solitary Pulmonary nodular lung adenocarcinoma

Original Research n Thoracic

Imaging

Note: This copy is for your personal non-commercial use only. To order presentation-ready copies for distribution to your colleagues or clients, contact us at www.rsna.org/rsnarights.

Solitary Pulmonary Nodular Lung Adenocarcinoma: Correlation of Histopathologic Scoring and Patient Survival with Imaging Biomarkers1 Ho Yun Lee, MD Ji Yun Jeong, MD Kyung Soo Lee, MD Hyo Jin Kim, MD Joungho Han, MD Byung-Tae Kim, MD Jhingook Kim, MD Young Mog Shim, MD Jae-Hun Kim, PhD Inyoung Song, MD

Purpose:

To evaluate the usefulness of histopathologic scoring for survival prediction in patients with solitary pulmonary nodular (SPN) lung adenocarcinomas and to correlate the histopathologic scoring with the results of computed tomography (CT) and fluorine 18 fluorodeoxyglucose positron emission tomography (PET)/CT.

Materials and Methods:

This retrospective study was institutional review board approved and the requirement for informed consent was waived. A total of 148 patients with SPN lung adenocarcinoma underwent PET/CT and CT. Correlations between histopathologic scores estimated by using two predominant histologic subtypes from each surgically resected specimen and the mass of the nodule at CT or maximum standardized uptake value (SUVmax) at PET/CT were assessed. Disease-free survival (DFS) was estimated by using the Kaplan-Meier method, and the log-rank test was used to evaluate differences in each histopathologic subtype.

Results:

In 135 (91%) patients, tumors had a mixed subtype. The most frequently observed histologic subtypes, in decreasing order, were acinar (51%), lepidic (18%), solid (10%), and papillary (9%). DFS rates at 5 years were higher than 90% for the group of patients with nodules that showed the lepidic growth pattern, and 50% for patients with nodules that showed the micropapillary pattern. The pathologic score proved to be a significant predictor of DFS (P , .001). Both SUVmax and the mass of the nodule were closely correlated with pathologic score.

Conclusion:

Pathologic scoring appears to help predict DFS in patients with SPN lung adenocarcinoma and shows close correlation with imaging biomarkers including the mass of the nodule at CT and SUVmax at PET/CT. RSNA, 2012

q

1

From the Department of Radiology and Center for Imaging Science (H.Y.L., K.S.L., H.J.K., J.H.K., I.S.), and Departments of Pathology (J.Y.J., J.H.), Nuclear Medicine (B.T.K.), and Department of Thoracic Surgery (J.K., Y.M.S.), Samsung Medical Center, Sungkyunkwan University School of Medicine, 50 Ilwon-Dong, Kangnam-Ku, Seoul 135-710, Korea. Received August 23, 2011; revision requested October 13; revision received November 18; accepted February 23, 2012; final version accepted March 26. Address correspondence to K.S.L. (e-mail: [email protected]). RSNA, 2012

q

884

radiology.rsna.org n Radiology: Volume 264: Number 3—September 2012

THORACIC IMAGING: Solitary Pulmonary Nodular Lung Adenocarcinoma

T

he 5-year overall survival rate for patients with stage 1 lung cancer approaches 67%, and 30%–40% of these patients have a recurrence of the disease at a later date (1–3). This indicates the need to identify robust prognostic biomarkers to help predict which patients with early-stage cancer are at the highest risk for recurrent disease and, therefore, are candidates for more aggressive surveillance or adjuvant therapy. There is increasing evidence that the characteristic histologic heterogeneity of lung adenocarcinomas means that there is also diversity in prognoses among individual tumors. (4–6). For example, mixed subtype tumors that have micropapillary or solid patterns are associated with worse outcomes than other subtypes (7–9). In 2011, the International Association for the Study of Lung Cancer, the American Thoracic Society, and the European Respiratory Society proposed a new international multidisciplinary classification system for lung adenocarcinoma (10). In this system, lesions that were formerly considered to be bronchioloalveolar carcinomas (BAC) are now classified as preinvasive lesions because patients with these 2–3-cm lesions have a 100% disease-free survival (DFS) rate. In addition, because most invasive lung adenocarcinomas consist of a mixture of histologic subtypes and the word “predominant” was appended to all categories of invasive adenocarcinomas, the classification of “adenocarcinoma, mixed subtype,” was removed.

Advances in Knowledge nn Of 148 tumors, 135 (91%) were composed of mixed patterns, and the most predominant subtypes were, in a decreasing order, acinar (51%), lepidic (18%), solid (10%), and papillary (9%) patterns. nn A pathologic scoring system appeared to help predict patient survival and had close correlation with both SUVmax at PET/CT (P , .001) and mass of the nodule at CT (P = .004).

