Symptomatic Parainfluenza Virus Infections in Children Undergoing Hematopoietic Stem Cell Transplantation

Abstract

Parainfluenza virus (PIV) infections cause significant mortality in adults undergoing hematopoietic stem cell transplantation (HSCT). Children are more prone than adults to develop PIV infections. However, data regarding the epidemiology of PIV infections in children undergoing HSCT is limited. This study sought to determine the incidence of symptomatic PIV infections, risk factors for lower respiratory tract infection (LRTI) and their impact on mortality after pediatric HSCT. A total of 1,028 children who underwent HSCT between1995 and 2009 were studied. PIV infections occurred in 46 (6.2%) of the 738 patients tested for respiratory infections. PIV infection was the most common symptomatic respiratory viral infection in this population. On multivariate logistic regression analysis, receipt of an allogeneic transplant (p<0.0001), and TBI-based conditioning (p<0.0001) was associated with an increased risk of acquiring symptomatic PIV infection. Of the 46 HSCT patients with PIV infection, 18 (39%) patients had LRTI. LRTI was associated with PIV infection in the first 100 days after transplant (p=0.006), use of steroids (p=0.035), and ALC <100 cells/μL at onset of infection (p<0.0001). ALC of <500 cells/μL was associated with prolonged viral shedding (p=0.045). Six (13%) HSCT patients died of PIV infection. Mortality was associated with African-American ethnicity (p=0.013), LRTI (p=0.002), use of steroids (p<0.0001), mechanical ventilation (p<0.0001) and ALC <100 cells/μL at onset of infection (p=0.01). PIV infection causes significant morbidity and mortality in children undergoing HSCT.

INTRODUCTION

The human parainfluenza viruses (PIV), which belong to the family Paramyxoviridae, are enveloped, single stranded, RNA viruses that cause upper respiratory tract infections (URTI) and lower respiratory tract infections (LRTI) in children and adults. PIV infections are relatively common and may occur in up to 7% of adults undergoing hematopoietic stem cell transplantation (HSCT) with a 180-day mortality as high as 75% [1]. In the largest study of PIV infections in adults undergoing HSCT to date, the only risk factor for developing PIV infection was transplantation from an unrelated donor [1]. Use of corticosteroids for graft versus host disease (GVHD) was associated with progression to LRTI. Mortality was increased in adults who had undergone transplantation within 100 days [1-3], by the presence of co-pathogens [1], URTI [1], LRTI [1, 4] and the need for mechanical ventilation [1].

Children are more prone than adults to develop PIV infections, which are second only to respiratory syncytial virus (RSV) as the cause of URTI and LRTI in those younger than 5 years of age [5]. PIV hospitalization rates in this group of children may be greater than those due to influenza viruses [5]. However, data regarding the epidemiology of PIV infections in children undergoing HSCT is limited.

This study sought to retrospectively determine the incidence of symptomatic PIV infections in a large cohort of children undergoing HSCT at our institution. Risk factors for progression to LRTI, and the impact on mortality after HSCT was studied.

PATIENTS AND METHODS

Patient population

This retrospective cohort study included a total of 1,028 children who underwent an allogeneic or autologous HSCT over a 15 year period, between January 1995 and December 2009 at St. Jude Children’s Research Hospital (SJCRH). A total of 738 children were tested for respiratory infections. The study was approved by the Institutional Review Board at SJCRH. SJCRH is a referral cancer hospital for children. Patients who have undergone HSCT are followed for a period of 10 years after transplant or until 18 years of age, whichever is longer. Virologic records were reviewed and recipients of HSCT in whom PIV 1-4 was detected by culture, direct fluorescent antibody (DFA) testing or polymerase chain reaction (PCR), from respiratory tract secretions were identified. Medical record review included analysis of clinical variables including age, race, gender, underlying malignancy, date of onset of respiratory symptoms, duration of viral shedding, whether infection was an URTI or LRTI, hospital or community-acquired, presence of fever, need for oxygen or mechanical ventilation, admission to an intensive care unit, mortality, absolute neutrophil count (ANC) and absolute lymphocyte count (ALC) at the time of infection and co-pathogens.

