Secondary Logo

Journal Logo

CLINICAL SCIENCE

Impact of HIV infection on baseline characteristics and survival of women with breast cancer

Brandão, Marianaa,b,c; Bruzzone, Marcod; Franzoi, Maria-Alicea; De Angelis, Claudiaa; Eiger, Daniela; Caparica, Rafaela; Piccart-Gebhart, Martinea; Buisseret, Laurencea; Ceppi, Marcellod; Dauby, Nicolase,f,g; Carrilho, Carlah,i; Lunet, Nunob,c; de Azambuja, Evandroa; Lambertini, Matteod,j

Author Information
doi: 10.1097/QAD.0000000000002810

Abstract

Introduction

The number of women living with HIV (WLWH) is increasing globally thanks to the use of effective antiretroviral treatment [1]. As these women become older, their risk of developing cancer increases, including non-AIDS-defining malignancies, such as breast cancer [2]. Breast cancer is still the leading cause of cancer-related death among women worldwide [3] and its burden is increasing among specific subgroups of women, such as those with HIV/AIDS [4]. Globally, WLWH constitute less than 1% of patients with breast cancer, but in high endemic regions, such as Southern Africa, WLWH represent 26% of breast cancers diagnosed among women less than 50 years. Indeed, multiple reports from the United States of America (USA) [5] and from sub-Saharan Africa [6,7] reveal that WLWH with breast cancer tend to be younger than HIV-negative patients, which is explained by differences in age distributions of the underlying HIV-positive and general populations [7,8]. Nonetheless, data are discordant concerning tumor stage at diagnosis, as some series showed that WLWH tend to present with later stage disease compared with HIV-negative women [9,10], whereas others did not show such a difference [7,11].

Breast cancer is a heterogeneous disease constituting various subtypes, with different prognosis and treatment approaches [12]. However, data on distribution of classic subtypes [estrogen receptor (ER)-positive/HER2-negative, triple-negative (ER-negative/HER2-negative) and HER2-positive] and of surrogate intrinsic subtypes (luminal A-like, luminal B-like, HER2-enriched and basal-like) [13] between WLWH and HIV-negative patients is scarce.

Additionally, although some reports showed that WLWH with breast cancer have an increased mortality compared with HIV-negative women, even after adjusting for stage and age [5,14], other studies did not demonstrate worse survival outcomes among WLWH [15–17]. These discrepancies may be because of the low number of WLWH with breast cancer and/or short follow-up time in most publications. Another reason may be the low accessibility to breast cancer diagnosis and treatment in some parts of the world, leading to a high proportion of late-stage disease and poor survival both in WLWH and HIV-negative patients.

As the number of WLWH with breast cancer is increasing worldwide [4], a comprehensive understanding of their characteristics at diagnosis and survival is paramount for better tailoring their cancer care. In this systematic review and meta-analysis, we aimed to compare clinicopathological characteristics at diagnosis and overall survival between WLWH and HIV-negative women with breast cancer.

Methods

This systematic review and meta-analysis was conducted and reported according to the MOOSE guidelines [18] and registered in the PROSPERO database (CRD42019141887).

Search strategy and selection criteria

We performed a systematic search of the literature without language restrictions from inception of each database up to 1 January 2020. We combined keywords like ‘breast cancer’, ‘breast’ and ‘HIV’ in the search strategy, which was run in MEDLINE, Scopus, ISI Web of Knowledge, LILACS and SciELO. In addition, abstracts from major conferences from 2016 to January 2020 were manually searched, in order to include unpublished studies (Supplementary Methods, https://links.lww.com/QAD/B974). We screened the reference lists of included studies and review articles to identify additional publications.

Cross-sectional or cohort studies were eligible if providing information on breast cancer patients’ baseline characteristics (stage and/or subtypes) and/or on overall survival, separately for WLWH and HIV-negative women, or measures of association for their comparison. Case--control studies, reports only including WLWH and studies with less than 10 WLWH with breast cancer were excluded. When more than one publication reported the same endpoint for the same/potentially overlapping study population, we used data preferentially from the full publication (vs. abstract), from the publication with the largest sample size or data corresponding to the longest follow-up period, as applicable.

WLWH vs. HIV-negative women with breast cancer were compared regarding the likelihood of being diagnosed with: locally advanced/metastatic breast cancer (tumour, node, metastasis stage III/IV); each classic subtype; each surrogate intrinsic subtype; and regarding: overall survival, defined as time from breast cancer diagnosis until death by any cause. Triple-negative and basal-like subtypes were analyzed together, as these subtypes are equivalent when assessed by immunohistochemistry/in-situ hybridization. The likelihood of being diagnosed with each stage at diagnosis (0/I, II, III and IV) and the ER status of the tumor (positive vs. negative) were also assessed. Subgroup analyses of each endpoint were performed according to region of the world, depending on data availability.

Data extraction

Two reviewers independently evaluated the screened titles/abstracts; in case of disagreement, a third author resolved it. After this initial screening, full articles were reviewed and data extracted by three authors, namely regarding: study characteristics (design, patient selection); race/ethnicity; CD4+ cell count and antiretroviral treatment at diagnosis of breast cancer; type of anticancer treatment received; age at diagnosis of breast cancer; adjusted odds ratio (OR) for the likelihood of being diagnosed with stage III/IV or, when adjusted ORs were not available, the number of patients with the combined III/IV stages and with each stage (0/I, II, III and IV); adjusted ORs or the number of patients with each classic and surrogate subtypes and estrogen receptor status positivity; and comparison of overall survival between WLWH vs. HIV-negative women, provided as hazard ratios and confidence intervals (CI). Both adjusted and unadjusted hazard ratios were collected.

Quality assessment and publication bias

A quality assessment was performed based on study design, patient selection, definition of stage according to the TNM classification, type of estrogen receptor/HER2 status assessment (if according to ASCO/CAP guidelines) [19–21], median follow-up, proportion of loss to follow-up and adjustment of survival analysis to confounding factors. In each category, risk was classified from ‘low’ to ‘high’. Egger's test was used for assessing small studies effects/publication bias [22].

