CURRENT STATUS
Lee E. Smith, M.D., Editor
Flow Cytometric Analysis of DNA Content in Colorectal Carcinoma Phillip A. Dean, M.D., Anthony M. Vernava III, M.D. From the D e p a r t m e n t of Surgery, Section o f Colon a n d Rectal Surgery, St. Louis University Medical Center, St. Louis, Missouri Over the past decade, flow cytometric DNA analysis has been employed by a number of investigators in an attempt to further define patient prognosis beyond classic pathologic staging. The results of these studies taken independently have been confusing; however, their cumulative effect suggests that flow cytometry is a useful prognostic indicator and can be used to further delineate prognosis within a specific pathologic stage. DNA nondiploid tumors are more likely to recur than diploid tumors, and patients with DNA nondiploid tumors have a poorer fiveyear survival than patients with DNA diploid tumors. There appears to be a weak relationship between advanced pathologic stage and DNA anenploid tumors, although there is no clear and consistent relationship between tumor ploidy and histology. Therefore, all patients with colorectal tumors should undergo DNA ploidy analysis. Patients with DNA nondiploid tumors should be treated for biologically more aggressive disease independent of other prognostic variables. Ploidy status should be employed as a variable by which to randomize patients to both primary treatment schemes and adjuvant therapies in clinical trials. [Key words: Flow cytometry; DNA analysis; Colorectal cancer]
pathologic stage and histologic grade are important prognostic indicators in colorectal carcinoma. Equally clear, however, is that t u m o r b e h a v i o r varies within each prognostic g r o u p 7 and that further selection of higher-risk patients b a s e d on the biologic characteristics of t u m o r cells w o u l d b e helpful in selecting treatment. Flow cytometric m e a s u r e m e n t of DNA content can p r o v i d e valuable information a b o u t the biologic b e h a v i o r of neoplastic cells. 8 The usefulness of flow cytometric DNA analysis in different h u m a n solid t u m o r s was extensively r e v i e w e d by Barlogie et aL 9 and was m o r e r e c e n t l y r e v i e w e d b y Seckinger et aL 1~ Both have d o c u m e n t e d its value in t u m o r prognostication. Initial reports b y Wolley et aL 1~ s u g g e s t e d a striking correlation b e t w e e n tum o r cell DNA content and survival i n d e p e n d e n t of other p r o g n o s t i c variables. T h e s e reports have stimulated multiple attempts to define the prognostic significance of DNA analysis in colorectal carcinomas using flow c y t o m e t r y ~1-34 as well as m o r e conventional m e t h o d s ? 4-36 While recent reviews have s u g g e s t e d a relationship b e t w e e n abn o r m a l DNA c o n t e n t and m o r e aggressive t u m o r behavior, ~~ the results in these studies have b e e n inconsistent and have left the clinical significance of such DNA analysis questionable. 29 This article will r e v i e w the use of flow c y t o m e t r y in DNA analysis of colorectal c a r c i n o m a and evaluate the role of DNA content as a p r o g n o s t i c indicator.
