SEARCH

SEARCH BY CITATION

Keywords:

  • Donor-to-host transmission;
  • HIV;
  • HCV;
  • nucleic acid diagnostics

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Disclosure
  8. References
  9. Appendices

In 2007, a previously uninfected kidney transplant recipient tested positive for human immunodeficiency virus type 1 (HIV) and hepatitis C virus (HCV) infection. Clinical information of the organ donor and the recipients was collected by medical record review. Sera from recipients and donor were tested for serologic and nucleic acid-based markers of HIV and HCV infection, and isolates were compared for genetic relatedness. Routine donor serologic screening for HIV and HCV infection was negative; the donor's only known risk factor for HIV was having sex with another man. Four organs (two kidneys, liver and heart) were transplanted to four recipients. Nucleic acid testing (NAT) of donor sera and posttransplant sera from all recipients were positive for HIV and HCV. HIV nucleotide sequences were indistinguishable between the donor and four recipients, and HCV subgenomic sequences clustered closely together. Two patients subsequently died and the transplanted organs failed in the other two patients. This is the first recognized cotransmission of HIV and HCV from an organ donor to transplant recipients. Routine posttransplant HIV and HCV serological testing and NAT of recipients of organs from donors with suspected risk factors should be considered as routine practice.


Abbreviations: 
ALT

alanine aminotransferase

AST

aspartate aminotransferase

CDC

centers for disease control

EIA

enzyme immunoassay

HAART

highly active antiretroviral therapy

HIV

human immunodeficiency virus

HCV

hepatitis C virus

MMF

mycophenolate mofetil

MSM

men who have sex with other men

NAT

nucleic acid test

OPO

organ procurement organization

OPTN

organ procurement & transplant network

RT-PCR

reverse transcriptase polymerase chain reaction

Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Disclosure
  8. References
  9. Appendices

Although organ transplantation has become more successful over time (1,2), infection transmission from donor to recipient remains a risk of organ transplantation (3). Organ donor screening has reduced the risk for human immunodeficiency virus type 1 (HIV) and hepatitis C virus (HCV); however transmission still occurs (3–9). Screening organ donors for HIV became required in 1985 and for HCV in 1992 (4,6–9).

In October 2007, a deceased donor renal transplant recipient who was sero-negative for HIV and HCV pretransplant, tested positive for both viruses 10 months after transplantation. Since the recipient had no identified risk factors for either infection and infection was recognized shortly after transplantation, the case was reported as a potential donor-derived infection as required by current Organ Procurement and Transplant Network (OPTN) policy (10). The state and local health departments were informed, and an investigation was initiated with assistance from the Centers for Disease Control and Prevention (CDC).

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Disclosure
  8. References
  9. Appendices

Epidemiologic investigation

Medical records of the organ donor and transplant recipients were reviewed to characterize clinical history, including healthcare risk factors for HIV and HCV (e.g. transfusion receipt) and diagnostic test results. For the donor, additional clinical information and social history was collected through interview of his family and known social contacts to elicit behavioral risk factors for HIV and HCV.

Laboratory investigations

Serum or plasma samples of the donor and recipients were tested by the CDC for HIV antibodies using a third generation ELISA immunoassay (Abbott Recombinant HIV 1 + 2, Abbott Laboratories, Abbott Park, IL, USA). Viral RNA was extracted from available specimens and sequences from three HIV viral regions (pol, env and gag) were amplified by reverse transcription polymerase chain reaction (RT-PCR) as previously described (11–13). PCR amplicons were sequenced and phylogenetically analyzed (Figure 1). HIV-1 pol sequences were examined for drug-resistance mutations (see Appendix A) (14).

image

Figure 1. Phylogenetic clustering of donor and recipient human immunodeficiency viruses. Neighbor-joining (NJ) tree of concatenated gag (p17) and envelope (gp41) sequences (743-bp) using the general time reversible model of nucleotide substitution, proportion of invariable sites (0.348) and gamma distributed rates (0.547). Similar tree topologies were obtained with these parameters using maximum likelihood (ML) analysis (PhyML) and Bayesian inference (BEAST). Support for the clustering of the organ donor and recipient HIV-1 sequences was determined using 1000 bootstrap replicates in the NJ and ML analyses and posterior probabilities inferred using 5 million Markov Chain Monte Carlo generations and a relaxed, uncorrelated lognormal clock model and either the Yule or constant tree priors in the BEAST analysis. Highly significant NJ/ML/Bayesian bootstrap support and posterior probabilities are shown at selected nodes. Cluster defined by the sequences from the donor and four recipients is circled. Numbered sequences are random US HIV-1 subtype B-infected persons for comparison; HXB2, sequence from the HIV-1 HXB2 isolate.

