1. Epidemiology and Global Health
  2. Evolutionary Biology

Estimating the potential to prevent locally acquired HIV infections in a UNAIDS Fast-Track City, Amsterdam

  1. Alexandra Blenkinsop  Is a corresponding author
  2. Mélodie Monod
  3. Ard van Sighem
  4. Nikos Pantazis
  5. Daniela Bezemer
  6. Eline Op de Coul
  7. Thijs van de Laar
  8. Christophe Fraser
  9. Maria Prins
  10. Peter Reiss
  11. Godelieve J de Bree
  12. Oliver Ratmann  Is a corresponding author
  13. On behalf of HIV Transmission Elimination AMsterdam (H-TEAM) collaboration
  1. Department of Mathematics, Imperial College London, United Kingdom
  2. Amsterdam Institute for Global Health and Development, Netherlands
  3. Stichting HIV Monitoring, Netherlands
  4. Department of Hygiene, Epidemiology and Medical Statistics, University of Athens, Greece
  5. Center for Infectious Diseases Prevention and Control, National Institute for Public Health and the Environment (RIVM), Netherlands
  6. Department of Donor Medicine Research, Sanquin, Netherlands
  7. Department of Medical Microbiology, Onze Lieve Vrouwe Gasthuis, Netherlands
  8. Big Data Institute, Nuffield Department of Medicine, University of Oxford, United Kingdom
  9. Academic Medical Center, Netherlands
  10. Department of Global Health, Amsterdam University Medical Centers, Netherlands
  11. Division of Infectious Diseases, Department of Internal Medicine, Amsterdam Infection and Immunity Institute, Netherlands
https://doi.org/10.7554/eLife.76487
Share this article
Cite this article
  1. Alexandra Blenkinsop
  2. Mélodie Monod
  3. Ard van Sighem
  4. Nikos Pantazis
  5. Daniela Bezemer
  6. Eline Op de Coul
  7. Thijs van de Laar
  8. Christophe Fraser
  9. Maria Prins
  10. Peter Reiss
  11. Godelieve J de Bree
  12. Oliver Ratmann
  13. On behalf of HIV Transmission Elimination AMsterdam (H-TEAM) collaboration
(2022)
Estimating the potential to prevent locally acquired HIV infections in a UNAIDS Fast-Track City, Amsterdam
eLife 11:e76487.
https://doi.org/10.7554/eLife.76487
  2. Download BibTeX
  3. Download .RIS
49 figures, 14 tables and 1 additional file

Figures

Figure 1
Download asset Open asset
Approach to analysis.

Input data includes patient baseline data at registration, clinical biomarker data and viral sequence data. Biomarker data is used to estimate infection times, the proportion of undiagnosed infections, and thus the total population size of people living with HIV. HIV sequence data is used to reconstruct phylogenetic trees. Groups of Amsterdam residents with distinct virus are determined phylogeographically with phyloscanner, and without considering genetic distances or bootstrap support. Each such group of Amsterdam residents with distinct virus is interpreted as the partially observed part of a distinct transmission chain among Amsterdam residents, and analysed in calendar time based on the infection times estimated from individual biomarker data, as well as clinical data on viral suppression. The partial observations are used to infer the number, size and growth of the actual transmission chains among Amsterdam residents, and derive key epidemic quantities of interest.

Figure 2
Download asset Open asset
HIV infections in Amsterdam residents in 2014–2018 that remained undiagnosed by 1 May 2019.

Posterior median estimates are shown as bars and 95% credible intervals as error bars. Estimates generated from time-to-diagnosis estimates for 535 MSM and 97 heterosexuals.

Figure 3
Download asset Open asset
Phylogenetically observed parts of Amsterdam transmission chains.

(A) All chains. Horizontal lines connect individuals in reconstructed transmission chains in Amsterdam by chains which had no new case since 2014, and those which continued to grow or emerged, among MSM (top) and heterosexuals (bottom), in order of last diagnosis per chain. (B) Subset of chains in which at least one individual was estimated to have been infected since 2014. Data are presented as in subfigure A.

Figure 4
Download asset Open asset
Estimated number of locally preventable infections in 2014–2018 along with 95% credible intervals, for MSM and heterosexuals stratified by region of birth.

Posterior median estimates of proportion (%) of preventable infections shown above bars. Estimates generated from 203 phylogenetic subgraphs among Amsterdam MSM, containing 297 individuals, and 41 subgraphs among Amsterdam heterosexuals, containing 44 individuals.

Appendix 1—figure 1
Download asset Open asset
Distribution of individual level posterior median estimated times to diagnosis by place of birth, for Amsterdam MSM and heterosexuals.
Appendix 1—figure 2
Download asset Open asset
Diagnosis date and posterior median estimated infection date (with 95% credible interval) of individuals in Amsterdam diagnosed between January 2014 and May 2019.
Appendix 1—figure 3
Download asset Open asset
Posterior median estimated time to diagnosis (with 95% credible interval) of HIV infections in Amsterdam occurring in 2014-2018, stratified by risk group (MSM and heterosexuals) and place of birth.

Estimates generated from time-to-diagnosis estimates for 535 MSM and 97 heterosexuals.

