HTML
-
The demographic characteristics of all participants were summarized in Table 1. In 2017, 191 convalescent sera were collected, 63 (33.0%) of which were males, the other 128 (67.0%) were females, with the male-female ratio of 0.49. The mean age of participants was 45 ± 14.8 which ranged from 8 to 76 years. We grouped them in decade increments. The highest proportion of participants was in the 41–50 years group (27.7%) and the lowest was in 0–10 years group (1.6%). Of the 99 participants enrolled in 2019, 32 (32.3%) were males, 67 (67.7%) were females, and the male-female ratio was 0.48. The mean age of participants was 45.7 ± 12.1 which ranged from 6 to 83 years. The 41–50 years group had the highest proportion (34.5%), while the 0–10 years group had the lowest (1.0%).
Variables Seropositive, n (%) P-valuec 2017a 2019b Gender Male 63/63 (100) 28/32 (87.5) ns Female 125/128 (97.7) 54/67 (80.6) < 0.0001 Age (y) 0–10 3/3 (100) 1/1 (100) ns 11–20 9/9 (100) 4/4 (100) ns 21–30 16/16 (100) 3/4 (75.0) ns 31–40 37/37 (100) 23/24 (95.8) ns 41–50 52/53 (98.1) 24/35 (68.6) 0.0001 51–60 40/42 (95.2) 21/24 (87.5) ns > 60 31/31 (100) 6/7 (85.7) ns Total 188/191 (98.4) 82/99 (82.8) 0.0004 a Four-year serum samples collected in 2017.
b Six-year serum samples collected in 2019.
c The statistical analysis was performed by Chi-square test.
IgG, immunoglobulin G; DENV, dengue virus; ns, no significance.Table 1. Demographic characteristics of participants and positive rates of DENV-specific IgG antibodies in patients at four-year and six-year convalescent phases for baseline serosurvey.
Demographic Characteristics of Participants
-
The seropositive rates of serum IgG against DENV were detected by ELISA and were shown in Table 1. Firstly, of 191 serum samples collected in 2017, the overall dengue IgG-positive rate was still as high as 98.4% (188/191) four years after DENV infection. In 2019, the positive rate of serum IgG was 82.8% (82/99) in those with a prognosis of six years. It meant that the DENV-specific IgG seroconversion rate decreased with the time elapsed post-infection (P = 0.0004). Moreover, the waning trend of IgG-positive rate was more pronounced in female (from 97.7% to 80.6%, P < 0.0001) than in male (from 100% to 87.5%). Meanwhile, in different age groups, significant decreasing of seropositivity was only observed in 41–50 age groups after two years (98.1% vs. 68.6%, P = 0.0001).
-
To investigate the change of DENV-3 specific NAb in convalescent sera, overall samples collected at four- or six-years post-infection were assayed by PRNT50. The distribution of NAb titers in overall sample population was shown in Fig. 1. At the fourth year of prognosis, NAb against DENV-3 could still be detected in 98.4% (188/191) participants. Among them, the highest percentage of 27.2% was found at the NAb titer of 1:320. The NAb titer of 1:1280 owned the lowest percentage of 0.5% (Fig. 1A). In three patients' sera, the PRNT50 titers were less than 1:10, indicating that 1.6% of four-year serum specimens had completely lost their neutralizing activity.
Figure 1. The titer distribution of NAb against DENV-3 in overall serum samples for baseline serosurvey. A NAb titers in four-year sera collected in 2017 (n = 191). B NAb titers in six-year sera collected in 2019 (n = 99). NAb, neutralizing antibody; DENV, dengue virus.
In six-year serum samples, the peak percentage of PRNT50 titer achieved 20.2% which occurred at the NAb titer of 1:80 (Fig. 1B), and the lowest percentage was 3.0% which was observed at the titer of 1:640. NAb against DENV-3 was detected in 82.8% (82/99) of samples, which was less than that in four-year sera (P < 0.0001). NAb against DENV-3 was undetectable (titer < 1:10) in 17.2% of the patients. In addition, as compared with the highest titer in four-year sera, the titer of 1:1280 was not observed in six-year sera. These results demonstrated that the neutralizing activity decreased significantly in two-year interval.
