was blood. It is known that 51% to 72% of blood donors have a positive complement fixation titer for CMV.:,(i1819 Yeager14 has reported that, in preterm infants examined at 7 months of age for seroconversion for CMV, 11% of nontransfused preterm infants converted, compared with 25% of infants who had received transfu¬ sions. In addition, Pass et al2" exam¬ ined CMV excretion in infants receiv¬ ing blood from two separate groups of donors: one with positive CMV titers (by complement fixation) of > 1:8 and the other with titers of < 1:8. In the first group, 35% of the infants excreted the virus on follow-up, as compared with 7% of the infants receiving blood from donors with low titers. It is possible that in the future
donors supplying blood for preterm infants should be screened for evi¬ dence of CMV infection. In summary, we have described a
relatively
common,
recognizable
symptom complex found in very sick, preterm infants, resulting from CMV
infection. The clinical deterioration associated with the disease seems to have a predictable course lasting approximately two weeks and seems to be self-limited. The most likely sources of infection in these infants seem to be exposure to maternal cervi¬ cal infection at the time of birth or, more probably, the multiple blood transfusions received by the infants in the course of treatment for their primary disease.
References 1. Weller TH: The cytomegaloviruses: Ubiquiagents with protean clinical manifestations: II. N Engl J Med 285:267-274, 1971. 2. Carlstr\l=o"\mG: Virologic studies on cytomegalic inclusion disease. Acta Pediatr Scand 54:17\x=req-\ 23, 1965. 3. Haldane EV, Embil JA, Wall AD: A serological study of cytomegalovirus infection in the population of Easter Island. Bull WHO 40:969\x=req-\ 973, 1969. 4. Levinsohn EM, Foy HM, Kenny GE, et al: Isolation of cytomegalovirus from a cohort of 100 infants throughout the first year of life. Proc Soc Exp Biol Med 132:957-962, 1969. 5. Wentworth BB, Alexander ER: Serioepidemiology of infections due to members of herpesvirus group. Am J Epidemiol 94:496-507, 1971. 6. Holzel A, Jacobs SI, Keen J, et al: Cytomegalovirus infection in the northwest of England: A report on a two-year study. Arch Dis Child 45:513-522, 1970. 7. Hanshaw JB, Dudgeon JD: Congenital cytomegalovirus, in Hanshaw JB, Dudgeon JD tous
(eds): Viral Diseases of the Fetus and Newborn. Philadelphia, WB Saunders Co, 1978, pp 97-152. 8. Granstrom M-L, Leinikki P, Santavuori P, et al: Perinatal cytomegalovirus infection in man. Arch Dis Child 52:354-359, 1977. 9. Whitley RJ, Brasfield D, Reynolds DW, et al: Protracted pneumonitis in young infants associated with perinatally acquired cytomegaloviral infection. J Pediatr 89:16-22, 1976. 10. McCracken GH, Shinefield HR, Cobb K, et al: Congenital cytomegalic inclusion disease. Am
J Dis Child 117:522-539, 1969. 11. Yeager AS, Jacobs H, Clark J: Nursery\x=req-\ acquired cytomegalovirus infection in two premature infants. J Pediatr 81:332-335, 1972. 12. McEnery G, Stern H: Cytomegalovirus infection in early infancy: Five atypical cases. Arch Dis Child 45:669-673, 1970. 13. Hufnagle K, Drew L, Ballard R: Cytomegalovirus in sick preterm infants, abstracted. Pediatr Res 10:399, 1976. 14. Yeager AS: Transfusion-acquired cytomegalovirus infection in newborn infants. Am J Dis Child 128:478-483, 1974.