Appendix 1—figure 4
Download asset Open asset
Annotated phylogeny of viral sequences of subtype A1 of Amsterdam MSM and background individuals.

Colours of tips show the observed states of each observed sequence, and colours of lineages represent inferred states. States were assigned to each sequence as described in Equation S22, and represent both transmission group (MSM, non-MSM) and place of birth or residence.

Appendix 1—figure 5
Download asset Open asset
Annotated phylogeny of viral sequences of circulating recombinant form 02AG of Amsterdam MSM and background individuals.

Colours of tips show the observed states of each observed sequence, and colours of lineages represent inferred states. States were assigned to each sequence as described in Equation S22, and represent both transmission group (MSM, non-MSM) and place of birth or residence.

Appendix 1—figure 6
Download asset Open asset
Annotated phylogeny of viral sequences of circulating recombinant form 01AE of Amsterdam MSM and background individuals.

Colours of tips show the observed states of each observed sequence, and colours of lineages represent inferred states. States were assigned to each sequence as described in Equation S22, and represent both transmission group (MSM, non-MSM) and place of birth or residence.

Appendix 1—figure 7
Download asset Open asset
Annotated phylogeny of viral sequences of circulating recombinant form 06cpx of Amsterdam MSM and background individuals.

Colours of tips show the observed states of each observed sequence, and colours of lineages represent inferred states. States were assigned to each sequence as described in Equation S22, and represent both transmission group (MSM, non-MSM) and place of birth or residence.

Appendix 1—figure 8
Download asset Open asset
Annotated phylogeny of viral sequences of a sub-clade of subtype B of Amsterdam MSM and background individuals.

Colours of tips show the observed states of each observed sequence, and colours of lineages represent inferred states. States were assigned to each sequence as described in Equation S22, and represent both transmission group (MSM, non-MSM) and place of birth or residence.

Appendix 1—figure 9
Download asset Open asset
Annotated phylogeny of viral sequences of a sub-clade of subtype B of Amsterdam MSM and background individuals.

Colours of tips show the observed states of each observed sequence, and colours of lineages represent inferred states. States were assigned to each sequence as described in Equation S22, and represent both transmission group (MSM, non-MSM) and place of birth or residence.

Appendix 1—figure 10
Download asset Open asset
Annotated phylogeny of viral sequences of a sub-clade of subtype B of Amsterdam MSM and background individuals.

Colours of tips show the observed states of each observed sequence, and colours of lineages represent inferred states. States were assigned to each sequence as described in Equation S22, and represent both transmission group (MSM, non-MSM) and place of birth or residence.

Appendix 1—figure 11
Download asset Open asset
Annotated phylogeny of viral sequences of a sub-clade of subtype B of Amsterdam MSM and background individuals.

Colours of tips show the observed states of each observed sequence, and colours of lineages represent inferred states. States were assigned to each sequence as described in Equation S22, and represent both transmission group (MSM, non-MSM) and place of birth or residence.

Appendix 1—figure 12
Download asset Open asset
Annotated phylogeny of viral sequences of subtype C of Amsterdam MSM and background individuals.

Colours of tips show the observed states of each observed sequence, and colours of lineages represent inferred states. States were assigned to each sequence as described in Equation S22, and represent both transmission group (MSM, non-MSM) and place of birth or residence.

Appendix 1—figure 13
Download asset Open asset
Annotated phylogeny of viral sequences of subtype D of Amsterdam MSM and background individuals.

Colours of tips show the observed states of each observed sequence, and colours of lineages represent inferred states. States were assigned to each sequence as described in Equation S22, and represent both transmission group (MSM, non-MSM) and place of birth or residence.

Appendix 1—figure 14
Download asset Open asset
Annotated phylogeny of viral sequences of subtype G of Amsterdam MSM and background individuals.

Colours of tips show the observed states of each observed sequence, and colours of lineages represent inferred states. States were assigned to each sequence as described in Equation S22, and represent both transmission group (MSM, non-MSM) and place of birth or residence.

Appendix 1—figure 15
Download asset Open asset
Annotated phylogeny of viral sequences of subtype A1 of Amsterdam heterosexual and background individuals.

Colours of tips show the observed states of each observed sequence, and colours of lineages represent inferred states. States were assigned to each sequence as described in Equation S23, and represent both transmission group (heterosexual, non-heterosexual) and place of birth or residence.

Appendix 1—figure 16
Download asset Open asset
Annotated phylogeny of viral sequences of circulating recombinant form 02AG of Amsterdam heterosexual and background individuals.

Colours of tips show the observed states of each observed sequence, and colours of lineages represent inferred states. States were assigned to each sequence as described in Equation S23, and represent both transmission group (heterosexual, non-heterosexual) and place of birth or residence.

Appendix 1—figure 17
Download asset Open asset
Annotated phylogeny of viral sequences of circulating recombinant form 01AE of Amsterdam heterosexual and background individuals.

Colours of tips show the observed states of each observed sequence, and colours of lineages represent inferred states. States were assigned to each sequence as described in Equation S23, and represent both transmission group (heterosexual, non-heterosexual) and place of birth or residence.

Appendix 1—figure 18
Download asset Open asset
Annotated phylogeny of viral sequences of circulating recombinant form 06cpx of Amsterdam heterosexual and background individuals.