Data synthesis and statistical analysis

Aggregated data from each study were used for the quantitative synthesis of results. For categorical variables (stage and subtypes), when adjusted ORs were not provided, ORs were computed from the frequencies provided in the manuscript and then summary risk estimates [pooled OR (pOR)] with 95% CI were calculated using random-effects models [23]. Overall survival data were pooled as hazard ratios and 95% CI and modeled using a random-effects model, separately for adjusted and unadjusted hazard ratios. All analyses were performed in logarithmic scale. Heterogeneity was assessed by visual inspection of forest plots and calculation of the I2 statistic. Leave-one-out sensitivity analyses and sensitivity analyses excluding poor-quality studies (with ‘high-risk’) were performed for each endpoint, whenever applicable. All analyses were run on Stata/SE 13.1 (StataCorp LP, Texas, USA).

Results

Characteristics of included studies

The systematic search of literature provided 10 174 unique citations, of which 62 were reviewed as full text (Fig. 1). Among them, 20 publications reporting on 18 studies [including 3174 WLWH (0.13%) and 2 394 598 HIV-negative women (99.87%)] were included: 15 assessed the likelihood of stage III/IV [6,7,9–11,14,17,24–31], 10 described each stage at diagnosis [6,7,9,14,17,24,25,27,29,31], nine reported on classic subtypes and estrogen receptor status [7,9,14,17,25,26,29,31,32], four described surrogate subtypes [17,26,29,31] and seven assessed overall survival [14–17,27,33,34] (Table 1).

F1
Fig. 1:
Study-selection flow chart.
Table 1 - Characteristics of included studies.
Author (year) [reference no.] Location BC only? Study design (study name) Inclusion dates WLWH (%) WLWH: % ARTa WLWH: CD4+ cell counta WLWH: race/ethnicity HIV-neg women (%)b HIV-neg women: race/ethnicity Variables provided
 Biggar et al. (2005) [34] USA/New York No Retrospective cohort 1996--2000 28 (0.18%) NR NR NR 15 225 (99.8%) NR Survival
 Shiels et al. (2015) [10] USA No Retrospective cohort (HACM study) 1996--2010 429 (0.05%) NR NR Overallc: 52% black; 47% white; 1% other 921 171 (99.95%) Overallc: 11% black; 85% white; 4% other Stage (III/IV)d
 Coghill et al. (2015) [33] 314 (0.08%) Overallc: 45% black; 39% WNH; Hispanic 15% 386 041 (99.92%) Overallc: 12% black; 77% WNH; 8% Hispanic Survival
 Presti et al. (2017) [29] USA/Baltimore Yes Retrospective, cross-sectional 2004--2014 43 (1.4%) NR NR 91% black, 9% other 3012 (98.6%) 81% black, 19% other Stage (III/IV and detailedc), classic (TNBC only) and surrogate subtypes, ER status
 Coghill et al. (2019) [27] USA No Retrospective cohort (NCDB) 2004--2014 1197 (0.08%) NR NR NR 1 448 757 (99.92%) NR Stage (III/IV and detailedf), survival
 Sseggwanyi et al. (2011) [24] Uganda/Kampala Yes Prospective cohort 2011 22 (35.5%) 36% <250 cells/μl: n = 3 (13.6%) 100% black 40 (64.5%) 100% black Stage (III/IV and detailedf)
 Coghill et al. (2013) [16] Uganda/Kyadondo County No Retrospective cohort 2003--2010 24 (10.9%) NR NR NR 196 (89.1%) NR Survival
 Cubasch et al. (2013) [7] South Africa/Johannesburg Yes Retrospective cohort 2006--2012 151 (19.7%) 17%e Median count: 316 cells/μl (range: 7–1203; IQR 196–487); count <200 cells/μl: n = 37 (26%) 100% black 614 (80.3%) 100% black Stage (IIII/IV and detailedf), classic and surrogate subtypes, ER status
 Cubasch et al. (2018) [15] 2009--2011 88 (17.6%) NR NR 411 (82.4%) Survival
 Traore et al. (2015) [28] Guinea/Conakry Yes Retrospective, cross-sectional 2007--2012 14 (5.0%) NR Count <350 cells/μl: n = 2 (33%) NR 264 (95.0%) NR Stage (III/IV)
 Phakathi et al. (2016) [6] South Africa/Pretoria Yes Prospective cohort 2009--2014 31 (19.4%) 55%e Mean count: 406 cells//μl (SD 206); Count <200 cells//μl: n = 5 (16%) 81% black; 6% mixed; 13% white 129 (80.6%) 44% black; 8% mixed; 47% white; 1% Indian Stage (III/IV and detailedf)
 Langenhoven et al. (2016) [9] South Africa/Cape Town Yes Retrospective, cross-sectional 2010--2011 31 (6.9%) 39%e Mean count: 477 cells/μl (SD 160) 58% black; 42% mixed 420 (93.1%) 10% black; 65% mixed; 25% white Stage (III/IV and detailedf), classic subtypes, ER status
 Ngidi et al. (2017) [25] South Africa/Durban Yes Retrospective, cross-sectional 2012--2015 21 (32.3%) 95% Mean count: 435 cells/μl (SD 248; range 80–945) NR 44 (67.7%) NR Stage (III/IV and detailedf), classic subtypes (HER2+ only), ER status
 McKenzie et al. (2018) [11] Nigeria, Uganda, Zambia, Namibia and South Africa Yes Prospective cohort (ABC-DO study) 2014--2016 178 (9.9%) NR NR NR 1617 (90.1%) NR Stage (III/IV)
 Van Zyl et al. (2018) [26] South Africa/Pretoria Yes Retrospective, cross-sectional 2013--2017 49 (38.0%) NR Count <200 cells/μl: n = 10 (26%) NR 80 (62.0%) NR Stage (III/IV), classic and surrogate subtypes, ER status
 Sadigh et al. (2019) [14] Botswana Yes Prospective cohort (Thabatse Cancer Cohort) 2010--2018 151 (31.6%) 85% Median count: 457 cells/μl (IQR 304–618); Count <200 cells/μl: n = 17 (11%) NR 327 (68.4%) NR Stage (III/IV and detailedf), classic subtypes (TNBC only), ER status, survival
 Brandao et al. (2019) [17] Mozambique Yes Prospective cohort (Moza-BC Cohort) 2015--2017 52 (25.5%) 74%e Median count: 439 cells/μl (range 43–1104); Count <200 cells/μl: 3 = 5 (10.9%) 98% black; 2% other 152 (74.5%) 98% black; 2% other Stage (III/IV and detailedf), classic and surrogate subtypes, ER status, survival
 Phakathi et al. (2019) [31] South Africa/Johannesburg Yes Retrospective, cross-sectional 2015--2017 226 (22.2%) 72% Median count: 477 cells/μl (IQR 287–670) 97% black; 0.4% white; 2% mixed; 0% Asian 790 (77.8%) 81% black; 10% white; 7% MIXED; 2% Asian Stage (III/IV and detailedf), classic and surrogate subtypes, ER status
 Bhatia et al. (2019) [32] Botswana/Gaborone Yes Retrospective, cross-sectional 2011--2015 49 (43.0%) NR Median count: 396 cells/μl (IQR 209–525.6; count <250 cells/μl: n = 13 (30%) NR 65 (57.0%) NR Classic subtypes, ER status
 Ayeni et al. (2019) [30] South Africa Yes Prospective cohort (SABCHO) 2016--2018 478 (22.0%) NR NR Overall cohortg: 77% black; 11% Asian; 7% white; 5% mixed 1695 (78.0%) Overall cohortg: 77% black; 11% Asian; 7% white; 5% mixed Stage (III/IV)
ABC-DO, African Breast Cancer – Disparities in Outcomes study; ART, antiretroviral treatment; BC, breast cancer; ER, estrogen receptor; HACM, HIV/AIDS Cancer Match study; HIV-neg, HIV-negative; NCDB, National Cancer Database; NR, not reported; SABCHO, South African Breast Cancer and HIV Outcomes Study; TNBC, triple negative breast cancer; USA, United States of America; WLWH, women living with HIV; WNH, white non-Hispanic.
aCD4+ cells count and antiretroviral treatment uptake was assessed at the time of diagnosis of breast cancer.
bOnly considering breast cancer patients.
cIncluding all tumor types (not restricted to breast cancer).
dIn this study, breast tumors were classified into ‘regional stage’, which we considered as stage III and into ‘distant stage’, which we considered as stage IV.
ePatients not taking antiretroviral treatment at the time of breast cancer diagnosis were referred to a HIV clinic in order to start antiretroviral treatment.
fDetailed stage meaning: stage 0/I vs. II vs. III vs. IV at diagnosis.
gIncluding both HIV-positive and HIV-negative patients.