Downloaded From: http://archpedi.jamanetwork.com/ by a Oakland University User on 06/08/2015
15. Reynolds DW, Stagno S, Hosty TS, et al: Maternal cytomegalovirus excretion and perinatal infection. N Engl J Med 289:1-5, 1973. 16. Kumar ML, Gole E, Nandervi G: Risk of acquired cytomegalovirus infection in infants of maternal CMV excretors, abstracted. Pediatr Res 9:342, 1975. 17. Hayes K, Danks DM, Gibas H, et al: Cytomegalovirus in human milk. N Engl J Med 287:177-178, 1972. 18. Perham TGM, Caul EO, Conway PJ, et al: Cytomegalovirus infection in blood donors: A prospective study. Br J Hematol 20:307-320, 1971. 19. Klemola E, von Essen R, Paloheimo J, et al: Cytomegalovirus antibodies in donors of fresh blood to patients submitted to open-heart surgery. Scand J Infect Dis 1:137-140, 1969. 20. Pass MA, Johnson JD, Schulman IA, et al: Evaluation of a walking blood-donor blood transfusion program in an intensive care nursery. J Pediatr 89:646-651, 1976.
in
Preterm Infants Roberta A. Ballard, MD; W. Lawrence Drew, MD, PhD; Karen G.
\s=b\Small sick preterm infants requiring
in a neonatal intensive care unit for more than 28 days were cultured for cytomegalovirus in urine and/or nasopharynx during two periods lasting a total of 13 months. Sixteen of 51 such infants began excreting the virus at 28 to 148 days of age (mean, 55 days). In 14 of the 16, a recognizable, self-limited symptom complex developed that consisted of care
respiratory deterioration, hepatosplenomegaly, a remarkable gray pallor, and both an atypical and absolute lymphocy-
tosis. All of the infants with the clinical symptom complex had underlying chronic lung disease and all had received multiple blood transfusions during their hospitalization. Acquired cytomegalovirus may be relatively common in sick preterm infants and should be distinguished from other causes of rapid deterioration. (Am J Dis Child 133:482-485, 1979)
clinical of cy¬ infection have
incidence, features, The sequelae congenital tomegalovirus (CMV) and
been studied
extensively and re¬ viewed by numerous authors.1-7 In contrast, the impact of acquired CMV infection in the newborn is less clear.
From the Department of Pediatrics (Drs Ballard, Hufnagle, and Riedel) and Pathology and Laboratory Medicine (Dr Drew), Mount Zion Hospital and Medical Center, San Francisco. Dr Hufnagle is now at the William Beaumont Hospital, Royal Oak, Mich. Reprint requests to Department of Pediatrics, Mount Zion Hospital and Medical Center, PO Box 7921, San Francisco, CA 94120 (Dr Ballard).
Hufnagle, MD; Philip
A. Riedel, MD
Granstrom et alR has reported that newborns who acquired the disease perinatally (^ 5 months of age) have few symptoms and develop normally; however, protracted inter¬ stitial pneumonitis has been reported in two term infants9 who apparently acquired the virus from the maternal genital tract. In the immunologically compromised host, the disease may be severe, with hepatitis, hemolytic anemia, pneumonitis, myocarditis and pericarditis, and possibly involvement of the CNS.1 Because sick preterm infants are immunologically com¬ promised, they might be considered candidates for the more serious disease. Since 1968, there have been occa¬ sional reports of serious CMV infec¬ tion in preterm infants in whom the illness was presumed to be acquired rather than congenital.101- In July 1975, we noted a distinct syndrome associated with CMV infection in three sick, preterm infants. After this observation, we initiated a systematic survey for the incidence and clinical manifestations of CMV infection in preterm infants and found 16 infants to be excreting the virus.13 This report describes 14 of these infants in whom a recognizable symptom complex due to CMV infection developed that was thought to be acquired in the inten¬ sive care nursery.
healthy
Downloaded From: http://archpedi.jamanetwork.com/ by a Oakland University User on 06/08/2015
METHODS Patient Population The Mount Zion Hospital intensive care nursery is a 16-bed unit with 300 admis¬ sions a year of which 50% are inborn and 50% are transported from outlying hospi¬ tals. The population is composed primarily of preterm infants with hyaline membrane disease. After identification of three
cytomegalovirus disease, two prevalence of CMV coloniza¬ done. During the first survey,
infants with surveys for
tion were 100 consecutive sick infants admitted to the unit for at least four days during a seven-month period underwent weekly cul¬ tures of urine and nasopharynx for virus. Since none of these infants excreted virus prior to 28 days of age, in the second phase of the survey, weekly cultures were taken from all infants who had been in the unit for more than 28 days. Twenty-one addi¬ tional infants met the criteria for screen¬ ing during a six-month period.
