Abstract
Background
HIV infected persons have a two to five-fold increased unadjusted risk of lung cancer. In the National Lung Screening Trial (NLST), computed tomography (CT) screening was associated with a reduction in lung cancer mortality among high-risk smokers. These results may not generalize to HIV infected persons, particularly if they are more likely to have false-positive chest CT findings.
Methods
We utilized data including standardized chest CT scans from 160 HIV infected and 138 uninfected Veterans enrolled between 2009-2012 in the multicenter Examinations of HIV Associated Lung Emphysema (EXHALE) Study. Abnormal CT findings were abstracted from clinical interpretations of the scans and classified as positive by NLST criteria vs. other findings. Clinical evaluations and diagnoses that ensued were abstracted from the medical record.
Results
There was no significant difference by HIV in the proportion of CT scans classified as positive by NLST criteria (29% of HIV infected and 24% of HIV uninfected, p=0.3). However, HIV infected participants with CD4 counts <200 cells/mm3 had significantly higher odds of positive scans, a finding that persisted in multivariable analysis. Evaluations triggered by abnormal CT scans were also similar in HIV infected and uninfected participants (all p>0.05).
Conclusion
HIV status was not associated with an increased risk of abnormal findings on CT or increased rates of follow-up testing in clinically stable outpatients with CD4 cell count >200. These data reflect favorably on the balance of benefits and harms associated with lung cancer screening for HIV infected smokers with less severe immunodeficiency.
Keywords: HIV, lung cancer, non-AIDS malignancies, lung cancer screening, lung nodules
Background
Similar to the general population, lung cancer is now the leading cause of cancer death in HIV infected persons.[1, 2] Compared to HIV uninfected persons, there is a two- to five-fold unadjusted increase in the risk of lung cancer in persons infected with HIV.[3-7] Although some of this excess risk is attributed to higher smoking rates,[6, 8, 9] elevated lung cancer risks in HIV infected persons persist even after controlling for smoking and is increased among those with low CD4 cell count.[3, 4, 6] These data suggest that HIV infection is an independent risk factor for lung cancer.[3-6]
The National Lung Screening Trial (NLST) recently demonstrated a reduction in lung cancer mortality associated with computed tomography (CT) lung cancer screening in heavy smokers from the general population.[10] As a result, the National Comprehensive Cancer Network and other national organizations have published guidelines recommending low-dose CT (LDCT) screening in patients at high risk for lung cancer.[11-13] Additionally, lung cancer screening with LDCT has been adopted by some private health insurers and the Veterans Affairs Health System.[14, 15] As HIV infected heavy smokers may have more than twice the risk of lung cancer of HIV uninfected smokers,[3, 4] they may be a unique high-risk group that can be targeted for lung cancer screening interventions.
A potential concern in implementing widespread lung cancer screening is that approximately 20% of CTs have positive findings that require additional work-up, while only 1% of scans will reveal lung cancer.[16] Follow-up tests frequently include additional diagnostic CTs, but may include more invasive procedures such as fine needle aspiration or surgical biopsy that may lead to potentially severe complications.[17] As HIV infected patients are more likely to have a history of lung infections or other pulmonary diseases that may lead to structural lung changes, positive screening tests may be more common in HIV infected smokers.[18-21] The increased risk of lung cancer in HIV infected persons is less likely to affect the positivity rate, given the relatively low number of cancers expected to be detected by screening. Despite this, clinicians caring for HIV infected patients may be aware of the higher risk of lung cancer and other malignant and non-malignant lung diseases, and as a consequence be more likely to aggressively evaluate abnormal imaging findings. Therefore, determination of the rate of positive findings on chest CT scanning in HIV infected persons and the subsequent follow-up evaluations of these findings would provide important information on the applicability of NLST data to HIV infected smokers.
In this study, we used data from a prospective cohort of asymptomatic HIV infected and uninfected Veterans, most with a significant smoking history, to compare the frequency of incidental chest CT findings, particularly pulmonary nodules, observed on chest CT scans obtained for research purposes. We then estimated the proportion of CT scans in HIV infected and HIV uninfected participants that would have been considered positive by NLST criteria. We compared the clinical evaluations triggered by positive CT scans in order to determine whether HIV infected patients were more likely to undergo additional diagnostic procedures.
Methods
Study Cohort
We used data from the Examinations of HIV Associated Lung Emphysema (EXHALE) cohort, a multi-center sub-study of the Veterans Aging Cohort Study (VACS).[22] EXHALE is an ongoing observational, longitudinal study conducted at four Veteran Affairs (VA) Medical Centers (VAMC), namely the Atlanta, Bronx, Houston and Los Angeles VAMC. HIV infected outpatients in VACS are approached for enrollment in EXHALE, and are block-matched to HIV negative subjects by current smoking status to achieve a sample with similar prevalence of current smoking. Those with chronic lung diseases (as noted by history or clinical examination) other than chronic obstructive pulmonary disease (COPD) or asthma are excluded, as are patients with acute respiratory infections or illness in the four weeks prior to the baseline measurements. However, patients with respiratory infection more than 4 weeks prior to enrollment were not excluded. Enrollment began in 2009 and is ongoing. As a part of the study protocol, all participants received baseline chest CT scans. These CT scans were interpreted by radiologists at participating sites and results were reported in the participant’s electronic medical record.