Consequently, in this new classification system, common subtypes include lepidic-growth predominant, acinar predominant, papillary predominant, micropapillary predominant, and solid pattern predominant. However, the problem of histologic subtyping, subsequent tumor scoring or grading, and prognosis prediction for lung adenocarcinomas is that subtyping is estimated mainly by using a resected surgical specimen (whole tumor) postoperatively, not by using core biopsy or cytologic material preoperatively (7,9). Therefore, preoperative prognostic prediction by using a preoperative biopsy pathologic specimen in small lung adenocarcinomas seems incomplete, if it is even possible. Thus, it is desirable to predict patient prognosis by using a preoperative surrogate biomarker with imaging tools that can be substituted for a histopathologic scoring system. In the study by Lee et al (11), the pathologic non-BAC ratio proved to be the only independent risk factor for poor prognosis in patients with a solitary pulmonary nodular (SPN) lung adenocarcinoma. However, in the Lee et al study, the researchers evaluated only the extent of BAC versus non-BAC components. They did not analyze the tumor histologic subtypes. Therefore, the purpose of this study was to evaluate the usefulness of histopathologic scoring for survival prediction in patients with SPN lung adenocarcinoma and to correlate the histopathologic scoring with the results of computed tomography (CT) and fluorine 18 fluorodeoxyglucose (FDG) positron emission tomography (PET)/CT.

Implications for Patient Care nn Imaging biomarker study results are closely correlated with pathologic score and thus enable prediction of patient survival. nn Preoperative evaluation of tumor grade by using PET/CT or CT may allow the selection of further staging workup and appropriate therapeutic strategies for patients with small lung adenocarcinomas.

Radiology: Volume 264: Number 3—September 2012 n radiology.rsna.org

Lee et al

Materials and Methods Our institutional review board approved this retrospective study with a waiver of informed consent. However, CT and PET studies were performed with patients’ informed consent for safety issues.

Patients The study group included 148 consecutive patients with SPN adenocarcinoma who were treated with surgery alone or surgery and postoperative adjuvant therapy at Samsung Medical Center (Seoul, Korea) between July 2003 and December 2007. Patients were identified in a search of the lung cancer surgical registry database of the department of thoracic surgery. Inclusion criteria for our study were as follows: (a) SPN adenocarcinomas (3 cm or less in diameter at CT) of clinical stage 1A or 1B with no evidence of malignant satellite nodules (proved with imaging study or lung biopsy beforehand) and no hilar or mediastinal lymphadenopathy on imaging study or at mediastinoscopy, (b) first treatment with surgery alone, with or without postoperative adjuvant treatment, (c) no other malignant tumor history for 5 years before the Published online before print 10.1148/radiol.12111793 Content codes: Radiology 2012; 264:884–893 Abbreviations: AIS = adenocarcinoma in situ BAC = bronchioloalveolar carcinoma DFS = disease-free survival FDG = fluorine 18 fluorodeoxyglucose SPN = solitary pulmonary nodular SUVmax = maximum standardized uptake value Author contributions: Guarantor of integrity of entire study, K.S.L.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; approval of final version of submitted manuscript, all authors; literature research, H.Y.L., J.Y.J., K.S.L.; clinical studies, H.Y.L., K.S.L., H.J.K., J.H., B.T.K., J.K., Y.M.S., J.H.K., I.S.; experimental studies, J.Y.J.; statistical analysis, H.Y.L., K.S.L.; and manuscript editing, H.Y.L., J.Y.J., K.S.L., Y.M.S. Potential conflicts of interest are listed at the end of this article.

885


diagnosis of lung adenocarcinoma, (d) both integrated FDG PET/CT and chest CT studies acquired within 1 month before resection, (e) patient alive 30 days after surgery, and (f) no loss of patient to follow-up during the 12 months after surgery. The cases were reviewed according to International Multidisciplinary Lung Adenocarcinoma Classification criteria (10) and staged according to the seventh edition of the TNM classification for lung cancer (12,13).

Imaging and Interpretation Imaging characteristics of each primary lung tumor were evaluated by using chest CT and the PET component images of PET/CT. PET/CT and chest CT were performed within a 2-week period (average time interval, 7 days; range, 0–14 days). FDG PET/ CT images were acquired by using a PET/CT device (Discovery LS; GE Healthcare, Milwaukee, Wis), which consisted of a PET scanner (Advance NXi; GE Healthcare) and an eightsection CT scanner (Light-Speed Plus; GE Healthcare). The imaging methods were described in detail in a previous report (14). Helical CT images were obtained with an eight– (LightSpeed Ultra, GE Healthcare) or 16–detector row (LightSpeed16, GE Healthcare) CT scanner. Unenhanced CT images were obtained with the following parameters: detector collimation, 0.625 mm; field of view, 34.5 cm; beam pitch, 1.35 or 1.375; gantry speed, 0.6 second per rotation; 120 kVp; 150–200 mA; and section thickness, 1.25 mm for transverse images. All imaging data were reconstructed by using soft-tissue algorithms. A nuclear medicine physician (B.T.K., with 13 years of experience in PET/CT interpretation) who was unaware of clinical and pathologic results evaluated the PET images. For a semiquantitative analysis of FDG uptake, regions of interest were placed over the most intense areas of FDG accumulation. In some patients, nodular FDG uptake could not be identified on the PET component images of their PET/CT study. In those patients, 886

