The current study describes the clinicopathologic characteristics of 36 patients with lung carcinoma and human immunodeficiency virus (HIV) infection observed within the Italian Cooperative Group on AIDS and Tumors (GICAT).
The current study describes the clinicopathologic characteristics of 36 patients with lung carcinoma and human immunodeficiency virus (HIV) infection observed within the Italian Cooperative Group on AIDS and Tumors (GICAT).
Patients with lung carcinoma and HIV infection collected by the GICAT between 1986–1998 were evaluated retrospectively. As a control group, the authors analyzed 102 patients age < 60 years with lung carcinoma but without HIV infection who were seen at the CRO, National Cancer Institute, Aviano, Italy between 1995–1996.
Patients with lung carcinoma and HIV infection were younger (38 years vs. 53 years) and previously smoked more cigarettes per day (40 vs. 20) than the control group. The main histologic subtype was adenocarcinoma. TNM Stage III–IV disease was observed in 53% of the patients. The median CD4 cell count was 150/mm3. The median overall survival was significantly shorter in the patients with HIV compared with the control group (5 months vs. 10 months; P = 0.0001).
The results of the current study demonstrate that lung carcinoma in the HIV setting affects mainly young individuals with a history of heavy tobacco smoking and a moderately advanced immunodeficiency status. Lung carcinoma is associated with a more adverse outcome in HIV patients and represents the cause of death in the majority of these patients. Cancer 2000;88:563–9. © 2000 American Cancer Society.
There is clear evidence that both the incidence and the mortality of acquired immunodeficiency syndrome (AIDS) have declined significantly in developed countries due to the very active antihuman immunodeficiency virus (anti-HIV) combination therapy introduced in clinics in the last few years. Accordingly, HIV-related malignancies, i.e., Kaposi sarcoma (KS) and non-Hodgkin lymphoma (NHL), also are declining.1–6 Other non-AIDS defining tumors, such Hodgkin disease, squamous cell carcinoma of the skin and lip, testicular carcinoma, and lung carcinoma, have been reported recently in several cohort studies with increased frequency in patients with HIV infection.7–9 In this study, the clinicopathologic features and the outcomes of 36 patients with HIV infection and lung carcinoma are reported.
Since 1986, the Italian Cooperative Group on AIDS and Tumors (GICAT) has worked with the objective of studying the malignant tumors in patients with HIV infection. In May 1998, we decided to evaluate retrospectively patients with lung carcinoma and HIV infection seen between the beginning of the 1986 GICAT study and April 1998. For a control group, we evaluated patients with lung carcinoma but without HIV infection who were seen at the CRO, National Cancer Institute, Aviano, Italy during 1995 and 1996. Because the patients with HIV infection were young, we decided to select a control group with patients age < 60 years to make the two patient groups more comparable. We tried to have a control group with the same median age of the HIV group. Unfortunately, too few of these patients had been observed during the last 10 years at our institution. Therefore, we decided to use those patients who were age < 60 years (all HIV patients included were age < 60 years).
Data on HIV positive patients were collected through questionnaires sent out to the participating centers of the GICAT, and the following parameters were reviewed with special attention: demographic features, HIV risk factors, cigarette smoking habits, past exposure to environmental carcinogens, HIV immunologic and clinical features at the time of diagnosis of lung carcinoma, lung carcinoma histologic type and clinical stage, type of treatment received, response, survival, and causes of death. Data from the control group were collected by a careful review of hospital charts of the CRO, National Cancer Institute, Aviano, Northeast Italy.
All patients were evaluated for extent of lung carcinoma by physical examination, complete hematologic and blood biochemistry evaluation, chest radiography, computed tomography of the thorax, and abdominal ultrasound. Bone marrow aspiration, computed tomography of the brain, and a bone scan were performed for patients with small cell lung carcinoma (SCLC) or when clinically indicated. Bronchoscopy and/or computed tomography-guided per cutaneous fine-needle aspiration were performed for histologic diagnosis according to the localization of the pulmonary lesions.