-
The GMT of NAb against DENV-3 in overall samples was presented with subset variables of gender and age (Table 2). In six-year serum samples, the GMT decreased to 1:46.66 (95% C.I. 35.09–62.04) as two years passed, which was dramatically lower than that of the four-year sera (P < 0.0001). In the male group, the GMT in six-year sera was significantly lower than that in four-year sera [1:59.07 (95% C.I. 36.13-96.58) vs. 1:160 (95% C.I. 125.12-204.61), P = 0.0001]. Similarly, there was a sharp decline in the GMT of the female population [1:153.06 (95% C.I. 129.18-181.34) vs. 1:41.69 (95% C.I. 29.33-59.24), P < 0.0001]. In age groups, the highest GMT of the specimens collected in two different convalescent phases was both observed in 11–20 years group. Meanwhile, statistically significant decrease of the GMT was found in age groups of 31–40, 41–50 and 51–60 years (P = 0.0074, P < 0.0001, P = 0.0472, respectively).
Variables NAb against DENV-3a P-valued 2017b (95% C.I.) 2019c (95% C.I.) Gender Male 160.00 (125.12–204.61) 59.07 (36.13–96.58) 0.0001 Female 153.06 (129.18–181.34) 41.69 (29.33–59.24) < 0.0001 Age (y) 0–10 126.99 (47.74–337.79) 80.00 (80.00–80.00) ns 11–20 201.59 (125.07–324.92) 134.54 (100.25–180.55) ns 21–30 146.72 (107.09–201.02) 40.00 (13.49–118.59) ns 31–40 157.03 (110.21–223.74) 62.33 (38.13–101.88) 0.0074 41–50 164.32 (122.00–221.32) 24.16 (14.77–39.50) < 0.0001 51–60 134.54 (102.64–176.36) 63.50 (35.22–114.14) 0.0472 > 60 163.62 (119.66–223.72) 72.46 (24.65–212.96) ns Total 155.35 (135.20–178.51) 46.66 (35.09–62.04) < 0.0001 a NAb titers were quantified by PRNT50 and the data were shown as GMT (95% C.I.).
b Four-year serum samples collected in 2017.
c Six-year serum samples collected in 2019.
d The statistical analysis was performed by Wilcoxon rank-sum test.
PRNT50, the reciprocal highest serum dilution that corresponded to a 50% reduction of the average number of plaques on the virus infection wells; GMT, geometric mean titer; NAb, neutralizing antibody; DENV, dengue virus; C.I., confidence interval; ns, no significance.Table 2. The GMT of NAb against DENV-3 in different gender and age populations at two convalescent phases for baseline serosurvey.
-
To estimate the general change of NAb in overall samples who infected with DENV-3, PRNT50 titers were categorized as high (1:321–1:1280), medium (1:81–1:320), low-medium (1:21–1:80) and low (< 1:20) titer groups (Fig. 2). Firstly, in four-year sera, the majority (81.7%) of the samples exerted NAb titers of 1:21–1:80 (30.9%) and 1:81–1:320 (50.8%), respectively. Moreover, 12.6% of samples kept the high neutralizing activity with the titers of 1:321–1:1280 four years after the DENV infection. Only 10 (5.7%) of 191 specimens had low NAb titers less than 1:20. Secondly, in six-year sera, the proportion of the population with high, low-medium, medium and low titers were 3%, 30.3%, 29.3% and 37.4%, respectively. Obviously, the high level of NAb was transformed to medium and low levels with the convalescence prolonged.
Dynamic Seroprevalence of IgG in Overall Samples
The Titer Distribution of NAb Against DENV-3 in Overall Samples
The GMT of NAb Against DENV-3 in Different Gender and Age Populations at Two Convalescent Phases for Baseline Serosurvey
Classification of DENV-3 Specific NAb Titers in Overall Samples
-
Of all participants, 73 people were sampled both in 2017 and 2019. Their paired serum samples in different convalescent phases of dengue were analyzed as consecutive samples in this study. The demographic data were shown in Table 3. Totally, 22 (30.1%) participants were males, and 51 (69.9%) were females, with the male-female ratio as 0.43. The age distribution of consecutive participants ranged from 9 to 65 years, with a mean age of 44.2 ± 10.9 years. Of these 73 participants, 28 participants (38.4%) were 41–50 years old, accounted for the highest number of all age groups.
