 |
|
|
| |
Infection-related and -unrelated malignancies, HIV and the aging population....older age, low CD4, viral load, smoking, HBV, [HCV].....Non-HIV Cancers Predicted to Increase
|
| |
| |
Download the PDF here
"The incidence of IURMs was forecast to gradually increase in the near future, driven by aging of the HIV-positive population. The only exception was in nonsmokers, reflecting lower lung cancer rates and further supporting the need for cessation programmes"
-----------------
The burden of non-AIDS-defining malignancies (NADMs) is now surpassing that of ADMs in HIV-positive people in the USA and Europe [1, 13]. Furthermore, some NADMs occur more frequently than ADMs in HIV-positive people; for example, more cases of anal cancer and Hodgkin lymphoma are diagnosed than invasive cervical cancer [1, 14]. For these reasons, recent research has shifted towards defining malignancies as infection-related malignancies (IRMs) and infection-unrelated malignancies (IURMs) [8].
The incidence of IURMs surpassed that of IRMs from January to June 2009 onwards, and was forcast to continue to increase over the subsequent 5 years.
Lung (n = 55), prostate [28], colorectal [23] and breast [22] cancers were common IURMs.
--------------------------
Older age was the largest contributor to IURM [infection-unrelated malignancies] incidence.....The effects of aging, including reduced immune function, are thought to be accelerated in HIV-positive populations and may also contribute to an increased IURM incidence [22].....As the proportion of the cohort aged over 50, 60 and 70 years increases, the relative impact of characteristics associated with IURM risk is likely to continue to shift, highlighting the need for ongoing monitoring of the changing epidemiology of malignancies.
from Jules: aging in HIV is a greater contributor to not just cancers risk but for other comorbidities including frailty because it appears that although with all aging older people immune decline sets in as people reach around 65 for HIV+ this immune decline appears to be worse not only for those at around 65 but this immune deficit hits all with HIV regardless of age but appears to kick in in higher gear when HIV do intact get older to around 65 because they get hit with both - immune deficit from normal aging & immune deficit from HIV. In HIV immune senescence develops shortly after infection, HIV despite increases in CD4 with ART causes gaps in immunity, ongoing inflammation & immune activation persists forever despite successful ART causing a simmering immune deficiency which can increase risk for comorbidities at any age but kicks into higher gear around 65 as patients start to really age. Frailty, cognitive & neurologic dysfunction & deficit are emerging as older HIV+ age, kidney disease is increasing, heart disease may be increasing the most, bone disease: falls & fractures are a great concern. Exercise, good diet, clean lifestyle [no drug or excessive alcohol use], high CD4 & undetectable viral load are the best preventions. And authors say: "highlighting the need for ongoing monitoring of the changing epidemiology of malignancies".
A higher IRM [infection-related malignancies] incidence was strongly associated with traditional HIV factors such as a higher HIV-VL and a lower CD4 cell count and, to a lesser extent, older age. The incidence of IRMs steadily increased with lower CD4 count category. However, the proportions of IRMs attributable to older age, higher HIV-VL and lower CD4 count were similar to each other in EuroSIDA, as a result of ongoing aging and a low prevalence of uncontrolled HIV infection in this cohort. Those aged ≥ 51 years may have had longer exposure to oncogenic viruses, which could explain the increased IRM incidence.
Current smoking status was associated with a higher IURM incidence, driven by lung cancer, in people aged ≥ 50 years, probably reflecting the lag time between smoking and the onset of consequences. As the proportion of the cohort aged over 50, 60 and 70 years increases, the relative impact of characteristics associated with IURM risk is likely to continue to shift, highlighting the need for ongoing monitoring of the changing epidemiology of malignancies. This is a comparatively new research area in HIV infection and large, well-designed observational studies are crucial to understand both existing and emerging comorbidities in HIV-positive people. A higher IURM incidence was associated with a CD4 cell count < 200 cells/μL in people aged < 50 years only, which appeared to be attributable to a relatively high number of lung cancers in this group. However, numbers were small and further investigation was not possible.