Colours of tips show the observed states of each observed sequence, and colours of lineages represent inferred states. States were assigned to each sequence as described in Equation S23, and represent both transmission group (heterosexual, non-heterosexual) and place of birth or residence.

Appendix 1—figure 19
Download asset Open asset
Annotated phylogeny of viral sequences of a sub-clade of subtype B of Amsterdam heterosexual and background individuals.

Colours of tips show the observed states of each observed sequence, and colours of lineages represent inferred states. States were assigned to each sequence as described in Equation S23, and represent both transmission group (heterosexual, non-heterosexual) and place of birth or residence.

Appendix 1—figure 20
Download asset Open asset
Annotated phylogeny of viral sequences of a sub-clade of subtype B of Amsterdam heterosexual and background individuals.

Colours of tips show the observed states of each observed sequence, and colours of lineages represent inferred states. States were assigned to each sequence as described in Equation S23, and represent both transmission group (heterosexual, non-heterosexual) and place of birth or residence.

Appendix 1—figure 21
Download asset Open asset
Annotated phylogeny of viral sequences of a sub-clade of subtype B of Amsterdam heterosexual and background individuals.

Colours of tips show the observed states of each observed sequence, and colours of lineages represent inferred states. States were assigned to each sequence as described in Equation S23, and represent both transmission group (heterosexual, non-heterosexual) and place of birth or residence.

Appendix 1—figure 22
Download asset Open asset
Annotated phylogeny of viral sequences of a sub-clade of subtype B of Amsterdam heterosexual and background individuals.

Colours of tips show the observed states of each observed sequence, and colours of lineages represent inferred states. States were assigned to each sequence as described in Equation S23, and represent both transmission group (heterosexual, non-heterosexual) and place of birth or residence.

Appendix 1—figure 23
Download asset Open asset
Annotated phylogeny of viral sequences of subtype C of Amsterdam heterosexual and background individuals.

Colours of tips show the observed states of each observed sequence, and colours of lineages represent inferred states. States were assigned to each sequence as described in Equation S23, and represent both transmission group (heterosexual, non-heterosexual) and place of birth or residence.

Appendix 1—figure 24
Download asset Open asset
Annotated phylogeny of viral sequences of subtype D of Amsterdam heterosexual and background individuals.

Colours of tips show the observed states of each observed sequence, and colours of lineages represent inferred states. States were assigned to each sequence as described in Equation S23, and represent both transmission group (heterosexual, non-heterosexual) and place of birth or residence.

Appendix 1—figure 25
Download asset Open asset
Annotated phylogeny of viral sequences of subtype G of Amsterdam heterosexual and background individuals.

Colours of tips show the observed states of each observed sequence, and colours of lineages represent inferred states. States were assigned to each sequence as described in Equation S23, and represent both transmission group (heterosexual, non-heterosexual) and place of birth or residence.

Appendix 1—figure 26
Download asset Open asset
Growth of phylogenetically observed subgraphs by subtype.

First column (index cases = 0) are for emergent chains, where the index case is among the newly generated cases. (A) Subgraphs among Amsterdam MSM. (B) Subgraphs among Amsterdam heterosexuals.

Appendix 1—figure 27
Download asset Open asset
Posterior predictive check for Amsterdam MSM (top) and Amsterdam heterosexuals (bottom) for B and non-B subtypes.

Estimates generated from 203 phylogenetic subgraphs among Amsterdam MSM, containing 297 individuals, and 41 subgraphs among Amsterdam heterosexuals, containing 44 individuals.

Appendix 1—figure 28
Download asset Open asset
Estimates for the proportion of HIV infections acquired within 6 months of diagnosis from the bivariate linear mixed model (BLMM) approach (for infections diagnosed between 2013-2015), compared with estimates obtained from avidity assays in a study by Slurink et al., 2021 (for infections diagnosed between 2013-2015).
Appendix 1—figure 29
Download asset Open asset
Map of Amsterdam postal code (PC4) areas.
Appendix 1—figure 30
Download asset Open asset
Pairs plot of the joint posterior density of the model parameters for MSM time-to-diagnosis model.
Appendix 1—figure 31
Download asset Open asset
Pairs plot of the joint posterior density of the model parameters for heterosexual time-to-diagnosis model.
Appendix 1—figure 32
Download asset Open asset
Trace plot for parameter with smallest effective sample size in MSM time-to-diagnosis model.
Appendix 1—figure 33
Download asset Open asset
Trace plot for parameter with smallest effective sample size in heterosexual time-to-diagnosis model.
Appendix 1—figure 34
Download asset Open asset
Posterior median cumulative distribution functions (CDFs) (line in colours) and 95% credible intervals (ribbon in colours) are shown along with the empirical CDF (steps in black).
Appendix 1—figure 35
Download asset Open asset
Estimated Amsterdam infections in 2014-2018.

Estimates of the total number of individuals resident in Amsterdam that were infected in 2014-2018 are shown along with the subset of individuals that were diagnosed, and the subset of those for who at least one viral sequence is available. Posterior median estimates (bars, and number on top of bar) are shown along with 95% credible intervals. The posterior median proportion of individuals with a viral sequence is also shown (proportion on top of bar).

