Reports of health care–associated viral hepatitis transmission have been increasing in the United States. Transmission due to poor infection control practices during myocardial perfusion imaging (MPI) has not previously been reported. The aim of this study was to identify the source of incident hepatitis C virus (HCV) infection in a patient without identified risk factors who had undergone MPI 6 weeks before diagnosis. Practices at the cardiology clinic and nuclear pharmacy were evaluated, and HCV testing was performed in patients with shared potential exposures. Clinical and epidemiologic information was obtained for patients with HCV infection, and molecular testing was performed to assess viral relatedness. Evidence of HCV transmission among patients who had undergone MPI at the cardiology clinic on 2 separate dates was found, involving 2 potential source patients and a total of 5 newly infected patients. Molecular testing identified a high degree of genetic homology among viruses from patients with common procedure dates. The nuclear medicine technologist routinely drew up flush from multidose vials of saline solution using the same needle and syringe that had been used to administer radiopharmaceutical doses. Multipatient use of vials was not observed, but a review of purchasing invoices and interviews with staff members suggested that this had occurred. No evidence of transmission via contamination of radiopharmaceuticals at the nuclear pharmacy was found. In conclusion, transmission of HCV occurred because of unsafe injection practices during MPI. Cardiologists should carefully review their infection control practices and the practices of other staff members involved with these procedures.
In May 2008, the North Carolina Division of Public Health (NCDPH) was contacted by a patient who had been informed that a recent blood donation was positive for hepatitis C virus (HCV) ribonucleic acid (RNA). The patient reported no history of HCV risk behaviors but had undergone several medical procedures during the 6 months before the positive HCV test result, including a myocardial perfusion imaging (MPI) study at an outpatient cardiology clinic (clinic A). In light of increasing reports of HCV transmission in United States health care settings, NCDPH began an investigation to identify the source of the infection and to determine if an ongoing risk for blood-borne pathogen transmission remained. During the initial steps of the investigation, a second patient from clinic A was identified to have been diagnosed with acute HCV infection 7 weeks after undergoing an MPI study. In this report, we describe the subsequent investigation and implications for cardiologists and other health care personnel who assist with MPI studies.
Methods
We interviewed clinic A staff members and observed MPI studies. We searched state surveillance databases for reports of acute HCV infection among patients who had undergone MPI studies at clinic A around the same time as the 2 patients already identified; chronic HCV infections are not reportable in North Carolina. We contacted the other patients who had undergone MPI studies on the same dates as the 2 initial patients and requested consent to test for antibodies against HCV (anti-HCV) as part of the public health investigation. For patients with positive anti-HCV results, we conducted routine viral hepatitis case interviews to collect clinical and risk factor information and performed medical record reviews to assess history of hepatitis testing, illness, or diagnosis.
To evaluate the possibility that HCV infections were associated with the use of contaminated radiopharmaceuticals, we visited the nuclear pharmacy that prepared and supplied radiopharmaceuticals to clinic A, interviewed pharmacy staff members, and observed radiopharmaceutical preparation. We requested consent for anti-HCV testing from selected patients seen at other cardiology clinics who had received radiopharmaceuticals from the same lots as those received by the index patient. To assess the possibility that radiopharmaceuticals could have been contaminated with HCV during the compounding process (i.e., from blood received for radiolabeling of white or red blood cells), we requested consent for anti-HCV testing from all patients whose blood had been processed in the nuclear pharmacy within 1 week before the lots in question were prepared.
Initial anti-HCV testing was performed at multiple commercial and hospital-based laboratories in North Carolina. All positive results were reviewed by investigators at NCDPH. When available, serum specimens from anti-HCV-positive patients were sent to the Centers for Disease Control and Prevention Division of Viral Hepatitis Laboratory. Specimens were tested for anti-HCV with the Ortho VITROS anti-HCV chemiluminiscent immunometric assay (VITROS ECi Immunodiagnostic System; Ortho-Clinical Diagnostics, Inc., Rochester, New York). Positive and indeterminate samples were tested for HCV RNA using the COBAS Amplicor HCV Monitor test version 2.0 (Roche Molecular Systems, Inc., Branchburg, New Jersey), an in vitro nucleic acid amplification test for the quantitative determination of HCV RNA in human serum or plasma.
HCV genotype was determined from a 300-nucleotide segment of the NS5b region. The genetic relatedness among patient isolates was determined by quasispecies analysis of the HCV genome encompassing the E1 and hypervariable region 1 (E1-HVR1) using previously described methods. NS5b consensus sequences and E1-HVR1 quasispecies sequences from patient specimens were compared to each other and to corresponding sequences of the same HCV genotype from randomly selected HCV-infected patients identified through the Third National Health and Nutrition Examination Survey (NHANES), a representative sample of the noninstitutionalized civilian population of the United States. Phylogenetic analysis and pairwise comparison of nucleotide sequences were performed as described elsewhere.