FUTURE INCIDENCE: The incidence of IRMs was forecast to decline over time, probably driven by the low prevalence of advanced and untreated HIV infection, following improvements in treatment efficacy, uptake and adherence. Antiretroviral therapy controls HIV viraemia, leading to CD4 recovery, and reduces the risk of some IRMs, such as KS and NHL [1, 11, 13, 14, 25], but not all (e.g. HL, ICC and anal cancers) [11, 25-27]. This was consistent in all strata except IDUs, who are less likely to be on treatment and have poorer outcomes. The incidence of IURMs was forecast to gradually increase in the near future, driven by aging of the HIV-positive population. The only exception was in nonsmokers, reflecting lower lung cancer rates and further supporting the need for cessation programmes. The malignancy incidence in an American study was projected to increase by approximately 45% between 2010 and 2030, driven by malignant diagnoses in older age groups in the general population [28]. A similar result was found for the UK population [21].
Assuming current trends continue, the crude IRM incidence for those recruited before 2001 was forecast to decline from an incidence of 3.1 (95% CI 1.5, 5.9)/1000 PYFU in July to December 2011 to 2.2 (95% CI 0.9, 4.3)/1000 PYFU after 5 years (Fig. 2a). This was consistent in all strata, with the exception of IDUs, in whom the incidence was stable (Table 4). The forecasted crude IURM incidence increased from 4.1 (95% CI 2.2, 7.2)/1000 PYFU in July to December 2011 to 5.9 (95% CI 3.2, 10.2)/1000 PYFU after 5 years (Fig. 2b), and was consistent in all strata, except in never smokers for whom the IURM incidence was forecast to decrease from 1.7 (95% CI 0.0, 10.6)/1000 PYFU in July to December 2011 to 0.8 (95% CI 0.0, 7.0)/1000 PYFU after 5 years (Table 4). The incidence of IURMs surpassed that of IRMs from January to June 2009 onwards, and was forcast to continue to increase over the subsequent 5 years.
As the proportion of the cohort aged over 50, 60 and 70 years increases, the relative impact of characteristics associated with IURM risk is likely to continue to shift, highlighting the need for ongoing monitoring of the changing epidemiology of malignancies......from Jules: low CD4, detectable viral load, HBV and aging increased IRM risk.....aging, low CD4, smoking increased IURM risk
A higher IRM incidence was strongly associated with traditional HIV factors such as a higher HIV-VL and a lower CD4 cell count and, to a lesser extent, older age. The incidence of IRMs steadily increased with lower CD4 count category. However, the proportions of IRMs attributable to older age, higher HIV-VL and lower CD4 count were similar to each other in EuroSIDA, as a result of ongoing aging and a low prevalence of uncontrolled HIV infection in this cohort. Those aged ≥ 51 years may have had longer exposure to oncogenic viruses, which could explain the increased IRM incidence.
A higher IURM incidence was associated with a CD4 cell count < 200 cells/μL in people aged < 50 years only, which appeared to be attributable to a relatively high number of lung cancers in this group. However, numbers were small and further investigation was not possible.
Factors strongly related to a high IRM incidence were HIV associated. Specifically, HIV-VL > 400 copies/mL was associated with a higher IRM incidence(aRR 1.84; 95% CI 1.39, 2.43; P < 0.01) and accounted for 19% of excess malignancies, relative to those with well-controlled viraemia (HIV-VL ≤ 400 copies/mL). A lower current CD4 cell count was associated with a higher IRM incidence (Table 3), which was similar in those aged < 50 years and ≥ 50 years (Fig. 1c; P interaction = 0.82). CD4 counts < 200 and 200-349 cells/μL accounted for 21% and 11% of excess IRMs, respectively (Table 3), relative to those with a CD4 count of ≥ 500 cells/μL. There was no association between IRM incidence and current smoking (Table 3) in younger or older people (Fig. 1d; P for interaction = 0.31). Prior HBV coinfection was associated with a higher IRM incidence (aRR 1.70; 95% CI 1.24, 2.32), but only 5% of IRMs were attributable to HBV coinfection within the cohort (Table 3).