Study CT scans
Study CT scans were all obtained using a standardized acquisition protocol on multi-detector scanners that were calibrated across centers by scanning of a lung phantom. Radiation dose for EXHALE CT scans was ~3.5 millisievert (mSv) (as compared to 1.5 mSv for NSLT screening scans and 8 mSv for standard CTs).[23] All CT scans were read by clinical radiologists practicing at each site at the time they were obtained, with results communicated to the participants’ primary medical provider. As this study was not a formal screening protocol, all recommendations made by radiologists and any additional evaluation of findings by medical providers were based on their standard practice. Interpreting radiologists at some sites could access patients’ electronic medical records and were not blinded to their HIV status. No subsequent CT scans were included in the study protocol, and the performance of any follow-up imaging was at the discretion of patients’ medical providers.
Data Collection
Demographic data, current smoking status, and tobacco use history were collected as part of the baseline EXHALE survey. Data on comorbidities, including chronic obstructive pulmonary disease (COPD), asthma, hepatitis C infection, and prior infections and opportunistic illnesses were obtained from the VACS database and were based on International Classification of Disease, 9th Edition (ICD9) diagnostic codes using validated algorithms.[24] Laboratory values closest to the date of EXHALE baseline enrollment, including CD4 count, and HIV viral load, were collected from the VA electronic laboratory database. Combination anti-retroviral therapy (cART) use, as defined by the presence of a multidrug antiretroviral regimen filled by the pharmacy during a three-month baseline period, was assessed using pharmacy data.
CT scan reports were abstracted by a reviewer (KS) blinded to the HIV status of the study participants. Our primary outcome was the presence of a non-calcified nodule (NCN) ≥ 4mm in diameter or any findings suggestive of lung cancer (positivity as per NLST criteria) on baseline chest CT scan.[10] Data were collected on all pulmonary nodules described in the CT reports, including nodule size, number, and presence of calcification. Nodules that were reported without any size description were assumed to be less than 4mm in diameter. All nodules without reported size were present on CT scans with other NLST positive nodules, so this classification did not affect any study outcomes. Other outcomes of interest from the CT scan reports included other incidental findings such as emphysematous changes, pleural effusions, ground glass infiltrates, bronchiectasis, or granulomas. Adenopathy was noted if the radiologist described it as clinically significant or if pulmonary lymph nodes were described as greater than 10 mm. Recommendations for clinical or imaging follow-up made by the interpreting radiologist were also recorded.
We then reviewed the medical records of all participants to determine subsequent evaluations prompted by the scan results within 18 months following the study CT. In the clinical evaluations triggered by these CT scans, our outcomes of interest were follow-up examinations (including follow-up CT scans, Positron Emission Tomography (PET) scans, bronchoscopy, CT guided biopsies, and surgical biopsies) in the presence or absence of recommendations for follow-up in the CT scan reports, and eventual diagnoses made by these evaluations. In addition, we determined the number of primary care visits completed by the participants in the 6 months after their CT scan to determine if HIV infected participants were subject to more primary care contact.
Data Analysis
Demographics, smoking status and numbers of pack-year smoked were compared among HIV infected and uninfected participants using Chi squared or Fisher’s exact for categorical variables, t-tests for normally distributed continuous variables and the rank-sum test for non-normal continuous variables. The proportion of CT studies meeting NLST positivity criteria, other clinically significant findings, follow-up recommendations, and types of subsequent clinical evaluations were compared by HIV status using chi-square tests. After determining the rates of suspicious nodules in different, clinically relevant CD4 strata among HIV infected participants (namely >350, 200-350, and <200), we found that a CD4 count of 200 cells/mm3 was an important threshold for comparison. We therefore stratified the HIV infected participants by baseline CD4 count (<200 cells/mm3 versus ≥200 cells/mm3) and compared baseline characteristics among these two groups using the same methods as our analysis for the full cohort. We then compared the proportion of CT studies meeting NLST positivity by CD4 cell count strata in HIV infected participants using a chi-square test. We then fit a logistic regression model to determine predictors of a false positive CT scan, including HIV status (and CD4 cell count strata), age, race/ethnicity, sex, smoking status, and pack-years of smoking as covariates. As a sensitivity analysis we then limited this multivariable analysis to participants meeting NLST inclusion smoking history criteria (≥30 pack-years of smoking in current smokers or former smokers who quit within 15 years of study enrollment).
Based on the prevalence of pulmonary nodules observed among patients in the cohort, we estimated that the study had an 80% power to detect 15% difference in nodules meeting NLST criteria by HIV status at a 0.05 significance level. All analyses were performed using STATA 11 (College Station, TX).
