J Antimicrob Chemother 2015; 70: 249 – 256 doi:10.1093/jac/dku362 Advance Access publication 15 September 2014

Immuno-virological and clinical impacts of treating cancer in patients living with HIV Gwenae¨l Le Moal1,2†, Thierry Prazuck2,3, Mariam Saberan-Roncato2,4, Joe¨lle Julien2,5, Thierry Pasdeloup2,6, Jean-Michel Descamps2,7, Marc Lestelle2,8, Philippe Arsac2,9, Elise Champeaux-Orange10, Lydia Bouaraba2,3 and Laurent Hocqueloux2,3*† 1

Service de Maladies Infectieuses, CHU La Mile´trie, Poitiers, France; 2COREVIH Centre Poitou-Charentes, CHU Bretonneau, Tours, France; Service de Maladies Infectieuses, CHR d’Orle´ans—La Source, Orle´ans, France; 4Service de Maladies Infectieuses, CHG, La Rochelle, France; 5 Service de Pneumologie, CH Jacques Cœur, Bourges, France; 6Service de Me´decine Interne, CHG, Saintonge, France; 7Poˆle Me´decine Interne et Re´animation, CHG, Niort, France; 8Service de Pneumologie, CH Louis Pasteur, Chartres, France; 9Service de Me´decine Interne, CHR Porte Madeleine, Orle´ans, France; 10Service d’Onco-Radiothe´rapie, CHR d’Orle´ans—La Source, Orle´ans, France

3

Received 5 May 2014; returned 12 July 2014; revised 8 August 2014; accepted 16 August 2014 Background: Patients living with HIV (PLHIV) are increasingly being affected by cancer. However, data evaluating the long-term impact of cancer treatment on HIV course are sparse. Methods: To determine whether anticancer treatments detrimentally impact HIV course, we conducted a retrospective cohort study in seven hospitals in France. Adult PLHIV treated for haematological or solid malignancies were included and compared (1:1) with suitably matched (cancer-free) controls. The primary outcome was the risk of a ≥25% reduction in the absolute CD4+ count during follow-up. The risks for virological failure (i.e. a confirmed plasma viral load .50 copies/mL), incidental AIDS-related illnesses and death over time were also assessed. Multivariate Cox proportional-hazards regression analyses were used to identify the outcome predictors. Results: One-hundred-and-ten patients with cancer and 110 controls were followed for a median of 4.4 years. In a Cox model, the CD4+ depletion was strongly predicted by external radiotherapy (ERT) exposure (HR¼ 5.1, 95% CI, 3.0–8.6, P,0.0001) but not by chemotherapy. For patients exposed to ERT, the magnitude of the CD4+ depletion peaked 6 months after their cancer diagnosis (mean CD4+ drop at this time¼2283+370 cells/mm3). Overall, the cancer patients were also more likely to experience virological failure than the controls (HR¼1.7, 95% CI, 1.1–2.7, P¼0.03). Finally, the incidence of AIDS-related illnesses was similar for both groups. Conclusions: In PLHIV, cancer treatment increased the risk for prolonged CD4+ depletion and virological failure but had no impact on AIDS-related events when appropriate prophylaxes were implemented. Keywords: ART, CD4+ depletion, AIDS, virological failure

Introduction Soon after highly active ARTs were first marketed, the incidence of AIDS-defining illnesses (ADIs) suddenly fell, dramatically reducing patient mortality.1 As a result, ageing has become a common occurrence for patients living with HIV (PLHIV). AIDS-free survival exposes these patients to new health challenges.2 In this setting, cancer is a major concern. Indeed, cancer is increasingly diagnosed in PLHIV; most of these malignancies are non-ADIs.3,4 In France, cancers are currently the leading cause of death among PLHIV.5 Although many studies have described the spectrum, incidence and risk factors of malignancies in PLHIV, few have

specifically focused on the impact of cancer treatments on immuno-virological parameters and clinical outcomes. Thus, little is known about the course of HIV after cancer management. Reversible myelosuppression is a major side effect of intravenous cytotoxic chemotherapy and radiotherapy; it exposes patients to transient but potentially severe white blood cell depletion, including T cell depletion, and consequent opportunistic infections. Moreover, it is well known that intravenous chemotherapy causes nausea and vomiting. Finally, significant drug – drug interactions between ART and cytotoxic chemotherapy have been described. 6 Together, these issues could have a deleterious impact on the efficacy of and adherence to combination

# The Author 2014. Published by Oxford University Press on behalf of the British Society for Antimicrobial Chemotherapy. All rights reserved. For Permissions, please e-mail: [email protected]

249

Downloaded from http://jac.oxfordjournals.org/ at University of Florida on February 16, 2015

*Corresponding author. Service des Maladies Infectieuses et Tropicales, CHR d’Orle´ans—La Source, BP 86709, 45067 Orle´ans cedex 2, France. Tel: +33-2-3822-9588; Fax: +33-2-3851-4153; E-mail: [email protected] †Contributed equally to the work.