Lee et al

a region of interest was drawn in a presumed nodular location, taking into account the CT component images of PET/CT. FDG uptake in the regions of interest was analyzed by using the maximum standardized uptake value (SUVmax). Chest CT data were sent directly to a picture archiving and communication system (Path-Speed or Centricity 2.0; GE Healthcare, Mt. Prospect, Ill), which displayed all image data on two monitors (1536 3 2048 matrix, eight-bit viewable grayscale, 60–footlambert [205.6 candela per square meter] luminescence). The monitors were used to view both mediastinal (width, 400 HU; level, 20 HU) and lung (width, 1500 HU; level, 2700 HU) window images. The time between CT study and surgical tumor removal ranged from 1 to 28 days (mean, 17.5 days; median, 13.5 days). CT images were assessed retrospectively for nodule size, appearance (pure ground-glass opacity, part solid, and solid), volume, and mass of the nodule, independently by one chest radiologist (I.S., with 2 years of experience in thoracic CT interpretation) and one radiologic physicist (J.H.K., with 3 years of experience in radiological physics), who were unaware of the clinical PET findings and histologic diagnoses. For nodule size, the researchers measured the longest tumor diameter on the transverse lung window image where the largest nodule dimension appeared. For nodule volumetry, they delineated nodule outlines electronically on all transverse images on which any portion of the nodule appeared. Then, the computer automatically calculated the nodule volume by multiplying the number of voxels by the unit volume of a voxel (15). Physical density (in grams per cubic centimeter) is roughly linear with CT attenuation (Hounsfield units) (16). From this linear relationship, the physical density of the nodule could be extrapolated from the mean CT attenuation measurements of the nodule. Because the nodules in our study did not contain

calcification and were scanned without contrast medium injection, the conversion process was feasible in all cases. The mass of the nodule (in grams) was calculated by multiplying nodule volume (in cubic centimeters) by mean nodule density.

Pathologic Evaluation Each resected specimen (entire tumor) was evaluated with standard pathologic methods as described in the surgical pathologic dissection manual of the Department of Pathology (17). All resected specimens were designated R0 (no residual tumor at the primary tumor site after surgical resection). Two experienced lung pathologists (J.Y.J. and J.H., with 5 and 18 years of experience in lung pathology, respectively) jointly interpreted all tissue sections. Tumor tissue samples of approximately 10 mm in diameter were obtained and each was placed on a slide. First, for each case, comprehensive histologic subtyping was performed for the primary tumor in a semiquantitative manner. The extent of existent tumor histologic subtypes and central fibrosis was quantified to the nearest 5% level, adding up to a total of 100% subtype components per tumor as described in Figure 1 (10). As previously reported (11,18), the central fibrosis region was defined as the areas of fibroblastic focus in which a moderate or abundant amount of collagen or hyalinized tissue was clearly noted. The region did not contain any adenocarcinoma in situ (AIS) components. Next, the most predominant and second most predominant patterns in a mixed-type adenocarcinoma were defined as the histologic subtypes that comprised the highest and second highest percentage of the tumor. When evaluating the predominant pattern, the central fibrosis area and its extent were disregarded. Next, tumors were graded by using a three-tier grading system (9) (Fig 2). Grade 1 included histologic subtypes of AIS, minimally invasive adenocarcinoma, and the lepidic pattern of invasive



Lee et al

Figure 1

Figure 1: Photomicrographs show morphology of invasive adenocarcinoma subtypes, including (a) lepidic pattern, (b) acinar pattern, (c) papillary pattern, (d) micropapillary pattern, and (e) solid pattern (hematoxylineosin stain; magnification ×200).

adenocarcinoma in accordance with the 2011 international lung adenocarcinoma classification system. Grade 2 corresponded to tumors that showed acinar or papillary patterns. Grade 3 corresponded to the tumors that showed micropapillary or solid patterns (9). When tumors were composed predominantly of a variant pattern, the tumors were removed from our study because they were uncommon and heterogeneous in terms of biologic behavior and prognosis (10). Finally, pathologic tumor scores were calculated by adding the two most predominant grades in each tumor (9) (Table 1). When tumors were composed of a pure

histologic type, the primary grade was used, and the score was determined by doubling the primary tumor grade.