Appendix 1—figure 36
Download asset Open asset
Estimated proportion of Amsterdam infections in 2014-2018 which remained undiagnosed by 2019, by year of infection.

(A) Using all biomarker data from all individuals. (B) Using midpoint estimates from seroconverters.

Appendix 1—figure 37
Download asset Open asset
Estimated Amsterdam infections in 2014-2018, using midpoint estimates from seroconverters.

Estimates of the total number of individuals resident in Amsterdam that were infected in 2014-2018 are shown along with the subset of individuals that were diagnosed, and the subset of those for who at least one viral sequence is available. Posterior median estimates (bars, and number on top of bar) are shown along with 95% credible intervals. The posterior median proportion of individuals with a viral sequence is also shown (proportion on top of bar).

Appendix 1—figure 38
Download asset Open asset
Growth of pre-existing (left) and emergent (right) phylogenetically observed subgraph sizes using estimated date of infection.

Estimates generated from 203 phylogenetic subgraphs among Amsterdam MSM, containing 297 individuals, and 41 subgraphs among Amsterdam heterosexuals, containing 44 individuals. * pre-existing prior to 2014.

Appendix 1—figure 39
Download asset Open asset
Top: Composition of subtypes among total predicted new cases.

Bottom: Estimated local infections among MSM (left) and heterosexuals (right), stratified by place of birth between 2014-2018. N = number of sequences available, N* = estimated actual infections [95% credible interval].

Appendix 1—figure 40
Download asset Open asset
Trace plot of parameter with the smallest effective sample size for Amsterdam MSM model.
Appendix 1—figure 41
Download asset Open asset
Trace plot of parameter with the smallest effective sample size for Amsterdam heterosexual model.
Appendix 1—figure 42
Download asset Open asset
Pairs plot of the joint posterior density of the model parameters for Amsterdam MSM.
Appendix 1—figure 43
Download asset Open asset
Pairs plot of the joint posterior density of the model parameters for Amsterdam heterosexuals.
Appendix 1—figure 44
Download asset Open asset
Growth of pre-existing and emergent phylogenetically observed subgraph sizes using diagnosis date and estimated date of infection.

* pre-existing prior to 2014.

Author response image 1
Download asset Open asset

Tables

Table 1
HIV infections among Amsterdam residents in 2014-2018.
Risk groupObserved HIV diagnoses in Amsterdam residents in 2014-May 2019(n)Observed HIV diagnoses in Amsterdam residents in 2014-May 2019 with CD4 <350(n)Observed HIV diagnoses in Amsterdam residents, estimated to have been infected in 2014–2018(n)Estimated undiagnosed HIV infections in Amsterdam residents until May 2019(%)Estimated HIV infections in Amsterdam residents in 2014–2018(n)
Total84627551619% [17–21%]636 [620-656]
MSM (all)67119244614% [12–16%]516 [506-529]
MSM (Dutch-born)29810319011% [9–13%]214 [209-219]
MSM (Born in W. Europe, N. America and Oceania)10012809% [6–14%]88 [85-93]
MSM (Born in E. and C. Europe)5183216% [11–24%]38 [36-42]
MSM (Born in S. America and the Caribbean)124388317% [13–22%]100 [95-107]
MSM (Born in any other country)98316120% [14–27%]76 [71-83]
Heterosexuals (all)175837041% [35–48%]119 [107-135]
Heterosexuals (Dutch-born)51192330% [21–44%]33 [29-41]
Heterosexuals (Born in Sub-Saharan Africa)67361757% [47–67%]40 [32-51]
Heterosexuals (Born in S. America and the Caribbean)37182128% [19–42%]29 [26-36]
Heterosexuals (Born in any other country)2010940% [25–57%]15 [12-21]

  1. Posterior estimated median time from infection to diagnosis [95% CI].

Table 2
Growth distribution of transmission chains among Amsterdam residents in 2014–2018.
Observed*Predicted†
Pre-existing chainsEmerging chainsPre-existing chainsEmerging chains
Transmission groupNew cases(N)(%)(N)(%)(N)(%)(N)(%)
Amsterdam MSM022071.2%--198 [175-221]64.1% [56.6–71.5%]--
15919.1%9482.5%52 [37-69]16.8% [12.0–22.3%]137 [118-158]79.7% [72.3–86.1%]
2154.9%119.6%23 [14-35]7.4% [4.5–11.3%]19 [11-30]11.2% [6.3–17.0%]
361.9%76.1%13 [6-20]4.2% [1.9–6.5%]7 [2-13]4.1% [1.2–7.6%]
431.0%21.8%7 [3-14]2.3% [1.0–4.5%]3 [0–8]1.8% [0.0–4.3%]
520.6%00.0%4 [1-10]1.3% [0.3–3.2%]2 [0–5]1.1% [0.0–2.9%]
600.0%00.0%3 [0–7]1.0% [0.0–2.3%]1 [0–4]0.6% [0.0–2.1%]
7+41.3%00.0%7 [2-14]2.3% [0.6–4.5%]2 [0–6]1.1% [0.0–3.2%]
Total that grew89114111 [88-134]172 [154-195]
Total309114309 [309-309]172 [154-195]
Amsterdam heterosexual015090.9%--138 [123-150]83.6% [74.5–90.9%]-
1137.9%2596.2%17 [9-28]10.3% [5.5–17.0%]50 [35-72]86.4% [74.1–95.6%]
221.2%13.8%5 [1-11]3.0% [0.6–6.7%]5 [1-12]9.3% [2.0–19.0%]
300.0%00.0%2 [0–6]1.2% [0.0–3.6%]1 [0–5]2.0% [0.0–7.8%]
400.0%00.0%1 [0–3]0.6% [0.0–1.8%]0 [0–2]0.0% [0.0–4.3%]
500.0%00.0%0 [0–2]0.0% [0.0–1.2%]0 [0–2]0.0% [0.0–2.6%]
600.0%00.0%0 [0–2]0.0% [0.0–1.2%]0 [0–1]0.0% [0.0–2.0%]
7+00.0%00.0%0 [0–3]0.0% [0.0–1.8%]0 [0–1]0.0% [0.0–2.0%]
Total that grew152627 [15-42]58 [42-83]
Total16526165 [165-165]58 [42-83]
  1. *