Le Moal et al.

antiretroviral therapy (cART) and could worsen the prognosis of HIV infection. We aimed to investigate whether PLHIV are at a higher risk for severe CD4+ depletion, virological failure, new ADIs and death following the treatment of various haematological or solid tumours compared with a matched population without malignancies.

We also examined the factors predicting the CD4+ decrease. The secondary objectives were to assess the risk of virological failure (defined as two consecutive PVLs .50 copies/mL), incidental ADIs (according to the 1993 revised classification of the CDC) and death rates over time between the patients and controls. Virological failure was considered only for patients under cART for ≥6 months at baseline or for those patients who initiated cART within 1 month after baseline (data were censored for these patients until 6 months of cART were completed).

Methods Covariates

We conducted a retrospective cohort study to evaluate the immunological, virological and clinical outcomes in a group of HIV-infected patients who were diagnosed and treated for malignancies (i.e. the patient group) as compared with a matched group without malignancy (i.e. the control group). This retrospective study was embedded in an ongoing multicentre prospective cohort (the ‘COREVIH Centre et Poitou-Charentes’ cohort) following HIV-infected individuals at seven primary care and referral hospitals located in two different geographical areas in France. Given the retrospective and observational nature of this study, written informed consent was waived; however, when their data were entered into the electronic medical record database, all participants provided informed written consent for the anonymous use of their clinical and biological data for biomedical research. The study was approved by the medical data protection committee (CCTIRS) and the French computer watchdog (CNIL). Living patients were informed that they could access their computerized data and withdraw their consent at any time.

Standardized data collection forms, including characteristics of both malignancy and HIV infection, were completed (G. L. M. and L. B.). The baseline data that were collected included demographics (age, gender and ethnicity), characteristics of the malignancy (histological reports, localization, stage and any iconographic or biological results supporting the diagnosis), hepatitis B or C coinfection, current tobacco smoking, alcohol abuse, date of HIV diagnosis, CDC classification, medical and ART history, CD4+ cell count nadir and PVL zenith before treatment. Follow-up data included the therapeutics used for the cancer treatment (surgery, irradiation, systemic chemotherapy and any other therapy), evolution of the malignancy, death (and its cause), new ADIs, CD4+ cell counts and PVL curves (typically, blood samples were taken three to four times a year, according to national guidelines7).

Participants Using the electronic medical record database, we systematically screened adult (18 years of age and older) HIV-1-infected patients attending one of the participating centres to identify those patients who were diagnosed with any malignancy, according to hospital discharge ICD-10 codes. The study enrolment period ran from January 2005 to December 2010. Patients were analysed until the last follow-up visit preceding the study termination (in July 2013) or until they were lost to follow-up or died. Patients who were diagnosed with any malignancy (i.e. codes between C00 and C97 of ICD-10 and/or those beginning with B21), excluding in situ neoplasms, were included in the study (Figure S1, available as Supplementary data at JAC Online). Baseline was defined as the day that the patients were diagnosed with cancer. Only new cancer diagnoses were included. New episodes of previously diagnosed malignancies were excluded. Patients with multiple malignancies or relapse during follow-up were only entered once. Based on histological reports, the scientific committee (E. C.-O., G. L. M., L. H. and T. Prazuck) reviewed and validated all cancer diagnoses, except for hepatocellular carcinoma, which was based on imaging and alpha-fetoprotein assay. Therapeutic approaches were carefully reviewed by the same staff and compared with contemporary standards of care (i.e. recommendations for HIV-uninfected patients). Each patient with cancer was matched (1:1) with a control without cancer who consulted on the same day (+1 month) in the same hospital whenever possible. For the controls, baseline was defined as the day they were matched with the patients. The controls and patients were also matched for age (+2 years), gender, ethnicity (white versus others), duration since the HIV diagnosis (+2 years), CDC classification stage (A versus B or C), cART use (if any), nadir CD4+ T cell count (+50 cells/mm3) and baseline plasma HIV-RNA viral load (PVL; +0.5 log10/mL when .50 copies/mL). No control was included twice.

Outcomes The main objective was to compare the risk of a ≥25% reduction of CD4+ absolute count between the patients and controls during the study period.