Viral Studies Urine specimens were transported to the laboratory within four hours of being obtained; 10 mL of urine were centrifuged at 1,000 g for 15 to 20 minutes. The sedi¬ ment was resuspended in 0.5 mL of super¬ natant fluid and 0.2 mL of this was added to the media of each of two tubes of WI-38
cells. Cultures of WI-38 were observed for 38 days to detect CMV cytopathogenic effect. Throat cultures were obtained with ster¬ ile cotton-tipped swabs that were immedi¬ ately immersed in a tube of WI-38 tissue culture. The tube was capped and sent to
virology laboratory. After approxi¬ mately one hour for adsorption of the virus to the tissue culture, the swab was removed and 0.2-mL aliquots of fluid were removed for inoculation into tubes of primary monkey kidney, primary human embryonic kidney, and Hep-2 tissue cultures. Only the
WI-38 cultures were inoculated if the infant was not acutely ill; the additional cultures were inoculated if the infant was sick. Autopsy cultures were prepared by mincing centimeter-sized portions of vari¬ ous organs in Hanks' balanced salt solution containing gentamicin sulfate and amphotericin B. The specimen was then sonically disrupted in a sonic oscillator at 10,000 Hz for 40 s. It was then centrifuged at 1,000 g for five minutes, and 0.2 mL of superna¬ tant fluid was inoculated per tube of tissue culture. Cultures of pooled breast milk were performed exactly as for urine.
REPORT OF A CASE This infant of 1,100-g birthweight and 29 weeks' gestation was born to a 25-year-old, gravida 4, para 2 mother with fever and foul-smelling amniotic fluid and had Apgar scores of 4 and 5 at one and five minutes, respectively. The infant had respiratory distress at birth and required ventilation with 80% oxygen. At 5 hours of age, the infant was transported to Mount Zion Hospital with a respirator setting of 0.8 FIO.,, a pressure of 22/7 cm H,0, and a ventilation rate of 40/min. The infant had a
complicated course requiring a peak pres¬
H,0 and FIO, of 0.7. Chest at 10 days of age showed early development of bronchopulmonary dysplasia. The infant's condition improved and then stabilized with a ventilator setting of FIO, 0.35 and pressure of 24/6 cm H ,0 for the next four weeks. Because of multiple blood studies, the infant received frequent blood transfusions of 10 to 15 mL of packed RBCs obtained from 19 different donors. At 42 days of age, respiratory status worsened, requiring an increase in FIO, from 0.35 to 0.85 and an increase in inspiratory pressure from 24 to 30 cm H,0 over a period of nine days. At this time, the infant had hepatosplenomegaly, a remark¬ able gray pallor, and looked "septic"; feed¬ ing was poor. Laboratory studies showed a lymphocytosis of 76%, with 11% atypical lymphocytes, and viral cultures of urine were positive for CMV. The infant's condi¬ tion then improved, and 19 days after the initial deterioration, respiratory setting returned to 0.35 FIO, and pressure to 24/6 cm H,0. The infant's color returned to normal, the hepatosplenomegaly gradually disappeared, and the lymphocytosis re¬ solved. He was treated with high-dose sure
of 33
cm
roentgenograms
furosemide for chronic
lung disease, which gradually improved, so that FIO, and pres¬ sures were gradually reduced during the next three weeks until he
was extubated. Urine cultures were still positive for CMV six weeks after the initial positive culture.