    Parts of the actual Amsterdam transmission chains were observed in viral phylogenies of the major subtypes and circulating recombinant forms (B, 01AE, 02AG, C, D, G, A1 or 06 cpx).

  2. †

    Predicted based on the Bayesian branching process growth model and accounting for undiagnosed and unsampled individuals.

Table 3
Distribution of Amsterdam infections since 2014 in pre-existing and emerging transmission chains.
Observed*Predicted†
TotalIn pre-existing chainsIn emerging chainsTotalIn pre-existing chainsIn emerging chains
(N)(N)(%)(N)(%)(N)(N)(%)(N)(%)
MSM (Dutch)1458659.30%5940.70%254 [202-318]136 [95-188]53.6% [44.1–62.4%]117 [93-147]46.4% [37.6–55.9%]
MSM (W. Europe, N. America, Oceania)402562.50%1537.50%68 [49-91]37 [23-56]54.8% [40.5–68.1%]31 [20-43]45.2% [31.9–59.5%]
MSM (E. & C. Europe)17952.90%847.10%29 [18-42]15 [8-25]53.6% [34.2–72.7%]13 [7-21]46.4% [27.3–65.8%]
MSM (S. America & Caribbean)532445.30%2954.70%95 [72-126]50 [33-74]52.8% [40.3–64.8%]45 [31-61]47.2% [35.2–59.7%]
MSM (Other)421433.30%2866.70%76 [55-103]37 [22-57]48.4% [34.4–61.7%]39 [26-56]51.6% [38.3–65.6%]
MSM (All)29715853.20%13946.80%523 [427-647]276 [200-377]52.8% [44.6–60.7%]246 [206-300]47.2% [39.3–55.4%]
Heterosexual (Dutch)14214.30%1285.70%38 [23-59]14 [5-29]37.8% [17.5–58.9%]23 [13-38]62.2% [41.1–82.5%]
Heterosexual (Sub-Saharan Africa)11436.40%763.60%30 [17-51]10 [3-24]34.3% [11.3–58.6%]20 [11-34]65.7% [41.4–88.7%]
Heterosexual (S. America & Caribbean)14857.10%642.90%35 [20-58]14 [5-33]42.9% [18.6–65.8%]19 [10-34]57.1% [34.2–81.4%]
Heterosexual (Other)5360.0%240.0%13 [6-23]5 [1-12]39.1% [9.1–70.0%]8 [3-15]60.9% [30.0–90.9%]
Heterosexual (All)441738.60%2761.40%117 [80-173]45 [22-83]38.7% [22.6–54.9%]71 [49-105]61.3% [45.1–77.4%]
  1. *

    Parts of the actual Amsterdam transmission chains were observed in viral phylogenies of the major subtypes and circulating recombinant forms (B, 01AE, 02AG, C, D, G, A1 or 06 cpx).

  2. †

    Predicted based on the Bayesian branching process growth model and accounting for undiagnosed and unsampled individuals.

Appendix 1—table 1
Patient characteristics for Amsterdam residents with an estimated infection date between 2014 and 2018.
StrataAll patientsPatients with a sequence*
SexFemale40 (7.8%)24 (7%)
Male476 (92.2%)317 (93%)
Risk groupMSM446 (86.4%)297 (87.1%)
Heterosexual70 (13.6%)44 (12.9%)
Age group at estimated time of infection18–2474 (14.3%)48 (14.1%)
25–34209 (40.5%)124 (36.4%)
35–44113 (21.9%)76 (22.3%)
45–59110 (21.3%)87 (25.5%)
60+10 (1.9%)6 (1.8%)
Place of birthSub-Saharan Africa24 (4.8%)16 (4.8%)
Asia20 (4%)13 (3.9%)
Australia & New Zealand2 (0.4%)2 (0.6%)
Central Europe25 (5%)16 (4.8%)
Eastern Europe8 (1.6%)1 (0.3%)
Suriname, Curacao and Aruba41 (8.1%)32 (9.6%)
South America and Caribbean63 (12.5%)35 (10.5%)
Middle East and North Africa31 (6.1%)20 (6%)
Netherlands213 (42.2%)159 (47.6%)
North America23 (4.6%)14 (4.2%)
Western Europe55 (10.9%)26 (7.8%)
Estimated time to diagnosis (years)0.4 [0.04–3.2]0.41 [0.03–3.25]
  1. *

    Patients with sequence of a subtype or circulating recombinant form B, 01AE, 02AG, C, D, G, A1 or 06 cpx

Appendix 1—table 2
Number and size of phylogenetically observed transmission chains by transmission risk group and HIV subtype or circulating recombinant form (CRF) for central analysis.