Download figure to PowerPoint

Sera were tested for anti-HCV IgG using the ORTHO HCV version 3.0 ELISA (Ortho-Clinical Diagnostics Inc., Raritan, NJ, USA). Reactive results were confirmed with supplemental testing using HCV Matrix (Abbott Laboratories) or RIBA 3.0 (Chiron Corporation, Emeryville, CA, USA). Viral RNA was extracted from serum and a segment of HCV NS5b region, encompassing positions 8275–8616, was amplified using real-time nested PCR (15,16). Endpoint limiting-dilution PCR for HVR1 region was done, and its nucleotide sequence determined from single genomes of HCV as described previously (Appendix B) (16). HVR-1-based phylogenetic trees of HCV quasispecies (see Figure 2) were constructed (PHYLIP package, v.3.6, J. Felsenstein, University of Washington, Seattle, WA, USA) (Appendix C).

image

Figure 2. Phylogenetic dendrogram of recipient hepatitis C viruses. E1-HVR1 Region (291 bp in length) unique clonal sequences are compared to show relatedness. Maximum nucleotide identity between case-patients is 100%. The neighbor-joining algorithm based on distance matrices generated using the Kimura two-parameter model of nucleotide substitution was used.

Download figure to PowerPoint

Blood product traceback

Units of blood products transfused into the organ donor were identified and associated blood donors were contacted for serologic and nucleic acid testing (NAT).

Case definition

We defined a case as HIV and HCV coinfection in a recipient from the same donor in January 2007 who was uninfected by either HIV or HCV before transplantation.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Disclosure
  8. References
  9. Appendices

Donor

The organ donor was a Mexican-American male in his late 30s who sustained a blunt head trauma due to a motor vehicle accident in January 2007. He was subsequently declared brain dead, and consent was obtained for organ donation. Family members were unable to provide the donor's medical or behavioral history; in the course of his evaluation, a social contact eventually revealed a history of having sex with another man, and he was designated a ‘high risk’ donor due to this recognized risk factor. It was unknown whether the donor had ever been previously tested for HIV or HCV.

The donor received a total of 20 units of blood products during his hospitalization. Two serum specimens, pre- and posttransfusion, were tested, per protocol, locally by the Organ Procurement Organization (OPO) and were negative for HIV-1/HIV-2 antibody [Abbott HIVAB™ HIV-1/HIV2 [rDNA] enzyme immunoassay (EIA)] and negative for HCV (Abbott HCV EIA 2.0); results of serologic evaluation for cytomegalovirus, Epstein-Barr virus, hepatitis B and syphilis were all negative. The posttransfusion specimens were significantly hemodiluted (Table 1). A pretransfusion specimen was not available for repeat testing as part of the later investigation.

Table 1.  Detection of HIV and HCV through laboratory testing in organ donor samples
 
SpecimenHemodilution1Anti-HIV antibodyHIV Viral LoadAnti-HCV IgGHCV Viral Load
  1. 1Hemodilution was calculated using methods described by the FDA. (http://www.fda.gov/downloads/BiologicsBloodVaccines/GuidanceComplianceRegulatoryInformation/Guidances/Tissue/UCM188251.pdf; accessed 5/23/2010). The donor's estimated blood volume is 5333 mL and his estimated plasma volume is 3200 mL. He received 2 L of normal saline before collection of the pretransfusion specimen and 6818 mL of fluids and blood products before collection of the posttransfusion specimen.

  2. 2Qualitative testing was done initially and there was insufficient residual volume to perform quantitative testing.

  3. CDC = Centers for Disease Control and Prevention; OPO = organ procurement organization; ND = not done.

Pretransfusion (Routine screening results)NoneNegativeNDNegativeND
Pretransfusion Unavailable for testingUnavailable for testingUnavailable for testingUnavailable for testing
Investigation results
PosttransfusionYesNegativePositive2NegativePositive
Investigation results    (898 IU/mL)

Blood transfusion traceback testing

All 20 individuals who donated the 20 units of blood products received by the donor were identified and tested for HIV and HCV by EIA and RT-PCR and found negative.

