Preliminary report of the lung cancer detection program in New York

Automated Pulmonary Nodule Classification in Computed Tomography Images Using a Deep Convolutional Neural Network Trained by Generative Adversarial Networks

Lung cancer is a leading cause of death worldwide. Although computed tomography (CT) examinations are frequently used for lung cancer diagnosis, it can be difficult to distinguish between benign and malignant pulmonary nodules on the basis of CT images alone. Therefore, a bronchoscopic biopsy may be conducted if malignancy is suspected following CT examinations. However, biopsies are highly invasive, and patients with benign nodules may undergo many unnecessary biopsies. To prevent this, an imaging diagnosis with high classification accuracy is essential. In this study, we investigate the automated classification of pulmonary nodules in CT images using a deep convolutional neural network (DCNN). We use generative adversarial networks (GANs) to generate additional images when only small amounts of data are available, which is a common problem in medical research, and evaluate whether the classification accuracy is improved by generating a large amount of new pulmonary nodule images using the GAN. Using the proposed method, CT images of 60 cases with confirmed pathological diagnosis by biopsy are analyzed. The benign nodules assessed in this study are difficult for radiologists to differentiate because they cannot be rejected as being malignant. A volume of interest centered on the pulmonary nodule is extracted from the CT images, and further images are created using axial sections and augmented data. The DCNN is trained using nodule images generated by the GAN and then fine-tuned using the actual nodule images to allow the DCNN to distinguish between benign and malignant nodules. This pretraining and fine-tuning process makes it possible to distinguish 66.7% of benign nodules and 93.9% of malignant nodules. These results indicate that the proposed method improves the classification accuracy by approximately 20% in comparison with training using only the original images.

Screening for lung cancer

BACKGROUND

This is an updated version of the original review published in The Cochrane Library in 1999 and updated in 2004 and 2010. Population-based screening for lung cancer has not been adopted in the majority of countries. However it is not clear whether sputum examinations, chest radiography or newer methods such as computed tomography (CT) are effective in reducing mortality from lung cancer.

OBJECTIVES

To determine whether screening for lung cancer, using regular sputum examinations, chest radiography or CT scanning of the chest, reduces lung cancer mortality.

SEARCH METHODS

We searched electronic databases: the Cochrane Central Register of Controlled Trials (CENTRAL) (The Cochrane Library 2012, Issue 5), MEDLINE (1966 to 2012), PREMEDLINE and EMBASE (to 2012) and bibliographies. We handsearched the journal Lung Cancer (to 2000) and contacted experts in the field to identify published and unpublished trials.

SELECTION CRITERIA

Controlled trials of screening for lung cancer using sputum examinations, chest radiography or chest CT.

DATA COLLECTION AND ANALYSIS

We performed an intention-to-screen analysis. Where there was significant statistical heterogeneity, we reported risk ratios (RRs) using the random-effects model. For other outcomes we used the fixed-effect model.

MAIN RESULTS

We included nine trials in the review (eight randomised controlled studies and one controlled trial) with a total of 453,965 subjects. In one large study that included both smokers and non-smokers comparing annual chest x-ray screening with usual care there was no reduction in lung cancer mortality (RR 0.99, 95% CI 0.91 to 1.07). In a meta-analysis of studies comparing different frequencies of chest x-ray screening, frequent screening with chest x-rays was associated with an 11% relative increase in mortality from lung cancer compared with less frequent screening (RR 1.11, 95% CI 1.00 to 1.23); however several of the trials included in this meta-analysis had potential methodological weaknesses. We observed a non-statistically significant trend to reduced mortality from lung cancer when screening with chest x-ray and sputum cytology was compared with chest x-ray alone (RR 0.88, 95% CI 0.74 to 1.03). There was one large methodologically rigorous trial in high-risk smokers and ex-smokers (those aged 55 to 74 years with ≥ 30 pack-years of smoking and who quit ≤ 15 years prior to entry if ex-smokers) comparing annual low-dose CT screening with annual chest x-ray screening; in this study the relative risk of death from lung cancer was significantly reduced in the low-dose CT group (RR 0.80, 95% CI 0.70 to 0.92).