The IURM incidence was 7.33-fold (95% CI 4.07, 13.21; P < 0.01) and 2.37-fold (95% CI 1.31, 4.27; P < 0.01) higher in those aged ≥ 51 years and aged 41-50 years compared with those aged 36-40 years (Table 3), and explained 56% and 17% of excess IURMs within the cohort, respectively. This corresponds to a twofold increase in IURM incidence per 10 years older age (aRR 2.07; 95% CI 1.84, 2.32). Current smoking was associated with elevated IURM incidence and explained 16% of IURMs overall. Stratifying by age, IURMs were elevated in current smokers relative to nonsmokers in those aged ≥ 50 years (aRR 1.75; 95% CI 1.23, 4.49; P < 0.01; Fig. 1b), but not in those aged < 50 years (aRR 1.12; 95% CI 0.71, 1.77; P = 0.51), although the P-value of the interaction term was nonsignificant (P = 0.32). Current smoking was not associated with IURM incidence after the exclusion of lung cancers. A low current CD4 count was associated with a higher IURM incidence (CD4 count < 200 cells/μL: aRR 1.99; 95% CI 1.26, 3.17; P < 0.01, relative to CD4 count > 500 cells/μL). Despite this, the overall excess of IURMs attributable to a CD4 count < 200 cells/μL (6%) was small. The association between higher IURM incidence and low CD4 count was evident in those aged < 50 years (aRR 2.52; 95% CI 1.40, 4.54; P = 0.01; Fig. 1a), but not in those aged ≥ 50 years (aRR 1.14; 95% CI 0.62, 2.12; P = 0.56; Fig. 1a), although the P-value of the interaction term did not reach statistical significance (P = 0.09). Prior HBV coinfection was also associated with higher IURM incidence (aRR 1.73; 95% CI 1.17, 2.55; P < 0.01), but only 5% were attributable to HBV coinfection within the cohort (Table 3).
Infection-related and -unrelated malignancies, HIV and the aging population. HIV Medicine
Shepherd et al. on behalf of EuroSIDA in EuroCOORD determined the impact of aging on future infection-related malignancies (IRM) and infection-unrelated malignancies (IURM) incidence. Out of 15'648 persons contributing to 95'033 person-years of follow-up, a total of 610 people (3.9%) developed 643 malignancies, of which 60.3% were IRMs and the remaining 39.7% were IURMs.
The most common IRMs were non-Hodgkin lymphoma (n = 116), anal cancer (85), Kaposi sarcoma (62) and Hodgkin lymphoma (43). Lung (n = 55), prostate (28), colorectal (23) and breast (22) cancers were common IURMs.
A higher IRM incidence was strongly associated with traditional HIV factors such as a higher HIV-viral load and a lower CD4 cell count and, to a lesser extent, older age. The incidence of IRMs steadily increased with lower CD4 count category. For IURM, older age was the largest contributor with a twofold higher IURM incidence for a 10-years increase in age. Smoking was associated with IURMs compared with never smokers in people aged ≥ 50 years only, and not with IRMs.
In sum, these findings suggest the need for targeted preventive measures and evaluation of the cost-benefit of screening for IURMs in HIV-infected populations.
----------------------
Infection-related and -unrelated malignancies, HIV and the aging population
L Shepherd,1 AH Borges,2 B Ledergerber,3 P Domingo,4 A Castagna,5 J Rockstroh,6 B Knysz,7 J Tomazic,8 I Karpov,9
O Kirk,2 J Lundgren2 and A Mocroft1 on behalf of EuroSIDA in EuroCOORD*
Abstract
Objectives
HIV-positive people have increased risk of infection-related malignancies (IRMs) and infection-unrelated malignancies (IURMs). The aim of the study was to determine the impact of aging on future IRM and IURM incidence.