Variables Seropositive, n (%) P-valuec 2017a 2019b Gender Male 22/22 (100) 19/22 (86.4) ns Female 51/51 (100) 37/51 (72.5) < 0.0001 Age (y) 0–10 1/1 (100) 1/1 (100) ns 11–20 1/1 (100) 1/1 (100) ns 21–30 3/3 (100) 2/3 (66.7) ns 31–40 18/18 (100) 17/18 (94.4) ns 41–50 28/28 (100) 17/28 (60.7) 0.0002 51–60 18/18 (100) 15/18 (83.3) ns > 60 4/4 (100) 3/4 (75.0) ns Total 73/73 (100) 56/73 (76.7) < 0.0001 a Four-year serum samples collected in 2017.
b Six-year serum samples collected in 2019.
c The statistical analysis was performed by Chi-square test.
IgG, immunoglobulin G; DENV, dengue virus; ns, no significance.Table 3. Demographic characteristics of participants and positive rates of DENV-specific IgG antibodies in patients at four-year and six-year convalescent phases for follow-up serosurvey.
To further investigate the change of DENV-specific IgG antibody, the IgG positive rate of 73 pairs of consecutive sera was analyzed for follow-up serosurvey. The seroprevalence in four-year sera exerted a significant difference with that in six-year sera (100% vs. 76.7%, P < 0.0001) (Table 3). Meanwhile, in the female group, the positive rate of DENV-specific IgG in six-year sera decreased by 27.5% when compared with four-year sera (100% vs. 72.5%, P < 0.0001). In age group of 41–50 years old, after two years, the positive rate of DENV-specific IgG decreased from 100% to 60.7%. The dengue-specific IgG of 11 specimens converted to negative after two years of convalescence.
-
As shown in Table 4, in consecutive samples, the GMT of NAb in four-year sera was 1:167.70 (95% C.I. 134.55–209.22) which presented a nearly 3-fold potency than that in six-year sera with the GMT titer of 1:53.18 (95% C.I. 37.11–76.20). The greatest decrease was observed in the female group [1:183.29 (95% C.I. 140.80-238.61) vs. 1:46.45 (95% C.I. 30.42-70.94), P < 0.0001]. In six-year sera, GMTs decreased in all age groups, but only the decline of GMT in the age groups of 41–50 and 51–60 years old had statistical significance (P < 0.0001, P = 0.0443).
Variables DENV-1 P-valuec DENV-2 P-valuec DENV-3 P-valuec DENV-4 P-valuec 2017a
(95% C.I.)2019b
(95% C.I.)2017a
(95% C.I.)2019b
(95% C.I.)2017a
(95% C.I.)2019b
(95% C.I.)2017a
(95% C.I.)2019b
(95% C.I.)Gender Male 15.06
((10.40–20.70)11.71
(9.29–16.14)ns 22.69
(13.61–37.80)20.64
(12.40–34.35)ns 136.68
(92.55–201.85)82.56
(44.67–152.60)ns 56.57
(40.00–80.00)10.99
(7.81–15.47)< 0.0001 Female 13.12
(10.25–16.22)9.87
(9.29–16.14)0.0489 16.99
(13.63–21.18)12.60
(9.35–16.99)ns 183.29
(140.80–238.61)46.45
(30.42–70.94)< 0.0001 41.10
(32.30–52.31)7.22
(6.20–8.40)< 0.0001 Age (y) 0–10 10.00
(10.00–10.00)5.00
(5.00–5.00)ns 10.00
(10.00–10.00)5.00
(5.00–5.00)ns 160.00
(160.00–160.00)80.00
(80.00–80.00)ns 160.00
(160.00–160.0)10.00
(10.00–10.00)ns 11–20 40.00
(40.00–40.00)40.00
(40.00–40.00)ns 20.00
(20.00–20.00)80.00
(80.00–80.00)ns 640.00
(640.00–640.00)160.00
(160.00–160.00)ns 40.00
(40.00–40.00)20.00
(20.00–20.00)ns 21–30 25.20
(9.47–67.02)10.00
(5.27–18.97)ns 20.00
(10.54–37.95)6.30
(4.35–9.12)ns 126.99
(87.74–183.81)80.00
(22.22–287.98)ns 25.20
(12.03–52.79)10.00
(5.27–18.97)ns 31–40 12.60
(9.32–17.03)9.62
(7.05–13.13)ns 17.14
(11.75–25.03)17.14
(10.83–27.15)ns 148.14
(98.13–223.64)96.99
(57.61–163.26)ns 52.40
(34.30–74.05)8.57
(6.16–11.92)0.0002 41–50 11.90
(9.84–14.37)8.41
(6.80–10.40)0.0311 16.82
(12.05–23.48)9.28