Four of the studies were from North America (all from the USA) [10,27,29,33,34] and 14 from sub-Saharan Africa [6,7,9,11,14–17,24–26,28,30,31]. Patient inclusion dates spanned from 1996 to 2018, reflecting the timing of introduction of highly active antiretroviral treatment. Still, there was a wide variation on CD4+ cell counts and on the proportion of WLWH under antiretroviral treatment at the time of cancer diagnosis (17--95%).

Studies’ quality assessment is provided in the appendix (Supplemental Table 1, https://links.lww.com/QAD/B974). Of note, two studies from sub-Saharan Africa had a loss to follow-up of at least 40% and were considered to be of poor-quality for the overall survival analysis [15,16]. There was a suggestion of publication bias for estrogen receptor status (Egger's test: P = 0.047), but not for the other endpoints (Supplemental Figures 1–11, https://links.lww.com/QAD/B974).

Baseline characteristics

WLWH had an increased odds of presenting with stage III/IV disease compared with HIV-negative patients both in studies from North America (pOR 1.76; 95% CI, 1.58–1.95; I2 = 0%) and from sub-Saharan Africa (pOR 1.23; 95% CI, 1.06–1.42; I2 = 3.1%) (Fig. 2). Analyses of the likelihood of presenting with each stage at diagnosis (0/I, II, III and IV) are provided in the Supplemental Figures 12–15 (https://links.lww.com/QAD/B974).

F2
Fig. 2:
Meta-analysis of the odds of presenting with locally advanced/metastatic stage (III/IV) at diagnosis.

Regarding the distribution of classic subtypes, WLWH from sub-Saharan Africa had smaller odds of presenting with the estrogen receptor-positive/HER2-negative subtype (pOR 0.81; 95% CI, 0.66–0.99; I2 = 0%; Fig. 3). No differences regarding the distribution of the other classic subtypes, surrogate intrinsic subtypes or estrogen receptor-positive status were observed (Supplemental Figures 16–19, https://links.lww.com/QAD/B974). Leave-one-out sensitivity analyses showed similar findings for all of these endpoints (Supplemental Tables 2–15, https://links.lww.com/QAD/B974).

F3
Fig. 3:
Meta-analysis of the odds of presenting with each of the classic subtypes.

Overall survival

WLWH from North America had an increased risk of dying compared with HIV-negative women, with a pooled adjusted hazard ratio for overall survival of 2.45 (95% CI, 1.11–5.41; I2 = 98.0%) (Fig. 4). Although all studies from North America presented a hazard ratio superior to 1.0, heterogeneity was high, possibly because of the high hazard ratio of 4.62 on the report by Coghill et al.[33] in 2015. When removing this study in the leave-one-out sensitivity analysis, the pooled adjusted hazard ratio was 1.77 (95% CI, 1.62–1.93), with no evidence of heterogeneity (I2 = 0%; Supplemental Table 16, https://links.lww.com/QAD/B974).

F4
Fig. 4:
Meta-analysis of adjusted overall survival.

