Pathology of parainfluenza virus infection in patients with congenital immunodeficiency syndromes

Pathology of parainfluenza virus infection in patients with congenital immunodeficiency syndromes

Pathology of Parainfluenza Virus Infection in Patients With Congenital Immunodeficiency Syndromes JOHN F. MADDEN, MD, PHD, JAMES L. BURCHETTE, JR., HT...

670KB Sizes 1 Downloads 24 Views

Pathology of Parainfluenza Virus Infection in Patients With Congenital Immunodeficiency Syndromes JOHN F. MADDEN, MD, PHD, JAMES L. BURCHETTE, JR., HT(ASCP), AND LAURA P. HALE, MD, PHD Infection with parainfluenza virus typically produces a mild, self-limited upper respiratory infection. However, parainfluenza infections have become increasingly recognized as a source of severe morbidity and mortality in immunocompromised patients. In this retrospective study we identified 6 patients with congenital immunodeficiency and positive respiratory cultures for parainfluenza virus who died and underwent complete autopsy. Tissues obtained at autopsy were studied using hematoxylin and eosin–stained sections, immunoperoxidase staining for parainfluenza virus, and in selected cases, electron microscopy. All 6 patients exhibited typical cytopathic effects of parainfluenza virus, including giant cell formation, in lung and/or bronchial tissues. Parainfluenza virus infection was also documented by giant cell formation and immunohistochemistry in the pancreas (in 3 of 6 patients) and the kidney or bladder (in 2 of 4 patients). Anti-parainfluenza antibody also specifically reacted with cells in the gastrointestinal tract (in 2 of 4), spleen (in 4 of 6), thymus

and/or lymph nodes (in 4 of 4), and small blood vessels in various organs (in 4 of 6). Pancreatic, bladder, colon, and thymic epithelial cell lines were susceptible to experimental infections with clinical isolates of parainfluenza virus type 3 in vitro. Parainfluenza virus infection was serious in patients with congenital immunodeficiencies, contributing directly to death in 5 of the 6 patients studied. Because this virus is capable of infecting tissues in the gastrointestinal and urinary systems as well as in the respiratory tract, body secretions and fluids from each of these locations should be considered potentially infectious. HUM PATHOL 35:594-603. © 2004 Elsevier Inc. All rights reserved. Key words: paramyxovirus, severe combined immunodeficiency, cytopathic effects. Abbreviations: mAb, monoclonal antibody; SCID, severe combined immunodeficiency.

Parainfluenza viruses belong to the Paramyxoviridae family of enveloped RNA viruses containing nonsegmented, single-stranded, negative-sense genomes. This family also includes the respiratory syncytial, mumps, and measles viruses. Four types of parainfluenza viruses have been identified, each of which has a distinct propensity to involve upper versus lower respiratory tracts or to cause croup. In addition, each type of parainfluenza virus has a distinct epidemiological pattern of outbreaks.1 Population-based studies indicate that up to 2/3 of children become infected with parainfluenza type 3 in their first 2 years of life.2 Although most infections cause mild, self-limited illness, parainfluenza 3 infection is an important cause of bronchiolitis, pneumonia, and croup in infants and newborns.3 This virus has also become increasingly recognized as an important source of morbidity and mortality in immunocompromised patients, both adults4-13 and children,14-16 after bone marrow, umbilical cord blood, or solid organ transplantation and in the institutionalized elderly.17 Features of respiratory virus infections in immunocompromised patients include persistent infection and prolonged viral shedding, high frequency of progression to

pneumonia, and high mortality rate relative to the general population.15 The mortality of bone marrow transplant patients with parainfluenza 3 lower respiratory infection has been reported to be as high as 37% to 60%.6,18,19 Ribavirin is the only antiviral drug currently approved for treating paramyxovirus infections. However, 6 of 10 patients who developed parainfluenza 3 pneumonia after hematopoietic stem cell transplantation died despite receiving ribavirin at a median of 3 days after onset of symptoms, and thus ribavirin may not alter mortality from parainfluenza infection in this patient population.11 Severe combined immunodeficiency (SCID) is a clinical syndrome that results from mutations in genes necessary for development and function of both T and B lymphocytes. Patients with SCID exhibit no significant T- or B-cell function; thus they are highly susceptible to infection and typically die from infection within the first year of life if untreated.20 SCID is typically treated by lymphocyte-depleted haploidentical bone marrow transplantation.21,22 SCID patients who are treated by bone marrow transplantation within the first 4 weeks of life were found to have increased survival relative to SCID patients treated similarly after the neonatal period (95% of 21 vs 74% of 96 infants).23 Unfortunately, many patients with SCID are not identified until they have experienced multiple infections. Taylor et al24 reported that 7 of 27 children with SCID who underwent bone marrow transplantation at their center had parainfluenza infection at the time of admission. These infections may limit the range of possible treatments or cause death before transplantation is done or before engraftment occurs. Several re-

From the Department of Pathology, Duke University Medical Center, Durham, NC. Accepted for publication November 4, 2003. Supported by funds from the Department of Pathology, Duke University Medical Center. Address correspondence and reprint requests to John F. Madden, MD, PhD, DUMC 3712, Duke University Medical Center, Durham, NC 27710. 0046-8177/$—see front matter © 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.humpath.2003.11.012

594

PARAINFLUENZA IN CONGENITAL IMMUNODEFICIENCY SYNDROMES (Madden, Burchette, and Hale)

TABLE 1. Clinical Characteristics of Patients With Parainfluenza 3 Infection Patient

Age at death (months)

1 2 3

21 21 15

4 5

12 8

6

15

Treatment (age)