Treatment and Follow-up Evaluation For stage 1 cancers, all 127 patients had undergone either sublobar resections (wedge resections or segmentectomies) or lobectomies. None of the patients with stage 1 cancer underwent adjuvant therapy. Of the patients with stage 2 and 3A cancers, 17 had undergone sublobar resections and the remaining four had undergone either pneumonectomy or lobectomy. These 21 patients were treated with adjuvant concurrent combined chemotherapy and radiation


therapy (five cycles of cisplatin chemotherapy [one cycle: an intravenous injection of cisplatin, 25 mg/m2, on day 1 for 1 week] and 25 Gy of radiation in 1st week). Median follow-up time after surgery for all patients was 50 months (range, 36–92 months). Sixty-nine percent (102 of 148) of patients had a follow-up time of less than 5 years. By July 2011, 32 (22%) patients developed recurrent disease after surgical resection, and the median time to recurrence was 39 months (range, 10–76 months). Of those patients, 19 had pulmonary metastasis. Metastases to mediastinal lymph nodes, pleura, brain, liver, or bone were detected in other patients. None had local tumor recurrences.

Statistical Analysis For measuring tumor size, volume, and mass of the nodule, the means of values measured by two observers were recorded, and interobserver variability was calculated by using repeated measures data analysis for the intraclass correlation coefficient. DFS was defined as the time from surgery to recurrence, lung cancer–related death, or last follow-up evaluation. DFS was estimated by using the Kaplan-Meier method. The log-rank test was used to evaluate the statistical significance of differences in survival among patient groups with different scores. Cause of death was determined from death certificates or through correspondence with the physician in charge. 887


Lee et al

Figure 2

using software (SPSS, version 19.0, 2010; SPSS, Chicago, Ill). A P value less than .05 was considered to indicate a statistically significant difference.

Results

Figure 2: Lung adenocarcinoma in a 67-year-old woman. (a) Targeted view of transverse lung window CT image shows 25-mm solid nodule with peripheral ground-glass opacity (arrow). (b) On mediastinal window image, solid area remains, but ground-glass opacity area is not visible. Mass of nodule was estimated as 13.3 g. (c) PET/CT image shows FDG uptake with SUVmax of 7.8. (d) Photomicrograph shows internal scar tissue (*), surrounding areas of acinar (**) and solid (#) adenocarcinoma patterns, and lepidic pattern (arrows) showing uniform cuboid cellular proliferation along alveolar walls only at tumor periphery. (Hematoxylin-eosin stain; original magnification, 310.) (e) Schematic of tumor components shows estimated percentages of grade 1 (yellow area, 10%), grade 2 (blue area, 50%), grade 3 (green area, 30%), and central fibrosis (red area, 10%).

Causes of death other than recurrence of the tumor were censored from the survival analysis. Interrelationships among the mass of the nodule at CT, FDG uptake at PET, and the pathologic score determined on the basis of the two most predominant histologic subtypes were assessed by 888

using the Spearman correlation coefficient test. Differences in statistical significance among subcategorized groups that were contingent on pathologic scores were compared in terms of the mass of the nodules at CT and FDG uptake at PET by using the Kruskal-Wallis test. Statistical analyses were performed by

CT Findings and Patient Demographics Of the 148 nodules, 10 showed groundglass opacity, 131 were part solid, and seven were solid nodules. Tumors were 22 mm (range, 7–30 mm) in mean diameter. Their sizes were less than or equal to 10 mm in diameter in 15 patients and greater than 10 mm in diameter in 133 patients. Interobserver agreement for measurements of the volume and mass of the nodules were moderate, and agreement for size was high. Intraclass correlation coefficients were 0.71 (95% confidence interval [CI]: 0.67–0.75) for volume, 0.69 (95% CI: 0.63–0.75) for mass, and 0.88 (95% CI: 0.85–0.91) for the size of nodules. The clinicopathologic characteristics of the 148 patients with lung adenocarcinomas included in this study are summarized in Tables 2 and 3. Histopathologic Characteristics and Association with Outcome Among 148 patients, 95 (64%) had nodules that showed the acinar pattern,



Table 2

Table 1

Patient Characteristics

Pathologic Scoring of Lung Adenocarcinomas Score

Lee et al

Grades of Two Most Predominant Subtypes*

2 3

1 and 1 1 and 2 or 2 and 1

4

2 and 2, 1 and 3, or 3 and 1

5

2 and 3 or 3 and 2

6

3 and 3

Characteristic Representative Pattern

Sex Male 66 (45) Female 82 (55) Median age (y) 59 (37–80)* Smoking Nonsmoker 86 (58) Current or former smoker 62 (42) Lymph node metastases Mediastinal 7 (5) Hilar 11 (7) No lymph nodes 130 (88) Subtype predominance† AIS 16 (5) MIA 10 (3) Lepidic 53 (18) Acinar 152 (51) Papillary 25 (9) Solid 28 (10) Micropapillary 12 (4) Pathologic score† 2 13 (9) 3 26 (18) 4 80 (54) 5 24 (16) 6 5 (3) Median follow-up period (mo) 50 (36–92)* Recurrence Yes 32 (22) No 116 (78) DFS at 5 years 0.82§

AIS, minimally invasive adenocarcinoma Mixed subtype, well differentiated with lepidic and acinar or papillary patterns Mixed subtype, moderately differentiated with acinar and/or papillary patterns, or lepidic with solid or micropapillary patterns, pure acinar or pure papillary Mixed subtype, poorly differentiated with acinar or papillary and micropapillary or solid patterns Mixed subtype, poorly differentiated with predominant solid and micropapillary patterns, pure solid, or pure micropapillary patterns

Note.—Tumors were graded according to the International Multidisciplinary Lung Adenocarcinoma Classification and scored on the basis of the two most predominant subtypes (9). * Grade 1 = AIS or minimally invasive adenocarcinoma and lepidic pattern in mixed tumors; grade 2 = acinar and papillary patterns; grade 3 = micropapillary and solid patterns (10).