250

Statistical analysis Data were expressed as median (IQR), mean+SD or percentage values. Between-group differences were compared using x2 or Fisher’s exact tests for categorical variables and using the unpaired t-test, Mann – Whitney U-test or Kruskal – Wallis test for continuous variables. Spearman’s rank-order correlation was used to determine linear correlations. Kaplan – Meier curves were drawn for the primary and secondary outcomes, and P values were based on the log-rank test. Cox proportionalhazards regression models were used to compare times to events and to control for covariates. All demographic, baseline clinical and biological variables were evaluated as covariates and considered for inclusion in the final model on the basis of an unadjusted association with different outcomes. All P values were two-tailed, and statistical significance was set at the 0.05 level. Statistical analyses were performed using MedCalcw software version 12.7.5 (MedCalc Software, Ostend, Belgium).

Results Participant characteristics During the 6 year enrolment period, 2390 PLHIV were followed, accounting for 12 717 person-years. A total of 115 individuals were diagnosed with cancer, five of whom were excluded because of insufficient data. Finally, 110 patients (78 men and 32 women) were included in the study. Thus, the incidence of cancer in our cohort was 8.7 cases per 1000 person-years. At the time of their cancer diagnoses, 83 of 110 patients (75%) were on cART, but only 61 (55%) had undetectable PVL (,50 copies/mL). In 7 of 110 patients (6%), the cancer revealed HIV infection. The patients suffered from various solid and haematological malignancies (Figure 1). Although Kaposi sarcoma was the most frequently occurring malignancy, non-AIDS-defining cancers were much more common than AIDS-defining cancers, accounting for 73% and 27% of all diagnoses, respectively. Of 110 patients, 105 (95%) received specific therapeutics (Table S1), which consisted of systemic cytotoxic chemotherapy (SCC, n¼58), surgery (n¼43), external radiotherapy (ERT, n¼41),

Downloaded from http://jac.oxfordjournals.org/ at University of Florida on February 16, 2015

Study design and setting

JAC

Impact of cancer treatment on HIV infection

5

0

10 Number of cases

15

20

* Cutaneous (n = 13), cutaneous and visceral (n = 6). ** Bladder (n = 1), cholangiocarcinoma (n = 1), chronic lymphocytic leukaemia (n = 1), erythroleukaemia (n = 1), glioblastoma (n = 1), liposarcoma (n = 1), ovarian (n = 1), thyroid (n = 1). Figure 1. Distribution of malignancies.

Table 1. Baseline characteristics of the patients with cancer and controls Patients (n¼110)

Controls (n ¼110)

P valuea

Age (years), median (IQR) Male, n (%) White, n (%) Hepatitis B and/or C coinfection, n (%) Previous history of malignancy, n (%) Duration with HIV (years), median (IQR)

47.5 (39.6–55.9) 78 (71) 94 (86) 27 (25) 10 (9) 10.7 (3.8–16.5)

47.0 (39.4– 55.9) 78 (71) 94 (86) 24 (22) 10 (9) 10.0 (2.7 –13.8)

0.96 1.0 1.0 0.75 1.0 0.23

CDC clinical category, n (%) A B C

51 17 42

52 25 33

Nadir CD4+ count (cells/mm3), median (IQR) Highest PVL (copies/mL), median (IQR) On cART, n (%) CD4+ count (cells/mm3), median (IQR) PVL ,50 copies/mL, n (%) Current alcohol abuse, n (%) Current tobacco smoking, n (%)

122 (47 –269) 5.1 (5.0– 5.7) 83 (75) 359 (214– 595) 61 (55) 24 (22) 58 (53)

151 (65– 276) 5.3 (4.7– 5.6) 86 (78) 423 (267–591) 58 (53) 11 (10) 45 (41)

0.27

0.53 0.83 0.75 0.34 0.79 0.03 0.10

a

A t-test was used for normally distributed continuous data, the Mann– Whitney U-test was used for non-parametric continuous data and the x2 test was used for dichotomous data.

ART initiation (n ¼ 9, for Kaposi sarcoma), immunomodulation agents (n¼4) and hormonotherapy (n¼3). Of these 105 patients, 27 received a combination of SCC plus ERT. Finally, 5 of 110 patients (5%) received only comfort care: one patient refused specific therapeutics, and four had incurable tumours. Overall, the patients

received 1.4 (mean) specific cancer treatments. Compared with contemporary guidelines, the treatments aligned with the standard of care for all but one patient, who refused any specific treatment. The patients were matched with 110 controls. The main baseline characteristics of both groups are summarized in Table 1. The

251

Downloaded from http://jac.oxfordjournals.org/ at University of Florida on February 16, 2015

Kaposi sarcoma* Lung Hodgkin’s disease Oral cavity, pharynx and larynx Non-Hodgkin’s lymphoma Breast Anal Skin (non-melanoma) Colon Prostate Hepatocellular carcinoma Multiple myeloma Kidney Cervix Endometrial Unknown origin Miscellaneous**

Le Moal et al.