Underlying immunodeficiency

Concurrent infections

fetal liver (18 months) fetal liver (21 months) HBMT ⫻ 3 (5, 10, and 13 months) HBMT HBMT (7 months)

Unknown SCID RAG2⫺ SCID IL7RA⫺ SCID

Pseudomonas pneumonia Aspergillus pneumonia Pseudomonas infection

IL2RG⫺ SCID IL2RG⫺ SCID

Chemotherapy and BMT (13 months)

Unknown

Pseudomonas and candida infections Adenovirus, Pneumocystis carinii, respiratory syncytial virus pneumonia Disseminated aspergillus; candida gastroesophagitis

NOTE: All patients were male. Abbreviations: HBMT, haplo-identical T-cell– depleted bone marrow transplant; BMT, bone marrow transplant; RAG2⫺, recombinase activating gene 2 deficiency; IL7RA⫺, interleukin-7 receptor alpha deficiency; IL2RG⫺, X-linked, common cytokine receptor gamma chain (␥c) deficiency.

ports have described fatal parainfluenza virus pneumonia in infants with SCID,25,26 particularly in those with high viral loads.24 A study that prospectively obtained viral cultures from patients with primary immunodeficiencies before and after bone marrow transplantation found that no SCID patient cleared the parainfluenza virus before T-cell engraftment, even with ribavirin treatment.27 It has been suggested that a lack of T-cell–mediated immunity may attenuate the inflammatory lung damage caused by respiratory viruses.19 However, infection with parainfluenza 3 has been a significant cause of severe morbidity and mortality in our SCID patient population over the past 20 years. In this article we describe the range of pathology seen with systemic parainfluenza 3 virus infection in 6 patients with congenital immunodeficiency (5 with SCID) who died despite receiving treatment for the immunodeficiency. We focus on defining the pathological characteristics of virally infected cells and the types of tissues infected in these patients, who typically had no detectable inflammatory response to this serious viral infection.

MATERIALS AND METHODS

Viral Culture and Confirmation of Parainfluenza Virus Infection Antemortem or postmortem cultures were used to confirm the diagnosis of parainfluenza infection for all patients studied. Direct fluorescence assays were performed on clinical specimens before culturing using a fluorescent anti-parainfluenza monoclonal antibody (mAb) (Trinity Biotech, Wicklow, Ireland). Samples were inoculated into tissue cultures containing rhesus monkey kidney cells, cynomologus monkey kidney cells, human neonatal kidney cells, or human foreskin fibroblasts (ViroMed, Minnetonka, MN) and were kept for at least 14 days. Growth of parainfluenza 3 was confirmed using a type-specific fluorescent mAb (Light Diagnostics, Temecula, CA). For assays of experimental infection, human cell lines plated in 4-well LabTek chamber slides were exposed to supernatant from rhesus monkey kidney cells infected with 3 different clinical isolates of parainfluenza 3. Cultures were visually examined daily for cytopathic effects. Immunohistochemical staining with anti-parainfluenza antibody was performed on days 3, 5, and 7. Cell lines tested included the J82 bladder carcinoma, Caco-2 colon carcinoma, and PANC-1 pancreatic carcinoma cell lines obtained from the American Type Culture Collection (Manassas, VA). The TE750 thymic epithelial cell line was obtained from Dr. Dhavalkumar D. Patel (University of North Carolina, Chapel Hill, NC).

Tissue Analysis

Identification of Cases A retrospective search was performed to identify all patients with a diagnosis of SCID who underwent autopsy at Duke University Medical Center between January 1981 and October 2002. The cause of SCID was first defined clinically, then confirmed by genotyping of patients for mutations in the indicated genes.21 A total of 18 SCID patients were identified. Medical records, autopsy reports, and pathological slides from these cases were reviewed. Five patients with SCID and parainfluenza 3 infection were identified. One additional case of parainfluenza virus infection was identified in an infant with severe congenital immunodeficiency of unknown type. Detailed study of these tissues for research was approved by the Duke Institutional Review Board. The clinical characteristics of the patients studied are listed in Table 1.

The tissues examined were fixed in 10% neutral buffered formalin, then processed into paraffin blocks using standard protocols. Although parainfluenza virus has been thought to primarily affect the respiratory tract, it is possible that other tissues may become infected if viral titers are high, as often occurs in immunodeficient patients. To determine the range of tissues that could be infected in vivo by parainfluenza virus in these immunodeficient patients, we examined all tissues available from the autopsies of these patients for viral cytopathic effects and for immunohistochemical reactivity with anti-parainfluenza virus mAb. Hematoxylin and eosin–stained sections and anti-parainfluenza virus–immunostained sections were examined for each tissue reported. However, not all tissues were examined at autopsy for all patients, and some of the tissues examined at autopsy were not available for

595

HUMAN PATHOLOGY

Volume 35, No. 5 (May 2004)

additional studies. Immunohistochemical staining was performed as described previously,28 with digestion by 0.25% stable pepsin solution, pH 2.0, for 13 minutes at 37 to 40°C instead of trypsin. The mouse anti-parainfluenza I, II, and III mAb cocktail (catalog no. MAB819; Chemicon, Temecula, CA) was used at a dilution of 1:1500. This antibody cocktail contains mAb clones with specificities for F0/F1, NP, and HM proteins and reacts with all known serotypes of parainfluenza 1, 2, and 3. Cells grown on chamber slides were fixed in acetone for 5 minutes at room temperature before immunostaining, using a 1:2500 dilution of anti-parainfluenza antibody. Transmission electron microscopy was performed on formalin-fixed tissues from cases 4 and 6 after postfixation in glutaraldehyde and staining with uranyl acetate and lead citrate according to standard techniques.

serous otitis media, and failure to thrive. Parainfluenza 3 was isolated from the nasopharynx on several occasions. The patient eventually developed increased cough, fever, and bilateral perihilar pulmonary infiltrates. A lung biopsy specimen demonstrated extensive pulmonary aspergillosis. A fetal liver transplant from a 6- to 8-week-old fetus was performed; however, the patient died 2 weeks later from respiratory failure at age 21 months. Autopsy was limited to exclude the brain. Autopsy cultures of lung grew Aspergillus fumigatus and parainfluenza 3 virus, with no additional fungal, bacterial, or viral growth from cultures of liver, spleen, or blood. Genetic studies completed after the patient’s death revealed a deficiency of RAG2 as the cause of his immunodeficiency.