Methods
People enrolled in EuroSIDA and followed from the latest of the first visit or 1 January 2001 until the last visit or death were included in the study. Poisson regression was used to investigate the impact of aging on the incidence of IRMs and IURMs, adjusting for demographic, clinical and laboratory confounders. Linear exponential smoothing models forecasted future incidence.
Results
A total of 15 648 people contributed 95 033 person-years of follow-up, of whom 610 developed 643 malignancies [IRMs: 388 (60%); IURMs: 255 (40%)]. After adjustment, a higher IRM incidence was associated with a lower CD4 count [adjusted incidence rate ratio (aIRR) CD4 count < 200 cells/μL: 3.77; 95% confidence interval (CI) 2.59, 5.51; compared with ≥ 500 cells/μL], independent of age, while a CD4 count < 200 cells/μL was associated with IURMs in people aged < 50 years only (aIRR: 2.51; 95% CI 1.40-4.54). Smoking was associated with IURMs (aIRR: 1.75; 95% CI 1.23, 2.49) compared with never smokers in people aged ≥ 50 years only, and not with IRMs. The incidences of both IURMs and IRMs increased with older age. It was projected that the incidence of IRMs would decrease by 29% over a 5-year period from 3.1 (95% CI 1.5-5.9) per 1000 person-years in 2011, whereas the IURM incidence would increase by 44% from 4.1 (95% CI 2.2-7.2) per 1000 person-years over the same period.
Conclusions
Demographic and HIV-related risk factors for IURMs (aging and smoking) and IRMs (immunodeficiency and ongoing viral replication) differ markedly and the contribution from IURMs relative to IRMs will continue to increase as a result of aging of the HIV-infected population, high smoking and lung cancer prevalence and a low prevalence of untreated HIV infection. These findings suggest the need for targeted preventive measures and evaluation of the cost-benefit of screening for IURMs in HIV-infected populations.
Introduction
HIV-positive people are at increased risk of many malignancies compared with the general population [1, 2]; however, the exact mechanisms are poorly understood [3]. The increased risk could be attributable to a high prevalence of traditional cancer risk factors such as smoking [4] and alcohol use [5]. However, coinfection with other pro-oncogenic viruses [6], immunodeficiency [2, 7], activated inflammation and coagulation [8], a potential direct pro-oncogenic effect of HIV and combination antiretroviral therapy (cART) toxicity [3] may also contribute to the increased risk. The introduction of cART in 1996 led to restored immune function, a reduced incidence of AIDS-defining malignancies (ADMs) [9-11] and increased survival [12]. As a result, the burden of cancers traditionally associated with older age is becoming increasingly important in the HIV-positive population. There is a growing need to address the changing epidemiology of cancers as the population ages. This has been done in several studies in the USA [1, 13]; however, research in the European population is needed.
The burden of non-AIDS-defining malignancies (NADMs) is now surpassing that of ADMs in HIV-positive people in the USA and Europe [1, 13]. Furthermore, some NADMs occur more frequently than ADMs in HIV-positive people; for example, more cases of anal cancer and Hodgkin lymphoma are diagnosed than invasive cervical cancer [1, 14]. For these reasons, recent research has shifted towards defining malignancies as infection-related malignancies (IRMs) and infection-unrelated malignancies (IURMs) [8].
The changing epidemiology of malignancies and the impact of aging on cancer incidence need to be better characterized. The aim of this study was to investigate the impact of aging in HIV-positive people on the incidence of IRMs and IURMs within EuroSIDA, a large European cohort of HIV-positive people with a long follow-up period, and to estimate the likely impact of IRMs and IURMs in the HIV-positive population in the next 5 years for future health care planning, treatment and prevention.