Dean PA, Vernava AM III. Flow cytometric analysis of DNA content in colorectal carcinoma. Dis Colon Rectum 1992;35:95-102. uring the past decade, analysis of DNA content by flow cytometric t e c h n i q u e s has e m e r g e d as the m o s t recent tool in a c o n t i n u i n g effort to i m p r o v e our ability to define p r o g n o s t i c variables for colorectal carcinoma. Identification of patients at high risk for local r e c u r r e n c e and distant metastasis has b e c o m e m o r e i m p o r t a n t as increasing e v i d e n c e suggests that adjuvant radiation therapy 1 and c h e m o t h e r a p y 2'3 can b e valuable in select patient groups. Since the initial staging of rectal cancer by L o c k h a r t - M u m m e r y 4 and its subs e q u e n t revision and refining b y D u k e s 5 and Gabriel et al.,6 it has b e c o m e clear that clinical and
D
DNA ANALYSIS BY FLOW
CYTOMETRY
The e s s e n c e of flow c y t o m e t r y is the measurem e n t of DNA in a large n u m b e r of cells to determ i n e the distribution of DNA within the study cell population. In eukaryotic organisms, the largest p r o p o r t i o n of any cell p o p u l a t i o n has a diploid
Address reprint requests to Dr. Vernava: Department of Surgery, St. Louis University Medical Center, 3635 Vista Avenue at Grand Boulevard, P.O. Box 15250, St. Louis, Missouri 63110-0250. 95
DEAN AND VERNAVA
96
(2C) DNA content and is in the cycling (G1) or resting (Go) phase of the cell cycle (Fig. 1). During the synthetic (S) phase, DNA content is doubled tO 4C (O2 phase) and remains that way until mitosis restores the normal chromosome number of 2C. Flow cytometry allows the analysis of cells in suspension with a fluorescent stain for DNA. It permits rapid passage of these suspended ceils in single file past an excitation light source. As each cell passes, the emitted fluorescent light is detected; the amount of fluorescence is proportional to the DNA content of the stained cell. 3s In this manner, large numbers of cells can have their DNA content rapidly measured and then analyzed by computer. Utilizing the fluorescence information, the computer constructs a histogram, which represents the relative distribution of DNA in the sample population (Figs. 1 and 2). Tumor cell populations containing abnormal amounts of DNA represent clonal chromosomal abnormalities and are expressed as being DNA aneuploid, s These DNA aneuploid cell populations are expressed as an abnormal peak on the histogram and have an increased relative DNA content. They are classified by comparing the aneuploid cell population with the normal diploid cell population; this ratio is expressed as the DNA index. Additional information can be obtained by calcu-
A
G /G1
G2~
8
G~
Cell Cycle
J Relative DNA content (Fluorescence Intensity)
B
Relative DNA content (Fluorescence Intensity) Figure 1. DNA histograms with cell cycle histogram. A. DNA diploid histogram. B. DNA nondiploid (aneuploid) histogram.
Dis Colon Rectum, January 1992
Flow Cytometry Methodology Fluorescent...
stain\~ Cells ~\Q\ u~
Fluorescence detector
--
f Oomp,,ter 1 [ .,stogram 1 2000
g o 1000
O 0
k...,"~ 2C 4C Relative DNA Content
DNA Diploid Histogram Figure 2. Method of flow cytometric DNA analysis. lating the proliferative index, which is the fraction of cells in the S phase or the G2 + M phase as determined from the DNA histogram. Rapid analysis of the DNA content of a small tumor fragment can thus be made. Using flow cytometry, the synthetic phase fraction and the proliferative index can also be determined. A theoretic limitation to the interpretation of DNA ploidy studies in colorectal carcinoma is the firmly established existence of tumor heterogeneity and its implications for sampling error in flow cytometry.21'23'36'39-41 The occurrence of DNA diploid and aneuploid areas within the same tumor might lead to sampling errors and result in the inappropriate labeling of the tumor. Quirke e t aL 23 found that heterogeneous tumors containing both DNA diploid and aneuploid areas in separate biopsies behaved similarly to uniform DNA aneuploid tumors; they suggested that a tumor could be classified as DNA aneuploid if any samples were found to contain such areas. In a detailed study of the distribution of DNA content in colorectal carcinoma, Scott e t aL 27 demonstrated that most tumors were homogeneous with respect to DNA ploidy determined by flow cytometry if adequate sample
Vol. 35, No. 1
FLOW CYTOMETRY IN COLORECTAL CANCER
sections were used. While tumor heterogeneity remains a concern, careful and adequate tissue sampling would minimize the probability of incorrectly labeling a DNA aneuploid tumor as being DNA diploid. Many studies evaluating flow cytometry with DNA content analysis as a prognostic indicator in colorectal carcinoma have been performed in a prospective fashion using fresh biopsy or operative specimens. 