Results
Our study included 299 participants, 54% (n=160) of whom were HIV infected. HIV infected participants were older (p=0.03) and more likely to be male (p<0.001) than HIV uninfected participants (Table 1). There was no difference in the distribution of race/ethnicity, smoking habits, or rates of baseline chronic lung disease between HIV infected and uninfected persons (p>0.05 for all comparisons). The majority of cohort subjects were either current or former smokers (>80%) with significant pack-year exposure (median 24.6 pack-years for current or former smokers in cohort). HIV infected participants were more likely to have been diagnosed with prior pulmonary infections including pneumonia (19% vs. 4%; p<0.001) and tuberculosis (7% vs 1%; p=0.007). Most HIV infected participants were prescribed cART and had well-controlled HIV viremia (Table 2; 80% for patients with CD4≥200 cells/mm3 versus 64% for patients with CD4 cell counts <200 cells/mm3; p=0.08).
Table 1.
Baseline Cohort Characteristics by HIV status with NLST Baseline Characteristics for Comparison
| HIV+ (n=160) | HIV- (n=139) | P-value** | NLST (n=53,454) | |
|---|---|---|---|---|
| Age, median (IQR) | 55 (50-59) | 52 (48-58) | 0.03 | 62 |
| Male, n, % | 157 (98) | 122 (88) | <0.001 | 31,532 (59) |
| Race/ethnicity, n, % | 0.1 | |||
| Black | 115 (72) | 89 (64) | 2,376 (4) | |
| White | 19 (12) | 28 (20) | 48,549 (91) | |
| Hispanic | 26 (16) | 22 (16) | 935 (2)* | |
| Other | 0 (0) | 0 (0) | 2,529 (5) | |
| Smoking Status, n, % | 0.5 | |||
| Never Smoker | 23 (14) | 26 (19) | 0 (0) | |
| Former Smoker | 34 (21) | 31 (22) | 27,677 (52) | |
| Current Smoker | 103 (64) | 82 (59) | 25,779 (48) | |
| Pack-years by self-report, median (IQR) | 26 (14-42) | 22 (10-38) | 0.4 | 48 (39-66) |
| COPD, n, % | 22 (14) | 24 (17) | 0.4 | 5137 (10) |
| Asthma, n, % | 32 (20) | 24 (18) | 0.5 | 5,228 (10) |
| Hepatitis C, n, % | 58 (36) | 27 (19) | 0.001 | -- |
| Previous Infections or Opportunistic Illnesses | ||||
| Prior pneumonia, n, % | 30 (19) | 6 (4) | <0.001 | -- |
| Tuberculosis, n, % | 11 (7) | 1 (1) | 0.007 | -- |
| Histoplasmosis, n, % | 2 (1) | 0 (0) | 0.2 | -- |
| Cryptococcal Disease, n, % | 1 (1) | 0 (0) | 0.4 | -- |
| Pneumocystis Jirovecii Pneumonia, n, % | 2(1) | 0 (0) | 0.2 | -- |
| Kaposi’s Sarcoma, n, % | 7 (4) | 0 (0) | 0.01 | -- |
| HIV Viral Load, median (IQR) | 48 (48-185) | -- | -- | |
| Baseline CD4 Count, n, % | ||||
| <200 cells/mm3 | 22 (14) | -- | -- | |
| ≥200 cells/mm3 | 138 (86) | -- | -- | |
| Anti-retroviral Therapy Current Use, % | 84.0 | -- | -- |
IQR: Interquartile range,
The NLST did not report ethnicity in combination with race, therefore Hispanic ethnicity is not mutually exclusive.
p-value values are for comparisons of HIV+ and HIV- study participants.
Table 2.