27 (18%) showed the lepidic pattern, and eight (5%) were AIS. One-hundred thirty-five (91%) of the tumors were mixedsubtype adenocarcinomas. The most frequently observed histologic subtypes (considering the two most predominant subtypes we observed in each tumor) were acinar (51%), lepidic (18%), solid (10%), and papillary (9%). DFS at 5 years was higher than 90% for the group of patients with AIS, minimally invasive adenocarcinoma, and nodules that showed the lepidic pattern, and 50% for nodules that showed the micropapillary pattern (Table 4). DFS curves for patient groups according to pathologic score are shown in Figure 3. Pathologic scores, which were computed as the sum of the tumor grades of the two most predominant subtypes, were found to be significant predictors of DFS (P , .001).

Correlation among Radiologic, Metabolic, and Histopathologic Factors In Tables 4 and 5 and in Figure 4, the interrelationship between the mass of the nodule at CT, SUVmax at PET, and pathologic score are expressed. SPN adenocarcinomas with a high pathologic score were shown to have higher mass and higher FDG uptake than those with a low score. A significant difference in

SUVmax (P , .001) and the mass of the nodule (P , .004) were found among tumors of different pathologic scores. Moreover, post hoc Bonferroni comparisons showed all cross-score parings to be statistically distinct with regard to SUVmax and the mass of the nodule, except the paring of pathologic scores 2 and 3. SUVmax was strongly correlated with pathologic score. The mass of the nodule was also strongly correlated with pathologic score (Fig 4).

Discussion Lung adenocarcinoma is highly heterogeneous, and it usually has variable combinations of two or more histologic subtype patterns (19). In the international criteria proposal (10), new concepts were introduced, such as the classification of small solitary adenocarcinomas with pure lepidic growth as AIS and those with predominantly lepidic growth but with invasion of 5 mm or less as minimally invasive adenocarcinoma. Invasive adenocarcinomas were classified by the predominant pattern after using comprehensive histologic subtyping with lepidic (formerly most mixed subtype tumors with nonmucinous BAC), acinar, papillary, and solid patterns; micropapillary was added as a


No. of Patients

Note.—Unless otherwise indicated, data in parentheses are percentages. MIA = minimally invasive adenocarcinoma. * Data in parentheses are the range. †

Data include the two most predominant subtypes in each lesion (n = 296).

‡

Score is the sum of the two most predominant tumor grades in each lesion (9).

§

Number is proportion of patients with DFS at 5 years.

new histologic subtype. Variants include invasive mucinous adenocarcinoma (formerly mucinous BAC), colloid, fetal, and enteric adenocarcinoma. Our study results included three main findings: (a) The new International Multidisciplinary Lung Adenocarcinoma Classification scheme, along with our modifications, which reflect the quality and quantity of the pathologic pattern, can help in the prediction 889


Lee et al

Table 4 Imaging and Prognostic Features Predominant Subtype AIS MIA Lepidic Acinar Papillary Solid Micropapillary Total

No. of Patients

Size (cm)*

Volume (cm3)*

Mass (g)*

SUVmax*

Follow-up Period (mo)*

Recurrence†

DFS at 5 Years‡

8 5 27 95 7 4 2 148

1.3 6 0.7 1.3 6 0.6 2.0 6 0.9 2.3 6 0.8 1.7 6 0.6 2.8 6 0.3 2.6 6 0.6 2.2 6 0.8

2.2 6 1.0 2.0 6 1.7 3.6 6 0.8 5.9 6 1.1 4.1 6 1.0 6.8 6 2.1 4.7 6 1.5 5.0 6 1.2

3.6 6 2.1 3.2 6 1.8 5.4 6 1.3 11.5 6 2.1 8.1 6 1.5 14.4 6 4.5 10.6 6 2.0 9.5 6 3.0

0 1.2 6 1.3 2.3 6 2.3 5.4 6 3.6 3.9 6 1.1 11.0 6 6.6 4.7 6 3.1 4.7 6 3.7

58.9 6 15.2 56.4 6 15.5 57.2 6 14.3 52.1 6 11.7 54.1 6 17.6 58.2 6 13.5 64.3 6 9.7 49.7 6 15.1

0 0 3 (11) 24 (25) 2 (29) 2 (50) 1 (50) 32

1 1 0.93 0.78 0.71 0.75 0.50 0.82

Note.—Classified According to the International Multidisciplinary Lung Adenocarcinoma Classification system. MIA = minimally invasive adenocarcinoma. * Data are means 6 standard deviation. †

Data in parentheses are percentages.