Left kidney recipient

The index case-patient was a male in his early 30s with a history of childhood-onset nephrotic syndrome of unknown etiology and failed deceased donor renal transplantation in 1997. He was HIV- and HCV-negative by serology in 2003 as part of his evaluation for his second transplant. He received his second kidney transplant and was given methlyprednisolone, alemtuzumab (30 mg) and a single dose of rituximab (375 mg/m2) for induction, and maintained on tacrolimus, mycophenolate mofetil (MMF) and prednisone. He denied any risk factors for acquiring HIV or HCV other than hemodialysis. Three months following the transplant, routine laboratory testing revealed elevated liver enzymes (aspartate aminotransferase [AST], 414 IU/mL; alanine aminotransferase [ALT], 660 IU/mL and alkaline phosphatase, 202 IU/mL). Serologic test results for infection by hepatitis A, B and C viruses were negative; HCV RNA testing was not performed. Subsequently, a liver biopsy indicated inflammation and stage II/III fibrosis, and HCV RNA (>10 000 000 IU/mL) was detected in serum. Ten months after the transplant, a kidney biopsy showed Banff 1A acute cellular rejection and a proliferative glomerulonephritis, possibly consistent with HIV infection. Serologic testing revealed positivity for HIV 1/2 antibody by EIA and Western blot. HIV viral load was 520 copies/mL and baseline CD4+ cell count was 16/μL. He was subsequently placed on highly active antiretroviral therapy (HAART) consisting of efavirenz, emtricitabine and tenofovir, resulting in a virologic and immunologic response. Given the diagnosis of newly acquired HIV and HCV infection, immunosuppressive therapy, including tacrolimus and MMF, was decreased. The patient's renal function progressively declined, resulting in reinstitution of hemodialysis, transplant nephrectomy and discontinuation of immunosuppressive medications 14 months posttransplant. Anti-HCV IgG was detected shortly after this, and he was started on anti-HCV treatment with peginterferon α-2a and ribavirin with complete and sustained virologic response. He has subsequently been listed for a combined kidney/liver transplant. His spouse tested negative for both HIV and HCV.

Right-kidney recipient

The right-kidney recipient was a young-adult female with a history of renal disease caused by glomerulonephritis. She was transplanted and received daclizumab for induction and was maintained on tacrolimus, mycofenolate mofetil and prednisone. Her pretransplant HIV and HCV serologies had negative results at the time of transplant. Serum drawn 10 months posttransplant tested positive for HIV antibodies by EIA and Western blot. Her HIV viral load was 34 100 copies/mL and CD4+ cell count was 556/μL. Anti-HCV IgG was negative, but HCV RNA was detected at greater than 5 million IU/mL. AST was 43 IU/mL and ALT 50 IU/mL. She was admitted in January 2008 for complications related to organ rejection and was given zidovudine, lamivudine and efavirenz. Continued rejection led to progressive azotemia requiring reinstitution of hemodialysis and transplant nephrectomy at 19 months posttransplant, at which time immunosuppression was discontinued and she was relisted for kidney transplant. She failed to achieve an HCV virologic response (<6 million IU copies/mL) with 16 weeks of treatment with peginterferon α-2a and ribavirin. As of January 2009 the patient remained clinically stable, with normal liver enzyme activities (AST 16 IU/mL and ALT 20 IU/mL). The patient responded virologically and immunologically to HAART; her regimen was changed to lamivudine, tenofovir and efavirenz because of a detectable viral load; as of May 2010, her CD4 count is 880 cells/μL with an HIV viral load <50 copies/mL. She had no contacts who warranted HIV testing.

Liver recipient

The liver recipient was a female in her late 60s with a history of cryptogenic cirrhosis, who underwent liver transplantation; some details of this donor have been previously published (17). She received four doses of methylprednisolone followed by daily tacrolimus and prednisone. Approximately 6 weeks after transplantation, marked allograft dysfunction developed and liver biopsies demonstrated changes consistent with acute cellular rejection with mild lobular hepatitis. Corticosteroid treatment of rejection led to some improvement in liver enzyme elevation, but persisted with hyperbilirubinemia. One month prior to transplantation, HIV and HCV serologic test results were negative and HCV RNA had been undetectable. Eleven months after transplantation, she was serologically positive for HIV, with an HIV viral load was >500 000 copies/mL and CD4 count of 34 cells/mL. Her HCV serologic testing was negative but her HCV viral load was 5 770 000 IU/mL. Antiretroviral therapy consisting of emtricitabine, efavirenz and tenofovir was initiated with virologic response (HIV RNA <400 copies/mL). Anti-HCV therapy was not given. Multiorgan failure developed 389 days after transplantation secondary to biliary sepsis, and the patient expired. She had no contacts who warranted HIV/HCV testing.