RESULTS Report of Cases

Case 3

Case 1

This male patient had 2 male siblings, 1 who died at age 10 months with pneumonia and 1 who died at age 6 weeks with complications of situs inversus. This patient’s neonatal course included chronic mild diarrhea, oral thrush, and mucoid rhinorrhea. He developed persistent right middle lobe pneumonia at age 6 months. Serum immunoglobulin levels were low (IgG, 100 mg/dL; IgM, 36 mg/dL; IgA, 7 mg/dL), and he was referred to this institution at age 8 months. Physical examination revealed no tonsillar or adenoidal tissue. Chest radiographs showed a possible small thymic shadow. T cell levels were low, with near absence (⬍1%) of CD8⫹ cells. There was no significant proliferation to phytohemagglutinin, concanavalin A, or pokeweed mitogens. The patient was diagnosed with SCID. A fetal liver transplant from a 9.5-week-old fetus failed to show evidence of engraftment. The patient’s clinical course was characterized by multiple infections. Cultures were consistently positive for Candida and Klebsiella pneumoniae. Later, Pseudomonas was isolated from throat and stool. Adenovirus was found in the stool. The patient developed increased pulmonary infiltrates and an enlarged cardiac shadow. Parainfluenza 3 was isolated from nasopharynx, pericardial fluid, cerebrospinal fluid, blood, tracheal aspirates, and urine. The patient developed refractory hypotension and hypoxia, leading to his death at age 21 months. A full autopsy was performed. The molecular cause of SCID in this patient is still unknown. Case 2

This patient did well until about age 4 months, when he developed chronic otitis media refractory to antibiotics. Laboratory studies revealed lymphopenia and severe hypoimmunoglobulinemia with no detectable IgM or IgA. Peripheral T cells were markedly decreased and no mitogen response was present. A probable small thymic shadow was visible on chest Xray. The clinical picture was diagnostic of SCID. The clinical course was complicated by anemia, chronic

Although this patient seemed normal at birth, by age 2 to 3 months he had developed bronchiolitis, with persistent cough, wheeze, nasal discharge, and oral thrush. A direct fluorescent antibody test for parainfluenza virus 3 was positive. Serum immunoglobulin levels were markedly decreased (IgG, 54 mg/dL; IgA, 0 mg/ dL; IgM, 19 mg/dL; IgE, 14 mg/dL; isohemagglutinins, negative). The absolute lymphocyte count was also low (white blood cell count, 4.3, with 47% polymorphonuclear leukocytes, 20% bands, 20% lymphocytes, 13% monocytes, and 0% eosinophils). Genetic studies showed that the patient had deficiency of the alpha chain of the interleukin-7 receptor, a known cause of SCID. Over the next 7 months, the patient received 3 separate haploidentical T-cell– depleted bone marrow transplants (the first 2 from his mother and the third from his father) without evidence of engraftment. His 11-month hospitalization was complicated by persistent sinusitis, pneumonia, and several episodes of sepsis. Parainfluenza 3 virus was repeatedly cultured from nasopharyngeal aspirates, 2 separate bronchoalveolar lavage specimens, and a sinus aspirate. Pseudomonas was also found repeatedly in cultures of urine, blood, ear, endotracheal suction, sinus aspirate, and nasal washes. Enterovirus and adenovirus were documented in the stool. Despite aggressive antibiotic treatment of these infections, the patient died of respiratory failure at age 15 months. An autopsy limited to the chest and abdomen was performed. Case 4

This 12-month-old male infant had X-linked SCID due to IL2RG deficiency. His immunodeficiency was diagnosed at age 8 months after 4 months of persistent cough and oral and diaper-area candidiasis followed by an acute febrile illness. Lymphopenia was identified, with 4% peripheral T cells and no response to mitogens. IgG and IgM levels were low, with absence of isohemagglutinins and antibodies to diptheria and tetanus. The patient was treated with a haploidentical bone marrow transplant from his mother. His subsequent course was complicated by multiple episodes of

596

PARAINFLUENZA IN CONGENITAL IMMUNODEFICIENCY SYNDROMES (Madden, Burchette, and Hale)

bacterial and candidal sepsis, with progressive respiratory and hepatic failure. He died of respiratory failure without evidence of bone marrow engraftment 4 months after his bone marrow transplant. Postmortem blood and lung cultures grew Pseudomonas aeruginosa. Candida albicans was additionally isolated from the blood culture. Immunofluorescence studies performed in the virology lab were positive for parainfluenza 3. Electron microscopy of lung tissue revealed viral particles consistent with paramyxovirus. Postmortem genetic analysis confirmed a mutation in the IL2RG gene. Case 5

This patient’s clinical course has been described in detail elsewhere.29 Briefly, the patient was an 8-monthold male who was diagnosed with X-linked SCID at age 6 months. Streptococcus pneumoniae bacteremia/pneumonia, parainfluenza 3, and respiratory syncytial virus infections were documented by cultures, and Pneumocystis carinii pneumonia was diagnosed by bronchoalveolar lavage. He was treated for these infections but subsequently developed disseminated adenoviral infection. A haploidentical bone marrow transplant from his mother was administered, but the patient died from pulmonary hemorrhage before engraftment occurred. Autopsy was limited to the chest and abdomen. Postmortem genetic analysis confirmed a mutation in the IL2RG gene. Case 6