The disease was staged according to TNM for nonsmall cell lung carcinoma (NSCLC).10 The patients with SCLC were classified according to the two-stage systems of the Veterans Administration Lung Group.11 Limited stage disease was defined as tumor confined to one hemithorax and the regional lymph nodes that could be encompassed within a tolerable radiotherapy port. Extended stage disease was defined as disease beyond these bounds.
The criteria used to diagnose AIDS-related complex (ARC) and AIDS were those defined by the Centers for Disease Control (CDC) classification.12 HIV serology was performed by enzyme-linked immunoabsorbent and was reconfirmed by Western blot analysis (immunoblot). The subsets and the absolute counts of CD4 and CD8 cells in the peripheral blood were determined only in HIV positive patients. In the 36 patients from the study group, HIV infection already had been diagnosed in all cases at the time of diagnosis of lung carcinoma, whereas, in the control group, lung carcinoma was diagnosed in patients who had never been tested for HIV infection and who were not actually tested after the diagnosis of lung carcinoma. This may be explained by the retrospective nature of this study and by the fact that, in Italy, the ongoing policy does not allow HIV testing without the presence of any risk factors and/or the presence of suspicious HIV medical complications (i.e., KS, Burkitt lymphoma, Pneumocycistis carinii pneumonia [PCP], etc.).
The 36 patients in the study group received various treatment modalities, such as surgery alone, radiotherapy (RT) alone, chemotherapy (CT) alone, or a combined treatment. CT regimens were administered according to the policy and the ongoing protocols of each oncologic center. In particular, the following regimens were employed: Single-agent CT included cisplatin 100 mg/m2 intravenously (IV) on Day 1 every 4 weeks, vinorelbine 30 mg/m2 IV on Day 1 weekly, or etoposide 120 mg/m2 IV on Days 1–5 every 3 weeks. CT with two drugs included carboplatin 300 mg/m2 IV on Day 1 plus etoposide 100 mg/m2 IV on Days 1–3 every 4 weeks. CT with more than two drugs included doxorubicin, cyclophosphamide, and etoposide (ACE)13; cyclophosphamide, doxorubicin, and vincristint (CAV)14; and cyclophosphamide, doxorubicin, and cisplatin (CAP).15 The treatment of patients with HIV infection included zidovudine alone; zidovudine and didanosine; zidovudine and lamivudine; zidovudine, lamivudine, and indinavir; and stavudine, lamivudine, and saquinavir. PCP prophylaxis was administered with oral trimetoprim 160 mg and sulfametoxazole 800 mg daily or with aerosolized pentamidine-isothionate 300 mg once per month. In the control group, patients were treated with the standard CT regimens used for the management of lung carcinoma in the general population (for example, carboplatin and etoposide, CAV regimen, ACE regimen, or CAP regimen).
RT was administered between 3000 rads and 6000 rads over 4–6 weeks with radical or palliative intent. In particular, it was given at a dose ≤ 4000 rads with palliative intent or at a dose ≥ 4000 rads with curative intent. Patients with HIV infection who received palliative RT were treated with a median dose of 2500 rads (range, 800–3900 rads) with a median dose per fraction of 200 rads (range, 200–800 rads), whereas those who were treated with curative RT received a median dose of 4800 rads (range, 4500–6000 rads), with a median dose per fraction of 250 rads (range, 200–300 rads). RT in the control group was given with palliative intent in doses ranging between 1500 rads and 4000 rads in 10–15 fractions or with curative intent at a median dose of 5000 rads (range, 4500–6000 rads) in 20–30 fractions. Surgery, when indicated, was performed according to the extent of the disease in patients from both groups.
A complete response (CR) was defined as the complete disappearance of all evaluable disease for at least 4 weeks. A partial response (PR) was defined as a reduction ≥ 50% in the sum of the products of the cross-section greatest dimensions of all known lesions. No response (NR) was defined as less than PR or as disease progression.
Survival was calculated from the date of the diagnosis of lung carcinoma to that of the last follow-up visit estimated according to the product-limit method of Kaplan and Meyer.16 The distribution comparisons were made with the method of Mantel.17 Significant tests for proportions were computed with the chi-square test,18 and continuous variables were analyzed with the Mann–Whitney test.