(b) 100 Controls

75

50 Cases 25 P < 0.0001, by log-rank test 0 1

2

6 7 3 4 5 Time since baseline, years

8

75 Controls Cases

50

25 P = 0.02, by log-rank test 0 0

9

1

2

6 7 3 4 5 Time since baseline, years

8

9

(d)

(c) 100

100

Cases Controls Percentage of survivors

Percentage without new ADI

100

75

50

25

Controls

75 Cases 50

25 P < 0.0001, by log-rank test

P = 0.5, by log-rank test 0

0 0

1

2

6 7 3 4 5 Time since baseline, years

8

9

0

1

2

3 4 5 6 7 Time since baseline, years

8

9

Figure 2. Kaplan – Meier survival estimates for participants without CD4+ depletion, virological failure, new ADI and death. Curves represent the percentages of patients and controls that remained free of pre-defined CD4+ depletion (main outcome) (a), virological failure (b) and new ADI (c), and who survived (d) from baseline to the last visit.

patients and controls had similar characteristics at baseline, except for current alcohol abuse, which was more frequently reported among the patients than among the controls (22% versus 10%, P ¼ 0.03).

Main outcome The median follow-up time was 3.6 years (IQR 2.0 – 5.3) for the patients and 4.8 years (IQR 3.3 –6.0) for the controls (P ¼ 0.001). No patient (or control) was lost to follow-up. During this period, 53 of 110 patients (48%) and 20 of 110 (18%) controls experienced a ≥25% CD4+ depletion compared with the respective baseline value (P, 0.0001). In an unadjusted analysis, the patients were more likely to experience CD4+ depletion (HR ¼ 3.5; 95% CI, 2.1 – 5.8, P, 0.0001) than the controls (Figure 2a). In a Cox proportional-hazards regression model, two variables remained predictive of pre-defined CD4+ depletion: exposure to ERT (HR ¼ 5.1; 95% CI: 3.0 – 8.6; P, 0.0001), and a .11 year duration of living with HIV (HR ¼ 2.0, 95% CI, 1.2 – 3.3, P ¼ 0.007) (Table 2 and Figure 3a). Notably, type of cancer (AIDS-defining versus non-AIDS-defining, haematological versus

252

non-haematological, infection-related versus non-infectionrelated), number of myelotoxic drugs used in SCC (0, 1, 2, ≥3) and exclusion of deceased patients had no impact on the incidence of CD4+ depletion in this model (data not shown). For the patients who received ERT, the magnitude of the CD4+ depletion peaked at 6 months after their cancer diagnoses [mean CD4+ drop at this time ¼ 2283+370 cells/mm3, median ¼ 2174 cells/mm3 (IQR 2352 to 292)]. This decrease in the CD4+ count was significantly greater than for the patients unexposed to ERT and the controls at every follow-up time from month 3 to month 36 (P ≤ 0.0001 for all comparisons) (Figure 3b). At the peak of the CD4+ depletion (i.e. month 6), the patients who were exposed to ERT also had a greater decrease in their total lymphocyte count, CD8+ count and CD4+/CD8+ ratio compared with the patients who were unexposed to ERT and the controls (data not shown). Finally, in the ERT-exposed patients, the magnitude of the CD4+ drop at month 6 was highly correlated with the baseline CD4+ count (r ¼ 20.8, P,0.0001) but not with age, type of cancer, co-administration of SCC, localization of radiotherapy, total dose received (Grays) and mode of administration (hypo- versus normal fractionation, data not shown).

Downloaded from http://jac.oxfordjournals.org/ at University of Florida on February 16, 2015

0

Percentage without virological failure

Percentage without CD4+ depletion

(a)

JAC

Impact of cancer treatment on HIV infection

Table 2. Baseline factors associated with pre-defined CD4+ depletion (main outcome) or virological failure in Cox models ≥25% CD4+ depletion (n ¼220)

Virological failure (n ¼205)

P value univariate

P value

multivariatea

0.18 0.02 1.0 0.22 0.05

— — — — —

Duration with HIV (years) duration with HIV .11 years

0.004 0.001

— 0.007

Highest PVL (log10 copies/mL) Nadir CD4+ (cells/mm3) Baseline CD4+ count (cells/mm3) Baseline CD4+/CD8+ ratio

0.26 0.02 0.11 0.72

Time elapsed since first undetectable PVLb (years) time elapsed since first undetectable PVL .1 year

95% CI

univariate

multivariatea

HR

95% CI

0.14 0.75 0.08 0.06 0.21

— — — — —

0.19 —

— —

— — — —

0.01 0.15 0.02 0.0001

— — — —

0.19 —

— —

0.0002 ,0.0001

— ,0.0001

0.15

0.06 –0.39

cART regimen class based on NNRTI (versus any other)

0.01 0.16

— —

0.002 0.0002

— 0.02

0.43

0.21 –0.86

Belongs to patient group (versus control) Received ERT Received cytotoxic chemotherapy