‡

Data are proportions of total number of patients.

Table 3 Pathologic Stage versus Predominant Histologic Subtype Predominant Subtype AIS MIA Lepidic Acinar Papillary Solid Micropapillary Total

Stage 1A

Stage 1B

Stage 2A

Stage 2B

Stage 3A

Total

8 5 25 51 3 3 0 95

0 0 2 23 4 1 2 32

0 0 0 4 0 0 0 4

0 0 0 10 0 0 0 10

0 0 0 7 0 0 0 7

8 5 27 95 7 4 2 148

Note.—Data are number of patients. Lesions were staged according to the 7th revision of the TNM classification. MIA = minimally invasive adenocarcinoma.

of tumor malignancy grade and patient survival. (b) Both the SUVmax on PET images and the mass of the nodule on the CT images are closely correlated with the pathologic score; and (c) both SUVmax and the mass of the nodule at CT could help to better stratify the prognoses of patients, particularly for those with higher grade tumors. Our results may have practical implications. With the help of preoperative imaging biomarker surrogate study results, we were able to determine which patients with small lung adenocarcinomas may develop recurrent disease after the initial treatment and which ones may not, even without examining a surgical tumor specimen. The histologic subtyping and scoring of 890

SPN lung adenocarcinoma appear to correlate with the mass of the nodule at CT or the SUVmax at PET. In addition, the mass of the nodule or SUVmax can be linked to the biologic aggressiveness of the tumor, and eventually to the surgical outcome and patient survival (11,20–23). The tumor grading system has been developed to correlate histopathologic characteristics of SPN lung adenocarcinomas with prognostic importance, and subsequently to compare the results with overall survival (7,9). In a study by Barletta et al (7), the percentage of solid patterns as a reflection of tumor architecture, the degree of cytologic atypia, and the mitotic count were evaluated to seek a prognostically

relevant grading system for lung adenocarcinomas. In the Barletta et al study, a grading system that incorporated the percentage of solid pattern and the degree of cytologic atypia, which was computed as the sum of the architecture score and cytologic atypia, appeared to be an independent predictor of survival. Moreover, there was a direct correlation between the degree of cytologic atypia and the amount of solid pattern. However, overall survival was not associated with mitotic count. In their study, the predominant histologic pattern versus the patient prognosis was not assessed because of a lack of tumors with dominant BAC (or AIS) or micropapillary patterns in their cohort. In another study, Sica et al (9) developed several separate scoring systems, including (a) the sum of the two most predominant grades (grade 1 = BAC or AIS, grade 2 = acinar and papillary patterns, and grade 3 = solid and micropapillary patterns) (b) the sum of the two highest grades, and (c) the sum of the predominant and the highest grade. They tried to evaluate both the quantity and combination of specific patterns of histopathologic subtypes for tumor metastatic potential. A scoring system based on the two most predominant grades was the best for categorizing patients as at low or high risk for recurrence of disease or death. In our study, a pathologic scoring system reflecting the two most



predominant grades was adopted for patient prognosis determination, and that scoring system showed good performance in separating the groups for the risk of tumor recurrence. The identification of patients with SPN lung adenocarcinoma who have a higher risk for recurrence and who may benefit from adjuvant therapy has been a target of intense investigation. There have been several reports suggesting that tumor gene expression profiling has prognostic relevance and can be used to help predict disease recurrence (2,24–26). Most of the gene expression studies rely on microarray technology, which may limit the practicability of this approach. We found both the SUVmax at PET and the mass of the nodule at CT help in the categorization of patients, particularly in those with high-grade tumors (more solid tumor at CT). That result may suggest the necessity of the combined use of both the mass of the nodule and SUVmax as preoperative imaging biomarkers (as imaging prognostic surrogates). It is also important to document the subtypes of adenocarcinoma in surgical pathologic reports, because the subtypes BAC or AIS have been shown to correlate with molecular abnormalities that help predict response to targeted therapies (27,28). Recently, Russell et al (29) reported that the new classification has advantages not only for individual patient care but also for better selection and stratification of clinical trials and molecular studies. Our study had several limitations. First, the small cohort size and different proportions of each predominant pattern limited our study; in the future, the results should be prospectively collected and validated, ideally in population-based studies. Second, the semiquantitative approach to the assessment of adenocarcinomas that is described by the International Association for the Study of Lung Cancer, the American Thoracic Society, and the European Respiratory Society has yet to be validated further. Until then, the international classification system has been given a weak recommendation with low-quality evidence (10). Third, survival in patients with SPN lung adenocarcinoma is not determined

Lee et al

Figure 3

Figure 3: DFS curves for groups according to scores based on most frequently observed histologic subtype in consideration of two most predominant histologic subtypes.