Between November 1986 and April 1998, in 20 centers of the GICAT, 36 patients were diagnosed with lung carcinoma and HIV infection. The median age was 38 years (range, 28–58 years), and 32 of 36 patients (89%) were males. Risk factors for HIV infection were intravenous drug use in 25 of 36 patients (69%), homosexual transmission in 6 of 36 patients (17%), and heterosexual transmission in 5 of 36 patients (14%). At the time of diagnosis of lung carcinoma, HIV infection was asymptomatic in 16 of 36 patients (44%), whereas 10 of 36 patients (28%) had ARC, and 10 of 36 patients (28%) had CDC-defined AIDS. In particular, 3 of the latter patients had an esophageal candidiasis, 3 patients had disseminated tuberculosis (TB), 1 patient had PCP, 1 patient had a pulmonary candidiasis, and 1 patient had neurotoxoplasmosis. Overall, 6 of 10 patients with a diagnosis of AIDS before lung carcinoma had had a pulmonary infection. Moreover, 7 of 13 patients with a histologic diagnosis of adenocarcinoma had concomitant clinical and/or radiographic evidence of TB. The median CD4 cell count was 150/mmc (range, 3–835 mmc). Overall, 32 of 34 patients (94%) were heavy tobacco smokers, with a median number of 40 cigarettes smoked per day (range, 5–60 cigarettes per day).
NSCLC was observed in 31 of 36 patients (86%), and SCLC was observed in 5 of 36 patients(14%). Among the patients with NSCLC, adenocarcinoma was the most frequently observed histologic subtype, with 13 of 31 patients (42%), followed by squamous cell carcinoma in 12 of 31 (39%) patients, and by anaplastic large cell carcinoma in 5 of 31 patients (16%). One patient (3%) had a diagnosis of bronchioalveolar carcinoma. With regard to disease stage in patients with NSCLC, Stage I was diagnosed in 3 of 31 patients (10%), Stage II was diagnosed in 2 of 31 patients (6%), Stage IIIA was diagnosed in 4 of 31 patients (13%), Stage IIIB was diagnosed in 5 of 31 patients (16%), and Stage IV was diagnosed in 17 of 31 patients (55%). Patients with SCLC had limited disease in 3 patients and extended disease in the 2 other patients. Table 1 summarized the demographic data from our patients.
|Characteristic||HIV positive (%)||Control group (%)|
|Male||32 of 36 (89)||90 of 102 (88)|
|Female||4 of 36 (11)||12 of 102 (12)|
|0–1||15 of 35 (43)||52 of 102 (51)|
|≥2||20 of 35 (57)||50 of 102 (49)|
|NSCLC||31 of 36 (86)||83 of 102 (81)|
|Adenocarcinoma||13 of 31 (42)||40 of 183 (48)|
|Squamous cell carcinoma||12 of 31 (39)||37 of 83 (44)|
|Anaplastic large cell carcinoma||5 of 31 (16)||5 of 83 (6)|
|Bronchioalveolar carcinoma||1 of 31 (3)||1 of 83 (2)|
|SCLC||5 of 36 (14)||19 of 102 (19)|
|I||3 of 31 (10)||10 of 83 (12)|
|II||2 of 31 (6)||9 of 83 (11)|
|IIIA||4 of 31 (13)||9 of 83 (11)|
|IIIB||5 of 31 (16)||15 of 83 (18)|
|IV||17 of 31 (55)||40 of 83 (48)|
|Limited||3 of 5||10 of 19 (53)|
|Extended||2 of 5||9 of 19 (47)|
Twenty of 36 patients (55%) were evaluable for anti-HIV therapy. In particular, 13 patients received antiretroviral therapy, including zidovudine alone in 7 patients; zidovudine and didanosine in 1 patient; zidovudine and lamivudine in 2 patients; zidovudine, lamivudine, and indinavir in 2 patients; and stavudine, lamivudine, and saquinavir in 1 patient. The majority of our patients already were receiving anti-HIV therapy at the time of diagnosis of lung carcinoma, but only 2 patients received concomitant anti-HIV therapy and antineoplastic CT. Seven patients did not receive any anti-HIV therapy: three patients because the diagnosis of lung carcinoma was made before any therapy for HIV infection was available, 2 patients because of the concomitant IV use of illicit drugs, and two patients because they refused anti-HIV therapy. Forty-four percent of patients had a CD4 cell count < 200/mmc and were eligible to receive the PCP prophylaxis. Seven patients were not evaluable for PCP prophylaxis. Two patients did not receive any prophylaxis because they died before the PCP prophylaxis was recommended for patients with CD4 cell count < 200/mmc. Among the 7 patients who received PCP prophylaxis, 5 patients received trimetoprim and sulfametoxazole, and 2 patients received aerosol pentamidine.