1L12':4':5'19'2~ The development of a technique for analysis of paraffin-embedded (archived) tissue samples 42 has allowed for retrospective analysis of DNA ploidy status in colorectal cancers from patients with known longterm follow-up. Some studies have suggested that the detection of small DNA aneuploid populations is lower in paraffin-embedded tissue samples. *~ However, Emdin e t a L = compared fresh and paraffin-embedded specimens from the same tumor samples and found nearly uniform agreement in DNA content. This finding confirms the use of archived tissue blocks to retrospectively evaluate DNA content using flow cytometry. F L O W C Y T O M E T R I C DNA ANALYSIS AND T U M O R P R O G N O S I S Studies utilizing flow cytometric analysis of DNA content in human colorectal adenocarcinoma fall into two groups. The first group consists of prospective studies done using fresh or frozen tumor specimens and generally involves a more complete multiparametric analysis of patient outcome. This group of studies is limited by slow patient accession rates, resulting in smaller study groups. In contrast, retrospective studies using archived tissue blocks generally consist of a much larger number of patients with longer follow-up periods but often lack important patient demographic material or tumor characteristics. The demonstrated correlation between fresh and archived tissues, however, allows these two study types to be simultaneously reviewed. = As is common in colon cancer, 43 staging terminology is often poorly defined and confusing, but, when possible, pathology was translated to the appropriate Dukes' stage (including Stage D, representing metastatic disease). Most studies included both colon and rectal carcinomas, although in some cases this was not specified. In addition, some prospective studies did not clearly state the duration of follow-up.
97
RELATIONSHIP O F DNA C O N T E N T T O P A T H O L O G I C STAGE There is considerable variation in the reported incidence of DNA aneuploid tumors in the 24 studies evaluating colorectal carcinoma DNA ploidy status (Table 1). The observed incidence of DNA aneuploidy ranged from a low of 35 percent 15 to a high of 76 percentY The average incidence of DNA aneuploidy in this group of studies using flow cytometry was 58 percent; this is identical to the incidence of DNA aneuploidy in colorectal cancers measured using image cytometry35 and is similar to the incidence of DNA aneuploidy obtained from cytogenetic analysis.36 In addition, the incidence of DNA aneuploidy appears to be remarkably similar between colon 21 and rectal 23'2
Lee E. Smith, M.D., Editor
Flow Cytometric Analysis of DNA Content in Colorectal Carcinoma Phillip A. Dean, M.D., Anthony M. Vernava III, M.D. From the D e p a r t m e n t of Surgery, Section o f Colon a n d Rectal Surgery, St. Louis University Medical Center, St. Louis, Missouri Over the past decade, flow cytometric DNA analysis has been employed by a number of investigators in an attempt to further define patient prognosis beyond classic pathologic staging. The results of these studies taken independently have been confusing; however, their cumulative effect suggests that flow cytometry is a useful prognostic indicator and can be used to further delineate prognosis within a specific pathologic stage. DNA nondiploid tumors are more likely to recur than diploid tumors, and patients with DNA nondiploid tumors have a poorer fiveyear survival than patients with DNA diploid tumors. There appears to be a weak relationship between advanced pathologic stage and DNA anenploid tumors, although there is no clear and consistent relationship between tumor ploidy and histology. Therefore, all patients with colorectal tumors should undergo DNA ploidy analysis. Patients with DNA nondiploid tumors should be treated for biologically more aggressive disease independent of other prognostic variables. Ploidy status should be employed as a variable by which to randomize patients to both primary treatment schemes and adjuvant therapies in clinical trials. [Key words: Flow cytometry; DNA analysis; Colorectal cancer]
pathologic stage and histologic grade are important prognostic indicators in colorectal carcinoma. Equally clear, however, is that t u m o r b e h a v i o r varies within each prognostic g r o u p 7 and that further selection of higher-risk patients b a s e d on the biologic characteristics of t u m o r cells w o u l d b e helpful in selecting treatment. Flow cytometric m e a s u r e m e n t of DNA content can p r o v i d e valuable information a b o u t the biologic b e h a v i o r of neoplastic cells. 8 The usefulness of flow cytometric DNA analysis in different h u m a n solid t u m o r s was extensively r e v i e w e d by Barlogie et aL 9 and was m o r e r e c e n t l y r e v i e w e d b y Seckinger et aL 1~ Both have d o c u m e n t e d its value in t u m o r prognostication. Initial reports b y Wolley et aL 1~ s u g g e s t e d a striking correlation b e t w e e n tum o r cell DNA content and survival i n d e p e n d e n t of other p r o g n o s t i c variables. T h e s e reports have stimulated multiple attempts to define the prognostic significance of DNA analysis in colorectal carcinomas using flow c y t o m e t r y ~1-34 as well as m o r e conventional m e t h o d s ? 4-36 While recent reviews have s u g g e s t e d a relationship b e t w e e n abn o r m a l DNA c o n t e n t and m o r e aggressive t u m o r behavior, ~~ the results in these studies have b e e n inconsistent and have left the clinical significance of such DNA analysis questionable. 29 This article will r e v i e w the use of flow c y t o m e t r y in DNA analysis of colorectal c a r c i n o m a and evaluate the role of DNA content as a p r o g n o s t i c indicator.