Immunodeficiency was first suspected in this male infant at age 2 months, when he was hospitalized with bronchiolitis and enteroviral meningitis. Serum IgG was 98.1 mg/dL, IgM was 36.2 mg/dL, and IgA was undetectable. No B cells or B-cell precursors were identified on flow cytometry. Cytogenetic analysis identified trisomy 20 in bone marrow cells. A diagnosis of atypical Bruton’s agammaglobulinemia was considered; however, genetic studies were not performed. The patient became pancytopenic and was treated with transfusion and intravenous immunoglobulin. Cultures documented parainfluenza 3 respiratory infection at age 8 months that required hospitalization including mechanical ventilation for 2 days. Other infections included salmonella enteritis and candidal sepsis. Chemotherapy and bone marrow transplantation were performed at age 14 months to treat increasing bone marrow dysplasia. The patient died at age 15 months from refractory hypotension. Autopsy documented invasive candida and aspergillus infections, in addition to disseminated parainfluenza virus. Respiratory Infection With Parainfluenza Virus All 6 patients had cytopathic effects consistent with parainfluenza virus infection in either the lungs (in 5 of the 6), trachea and bronchi (in 4 of the 6), or both. Infected alveolar epithelial cells typically appeared highly reactive, with enlarged, hyperchromatic nuclei

and prominent nucleoli. Most cases contained multinucleated giant cells derived from alveolar epithelium (Fig 1A and B). Eosinophilic cytoplasmic inclusions were prominent within giant cells. Cytopathic effects consistent with parainfluenza virus infection were typically present in multiple discrete microscopic foci scattered throughout both lungs, separated by uninfected lung tissue. Infected foci were often associated with granulation tissue and fibrosis (Fig 1C). These foci stained strongly with anti-parainfluenza virus mAb, with a granular cytoplasmic pattern (Fig 1B). Although many multinucleated giant cells reacted strongly with anti-parainfluenza virus mAb, some appeared to be nonreactive. In case 6, cytopathic effects, including alveolar epithelial giant cells, were not observed in lung parenchyma, but the alveoli immediately adjacent to major bronchi reacted with anti-parainfluenza antibody (Fig 1D). Inflammatory infiltrates were typically absent in these immunocompromised patients. Bronchial involvement varied from intense, with highly reactive bronchial epithelium containing multinucleated giant cells (Fig 1E) and strong immunostaining, to focal, with only a few bronchial epithelial cells involved. Electron microscopy of reactive bronchial epithelium in patient 6 showed cellular membranes thickened with viral proteins, budding virus, and mature enveloped paramyxovirus virions, consistent with the diagnosis of parainfluenza infection (Fig 2). Bronchial mucous glands reacted with anti-parainfluenza virus mAb in 2 of the 4 cases with bronchial epithelial involvement (Fig 1F). Parainfluenza Virus Infection in the Digestive System We found that pancreatic acinar cells were infected with parainfluenza virus in 3 of 4 patients for whom pancreas tissues were available. Two cases demonstrated cytopathic effects that included giant cell formation (Fig 1G; Table 2). The presence of giant cell pancreatitis in one of these patients was described previously.30 The remaining 2 cases exhibited granular cytoplasmic immunostaining for parainfluenza virus in multiple acini scattered throughout the pancreas (Fig 1H). The combination of concurrent cytopathic effects and the positive immunostaining is consistent with true infection of pancreatic acinar cells rather than nonspecific staining. Positive staining was not seen in pancreas from patient 6, who had positive tracheal immunostaining for parainfluenza virus and lacked pancreatic giant cell formation. Rare giant cells were identified within gastric glands in both cases where the stomach was examined (Fig 1I). These cases also demonstrated granular cytoplasmic anti-parainfluenza immunostaining of the surface epithelium and/or mucosal glands of the stomach (Fig 1J and K) that appeared identical to the pattern of immunostaining seen in the pancreas and bronchial mucus glands. Two of 4 cases had positive immunostaining of the glands at the base of the villi in the small intestine and/or similar staining of colonic crypts (not

597

HUMAN PATHOLOGY

Volume 35, No. 5 (May 2004)

598

PARAINFLUENZA IN CONGENITAL IMMUNODEFICIENCY SYNDROMES (Madden, Burchette, and Hale)

TABLE 2. Organs Affected by Parainfluenza Virus Organ

FIGURE 2. Parainfluenza virus infection of the bronchus. Electron microscopy of bronchial epithelium (bar ⫽ 100 nm) shows cellular membranes thickened by virus proteins (left), budding virus, and enveloped virons (*) characteristic of parainfluenza virus infection. The normal membrane of an adjacent uninfected cell is indicated by arrowheads.

shown). Again, the absence of intestinal staining in 2 cases with documented parainfluenza infection of the respiratory tract and negative internal controls in cases with positive intestinal immunostaining suggests true, rather than nonspecific, staining. To further address the issue of possible antibody cross-reactivity with antigens present on noninfected gastrointestinal tissues, we immunostained stomach, small bowel, and colon tissue specimens from a series of control patients not clinically suspected to be infected with parainfluenza virus. Kidney tissues from a 35-yearold female who died from a motor vehicle accident and a 59-year-old female who died from stroke, along with uninvolved stomach from a 67-year-old with gastric cancer, did not react with the anti-parainfluenza virus antibody. However, we found positive immunostaining without evidence of cytopathic effects in a subset of gastric glands, at the base of small intestinal villi, and in colonic crypts in a 23-year-old male who died from trauma and in a 45-year-old male who died from myocardial infarction. The staining in these tissues was specific, affecting only subsets of glands, in patterns