Table 5 Correlation of Imaging Biomarker Features with Pathologic Grading System and Scores Semiquantification of Tumor Components† Biomarker Feature Tumor size 1 cm .1 cm Tumor volume Tumor mass SUVmax

Pathologic Score*

Central Fibrosis

Grade 1

Grade 2

Grade 3

0.333§ 20.417 0.240§ 0.215‡ 0.479§ 0.559§

0.031 ... 20.084 0.035 0.111 0.043

20.296§ 0.165 20.158 20.256§ 20.581§ 20.491§

0.203‡ 20.177 0.093 0.198‡ 0.419§ 0.289§

0.192‡ 0.196 0.147 0.169 0.235§ 0.350§

Note.—Data are R values (correlation coefficients). * Based on two most predominant subtypes (10). †

Grade 1 = AIS; grade 2 = acinar and papillary patterns; and grade 3 = micropapillary and solid patterns (9).

‡

P , .05.

§

P , .01.

by histopathologic subtypes and imaging biomarker studies alone. Prognoses may also be influenced by other factors such as tumor stage, molecular features such as epidermal growth factor receptor and KRAS mutation positivity, and treatment modalities including adjuvant chemotherapeutic agents. Therefore,


prognosis for patients with SPN lung adenocarcinomas may be determined by integrating histopathologic subtypes, imaging biomarker studies, tumor stage, tumor molecular features, and given or scheduled treatment methods. In our study of 148 patients with clinical stage 1 SPN lung adenocarcinomas, 21 (14%) 891


Lee et al

Figure 4

Figure 4: Box and whisker plots show correlation between preoperative imaging biomarkers and postoperative pathologic scores in SPN of the lung. (a) Tumor mass of the nodule volume at CT versus pathologic scores (P = .004), (b) SUVmax at PET/CT versus pathologic scores (P , .001). Post hoc Bonferroni comparisons showed that all between-group differences were significant except those between scores 2 and 3 (P , .01 for nodule mass and P , .001 for SUVmax).

patients proved to have pathologic stage 2 or higher disease and received adjuvant chemotherapy. Finally, in our study, the range of follow-up time after surgery was relatively wide (36–92 months). This might have been an influence on patient prognosis. Selecting a proper follow-up time is important. In conclusion, pathologic scoring by using the newly proposed lung adenocarcinoma classification system (grading the two most predominant histologic subtypes of the carcinoma) appears to help predict patient survival in SPN lung adenocarcinoma and shows close correlation with imaging biomarker studies. Because imaging biomarker study results are closely correlated with pathologic score, and thus, enable prediction of patient survival, the preoperative evaluation of tumor grade by using PET/CT or CT may allow selection of further staging workup and appropriate therapeutic strategies for patients with small lung adenocarcinomas. Disclosures of Potential Conflicts of Interest: H.Y.L. No potential conflicts of interest to disclose. J.Y.J. No potential conflicts of interest to disclose. K.S.L. No potential conflicts of inter-

892

est to disclose. H.J.K. No potential conflicts of interest to disclose. J.H. No potential conflicts of interest to disclose. B.T.K. No potential conflicts of interest to disclose. J.K. No potential conflicts of interest to disclose. Y.M.S. No potential conflicts of interest to disclose. J.H.K. No potential conflicts of interest to disclose. I.S. No potential conflicts of interest to disclose.

6. Yim J, Zhu LC, Chiriboga L, Watson HN, Goldberg JD, Moreira AL. Histologic features are important prognostic indicators in early stages lung adenocarcinomas. Mod Pathol 2007;20(2):233–241.

References

8. Makimoto Y, Nabeshima K, Iwasaki H, et al. Micropapillary pattern: a distinct pathological marker to subclassify tumours with a significantly poor prognosis within small peripheral lung adenocarcinoma (,/=20 mm) with mixed bronchioloalveolar and invasive subtypes (Noguchi’s type C tumours). Histopathology 2005;46(6): 677–684.

1. Hoffman PC, Mauer AM, Vokes EE. Lung cancer. Lancet 2000;355(9202):479–485. 2. Director’s Challenge Consortium for the Molecular Classification of Lung Adenocarcinoma, Shedden K, Taylor JM, et al. Gene expression-based survival prediction in lung adenocarcinoma: a multi-site, blinded validation study. Nat Med 2008;14(8):822–827. 3. Mountain CF. Staging classification of lung cancer. A critical evaluation. Clin Chest Med 2002;23(1):103–121. 4. Kobayashi N, Toyooka S, Ichimura K, et al. Non-BAC component but not epidermal growth factor receptor gene mutation is associated with poor outcomes in small adenocarcinoma of the lung. J Thorac Oncol 2008;3(7):704–710. 5. Miyoshi T, Satoh Y, Okumura S, et al. Early-stage lung adenocarcinomas with a micropapillary pattern, a distinct pathologic marker for a significantly poor prognosis. Am J Surg Pathol 2003;27(1):101–109.