AUTHORS' CONCLUSIONS

The current evidence does not support screening for lung cancer with chest radiography or sputum cytology. Annual low-dose CT screening is associated with a reduction in lung cancer mortality in high-risk smokers but further data are required on the cost effectiveness of screening and the relative harms and benefits of screening across a range of different risk groups and settings.

Nael Martini: a leader in thoracic surgical oncology

Nael Martini was one of the leading academic general thoracic surgeons of the late 20th century. His most notable contributions related to the surgical and multimodality treatment of lung cancer.

Randomized controlled trials of the efficacy of lung cancer screening by sputum cytology revisited: a combined mortality analysis from the Johns Hopkins Lung Project and the Memorial Sloan-Kettering Lung Study

BACKGROUND

: Two randomized controlled trials of lung cancer screening initiated in the 1970s, the Johns Hopkins Lung Project and the Memorial Sloan-Kettering Lung Study, compared 1 arm that received annual chest X-ray and 4-monthly sputum cytology (dual-screen) to a second arm that received annual chest X-ray only. Previous publications from these trials reported similar lung cancer mortality between the 2 groups. However, these findings were based on incomplete follow-up, and each trial on its own was underpowered to detect a modest mortality benefit.

METHODS

: The authors estimated the efficacy of lung cancer screening with sputum cytology in an intention-to-screen analysis of lung cancer mortality, using combined data from these trials (n = 20,426).

RESULTS

: Over (1/2) of squamous cell lung cancers diagnosed in the dual-screen group were identified by cytology; these cancers tended to be more localized than squamous cancers diagnosed in the X-ray only arm. After 9 years of follow-up, lung cancer mortality was slightly lower in the dual-screen than in the X-ray only arm (rate ratio [RR], 0.88; 95% confidence interval [CI], 0.74-1.05). Reductions were seen for squamous cell cancer deaths (RR, 0.79; 95% CI, 0.54-1.14) and in the heaviest smokers (RR, 0.81; 95% CI, 0.67-1.00). There were also fewer deaths from large cell carcinoma in the dual-screen group, although the reason for this is unclear.

CONCLUSIONS

: These data are suggestive of a modest benefit of sputum cytology screening, although we cannot rule out chance as an explanation for these findings. Cancer 2009. (c) 2009 American Cancer Society.

Screening for lung cancer

BACKGROUND

While population based screening for lung cancer has not been adopted by most countries, it is not clear whether sputum examinations, chest radiography or newer methods such as computed tomography are effective in reducing mortality from lung cancer.

OBJECTIVES

To determine whether screening for lung cancer using regular sputum examinations or chest radiography or CT chest reduces lung cancer mortality.

SEARCH STRATEGY

Electronic databases (the Cochrane Central Register of Controlled Trials, MEDLINE, PREMEDLINE and EMBASE; 1966 to July 2000) ), bibliographies, hand searching of a journal and discussion with experts were used to identify published and unpublished trials.

SELECTION CRITERIA

Controlled trials of screening for lung cancer using sputum examinations, chest radiography or CT chest.

DATA COLLECTION AND ANALYSIS

Intention to screen analysis was performed. Where there was significant statistical heterogeneity relative risks were reported using the random effects model, but for other outcomes the fixed effect model was used.

MAIN RESULTS

Seven trials were included (6 randomised controlled studies and 1 non-randomised controlled trial) with a total of 245,610 subjects. There were no studies with an unscreened control group. Frequent screening with chest x-rays was associated with an 11% relative increase in mortality from lung cancer compared with less frequent screening (RR 1.11, CI: 1.00-1.23). A non statistically significant trend was observed to reduced mortality from lung cancer when screening with chest x-ray and sputum cytology was compared with chest x-ray alone (RR 0.88, CI:0.74-1.03). Several of the included studies had potential methodological weaknesses. There were no controlled studies of spiral CT.