Lung Bronchi Bronchial mucus glands Heart Pancreas Adrenal medulla Adrenal cortex Thyroid Parathyroid Kidney Transitional epithelium of renal pelvis Bladder epithelium Prostate Seminal vesicle Testis Liver Stomach Small intestine Colon Lymph node Spleen Thymus

Cytologically or immunohistochemically positive cases/cases examined (%) 6/6 (100%) 4/6 (67%) 2/3 (67%) 0/3 (0%) 3/4 (75%) 3/4 (75%) 0/4 (0%) 0/4 (0%) 1/2 (50%) 2/4 (50%) 2/2 (100%) 1/2 (50%) 0/2 (0%) 1/2 (50%) 0/3 (0%) 0/4 (0%) 2/2 (100%) 2/4 (50%) 2/4 (50%) 2/3 (67%) 4/4 (100%) 4/4 (100%)

similar to those seen in the immunodeficient children that we studied. This positive immunostaining in the absence of cytopathic effects could be due to crossreactivity with 1 or more of the antibody clones present within the anti-parainfluenza virus mAb cocktail, or to asymptomatic parainfluenza virus carriage or unsuspected infection. If antibody cross-reactivity exists, then it is to antigens expressed only in a subset of patients and not to antigens universally expressed by gastrointestinal tissues. Prospective studies, including viral cultures, are needed to further address the specificity of immunostaining in control tissues. Parainfluenza Virus Infection in the Urinary Tract Two of 4 cases also had parainfluenza infection of the kidney and/or bladder (Table 2). The collecting tubules of the kidney appeared highly reactive on hematoxylin and eosin–stained sections, but did not exhibit classic multinucleated giant cells. Rare multinu-

4™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™ FIGURE 1. Pathology of parainfluenza virus infection. (A and B) Lung was the major organ affected in the immunocompromised patients studied, with prominent alveolar giant cell formation. (C) Cytopathic effects were more frequent in areas of lung with associated fibrosis. (D) In case 6, infection of lung parenchyma was limited to alveoli directly bordering infected bronchi. (E) Epithelial cells of the trachea and bronchi also were infected in 4 of 6 patients, with reactive change and rare giant cells. (F) Bronchial mucus glands were reactive with anti-parainfluenza mAb in 2 of 3 cases where they could be evaluated. (G and H) Pancreatic acinar cells also demonstrated infection with parainfluenza virus. Giant cell formation was present focally (G, arrowheads); however, many of the pancreatic acinar cells that were immunohistochemically reactive with anti-parainfluenza virus mAb (H) did not show visible cytopathic effects. Rare giant cells (I) and focal reactivity with anti-parainfluenza virus mAb (J and K) were similarly observed in mucosal glands of the stomach. Epithelial cells of the collecting tubules of the kidney (L and M) and bladder (N and O) showed nuclear enlargement, prominent nucleoli, and strong granular reactivity with anti-parainfluenza virus mAb. Immunohistochemical reactivity with anti-parainfluenza virus mAb was also observed in cells present within the spleen (P) and thymus (Q). Small blood vessels were focally reactive with anti-parainfluenza virus antibody in 4 of 6 cases examined (Q). (Panels A, C, D, G, I, and N were stained with hematoxylin and eosin; the remaining panels were immunohistochemically stained with anti-parainfluenza virus mAb, with a hematoxylin counterstain. Brown color indicates positive reaction with mAb. Original magnifications were ⫻25 for D and J; ⫻50 for L; ⫻100 for A, B, H, and P; ⫻200 for C; and ⫻250 for E, F, G, I, K, M, N, O, and Q.)

599

HUMAN PATHOLOGY

Volume 35, No. 5 (May 2004)

cleated giant cells, sometimes containing eosinophilic inclusions, were seen in the transitional epithelium of the bladder (Fig 1N) or renal pelvis. The reactive collecting tubules and the transitional epithelium of the renal pelvis and the bladder from these patients stained strongly with anti-parainfluenza virus mAb (Fig 1L, M, and O). The concurrence of immunostaining and cytopathic effects observed on hematoxylin and eosin– stained sections suggest the specificity of the immunostaining for parainfluenza virus infection. We did not observe positive immunostaining in kidneys or bladder of 2 SCID patients with documented parainfluenza virus infections in other tissues or in kidney tissues obtained from from a 35-year-old female who died from trauma sustained in a motor vehicle accident and a 59-year-old female who died from stroke. Thus the anti-parainfluenza virus antibody that we used does not cross-react with an antigen present in all collecting tubules and transitional epithelium. However, we did observe positive immunostaining of collecting tubules in kidneys from a 23-year-old male who died from trauma and a 45-year-old male who died from myocardial infarction, and in 1 of 3 renal transplant nephrectomies examined. Further studies will be required to determine the significance of this staining.

and gastrointestinal tracts and lymphoid organs have, to our knowledge, not yet been reported. Although we found classical parainfluenza virus cytopathic effects in most of the tissues that we classified as infected, some tissues (eg, colon and thymus) that were immunoreactive with anti-parainfluenza virus mAb failed to exhibit cytopathic effects. To further investigate whether these tissues were truly susceptible to parainfluenza virus infection, we determined the susceptibility of epithelial cell lines derived from thymus and colon to experimental infection with parainfluenza virus. Cell lines derived from pancreas and bladder, organs that did exhibit typical cytopathic effects in vivo, were used as positive controls. The Caco-2 colon, TE750 thymic epithelial, PANC-1 pancreatic, and J82 bladder cell lines were each exposed to 3 different clinical isolates of parainfluenza virus type 3 in vitro. Each of these cell lines was readily susceptible to infection by parainfluenza virus infection in vitro (Fig 3). However, PANC-1 and J82 cells exhibited strong immunoreactivity with anti-parainfluenza virus mAb by day 3 postinfection, whereas similarly strong staining was not observed for TE750 and Caco-2 cells until at least day 5 postinfection. DISCUSSION