RESULTS Viral Cultures In addition to the three original eight of 100 infants screened on a weekly basis during a seven-month period began excreting CMV in urine and /or nasopharyngeal culture speci¬ mens some time after 28 days of age. Cytomegalovirus was recovered only from infants who had been in the nursery for four weeks or more; of 30 such infants, 11 (37%) were culturepositive. Since CMV acquired at or shortly after birth does not become detectable in urine until 28 days of age, routine clinical surveying of infants was resumed, but only among those 1 month of age or older. Of 21 such infants tested over the next sixmonth period, five (24%) were identi¬ fied as having acquired CMV infec¬ tion. Altogether, a total of 16 CMVinfected babies were identified, of whom eight had urine cultures for CMV within ten days of birth, all of which were negative. No other viruses were recovered in any of the infants, despite frequent use of tissue culture
systems that would have permitted
their recovery. In the three infants who died, CMV inclusions were found in the cells of liver, pancreas, kidney, testes, and lung (Fig 1). The virus was cultured from all of these organs, as well as from cerebral cortex, cerebellum, ileum, adrenal gland, esophagus, and
spleen.
cases,
Clinical Features
Cytomegalovirus was documented in 16 infants during the 18-month period of this study, and 14 of these demonstrated clinical features that were part of a recognizable symptom complex. The most frequent clinical feature was hepatosplenomegaly, which was found in 13 (93%) of the 14 infants; 12 (88%) infants appeared to be "septic"; ten (71%) experienced deterioration in respiratory status at the time of infection; and five (36%) had fevers above 38 °C. Ten (71%) of the infants had a peculiar gray pallor at the time they first began excreting virus. Three of the sickest infants died. An
Manifestations
atypical
lymphocytosis
of
greater than 8% was found in 13 (93%) of the 14 infants, a distinctly abnor¬ mal
Lung of infant dying with cytomegalovirus infection alovirus intranuclear inclusion (x320).
Downloaded From: http://archpedi.jamanetwork.com/ by a Oakland University User on 06/08/2015
Laboratory
finding
showing
in infants 28
cells with
days of age
typical cytomeg¬
or
older. An absolute
greater than 75%
lymphocytosis of
also present in nine (64%) infants. In four (29%), the clinical impression of bacterial sepsis was strengthened by an absolute increase in the nonsegmented polywas
morphonuclear leukocyte
greater than 1,400/cu
mm.
count to Four (29%)
infants also had thrombocytopenia of less than 75,000/cu mm. Repeated cultures of blood, urine, and CSF for bacteria and fungi were negative. General Characteristics of Infected Infants
The Table compares the general characteristics of the infected infants with those of noninfected infants who were cultured during the initial screening period. The mean gestation¬ al age was 28.5 weeks, and the mean birth weight was 1,058 g compared with a mean gestational age of 33.6 weeks and mean birth weight of 1,976 g in the noninfected infants. The infants were 28 to 148 days of age when the first positive CMV culture was obtained. All the infected infants had received multiple blood transfu¬ sions prior to the first positive CMV culture (mean, 21 separate units). Associated conditions or procedures in the infected infants were respiratory distress syndrome, 12; ventilatory support, 13; chronic lung disease, 14; multiple blood transfusions, 14; patent ductus arteriosus ligation, five; and pneumothoraces, five. COMMENT We first became aware of acquired CMV infection in our intensive care nursery in July 1975, during an inves¬ tigation for persistent conjugated hyperbilirubinemia in a 6-week-old preterm infant who had experienced an extremely stormy neonatal course. Two additional cases were identified within two weeks in similar, very sick, preterm infants. At that time, the
program of all infants admitted to the nursery was under¬ taken. The majority of CMV-infected infants manifested hepatosplenomeg¬ aly, a remarkable gray pallor, a selflimited respiratory deterioration, and a lymphocytosis, with an increase in atypical lymphocytes. The infants affected could all be considered as compromised hosts since they were all very small, preterm infants suffering from severe respiratory distress and its complications, including chronic lung disease. All had survived their initial illness, usually severe respirato¬ ry distress, and were progressing satisfactorily until approximately 4 to 6 weeks of age when they either became ill again or suddenly suffered abrupt deterioration in respiratory status. Prior to this study, infants manifesting this symptom complex would have been thought by our staff to be "septic," or to be suffering from severe respiratory deterioration sec¬ ondary to their chronic pulmonary disease. Recognition of this self-
screening
limited, predictable symptom complex a course of ten to 14 days, followed by return to the baseline
with
respiratory status has greatly aided in