Results
Characteristics
A total of 15 648 persons contributed 95 033 PYFU with a median follow-up time of 6.0 [interquartile range (IQR) 2.5, 10.7] years. Baseline characteristics according to malignancy status are shown in Table 1. At baseline, 16.0% of patients were aged ≥ 50 years, 72.6% of the population were male, 88.3% were of white ethnic origin and 38.7% of men were infected with HIV through homosexual exposure. Approximately one-third of patients were current smokers and one-third had never smoked. The median CD4 count at baseline was 410 (IQR 265, 588) cells/μL, with 14.8% having a CD4 count ≤ 200 cells/μL, and HIV-VL was 123 (IQR < 50, 5200) HIV-1 RNA copies/mL, with 54.5% having HIV-VL ≤ 400 copies/mL. Prior AIDS-defining and non-AIDS-defining events were present in 24.4% and 1.9% of patients, respectively. Coinfection with HCV and HBV was prevalent in 23.1% and 5.5% of people, respectively.
A total of 610 people developed 643 malignancies, of which 60.3% were IRMs and the remaining 39.7% were IURMs (Table 2). The most common IRMs were NHL (n = 116), anal cancer (85), KS (62) and HL (43). Lung (n = 55), prostate [28], colorectal [23] and breast [22] cancers were common IURMs. At diagnosis, those who developed IURMs relative to those who developed IRMs were older [median age 54 (IQR 46, 61) years for IURMs vs. 46 (IQR 39, 52) years for IRMs] and had a higher CD4 count [median 446 (IQR 295, 608) cells/μL vs. 342 (IQR 182, 546) years, respectively] and a lower HIV-VL [median < 50 (IQR < 50, 86) copies/mL vs. 61 (IQR < 50, 20 002) copies/mL, respectively]. A lower proportion of those with IURMs had prior HCV coinfection and the proportions of those with prior HBV coinfection were similar. Proportions of ever smokers with IRMs (45%) and IURMS (46%) were similar. The majority of cancers diagnosed in those younger than 50 years (n = 353 cancers) were IRMs (75%), with EBV related malignancies accounting for 31% of IRMs and HPV related malignancies accounting for 25%. In those aged ≥ 50 years (290 cancers), more than half had IURMs (57%), with lung cancer accounting for 13% of IURMs and prostate cancer accounting for 10%.
Adjusted incidence of IRM and IURM
In adjusted models, those aged ≥ 50 years had a 1.62 (95% CI 1.14, 2.30) times higher IRM incidence compared with those aged 36-40 years (Table 3), corresponding to a 17% higher incidence per 10 years older age (adjusted rate ratio (aRR) 1.17; 95% CI 1.05, 1.32). The percentage of excess IRMs attributable to being aged ≥ 51 years compared with 36-40 years was 12%. Factors strongly related to a high IRM incidence were HIV associated. Specifically, HIV-VL > 400 copies/mL was associated with a higher IRM incidence (aRR 1.84; 95% CI 1.39, 2.43; P < 0.01) and accounted for 19% of excess malignancies, relative to those with well-controlled viraemia (HIV-VL ≤ 400 copies/mL). A lower current CD4 cell count was associated with a higher IRM incidence (Table 3), which was similar in those aged < 50 years and ≥ 50 years (Fig. 1c; P interaction = 0.82). CD4 counts < 200 and 200-349 cells/μL accounted for 21% and 11% of excess IRMs, respectively (Table 3), relative to those with a CD4 count of ≥ 500 cells/μL. There was no association between IRM incidence and current smoking (Table 3) in younger or older people (Fig. 1d; P for interaction = 0.31). Prior HBV coinfection was associated with a higher IRM incidence (aRR 1.70; 95% CI 1.24, 2.32), but only 5% of IRMs were attributable to HBV coinfection within the cohort (Table 3).