Other Pathological Findings Spleen (4 of 4 cases), adrenal medulla (3 of 4 cases), parathyroid gland (1 of 2 cases), and seminal vesicle (1 of 2 cases) also contained cells that were immunohistochemically reactive with anti-parainfluenza virus mAb (Table 2). In the spleen, reactivity with anti-parainfluenza virus mAb was typically focal and limited to macrophage-like cells; however, it affected the red pulp globally in 1 patient (patient 4) (Fig 1P). Scattered cells in thymus and/or lymph node were also reactive with anti-parainfluenza mAb in 4 of 4 cases examined. The positive cells in these organs had a morphology consistent with macrophages (Fig 1Q); however, it was not possible to rigorously exclude infection of thymic epithelial cells in thymic tissues. Small blood vessels, particularly capillaries, reacted with antiparainfluenza virus antibody in various organs in 4 of 6 cases examined (Fig 1Q, arrowheads). Organs with parenchymal cells that consistently failed to stain with anti-parainfluenza virus mAb included liver (0 of 4 cases) heart muscle (0 of 3 cases), thyroid (0 of 4 cases), adrenal cortex (0 of 4 cases), testis (0 of 3 cases), and prostate (0 of 2 cases) (Table 2). Five of the 6 cases examined also exhibited prominent erythrophagocytosis in lymph nodes and spleen and/or in the virally infected giant cells within the lungs (data not shown). Susceptibility of Cell Lines to Parainfluenza Virus Infection in Vitro Although giant cell pancreatitis has been previously reported in a patient with parainfluenza virus infection,30 infections of other organs in the urinary

The histopathology of most viral infections includes damage from the infectious organism (so-called “cytopathic effects”), as well as tissue damage resulting from the host inflammatory response. In the immunodeficient patients described here, the host inflammatory response is negligible. Thus the viral cytopathic effects can be visualized in the absence of host-induced damage due to inflammation. Although parainfluenza is considered primarily a respiratory virus, we found that it infected a much wider range of tissues in vivo in the immunodeficient patients studied, including tissues of the digestive and urinary systems. The concordance of cytopathic effects, such as reactive-appearing nuclei, eosinophilic inclusions, and giant cell formation with positive immunostaining, with anti-parainfluenza virus antibody suggest that these tissues are truly infected rather than simply exhibiting cross-reactivity with the diagnostic antibody. Although thymus and colon demonstrated at least focal reactivity with anti-parainfluenza virus mAb in some patients, classic cytopathic effects were not observed in these organs in vivo. Epithelial cell lines derived from colon and thymus were susceptible to experimental infection with clinical isolates of parainfluenza in vitro; however, clear evidence of viral infection was not observed until 5 days postinfection, versus 3 days for the pancreas and bladder cell lines. Differences in the relative susceptibility to infection and permissiveness for viral replication of various tissue types may explain the numbers of cells that exhibit viral cytopathic effects among the tissues examined in this study. The infectivity of parainfluenza viruses has been shown to depend on the availability of host trypsin-like enzymes to proteolytically cleave the viral F0 protein.31

600

PARAINFLUENZA IN CONGENITAL IMMUNODEFICIENCY SYNDROMES (Madden, Burchette, and Hale)

FIGURE 3. Parainfluenza 3 infection of epithelial cells in vitro. Epithelial cell lines derived from (A) thymus (TE750), (B) pancreas (PANC-1), (C) bladder (J82), and (D) colon (Caco-2) were highly susceptible to infection with parainfluenza 3 virus in vitro, confirming the potential susceptibility of these cell types to infection in vivo. (Anti-parainfluenza virus immunoperoxidase stain, hematoxylin counterstain; original magnification ⫻ 100.)

Production of trypsin by pancreatic acinar cells may thus contribute to their susceptibility to infection by parainfluenza virus. The range of tissues infected in vivo may reflect the relative expression of appropriate proteinases by these cell types. Infection of tissues not previously known to be targets for parainfluenza virus may explain pathophysiology and previously unexplained morbidity (eg, abdominal pain) or organ dysfunction in these patients. The potential susceptibility of cells and tissues of the urinary tract to parainfluenza virus infection perhaps should be not surprising, because parainfluenza virus is typically isolated in virology laboratories using monkey and/or human fetal or neonatal kidney cells. In addition to the immunohistochemical reactivity of collecting tubules of the kidney and transitional epithelium of the bladder with antiparainfluenza antibody, parainfluenza virus was cultured from the urine of patient 1. Viral culture of urine was not performed in the other 5 patients. Our identification of parainfluenza virus infection in tissues in the digestive and urinary systems as well as in the respira-

tory tract also suggests that body secretions and fluids (mucus, urine, vomitus, feces) from respiratory, gastrointestinal, and urinary tracts should be considered potentially infectious. Foci of parainfluenza-infected giant cells in the lungs often were located in sites with granulation tissue and fibrosis, raising the possibility that parainfluenza virus infection can result in permanent lung injury. However, many of the patients studied also suffered previous or concurrent bacterial infections (eg, Pseudomonas) that could result in lung scarring. It is impossible to determine from examination of the available tissue whether parainfluenza virus itself causes the observed tissue damage or whether the virus simply preferentially infects sites of previous or current bacterial infection or ongoing tissue repair. Lewis et al6 described intracytoplasmic inclusions in the lung of 6 of 7 adult bone marrow transplant patient with parainfluenza pneumonia (1 open lung biopsy specimen and 6 autopsy specimens). These authors did not describe the presence or absence of giant