Twenty-five of 36 patients (69%) were treated for lung carcinoma. Eleven patients were not treated: two patients because the diagnosis was made postmortem; five patients because of poor performance status; two patients because of concomitant opportunistic infections (OI); and two patients because they refused treatment. Of the 25 treated patients, 1 patient underwent surgery, 1 patient underwent surgery plus RT, and 1 patient underwent surgery plus CT. Eight patients received RT alone, 12 patients received CT alone, and 2 patients received both treatments. All 3 of the patients who underwent surgery underwent lobectomy. However, 9 of 31 patients (29%) with NSCLC gad Stage I–IIIA disease, but only 3 of them underwent surgery. The other 6 patients did not undergo surgery because of a poor performance status (3 patients) or previous pulmonary opportunistic infections (3 patients).
Fifteen of 25 treated patients (60%) received CT (alone or as combined modality treatment). In particular, the following CT regimens were used: 1 of 15 patients (7%) received etoposide alone, 1 of 15 patients (7%) received cisplatin alone, 1 of 15 patients (7%) received vinorelbine alone, 7 of 15 patients (46%) received carboplatin and etoposide, 3 of 15 patients (20%) received the ACE regimen,13 1 patient (7%) received the CAV regimen,14 and 1 patient (7%) received the CAP regimen.15
Eleven of 25 patients (44%) received RT, 5 patients with palliative intent and 6 with curative intent. Twenty-three of 25 treated patients were evaluable for response. Overall, 11 of 23 patients (48%) had an objective response. In particular, 4 of 23 patients (18%) obtained a CR, and 7 of 23 patients (30%) obtained a PR. Twelve of 23 patients (52%) experienced disease progression.
Nineteen of 25 patients (76%) were assessable for toxicity. We did not observe any side effects in the 3 patients who were treated with surgery. Among the 6 patients who were treated with RT, we observed 4 patients with of radiation-induced esophagitis (Grade 3 in 3 patients and Grade 2 in 1 patient). In these patients, we observed 2 case of TB 2 months after the end of RT. The remaining 10 patients received CT. Severe hematologic toxicity (World Health Organization Grade 3 and 4) was observed in 5 patients with a toxic death due to a cerebral hemorrhage during thrombocytopenia after carboplatin and etoposide. One patient developed Grade 2 mucositis. Eight of 25 treated patients (32%) developed OI during treatment or follow-up. In particular, 5 cases were observed among patients who received CT (2 PCP, 1 toxoplasmosis, 1 TB, and 1 cytomegalovirus [CMV] retinitis); 2 cases were diagnosed in patients who received RT (1 TB and 1 CMV retinitis), and 1 case (TB) was observed in a patient who received CT and RT.
The median survival was 5 months, and the actuarial survival rate at 1 year was 10% (95 confidence interval [CI], 0–20%). We compared the survival of the patients with CD4 cell count >100/mmc or <100/mmc and >200 mmc or <200/mmc. Patients with a CD4 cell count ≥ 100/mmc had a median survival of 5.2 months compared with a median survival of 4.4 months for the patients with CD4 cell count < 100/mmc (P = 0.75). Similarly, the patients with CD4 cell count ≥ 200/mmc had a median of survival of 5.1 months compared with a median survival of 5 months for the patients with CD4 cell count < 200/mmc (P = 0.98).