95% confidence intervals are obtained from 100 bootstrap analyses for each subtype alignment.

Risk groupSubtype or CRFTotal number of chainsChains of size 1Chains of size 2-5Chains of size 5-10Chains of size ≥10
Amsterdam MSMB1237 [1259-2097]856 [872-1446]276 [264-479]64 [58-116]41 [32-66]
01AE41 [37-46]24 [21-32]15 [12-17]2 [0-3]0 [0-1]
02AG26 [21-34]17 [14-27]7 [2-9]1 [0-4]1 [0-2]
C26 [24-28]22 [18-25]4 [3-6]0 [0-0]0 [0-0]
A121 [18-25]13 [10-18]6 [4-7]0 [0-3]2 [0-2]
G9 [8-9]0 [0-0]8 [6-8]1 [1-2]0 [0-0]
D6 [6-6]6 [6-6]0 [0-0]0 [0-0]0 [0-0]
06cpx2 [2-2]2 [2-2]0 [0-0]0 [0-0]0 [0-0]
Amsterdam heterosexualsB277 [272-482]225 [217-392]45 [39-77]6 [2-9]1 [1-3]
01AE23 [20-24]19 [15-21]4 [3-6]0 [0-0]0 [0-0]
02AG111 [106-126]77 [77-100]30 [20-31]4 [1-6]0 [0-1]
C87 [82-89]72 [63-75]15 [13-19]0 [0-1]0 [0-0]
A143 [37-49]34 [30-42]8 [3-12]1 [0-2]0 [0-1]
G28 [28-33]22 [20-29]6 [4-8]0 [0-0]0 [0-0]
D16 [15-18]12 [10-16]4 [2-5]0 [0-0]0 [0-0]
06cpx17 [14-21]12 [8-15]4 [2-8]1 [0-2]0 [0-1]
Appendix 1—table 3
Estimated numbers of phylogenetic transmission chains with ancestral origins in each geographic region from central analysis.

95% confidence intervals obtained from 100 bootstrap analyses for each subtype alignment.

Subtype or CRFEstimated ancestral originAmsterdam MSMAmsterdam heterosexuals
BAmsterdam - other risk group16 [8-27]73 [59-124]
Netherlands699 [721-1238]110 [113-199]
Western Europe147 [133-253]18 [6-24]
Eastern Europe and Central Asia27 [21-46]1 [1-3]
North America84 [71-151]7 [4-20]
South America and Caribbean21 [16-43]1 [1-4]
Middle East and North Africa2 [1-5]-
South and South-East Asia3 [2-8]-
Oceania1 [1-3]-
01AEAmsterdam - other risk group-2 [1-4]
Netherlands11 [5-17]10 [5-14]
Middle East and North Africa1 [1-1]-
South and South-East Asia21 [14-24]8 [3-9]
02AGAmsterdam - other risk group-5 [3-8]
Netherlands11 [6-20]29 [20-39]
Sub-Saharan Africa4 [1-7]39 [29-51]
Western Europe5 [1-4]2 [1-9]
CAmsterdam - other risk group2 [1-3]1 [1-2]
Netherlands8 [3-9]21 [15-26]
Sub-Saharan Africa4 [2-7]29 [25-39]
Western Europe1 [1-3]2 [1-7]
South America and Caribbean2 [1-3]1 [1-1]
South and South-East Asia3 [1-3]1 [1-2]
A1Amsterdam - other risk group1 [1-2]3 [1-5]
Netherlands10 [6-13]19 [12-24]
Sub-Saharan Africa1 [1-2]11 [9-17]
Western Europe2 [1-3]-
Eastern Europe and Central Asia1 [1-2]-
A1South and South-East Asia3 [1-3]-
Netherlands2 [1-3]5 [1-7]
Sub-Saharan Africa1 [1-3]12 [9-18]
Western Europe1 [1-2]3 [1-6]
GEastern Europe and Central Asia2 [1-2]1 [1-1]
Netherlands1 [1-2]2 [1-6]
DSub-Saharan Africa2 [1-3]9 [5-11]
Netherlands-1 [1-4]
Sub-Saharan Africa1 [1-1]9 [6-14]
Western Europe1 [1-1]-
Appendix 1—table 4
Viral suppression status of the phylogenetically observed pre-2014 Amsterdam transmission chains.
Risk groupSubtypeAll sampled individuals virally suppressed by 2014*Pre-2014 chainsPre-2014 chains that grewIndividuals (Total)Individuals (infected before 2014)Individuals (infected before 2014 and not virally suppressed)
(n)(n)(%)(n)(n)(%)(n)(%)
Amsterdam MSMBYes866354%1432127989%00%
BNo2864415%1740130375%35220%
Non-BYes83810%17211969%00%
Non-BNo18211%805164%2329%
Total1253897%3424275280%37511%
Amsterdam heterosexualBYes18053%21820092%00%
BNo8545%28418967%9032%
Non-BYes22152%30128193%00%
Non-BNo9011%23514260%9239%
Total576153%103881278%18218%
Total18291046%4462356480%55712%
  1. *

    Individuals infected prior to 2014, with last viral load measurement before 2014 below 100copies/ml.