WLWH from sub-Saharan Africa presented a pooled adjusted hazard ratio for overall survival of 1.58 (95% CI, 1.25–1.98; I2 = 0) compared with HIV-negative women (Fig. 4 and Supplemental Table 17, https://links.lww.com/QAD/B974). Studies reporting unadjusted hazard ratios for survival were all from sub-Saharan Africa and provided a pooled hazard ratio of 1.43 (95% CI, 1.06–1.92) (Supplemental Figure 20, https://links.lww.com/QAD/B974 and Table 18, https://links.lww.com/QAD/B974). In the sensitivity analysis, only including studies with ‘low/medium-risk’ for survival follow-up, the respective pooled hazard ratios were similar (Supplemental Table 19, https://links.lww.com/QAD/B974).

Discussion

This is the first meta-analysis to quantify the differences in baseline clinicopathological characteristics and survival outcomes between WLWH and HIV-negative women with breast cancer. Our findings suggest that WLWH are diagnosed with breast cancer at a more advanced stage and have a lower likelihood of presenting with the estrogen receptor-positive/HER2-negative subtype. In addition, even after adjusting for known prognostic factors, WLWH have worse overall survival compared with HIV-negative women, both in sub-Saharan Africa and in North America.

It is now understood that the classic risk factors (parity, age at first birth, breastfeeding and age at menarche/menopause) have different influences on the risk of developing each of the breast cancer subtypes, being mostly associated with the occurrence of estrogen receptor-positive/HER2-negative tumors [35,36]. However, no study has yet evaluated the association of HIV infection with the risk of developing each subtype. This could be interesting, as it is well known that they present distinct immunogenic profiles: estrogen receptor-positive/HER2-negative tumors have a lower proportion of tumor infiltrating lymphocytes and a lower tumor mutational burden compared with HER2-positive or triple-negative tumors [37,38]. Additionally, higher levels of tumor-infiltrating lymphocytes are associated with favorable outcomes among patients with early stage HER2-positive and triple-negative subtypes, but not with estrogen receptor-positive/HER2-negative tumors [37]. There was a wide variation on the proportion of WLWH under antiretroviral treatment at the time of cancer diagnosis in the included studies but it is known that even after antiretroviral treatment initiation, persisting immune alterations can affect local immune responses. Therefore, differences in the distribution of subtypes between WLWH vs. immunocompetent patients (HIV-negative) would be expected. Yet, the only difference observed was regarding the estrogen receptor-positive/HER2-negative subtype, which was less likely to be diagnosed among WLWH. Given that all included ORs were unadjusted to other clinicopathological characteristics, this difference could be partly explained by the younger age at diagnosis of WLWH (Supplemental Table 20, https://links.lww.com/QAD/B974) [5–7], as younger patients have a lower prevalence of estrogen receptor-positive/HER2-negative tumors compared with older patients [39]. On the other hand, WLWH under antiretroviral treatment may have lower blood estradiol levels and prolonged periods of amenorrhea [40,41] and this might also be a reason for their smaller likelihood of presenting with estrogen receptor-positive/HER2-negative tumors. This intriguing finding could also potentially justify why studies from the USA (in which estrogen receptor-positive/HER2-negative tumors represent more than 70% of cases [39]) report a lower incidence of breast cancer among WLWH compared with the general population [42], whereas studies from sub-Saharan Africa (in which the proportion of estrogen receptor-positive/HER2-negative tumors is much lower [43]) suggest that HIV infection does not increase nor decrease the likelihood of developing breast cancer [6,7]. Despite the fact that heterogeneity was low for the estrogen receptor-positive/HER2-negative meta-analysis, part of these studies did not follow ASCO/CAP guidelines on estrogen receptor/HER2 assessment [19–21], thus further data are needed in order to confirm these findings.

Our results also suggest that when WLWH develop breast cancer, they present with a more advanced disease. One argument is that this may be because of a lower access to healthcare services/surveillance among WLWH [10,44]. As none of the North American studies provided the adjusted ORs specifically for the likelihood of presenting with stage III/IV vs. stage 0/II, our pooled OR was computed using unadjusted ORs, which is a limitation. However, in the large study by Coghill et al. (2019), after adjustment for age, race/ethnicity, year of cancer diagnosis, median household income, type of healthcare facility and individual health insurance, the association between HIV infection and stage IV breast cancer (vs. stage I) was significant, with a OR of 2.06 (95% CI, 1.70–2.50) [27]. Interestingly, in subgroup analysis, this association was more pronounced among women with private insurance (OR 2.22) or Medicare (OR 2.27), expected to have good access to cancer care, whereas the OR was 0.77 among women with Medicaid. Likewise, the other large study from the USA showed that WLWH had higher odds of developing stage IV (vs. stage 0--II) breast cancer, with an OR of 1.99 (95% CI, 1.40–2.83), adjusted for age, race/ethnicity and year of cancer diagnosis [10].

In studies from sub-Saharan Africa, most of which were conducted in public hospitals from low/middle-income countries with low or unavailable access to breast cancer screening, the proportion of stage III/IV disease was high among HIV-negative patients, as expected [45], but it was slightly higher among WLWH. Unfortunately, only two of those studies provided adjusted ORs for this comparison, taking into account age, race/ethnicity, education, among others, and those ORs were nonsignificant [11,30]. Thus, it is possible that among this patient population, with an already high baseline proportion of stage III/IV, that the presence of HIV infection does not significantly increase the likelihood of advanced/metastatic breast cancer if adjusted to other clinicopathological characteristics. However, more detailed data from sub-Saharan African studies would be needed to fully clarify this question.