Dean PA, Vernava AM III. Flow cytometric analysis of DNA content in colorectal carcinoma. Dis Colon Rectum 1992;35:95-102. uring the past decade, analysis of DNA content by flow cytometric t e c h n i q u e s has e m e r g e d as the m o s t recent tool in a c o n t i n u i n g effort to i m p r o v e our ability to define p r o g n o s t i c variables for colorectal carcinoma. Identification of patients at high risk for local r e c u r r e n c e and distant metastasis has b e c o m e m o r e i m p o r t a n t as increasing e v i d e n c e suggests that adjuvant radiation therapy 1 and c h e m o t h e r a p y 2'3 can b e valuable in select patient groups. Since the initial staging of rectal cancer by L o c k h a r t - M u m m e r y 4 and its subs e q u e n t revision and refining b y D u k e s 5 and Gabriel et al.,6 it has b e c o m e clear that clinical and
D
DNA ANALYSIS BY FLOW
CYTOMETRY
The e s s e n c e of flow c y t o m e t r y is the measurem e n t of DNA in a large n u m b e r of cells to determ i n e the distribution of DNA within the study cell population. In eukaryotic organisms, the largest p r o p o r t i o n of any cell p o p u l a t i o n has a diploid
Address reprint requests to Dr. Vernava: Department of Surgery, St. Louis University Medical Center, 3635 Vista Avenue at Grand Boulevard, P.O. Box 15250, St. Louis, Missouri 63110-0250. 95
DEAN AND VERNAVA
96
(2C) DNA content and is in the cycling (G1) or resting (Go) phase of the cell cycle (Fig. 1). During the synthetic (S) phase, DNA content is doubled tO 4C (O2 phase) and remains that way until mitosis restores the normal chromosome number of 2C. Flow cytometry allows the analysis of cells in suspension with a fluorescent stain for DNA. It permits rapid passage of these suspended ceils in single file past an excitation light source. As each cell passes, the emitted fluorescent light is detected; the amount of fluorescence is proportional to the DNA content of the stained cell. 3s In this manner, large numbers of cells can have their DNA content rapidly measured and then analyzed by computer. Utilizing the fluorescence information, the computer constructs a histogram, which represents the relative distribution of DNA in the sample population (Figs. 1 and 2). Tumor cell populations containing abnormal amounts of DNA represent clonal chromosomal abnormalities and are expressed as being DNA aneuploid, s These DNA aneuploid cell populations are expressed as an abnormal peak on the histogram and have an increased relative DNA content. They are classified by comparing the aneuploid cell population with the normal diploid cell population; this ratio is expressed as the DNA index. Additional information can be obtained by calcu-
A
G /G1
G2~
8
G~
Cell Cycle
J Relative DNA content (Fluorescence Intensity)
B
Relative DNA content (Fluorescence Intensity) Figure 1. DNA histograms with cell cycle histogram. A. DNA diploid histogram. B. DNA nondiploid (aneuploid) histogram.