Results
The index patient was a 49-year-old man who had undergone MPI at clinic A in December 2007. The patient was first found to be HCV RNA positive 6 weeks later, during screening of donated blood ( Figure 1 ). Anti-HCV antibodies were not detectable for an additional 4 weeks, indicating recent infection. The patient denied experiencing any symptoms suggestive of acute hepatitis during this period. No behavioral HCV risk factors were identified via patient interview or review of medical records. The patient’s spouse tested anti-HCV negative. During the patient’s likely HCV exposure period (i.e., the 6 months before HCV infection was identified), he reported several additional health care exposures, including a colonoscopy, a dental cleaning, and monthly allergy injections. A review of practices at the facility at which the colonoscopy was performed revealed no apparent opportunities for transmission of blood-borne pathogens. Because of accumulating evidence suggesting clinic A as a potential venue for transmission, most notably the identification of acute HCV infection after MPI in a second clinic A patient, other potential health care exposures were not further evaluated.
The second patient identified with acute HCV infection was an 81-year-old woman who had undergone MPI at clinic A in June 2007. Seven weeks after MPI, she developed fatigue, abdominal pain, nausea, diarrhea, and amber-colored urine. Nine weeks after MPI, the patient’s serum alanine aminotransferase level was elevated (861 IU/L), and she was found to be anti-HCV positive. The patient denied behavioral HCV risk factors and had undergone no other medical procedures during the 6 months before the onset of symptoms. Her spouse tested negative for anti-HCV.
Clinic A is a single-physician cardiology clinic that began offering MPI studies in 2006 and typically performed 3 to 8 studies per day. No other procedures involving percutaneous exposures were performed. One nuclear medicine technologist was responsible for all aspects of MPI studies, including peripheral intravenous catheter placement and medication administration. This technologist began employment at clinic A in June 2007, on the same date the second patient identified with acute HCV infection underwent her procedure. No other technologists had worked at clinic A since that time. Two parenteral medications were routinely used during MPI studies: the radiopharmaceutical (technetium-99m–labeled sestamibi) prepared at the nuclear pharmacy and a sodium chloride 0.9% (normal saline) solution. The clinic did not stock or administer any parenteral narcotics. No written infection control policies were in place at the time of our assessment.
During observations of MPI studies, each patient received 2 doses of the radiopharmaceutical: a rest dose and a stress dose. Each dose arrived from the nuclear pharmacy on the date of the procedure in a-3 mL syringe labeled with the patient’s name and packaged in a lead container. The rest dose was administered into an intravenous catheter using the prefilled 3-mL syringe and needle. The technician was then observed to withdraw 3 mL of flush from a new 30-mL multidose vial of normal saline using the same needle and syringe used to administer the rest dose; this flush was then administered to the patient. The process was repeated using the same needle, syringe, and saline vial to administer a second 3-mL saline flush (6 mL total flush volume per rest dose). The saline vial and syringe were then observed to be discarded. The technician reported that 2 saline flushes were routinely administered to be sure that all residual radiopharmaceutical would be injected into the patient. However, this practice was neither recommended nor endorsed by the nuclear pharmacy. For the stress dose (usually administered 30 minutes to 2 hours after the rest dose), the technologist was observed to administer the radiopharmaceutical using a second prefilled 3-mL syringe provided by the pharmacy, followed this time by a 10-mL saline flush drawn from a new 30-mL vial with a new needle and syringe, both of which were then observed to be discarded.
The technologist stated that a new 30-mL saline vial was used to supply flush for each dose of radiopharmaceutical and then discarded (i.e., 2 vials discarded per patient with 24 and 20 mL of residual saline discarded for rest and stress doses, respectively). However, a review of saline vial purchasing invoices indicated that clinic A was ordering <1/2 the number of vials needed to perform the procedures in this manner. When asked about this discrepancy, the technologist stated that saline vials had not always been discarded after flushing the rest dose but had sometimes been used to also flush the stress dose for the same patient. When this was done, the technologist reported that each vial was dedicated to a single patient, labeled with that patient’s initials, and discarded after flushing the stress dose. The technologist reported that care was taken to ensure the same vial was not used for multiple patients but acknowledged that there was a possibility for errors to occur, because multiple patients were undergoing stress tests at the same time.