Appendix 1—table 5
Observed and estimated ancestral origins of phylogenetic subgraphs and estimated complete transmission chains with new cases in 2014-2018.
Risk groupSubtypeOrigin of chainsObserved (N)Observed (%)Predicted (N)Predicted (%)
Amsterdam MSMBAmsterdam - other risk group1 [1-3]0.8% [0.5-2%]2 [1-6]0.5% [0.2-1.4%]
Asia2 [2-4]1.5% [1-2.3%]6 [2-12]1.5% [0.5-2.8%]
Eastern Europe and Central Asia7 [4-13]5% [2.9-7.3%]21 [12-30]5% [3-7.3%]
South America and Caribbean5 [2-12]3.2% [1.5-5.9%]14 [8-22]3.4% [1.9-5.4%]
Middle East and North Africa1 [1-2]0.8% [0.5-1.3%]3 [1-7]0.7% [0.2-1.7%]
Netherlands96 [84-159]71.1% [64-77.1%]294 [272-317]71.1% [66.8-75.4%]
North America8 [4-17]5.7% [2.5-9.3%]23 [15-33]5.7% [3.6-8%]
Oceania2 [2-2]1% [1-1%]1 [1-2]0.2% [0.2-0.5%]
Western Europe16 [11-29]11.7% [8-15.9%]48 [36-61]11.6% [8.7-14.9%]
Non-BSub-Saharan Africa3 [1-5]10.7% [3.6-19.6%]7 [3-13]10.8% [4.2-19%]
Amsterdam - other risk group1 [1-3]3.9% [3.3-11.4%]2 [1-4]2.5% [1.3-6.2%]
Asia8 [6-11]31% [22.2-42.3%]21 [13-30]31.3% [20.3-43.1%]
Eastern Europe and Central Asia1 [1-1]3.5% [3.3-3.6%]1 [1-2]1.5% [1.3-2.8%]
South America and Caribbean1 [1-2]4% [3.3-8.2%]3 [1-7]4.4% [1.4-10%]
Middle East and North Africa1 [1-1]3.6% [3.3-4%]1 [1-3]1.5% [1.3-4.2%]
Netherlands12 [8-16]46.4% [32.1-59.5%]31 [22-41]45.9% [34.2-57.8%]
Amsterdam heterosexualBAmsterdam - other risk group3 [1-7]21.4% [7.4-38.5%]22 [14-30]21.4% [13.8-29.4%]
Eastern Europe and Central Asia1 [1-1]7.2% [6.7-7.7%]1 [1-2]1% [0.9-1.9%]
Netherlands11 [8-17]75% [54.8-92%]75 [64-89]74.8% [66.3-82.8%]
North America1 [1-3]6.7% [4.7-10.6%]2 [1-4]1.9% [0.9-4.2%]
Western Europe1 [1-3]7.1% [5.3-20.3%]2 [1-6]2.1% [0.9-5.5%]
Non-BSub-Saharan Africa5 [2-8]33.3% [9.4-51.9%]39 [29-51]31.9% [24-40.5%]
Amsterdam - other risk group1 [1-2]6.7% [5.4-12.5%]9 [3-15]7% [2.7-11.8%]
Asia1 [1-1]6.7% [5.7-9.8%]2 [1-6]1.7% [0.8-4.7%]
Netherlands8 [4-12]50% [28.9-74.2%]62 [50-77]50.4% [41.7-59.7%]
North America1 [1-1]5.6% [5.6-5.6%]1 [1-2]0.8% [0.7-1.6%]
Appendix 1—table 6
Patient characteristics for individuals with an estimated infection date between 2010-2012.
Risk groupPlace of birthAmsterdam infections 2010-2012Median estimated time to diagnosis (years) [95% quantiles]
Amsterdam MSMW.Europe, N.America, Oceania720.42 [0.05-3.41]
E. & C. Europe310.88 [0.13-6.04]
S. America & Caribbean811.04 [0.05-5.57]
Netherlands2950.56 [0.04-4.77]
Other561.38 [0.12-4.97]
All5350.64 [0.04-4.97]
Amsterdam heterosexualSub-Saharan Africa353.86 [0.33-6.8]
S. America & Caribbean221.37 [0.14-5.68]
Netherlands271.42 [0.07-6.16]
Other131.6 [0.99-6.12]
All972.22 [0.1-6.67]
Appendix 1—table 7
Estimated undiagnosed HIV infections in Amsterdam until May 2019 using equal weights, or weighting by diagnosis rates or estimated infection rates.
Estimated undiagnosed HIV infections
Risk groupRegion of birthEqual weightsWeighted by diagnosis ratesWeighted by infection rates
Amsterdam MSMNetherlands17% [15-20%]11% [9-13%]11% [9-13%]
W. Europe, N. America, Oceania16% [11-21%]9% [6-13%]9% [6-14%]
E. & C. Europe22% [16-32%]14% [9-22%]16% [11-24%]
S. America and Caribbean23% [19-30%]19% [14-25%]17% [13-22%]
Other27% [20-34%]23% [16-31%]20% [14-27%]
All20% [18-22%]14% [13-17%]14% [12-16%]
Amsterdam heterosexualNetherlands34% [23-47%]28% [18-39%]30% [21-44%]
Sub-Saharan Africa60% [48-69%]48% [37-59%]57% [47-67%]
S. America and Caribbean30% [19-45%]25% [16-38%]28% [19-42%]
Other44% [31-59%]31% [18-50%]40% [25-57%]
All44% [37-50%]34% [28-41%]41% [35-48%]
All24% [22-27%]18% [16-20%]19% [17-21%]
Appendix 1—table 8
Input quantities used to estimate proportion of infections acquired locally in Amsterdam.
Risk groupSubtypeChains of non-Amsterdam origin (1 − λ)Phylogenetically observed emergent subgraphs (|x~|)Emergent transmission chains (unobserved) (N∗not-obs)Total emergent chains (partially observed + unobserved |x~|+Nnot-obs∗|)Individuals in pre-existing and emergent chains (x∗)Proportion of infections that are importations ((1−λ)(|x~|+Nnot-obs∗)x∗)External importations ((1−λ)(|x~|+Nnot-obs∗)x∗)Locally acquired infections (γ)
Amsterdam hetersexualB78.6% [70.6-86.2%]12 [12-12]14 [5-30]26 [17-42]58 [35-95]0.47 [0.27-0.7]36.6% [21.1-55.6%]63.4% [44.4-78.9%]
Amsterdam hetersexualNon-B93% [88.2-97.3%]14 [14-14]17 [7-35]31 [21-49]58 [37-93]0.55 [0.35-0.78]51% [32.1-72.5%]49% [27.5-67.9%]
Amsterdam MSMB99.5% [98.6-100%]85 [85-85]45 [30-64]130 [115-149]412 [332-521]0.32 [0.25-0.4]31.5% [24.8-39.3%]68.5% [60.7-75.2%]
Amsterdam MSMNon-B98.5% [94.1-100%]29 [29-29]13 [5-24]42 [34-53]106 [72-172]0.4 [0.24-0.58]38.7% [23.5-57.2%]61.3% [42.8-76.5%]
Appendix 1—table 9
Empirical results from partially observed subgraphs in phylogenetic trees, and model estimates based on complete transmission chains, adjusting for sampling (those in study with a sequence available) for new infections since 2014.