Virology and microbiology procedures

A nasopharyngeal swab or wash for culture and viral DFA staining was performed on patients with clinically suspected URTI or LRTI. Respiratory samples also included broncho-alveolar lavage (BAL), tracheal secretions, lung biopsy and autopsy specimens. All patients with LRTI underwent a BAL. Samples were inoculated into tissue culture containing rhesus monkey kidney, A549 and Madin-Darby canine kidney (MDCK) cells. Respiratory cultures were kept for 14 days before reporting as negative. Virus causing hemadsorption with guinea pig red cells on day 7 and 14 or cytopathic effect at any time was confirmed as PIV 1-4 using type-specific respiratory DFA smears prepared using commercially available type-specific antisera against PIV 1-4 (Bartels, County Wicklow Ireland; Pathodx, KS,USA; Diagnostic Hybrid, OH,USA). PCR was performed on respiratory samples obtained after February 2007, using a laboratory developed test panel of individual real-time PCR assays, targeting PIV 1-3, run on SmartCycler real-time PCR instrumentation (Smart cycler, Cepheid, Sunnyvale, CA).

Definitions

PIV infection was defined as an URTI when signs and symptoms were consistent with the diagnosis, with a normal chest examination and chest roentgenogram. PIV infection was defined as a LRTI when symptoms were accompanied with signs on lung auscultation and a new pulmonary infiltrate as defined by a plain chest radiograph, high resolution computerized-tomography or both, unexplained by a non-infectious etiology, was present. The day of onset of PIV infection was defined as the day when the positive diagnostic sample was collected. Mortality was considered attributable to PIV, if the patient died from a LRTI, as a result of respiratory failure, with isolation of PIV from respiratory secretions, and with no other ascertainable cause. The presence of a co-pathogen was defined by the isolation or detection of pathogenic bacterial or fungal species or by the identification of other opportunistic pathogens in addition to PIV from any site. Hospital-acquired (HA) infections were defined as symptomatic PIV infections occurring 7 days or more after admission to the inpatient unit. Duration of viral shedding was defined as the interval between the day of the first positive culture or diagnostic test and the first negative culture or diagnostic test. Recurrent infections were defined by isolation or detection of PIV from the same patient, 14 days or more after the last negative culture or diagnostic test. Only the first infection from each patient and the first transplant, was included in the analyses. Day of engraftment after HSCT was defined as the first of three consecutive days of achieving an ANC greater than 500 cells/μL. Assessment of Graft- versus-Host disease (GVHD) was made according to consensus criteria [6].

Management

Children diagnosed with PIV infection were placed in respiratory isolation. Nasopharyngeal swabs or washes were obtained once weekly in most patients to document clearance of the virus. Ribavirin was used for the patients in this cohort, at the discretion of the treating physician at a dose of 2 grams (60mg/mL) by intermittent aerosolization for 2 hours, three times a day for up to 7 days. Treatment with antibiotics and antifungals was as indicated.

Patients were isolated post HSCT in individual rooms with positive pressure with respect to the hallway and high efficiency particulate air filtration. Visitors were screened and strongly encouraged to perform hand hygiene before and after patient contact. Intravenous immune globulin was routinely given weekly to all patients post HSCT until day + 100, to maintain trough serum concentrations >400 mg/dL. Prophylaxis against RSV was given monthly with palivizumab or RespiGam (MedImmune, Gaithersburg, MD), during the RSV season, for children who were transplant recipients and less than 2 years of age.

Statistical Analysis

Descriptive statistics were obtained for the demographic variables of children with PIV infection who had undergone HSCT. The marginal association between PIV infection and independent variables were determined by Fischer’s exact test and Kruskal-Wallis test. Furthermore, when appropriate exact multiple logistic regression models were used to investigate the independent association of one factor after controlling for other variables. Wilcoxon rank sum test was used to compare continuous variables between two groups. All analyses were performed in statistical software package SAS 9.2 (SAS Institute, Inc., Cary, NC).