(6.81–12.65)0.0013 172.33
(112.68–263.57)27.59
(15.70–48.48)< 0.0001 42.03
(29.44–60.00)7.07
(5.72–8.74)< 0.0001 51–60 15.27
(9.35–24.95)13.61
(8.52–21.73)ns 20.79
(12.82–33.69)31.75
(13.75–58.07)ns 186.64
(129.85–268.29)80.00
(38.48–166.30)0.0443 46.66
(32.46–67.07)8.57
(5.96–12.32)< 0.0001 > 60 16.81
(9.57–29.53)16.82
(6.96–40.64)ns 33.64
(9.13–123.97)8.41
(4.79–14.77)ns 134.54
(64.18–282.06)56.57
(9.08–352.30)ns 47.57
(22.69–99.72)10.00
(6.19–16.17)ns Total 13.80
(11.60–16.41)10.30
(8.66–12.46)0.0136 18.54
(14.93–23.01)14.60
(11.27–18.97)ns 167.70
(134.55–209.22)53.18
(37.11–76.20)< 0.0001 45.26
(37.01–55.33)8.17
(7.00–9.53)< 0.0001 NAb titers were quantified by PRNT50 and the data were shown as GMT (95% C.I.).
a Four-year serum samples collected in 2017.
b Six-year serum samples collected in 2019.
c The statistical analysis was performed by Wilcoxon rank sum test.
GMT, geometric mean titer; NAb, neutralizing antibody; DENV, dengue virus; C.I., confidence interval; ns, no significance; PRNT50, the reciprocal highest serum dilution that corresponded to a 50% reduction of the average number of plaques on the virus infection wells.Table 4. The GMT distribution of NAb against DENV-3 and cross-reactive antibody against DENV-1, DENV-2 and DENV-4 in different populations in consecutive samples for follow-up serosurvey (n = 73).
The change of NAb titers against DENV-3 in consecutive sera was similar with that in overall samples. Six years after infection, the GMT of NAb against DENV-3 in 69.9% (51/73) of the specimens displayed a decline from 1:204.35 (95% C.I. 155.81–268.01) to 1:31.32 (95% C.I. 20.99–46.73) (Fig. 3A-a). Meanwhile, the titer of NAb in 12.3% (9/73) of the samples remained unaltered (Fig. 3A-b). However, we found that the NAb titer of 17.8% (13/73) of the samples rose from 1:89.00 (95% C.I. 58.83–134.65) to 1:272.70 (95% C.I.189.00–393.47) (Fig. 3A-c). Among them, the maximum increase of NAb titer (No.71) changed from 1:40 to 1:640. We speculated that the most common reason was subsequent infection during the convalescent phase.
Figure 3. The titer distribution of neutralizing and cross-reactive antibodies in consecutive serum samples for follow-up serosurvey (n = 73). A The titer of NAb against DENV-3. B–D Cross-reactive antibodies against DENV-1 (B) DENV-2 (C) and DENV-4 (D). The first row was ordered by the titers of NAb against DENV-3 and grouped into NAb-decreased samples (a), NAb-unaltered samples (b) and NAb-increased samples (c). Every longitudinal column including eight bars (DENV-1–4 of 2017 and 2019) represented the paired samples from the same participant. NAb, neutralizing antibody; DENV, dengue virus.
-
When infected with DENV, recovered patients would generate specific NAb as well as cross-reactive antibodies simultaneously (St John and Rathore 2019). We paired consecutive serum samples at two follow-up visits to investigate the variation of cross-reactive antibodies against DENV serotype 1, 2 and 4 (Table 4).
The total GMT of the cross-reactive antibody against DENV-1 decreased from 1:13.80 (95% C.I. 11.60–16.41) to 1:10.30 (95% C.I. 8.66–12.46). In 52.1% (38/73) samples, after two years, the titer of the antibody displayed a decline. Meanwhile, the titer of cross-reactive antibody remained unaltered in 30.1% (22/73) of the samples, and rose in 17.8% (13/73) of the samples (Fig. 3B).