601

HUMAN PATHOLOGY

Volume 35, No. 5 (May 2004)

cells in these tissues. Giant cell pneumonia due to parainfluenza is not limited to patients with primary immunodeficiency, but has been reported in a patient with metastatic carcinoma4 and in pediatric patients who underwent liver transplantation for alpha-1-antitrypsin deficiency14 or chemotherapy and cord blood transplantation for treatment of malignancy.16 We found parainfluenza pneumonia or bronchitis with giant cells in all 6 of the cases that we examined. Lymphopenia has been reported as a factor that favors development of pneumonia rather than uncomplicated upper respiratory infection due to parainfluenza virus.6 The mortality of parainfluenza and influenza infections in pediatric organ transplant patients was increased if immunosuppression (steroid bolus or OKT3) had been received within 1 week of infection.14 Parainfluenza virus type 3 has previously been reported as a cause of viral meningitis and encephalitis.32-34 Patient 1 had parainfluenza 3 cultured from cerebrospinal fluid antemortem, confirming viral infection of the central nervous system. Unfortunately, we were unable to examine brain tissues immunohistologically for effects of parainfluenza infection, due to tissue availability and/or autopsy permit limitations. Although myopericarditis associated with parainfluenza virus type 3 infection has been previously reported35 and virus was cultured from pericardial fluid in patient 1, pericardial tissue was not examined at autopsy in this patient or any of the other 5 patients in the present study. Parainfluenza myocarditis was not seen in the 3 heart specimens that were available for study. The 5 SCID patients reported here represent 28% (5 of 18) of the patients with SCID who died and underwent autopsy during the time period studied. Each of these patients had been treated for SCID, either by fetal liver transplant (for the earliest patients) or by bone marrow transplant, but had not engrafted by the time of infection. During the period studied, the overall survival of 117 SCID patients receiving bone marrow transplants at this institution was ⬃78%.23 Parainfluenza virus infection also contributed to the death of patient 6 who also died following bone marrow transplantation for treatment of his (non-SCID) immunodeficiency and bone marrow dysplasia but before engraftment. Failure of immunologic reconstitution therapy is thus a strong risk factor for fatal parainfluenza virus infection in patients with congenital immunodeficiency. Before closing, it is important to note that we identified immunohistochemical reactivity for parainfluenza virus in kidney and gastrointestinal tract in the absence of cytopathic effects in surgical and autopsy tissues from a subset of control patients not clinically suspected to have parainfluenza virus infection. Viral cultures were not obtained in these patients. Some of these tissues were derived from patients who were immunosuppressed due to transplant or chemotherapy for cancer; however, others were from patients with sudden death due to cardiovascular disease or trauma. Further studies, including prospective viral cultures with or without electron microscopy, will be necessary

to determine whether these findings represent asymptomatic parainfluenza virus carriage, unsuspected infection, or are due to antibody cross-reactivity. We emphasize that all of the 6 immunodeficient patients reported here had antemortem cultures that grew parainfluenza virus type 3. Classic cytopathic effects were also observed at least focally in most of the tissues that immunohistochemically reacted with anti-parainfluenza virus antibody. Thus the diagnosis of disseminated parainfluenza virus infections in these patients appears to be well justified. However, as for all pathological diagnoses, it is important that a diagnosis of parainfluenza virus infection not be made solely on the basis of immunohistochemical results in the absence of viralinduced cytopathic effects or other corroborative evidence of clinically significant infection. In summary, failure of immune reconstitution therapy is a strong risk factor for severe and ultimately fatal parainfluenza virus infections in patients with congenital immunodeficiencies. The pathology of these infections is most severe in the respiratory tract, with giant cell pneumonitis and lack of host inflammatory response. However, the range of tissues infected may also include those in the digestive, urinary, and hematopoietic systems. Patients with congenital immunodeficiencies and parainfluenza 3 infections may have multiple infections, so careful studies must also be done to rule out concurrent infections with other viral, bacteria, and fungal organisms. Acknowledgment. We thank Dr. Sara Miller for electron microscopy services and Carolyn Blount for providing clinical isolates of parainfluenza virus.

REFERENCES 1. Hall CB: Respiratory syncytial virus and parainfluenza virus. New Engl J Med 344:1917-1928, 2001 2. Glenzen WP, Frank AL, Taber LH, et al: Parainfluenza virus type 3: Seasonality and risk of infection and reinfection in young children. J Infect Dis 150:851-857, 1984 3. Counihan ME, Shay DK, Holman RC, et al: Human parainfluenza virus-associated hospitalization among children less than five years of age in the United States. Ped Infect Dis J 20:646-653, 2001 4. Akizuki S, Nasu N, Setoguchi M, et al: Parainfluenza virus pneumonitis in an adult. Arch Pathol Lab Med 115:824-826, 1991 5. Wendt CH, Weisdorf DJ, Jordan MC, et al: Parainfluenza virus respiratory infection after bone marrow transplantation. New Engl J Med 326:921-926, 1992 6. Lewis VA, Champlin R, Enflund J, et al: Respiratory disease due to parainfluenza virus in adult bone marrow transplant recipients. Clin Infect Dis 23:1033-1037, 1996 7. Ljungman P: Respiratory virus infections in bone marrow transplant recipients: The European perspective. Am J Med 102:4447, 1997 8. Matar LD, McAdams HP, Palmer SM, et al: Respiratory infections in lung transplant patients: Radiologic findings with clinical correlation. Radiology 213:735-742, 1999 9. Glenzen WP, Greenberg SB, Atmar RL, et al: Impact of respiratory virus infections on persons with chronic underlying conditions. JAMA 283:499-505, 2000 10. Cortez KJ, Erdman DD, Peret TC, et al: Outbreak of parainfluenza virus 3 infections in a hematopoietic stem cell transplant population. J Infect Dis 184:1093-1097, 2001 11. Nichols WG, Corey L, Gooley T, et al: Parainfluenza virus infections after hematopoietic stem cell transplantation: Risk factors,