WLWH had worse overall survival, even when adjusting for other factors, such as age, stage, ethnicity, and in some studies, breast cancer subtype, income and treatments. It may be argued that this increased mortality may be because of nonbreast cancer-related deaths (e.g. HIV-related). Nonetheless, in the Coghill et al.[33] study, breast cancer-specific survival was worse among WLWH compared with HIV-negative patients diagnosed from 2004 to 2010, with an adjusted hazard ratio of 3.43 (95% CI, 2.35–5.01). Another study from the USA, which was not included in this meta-analysis because of potentially overlapping population, also showed a worse breast cancer-specific survival among WLWH (adjusted hazard ratio 2.84; 95% CI, 2.29–3.52) [46]. Likewise, the Coghill et al.[27] study demonstrated that breast cancer patients with similar age, race, year of cancer diagnosis, median household income, stage, treatment, type of health insurance and treating cancer facility still had a higher mortality likelihood if they had HIV infection. On the other hand, patients with cancer from sub-Saharan Africa usually have worse access to effective treatments compared with patients in developed countries, leading to worse survival rates [47]. Yet, despite this baseline poor prognosis among the general breast cancer population, this meta-analysis showed that WLWH in sub-Saharan Africa do even worse in terms of survival compared with HIV-negative sub-Saharan African women, despite adjustment for age, stage and subtype. This has been recently confirmed by the survival analysis of the ABC-DO study, showing that age-stage-adjusted hazard ratio for 3-year all-cause mortality was 1.48 (95% CI, 1.22–1.81) among WLWH compared with HIV-negative women, which is similar to our findings [48].

Therefore, in addition to the potential worse socioeconomic/treatment factors and the HIV-induced morbidity/mortality, WLWH may indeed present with breast cancer that is biologically more aggressive, translating into more advanced stage at diagnosis and worse survival. This is line with our current knowledge about the importance of immune system functioning on breast cancer prognosis, as differences in the composition and function of the immune cell infiltrate can lead to distinct survival outcomes in breast cancer [49]. In addition, recent clinical trials have shown the positive effect of stimulating the immune system with immune checkpoint inhibitors, further reinforcing the importance of an intact immune system to combat the tumor [50,51]. Moreover, it is also recognized that immunosuppressed solid organ transplant recipients have worse breast cancer-specific survival comparing to the general cancer population [52].

In addition to the limitations already mentioned, performing this meta-analysis posed other challenges. High heterogeneity was found in the meta-analysis for adjusted overall survival in North America. Nonetheless, heterogeneity was low for most of the other meta-analyses. Data regarding cancer treatment received by WLWH vs. HIV-negative patients were scarce (Supplemental Table 20, https://links.lww.com/QAD/B974). Still, unlike what has been reported in the USA [44], no differences were observed in terms of receipt of surgery, radiotherapy or chemotherapy between WLWH vs. HIV-negative women in the sub-Saharan African studies [6,14,17], but only one of them adjusted survival estimates for treatment [17]. Similarly, no study adjusted survival estimates to CD4+ cell count, antiretroviral therapy or to time since HIV diagnosis. This is a limitation as a low CD4+ cell count at cancer diagnosis seems to be associated with worse survival among patients with non-AIDS-defining malignancies and could be a potential factor to explain why survival outcomes are worse among some WLWH with breast cancer. On the other hand, the influence of antiretroviral therapy or time since HIV diagnosis on cancer-related survival is less clear [53,54].

All included studies were conducted in North America and sub-Saharan Africa, which may limit the generalization of these results to parts of the world with a different type/access to healthcare system. Although there are studies on WLWH with breast cancer in Latin America and Europe, they do not make a comparison to HIV-negative women [54–57]. On the other hand, sub-Saharan Africa is a very large region, in which countries’ income classification vary from low-income (e.g. Mozambique) to upper middle income (e.g. South Africa), which may affect patients’ outcomes. The small number of studies within each of these income categories precluded a meaningful pooled analysis of each endpoint, but reassuringly, heterogeneity was low/moderate in most of the different meta-analyses, which implies that these diverse economic backgrounds do not have a large influence in the differences between WLWH and HIV-negative women with breast cancer in sub-Saharan Africa. Lastly, an individual patient-level data meta-analysis would allow a more precise and robust analysis of our endpoints as it would account for more confounding factors.

Our results have implications for the care of WLWH. The European AIDS Clinical Society already recommends mammographic screening for WLWH with an age of 50 years or more [58]. Nonetheless, in most regions of the world, 70% of WLWH are diagnosed with breast cancer under the age of 50 years [4] and the implementation of population-mammographic screening in low-resource settings, such as sub-Saharan Africa is very low [59]. In this region, women often present with late-stage disease, thus the use of clinical breast examinations by a health professional could be considered. These examinations detect breast cancer in symptomatic women at earlier stages (i.e. ‘clinical downstaging’) and could raise awareness regarding breast cancer among WLWH and health professionals; however, their survival benefit has not yet been demonstrated [60]. For WLWH who already developed breast cancer, an effort should be made to provide a similar care to the one given to HIV-negative patients, namely in terms of access to effective treatment and adequate follow-up. These patients should also be included in clinical trials testing new anticancer therapies, including immune checkpoint inhibitors, as a way to assess the safety and efficacy of innovative treatments in this population [61–63].

Conducting a systematic review and meta-analysis using data from observational studies is often more difficult to perform compared with the use of data from randomized clinical trials, given the potentially higher heterogeneity in terms of study design and methodology. Nonetheless, there are several research questions that can only be answered through the pooling of observational studies and the comparison of WLWH vs. HIV-negative women with breast cancer is one of them. Therefore, this study provides the best available evidence on this topic for the time being, and the results suggest that WLWH are diagnosed with breast cancer at a more advanced stage and have a lower likelihood of presenting with estrogen receptor-positive/HER2-negative subtype. In addition, even after adjusting for other prognostic factors, WLWH have worse overall survival compared with HIV-negative patients, both in sub-Saharan Africa and in North America. These results should raise the awareness of clinicians and policy makers regarding the detection and survival gap among WLWH who develop breast cancer. This underserved population should be the focus of more research efforts, in order to better understand the reasons behind these disparities, which can be related to viral factors, distinct breast cancer biology and anticancer immune response, and/or to a lower access to timely diagnosis of breast cancer and effective treatments.