The total GMT of the cross-reactive antibody against DENV-2 decreased from 1:18.54 (95% C.I. 14.93–23.01) to 1:14.60 (95% C.I. 11.27–18.97). In 53.4% (39/73) specimens, the titers declined, and the proportions of titer-unaltered and titer-increased samples were 23.3% (17/73) and 23.3% (17/73), respectively (Fig. 3C).
In six-year sera, the GMT of cross-reactive antibodies against DENV-4 was dramatically lower than four-year sera [1:8.17 (95% C.I. 7.00-9.53) vs. 1:45.26 (95% C.I. 37.01-55.33), P < 0.0001] (Table 4). In gender groups, the significant decrease was observed both in the male and female groups (P < 0.0001, P < 0.0001, respectively). Statistical significance was only observed in the age group of 31–40, 41–50 and 51–60 years old when comparing the four-year and six-year sera (P = 0.0002, P < 0.0001, P < 0.0001, respectively). As shown in Fig. 3D, in 93.2% (68/73) of the samples, the level of cross-reactive antibodies against DENV-4 decreased after two years. Among them, the antibody was undetectable in 40 samples. Simultaneously, the level of the antibody remained unchanged in 4.1% (3/73) of the samples and increased in 2.7% (2/73) of the samples.
The results suggested that in patients infected with DENV-3, cross-reactive antibodies against DENV-1, DENV-2 and DENV-4 were detected in four-year sera. However, the decreasing trends of the cross-reactive antibodies were different with the convalescence being prolonged. The GMT of the cross-reactive antibody against DENV-4 declined the most.
-
ADE hypothesis was associated with the concentration of sub-neutralizing antibody and the interval between primary and secondary infection (Cedillo-Barron et al. 2014; Screaton et al. 2015). To investigate the risk of secondary heterologous DENV infection in consecutive samples from different convalescent phases, we evaluated the in vitro enhancement of DENV-1, DENV-2, and DENV-4 for follow-up serosurvey using K562 cells bearing Fc-γ receptor (Supplementary Fig. S1). The ADE enhancement was calculated as AUC based on serially-diluted sera and was showed in Fig. 4.
Figure 4. Enhancement (calculated as AUC) of DENV-1, DENV-2 and DENV-4 infections in K562 cells triggered by consecutive serum samples (n = 73 in 2017 and 2019 groups, n = 15 in naive group). A–C Scatterplots showing the enhancement in each sample. Data are shown as mean with SD. D–F Line plots (connecting means) showing the enhancement relationship and change between each paired consecutive serum sample. (**P < 0.01, ***P < 0.001). Dengue-negative serum samples from healthy subjects served as naive group. Data were non normal distribution and were analyzed by Wilcoxon rank sum test. AUC, area under curve; DENV, dengue virus; ns, no significance.
After infection of DENV-1, an obvious enhancement was exhibited in four-year sera, while the mean level of which was 3-fold greater than naïve sera, and then, the effect almost disappeared after two years (P < 0.001, Fig. 4A). We paired the 73 consecutive sera, and found that enhancement in 91.8% (67/73) samples decreased after two years (Fig. 4D).
Fig. 4B showed that, an 1.8-fold decline of the mean level of ADE, caused by DENV-2 infection, was observed in six-year sera as compared with four-year sera (P < 0.001). Among them, 67.1% (49/73) samples showed the attenuated enhancement, while 32.9% (24/73) of the samples appeared a slight increase two years later (Fig. 4E).
After infection of DENV-4, a high level of ADE was observed in both of four-year and six-year sera when compared with naïve sera (P < 0.001, Fig. 4C). Interestingly, as shown in Fig. 4F, although the mean level of ADE effect decreased in six-year sera as compared with four-year sera (Fig. 4C), 52 (71.2%) samples exerted the increasing enhancement effect with the extension of convalescent phase (Fig. 4F).
-
To quantify the risk of ADE after infection with heterotypes of DENV, we classified the values of enhancement AUC to assess the proportion of ADE in consecutive serum samples. The AUC ≤ 5, between 6 to 15, and ≥ 16 were categorized as low, medium and high levels of ADE effect, respectively. Then we used four-year samples as a reference to estimate the risk in six-year samples (Table 5). The OR value was used to evaluate the strength of the association between the occurrence of ADE effect and the time-varying risk factor.