REVIEWER'S CONCLUSIONS

The current evidence does not support screening for lung cancer with chest radiography or sputum cytology. Frequent chest x-ray screening might be harmful. Further, methodologically rigorous trials are required.

Screening for lung cancer

BACKGROUND

The effectiveness of screening for lung cancer with chest radiography, sputum cytology or spiral CT has not been established.

OBJECTIVES

To determine whether screening for lung cancer using regular sputum examinations or chest radiography or CT chest reduces lung cancer mortality.

SEARCH STRATEGY

Electronic databases, bibliographies, hand searching of a journal and discussion with experts were used to identify published and unpublished trials.

SELECTION CRITERIA

Controlled trials of screening for lung cancer using sputum examinations, chest radiography or CT chest.

DATA COLLECTION AND ANALYSIS

Intention to screen analysis was performed. Where there was significant statistical heterogeneity relative risks were reported using the random effect model, but for other outcomes the fixed effect model was used.

MAIN RESULTS

Seven trials were included (6 randomised controlled studies and 1 non-randomised controlled trial) with a total of 245,610 subjects. There were no studies with an unscreened control group. Frequent screening with chest x-rays was associated with an 11% relative increase in mortality from lung cancer compared with less frequent screening (RR 1.11, CI: 1.00-1.23). A non statistically significant trend was observed to reduced mortality from lung cancer when screening with chest x-ray and sputum cytology was compared with chest x-ray alone (RR 0.88, CI:0.74-1.03). Several of the included studies had potential methodological weaknesses. There were no controlled studies of spiral CT.

REVIEWER'S CONCLUSIONS

The current evidence does not support screening for lung cancer with chest radiography or sputum cytology. Frequent chest x-ray screening might be harmful. Further, methodologically rigorous trials are required.

Mass screening for lung cancer with mobile spiral computed tomography scanner

BACKGROUND

The incidence of and mortality from lung cancer have increased steadily. Most lung cancers are not localised when first detected, but early detection is mandatory to improve prognosis. Since curable early cases are hard to visualise with conventional chest radiography, a new diagnostic means must be found. We assessed whether population-based mass screening with a spiral computed tomography scanner could contribute substantially to detection of smaller cancers, and decrease mortality.

METHODS

In 1996, we screened in a mobile unit 5483 individuals from the general population of Matsumoto, Japan, aged between 40 years and 74 years who had undergone annual chest radiography (miniature fluorophotography) and cytological assessment of sputum. All participants had a low-dose X-ray spiral computed tomography (CT) scan of the thorax; 3967 also underwent miniature fluorophotography. We compared smokers and non-smokers. Further assessments were done for probably benign but suspicious lesions; suspicion of cancer; and indeterminate small nodules by chest radiography and conventional CT, with additional transbronchial biopsy when possible. Thoracotomy was recommended when it was strongly suspected that the patients had lung cancer.

FINDINGS

19 patients were diagnosed as having lung cancer--14 with suspicion of lung cancer, three with benign but suspicious lesions, and two with indeterminate small nodules. 18 cases were surgically confirmed, and one was clinically diagnosed. The mean size of lesions was 17 mm (range 6-47). In four of 19 patients, lung abnormality was seen on CT and miniature fluorophotography. The lung-cancer detection rate with CT was 0.48%, significantly higher than the 0.03-0.05% for standard mass assessments done previously in the same area. CT missed one case that was found solely on a sputum cytology examination.

INTERPRETATION

Our results show that miniature fluorophotography or conventional chest radiography, which have been the main diagnostic techniques for lung cancer, showed few small cancers. CT was more accurate in mass screening for lung cancer and led to early detection and an accurate diagnosis of lung cancer, and should be considered in future health plans.

