Screening HIV-positive people for lung cancer with low-dose computed tomography (LDCT) yielded a positive-image rate similar to that in the general population, according to the results of a 901-person study in Denmark. CT images proved positive in almost 10% of high-risk HIV patients, and nearly 3% had lung cancer diagnosed.
Lung cancer remains the leading cause of cancer deaths worldwide and in the U.S. and has become the most prevalent non-AIDS cancer in some HIV populations. U.S. experts recommend annual lung cancer screening with LDCT for current or former smokers in the general population. But, whether the same advice should extend to HIV-positive smokers remains unclear because the potentially high prevalence of nodules in this group may lead to inappropriate, risky interventions.
In a large HIV group receiving LDCT, researchers in Denmark aimed to determine lung cancer prevalence, positive CT images and related risk factors. The Copenhagen Comorbidity in HIV Infection Study (COCOMO) is an observational prospective analysis of non-AIDS comorbidities in people with HIV infection. This study involved people who had LDCT read by two radiologists, who recommended follow-up for cases of suspected malignancy. Participants’ clinicians decided whether to repeat imaging or perform invasive procedures.
Primary outcomes were a positive LDCT image and histologically proven lung cancer. Researchers assessed outcomes for the entire study group and for high-risk patients, defined as those 50 to 74 years old who currently or formerly smoked cigarettes at a rate of more than 30 pack-years.
Of 901 HIV-positive people who had chest LDCT, 113 (12.5%) fit into the high-risk group. Median ages were 50.4 years in the entire group and 60.3 years in the high-risk cluster. Most study participants were men (86.7%) and white (86.2%), and only 4.9% had a viral load above 50 copies/mL. Current smokers made up 27.9% of the study population, and previous smokers made up 35.6%.
While 3.1% of the whole cohort had a positive LDCT image, 9.7% of the high-risk group had a positive image. Through a median follow-up of 14.6 months in the 28 people with a positive image, three (all in the high-risk group) had a confirmed lung cancer diagnosis, representing 0.3% of the entire population and 2.7% of the high-risk group. Two other people were diagnosed with nonpulmonary cancers, and 12 had clinically significant LDCT findings not suggesting malignancy.
Nine people underwent invasive diagnostic procedures, including seven bronchoscopies with transbronchial biopsies, one CT-guided transthoracic fine-needle aspiration biopsy, one ultrasonically guided needle biopsy and five video-assisted thoracic surgeries. Pneumothorax developed during three procedures, but all cases were self-resolving and required no intervention. The three diagnosed lung cancers all had surgical treatment, with no relapses during follow-up.
Logistic regression analysis adjusted for age and cumulative smoking identified three independent predictors of a positive image: current CD4 count below 500 cells/mm3 (odds ratio [OR] 2.32, 95% confidence interval [CI] 1.01 to 5.13, P = .04), nadir CD4 count below 200 cells/mm3 (OR 2.63, 95% CI 1.13 to 6.66, P = .03) and a history of Pneumocystis pneumonia (OR 4.32, 95% CI 1.34 to 11.9, P = .01).
The researchers calculated that 38 LDCTs were needed to detect one lung cancer. They noted that the 9.7% positive-LDCT rate in the high-risk study population is comparable to screening results in the Danish general population. A recent French study of lung cancer screening in people with HIV had a comparable lung cancer diagnosis rate of 2% but a higher positive-image rate of 20%. The Danish team added that no interventional trials have assessed the value of LDCT screening in HIV populations, “and barriers to successful implementation of LDCT screening may be different” in this group.
By Mark Mascolini