Heart recipient

The heart recipient was a male in his early 60s with a history of ischemic cardiomyopathy. He underwent a heart transplant and received basiliximab and thymoglobulin for induction and was maintained on tacrolimus, mycophenolate mofetil and prednisone. HIV and HCV serologic testing, on serum drawn 1 month pretransplant, were negative. In November 2007 he was found to be positive for HIV 1/2 antibody by EIA and Western blot, with an HIV viral load >100 000 copies/mL and a CD4+ count of 48 cells/μL. Anti-HCV IgG was negative, but the HCV viral load was >4 million IU/mL. Serum AST was 43 IU/mL and ALT was 28 IU/mL. He was subsequently placed on therapy with efavirenz, emtricitabine and tenofovir and achieved rapid and sustained virologic and immunologic responses. He was admitted in December 2007 for biopsy-proven moderate acute cellular rejection, which was treated with a course of prednisone. As of August 2009 the HIV viral load remained undetectable, the CD4+ cell count was 50/μL, and HCV viral load remained >4 million copies/mL (AST of 117, ALT of 47). In August 2009, the patient became noncompliant with his medical care and refused subsequent follow-up. He expired of multiorgan failure in November 2009. His spouse tested negative for HIV and HCV.

Laboratory results

HIV testing:  Post-transfusion serum obtained from the organ donor tested at CDC was negative for antibodies to HIV; however, HIV RNA was positive by qualitative PCR. There was insufficient residual sample to perform quantitative PCR. All recipients were confirmed positive for HIV antibodies in their posttransplant samples (Table 2). HIV pol sequences from the organ donor and the liver recipient were not amplifiable; pol sequences from the other three recipients failed to show drug-resistance mutations. The gag and env sequences from all five patients were identical except for a single nucleotide difference found in the liver recipient. Phylogenetic analysis showed close clustering of the HIV sequences from all five patients (Figure 1).

Table 2.  Investigation results of HIV and HCV laboratory testing in four organ recipients
OrganPretransplantPosttransplant values at initial assessment—November 2007
DateAnti-HIV IgGAnti-HCV IgGDateAnti-HIV IgGHIV Viral Load (copies/mL)Anti-HCV IgGHCV RNAHCV Viral Load (IU/mL)
  1. 1Approximately 3 years pretransplant.

  2. NR = nonreactive.

Left kidney20031NRNR10/07Reactive   520NRReactive23 million
Right kidney1/07NRNR11/07Reactive 35,000NRReactive4–5 million
Liver12/06NRNR11/07Reactive500,000NRReactive5 million
Heart12/06NRNR11/07Reactive1 millionNRReactive4–5 million

HCV testing; Post-transfusion serum obtained from the organ donor tested at CDC was negative for anti-HCV IgG, but the HCV viral load was 898 IU/mL. The four recipients were negative by HCV serologic testing in all posttransplantation samples but were positive by HCV RT-PCR. HCV RNA from the donor's serum could not be amplified for sequencing analysis, but HCV in sera of all the recipients was genotype 1a. Quasispecies analysis of HCV in recipient samples showed close clustering (Figure 2).

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Disclosure
  8. References
  9. Appendices

We describe cotransmission of HIV and HCV from an organ donor to four transplant recipients. Genetic sequences were highly related between HIV and HCV strains infecting all four recipients, confirming transmission from a common donor. The donor was ‘high risk’ by the OPTN definition, due to history of having sex with another man, and was screened according to current OPTN policy (18). However, serologic screening for HIV and HCV yielded negative results, presumably due to sampling during the ‘window period’ which follows acute HIV and HCV infection, before seroconversion. However, other possibilities for a falsely negative serology result include hemodilution and assay interference from substances in the specimen (5). An aliquot of the donor's pretransfusion blood was available, but this sample could not be accessed for HIV and HCV retesting due to ongoing litigation, thereby limiting the ability to understand the reason for the false-negative results. Transmission through transfusion of the organ donor was excluded through lookback tracing and testing.