Non-small cell lung cancer. Part I: Biology, diagnosis, and staging

Squamous, large cell, and adenocarcinoma, collectively termed non-small cell lung cancer (NSCLC), are diagnosed in approximately 75% of patients with lung cancer in the United States. The treatment of these three tumor cell types is approached in virtually identical fashion because, in contrast to small cell carcinoma of the lung, NSCLC more frequently presents with localized disease at the time of diagnosis and is thus more often amenable to surgical resection but less frequently responds to chemotherapy and irradiation. Cigarette smoking is etiologically related to the development of NSCLC in the great majority of cases. Genetic mutations in dominant oncogenes such as K-ras, loss of genetic material on chromosomes 3p, 11p, and 17p, and deletions or mutations in tumor suppressor genes such as rb and p53 have been documented in NSCLC tumors and tumor cell lines. NSCLC is diagnosed because of symptoms related to the primary tumor or regional or distant metastases, as an incidental finding on chest radiograph, or rarely because of a paraneoplastic syndrome such as hypercalcemia or hypertrophic pulmonary osteoarthropathy. Screening smokers with periodic chest radiographs and sputum cytologic examination has not been shown to reduce mortality. The diagnosis of NSCLC is usually established by fiberoptic bronchoscopy or percutaneous fine-needle aspiration, by biopsy of a regional or distant metastatic site, or at the time of thoracotomy. Pathologically, NSCLC arises in a setting of bronchial mucosal metaplasia and dysplasia that progressively increase over time. Squamous carcinoma more often presents as a central endobronchial lesion, while large cell and adenocarcinoma have a tendency to arise in the lung periphery and invade the pleura. Once the diagnosis is made, the extent of tumor dissemination is determined. Since most NSCLC patients who survive 5 years or longer have undergone surgical resection of their cancers, the focus of the staging process is to determine whether the patient is a candidate for thoracotomy with curative intent. The dominant prognostic factors in NSCLC are extent of tumor dissemination, ambulatory or performance status, and degree of weight loss. Stages I and II NSCLC, which are confined within the pleural reflection, are managed by surgical resection whenever possible, with approximate 5-year survival of 45% and 25%, respectively. Patients with stage IIIa cancers, in which the primary tumor has extended through the pleura or metastasized to ipsilateral or subcarinal lymph nodes, can occasionally be surgically resected but are often managed with definitive thoracic irradiation and have 5-year survival of approximately 15%.(ABSTRACT TRUNCATED AT 400 WORDS)

The effect of tumor size and tumor cell DNA content on the survival of patients with stage I adenocarcinoma of the lung

Tumor size at resection and flow cytometric (FCM) DNA distribution of 93 pathologically confirmed AJC Stage I adenocarcinomas of lung were compared with survival. All lung cancers had been treated by lobectomy and mediastinal lymph node dissection at this Institution, and the pathology of all cases was reviewed. Median tumor size was 2.5 cm. FCM DNA assays were carried out on the formalin-fixed, paraffin-embedded tumor tissue. Seventy-nine (85%) of the tumors were aneuploid and 14 were diploid. There was no significant difference in survival between patients with diploid and aneuploid tumors. The DNA index, proliferation fraction, and fraction of aneuploid cells also had no effect on survival. However, patients with tumors smaller than 3 cm in diameter had a significantly better prognosis than those with larger tumors. Thus, in Stage I adenocarcinoma of lung, tumor size but not cellular DNA content distinguishes subgroups with favorable versus unfavorable prognosis.

Impact of early detection on the clinical course of lung cancer

The Memorial Sloan-Kettering Lung Cancer Detection Program is described, and data from this study are reported and analyzed. The program enrolled 10,040 cigarette-smoking men who were screened for at least 5 years by annual chest radiographs; half the men randomly selected also had 4-monthly sputum cytologic examinations. Of the 354 lung cancers that developed in these men during the screening period and a 2-year post-screening period, nearly half were adenocarcinomas and nearly one third were epidermoid (or squamous) carcinomas. Lung cancers detected by screening were more likely to be early stage (53 per cent) and resectable (65 per cent) than were those discovered by other means (for example, symptoms) or in the post-screening period (20 per cent early stage, 32 per cent resectable). Sputum cytology was most effective in detecting squamous carcinomas early; chest radiographs were most effective for detecting adenocarcinoma. However, sputum cytology did not reduce lung cancer mortality among these men who were in a program of annual chest radiographic examinations. Overall 5-year survival for all of the men who developed lung cancer was 35 per cent, compared with the national average of 13 per cent.