Acknowledgements

N.D. is a postdoctorate clinical master specialist of the F.R.S-FNRS. L.B. is supported by ‘Les Amis de l’Institut Bordet’ Foundation. D.E. acknowledges the support of the European Society for Medical Oncology (ESMO) for a research fellowship (2018–2019) at Institut Jules Bordet (Brussels, Belgium).

Authors’ contributions: Conceptualization: M.B., E.A., M.L. Data curation: M.B., M.A.D., C.D.A., D.E., R.C. Formal analysis: M.Br., M.C. Investigation: all authors. Methodology: M.B., M.Br., M.C., N.L., M.L. Supervision: M.L., N.L. Writing – original draft: M.B. Writing – review and editing: all authors.

Conflicts of interest

M.P., L.B., Ed.A.: research grants for their Institute from Radius, AstraZeneca, Lilly, MSD, GSK/Novartis, Roche/GNE, Synthon, Servier and Pfizer. M.B., M.A.F., C.D.A., D.E., R.C.: research grants for their Institute, where they work as medical research fellows: from Radius, AstraZeneca, Lilly, MSD, GSK/Novartis, Roche/GNE, Synthon, Servier and Pfizer (none in their name). M.B.: travel grant and speaker honoraria from Roche. D.E.: funding for his ESMO fellowship (2018–2019) from Novartis. R.C.: speaker honoraria from Boehringer-Ingelheim, AstraZeneca and Janssen; travel grants from Pfizer and AstraZeneca. M.P.: consultancy honoraria for AstraZeneca, Camel-IDS, Crescendo Biologics, Debiopharm, G1 Therapeutics, Genentech, Huya, Immunomedics, Lilly, Menarini, MSD, Novartis, Odonate, Periphagen, Pfizer, Roche, Seattle Genetics; scientific board member: Oncolytics. L.B.: travel grant from Roche; speaker honoraria from BMS. N.D.: travel grants from Pfizer, MSD and Janssen's. Ed.A.: speaker honoraria and/or advisory board for Roche/GNE, Novartis, SeaGen and Zodiac; travel grants from Roche/GNE and GSK/Novartis. M.L.: consultancy for Roche and Novartis; speaker honoraria from Theramex, Roche, Novartis, Pfizer, Lilly and Takeda. All other authors declare that they have no conflict of interest (M.Br., M.C., C.C. and N.L.).