7. Barletta JA, Yeap BY, Chirieac LR. Prognostic significance of grading in lung adenocarcinoma. Cancer 2010;116(3):659–669.

9. Sica G, Yoshizawa A, Sima CS, et al. A grading system of lung adenocarcinomas based on histologic pattern is predictive of disease recurrence in stage I tumors. Am J Surg Pathol 2010;34(8):1155–1162. 10. Travis WD, Brambilla E, Noguchi M, et al. International association for the study of lung cancer/american thoracic society/ european respiratory society international multidisciplinary classification of lung adenocarcinoma. J Thorac Oncol 2011;6(2): 244–285. 11. Lee HY, Han J, Lee KS, et al. Lung adenocarcinoma as a solitary pulmonary nodule: prognostic determinants of CT, PET, and



histopathologic findings. Lung Cancer 2009; 66(3):379–385. 12. Goldstraw P, ed. International Association for the Study of Lung Cancer staging manual in thoracic oncology. Orange Park, Fla: Editorial Rx Press, 2009. 13. Edge SB, Byrd DR, Compton CC, Fritz AG, Greene FL, Trotti A 3rd, eds. AJCC cancer staging manual. 7th ed. New York, NY: Springer, 2009. 14. Shim SS, Lee KS, Kim BT, et al. Non-small cell lung cancer: prospective comparison of integrated FDG PET/CT and CT alone for preoperative staging. Radiology 2005;236 (3):1011–1019. 15. de Hoop B, Gietema H, van de Vorst S, Murphy K, van Klaveren RJ, Prokop M. Pulmonary ground-glass nodules: increase in mass as an early indicator of growth. Radiology 2010;255(1):199–206. 16. Mull RT. Mass estimates by computed tomography: physical density from CT numbers. AJR Am J Roentgenol 1984;143(5):1101–1104.

Lee et al

19. Beasley MB, Brambilla E, Travis WD. The 2004 World Health Organization classification of lung tumors. Semin Roentgenol 2005;40(2):90–97.

25. Lau SK, Boutros PC, Pintilie M, et al. Three-gene prognostic classifier for earlystage non small-cell lung cancer. J Clin Oncol 2007;25(35):5562–5569.

20. Okada M, Tauchi S, Iwanaga K, et al. Associations among bronchioloalveolar carcinoma components, positron emission tomographic and computed tomographic findings, and malignant behavior in small lung adenocarcinomas. J Thorac Cardiovasc Surg 2007;133(6):1448–1454.

26. Potti A, Mukherjee S, Petersen R, et al. A genomic strategy to refine prognosis in earlystage non-small-cell lung cancer. N Engl J Med 2006;355(6):570–580. [Published retraction appears in Potti A, Mukherjee S, Petersen R, et al. N Engl J Med 2011;364(12):1176.]

21. Travis WD, Garg K, Franklin WA, et al. Evolving concepts in the pathology and computed tomography imaging of lung adenocarcinoma and bronchioloalveolar carcinoma. J Clin Oncol 2005;23(14): 3279–3287.

27. Ohtsuka K, Ohnishi H, Furuyashiki G, et al. Clinico-pathological and biological significance of tyrosine kinase domain gene mutations and overexpression of epidermal growth factor receptor for lung adenocarcinoma. J Thorac Oncol 2006;1(8): 787–795.

22. Shimizu K, Yamada K, Saito H, et al. Surgically curable peripheral lung carcinoma: correlation of thin-section CT findings with histologic prognostic factors and survival. Chest 2005;127(3):871–878.

28. Kim YH, Ishii G, Goto K, et al. Dominant papillary subtype is a significant predictor of the response to gefitinib in adenocarcinoma of the lung. Clin Cancer Res 2004; 10(21):7311–7317.

17. Lester SC. Manual of surgical pathology. 2nd ed. New York, NY: Elsevier Churchill Livingstone, 2006.

23. Chung CK, Zaino R, Stryker JA, O’Neill M Jr, DeMuth WE Jr. Carcinoma of the lung: evaluation of histological grade and factors influencing prognosis. Ann Thorac Surg 1982;33(6):599–604.

18. Suzuki K, Yokose T, Yoshida J, et al. Prognostic significance of the size of central fibrosis in peripheral adenocarcinoma of the lung. Ann Thorac Surg 2000;69(3):893–897.

24. Chen HY, Yu SL, Chen CH, et al. A five gene signature and clinical outcome in nonsmall-cell lung cancer. N Engl J Med 2007; 356(1):11–20.


29. Russell PA, Wainer Z, Wright GM, Daniels M, Conron M, Williams RA. Does lung adenocarcinoma subtype predict patient survival?: A clinicopathologic study based on the new International Association for the Study of Lung Cancer/American Thoracic Society/European Respiratory Society international multidisciplinary lung adenocarcinoma classification. J Thorac Oncol 2011; 6(9):1496–1504.

893