Observations at the nuclear pharmacy did not reveal any infection control breaches that would have facilitated the transmission of blood-borne pathogens. No blood specimens were being processed during the time of our observations, but our review of procedures revealed no apparent opportunities for blood contamination of other pharmaceutical products. Specifically, blood received in the pharmacy for radiolabeling was processed in a designated blood processing biosafety cabinet using dedicated equipment with colored labels to prevent inadvertent use for other purposes. Used equipment was placed in plastic bags before removal from the blood processing cabinet and discarded in a medical waste bin. We identified 2 patients who had blood processed in the nuclear pharmacy during the week before the lots under investigation were prepared, 1 for each procedure date. Both patients were tested and found to be anti-HCV negative.
All patients who underwent MPI studies at clinic A on the 2 dates of interest were tested for anti-HCV, and 5 additional patients with evidence of HCV infection were identified. Including the 2 initial cases, anti-HCV was detected in 4 of 4 patients from the June 2007 date (cluster X) and 3 of 6 patients from the December 2007 date (cluster Y) ( Table 1 ). No patient from either cluster had been diagnosed with hepatitis C before undergoing the MPI study. A review of state surveillance databases identified no additional reports of acute HCV infection among patients who had undergone MPI studies at clinic A within 1 day before or after these 2 dates.
| Cluster X (June, 2007) | Cluster Y (December 2007) | ||||
|---|---|---|---|---|---|
| Patient ID | Dose Type | Time | Patient ID | Dose Type | Time |
| X1 ⁎ | Rest | 8:30 am | Y1 ⁎ | Rest | 9:05 am |
| X2 ⁎ | Rest | 9:50 am | Y2 ⁎ † | Rest | 9:20 am |
| X3 ⁎ | Rest | 10:10 am | Y3 ⁎ | Rest | 9:35 am |
| X4 ⁎ | Rest | 10:15 am | Y1 ⁎ | Stress | 10:10 am |
| X2 ⁎ | Stress | 10:25 am | Y2 ⁎ † | Stress | 10:30 am |
| X1 ⁎ | Stress | 10:35 am | Y4 | Rest | 10:40 am |
| X3 ⁎ | Stress | 11:00 am | Y3 ⁎ | Stress | 11:10 am |
| X4 ⁎ | Stress | 11:40 am | Y5 | Rest | 11:25 am |
| Y4 | Stress | 12:40 pm | |||
| Y6 | Rest | 12:55 pm | |||
| Y5 | Stress | 1:15 pm | |||
| Y6 | Stress | 1:50 pm | |||
Among the 4 HCV-positive patients in cluster X ( Table 1 ), 3 had evidence of recent infection; 1 (X3) had onset of dark urine, and 3 (X2, X3, and X4) had elevated serum aminotransferase levels within 6 months after the procedure. The remaining patient (X1) reported nonspecific symptoms during this time frame (i.e., fatigue, intermittent nausea, and abdominal pain); serum aminotransferase levels were normal 10 months after the procedure. Among the 3 HCV-positive patients in cluster Y, 2 had evidence of recent infection; 1 (Y1) had onset of dark urine, nausea, and fatigue but did not seek medical care, and 1 (Y2, the index patient) had elevated serum aminotransferase levels but no clinical manifestations of acute hepatitis within 6 months after the procedure date. The remaining patient (Y3) had no clinical manifestations of hepatitis; results of post-MPI serum aminotransferase testing were not available. Jaundice was not noted among patients in either cluster. No traditional risk factors for HCV infection were identified among patients in either cluster through interviews or medical record reviews; all patients tested negative for hepatitis B virus (HBV) and human immunodeficiency virus.
Among the HCV-infected patients in cluster Y, all had received radiopharmaceutical from the same 2 lots: lot A (rest doses) and lot B (stress doses). We identified 7 additional patients who had undergone procedures at 3 other clinics and had received radiopharmaceutical from these same lots. All 7 were tested and found to be anti-HCV negative.
Specimens were obtained from all 4 patients in cluster X and in 2 of 3 patients in cluster Y. All 4 specimens from cluster X were positive for HCV RNA of subgenotype 1a. Both specimens from cluster Y were positive for HCV RNA of subgenotype 1b. There was a high degree of sequence homology among HCV strains from patients in cluster X, with 97.3% to 99% nucleotide sequence identity in NS5b ( Figure 2 ). HCV strains from specimens in cluster Y also demonstrated a high degree of homology, with 99.7% nucleotide sequence identity in NS5b ( Figure 2 ) and 99.3% maximum nucleotide sequence identity in E1-HVR1 ( Figure 3 ).