The IURM incidence was 7.33-fold (95% CI 4.07, 13.21; P < 0.01) and 2.37-fold (95% CI 1.31, 4.27; P < 0.01) higher in those aged ≥ 51 years and aged 41-50 years compared with those aged 36-40 years (Table 3), and explained 56% and 17% of excess IURMs within the cohort, respectively. This corresponds to a twofold increase in IURM incidence per 10 years older age (aRR 2.07; 95% CI 1.84, 2.32). Current smoking was associated with elevated IURM incidence and explained 16% of IURMs overall. Stratifying by age, IURMs were elevated in current smokers relative to nonsmokers in those aged ≥ 50 years (aRR 1.75; 95% CI 1.23, 4.49; P < 0.01; Fig. 1b), but not in those aged < 50 years (aRR 1.12; 95% CI 0.71, 1.77; P = 0.51), although the P-value of the interaction term was nonsignificant (P = 0.32). Current smoking was not associated with IURM incidence after the exclusion of lung cancers. A low current CD4 count was associated with a higher IURM incidence (CD4 count < 200 cells/μL: aRR 1.99; 95% CI 1.26, 3.17; P < 0.01, relative to CD4 count > 500 cells/μL). Despite this, the overall excess of IURMs attributable to a CD4 count < 200 cells/μL (6%) was small. The association between higher IURM incidence and low CD4 count was evident in those aged < 50 years (aRR 2.52; 95% CI 1.40, 4.54; P = 0.01; Fig. 1a), but not in those aged ≥ 50 years (aRR 1.14; 95% CI 0.62, 2.12; P = 0.56; Fig. 1a), although the P-value of the interaction term did not reach statistical significance (P = 0.09). Prior HBV coinfection was also associated with higher IURM incidence (aRR 1.73; 95% CI 1.17, 2.55; P < 0.01), but only 5% were attributable to HBV coinfection within the cohort (Table 3).
Future incidence
There were 6111 people enrolled prior to 1 January 2001, contributing 54 030 PYFU [median of 11.1 (IQR 5.8-11.3) PYFU per person] who developed 243 IRMs and 161 IURMs during follow-up. At baseline, 82% of patients were aged < 50 years, 78% were male and 41% had a CD4 count < 350 cells/μL. Forty-seven per cent were homosexual, 25% were heterosexual and 21% were IDUs. Twenty-six per cent were smokers at baseline, 21% were nonsmokers at baseline and 52% had unknown smoking status.
Assuming current trends continue, the crude IRM incidence for those recruited before 2001 was forecast to decline from an incidence of 3.1 (95% CI 1.5, 5.9)/1000 PYFU in July to December 2011 to 2.2 (95% CI 0.9, 4.3)/1000 PYFU after 5 years (Fig. 2a). This was consistent in all strata, with the exception of IDUs, in whom the incidence was stable (Table 4). The forecasted crude IURM incidence increased from 4.1 (95% CI 2.2, 7.2)/1000 PYFU in July to December 2011 to 5.9 (95% CI 3.2, 10.2)/1000 PYFU after 5 years (Fig. 2b), and was consistent in all strata, except in never smokers for whom the IURM incidence was forecast to decrease from 1.7 (95% CI 0.0, 10.6)/1000 PYFU in July to December 2011 to 0.8 (95% CI 0.0, 7.0)/1000 PYFU after 5 years (Table 4). The incidence of IURMs surpassed that of IRMs from January to June 2009 onwards, and was forcast to continue to increase over the subsequent 5 years.
Discussion
Demographic and HIV-related risk factors for IURMs and IRMs differ markedly. The incidence of IURMs has exceeded that of IRMs since January-June 2009 (in those enrolled in EuroSIDA prior to 2001) and the contribution from IURMs is forecast to increase over the subsequent 5 years as a result of aging of the HIV-positive population, a high smoking prevalence [4], and a low prevalence of untreated and advanced HIV infection. These findings suggest the need to develop targeted preventive measures and evaluate the cost-benefit of screening for IURMs in HIV-positive populations. Ours is one of the few large prospective studies in a European population with free access to care. The majority of the research to date has focussed largely on patients within the USA.
A higher IRM incidence was strongly associated with traditional HIV factors such as a higher HIV-VL and a lower CD4 cell count and, to a lesser extent, older age. The incidence of IRMs steadily increased with lower CD4 count category. However, the proportions of IRMs attributable to older age, higher HIV-VL and lower CD4 count were similar to each other in EuroSIDA, as a result of ongoing aging and a low prevalence of uncontrolled HIV infection in this cohort. Those aged ≥ 51 years may have had longer exposure to oncogenic viruses, which could explain the increased IRM incidence.