Previous reports of viral pathogens transmitted through organ transplantation underscore the potential for unexpected adverse outcomes of transplantation (19–21). Residual risk of disease transmission should be understood since, unlike many tissues used for transplantation, solid organs cannot be processed, disinfected or otherwise modified to inactivate infectious pathogens, and there is no absolute donor exclusion policy based on behavioral risk history. Further, the timeline for donor evaluation and screening must be done within a few hours after declaration of brain death, and the quality of donor medical and social history assessment is often limited. If only the parents had been interviewed, donor risk would not have been identified.

Laboratory assessment of donors includes serological screening for markers of infection (18). Although NATs for HIV and HCV shortens the interval between infection and detection, their use is currently limited by lack of availability, prolonged turnaround times and false-positive results which may delay or prevent transplantation and is not required by current OPTN policy (22).

The risk of disease transmission through organ transplantation must be balanced against the risk of potential recipients dying while waiting for an organ (23). However, due to organ shortage and because organ transplantation often is lifesaving, organs from donors with known elevated risk including men who have sex with other men (MSM) history often are transplanted whenever the benefit of transplantation is considered to outweigh the risk of potential disease transmission (24–28). Risk-benefit may not be the same, however, for all organ types, or for all candidates on the waiting list for a given organ type, emphasizing the need for careful candidate informed consent.

Current OPTN policy requires that when an organ donor is classified as ‘increased risk’, the recipients must be informed of, and give consent to, the risk in order to be transplanted with organs from that donor; such a policy was not in place at the time of this transmission event (10). Guidelines recommend that recipients of organs from increased-risk donors be tested at 1, 3 and 12 months posttransplantation (5,22). However, many centers have not implemented such screening practices or rely on serologic testing alone which may yield false negative results; as such, routine testing of recipients with assays that directly detect infection, such as NAT, should be considered. None of the recipients involved in this case were screened posttransplantation (29). Improved screening using NAT for recipients of organs from high-risk donors should allow earlier therapeutic intervention and may prevent further disease transmission through quarantine of residual tissue from the same donor, and could have led to recognition of this transmission event earlier (30). This investigation also highlights the importance of NAT for diagnosis of donor-derived HIV and HCV in the setting of immunosuppression. In this disease cluster, all recipients were seronegative but NAT-positive for HCV, and one had an indeterminate HIV western blot; similar results have been found in other unpublished HCV transmissions.

This transmission event also suggests that screening donors with NAT or combined antibody–antigen assays may identify infectious donors missed by serologic testing alone. Existing data suggest that 0.9% of seronegative organ donors have detectable HCV RNA, in part due to the long window period for HCV (31). Further, organs from HCV-infected donors can be safely used in HCV-infected recipients, minimizing the impact of false-positive results on organ availability (32). In contrast, organs from HIV-infected donors cannot be recovered by OPTN policy, and therefore false-positive results of NAT may result in a net loss of usable organs (22). It also is possible, however, that use of NAT may improve organ acceptance from high-risk donors with negative testing results, expanding the donor pool and balancing or exceeding organ discard from false-positive results.

The lack of recognition and real-time communication of recipient status have been common features of many of the clinically significant donor-derived disease transmissions (3,19,21,33). New diagnosis of HCV in our index patient was not considered to be of donor origin until recognition 7 months after the patient's transplantation. Further, although at least two recipients had liver abnormalities that were noted, it was unknown that similar problems existed in other recipients until the investigation began several months later. Testing of other recipients was only conducted after the index case was reported to the OPTN and the public health system; such testing led to recognition of infection of all recipients. Although the OPTN/UNOS Patient Safety System facilitates communication and categorization of potential disease transmissions, through review by the Disease Transmission Advisory Committee, it is currently limited to organ transplantation in an oversight role and does not include public health investigation. Since numerous tissues may be procured from the same individual who donates organs, broader systems, similar to the recently piloted Transplantation Transmission Sentinel Network (TTSN) (34), are needed to facilitate communication and investigation of potential disease transmissions for events involving other tissues, and for those events of public health significance.

In summary, both HIV and HCV were transmitted from an organ donor to four transplant recipients despite routine negative serologic screening. Recognition of recipient infection was delayed by lack of awareness of potential donor-origin of the infection and by reliance on serology instead of including NAT for recipient testing. Standardization of donor risk assessment, use of NAT in donor screening, routine posttransplant testing of recipients for HIV and HCV by serology and NAT of recipients from high-risk donors may help to further improve the outcomes of organ transplantation.