the management of these infants, as well as in working with their fami¬ lies. The three preterm infants who died early in the study had the most severe underlying respiratory disease, and at autopsy had multiple organ cultures positive for the virus. Heretofore, there have been only occasional reports of clinical sympto¬ matology in infants with acquired CMV infection. In 1969, McCracken et al1" described two infants (one pre¬ term) who were thought to have symptomatic CMV infection 39 and 58
days postnatally, possibly secondary to multiple blood transfusions in the immediate neonatal period. At autop¬ sy of the infant who died, hydrocepha-
General Characteristics of Infected and Noninfected Infants Infected
Noninfected
(Mean & Range) 1,976 (830-3,969) (26-32)_33.6 (26-42)
(Mean & Range) Birth weight, g_1,058 (720-1,600) Gestational age, wk_28.5 at first positive cytomegalovirus culture,
Age
days_55 (28-148)_None_
No. of separate blood units transfused
21
(8-36)
4
(0-30)
Downloaded From: http://archpedi.jamanetwork.com/ by a Oakland University User on 06/08/2015
lus and hepatosplenomegaly were not¬ ed, and CMV inclusions were present in the pancreas, parotid, and kidney. McEnery and Stern1- described another CMV-infected, premature in¬ fant who received intrauterine blood transfusions and died seven weeks after birth with pneumonia and hepa¬
tosplenomegaly bilaterally. Yeager et
al11 described two preterm infants who apparently acquired CMV infec¬ tion in the intensive care nursery and clinically manifested hepatospleno¬ megaly and lymphocytosis. In a later study, Yeager11 described six infants (three preterm) who probably ac¬ quired CMV infection in the nursery. In this group, a febrile illness devel¬ oped in one, two had hepatospleno¬ megaly, and two had splenomegaly. Pneumonia was noted in one and thrombocytopenia in one. We found 16 of 51 (31%) infants hospitalized in our intensive care nursery for more than 28 days to be excreting CMV. The reasons for such a high prevalence of CMV is not entirely clear. It has been reported that 1% to 2% of all newborn infants and 3% to 4% of premature infants excrete CMV at birth,1 i; suggesting congenital infection with the virus. This does not seem to be the major route of infection in our infants, since, of the 16 infected infants, cultures were made in eight within a few days of birth and all were negative. In a study of the effect of maternal cyto¬ megalovirus excretion on early peri¬ natal infection, Reynolds et al1"' found that 37% of infants born to thirdtrimester cervical CMV excreters be¬ came infected by 3 months of age compared with 4% of infants born to nonexcreters. Viral excretion in the infected infants was found to begin between 30 and 60 days of age. Kumar et alie also found that 20% of CMVnegative infants born to mothers with cytomegaloviruria during pregnancy will begin excreting virus before 14 weeks of age. Although we do not have data on CMV excretion in the mothers in this report, this route of infection is possible for at least some of our infants. Another possible source of infection is breast milk. Cytomegalovirus has been recovered from the breast milk
of 27% of
seropositive mothers,17 and it, therefore, might well be expected that virus would be present in pooled specimens. However, 25 samples from pools of such milk from our breast milk bank were cultured for CMV, and all were negative. Nosocomial spread might also have contributed to a high prevalence of CMV infection, although there was no
evidence for increased occurrence in cohorts or other suggestion of patientto-patient spread. In addition, during this time, there were periods of one to two months when there were no infected infants in the unit. Finally, all of these infants received multiple transfusions with either packed RBCs or whole, heparinized blood, and it is possible that the source
of infection was blood. It is known that 51% to 72% of blood donors have a positive complement fixation titer for CMV.:,(i1819 Yeager14 has reported that, in preterm infants examined at 7 months of age for seroconversion for CMV, 11% of nontransfused preterm infants converted, compared with 25% of infants who had received transfu¬ sions. In addition, Pass et al2" exam¬ ined CMV excretion in infants receiv¬ ing blood from two separate groups of donors: one with positive CMV titers (by complement fixation) of > 1:8 and the other with titers of < 1:8. In the first group, 35% of the infants excreted the virus on follow-up, as compared with 7% of the infants receiving blood from donors with low titers. It is possible that in the future
donors supplying blood for preterm infants should be screened for evi¬ dence of CMV infection. In summary, we have described a
relatively
common,
recognizable
symptom complex found in very sick, preterm infants, resulting from CMV
infection. The clinical deterioration associated with the disease seems to have a predictable course lasting approximately two weeks and seems to be self-limited. The most likely sources of infection in these infants seem to be exposure to maternal cervi¬ cal infection at the time of birth or, more probably, the multiple blood transfusions received by the infants in the course of treatment for their primary disease.