Baseline Cohort Characteristics for HIV+ Participants by CD4 Count Strata
| HIV+ CD4 < 200 (n=22) | HIV+ CD4 ≥ 200 (n=138) | P-value | |
|---|---|---|---|
| Age, median (IQR) | 56 (51-59) | 55 (50-59) | 0.9 |
| Male, n, % | 22 (100) | 135 (98) | 0.5 |
| Race/ethnicity, n, % | 0.9 | ||
| Black | 16 (73) | 99 (72) | |
| White | 2 (9) | 17 (12) | |
| Hispanic | 4 (18) | 22 (16) | |
| Smoking Status, n, % | 0.9 | ||
| Never Smoker | 4 (18) | 19 (14) | |
| Former Smoker | 4 (18) | 28 (21) | |
| Current Smoker | 14 (64) | 87 (65) | |
| Pack-years by self-report, median (IQR) | 27 (13-42) | 26 (14-42) | 0.8 |
| COPD, n, % | 2 (9) | 20 (15) | 0.5 |
| Asthma, n, % | 4 (18) | 28 (21) | 0.8 |
| Hepatitis C, n, % | 11 (50) | 47 (34) | 0.15 |
| Previous Infections or Opportunistic Illnesses | |||
| Prior pneumonia, n, % | 7 (32) | 23 (17) | 0.09 |
| Tuberculosis, n, % | 3 (14) | 8 (6) | 0.2 |
| Histoplasmosis, n, % | 0 (0) | 2 (2) | 0.6 |
| Cryptococcal Disease, n, % | 0 (0) | 1 (1) | 0.7 |
| Pneumocystis Jirovecii Pneumonia, n, % | 0 (0) | 2 (2) | 0.6 |
| Kaposi’s Sarcoma, n, % | 1 (5) | 6 (4) | 0.9 |
| HIV Viral Load <400 Copies/mL | 14 (64) | 111 (80) | 0.08 |
| Anti-retroviral Therapy Current Use, % | 22 (100) | 113 (82) | 0.03 |
CT Findings
Prevalent lung cancer was diagnosed by lung biopsy in 3 of the HIV infected and 1 of the HIV uninfected participants (2.0% vs. 0.7%; p=0.3). However, a substantially larger number of scans had incidental findings. The proportion of CT scans that met the positive criteria as defined in the NLST did not differ when comparing HIV infected to HIV uninfected persons (Table 3; 29% vs. 24%; p=0.3). HIV infected participants with baseline CD4 cell counts <200 cells/mm3 had a greater frequency of positive scans than HIV infected participants with CD4 counts ≥200 cells/mm3 (55% vs. 25%; p=0.008). There was no significant difference in other clinically significant findings on study CT scans including adenopathy and pleural effusions, but there was a trend towards more emphysematous changes in HIV infected participants (41% vs. 30%; p=0.05).
Table 3.
Findings on Computed Tomography by HIV Status
| HIV+ (n=160) | HIV- (n=139) | P-value | |
|---|---|---|---|
| CT Scans Meeting NLST Positivity Criteria, Overall, % | 29 | 24 | 0.3 |
| By Baseline CD4 Count, % | |||
| <200 cells/mm3 | 55 | -- | 0.008 |
| ≥200 cells/mm3 | 25 | -- | |
| Any Nodules, % | 48 | 48 | 0.9 |
| Number of Nodules, median, (IQR) | 2 (1-4) | 1 (1-3) | 0.2 |
| Granulomas, % | 24 | 18 | 0.2 |
| Lymphadenopathy, % | 13 | 6.5 | 0.1 |
| Carcinoma Suggested, % | 4 | 3 | 0.8 |
| Emphysematous Changes, % | 41 | 30 | 0.05 |
| Pleural Effusion, % | 0.0 | 0.7 | 0.5 |
| Ground Glass Infiltrates, % | 15 | 14 | 0.9 |
| Bronchiectasis, % | 6 | 6 | 0.8 |
IQR: Interquartile range.
In our multivariable analysis, being HIV infected with a CD4 count <200 cells/mm3 was independently associated with increased odds of a false positive scan (Table 4; odds ratio [OR]: 3.6, 95% confidence interval [CI]: 1.4-9.4), as was increasing age (OR 1.3 for each 5-year age increase, 95% CI: 1.0-1.6). In a sensitivity analysis using an identical model limited to participants meeting NLST inclusion criteria for smoking history, HIV infection with CD4 count <200 cells/mm3 was the only significant predictor of scan false positivity (OR: 4.9, 95% CI: 1.3-19.0).
Table 4.
Results of Logistic Regression Model Evaluating the Association of Predictors of Computed Tomography Scan False Positivity
| Characteristic | Multivariable Odds Ratio for False Positive Scan | 95% CI |
|---|---|---|
| HIV Status | ||
| Uninfected | -- | |
| Infected, Baseline CD4 <200 cells/mm3 | 3.1 | 1.2-8.2 |
| Infected, Baseline CD4 ≥200 cells/mm3 | 1,0 | 0.5-1.8 |
| Age* | 1.2 | 1.0-1.4 |
| Male Gender | 1.2 | 0.4-4.0 |
| Race/Ethnicity | ||
| White | -- | -- |
| Black | 1.4 | 0.6-2.9 |
| Hispanic | 0.6 | 0.2-1.3 |
| Smoking Status | ||
| Never Smoker | -- | |
| Former Smoker | 0.5 | 0.2-1.4 |
| Current Smoker | 1.3 | 0.5-3.0 |
| Pack-years of Smoking** | 1.0 | 0.9-1.1 |
5 year increments.
1 year increments.
CI: Confidence Interval.