,0.0001 ,0.0001 0.0006

— ,0.0001 —

0.08 0.26 0.25

— — —

c

2.02

5.10

1.22– 3.33

3.04– 8.58

a

Stepwise regression. PVL ,50 copies/mL. c Regimen based on PIs, integrase inhibitors or NNRTIs. b

Other outcomes Virological failure A total of 205 patients were eligible for the virological outcome evaluations: 169 patients were undergoing cART at baseline, and 36 started cART within 1 month after baseline. During follow-up, 44 of 104 patients (42%) and 30 of 101 (30%) controls fulfilled the criteria of virological failure (P¼0.08). In an unadjusted analysis, the patients were more likely than the controls to experience virological failure over time (HR ¼ 1.7; 95% CI, 1.1 – 2.7, P ¼ 0.03) (Figure 2b). This difference was no longer significant after adjusting for other baseline characteristics; in the multivariate analysis, only the use of NNRTIs and the duration with undetectable PVL at baseline (.1 year) remained predictive for viral control (Table 2). Type of cancer (AIDS-defining versus non-AIDS-defining, haematological versus non-haematological, infection-related versus noninfection-related), number of myelotoxic drugs used in SCC (0, 1, 2, ≥3) and exclusion of patients had no impact on the incidence of virological failure in this model (data not shown).

New ADIs During follow-up, no significant between-group difference was observed concerning new ADIs (HR ¼ 0.6, 95% CI, 0.1 – 2.9, P ¼ 0.5) (Figure 2c). Overall, six PLHIV experienced new ADIs: two patients (candidal oesophagitis, n¼ 1; Mycobacterium avium

infection, n ¼ 1) and four controls (candidal oesophagitis, n ¼ 1; extrapulmonary tuberculosis, n ¼ 1; Pneumocystis jirovecii pneumonia, n ¼ 1; recurrent pneumonia plus wasting syndrome, n ¼ 1). Notably, in both groups, all PLHIV with CD4+ T cell counts ,200 cells/mm3 (,15%) were prescribed systematic prophylaxis against P. jirovecii pneumonia and toxoplasmosis. Antimycobacterial prophylaxis was also administered to PLHIV with a CD4+ count ,50 cells/mm3.

Deaths Overall, 31 of 110 patients (28%) died during follow-up; there were no deaths in the control group (P,0.0001) (Figure 2d). The deaths were directly attributable to the progression of cancer in all but two patients (end-stage renal insufficiency, n ¼ 1; aspiration pneumonia complicating severe chronic respiratory failure, n¼ 1). The proportion of death, according to each type of cancer, was as follows (in order of decreasing frequency, for cancers whose number is ≥4): liver, 4 of 4 (100%); lung, 9 of 12 (75%); throat, 4 of 10 (40%); anal, 2 of 7 (29%); prostate, 1 of 4 (25%); colon, 1 of 4 (25%); Hodgkin’s lymphoma, 1 of 5 (20%); breast, 1 of 8 (13%); and non-Hodgkin’s lymphoma, 1 of 14 (7%).

Long-term immuno-virological outcomes The remaining 79 patients who survived and 110 controls had similar follow-up durations (median 4.6 versus 4.8 years,

253

Downloaded from http://jac.oxfordjournals.org/ at University of Florida on February 16, 2015

Age (years) Male White (versus any other ethnicity) Current alcohol abuse Current tobacco smoking

HR

Le Moal et al.

(a)

(b) 300

P £ 0.0001*

Controls

75

Cases without ERT

50

P < 0.0001, by log-rank test 25 Cases with ERT 0 0

1

2

3 4 5 6 7 Time since baseline, years

8

9

200

100

0

–100

–200

–300 Cases with ERT (n = 41)

Cases without ERT (n = 69)

*At every time of follow-up from M3 to M36

Controls (n = 110)

Time after baseline 3 months 6 months 9 months 12 months 24 months 36 months

Figure 3. Risk and quantitative aspects of CD4+ depletion in patients who underwent ERT compared with patients who did not and controls. (a) Kaplan – Meier survival estimates for the percentage of patients who remained free of pre-defined CD4+ depletion, according to the group and use of ERT: patients with ERT, patients without ERT and controls. (b) Mean CD4+ changes over time in patients with or without ERT and controls.

respectively, P ¼ 0.75). At the last visit, these patients and controls demonstrated comparable outcomes regarding the use of cART (97% versus 99%, respectively, P ¼ 0.75), immuno-virological parameters (median CD4+ count: 476 versus 496 cells/mm3, respectively, P ¼ 0.26) and percentage with PVL ,50 copies/mL (90% versus 92%, respectively, P ¼0.8). Nevertheless, the patients who had been exposed to ERT still had a slightly lower CD4+ count than the non-ERT exposed patients and controls (median, 431, 540 and 496 cells/mm3, respectively, P ¼ 0.056 for overall comparisons) because of the persistence of a significant CD4+ decrease compared with the baseline values [DCD4+ count: 253, +163 and +82 cells/mm3, respectively, P, 0.05 (overall comparison)].