Lung cancer detection. Results of a randomized prospective study in Czechoslovakia

A randomized prospective study of lung cancer detection was begun in 1976 to evaluate semiannual screening by radiologic and sputum cytologic study in comparison to screening at a 3-year interval, and to no screening. In a high-risk population of 6364 men (aged 40 to 64 years), the initial prevalence of lung cancer was 0.28% (18 cases), the annual incidence was 0.35% per year (66 cases during 3 years), the proportion of Stage I cases was 31% (26/84), and Stage II was 17% (14/84), "curative" resections were 27% (23/84), and 5-year survival was 23% (19/84). The study confirmed the ability of radiologic screening to detect lung cancer at an earlier stage when treatment by resection can be accomplished. The fate of a high-risk population submitted to screening was better than that of a population with no screening where lung cancer was discovered by symptoms, accidental x-rays, or at autopsy. A matter of lesser importance was the frequency of screening. The absolute numbers of 5-year survivors detected by screening were practically the same for either compared screening frequency.

Screening by nurse clinicians in cancer prevention and detection

Cancer screening by definition as an organized effort to detect cancer at an early stage when reduced mortality is possible. Cancer screening should not be the exclusive responsibility of comprehensive cancer centers or large urban medical centers. Rather, every health professional should play a significant role in the early detection and education of patients with cancer. As seen here, some of the most common and life-threatening forms of cancer are the most preventable. Practical and inexpensive steps exist for thorough and effective screening. Nurses educated in the early detection of cancer are in a unique position; they are on the front line of patient care in the promotion and education of patients to the many benefits of early detection. Preventive efforts, such as screening and early detection, coupled with vastly improved treatment modalities, provide a very real and potent weapon against diseases that only 20 years ago were considered incurable.

Screening for early lung cancer. Results of the Memorial Sloan-Kettering study in New York

The Memorial Sloan-Kettering lung cancer screening program was begun in 1974 to evaluate sputum cytology as a supplement to the annual chest x-ray examination for early detection and diagnosis. The 10,040 adult, male cigarette smokers who enrolled were randomly assigned to receive annual chest x-ray examinations only or a dual screen with annual chest x-ray examination and four monthly sputum cytology evaluation. Over 40 percent of the 288 who developed lung cancer were diagnosed in stage I, and their survival was 76 percent at five years; overall survival was 35 percent. Nearly one third of the lung cancers detected on first examination on the dual screen, and 14 percent of those on subsequent examinations were found by cytologic examination. The same number of cancers developed in the x-ray screen only group, and were diagnosed at a later date. Despite the delay, survival and mortality were the same, suggesting that the squamous carcinomas detected by cytologic examination alone are very slow growing and tend to remain localized until detectable by x-ray examination.

Changes in physician behavior and cost savings associated with organizational recommendations on the use of "routine" chest X rays and multichannel blood tests

Group Health Cooperative (GHC) of Puget Sound is a prepaid health plan with 285,000 enrollees and 340 physicians. In 1978 a "criterion analysis" approach was instituted to develop medical staff recommendations about preventive care. By 1979, two recommendations evolved discouraging the "routine" use of: (a) chest X rays (CXR) and (b) multichannel blood tests (MCBT) in asymptomatic adults at "routine" physical exams. An extensive educational campaign was conducted to change physician behavior in use of these tests. Usage patterns were measured both before and after the recommendations and educational programs. A fivefold fall in "nonindicated" use of CXR and a 1.5-fold fall in MCBT occurred. A cost savings of $166,582 (annually) resulted from this diminution in "nonindicated" testing. The decision-making process and educational campaign technology hold promise for GHC's preventive medicine program in the future. These results may have applicability to other health care organizations.