Clinical Evaluation Following Study CT Scans
Clinical follow-up recommendations and patterns after study CT scan completion were similar among HIV infected and uninfected cohort participants (Table 5). Follow-up recommendations by the interpreting radiologists (23% vs. 30%; p=0.2) and subsequent follow-up completion rates (50% vs. 54%; p=0.8) were similar in the HIV infected and uninfected groups. HIV infected participants were more likely to have a routine medical visit in the 6 months after the study CT scan (89% vs. 63%; p<0.001). Positive study CT scans led to similar rates of follow-up procedures including subsequent CT scans, PET scans, and biopsies (p>0.05 for all comparisons). No bronchoscopies were performed in response to CT findings in either study group. All biopsies led to lung cancer diagnoses (all adenocarcinomas). Final diagnoses triggered by study CT scans did not differ by HIV status.
Table 5.
Follow-up Patterns After Study Computed Tomography Scan, Follow-up Testing Triggered by Computed Tomography Scan Results and Final Diagnoses by HIV Status
| HIV+ (n=160) | HIV-(n=139) | p value | |
|---|---|---|---|
| Follow-up Recommended by Radiologist, % | 23 | 30 | 0.2 |
| Follow-up Carried Out When Recommended, % | 50 | 54 | 0.8 |
| Follow-up Carried Out When Not Recommended, % | 8 | 10 | 0.8 |
| Previous CT Scan, n, % | 39 (24) | 23 (17) | 0.1 |
| Routine Medical Visits in 6 Months After CT, n, % | 142 (89) | 87 (63) | <0.001 |
| Number of Routine Visits if Performed, median | 2 (1-2) | 1 (1-2) | 0.03 |
| Actions Resulting From CT Scans, n, % | |||
| Follow-up CT Scan | 26 (16) | 31 (22) | 0.2 |
| CT Guided Biopsy | 1 (0.7) | 0 (0.0) | 0.4 |
| Positron Emission Tomography Scan | 5 (3) | 2 (1) | 0.3 |
| Surgical biopsy | 2 (1) | 1 (0.7) | 0.6 |
| Final Diagnoses, n, % | 0.3 | ||
| None | 138 (86) | 111 (80) | |
| Lung Cancer | 3 (2) | 1 (0.7) | |
| Inflammation, Infection, or Stable Pulmonary Nodules | 17 (11) | 24 (17) | |
| Non-cancer Pathology | 1 (0.70) | 1 (0.6) | |
| New or Progressed Lesions | 1 (0.6) | 2 (1) |
Discussion
HIV infected smokers are a unique high-risk group that may benefit from CT based lung cancer screening. However, chronic lung changes resulting from immunosuppression-related pulmonary infections have raised concerns as to whether the frequency of false positive tests may be higher in this patient group. In this study, we found a similar likelihood of pulmonary nodules meeting NLST criteria for a positive CT scan among asymptomatic HIV infected and uninfected persons. We also found similar patterns of clinical evaluation triggered by study CT scans, suggesting that follow-up may not be more aggressive in HIV infected patients with abnormal chest imaging findings. This provides preliminary evidence that lung cancer screening may have a favorable harm/benefit profile in certain HIV infected smokers.
A limited number of studies have systematically described imaging findings on chest CTs in HIV infected patients.[20, 25] Jasmer et al. reviewed 242 lung CTs (mostly obtained in the pre-cART era) describing the frequency of pulmonary nodules among other findings noted on imaging. In this study of younger patients (mean age 40), the authors found 36% of CTs had pulmonary nodules, however patients were often symptomatic from opportunistic infections at the time of their CT scans.[25] A more recent study described the prevalence of incidental findings on lung CTs obtained for the calculation of coronary artery calcification scores in a cohort of asymptomatic HIV infected patients. The proportion of patients with incidental findings requiring additional workup or medical referral was 43% in that cohort, the majority of whom were on stable cART.[20] Our study is the first to compare rates of incidental chest imaging findings in HIV infected persons to a matched, HIV uninfected control group, and we find similar rates of pulmonary nodules and other incidental findings in both groups. This is also the first study to investigate the applicability of lung cancer screening using NLST positivity criteria in HIV infected persons. In our cohort, HIV infected participants with CD4 counts ≥200 cells/mm3 had a similar frequency of positive scans to the NLST (25% vs. 24%),
Inflammation has been linked to lung cancer risk in HIV infected persons.[26] However, higher rates of previous pulmonary infections did not increase the rate of suspicious pulmonary nodules or eventual inflammatory or infectious final diagnoses in our HIV infected participants compared to uninfected participants. We did, however, observe higher rates of emphysematous changes on imaging in HIV infected patients. The development of emphysema has been tied closely to inflammation, especially in HIV infected patients[27], and these findings continue to support a potential role of inflammatory processes in the development of lung nodules in these patients. Emphysema is also a known risk factor for lung cancer. Future analyses will include standardized quantitative and semi-quantitative assessment of emphysema severity and other markers of inflammation and their associations with lung nodules.