Thirty-one of 36 patients (86%) died: 23 of 31 patients (74%) from lung carcinoma, 4 of 31 patients (13%) from OI (2 patients with PCP, 1 patient with neurotoxoplasmosis, and 1 patient with TB), 1 of 31 patients (3%) from wasting syndrome, and 3 of 31 patients (10%) died from other causes (overdose in 1 patient, stroke in 1 patient, and toxic death in 1 patient). Among the 31 patients who died, 7 patients underwent autopsy. For the remaining 24 patients, the cause of death was determined based on clinical evidence.
We evaluated 102 patients with lung carcinoma and without HIV infection age < 60 years who wee seen at CRO, Aviano, during 1995 and 1996. The median age was 53 years (range, 37–59 years), and 88% of patients were males. The median number of cigarettes smoked per day was 20 (range, 10–70 cigarettes per day). NSCLC was diagnosed in 83 of 102 patients (81%: 48% adenocarcinoma, 44% squamous cell carcinoma, 6% large cell carcinoma, and 2% bronchioalveolar carcinoma), and SCLC was diagnosed in 19 of 102 patients (19%). Forty-eight percent of patients with NSCLC had metastatic disease, and 47% of patients with SCLC had extended disease. Eighty-three of 102 patients (81%) were treated for lung carcinoma with a standard approach according to the extent of their disease. In particular, 10 patients underwent surgery, 9 patients underwent surgery plus RT, and 3 patients underwent surgery plus CT. Twenty-eight patients received CT alone, 14 patients received RT alone, and 19 patients received both treatment. With regard to the surgery employed, 16 patients underwent lobectomy, and 7 patients underwent pneumectomy. The following CT regimens were used: 2 of 50 patients (4%) received doxorubicin alone, 3 of 50 patients (6%) received vinorelbine alone, 3 of 50 patients (6%) received cisplatin alone, 4 of 50 patients (18%) received cisplatin and paclitaxel, 5 of 50 patients (10%) received carboplatin and vinorelbine, 28 of 50 patients (56%) received carboplatin and etoposide, and 5 of 50 patients (10%) received the CAV regimen. All patients received RT with curative intent. The outcome was better compared with that of patients with HIV infection, and the median survival was 10 months (P = 0.0001).
The current study represents the largest series of patients with lung carcinoma and HIV infection published in the literature. Sixty-nine per cent of HIV patients were intravenous drug users in accordance to the epidemiology of HIV infection in Italy. It is noteworthy that ≈50% of the HIV seropositive patients were asymptomatic for HIV infection at the time of diagnosis of lung carcinoma despite a CD4 cell count that was quite low, i.e., 150/mmc. However, lung carcinoma may appear earlier in the course of HIV infection than other, more commonly encountered neoplasms, i.e., KS and NHL. The changing pattern of the HIV epidemic could increase the incidence of lung carcinoma in patients with HIV infection in the near future.
The median age of our study group was 38 years, similar to the median age of patients with AIDS in Italy (32 years among the 40,950 CDC-defined AIDS patients registered in Italy as of December 1997) but significantly younger than that observed in the control group (53 years), although we selected only patients age < 60 years in the latter group.
Our study shows that the patients with HIV infection used to smoke more cigarettes per day than the patients in the control group (40 vs. 20 cigarettes per day), although the duration of tobacco exposure evidently was shorter. Whether the greater number of cigarettes smoked daily or the longer tobacco exposure is more relevant to lung carcinogenesis it is a matter of discussion.
Adenocarcinoma was the main subtype (42%) in our study group, and this rate was similar to that observed in the control group (48%). Adenocarcinoma has been shown to be the main subtype in young patients from the general population, and its increased frequency in these patients suggests that its development may require less overall carcinogen exposure and/or fewer genetic derangements compared with other histologic subtypes.19 In addition, we can speculate that the high incidence of adenocarcinoma in HIV-infected individuals from our study group also may be related to the fact that lungs are a common site for infections in these patients. It must be remembered that our study group was composed mainly of individuals who had self-injected heroin intravenously. Moreover, there may be an increase in cancer arising in preexisting scar tissue as a result of the increased incidence of pulmonary infections and, in particular, TB. The rate of TB occurrence has been associated with an adenocarcinoma subtype in several studies.20–22 Furthermore, it is well know that normal pulmonary immunologic surveillance is impaired in the HIV setting. In particular, natural killer cell activity may allow an unchecked proliferation of spontaneously developing tumor cells.23 Moreover, the HIV virus itself may stimulate the release of aberrant growth factors, resulting in oncogenesis.24 This can be emphasized in intravenous drug users due to the fact that recurrent bacterial pulmonary infections in these patients and the drug-induced changes in the pulmonary stroma could increase the production of growth factors. In our series, 6 of 10 patients with a diagnosis of AIDS before lung carcinoma had had a pulmonary infection. Moreover, 7 of 13 patients with adenocarcinoma had concomitant clinical and/or radiographic evidence of TB.