602

PARAINFLUENZA IN CONGENITAL IMMUNODEFICIENCY SYNDROMES (Madden, Burchette, and Hale) response to antiviral therapy, and effect on transplant outcome. Blood 98:573-578, 2001 12. Vilchez RA, McCurry K, Dauber J, et al: The epidemiology of parainfluenza virus infection in lung transplant patients. Clin Infect Dis 33:2004-2008, 2001 13. Vilchez R, McCurry K, Dauber J, et al: Influenza and parainfluenza respiratory virus infection requiring admission in adult lung transplant patients. Transplantation 73:1075-1078, 2002 14. Apalsch AM, Green M, Ledesma-Medina J, et al: Parainfluenza and influenza infections in pediatric organ transplant recipients. Clin Infect Dis 20:394-399, 1995 15. Lujan-Zilbermann J, Benaim E, Tong X, et al: Respiratory virus infections in pediatric hematopoietic stem cell transplantation. Clin Infect Dis 33:962-968, 2001 16. Butnor KJ, Sporn TA Jr: Human parainfluenza virus giant cell pneumonia following cord blood transplant associated with pulmonary alveolar proteinosis. Arch Pathol Lab Med 127:235-238, 2003 17. Todd Faulks J, Drinka PJ, Shult P: A serious outbreak of parainfluenza type 3 on a nursing unit. J Am Geriatr Soc 48:12161218, 2000 18. Elizaga J, Olavarria E, Apperly J, et al: Parainfluenza virus 3 infection after stem cell transplant: Relevance to outcome of rapid diagnosis and ribavirin treatment. Clin Infect Dis 32:413-418, 2001 19. Chakrabarti S, Collingham KE, Holder K, et al: Preemptive oral ribavirin therapy of paramyxovirus infections after haematopoietic stem cell transplantation: A pilot study. Bone Marrow Transplant 28:759-763, 2001 20. Buckley RH: Primary immunodeficiency diseases due to defects in lymphocytes. N Engl J Med 343:1313-1324, 2000 21. Buckley RH, Schiff SE, Schiff RI, et al: Hematopoietic stemcell transplantation for the treatment of severe combined immunodeficiency. N Engl J Med 340:508-516, 1999 22. Patel DD, Gooding M, Parrott RE, et al: Thymic function after hematopoietic stem-cell transplantation for the treatment of severe combined immunodeficiency. N Engl J Med 342:1325-1332, 2000 23. Myers LA, Patel DD, Puck JM, et al: Hematopoietic stem cell transplantation for severe combined immunodeficiency in the neonatal period leads to superior thymic output and improved survival. Blood 99:872-878, 2002

24. Taylor CE, Osman HK, Turner AJ, et al: Parainfluenza and respiratory syncytial virus infection in infants undergoing bone marrow transplantation for severe combined immunodeficiency. Commun Dis Public Health 1:202-203, 1998 25. Jarvis WR, Middleton PJ, Gelfand EW: Parainfluenza pneumonia in severe combined immunodeficiency disease. J Pediatr 94: 423-425, 1979 26. Delage G, Brochu P, Pelletier M, et al: Giant cell pneumonia caused by parainfluenza virus. J Pediatr 94:426-429, 1979 27. Crooks BN, Taylor CE, Turner AJ, et al: Respiratory viral infections in primary immune deficiencies: Significance and relevance to clinical outcome in a single BMT unit. Bone Marrow Transplant 26:1097-1102, 2000 28. Pham TTN, Burchette JL Jr, Hale LP: Fatal disseminated adenovirus infections in immunocompromised patients. Am J Clin Pathol 120:575-583, 2003 29. Hale LP, Markert ML: Corticosteroids regulate epithelial cell differentiation and Hassall body formation in the human thymus. J Immunol 172:617-624, 2004 30. Washington K, Gossage DL, Gottfried MR: Pathology of the pancreas in severe combined immunodeficiency and DiGeorge syndrome: Acute graft-versus-host disease and unusual viral infections. HUM PATHOL 25:908-914, 1994 31. Chanock RM, Murphy BR, Collins PL: Parainfluenza viruses, in Knipe DM, Howley PM, Fields BN (eds): Fields Virology (ed 4). Philadelphia, PA, Lippincott Williams & Wilkins, 2001, pp 1341-1379 32. Arisoy ES, Demmler GJ, Thakar S, et al: Meningitis due to parainfluenza virus type 3: Report of two cases and review. Clin Infect Dis 17:995-997, 1993 33. Craver RD, Gohd RS, Sundin DR, et al: Isolation of parainfluenza type 3 from cerebrospinal fluid associated with aseptic meningitis. Am J Clin Pathol 99:705-707, 1993 34. McCarthy VP, Zimmerman AW, Miller CA: Central nervous system manifestations of parainfluenza virus type 3 infections in childhood. Pediatr Neurol 6:197-201, 1990 35. Wilks D, Burns SM: Myopericarditis associated with parainfluenza virus type 3 infection. Eur J Clin Microbiol Infect Dis 17:363365, 1998

603