References 1. Weller TH: The cytomegaloviruses: Ubiquiagents with protean clinical manifestations: II. N Engl J Med 285:267-274, 1971. 2. Carlstr\l=o"\mG: Virologic studies on cytomegalic inclusion disease. Acta Pediatr Scand 54:17\x=req-\ 23, 1965. 3. Haldane EV, Embil JA, Wall AD: A serological study of cytomegalovirus infection in the population of Easter Island. Bull WHO 40:969\x=req-\ 973, 1969. 4. Levinsohn EM, Foy HM, Kenny GE, et al: Isolation of cytomegalovirus from a cohort of 100 infants throughout the first year of life. Proc Soc Exp Biol Med 132:957-962, 1969. 5. Wentworth BB, Alexander ER: Serioepidemiology of infections due to members of herpesvirus group. Am J Epidemiol 94:496-507, 1971. 6. Holzel A, Jacobs SI, Keen J, et al: Cytomegalovirus infection in the northwest of England: A report on a two-year study. Arch Dis Child 45:513-522, 1970. 7. Hanshaw JB, Dudgeon JD: Congenital cytomegalovirus, in Hanshaw JB, Dudgeon JD tous
(eds): Viral Diseases of the Fetus and Newborn. Philadelphia, WB Saunders Co, 1978, pp 97-152. 8. Granstrom M-L, Leinikki P, Santavuori P, et al: Perinatal cytomegalovirus infection in man. Arch Dis Child 52:354-359, 1977. 9. Whitley RJ, Brasfield D, Reynolds DW, et al: Protracted pneumonitis in young infants associated with perinatally acquired cytomegaloviral infection. J Pediatr 89:16-22, 1976. 10. McCracken GH, Shinefield HR, Cobb K, et al: Congenital cytomegalic inclusion disease. Am
J Dis Child 117:522-539, 1969. 11. Yeager AS, Jacobs H, Clark J: Nursery\x=req-\ acquired cytomegalovirus infection in two premature infants. J Pediatr 81:332-335, 1972. 12. McEnery G, Stern H: Cytomegalovirus infection in early infancy: Five atypical cases. Arch Dis Child 45:669-673, 1970. 13. Hufnagle K, Drew L, Ballard R: Cytomegalovirus in sick preterm infants, abstracted. Pediatr Res 10:399, 1976. 14. Yeager AS: Transfusion-acquired cytomegalovirus infection in newborn infants. Am J Dis Child 128:478-483, 1974.
Downloaded From: http://archpedi.jamanetwork.com/ by a Oakland University User on 06/08/2015
15. Reynolds DW, Stagno S, Hosty TS, et al: Maternal cytomegalovirus excretion and perinatal infection. N Engl J Med 289:1-5, 1973. 16. Kumar ML, Gole E, Nandervi G: Risk of acquired cytomegalovirus infection in infants of maternal CMV excretors, abstracted. Pediatr Res 9:342, 1975. 17. Hayes K, Danks DM, Gibas H, et al: Cytomegalovirus in human milk. N Engl J Med 287:177-178, 1972. 18. Perham TGM, Caul EO, Conway PJ, et al: Cytomegalovirus infection in blood donors: A prospective study. Br J Hematol 20:307-320, 1971. 19. Klemola E, von Essen R, Paloheimo J, et al: Cytomegalovirus antibodies in donors of fresh blood to patients submitted to open-heart surgery. Scand J Infect Dis 1:137-140, 1969. 20. Pass MA, Johnson JD, Schulman IA, et al: Evaluation of a walking blood-donor blood transfusion program in an intensive care nursery. J Pediatr 89:646-651, 1976.