Discussion To the best of our knowledge, the present study is the first to specifically focus on the long-term impact of treating cancer in HIV-infected patients, encompassing both immuno-virological and clinical outcomes. Moreover, our results are reinforced by the fact that we compared the cancer cases (i.e. PLHIV and cancer) with well-matched controls (i.e. PLHIV without malignancies). Mainly, we found that radiotherapy (and not chemotherapy)

254

promptly induced a steep and protracted decline in the peripheral CD4+ count, whereas the patients generally were at increased risk of virological failure compared with the controls. Nevertheless, no harmful consequence (in terms of new ADIs) was observed in the patients undergoing appropriate prophylaxes. The deleterious effect of localized irradiation on circulating T cells was well described in the late 1980s in HIV-uninfected adults treated for various cancers.8,9 Irradiation transiently affects all circulating lymphocytes,8,9 but in the long term it affects CD4+ T cells, particularly naive cells.10 Two studies have reported that lymphocyte declines can last for up to 5– 11 years after radiotherapy.10,11 Radio-induced cell activation and apoptosis are likely to explain such rapid T cell depletion, whereas thymic involution (in adults) can explain its sustainability, particularly for naive cells.10 More recently, T cell decline has also been described in small studies of HIV-infected adults,12 – 14 but a single study reported longterm outcomes (up to 3 years) for CD4+ count and demonstrated results that were comparable to those for HIV-uninfected patients.15 The magnitude of CD4+ loss reported in most previous studies (regardless of HIV status) was 150 –200 cells/mm3, which is similar to our findings.10,12,13,15 Until the present report, it has not been clearly demonstrated whether radio-induced CD4+ T cell depletion clinically could lead to an increased risk of opportunistic infections, particularly in the context of PLHIV who are on

Downloaded from http://jac.oxfordjournals.org/ at University of Florida on February 16, 2015

Percentage without CD4+ depletion

100

Mean CD4+ change since baseline, per mm3

P £ 0.0001*

JAC

Impact of cancer treatment on HIV infection

decrease or to compare this decline with that in HIV-uninfected patients treated for the same malignancies. Compared with large, nationwide studies in PLHIV, we found similar patterns concerning the type and relative frequency of cancers.3,4,26 Nevertheless, our case collection was not exhaustive and probably missed some relevant cases, as we did not systematically approach all adult oncology departments or paediatric wards. Indeed, our cancer rate was slightly lower than those previously reported in France and the USA (i.e. 13 – 14/1000 personyears).4,26 Our goal was only to assess the impact of cancer on HIV course and not to describe the incidence and risk factors of cancers. Thus, we valued the completeness of medical records rather than the number of cases. Another limitation is that we cannot rule out that the CD4+ depletion was due to the cancer itself rather than the treatment that was received, in particular with patients with haematological malignancy. An increasing number of PLHIV face cancer because of ageing. Our results support the conclusion that these patients should be treated in a manner similar to HIV-uninfected people and certainly should not receive less vigorous treatment, particularly because nearly all deaths were directly attributable to the evolution of malignancy and not to ADIs. Nevertheless, irradiation strongly impacts CD4+ count for years and warrants special attention to prevent opportunistic infections. Such a CD4+ T cell decline is intriguing and warrants further investigation and better characterization.

Acknowledgements A portion of this work was presented at the Fourteenth European AIDS Conference, Brussels, Belgium, 2013 (Abstract #PE16/12). We are indebted to all study participants. We also thank American Journal Experts for English language editing.

Funding This work was supported by grants from COREVIH Centre—Poitou Charentes, France. The sponsor of the study had no role in: the design and the conduct of the study; the collection, management, analysis and interpretation of the data; or the preparation, review or approval of the manuscript. COREVIH funded the English language editing by American Journal Experts.

Transparency declarations No author of this study has any financial or personal relationships with people or organizations that could inappropriately influence this work, although some authors (G. L. M., T. Prazuck and L. H.) have, at some stage in the past, received funding from a variety of pharmaceutical companies for research, travel grants, speaking engagements or consultancy fees. American Journal Experts edited this manuscript for English language.