The rate of false positive tests in asymptomatic smokers undergoing chest CT scanning is a critical determinant in establishing the harms and cost-effectiveness of lung cancer screening.[28, 29] Positive screening tests can lead to invasive procedures such as bronchoscopy, needle biopsy, and surgical lung biopsy, which can all have serious complications, including respiratory failure and death.[10] Moreover, the morbidity associated with diagnostic tests for lung nodules may potentially be higher in HIV+ patients.[30] Specifically, the increased risk of emphysema in HIV+ patients[31-33] may confer greater risk of pneumothorax, the most common serious complication of lung biopsy.[34] No published data exist regarding lung cancer screening with chest CT in HIV infected persons, and therefore our findings represent a preliminary evaluation of the safety of this potentially important screening modality. The overall efficacy and cost-effectiveness of lung cancer screening in HIV infected persons could be further explored using simulation modeling, and our findings are critical to inform such models.
We found that HIV infected participants with CD4 counts <200 cells/mm3 were more likely to have a positive CT scan, even after adjusting for smoking exposure. This finding is likely related to a higher risk of lung infections associated with significant immunosuppression. It is also possible that radiologists reviewing the study CTs were aware of patients’ CD4 count, potentially biasing the interpretations of the scans. This finding suggests that patients with CD4 counts <200 cells/mm3 may have higher lung cancer screening false positivity rates, and therefore may be subject to increased harms associated with screening. This finding deserves further investigation in prospective screening studies.
Clinicians caring for HIV infected persons may be aware of their increased risk of lung cancer. This may affect their pretest probability estimates for cancer when evaluating chest imaging. Despite this, the patterns of clinical evaluation triggered by CT findings in our study did not differ by HIV status. Although HIV infected participants had more frequent healthcare contact, abnormal CT findings were not investigated with more frequency. As radiologists often base their follow-up recommendations for pulmonary nodules on Fleischner society criteria,[35] which does not incorporate HIV infection as a prognostic factor, it is possible that clinicians in our study based their follow-up evaluations on recommendations made by the radiologists interpreting our CT scans. HIV infected participants were also more likely to have had prior chest CT scans, which may have allowed interpreting clinicians to determine the stability of existing nodules, thereby limiting the need for further evaluations.
Our study has several strengths. This was a multi-site, prospective study design with demographically similar HIV infected and uninfected individuals with comparable smoking exposure. The imaging studies performed in our protocol were similar to the low dose CT scans used in lung cancer screening trials. The study collected data on “real-world” processes and therefore represents how clinicians would follow-up lung cancer screening scans in HIV infected patients in actual practice. Our cohort was small compared to other studies of lung cancer screening, which may have limited our ability to find differences in the overall rates of suspicious nodules between HIV infected and uninfected participants. Another limitation of our study was the exclusion of persons with prior lung diseases, besides COPD and asthma. This criterion likely excluded more HIV infected persons than HIV uninfected, potentially affecting the generalizeability of our findings. However, patients with prior lung diseases were also excluded from the NLST, which improves the validity of the extrapolation of our findings to the screening results of that trial. Other limitations include the unavailability of details on follow-up studies performed at non-VA medical centers. Additionally, not all patients in our study were smokers, and therefore would not meet NLST inclusion criteria; in exploratory analyses limited only to smokers, however, we found similar results to those from the overall cohort. Last, this study used data from a longitudinal study of lung function and was not a formal lung cancer screening protocol. Although the study CTs were included in participants’ electronic medical record, there were no study procedures that dictated a protocol for follow-up of CT results. Rather, follow-up was at the discretion of the patient’s medical provider. This may have affected the rates of clinical evaluation of CT findings as well as the determination of final diagnoses in patients with abnormal scans.
HIV infected patients with CD4 counts >200 in our study did not have higher likelihood of findings meeting NLST criteria for a positive chest CT compared to uninfected controls, despite more previous lung infections, and a greater risk of lung cancer. These data suggest a similar frequency of positive screening scans among well controlled HIV-infected compared to HIV-uninfected persons, suggesting that lung cancer screening could be favorable in HIV infected smokers. The risk/benefit profile of lung cancer screening requires further evaluation in high-risk HIV-infected persons.
Acknowledgments
Funding:
This study was supported by the National Heart, Lung, and Blood Institute (R01HL090342 to KC), the National Cancer Institute (R01CA173754 to KC and JW), the National Center for Research Resources (KL2TR000069 to KS), and the National Institute on Alcohol and Alcohol Abuse (3U01AA13566). The funder provided funding for the study and salary support for investigators. The funder did not play any role in the data collection, analysis, interpretation or manuscript preparation.
Footnotes
Contributorship:
KS and KC participated in the design, data collection, analysis and manuscript writing. JW and SS participated in the analysis and manuscript writing. SB, AJ, JK, MRB, KA, DR and GSH all participated in the design, data collection, and manuscript writing portions of the study.
Competing Interests:
No authors report any relevant financial conflicts of interest to this study.