RESULTS

Incidence of symptomatic PIV infections

PIV infection was documented in 46 (6.2%) of the 738 HSCT patients tested between January 1995 and December 2009. PIV infection was the most common respiratory viral infection in our HSCT population during this 15 year period, followed by adenovirus in 39 (5.3%), influenza A in 36 (4.9%), cytomegalovirus in 33 (4.5%), RSV in 29 (3.9%), HSV in 19 (2.6%), influenza B in 15 (2%), rhinovirus in 7(1%), human metapneumovirus in 2 patients and human enterovirus in 1 patient.

PIV-3 accounted for the majority of PIV infections (30 of 46 patients or 65%), PIV-1 was isolated in 8 (17%), PIV-2 in 5 (11%) and PIV-4 in 2 (4%) patients. The isolate was not typed in 1 patient. PIV-3 infections occurred throughout the year with a peak in spring and summer seasons (30 of 45 patients or 66%), while more than 60% of PIV-1, 2 and 4 infections occurred in fall and winter (Fig 1). PIV-1 was seen predominantly in odd-numbered years including 1995,1997,2001,2003 and 2009. PIV-2 was also seen predominantly in odd-numbered years including 1997, 1999 and 2003. There were 21 PIV infections between 1995-1999, 15 between 2000-2004 and 10 between 2005-2009, with no trend in frequency observed over time (p=0.31). No apparent outbreaks of infection were observed. Of the 46 patients, 33 (72%) had community-acquired (CA) infections and 13 (28%) had HA infections. These were interspersed among CA infections with no evidence of clustering. A single cluster of 3 CA infections occurred over one month during the study period.

Figure 1.

Monthly distribution of the incidence of symptomatic parainfluenza virus respiratory infections in children who had undergone HSCT at SJCRH between 1995 and 2009.

Risk factors for development of symptomatic PIV infection

The characteristics of the HSCT recipients with and without PIV infection are shown in Table I. The entire cohort of 1028 patients were followed for a total of 3472 patient years giving an incidence of 13.2 cases of symptomatic PIV infection per 1000 patient years of observation. The median age of the children at the time of infection was 5 years and 6 months (range: 8 months to 20 years and 8 months). Of the 46 patients, 5 (11%) were ≤ 2 years of age, 23 (50%) were between 2 and 10 years of age, 16 (35%) were between 10 and 18 years of age and 2 (4%) were 18 to 21 years of age.

Table I.

Characteristics of the HSCT recipients with and without symptomatic PIV infection

Characteristic	Study Cohort n=1028	Children without PIV infection n= 982	Children with PIV infection n=46	p-value odds ratio 95% CI
Median age (years)	8.6	8.6	5.5	0.20
Males	630 (61)	602 (61)	28 (61)	0.95 1.0 (0.56-1.87)
Race
White	752 (73)	722 (74)	30 (65)	0.96
AA	171 (17)	162 (16)	9 (20)	1.34 (0.62-2.87)
Others	105 (10)	98 (11)	7 (6)
Diagnosis
Leukemia	497 (48)	468 (48)	29 (63)	0.04
Lymphoma	94 (9)	92 (9)	2 (4)	1.87 (1.0-3.5)
MDS	46 (4)	43 (4)	3 (7)
Solid tumors	253 (25)	251(26)	2 (4)	0.005
Hematologic disorders	62 (6)	56 (6)	6 (13)	0.13 (0.03-0.6)
Immunologic disorders	44 (4)	40 (4)	4 (9)
Metabolic disorders	32 (3)	32 (3)	0
Donor type
Autologous	333 (32)	329 (34)	4 (9)	0.0016
Allogeneic	695 (68)	653 (66)	42 (91)	5.3(1.89-14.88)
Unrelated	278 (22)	255 (21)	23 (50)	0.40
Matched unrelated	225 (22)	202 (21)	23 (50)	0.69 (0.30-1.47)
Mis-matched unrelated	53 (5)	53 (5)	0
Sibling related	214 (17)	204 (17)	10 (22)
Matched sibling	207 (16)	197 (16)	10 (22)
Mis-matched sibling	7 (1)	7 (1)	0
Haplo-identical	194 (15)	186 (15)	8 (17)
Cord blood	9 (1)	8 (1)	1 (2)
CMV positive recipient	543 (52)	524 (55)	19 (41)	0.11 0.61 (0.34-1.12)
Conditioning
TBI	458(45)	425 (43)	33 (72)	0.0003 0.30 (0.16-0.58)
Non-myeloablative	27 (3)	25 (2)	2 (4)	0.46
Myeloablative	1001(98)	957 (97)	44 (95)	0.57 (0.13-2.50)
GVHD
Acute	267 (26)	259 (26)	8 (17)	0.006
Grade I-II	205	199	6
Grade I-IV	62	60	2
Chronic	120 (12)	120 (12)	0	0.33 (0.15-0.72)
Limited	70	70	0
Extensive	50	50	0