Author contributions L. H. and T. Prazuck had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. G. L. M., T. Prazuck and L. H. participated in the study conception and design. G. L. M., T. Prazuck, M. S.-R., J. J., T. Pasdeloup, J.-M. D., M. L., P. A. and L. H. included patients in the study. G. L. M. and L. B. collected

255

Downloaded from http://jac.oxfordjournals.org/ at University of Florida on February 16, 2015

a successful cART.16 We concluded that there was no increased risk for opportunistic infections when the prophylactic conditions aligned with the CD4+ count. The effect of SCC on T cell count is more controversial. Hakim et al.17 reported a profound decline in CD4+ (.CD8+) T cells for more than 1 year following chemotherapy in the context of HIV-uninfected patients with breast cancer. Conversely, Powles et al. 18 reported a rapid recovery in CD4+ count following the completion of chemotherapy in PLHIV treated for AIDS-related lymphoma and undergoing cART. These authors postulated that the constant thymic stimulation caused by cART19 was responsible for the rapid rise in CD4+ count following the cessation of chemotherapy. Moreover, Sankatsing et al. 15 concluded that ERT resulted in a significant and prolonged decline in CD4+ counts in PLHIV, whereas SCC did not significantly influence CD4+ cell count recovery. In our study, we found no significant association between SCC use and the incidence of a ≥25% CD4+ depletion in the multivariate analysis (after adjusting for the use of ERT). Assessing the specific decay of cells carrying HIV-DNA could be of particular interest. In fact, an increasing body of evidence indicates that a cure for HIV is an attainable goal.20,21 The combination of irradiation, systemic chemotherapy and bone marrow graft has been proven to dramatically reduce HIV reservoirs and to sometimes lead to viral control after cART cessation.22 – 24 In this context, PLHIV undergoing treatments for cancer could be a good model to assess the impact of various antitumoral strategies on the biomarkers of HIV reservoirs (as HIV-DNA) in a large number of patients, particularly as such strategies are not ethically evaluable in many patients without malignancy. We showed that cancer patients had an increased risk of experiencing virological failure compared with matched controls. The use of cART during SCC entails a number of potential complications, including pharmacokinetic interactions and overlapping toxicities that might compromise both cancer treatment and HIV-1 control.6 Indeed, as either inducers or inhibitors of the cytochrome P450 pathway, many antiretroviral drugs are metabolized through it. Thus, the co-administration of cART and SCC metabolized through this pathway can result in drug accumulation (and possible toxicity) or, in contrast, decreased concentration (and efficacy) of one or both drugs. Moreover, treatment compliance of PLHIV with cancer may be altered because of the increased side effects resulting from taking drugs for both HIV and cancer. Together, these constraints could explain part of the virological failure that we observed in our patients. Unfortunately, because the data collection was retrospective, we are unable to confirm these hypotheses in the absence of specific drug monitoring or adherence measures. Interestingly, multivariate analysis demonstrated that belonging to the cancer group was no longer associated with virological failure, whereas the class regimen of the third agent was a major predictor of success, along with the duration of virological control before baseline. Recently, Torres et al. 25 reported that the use of NNRTIs or an integrase strand transfer inhibitor (raltegravir) as the third agent was associated with better efficacy and safety and lower levels of drug – drug interactions during the treatment of cancer than regimens containing PIs. We found similar results concerning the use of NNRTIs, but we had insufficient data to specifically draw any conclusions regarding raltegravir. This study has some limitations. Because of its retrospective nature, we were unable to explore the mechanism of CD4+

Le Moal et al.

clinical data. E. C.-O. reviewed the anticancer treatment options. L. H. and T. Prazuck analysed and interpreted the data. G. L. M. and L. H. wrote the manuscript. All authors reviewed this paper, revised it for content and approved the final version.

12 Siraprapasiri P, Tharavichitkul E, Suntornpong N et al. Effect of radiation therapy to immunological and virological status in HIV/AIDS-cancer patients, preliminary report. J Med Assoc Thai 2011; 94 Suppl 2: S88– 93.

Supplementary data

14 Schreiber D, Chhabra A, Rineer J et al. Outcomes and tolerance of human immunodeficiency virus-positive U.S. veterans undergoing dose-escalated external beam radiotherapy for localized prostate cancer. Clin Genitourin Cancer 2014; 12: 94– 9.

Figure S1 and Table S1 are available as Supplementary data at JAC Online (http://jac.oxfordjournals.org/).

References

13 Alfa-Wali M, Allen-Mersh T, Antoniou A et al. Chemoradiotherapy for anal cancer in HIV patients causes prolonged CD4 cell count suppression. Ann Oncol 2012; 23: 141–7.

15 Sankatsing SUC, Hillebregt MMJ, Gras L et al. Prolonged decrease of CD4+ T lymphocytes in HIV-1-infected patients after radiotherapy for a solid tumor. J Acquir Immune Defic Syndr 2013; 62: 546– 9.

17 Hakim FT, Cepeda R, Kaimei S et al. Constraints on CD4 recovery postchemotherapy in adults: thymic insufficiency and apoptotic decline of expanded peripheral CD4 cells. Blood 1997; 90: 3789 –98.