References
- 1.Simard EP, Engels EA. Cancer as a cause of death among people with AIDS in the United States. Clin Infect Dis. 2010;51:957–962. doi: 10.1086/656416. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Shiels MS, Pfeiffer RM, Gail MH, Hall HI, Li J, Chaturvedi AK, et al. Cancer burden in the HIV-infected population in the United States. J Natl Cancer Inst. 2011;103:753–762. doi: 10.1093/jnci/djr076. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Shiels MS, Cole SR, Mehta SH, Kirk GD. Lung cancer incidence and mortality among HIV-infected and HIV-uninfected injection drug users. J Acquir Immune Defic Syndr. 2010;55:510–515. doi: 10.1097/QAI.0b013e3181f53783. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Sigel K, Wisnivesky J, Gordon K, Dubrow R, Justice A, Brown ST, et al. HIV as an independent risk factor for incident lung cancer. AIDS. 2012;26:1017–1025. doi: 10.1097/QAD.0b013e328352d1ad. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Engels EA, Brock MV, Chen J, Hooker CM, Gillison M, Moore RD. Elevated incidence of lung cancer among HIV-infected individuals. J Clin Oncol. 2006;24:1383–1388. doi: 10.1200/JCO.2005.03.4413. [DOI] [PubMed] [Google Scholar]
- 6.Kirk GD, Merlo C, P OD, Mehta SH, Galai N, Vlahov D, et al. HIV infection is associated with an increased risk for lung cancer, independent of smoking. Clin Infect Dis. 2007;45:103–110. doi: 10.1086/518606. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Patel P, Hanson DL, Sullivan PS, Novak RM, Moorman AC, Tong TC, et al. Incidence of types of cancer among HIV-infected persons compared with the general population in the United States, 1992-2003. Ann Intern Med. 2008;148:728–736. doi: 10.7326/0003-4819-148-10-200805200-00005. [DOI] [PubMed] [Google Scholar]
- 8.Clifford GM, Lise M, Franceschi S, Egger M, Bouchardy C, Korol D, et al. Lung cancer in the Swiss HIV Cohort Study: role of smoking, immunodeficiency and pulmonary infection. Br J Cancer. 2012;106:447–452. doi: 10.1038/bjc.2011.558. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Lifson AR, Neuhaus J, Arribas JR, van den Berg-Wolf M, Labriola AM, Read TR. Smoking-related health risks among persons with HIV in the Strategies for Management of Antiretroviral Therapy clinical trial. Am J Public Health. 2010;100:1896–1903. doi: 10.2105/AJPH.2009.188664. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Aberle DR, Adams AM, Berg CD, Black WC, Clapp JD, Fagerstrom RM, et al. Reduced lung-cancer mortality with low-dose computed tomographic screening. N Engl J Med. 2011;365:395–409. doi: 10.1056/NEJMoa1102873. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Jaklitsch MT, Jacobson FL, Austin JH, Field JK, Jett JR, Keshavjee S, et al. The American Association for Thoracic Surgery guidelines for lung cancer screening using low-dose computed tomography scans for lung cancer survivors and other high-risk groups. J Thorac Cardiovasc Surg. 2012;144:33–38. doi: 10.1016/j.jtcvs.2012.05.060. [DOI] [PubMed] [Google Scholar]
- 12.Network NCC. Lung Cancer Screening Guidelines. 2011 [Google Scholar]
- 13.Bach PB, Mirkin JN, Oliver TK, Azzoli CG, Berry DA, Brawley OW, et al. Benefits and harms of CT screening for lung cancer: a systematic review. JAMA. 2012;307:2418–2429. doi: 10.1001/jama.2012.5521. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Journal WS. Wellpoint to cover lung CT scans for heavy smokers. 2011 [Google Scholar]
- 15.VVA.org. VA to Screen Vets for Lung Cancer in Early Stages. 2012 [Google Scholar]
- 16.Church TR, Black WC, Aberle DR, Berg CD, Clingan KL, Duan F, et al. Results of initial low-dose computed tomographic screening for lung cancer. N Engl J Med. 2013;368:1980–1991. doi: 10.1056/NEJMoa1209120. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Sigel K, Dubrow R, Silverberg M, Crothers K, Braithwaite S, Justice A. Cancer Screening in Patients Infected with HIV. Curr HIV/AIDS Rep. 2011 doi: 10.1007/s11904-011-0085-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Castaner E, Gallardo X, Mata JM, Esteba L. Radiologic approach to the diagnosis of infectious pulmonary diseases in patients infected with the human immunodeficiency virus. Eur J Radiol. 2004;51:114–129. doi: 10.1016/j.ejrad.2004.03.008. [DOI] [PubMed] [Google Scholar]
- 19.Gold JA, Rom WN, Harkin TJ. Significance of abnormal chest radiograph findings in patients with HIV-1 infection without respiratory symptoms. Chest. 2002;121:1472–1477. doi: 10.1378/chest.121.5.1472. [DOI] [PubMed] [Google Scholar]