Advanced disease stage (i.e., Stage III–IV) was diagnosed in the vast majority of patients in our study group (84%), with ≈50% of the patients presenting with distant metastases at the time of diagnosis. This rate was similar to that observed in the control group. However, in the older general population with lung carcinoma, there is a smaller group of patients that presented with advanced disease.19 In the HIV setting, it must be considered that some physicians may have misdiagnosed lung carcinoma as an HIV-related infection, with a consequent delay in the diagnosis of lung carcinoma.
The median overall survival was significantly shorter for our study group population compared with that of the control group (5 months vs. 10 months; P = 0.0001; Fig. 1). Moreover, lung carcinoma was the cause of death in the vast majority of our study group population (74%), whereas it was the only cause of death in the patients from the control group. The shorter survival of the study group patients may be explained by both a synergistic effect and/or an additive effect of HIV in the more aggressive course of lung carcinoma and by the fact that such patients were not considered eligible for curative surgical resection. In fact, only 3 of 9 patients with Stage I–IIIA NSCLC underwent surgery, whereas the other 6 patients were not eligible because of poor performance status or previous lung OI. It is not yet known, although it may be logical, whether surgery could be of some benefit even for patients with Stage I–IIIA disease, a high CD4 cell count, good performance status, and adequate pulmonary reserve and without a previous diagnosis of AIDS. Furthermore, in our group, only 3 patients received highly active antiretroviral therapy (HAART) before or during chemotherapy, and it is not possible to evaluate the tolerance to HAART and the toxicity of the combined treatment. However, taking into consideration that the majority of our patients had metastatic disease at the time of diagnosis and that 74% of them died of lung carcinoma, we believe that the impact of HAART on the natural history of lung carcinoma should be negligible. However, further epidemiologic and clinical prospective studies are needed to answer these questions.
In conclusion, our results demonstrate that lung carcinoma in the HIV setting affects mainly young individuals with a heavy tobacco smoking history and prevalently those with a moderately advanced immunodeficiency status. However, lung carcinoma is associated with a more adverse outcome in the HIV setting and represents the cause of death in the majority of these patients.
The authors thank Dr. Maddalena Mosconi for expert assistance in the preparation of the article. The authors also thank all of the contributors to the study: F. Betti, Division of Infectious Diseases, General Hospital, Macerata; M. Clerici, Division of Medical Oncology, San Carlo Borromeo Hospital, Milan; G. M. Fiorentini, Division of Medical Oncology, General Hospital, Ravenna; E. Galligioni, Division of Medical Oncology, S. Chiara Hospital, Trento; F. Giordano, Division of Pneumology, Forlanini Hospital, Rome; G. Landonio, Division of Medical Oncology, Niguarda Ca' Granda Hospital, Milan; A. Lazzarin, Division of Infectious Diseases, San Raffaele Hospital, Milan; M. Libanore, Division of Infectious Diseases, S. Anna Hospital, Ferrara; S. Marra, Division of Infectious Diseases, Casa del Sole Hospital, Palermo; L. Nigro, Division of Infectious Diseases, Garibaldi Hospital, Catania; A. Parina, Division of Pathology, Riuniti Hospitals, Bergamo; E. Raise, Division of Infectious Diseases, General Hospital, Venice; G. Riazzardini, I Division of Infectious Diseases, Sacco Hospital, Milan; and S. Sabbatani, Division of Infectious Diseases, Maggiore Hospital, Bologna.