References

1. GBD 2017 HIV collaborators. Global, regional, and national incidence, prevalence, and mortality of HIV, 1980–2017, and forecasts to 2030, for 195 countries and territories: a systematic analysis for the Global Burden of Diseases, Injuries, and Risk Factors Study 2017. Lancet HIV 2019; 6:e831–e859.
2. Yarchoan R, Uldrick TS. HIV-associated cancers and related diseases. N Engl J Med 2018; 378:1029–1041.
3. GLOBOCAN Cancer Fact Sheets: Breast cancer. Available at: http://globocan.iarc.fr/old/FactSheets/cancers/breast-new.asp. [Accessed 7 August 2018]
4. McCormack V, Febvey-Combes O, Ginsburg O, Dos-Santos-Silva I. Breast cancer in women living with HIV: a first global estimate. Int J Cancer 2018; 143:2732–2740.
5. Coghill AE, Engels EA, Schymura MJ, Mahale P, Shiels MS. Risk of breast, prostate, and colorectal cancer diagnoses among HIV-infected individuals in the United States. J Natl Cancer Inst 2018; 110:959–966.
6. Phakathi BP, Basson G, Karusseit VOL, Olorunju SAS, Mokoena T. The effect of HIV infection on the surgical, chemo- and radiotherapy management of breast cancer. A prospective cohort study. Int J Surg Lond Engl 2016; 34:109–115.
7. Cubasch H, Joffe M, Hanisch R, Schuz J, Neugut AI, Karstaedt A, et al. Breast cancer characteristics and HIV among 1,092 women in Soweto, South Africa. Breast Cancer Res Treat 2013; 140:177–186.
8. Shiels MS, Pfeiffer RM, Engels EA. Age at cancer diagnosis among persons with AIDS in the United States. Ann Intern Med 2010; 153:452–460.
9. Langenhoven L, Barnardt P, Neugut AI, Jacobson JS. Phenotype and treatment of breast cancer in HIV-positive and -negative women in Cape Town, South Africa. J Glob Oncol 2016; 2:284–291.
10. Shiels MS, Copeland G, Goodman MT, Harrell J, Lynch CF, Pawlish K, et al. Cancer stage at diagnosis in patients infected with the human immunodeficiency virus and transplant recipients. Cancer 2015; 121:2063–2071.
11. McKenzie F, Zietsman A, Galukande M, Anele A, Adisa C, Parham G, et al. Drivers of advanced stage at breast cancer diagnosis in the multicountry African breast cancer – disparities in outcomes (ABC-DO) study. Int J Cancer 2018; 142:1568–1579.
12. Cardoso F, Kyriakides S, Ohno S, Penault-Llorca F, Poortmans P, Rubio IT, et al. Early breast cancer: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up. Ann Oncol 2019; 30:1194–1220.
13. Goldhirsch A, Winer EP, Coates AS, Gelber RD, Piccart-Gebhart M, Thürlimann B, Senn HJ. Panel members. Personalizing the treatment of women with early breast cancer: highlights of the St Gallen International Expert Consensus on the Primary Therapy of Early Breast Cancer 2013. Ann Oncol 2013; 24:2206–2223.
14. Sadigh K, Hodgeman R, Tapela N. HIV is associated with decreased breast cancer survival: a Prospective Cohort Study | CROI Conference. 2019. Available at: http://www.croiconference.org/sessions/hiv-associated-decreased-breast-cancer-survival-prospective-cohort-study. [Accessed 17 April 2019]
15. Cubasch H, Dickens C, Joffe M, Duarte R, Murugan N, Tsai Chih M, et al. Breast cancer survival in Soweto, Johannesburg, South Africa: a receptor-defined cohort of women diagnosed from 2009 to 11. Cancer Epidemiol 2018; 52:120–127.
16. Coghill AE, Newcomb PA, Madeleine MM, Richardson BA, Mutyaba I, Okuku F, et al. Contribution of HIV infection to mortality among cancer patients in Uganda. AIDS Lond Engl 2013; 27:2933–2942.
17. Brandão M, Guisseve A, Bata G, de Morais G, Alberto M, Schmitt F, et al. 213PHIV infection in breast cancer patients from Mozambique: a prospective cohort study. Ann Oncol 2019; 30: doi:10.1093/annonc/mdz101.014.
18. Stroup DF, Berlin JA, Morton SC, Olkin I, Williamson GD, Rennie D, et al. Meta-analysis of observational studies in epidemiology: a proposal for reporting. JAMA 2000; 283:2008–2012.
19. Wolff AC, Hammond MEH, Allison KH, Harvey BE, Mangu PB, Bartlett JMS, et al. Human Epidermal Growth Factor Receptor 2 Testing in Breast Cancer: American Society of Clinical Oncology/College of American Pathologists Clinical Practice Guideline Focused Update. J Clin Oncol 2018; 36:2105–2122.
20. Hammond MEH, Hayes DF, Dowsett M, Allred DC, Hagerty KL, Badve S, et al. American Society of Clinical Oncology/College of American Pathologists Guideline Recommendations for Immunohistochemical Testing of Estrogen and Progesterone Receptors in Breast Cancer. J Clin Oncol 2010; 28:2784–2795.
21. Dowsett M, Nielsen TO, A’Hern R, Bartlett J, Coombes RC, Cuzick J, et al. International Ki-67 in Breast Cancer Working Group. Assessment of Ki67 in breast cancer: recommendations from the International Ki67 in Breast Cancer Working Group. JNCI J Natl Cancer Inst 2011; 103:1656–1664.
22. Egger M, Davey Smith G, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical test. BMJ 1997; 315:629–634.
23. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials 1986; 7:177–188.
24. Sseggwanyi J, Galukande M, Fualal J, Jombwe J. Prevalence of HIV/AIDS among breast cancer patients and the associated clinico-pathological features. Ann Afr Surg 2011; 8:22–27.
25. Ngidi S, Magula N, Sartorius B, Govender P, Madiba TE. Incidence of chemotherapy-induced neutropenia in HIV-infected and uninfected patients with breast cancer receiving neoadjuvant chemotherapy. SAMJ South Afr Med J 2017; 107:595–601.
26. van Zyl N, Minné C, Mokone DH. Human immunodeficiency virus infection in breast cancer patients: the prevalence thereof and its effect on breast cancer characteristics at Dr George Mukhari Academic Hospital Breast Clinic, Ga-Rankuwa, South Africa. SA J Radiol 2018; 22:1–7.
27. Coghill AE, Han X, Suneja G, Lin CC, Jemal A, Shiels MS. Advanced stage at diagnosis and elevated mortality among US patients with cancer infected with HIV in the National Cancer Data Base. Cancer 2019; 125:2868–2876.
28. Traore B, Diane S, Sow MS, Keita M, Conde M, Traore FA, et al. HIV infection in patients with breast cancer in Guinea (West Africa). Pan Afr Med J 2015; 21:261.
29. Presti C, Haslinger ML, Wehner P. Breast cancer in HIV-positive patients: a multi-institutional retrospective review. J Prev Med Heal 2017; 1:1009.
30. Ayeni OA, Joffe M, Cubasch H, Rinaldi S, Taljaard C, Vorster E, et al. Prevalence of comorbidities in women with and without breast cancer in Soweto, South Africa: results from the SABC study. S Afr Med J 2019; 109:264–271.
31. Phakathi B, Cubasch H, Nietz S, Dickens C, Dix-Peek T, Joffe M, et al. Clinico-pathological characteristics among South African women with breast cancer receiving antiretroviral therapy for HIV. Breast 2019; 43:123–129.
32. Bhatia RK, Narasimhamurthy M, Martei YM, Prabhakar P, Hutson J, Chiyapo S, et al. Report of clinico-pathological features of breast cancer in HIV-infected and uninfected women in Botswana. Infect Agent Cancer 2019; 14:28.