Variables DENV-1 OR (95% C.I.) P-valuec DENV-2 OR (95% C.I.) P-valuec DENV-4 OR (95% C.I.) P-valuec 2017a 2019b 2017a 2019b 2017a 2019b Gender Male 4 1 4.67 (0.48–45.62) ns 10 4 0.27 (0.07–1.05) ns 10 19 7.60 (1.73–33.35) 0.007 Female 26 5 0.11 (0.04–0.31) < 0.001 29 18 0.41 (0.19–0.92) 0.030 14 29 3.48 (1.52–7.97) 0.003 Age (y) 0–10 0 1 9.00 (0.10–831.78) ns 1 1 1.00 (0.01–92.42) ns 0 1 9.00 (0.10–831.78) ns 11–20 0 0 1.00 (0.01–92.42) ns 1 0 0.11 (0.00–10.27) ns 1 1 1.00 (0.01–92.42) ns 21–30 1 1 1.00 (0.03–29.81) ns 1 2 4.00 (0.13–119.23) ns 1 3 11.67 (0.32–422.14) ns 31–40 9 1 0.06 (0.01–0.54) 0.012 7 2 0.20 (0.03–1.13) ns 3 13 13.00 (2.59–65.20) 0.002 41–50 12 2 0.10 (0.02–0.52) 0.006 18 10 0.31 (0.10–0.92) 0.035 11 16 2.06 (0.71–5.98) ns 51–60 5 1 0.15 (0.02–1.47) ns 11 5 0.25 (0.06–0.99) 0.049 7 13 4.09 (1.01–16.58) 0.049 > 60 2 0 0.10 (0.00–3.35) ns 0 2 9.00 (0.30–271.70) ns 1 2 3.00 (0.15–59.89) ns AUC ≤ 5 25 4 0.07 (0.02–0.22) < 0.001 21 19 0.62 (0.29–1.31) ns 10 29 5.68 (2.39–13.50) < 0.001 6–15 15 2 0.16 (0.00–10.14) ns 12 3 0.09 (0.00–2.83) ns 4 18 4.11 (0.07–236.69) ns ≥ 16 3 0 0.14 (0.00–19.45) ns 6 0 0.08 (0.00–9.95) ns 10 1 0.14 (0.00–10.29) ns a The number of ADE-positive sera in four-year samples collected in 2017.
b The number of ADE-positive sera in six-year samples collected in 2019.
c The statistical analysis was performed by logistic regression.
DENV, dengue virus; AUC, area under curve; OR, odd ratio; C.I., confidence interval; ns: no significance.Table 5. Risk factor analysis of enhancement of DENV-1, DENV-2 and DENV-4 infections triggered by consecutive sera in different populations and AUC groups (n = 73).
In gender-associated analysis, the valid data shown in Table 5 suggested that, the OR of female group in DENV-1 and DENV-2 infection was 0.11 (95% C.I. 0.04–0.31) and 0.41 (95% C.I. 0.19–0.92) in six-year sera when compared with four-year sera (P < 0.001, P = 0.030, respectively). The risk of ADE was negative correlated with the extension of convalescent phase in DENV-1 and DENV-2 infection. However, in DENV-4 infection, the ADE risk both in female and male groups showed 3.48-fold and 7.60-fold rise (OR = 3.48, 95% C.I. 1.52–7.97, P = 0.003; OR = 7.60, 95% C.I. 1.73–33.35, P = 0.007, respectively).
In age-stratified analysis, the risk of ADE after infection of DENV-4 was positively associated with the extension of convalescent phase only in age groups of 31–40 and 51–60 years old (OR = 13, 95% C.I. 2.59–65.20, P = 0.002; OR = 4.09, 95% C.I. 1.01–16.58, P = 0.049, respectively).
The AUC ≤ 5 group conducted a significantly decreasing risk in enhancement of DENV-1 infection (OR = 0.07, 95% C.I. 0.02–0.22, P < 0.001). However, on the contrary, in secondary DENV-4 infection, the high OR implied an obviously elevated risk of ADE (OR = 5.68, 95% C.I. 2.39–13.50, P < 0.001).