NOTE. Data are no. (%) of patients, unless otherwise indicated. HSCT, Hematopoietic Stem Cell transplantation; PIV, parainfluenza virus; CI, confidence interval; AA, African-American; MDS, myelodysplastic syndrome; CMV, cytomegalovirus; TBI, total body radiation; GVHD, graft versus host disease.

Univariate logistic regression analysis revealed receipt of an allogeneic HSCT (p=0.0016) and Total Body Irradiation (TBI; p=0.0003) to be significantly associated with PIV infection. An underlying diagnosis of leukemia (p=0.04) was of marginal significance (Table I). On multivariate logistic regression analysis, patients who underwent an allogeneic transplant (p<0.0001) and those who received TBI (p<0.0001) were more likely to acquire PIV infection. Patients with an underlying diagnosis of a solid tumor (p=0.005) and GVHD (p=0.006) were less likely to have PIV infection. Age, gender, ethnicity, donor type, recipient CMV positivity, myeloablative conditioning and presence of GVHD, were not significantly associated with symptomatic PIV infection (Table I).

Clinical features of symptomatic PIV infection

Of the 46 patients with PIV infections, 28 (61%) presented with URTI without any evidence of LRTI. The most common symptoms were cough and rhinorrhea which was accompanied with fever in only 1 patient. One patient with URTI presented with croup. Eighteen patients (39%) had signs of LRTI on lung auscultation confirmed by chest radiography in 14 patients, with unilateral alveolar infiltrates in 3 patients and bilateral interstitial infiltrates in 11 patients. In 4 patients who had normal chest radiography, computerized chest tomography (CT) revealed bilateral interstitial infiltrates in 2 patients and multiple bilateral lung nodules in 2 patients. Sinusitis as confirmed by CT was present in 6 patients with LRTI and 3 patients with URTI. Fever accompanied LRTI in 7 (39%) patients. LRTI was preceded by URTI in all but 1 patient.

Diagnosis of PIV infection

The median duration between onset of symptoms and diagnosis was 2 days (range 1-7 days). The median duration of viral shedding evaluated in 32 patients who had at least one negative follow-up testing was 8 days (range 2-38 days). The duration of viral shedding was not significantly different between patients who had URTI or LRTI (p=0.59). Of the remaining 14 patients, 4 did not have negative follow-up testing as they died from the infection and in 10 patients follow up testing was not obtained. For diagnosis of the initial infection, the virus was isolated from culture in all but 1 patient who was identified by PCR only. A positive DFA was obtained with the culture in 18 (39%) patients. PCR tested positive with culture in all 6 patients where both methods were used. Co-pathogens were identified from the respiratory secretions in 8 (17%) patients. These included adenovirus in 3 patients and Candida parapsilosis, HSV, Pseudomonas aeruginosa and RSV in 1 patient each. Cytomegalovirus (CMV) viremia and adenoviremia was seen in 1 patient each. Two patients had recurrent infections with a different serotype, 8 months and 4 months after the initial infection.