Disclosure

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Disclosure
  8. References
  9. Appendices

No commercial organization prepared or funded this document. The authors of this manuscript have no conflicts of interest to disclose as described by the American Journal of Transplantation, except for MG Ison who discloses research funding, paid to Northwestern University, by ViraCor and paid consultation by Abbott Molecular, Biogen Idec, and ViraCor. The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Disclosure
  8. References
  9. Appendices

Appendices

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Disclosure
  8. References
  9. Appendices

Appendix A

Unequal nucleotide composition and measurement of phylogenetic signal were determined with the likelihood mapping implemented in the TREE-PUZZLE program (35). Sequences were aligned using Clustal W (36), edited manually and all indels were removed. The best fitting evolutionary model for the aligned sequences was determined using a hierarchical likelihood ratio test in the program Modeltest v3.7 (37). A variant of the GTR model, allowing six different substitution rate categories (rA[LEFT RIGHT ARROW]C= 2.50, rA[LEFT RIGHT ARROW]T= 0.87, rG[LEFT RIGHT ARROW]T= 1, rA[LEFT RIGHT ARROW]G= 6.27, rC[LEFT RIGHT ARROW]G= 0.96, rC[LEFT RIGHT ARROW]T= 6.27), with gamma-distributed rate heterogeneity (α= 0.6041) and an estimated proportion of invariable sites (0.3652), was determined to best fit the data. Phylogenetic relatedness of concatenated donor and recipient HIV env (360-bp) and gag (p17; 396-bp) sequences was inferred using Bayesian inference implemented in the BEAST software program (38). These two HIV regions are commonly used in phylogenetic analyses and comparisons of HIV strains in transmission cases (39,40). The Bayesian calculation consisted of two independent 10 000 000 Markov Chain Monte Carlo (MCMC) generations with sampling every 1000th generation. Convergence of the MCMC was assessed by calculating the effective sampling size (ESS) of the runs using the program Tracer available at http://beast.bio.ed.ac.uk/Tracer. All parameter estimates showed significant ESSs (>150). The tree with the maximum product of the posterior clade probabilities (maximum clade credibility tree) was chosen from the posterior distribution of 5000 sampled trees (after burning in the first 5001 sampled trees) with the program TreeAnnotator version 1.4.6 included in the BEAST software package (38). Both the constant coalescent and Yule process were used as tree priors with a relaxed molecular clock model. Trees were viewed and edited using FigTree v1.1.2 available at http://tree.bio.ed.ac.uk/software/figtree. Resistance was determined using the Stanford genotypic resistance interpretation algorithm available at http://hivdb.stanford.edu/pages/algs/HIVdb.html.

Appendix B

At the limiting dilution, the DNA target templates are presumed to be distributed in a Poisson manner, so that 50% of reactions do not carry template molecules, and thus generate no PCR product. Under such conditions, the positive reactions are most likely to have been initiated from a single-template molecule. Multiple, nested real-time PCR amplifications were performed at the endpoint limiting-dilution determined for each specimen, following which about 50–60 positive amplicons were identified using melting peak analysis. All the HVR1 PCR positive products were sequenced. Sequencing reactions were performed using the BigDye v3.1 chemistry sequencing kit (Applied Biosystems, Foster City, CA, USA), and products were sequenced using an automated sequencer (3130xl Genetic Analyzer, Applied Biosystems). Preliminary sequence analysis was conducted using SeqMan and MegAlign programs from the Lasergene DNA & Protein analysis software (version 7.0, DNASTAR Inc., Madison, WI, USA). The Accelrys GCG Package (Genetic Computer Group, version 11.1-UNIX, Accelrys Inc., San Diego, CA, USA) was used for further analysis. Nucleotide sequences were aligned using the GCG multiple alignment program Pileup.

Appendix C

Frequency distributions of pair-wise distances between nucleotide sequences were estimated using the evolution program in the Accelrys GCG Package. Shannon entropy analysis was applied to measure the extent of quasispecies complexity at suboptimal PCR conditions. This entropy measure was calculated as S =–Σ(pi ln pi), which takes into account the frequency of each sequence (pi) in a given set of quasispecies. The entropy value was then normalized as Sn= S/lnN, to take into consideration the total number of quasispecies sequences (N) analyzed during each PCR. SAS for Windows (Version 9.12, SAS Institute Inc., Cary, NC, USA) was used for the comparison of Shannon entropy and genetic distances.