Older age was the largest contributor to IURM incidence. Our finding of a twofold higher IURM incidence for a 10-year increase in age is similar to findings of the The Strategies for Management of Antiretroviral Therapy (SMART) study [24] and to data published online by the European Cancer Observatory, which showed a 1.9-fold increase in the incidence of all malignancies in the general population [23]. The effects of aging, including reduced immune function, are thought to be accelerated in HIV-positive populations and may also contribute to an increased IURM incidence [22].
Current smoking status was associated with a higher IURM incidence, driven by lung cancer, in people aged ≥ 50 years, probably reflecting the lag time between smoking and the onset of consequences. As the proportion of the cohort aged over 50, 60 and 70 years increases, the relative impact of characteristics associated with IURM risk is likely to continue to shift, highlighting the need for ongoing monitoring of the changing epidemiology of malignancies. This is a comparatively new research area in HIV infection and large, well-designed observational studies are crucial to understand both existing and emerging comorbidities in HIV-positive people. A higher IURM incidence was associated with a CD4 cell count < 200 cells/μL in people aged < 50 years only, which appeared to be attributable to a relatively high number of lung cancers in this group. However, numbers were small and further investigation was not possible.
The incidence of IRMs was forecast to decline over time, probably driven by the low prevalence of advanced and untreated HIV infection, following improvements in treatment efficacy, uptake and adherence. Antiretroviral therapy controls HIV viraemia, leading to CD4 recovery, and reduces the risk of some IRMs, such as KS and NHL [1, 11, 13, 14, 25], but not all (e.g. HL, ICC and anal cancers) [11, 25-27]. This was consistent in all strata except IDUs, who are less likely to be on treatment and have poorer outcomes. The incidence of IURMs was forecast to gradually increase in the near future, driven by aging of the HIV-positive population. The only exception was in nonsmokers, reflecting lower lung cancer rates and further supporting the need for cessation programmes. The malignancy incidence in an American study was projected to increase by approximately 45% between 2010 and 2030, driven by malignant diagnoses in older age groups in the general population [28]. A similar result was found for the UK population [21].
This study has a number of limitations. EuroSIDA has a relatively large number of prospectively collected source validated malignancies, but, despite this, the frequencies of individual malignancies were small and could not be investigated or forecasted individually. Furthermore, this is an observational study and residual confounding cannot be ruled out. The predicted small increase in IURMs may be influenced by changes in and uptake of cancer screening practices, such as cervical, breast, colorectal and prostate screening, in clinics. However, no significant changes in the incidence of these cancers occurred in our study, although numbers were small. EuroSIDA does not routinely collect data on screening practices and therefore we cannot investigate the role of screening further. However, recent survey data on HIV management in active EuroSIDA sites found that screening rates for cervical and anorectal cancers were low [29]. Linear exponential smoothing models are a simple method of forecasting which assumes the continuation of previous population trends. A limited amount of historical data was available for forecasting, contributing to uncertainty around forecasts. HIV-specific population projections are not currently available, which prevents the use of more advanced methodologies, such as age-period-cohort models.
As the HIV-infected population ages, the incidence of IURMs is expected to increase as a result of aging of the HIV-infected population and the high prevalence of lung cancer caused by smoking, the effects of which generally manifest with longer term exposure and thus at older ages. Conversely, IRM incidence is expected to decline as a result of the low prevalence of advanced and untreated HIV infection and severe immunodeficiency. IURMs should therefore be a priority in the coming years as higher proportions of HIV-positive people live past 50, 60 and 70 years. Studies evaluating the cost-benefit of screening programmes for HIV-positive people and targeted preventive interventions, such as cessation programmes for smoking and alcohol use and vaccinations for oncogenic viruses, should be considered to reduce the burden of avoidable cancers in the long term.
|
| |
|
|
|
|
|