- 20.Crum-Cianflone N, Stepenosky J, Medina S, Wessman D, Krause D, Boswell G. Clinically significant incidental findings among human immunodeficiency virus-infected men during computed tomography for determination of coronary artery calcium. Am J Cardiol. 2011;107:633–637. doi: 10.1016/j.amjcard.2010.10.026. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Crothers K, Thompson BW, Burkhardt K, Morris A, Flores SC, Diaz PT, et al. HIV-associated lung infections and complications in the era of combination antiretroviral therapy. Proc Am Thorac Soc. 2010;8:275–281. doi: 10.1513/pats.201009-059WR. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Justice AC, Dombrowski E, Conigliaro J, Fultz SL, Gibson D, Madenwald T, et al. Veterans Aging Cohort Study (VACS): Overview and description. Med Care. 2006;44:S13–24. doi: 10.1097/01.mlr.0000223741.02074.66. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Aberle DR, Berg CD, Black WC, Church TR, Fagerstrom RM, Galen B, et al. The National Lung Screening Trial: overview and study design. Radiology. 2011;258:243–253. doi: 10.1148/radiol.10091808. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Justice AC, Lasky E, McGinnis KA, Skanderson M, Conigliaro J, Fultz SL, et al. Medical disease and alcohol use among veterans with human immunodeficiency infection: A comparison of disease measurement strategies. Med Care. 2006;44:S52–60. doi: 10.1097/01.mlr.0000228003.08925.8c. [DOI] [PubMed] [Google Scholar]
- 25.Jasmer RM, Edinburgh KJ, Thompson A, Gotway MB, Creasman JM, Webb WR, Huang L. Clinical and radiographic predictors of the etiology of pulmonary nodules in HIV-infected patients. Chest. 2000;117:1023–1030. doi: 10.1378/chest.117.4.1023. [DOI] [PubMed] [Google Scholar]
- 26.Borges AH, Silverberg MJ, Wentworth D, Grulich AE, Fatkenheuer G, Mitsuyasu R, et al. Predicting risk of cancer during HIV infection: the role of inflammatory and coagulation biomarkers. AIDS. 2013;27:1433–1441. doi: 10.1097/QAD.0b013e32835f6b0c. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Petrache I, Diab K, Knox KS, Twigg HL, 3rd, Stephens RS, Flores S, Tuder RM. HIV associated pulmonary emphysema: a review of the literature and inquiry into its mechanism. Thorax. 2008;63:463–469. doi: 10.1136/thx.2007.079111. [DOI] [PubMed] [Google Scholar]
- 28.Goulart BH, Bensink ME, Mummy DG, Ramsey SD. Lung cancer screening with low-dose computed tomography: costs, national expenditures, and cost-effectiveness. J Natl Compr Canc Netw. 2012;10:267–275. doi: 10.6004/jnccn.2012.0023. [DOI] [PubMed] [Google Scholar]
- 29.Black C, Bagust A, Boland A, Walker S, McLeod C, De Verteuil R, et al. The clinical effectiveness and cost-effectiveness of computed tomography screening for lung cancer: systematic reviews. Health Technol Assess. 2006;10:iii–iv. ix–x, 1–90. doi: 10.3310/hta10030. [DOI] [PubMed] [Google Scholar]
- 30.Hooker CM, Meguid RA, Hulbert A, Taylor JT, Shin J, Wrangle J, et al. Human immunodeficiency virus infection as a prognostic factor in surgical patients with non-small cell lung cancer. Ann Thorac Surg. 2012;93:405–412. doi: 10.1016/j.athoracsur.2011.11.012. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Diaz PT, King MA, Pacht ER, Wewers MD, Gadek JE, Nagaraja HN, et al. Increased susceptibility to pulmonary emphysema among HIV-seropositive smokers. Ann Intern Med. 2000;132:369–372. doi: 10.7326/0003-4819-132-5-200003070-00006. [DOI] [PubMed] [Google Scholar]
- 32.Kuhlman JE, Knowles MC, Fishman EK, Siegelman SS. Premature bullous pulmonary damage in AIDS: CT diagnosis. Radiology. 1989;173:23–26. doi: 10.1148/radiology.173.1.2781013. [DOI] [PubMed] [Google Scholar]
- 33.Morris A, George MP, Crothers K, Huang L, Lucht L, Kessinger C, Kleerup EC. HIV and chronic obstructive pulmonary disease: is it worse and why? Proc Am Thorac Soc. 2011;8:320–325. doi: 10.1513/pats.201006-045WR. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Wiener RS, Schwartz LM, Woloshin S, Welch HG. Population-based risk for complications after transthoracic needle lung biopsy of a pulmonary nodule: an analysis of discharge records. Ann Intern Med. 2011;155:137–144. doi: 10.1059/0003-4819-155-3-201108020-00003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Masciocchi M, Wagner B, Lloyd B. Quality review: Fleischner criteria adherence by radiologists in a large community hospital. J Am Coll Radiol. 2012;9:336–339. doi: 10.1016/j.jacr.2011.12.026. [DOI] [PubMed] [Google Scholar]