Dis Colon Rectum, January 1992
Flow Cytometry Methodology Fluorescent...
stain\~ Cells ~\Q\ u~
Fluorescence detector
--
f Oomp,,ter 1 [ .,stogram 1 2000
g o 1000
O 0
k...,"~ 2C 4C Relative DNA Content
DNA Diploid Histogram Figure 2. Method of flow cytometric DNA analysis. lating the proliferative index, which is the fraction of cells in the S phase or the G2 + M phase as determined from the DNA histogram. Rapid analysis of the DNA content of a small tumor fragment can thus be made. Using flow cytometry, the synthetic phase fraction and the proliferative index can also be determined. A theoretic limitation to the interpretation of DNA ploidy studies in colorectal carcinoma is the firmly established existence of tumor heterogeneity and its implications for sampling error in flow cytometry.21'23'36'39-41 The occurrence of DNA diploid and aneuploid areas within the same tumor might lead to sampling errors and result in the inappropriate labeling of the tumor. Quirke e t aL 23 found that heterogeneous tumors containing both DNA diploid and aneuploid areas in separate biopsies behaved similarly to uniform DNA aneuploid tumors; they suggested that a tumor could be classified as DNA aneuploid if any samples were found to contain such areas. In a detailed study of the distribution of DNA content in colorectal carcinoma, Scott e t aL 27 demonstrated that most tumors were homogeneous with respect to DNA ploidy determined by flow cytometry if adequate sample
Vol. 35, No. 1
FLOW CYTOMETRY IN COLORECTAL CANCER
sections were used. While tumor heterogeneity remains a concern, careful and adequate tissue sampling would minimize the probability of incorrectly labeling a DNA aneuploid tumor as being DNA diploid. Many studies evaluating flow cytometry with DNA content analysis as a prognostic indicator in colorectal carcinoma have been performed in a prospective fashion using fresh biopsy or operative specimens. 1L12':4':5'19'2~ The development of a technique for analysis of paraffin-embedded (archived) tissue samples 42 has allowed for retrospective analysis of DNA ploidy status in colorectal cancers from patients with known longterm follow-up. Some studies have suggested that the detection of small DNA aneuploid populations is lower in paraffin-embedded tissue samples. *~ However, Emdin e t a L = compared fresh and paraffin-embedded specimens from the same tumor samples and found nearly uniform agreement in DNA content. This finding confirms the use of archived tissue blocks to retrospectively evaluate DNA content using flow cytometry. F L O W C Y T O M E T R I C DNA ANALYSIS AND T U M O R P R O G N O S I S Studies utilizing flow cytometric analysis of DNA content in human colorectal adenocarcinoma fall into two groups. The first group consists of prospective studies done using fresh or frozen tumor specimens and generally involves a more complete multiparametric analysis of patient outcome. This group of studies is limited by slow patient accession rates, resulting in smaller study groups. In contrast, retrospective studies using archived tissue blocks generally consist of a much larger number of patients with longer follow-up periods but often lack important patient demographic material or tumor characteristics. The demonstrated correlation between fresh and archived tissues, however, allows these two study types to be simultaneously reviewed. = As is common in colon cancer, 43 staging terminology is often poorly defined and confusing, but, when possible, pathology was translated to the appropriate Dukes' stage (including Stage D, representing metastatic disease). Most studies included both colon and rectal carcinomas, although in some cases this was not specified. In addition, some prospective studies did not clearly state the duration of follow-up.
97
RELATIONSHIP O F DNA C O N T E N T T O P A T H O L O G I C STAGE There is considerable variation in the reported incidence of DNA aneuploid tumors in the 24 studies evaluating colorectal carcinoma DNA ploidy status (Table 1). The observed incidence of DNA aneuploidy ranged from a low of 35 percent 15 to a high of 76 percentY The average incidence of DNA aneuploidy in this group of studies using flow cytometry was 58 percent; this is identical to the incidence of DNA aneuploidy in colorectal cancers measured using image cytometry35 and is similar to the incidence of DNA aneuploidy obtained from cytogenetic analysis.36 In addition, the incidence of DNA aneuploidy appears to be remarkably similar between colon 21 and rectal 23'2