33. Coghill AE, Shiels MS, Suneja G, Engels EA. Elevated cancer-specific mortality among HIV-infected patients in the United States. J Clin Oncol 2015; 33:2376–2383.
34. Biggar R, Engels E, Ly S, Kahn A, Schymura M, Sackoff J, et al. Survival after cancer diagnosis in persons with AIDS. J Acquir Immune Defic Syndr 2005; 39:293–299.
35. Lambertini M, Santoro L, Mastro LD, Nguyen B, Livraghi L, Ugolini D, et al. Reproductive behaviors and risk of developing breast cancer according to tumor subtype: a systematic review and meta-analysis of epidemiological studies. Cancer Treat Rev 2016; 49:65–76.
36. Gaudet MM, Gierach GL, Carter BD, Luo J, Milne RL, Weiderpass E, et al. Pooled analysis of nine cohorts reveals breast cancer risk factors by tumor molecular subtype. Cancer Res 2018; 78:6011–6021.
37. Denkert C, von Minckwitz G, Darb-Esfahani S, Lederer B, Heppner BI, Weber KE, et al. Tumour-infiltrating lymphocytes and prognosis in different subtypes of breast cancer: a pooled analysis of 3771 patients treated with neoadjuvant therapy. Lancet Oncol 2018; 19:40–50.
38. The Cancer Genome Atlas Network. Comprehensive molecular portraits of human breast tumours. Nature 2012; 490:61.
39. Howlader N, Altekruse SF, Li CI, Chen VW, Clarke CA, Ries LAG, Cronin KA. US incidence of breast cancer subtypes defined by joint hormone receptor and HER2 status. J Natl Cancer Inst 2014; 106: dju055.
40. Karim R, Mack WJ, Kono N, Tien PC, Anastos K, Lazar J, et al. Gonadotropin and sex steroid levels in HIV-infected premenopausal women and their association with subclinical atherosclerosis in HIV-infected and -uninfected women in the Women's Interagency HIV Study (WIHS). J Clin Endocrinol Metab 2013; 98:E610–E618.
41. King EM, Albert AY, Murray MCM. HIV and amenorrhea: a meta-analysis. AIDS 2019; 33:483–491.
42. Robbins HA, Pfeiffer RM, Shiels MS, Li J, Hall HI, Engels EA. Excess cancers among HIV-infected people in the United States. J Natl Cancer Inst 2015; 107: dju503.
43. Eng A, McCormack V, dos-Santos-Silva I. Receptor-defined subtypes of breast cancer in indigenous populations in Africa: a systematic review and meta-analysis. PLoS Med 2014; 11:e1001720.
44. Suneja G, Shiels MS, Angulo R, Copeland GE, Gonsalves L, Hakenewerth AM, et al. Cancer treatment disparities in HIV-infected individuals in the United States. J Clin Oncol 2014; 32:2344–2350.
45. Jedy-Agba E, McCormack V, Adebamowo C, dos-Santos-Silva I. Stage at diagnosis of breast cancer in sub-Saharan Africa: a systematic review and meta-analysis. Lancet Glob Health 2016; 4:e923–e935.
46. Chhatre S, Schapira M, Metzger DS, Jayadevappa R. Association between HIV infection and outcomes of care among medicare enrollees with breast cancer. EClinicalMedicine 2019; 17:100205.
47. Vanderpuye V, Grover S, Hammad N, PoojaPrabhakar, Simonds H, Olopade F, et al. An update on the management of breast cancer in Africa. Infect Agent Cancer 2017; 12:1–12.
48. McCormack V, McKenzie F, Foerster M, Zietsman A, Galukande M, Adisa C, et al. Breast cancer survival and survival gap apportionment in sub-Saharan Africa (ABC-DO): a prospective cohort study. Lancet Glob Health 2020; 8:e1203–e1212.
49. Bense RD, Sotiriou C, Piccart-Gebhart MJ, Haanen JBAG, van Vugt MATM, de Vries EGE, et al. Relevance of tumor-infiltrating immune cell composition and functionality for disease outcome in breast cancer. J Natl Cancer Inst 2017; 109: djw192.
50. Schmid P, Adams S, Rugo HS, Schneeweiss A, Barrios CH, Iwata H, et al. IMpassion130 Trial Investigators. Atezolizumab and nab-paclitaxel in advanced triple-negative breast cancer. N Engl J Med 2018; 379:2108–2121.
51. Schmid P, Cortes J, Pusztai L, McArthur H, Kümmel S, Bergh J, et al. Pembrolizumab for early triple-negative breast cancer. N Engl J Med 2020; 382:810–821.
52. D’Arcy ME, Coghill AE, Lynch CF, Koch LA, Li J, Pawlish KS, et al. Survival after a cancer diagnosis among solid organ transplant recipients in the United States. Cancer 2019; 125:933–942.
53. Fink VI, Jenkins CA, Castilho JL, Person AK, Shepherd BE, Grinsztejn B, et al. Survival after cancer diagnosis in a cohort of HIV-positive individuals in Latin America. Infect Agent Cancer 2018; 13:16.
54. Gotti D, Raffetti E, Albini L, Sighinolfi L, Maggiolo F, Di Filippo E, et al. Master Cohort Group. Survival in HIV-infected patients after a cancer diagnosis in the cART Era: results of an Italian Multicenter Study. PloS One 2014; 9:e94768.
55. Dauby N, De Wit S, Delforge M, Necsoi VC, Clumeck N. Characteristics of non-AIDS-defining malignancies in the HAART era: a clinico-epidemiological study. J Int AIDS Soc 2011; 14:16.
56. Andrade ACV, de Luz PM, Veloso VG, Cardoso SW, Moreira RI, Grinsztejn B, et al. Breast cancer in a cohort of human immunodeficiency virus (HIV)-infected women from Rio de Janeiro, Brazil: a cases series report and an incidence rate estimate. Braz J Infect Dis 2011; 15:387–393.
57. Spano J-P, Lanoy E, Mounier N, Katlama C, Costagliola D, Heard I. Breast cancer among HIV infected individuals from the ONCOVIH study, in France: therapeutic implications. Eur J Cancer 2012; 48:3335–3341.
58. Cancer Screening. EACS Guide l. Available at: https://eacs.sanfordguide.com/prevention-noninfectious-co-morbidities/cancer-screening-methods. [Accessed 18 March 2020]
59. Black E, Richmond R. Improving early detection of breast cancer in sub-Saharan Africa: why mammography may not be the way forward. Glob Health 2019; 15:3.
60. Mittra I, Mishra GA, Singh S, Aranke S, Notani P, Badwe R, et al. A cluster randomized, controlled trial of breast and cervix cancer screening in Mumbai, India: methodology and interim results after three rounds of screening. Int J Cancer 2010; 126:976–984.
61. Uldrick TS, Ison G, Rudek MA, Noy A, Schwartz K, Bruinooge S, et al. Modernizing clinical trial eligibility criteria: recommendations of the American Society of Clinical Oncology–Friends of Cancer Research HIV Working Group. J Clin Oncol 2017; 35:3774–3780.
62. Gonzalez-Cao M, Morán T, Dalmau J, Garcia-Corbacho J, Bracht JWP, Bernabe R, et al. Assessment of the feasibility and safety of durvalumab for treatment of solid tumors in patients with HIV-1 infection: the Phase 2 DURVAST Study. JAMA Oncol 2020; 6:1063–1067.
63. Uldrick TS, Gonçalves PH, Abdul-Hay M, Claeys AJ, Emu B, Ernstoff MS, et al. Cancer Immunotherapy Trials Network (CITN)-12 Study Team. Assessment of the safety of pembrolizumab in patients with HIV and advanced cancer—a phase 1 study. JAMA Oncol 2019; 5:1332–1339.
Keywords:

breast neoplasms; cancer staging; HIV; meta-analysis; survival; women's health

Supplemental Digital Content

Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved.