Keith M. Sigel

Keith M. Sigel, M.D., Ph.D., M.P.H.
Mount Sinai Medical Center — New York, New York
Assistant Professor of Medicine
Division of General Internal Medicine
Division of Infectious Diseases
(Courtesy of The Center for AIDS Information & Advocacy)

Highly regarded for his contributions to cancer research in people with HIV, Dr. Sigel is Assistant Professor of Medicine in the Division of General Internal Medicine and the Division of Infectious Diseases at the Icahn School of Medicine at Mount Sinai in New York City. With a focus on two non-AIDS-defining malignancies — lung cancer and anal cancer — Dr. Sigel has led numerous innovative studies on these conditions and others in people with HIV. His work addresses unique aspects of HIV-related malignancy, including risk, prevention, treatment, and prognosis. He cochairs the Cancer Core of the Veterans Aging Cohort Study, a large longitudinal comparison of HIV-positive and -negative veterans with equivalent access to care and a high lung cancer burden. With an MD and a Masters in Public Health from the University of North Carolina and a PhD from Mount Sinai, Dr. Sigel devotes time to the primary care of people with HIV infection and hepatitis C infection. His many speaking invitations include talks at Yale University, the University of Massachusetts, and the University of North Carolina.

Mark Mascolini: Is suspicion of lung cancer high enough among HIV clinicians?

Keith Sigel: In general I think many HIV clinicians are aware of the prominence of non-AIDS-defining cancers in the HIV population. Several large cohort studies have shown that non-AIDS-defining cancers have become a leading source of morbidity and mortality in the HIV population.1-3 And it’s hard for any HIV clinician to ignore the persistently high rates of smoking in the US HIV population.

I’m not aware of any literature that specifically addresses HIV clinicians’ awareness of the higher risk of lung cancer with HIV, but my suspicion is that clinicians across the country have become highly aware of this risk. When I go to national meetings, lung cancer gets good coverage in well-attended sessions. Because HIV clinicians function in a specialty area with lots of intragroup communication, I do think they tend to be aware of the threat of lung cancer.

Lung Cancer Incidence and Risk

Mascolini: Is lung cancer incidence flat or falling in people with HIV, and what are the factors determining this trend?

Sigel: A study from the Kaiser Permanente healthcare system a few years ago looked at trends in lung cancer incidence over the antiretroviral era and found essentially a flat curve as we’ve had more time in the antiretroviral era.4 That being said, the HIV cohort in the United States is aging, so much of the current incidence trend reflects the graying of the HIV population and the increasing lung cancer risk with aging.

But lung cancer incidence rates are still high in relation to the general population,4-6 and I haven’t seen any data to refute that. There’s been some interesting data showing that, particularly in the early combination antiretroviral therapy era, immunosuppression with HIV was driving some portion of the excess in lung cancer incidence seen in people with HIV. I think the impact of immunosuppression has lessened and aging has become a big factor.

Mascolini: Smoking and prior lung disease are clear risk factors for lung cancer in people with HIV. But evidence on nadir or current CD4 count seems more complicated. How do you interpret those data?

Sigel: This is something I’ve spent a lot of time looking at, and there’s a decent amount of data out of Europe addressing this question as well.5,7,8 The relationship between CD4 count and lung cancer risk is complicated because looking at a single CD4 measurement in time can lead to the conclusion that CD4 count is not related to lung cancer risk or is very weakly related. It does seem that longitudinal trends — long periods of low CD4 count or low CD4/CD8 ratio — appear to have a much stronger association with lung cancer risk.

I published a paper from the Veterans Aging Cohort Study (VACS) in the US that showed a much stronger relationship between long periods of low CD4 count and lung cancer risk,9 and research on the large French HIV cohort found similar effects.5,7 These findings are not a big surprise because the immune system is responsible for immune surveillance for lung cancer. We always suspected that some degree of immunosuppression was going to be related to some portion of the excess lung cancer risk with HIV. So I do believe, based on the larger published studies, that longer periods with low CD4 count do play a role in lung cancer risk. It would be great to create an HIV-specific lung cancer risk index using some of that information. But such an index has not emerged yet.

Mascolini: In that VACS study you found an independent association between lower CD4/CD8 ratio and lung cancer risk.9 What’s driving that particular association?

Sigel: CD4/CD8 ratio is an interesting measure. The ratio was used a lot more clinically earlier in the HIV epidemic. But it’s gotten much more attention in the past few years because it does appear to be a strong marker of continued immune dysfunction during HIV suppression. You can have patients who are reasonably adherent to antiretroviral therapy and who may even return to a normal CD4 count. But some people nevertheless maintain a low CD4/CD8 ratio, and it’s not clear why.

We found a lower CD4/CD8 ratio was actually the strongest immunologic predictor of lung cancer incidence.

There’s some suggestion that a persistently low CD4/CD8 ratio reflects later initiation of antiretroviral therapy. But it certainly seems to be a marker of many poor outcomes. Several studies show that low CD4/CD8 ratio is associated with mortality in general as well as with non-AIDS-defining cancer risk.10,11 In the Veterans cohort we were able to look at it specifically with lung cancer because we have so many lung cancer cases.9 We found a lower CD4/CD8 ratio was actually the strongest immunologic predictor of lung cancer incidence. That finding is really interesting because it underlines the possibility that overall immune dysfunction — not just immunodeficiency — plays a role in lung cancer development. Whether the CD4/CD8 ratio could be a biomarker included in lung cancer risk assessments is still not clear. We still have this issue of whether or not a single measurement in time (as opposed to a longitudinal average) is an effective lung cancer predictor.

Mascolini: Why are prior pneumonia and chronic obstructive pulmonary disease (COPD) risk factors for lung cancer in people with HIV?

Sigel: COPD is an acknowledged risk factor for lung cancer in the general population because it signifies an inflammatory state in the lung that lies on the pathway to the local dysfunction required to create the mutations that drive lung cancer. So it’s not a surprise that COPD emerged as an independent lung cancer risk factor in HIV-infected patients.

We also have data suggesting that HIV is associated with a higher risk of COPD.12 And one thing that we’ve been trying to determine with spirometry data in the VACS is whether COPD is associated with excess lung cancer risk in people with HIV. We’re presenting a study at the American Thoracic Society in a few months looking at that issue. We found that COPD seems to be an equal-magnitude risk factor for lung cancer in both HIV-infected and uninfected veterans. So it appears to be an independent risk factor with HIV but doesn’t seem to be a stronger risk factor with HIV, which is something that hadn’t been assessed before.

Prior pneumonia also has some association with lung cancer in the general population. It’s a difficult exposure to untangle epidemiologically because sometimes pneumonia is the initial presenting factor for lung cancer. But our VACS study9 and others assessed the impact of pneumonias much more distant to the cancer diagnosis and did find a positive association. The suspicion has been that pneumonia may trigger a strong inflammatory injury that then spurs oncogenic mutations as one of the hits needed to cause cancer.

Lung Cancer Screening Advice in HIV Patients

Mascolini: There are lung cancer screening guidelines for the general population13 but not specifically for people with HIV. When do you consider screening people with HIV?

Sigel: I generally stick to the National Lung Screening Trial (NLST) inclusion criteria,14 which informed the Preventive Services Task Force guidelines13 and Centers for Medicare and Medicaid Services (CMS) guidelines. Those low-dose CT criteria are to start screening at age 55 in persons with a 30 pack-year smoking history. This includes current smokers or former smokers who quit within 15 years. With HIV infection, deciding on screening gets complicated because it’s hard to get your screening scan covered by insurance if you don’t meet these national guidelines.13 Clinicians will have trouble if they try to screen outside of those guidelines just because there’s such limited data on screening HIV-infected patients.

Until we have stronger evidence, I think it’s difficult to make a strong suggestion that we veer from those [general-population lung cancer screening] criteria.

One common complicating issue is that we have consistent epidemiologic evidence suggesting lung cancer emerges at an earlier age in people with HIV — on average about 3 to 5 years earlier.15 More limited evidence suggests that lung cancer develops with less smoking in people with HIV. This has prompted researchers and clinicians to ask whether we should be screening HIV patients with a little bit less than a 30-pack-year history and at an earlier age.

I don’t think we have a great answer to that question yet. We have a study coming out in AIDS in which we used a highly validated lung cancer simulation — the Lung Cancer Policy Model — to try to determine appropriate criteria for screening for lung cancer in people with HIV.16 What we came up with is that the most efficient solution based on available data is the NLST criteria14,15 — yearly screening starting at age 55, 30 pack-years, current and former smokers within 15 years. Until we have stronger evidence, I think it’s difficult to make a strong suggestion that we veer from those criteria.

Lung Cancer Treatment and Survival Disparities

Mascolini: Should clinicians use the same criteria to determine whether people should get treated for lung cancer regardless of HIV status?

Sigel: This is an area that has even less data than the screening question we just considered. We know there are lung cancer treatment disparities in people with HIV. Several large national studies suggest lower receipt of stage-appropriate treatment in HIV populations.17,18

Our own recent analysis of practice within the Veterans Administration system — where care is accessible regardless of HIV status — found very limited lung cancer treatment disparities by HIV status,19suggesting that a lot of these treatment disparities are probably structural. Gita Suneja from Duke did an interesting study a few years ago in which she surveyed US oncologists nationally about their perspective on treating HIV-infected patients with non-AIDS-defining cancer, and 20% of respondents said they do change practice when treating HIV-infected patients.20 So we do need more information in this area.

Table 1. Two Studies of Chemotherapy and Surgery for Lung Cancer With HIV
Chemotherapy study: Makinson et al.21

  • Design: Retrospective analysis of non-small-cell lung cancer patients in large French HIV cohort
  • Study period: 1996-2008
  • Participants: 52 people, 81% men, median age 48; 40 in toxicity substudy, 78% men, median age 50
  • Toxicity: 14 of 68 chemotherapeutic combinations (21%) complicated by grade 4 hematologic toxicity in 13 people (33%); 6 deaths (15%) due to hematologic toxicity
  • Toxicity factor: Protease inhibitor therapy associated with 5-fold higher odds of grade 4 hematologic toxicity

Surgery study: Sigel et al.22

  • Design: Linkage of VACS cancer data to Veterans Administration surgical database
  • Study period: 2000-2016
  • Participants: 424 early-stage lung resection patients, 151 with HIV, 273 without HIV; mean age 60 in both groups; no differences by sex, race, year of surgery/cancer diagnosis, surgical risk class, cancer stage or histologic type
  • Outcomes: Frequency of any complication did not differ by HIV status; no difference in 30-day mortality (2% with HIV); 180-day mortality nonsignificantly higher with HIV (11% versus 6%, P = 0.07)
In patients with well-controlled HIV, if you pay attention to potential antineoplastic drug interactions … you should be able to treat lung cancer as you ordinarily would.

Randomized controlled trial data on lung cancer therapy specific to HIV-infected patients are very limited. What data there are suggest lung cancer treatment is well tolerated and does not seem to cause excess harm in people with HIV. But there are some observational data from Alain Makinson in France on a large series of HIV-infected patients who received chemotherapy, and there was a high rate of serious complications in that HIV cohort (Table 1).21 So the question about appropriate lung cancer therapy for people with HIV is difficult to answer. My personal feeling is that in patients with well-controlled HIV, if you pay attention to potential antineoplastic drug interactions with antiretroviral therapy and other medications, you should be able to treat lung cancer as you ordinarily would.

In 2017 we presented data at the International Conference on Malignancies in HIV/AIDS looking at HIV-infected and uninfected patients with early-stage lung cancer who were undergoing surgery for their early-stage lung cancers.22 We found that the rate of surgical complications was identical by HIV status (Table 1), suggesting that lung-cancer surgery is very well tolerated in this group. So in terms of surgical treatment, in this era there should be little reluctance to treat, almost as if the HIV wasn’t there. We need more information on chemotherapy.

Mascolini: Is antiretroviral therapy improving the lung cancer survival disadvantage reported in people with HIV?

Sigel: I do think so. The association of HIV with poorer lung cancer survival has been complicated for a few reasons. Number one, a lot of the data is still from the early 2000s, when people were not having HIV-related outcomes as good as they do now. So competing risks were influencing the poor prognosis we saw in many of the bigger studies.

I also think that treatment disparities, which can be difficult to measure, were playing a role in survival outcomes. Some HIV-infected patients were not getting optimal lung cancer treatment and that led to the appearance of worse lung cancer outcomes. I presented survival data last year at CROI looking within the Veterans Administration, where there are many fewer lung cancer treatment disparities by HIV status than in the general population.19 In the latest study period, 2009-2015, we found no difference in lung cancer survival by HIV status in a large group of lung cancer patients, almost 600 with HIV and almost 900 without HIV. So I think antiretroviral therapy does mitigate the poorer-prognosis issues identified in some older studies.

Lung Cancer Pitfalls and Opportunities

Mascolini: What’s the biggest mistake HIV clinicians make in lung cancer management?

Sigel: I think it’s reluctance to offer stage-appropriate treatment. Based on our data,22 I definitely think there should be very limited reluctance to refer for surgery, which for early-stage lung cancer can be curative. Make that referral — don’t let the HIV stand in your way. In terms of prevention, the two big potential areas of benefit are focusing on smoking cessation and screening patients who meet national screening guidelines.13

Mascolini: Are there other issues we haven’t addressed that you would like to raise?

Sigel: There’s a big one: the role of immunotherapy. The question is whether immunotherapy is appropriate for HIV-infected patients with lung cancer. I’m not an oncologist, I’m an HIV clinician and infectious disease specialist. So I don’t treat cancer. But I’ve followed the immunotherapy story closely because it’s very interesting in relationship to lung cancer with HIV.

The checkpoint inhibitors are becoming very important tools,23,24 particularly in certain advanced lung cancer cases. They target PD-1, which has been an important immune exhaustion marker in HIV for a long time. It manifests itself prominently in patients with HIV. A small study in Japan suggested that HIV-infected lung cancer patients with higher PD-1 expression have worse outcomes.25

There has been some question about whether checkpoint inhibitors and other new immunotherapies are safe in HIV patients. Very preliminary case-study data — including a retrospective review of veterans with HIV and cancer26 — suggest checkpoint inhibitors do appear safe. But these early results need confirmation. An ongoing National Cancer Institute-sponsored trial is evaluating checkpoint inhibitors in HIV-infected patients.

There’s another large question beyond safety of checkpoint inhibitors. We think immune dysregulation plays at least a small role in driving the excess lung cancer incidence in people with HIV. Is it possible that checkpoint inhibitors may be even more effective than current therapies or that this might be a group of tumors that particularly exhibits local immune exhaustion? These are important and interesting areas to explore as we continue to learn more about the role of the immune system in managing lung cancer.

Finally, I should reiterate something we discussed before: Another big area that needs clarification is whether and how we should modify lung cancer screening recommendations to be more inclusive when it comes to people with HIV infection.

Author: Mark Mascolini