3 Grulich AE, van Leeuwen MT, Falster MO et al. Incidence of cancers in people with HIV/AIDS compared with immunosuppressed transplant recipients: a meta-analysis. Lancet 2007; 370: 59– 67.

18 Powles T, Imami N, Nelson M et al. Effects of combination chemotherapy and highly active antiretroviral therapy on immune parameters in HIV-1 associated lymphoma. AIDS 2002; 16: 531–6.

4 Lanoy E, Spano J-P, Bonnet F et al. The spectrum of malignancies in HIV-infected patients in 2006 in France: the ONCOVIH study. Int J Cancer 2011; 129: 467–75. 5 Roussillon C, He´nard S, Hardel L et al. Causes of death among HIV-infected patients in France in 2010. ANRS EN20 Mortalite´ 2010 Survey. Bulletin Epide´miologique Hebdomadaire 2012; 46– 47: 541–5.

16 Housri N, Yarchoan R, Kaushal A. Radiotherapy for patients with the human immunodeficiency virus: are special precautions necessary? Cancer 2010; 116: 273–83.

19 Douek DC, McFarland RD, Keiser PH et al. Changes in thymic function with age and during the treatment of HIV infection. Nature 1998; 396: 690–5. 20 Purcell DF, Elliott JH, Ross A-L et al. Towards an HIV cure: science and debate from the International AIDS Society 2013 symposium. Retrovirology 2013; 10: 134.

6 Makinson A, Pujol J-L, Le Moing V et al. Interactions between cytotoxic chemotherapy and antiretroviral treatment in human immunodeficiency virus-infected patients with lung cancer. J Thorac Oncol 2010; 5: 562–71.

21 Katlama C, Deeks SG, Autran B et al. Barriers to a cure for HIV: new ways to target and eradicate HIV-1 reservoirs. Lancet 2013; 381: 2109 –17.

7 Yeni P. Prise en charge me´dicale des personnes infecte´es par le VIH. Paris: La documentation franc¸aise, 2010.

22 Hu¨tter G, Nowak D, Mossner M et al. Long-term control of HIV by CCR5 delta32/delta32 stem-cell transplantation. New Engl J Med 2009; 360: 692–8.

8 Newman GH, Rees GJ, Jones RS et al. Changes in helper and suppressor T lymphocytes following radiotherapy for breast cancer. Clin Radiol 1987; 38: 191–3. 9 Yang SJ, Rafla S, Youssef E et al. Changes in T-cell subsets after radiation therapy. Radiology 1988; 168: 537–40. 10 De Ruysscher D, Waer M, Vandeputte M et al. Changes of lymphocyte subsets after local irradiation for early stage breast cancer and seminoma testis: long-term increase of activated (HLA-DR+) T cells and decrease of ‘naive’ (CD4-CD45R) T lymphocytes. Eur J Cancer 1992; 28A: 1729 –34. 11 Rotstein S, Blomgren H, Petrini B et al. Long term effects on the immune system following local radiation therapy for breast cancer. I. Cellular composition of the peripheral blood lymphocyte population. Int J Radiat Oncol Biol Phys 1985; 11: 921–5.

256

23 Henrich TJ, Hu Z, Li JZ et al. Long-term reduction in peripheral blood HIV type 1 reservoirs following reduced-intensity conditioning allogeneic stem cell transplantation. J Infect Dis 2013; 207: 1694 –702. 24 Hawley T, Spear M, Guo J et al. Inhibition of HIV replication in vitro by clinical immunosuppressants and chemotherapeutic agents. Cell Biosci 2013; 3: 22. 25 Torres HA, Rallapalli V, Saxena A et al. Efficacy and safety of antiretrovirals in HIV-infected patients with cancer. Clin Microbiol Infect 2014; doi:10.1111/1469-0691.12589. 26 Yanik EL, Tamburro K, Eron JJ et al. Recent cancer incidence trends in an observational clinical cohort of HIV-infected patients in the US, 2000 to 2011. Infect Agents Cancer 2013; 8: 18.

Downloaded from http://jac.oxfordjournals.org/ at University of Florida on February 16, 2015

1 Schwarcz L, Chen M-J, Vittinghoff E et al. Declining incidence of AIDS-defining opportunistic illnesses: results from 16 years of populationbased AIDS surveillance. AIDS 2013; 27: 597– 605. ¨ nen NF, Overton ET. A review of premature frailty in HIV-infected 2 O persons; another manifestation of HIV-related accelerated aging. Curr Aging Sci 2011; 4: 33– 41.

Immuno-virological and clinical impacts of treating cancer in patients living with HIV.

Patients living with HIV (PLHIV) are increasingly being affected by cancer. However, data evaluating the long-term impact of cancer treatment on HIV c...
317KB Sizes 0 Downloads 5 Views