Estimated reproduction number and proportion of locally acquired infections are also presented.

Phylogenetically observedModel estimates
Risk groupSubtypeNew casesSubgraphsAverage new casesTransmission chainsAverage new casesEffective reproduction numberVariance-to-mean ratioInfections acquired in Amsterdam
Amsterdam MSMB2413680.65413 [398-432]1.010.26 [0.22-0.31]1.69 [1.26-2.38]68.5% [60.7-75.2%]
Amsterdam MSMNon-B65551.1868 [60-79]1.620.39 [0.28-0.53]1.33 [1.02-2.53]61.3% [42.8-76.5%]
Amsterdam heterosexualNon-B211050.2122 [112-140]0.490.17 [0.09-0.26]1.26 [1.01-2.94]49% [27.5-67.9%]
Amsterdam heterosexualB23860.27100 [91-116]0.590.19 [0.11-0.3]1.25 [1.01-2.59]63.4% [44.4-78.9%]
Author response table 1
YearMSMHeterosexuals
20140.260.19
20150.230.19
20160.200.20
20170.170.21
20180.130.22
Author response table 2
MSMHSX
yearoriginweightoriginweight
2014W.Europe, N.America, Oceania0.28Sub-Saharan Africa0.25
20150.270.30
20160.190.20
20170.130.16
20180.130.09
2014E and C. Europe0.25S. America and Caribbean0.24
20150.250.16
20160.290.27
20170.120.19
20180.080.14
2014S. America and Caribbean0.20NL0.24
20150.200.24
20160.200.20
20170.250.14
20180.160.18
2014NL0.28Other0.20
20150.240.20
20160.200.35
20170.140.25
20180.150.00
2014Other0.23
20150.15
20160.23
20170.21
20180.19

Additional files

MDAR checklist
https://cdn.elifesciences.org/articles/76487/elife-76487-mdarchecklist1-v2.docx

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. Alexandra Blenkinsop
  2. Mélodie Monod
  3. Ard van Sighem
  4. Nikos Pantazis
  5. Daniela Bezemer
  6. Eline Op de Coul
  7. Thijs van de Laar
  8. Christophe Fraser
  9. Maria Prins
  10. Peter Reiss
  11. Godelieve J de Bree
  12. Oliver Ratmann
  13. On behalf of HIV Transmission Elimination AMsterdam (H-TEAM) collaboration
(2022)
Estimating the potential to prevent locally acquired HIV infections in a UNAIDS Fast-Track City, Amsterdam
eLife 11:e76487.
https://doi.org/10.7554/eLife.76487