Risk factors for LRTI

There were 13 patients who had PIV infections less than 30 days after transplant, 10 of these infections were LRTI. Another 13 patients had PIV infections more than 1 year after transplant, 5 of these infections were LRTI. The other patients were proportionately distributed in the period 30-100 days, 100-180 days and 6 months to 1 year after transplant. Infection acquired less than 100 days after transplant (p=0.006), and the use of corticosteroids at the time of onset of infection (p=0.035; Table 2) were significantly associated with LRTI. Steroids were used for treatment of acute GVHD of grade II-III severity in 8 patients. GVHD was limited to the skin in 4 patients, gut in 1 patient, combined skin and liver in 2 patients and skin and gut in 1 patient. There were 2 other patients who received steroids for engraftment syndrome. Eight patients received steroids at a dose of 2 mg/kg/day and 2 patients at a dose of 0.5 mg/kg/day. Other factors that increased the risk for LRTI included ANC <500 cells/μL (p<0.0001), ALC <100 cells/μL (p<0.0001) and ALC<500 cells/μL (p=0.002), at the time of infection (Table 2).

Table 2.

Risk factors for LRTI from symptomatic PIV infections

Characteristic	LRTI n=18	URTI n= 28	p-value
Infection <100 days post transplant	13	8	0.006
Infection >100 days post transplant	5	20
Use of steroids prior to infection	7	3	0.035
No steroids prior to infection	11	25
ANC<500 cells/μL	9	1	<0.0001
ANC>500 cells/μL	9	27
ALC<100 cells/μL	12	3	<0.0001
ALC>100 cells/μL	6	25

LRTI, lower respiratory tract infection; URTI, upper respiratory tract infection; PIV, parainfluenza virus; ANC, absolute neutrophil count; ALC, absolute lymphocyte count.

Patients with an ALC<500 cells/μL also had a more prolonged duration of viral shedding with a median of 14 days (range 2-38 days) as compared to a median of 7 days (range 2- 35 days) in patients with an ALC>500 cells/μL (p=0.045). Age, race, gender, seasonality of infection, underlying diagnosis, remission status, transplant and donor type, donor CMV status, presence of GVHD and serotype of PIV were not associated with LRTI. Pulmonary function testing was performed in 10 of the 12 patients with LRTI who survived one year after the episode. This demonstrated no abnormalities in 8 patients, decreased diffusion capacity in 1 patient and restrictive lung disease in the other.

Of the 46 patients with PIV infection, 32 (70%) received TBI-based conditioning. LRTI was diagnosed in only 11 (34%) of the 32 patients. Three of these patients succumbed to the infection. Of the 8 surviving patients, 2 had abnormalities of lung function as noted above, one year after transplant, showing evidence of decreased diffusion capacity in 1 patient and restrictive lung disease in the other.

Risk factors for mortality in patients with symptomatic PIV infections

Mortality was attributed to PIV infection in 6 (13%) patients. Of the 6 patients, 4 (67%) were African-American (AA), as compared to 10 (22%) in the study group (p=0.013). Children of AA-ethnicity were more likely to have received steroids prior to the onset of infection (p=0.008). Presence of GVHD (p=0.04) and ANC <100 cells/ μL (p=0.04) was noted to have a marginally significant association with PIV infection in the AA group. The occurrence of LRTI (p=0.12), infection <100 days post transplant (p=0.71) and ALC<100 cells/μL (p=0.24), was not significantly different in children with AA as compared to non-AA ethnicity.

All the deceased patients were recipients of allogeneic HSCT, had LRTI and died of respiratory failure. BAL was the source of the respiratory sample in all these patients. Mechanical ventilation was initiated in the 6 patients at a median duration of 2 weeks from the onset of symptoms (range 2 days-4 weeks). Death occurred at a median duration of 3 wks after ventilation (range 2-6 weeks). LRTI (p=0.002) and need for mechanical ventilation (p <0.0001) was significantly associated with mortality (Table 3). Acute GVHD was seen in 4 of the 6 deceased patients and 5 patients received steroids for management of GVHD. Use of steroids at the time of onset of infection was associated with mortality (p<0.0001). An ALC <100 cells /μL at onset of infection was also associated with mortality (p=0.01; Table 3). Age, gender, donor type, donor CMV status, time of infection after transplant, underlying disease, type of conditioning, duration of viral shedding and presence of co-pathogens were not associated with mortality. Infection was caused by PIV-1 in 1 patient, PIV-2 in 1 patient and PIV-3 in 4 patients. PIV-3 infections were not significantly associated with LRTI (p=0.75), or mortality (p=1.0). Co-infections were seen with adenovirus, RSV, Pseudomonas aeruginosa and CMV in 1 patient each. Of the 18 patients with LRTI, 3 received ribavirin within 24 hours of diagnosis and 2 after mechanical ventilation was initiated. There was no apparent effect of therapy with ribavirin on mortality (p=0.12) or duration of viral shedding.

Table 3.

Risk factors for mortality from symptomatic PIV infections

Characteristic	Mortality		p-value
	Yes n=6	No n=40
African-American race	4	5	0.013
White	2	28
Other	0	7
LRTI	6	12	0.002
URTI	0	28
Mechanical ventilation	6	0	<0.0001
No mechanical ventilation	0	40
Use of steroids prior to infection	5	4	<0.0001
No steroids prior to infection	1	36
ALC<100 cells/μL	5	10	0.01
ALC>100 cells/μL	1	30

PIV, parainfluenza; LRTI, lower respiratory tract infection; URTI, upper respiratory tract infection; ALC, absolute lymphocyte count.

DISCUSSION

This retrospective study describes the clinical characteristics of symptomatic PIV infections, in children who had undergone HSCT over a 15 year period. Although studies in adult HSCT patients have shown that PIV infections are relatively common and cause significant morbidity and mortality, their role in immune-compromised children remains less well understood.

In a population of otherwise healthy children, in an out-patient setting, PIV infection was cultured in 286 out of 5,099 (5.6%) samples, collected prospectively over a 20 year period. PIV infections caused 64% of croup, 22% of LRTI and 20% of URTI in these pre-school children [7]. In a population based study of the inpatient burden of PIV infections, Weinberg et al identified 191(6.8%) cases of PIV, 529 (18.7%) cases of RSV and 159 (5.7%) cases of influenza in 2,298 immune-competent children with acute respiratory illnesses or fever [5]. The type distribution was 50% PIV-3, 38% PIV-1 and 12% PIV-2.

In our population, PIV caused infection in 6.2% of patients with respiratory illnesses and was more frequent than influenza A (4.9%) and RSV (3.9%). The type distribution was 65% PIV-3, 17% PIV-1, 11% PIV-2 and 4% PIV-4. This is similar to the incidence in otherwise healthy children, but contrasts with the experience in adult recipients of HSCT where 90% or more patients had PIV-3 infections [1, 3].

In otherwise healthy children, PIV-1and PIV-2 tends to occur biennially in the fall of odd-numbered years, while PIV-3 occurs throughout the year with a peak in spring and early summer [7]. In our population, PIV maintained this seasonal and temporal distribution. CA infections were observed in 72% of our patients; however no clustering of episodes was seen. CA infections were seen in 78% of adult HSCT patients with PIV infection, 58% of these infections were in clustered episodes [1]. Outbreaks of HA infection have also been reported in adult HSCT patients [2, 3].

The incidence of LRTI has ranged from 25% in a series of 222 HSCT patients with PIV infection as shown by Nichols et al [1], to 70% in a series of 27 HSCT patients including 15 children as reported by Wendt et al [4]. Thirty nine per cent of our patients had LRTI confirmed mainly by the presence of diffuse interstitial infiltrates demonstrated by chest radiography or chest CT and multiple bilateral nodules in 2 patients. Multiple bilateral small nodules <10mm in diameter have been associated with a diagnosis of viral pneumonitis [8]. Only one child presented with croup. This contrasts with the incidence of 64% reported for croup, in otherwise healthy pre-school children with PIV infection [7 ]. The median duration between onset of symptoms and diagnosis of 2 days (range 1-7 days) compares well with a median duration of 9 days (range 3-60 days) using culture methods alone [4], and a median duration of 4 days (range 0-22 days) using culture and DFA [2], where over 50% of patients already had signs of LRTI before a diagnosis was made.

Culture and DFA detection were used for the entire duration of the study. There were no significant changes in the culture and DFA techniques during this period, which would have affected the sensitivity of the test. PCR was performed on respiratory samples obtained after February 2007. For diagnosis of the initial infection, the virus was isolated from culture in all but 1 patient who was identified by PCR only. PCR tested positive with culture in all 6 patients where both methods were used. One patient was detected by PCR alone. Only 7 patients were tested by culture, DFA and PCR, hence a method comparison of the sensitivity of PCR with DFA and culture was not made. Real-time PCR provides a much more sensitive test than conventional techniques to detect PIV as well as being more rapid than viral culture [9, 10]. This assay could improve management of immunecompromised patients by enabling prompt isolation after early detection.

Receipt of an allogeneic transplant significantly increased the risk of PIV infection in our patients. Receipt of an unrelated transplant has been reported as a risk factor for PIV-3 acquisition [1]. In a smaller series of adult HSCT patients, Elizaga et al [2], found a similar frequency of PIV-3 infections in allogeneic and autologous transplants and with matched sibling donors as compared to unrelated donors. PIV-3 pneumonia occurs after autologous transplantation mainly in the setting of CD34 selection or use of high-dose steroids [1].

TBI was given in 2 fractions of 1.5 Gy each daily, for 4 days, with lung shielding. There is evidence linking acute idiopathic interstitial pneumonitis with TBI even with fractionated doses [11]. However, an increased risk of acquiring respiratory viral infections has not been reported. The majority (66%) of patients with PIV infection, who had received TBI-based conditioning had URTI. The association between total body irradiation and PIV infection would need to be confirmed by further studies.

Infection acquired less than 100 days after transplant was significantly associated with LRTI in our study. Of the 24 adults who developed PIV-3 infection, 20 (83%) developed the infection within 100 days of transplant [2]. Infection acquired less than 100 days after transplant was also associated with a higher frequency of LRTI and increased mortality in a retrospective study of adult HSCT recipients by Lewis et al [3]. Use of corticosteroids at the time of onset of infection was significantly associated with LRTI. The dose of corticosteroids at the onset of PIV-3 infection was the primary factor associated with the development of PIV pneumonitis in the study by Nichols et al [1]. It is likely that the effect of corticosteroids is due to a decline in T cell immunity. Chakrabarti et al [12], have shown the probability of respiratory virus isolation to be as high as 35% in adults conditioned with a non-myeloablative regimen that included Campath-1H, which has a long half-life, depletes T cells and impedes recovery of CD4+ T cells in the early post-transplant period.

ALC<100 cells/μL was associated with development of LRTI. The association of LRTI with severe lymphopenia may reflect an increase in progression with severe myelosuppression. Infection less than 100 days post HSCT was highly correlated with an ANC <500 cells/μL in the statistical analysis. It is likely that time post HSCT is the more meaningful independent risk factor for LRTI. Patients with an ALC<500 cells/μL had a more prolonged duration of viral shedding. In a prospective study of respiratory viral infections in 70 adult allogeneic HSCT recipients of whom 23 had LRTI, severe lymphopenia was associated with the risk of developing LRTI [13]. This included 1 patient with PIV-3 infection. In the study by Lewis et al [3], a significantly higher number of patients with pneumonia were lymphopenic, compared to those with uncomplicated URTI. However the frequency of neutropenia in the two groups was similar. Late airflow obstruction has been linked to both PIV LRTI and URTI [14]. In our series only 2 of the 10 surviving patients with LRTI tested, had abnormalities on pulmonary function testing, performed one year after the infection, in the form of restrictive lung disease and decreased diffusion capacity.

LRTI and need for mechanical ventilation was associated with increased mortality. Mortality in adult HSCT patients was influenced by the need for mechanical ventilation and by the presence of co-pathogens which was seen in 53% of patients with LRTI [1], as compared to 17% of patients in the present series. Aspergillus fumigatus occurred in 24% of cases [1]. This may partly explain the high 30 day mortality of 35% and 180 day mortality of 75% as compared to 17% in the present series. A high mortality of 34% associated with LRTI was also observed by Lewis et al. where 70% of the deceased patients had other serious concurrent infections [3].Other factors contributing to the increased mortality in adults may include presence of co-morbid conditions and a higher incidence of GVHD with consequent T cell suppression. Mortality was also influenced by onset of infection before day100 [1-3] and presence of LRTI [1, 4]. Although no randomized studies have been published, neither aerosolized ribavirin nor intravenous immunoglobulin has led to an improved outcome [1]. In the present series there was no effect of ribavirin on mortality or duration of viral shedding. Whether pre-emptive treatment of URTI with aerosolized ribavirin may prevent progression to LRTI is unknown. Although the incidence of PIV infections after reduced-intensity conditioning was high, the mortality was only 8%, perhaps as a result of reduced regimen-related toxicity and low incidence of GVHD. Since use of corticosteroids has been observed in several studies to be a risk factor for LRTI, and was associated with mortality in our series, judicious use of steroids may be considered in the presence of PIV URTI.

Patients of AA-ethnicity were more likely to have received steroids prior to onset of infection, which was a risk factor for LRTI. However, the number of AA patients was small in the present series, and hence the association with increased mortality should be interpreted with caution. Further studies are needed to corroborate if AA ethnicity is a risk factor for mortality from respiratory virus infections, in children undergoing HSCT.

Asymptomatic carriage of PIV in adult recipients of HSCT, with a cumulative incidence at day 100 of 17.9% has been noted [15]. Asymptomatic shedding of the virus for long periods of time in immune-competent adults has been observed [16]. PIV may be present in the respiratory secretions of healthy children without symptoms [17]. Further, the virus tends to persist on environmental surfaces for long periods of time [18]. These factors may predispose to outbreaks in the hospital setting. Bivalent vaccines for PIV-3 and RSV have been developed which are safe and immunogenic [19], and are undergoing further trials. In addition, a vaccine for PIV-1 and a bivalent vaccine for PIV-1 and RSV are under development [20, 21]. Since the majority of infections are community-acquired, with young infants as the reservoir of infection, this strategy may prove to be useful.

Our study has several limitations including its retrospective nature and small number of events per variable. The latter along with the relatively large number of significant marginal risk factors precluded multiple logistic regression analysis that can adjust for possible confounding factors, for evaluating the risk for LRTI and mortality [22]. Further, there is ascertainment bias as only children with respiratory symptoms were screened, at the discretion of the treating physician.

In conclusion, this is the largest retrospective study of PIV infections to date in children, including 1028 patients, who had undergone HSCT over a 15 year period. PIV infections occurred in 46 (6.2%) of patients and were the most common respiratory virus infections in this population. Receipt of an allogeneic transplant and TBI-based conditioning increased the risk for PIV infection. Risk factors for LRTI which occurred in 18 (39%) patients included infection before day+100, use of corticosteroids at the time of onset of infection, and ALC <100 cells/μL. Croup was noticeably uncommon as a presenting manifestation of PIV infection. Risk factors for mortality included LRTI, use of corticosteroids, need for mechanical ventilation and ALC<100 cells/μL. Mortality from PIV pneumonia is high. New agents are needed, and until then preventive strategies may need to focus on surveillance of both asymptomatic and symptomatic patients, and compliance with hand hygiene practices to prevent spread of PIV infection.