The costs and effects of screening for cancer among asbestos-exposed workers

Several million workers previously exposed to asbestos are at risk each year for developing asbestosis or cancer as a result of this exposure. We have reviewed the data on the general effectiveness of screening for the two most common cancers found in these workers (lung cancer and colo-rectal cancer) and have compared the relative benefits expected from screening workers for colon or lung cancer 10, 15, 20 or 25 yr after their initial exposure. Although lung cancer is common, there is little evidence at present that screening for this disease is effective in reducing mortality. Colon cancer is less common, there is more evidence that screening is effective, and the relatively high risk of dying from this disease in asbestos workers makes screening useful. The cost per additional year of life extended by screening is only a few hundreds of dollars, making colon cancer screening programs for asbestos workers considerably more cost-effective than most other screening programs. We conclude that occupational safety agencies should consider implementing screening programs for colon cancer in workers exposed to asbestos over 10 yr ago.

Lung cancer as a complication of prolonged survival in patients with lymphoma

As a result of successful therapy with radiation, chemotherapy, or both, patients with lymphoma are living longer. The improved survival has led to an increased awareness of long term complications including second or subsequent malignancies. We have recently seen six cases of lung cancer--three adenocarcinomas, one squamous cell carcinoma, and two small cell carcinomas--among 655 patients with Hodgkin and non-Hodgkin lymphoma suggesting a possible association, although other carcinogenic or cocarcinogenic factors cannot be eliminated.

Detection of true pathologic stage I lung cancer in a screening program and the effect on survival

One-hundred-sixty-nine lung cancers have occurred to date among 10,040 cigarette smoking men who participated in the New York Lung Cancer Detection Program. Almost 40% of the cases, 65, were still Stage I when their disease was diagnosed; 62 had thoracotomy and resection, and in 57, mediastinal node dissection confirmed that the mediastinum was free of metastases ("true pathologic" Stage I). Fifty-four of the 62 (87%) are still alive at this time, while only 15 of 104 (14%) of those with Stage II and III lung cancers are alive. Only two patients of the 62 in Stage I who were treated by resection died of lung cancer, both with T2 tumors. Two others are alive with metastases, one died postoperatively, and five died of other causes without evidence of lung cancer. The estimated probability of survival for true Stage I lung cancer is over 90% at five years, and close to 40% of all lung cancers can be detected in this favorable stage by present radiologic and cytologic screening techniques.

Occult carcinomas of the lung

In 1974, we reported 26 patients with roentgenographically occult lung carcinomas. In 13 (50%) of them, the tumor was localized and treated by resection while the disease was still early (Stage I). These patients have done well, and the median survival has reached 8 years. We have added 21 patients to the series since our last report. Localization was by fiberoptic bronchoscopy in all 21. Fourteen of these patients had Stage I disease and were treated by resection. Of a total of 27 patients with early disease treated by resection, none has had recurrence after follow-up extending from 2 months to 20 years. However, in 21 of the entire 47 patients (45%) a second carcinoma developed, 15 (71%) of which were second lung carcinomas.

Lung cancer: the pathologist's role in management

The prognosis of lung cancer is poor, although there is hope for improvement based on early detection through cytologic screening and the use of newer treatment protocols. Although there are difficulties inherent in classifying tumors by type on the basis of very small samples of lesions that are not necessarily homogeneous, the pathologist should use a standard classification that sets forth definite criteria for the differentiated types. Among the common tumors, a small, well-differentiated epidermoid carcinoma with no evidence of lymph node involvement has the most favorable prognosis, and a small-cell anaplastic or oat cell carcinoma the worst. Results of therapy for tumors of other types show much variation. Future studies of lung cancer should include careful clinical staging and, when resection is done, pathologic staging. More study of the immunomorphology of lymph nodes is needed, as this may provide highly useful prognostic information.