- Email: [email protected]
The effect of tissue fixation and processing on breast cancer size
Human Pathology (2005) 36, 756 – 760
www.elsevier.com/locate/humpath
The effect of tissue fixation and processing on breast cancer sizeB,BB Bobbi Pritt MDa, Joseph J. Tessitore BHSa, Donald L. Weaver MDa,b, Hagen Blaszyk MDa,* a
Department of Pathology, University of Vermont College of Medicine, Burlington, VT 05405, USA Vermont Cancer Center, University of Vermont College of Medicine, Burlington, VT 05405, USA
b
Accepted 25 April 2005
Keywords: Breast; Cancer; Fixation; Measurement; Formalin
Summary Precise measurement of an invasive breast cancer is crucial for pathological staging and subsequent patient management. Formalin fixation and histological processing may change tissue size, but there is no agreement on which state of the specimen, fresh or fixed, should be used for final tumor measurement. To determine the influence of fixation and processing on breast tumor size, a specific 1-dimensional measurement from 50 invasive breast tumors was recorded in fresh, fixed, and processed/ mounted states. Tumors varied in maximum measured dimension from 4 to 20 mm and contained 10% to 90% estimated fibrous tissue (mean, 52.8%). In 96% of cases, there was no difference in measured size between fresh and fixed states. After final processing and mounting, a decrease in size from initial fresh measurement was noted in 40% of cases (mean difference, 2.4 mm; maximum difference, 7 mm). In 9 cases (18%), the measured size increased by a maximum of 3 mm (mean, 1.7 mm) after processing/ mounting. Twenty-one cases (42%) showed no change in measurement during the entire fixation and processing protocol. Increases in measured size were attributed largely to tissue expansion during histological sectioning/mounting. One can arguably measure the size of an invasive breast cancer from either the fresh or fixed state without affecting accuracy, but caution should be exercised in relying solely on the microscopic measurements. D 2005 Elsevier Inc. All rights reserved.
1. Introduction B
The views expressed in this article are solely those of the authors and do not necessarily represent the official views of the National Cancer Institute or the federal government. The data were presented in part at the United States and Canadian Academy of Pathology 93rd Annual Meeting, March 8, 2004, Vancouver, British Columbia. BB This work was supported by grant P30-CA22435 from the National Cancer Institute. T Corresponding author. E-mail address: [email protected] (H. Blaszyk). 0046-8177/$ – see front matter D 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.humpath.2005.04.018
Precise measurement of a primary invasive breast cancer is crucial for accurate staging and subsequent patient management. In some instances, the difference of only 1 mm in maximum size could change the final staging category in which a patient is placed and alter patient management. For example, changing T classification from pT1a to pT1b (0.5 versus 0.6 cm maximum tumor size) could make axillary evaluation probable, and changing T classification from pT1b to pT1c (1.0 versus 1.1 cm
The effect of tissue fixation and processing on breast cancer size maximum tumor size) could make the patient a candidate for adjuvant therapy. The American Joint Committee on Cancer staging manual [1] sets guidelines for assigning pathological T classification in breast cancer based on maximum measurement of the invasive tumor. However, the state of the specimen (fresh, formalin-fixed, or processed) from which the measurements are to be obtained is not specified. The 3-dimensional measurements are most easily obtained from the fresh gross specimen [2]. However, the opportunity exists to measure maximum tumor dimensions at several other stages during the normal course of processing in the pathology laboratory, such as after overnight fixation in formalin and after processing and mounting on a glass slide. Indeed, many pathology laboratories serving smaller health-care institutions obtain gross tumor measurements after formalin fixation because of geographic or temporal constraints in the logistics of tissue processing. Furthermore, situations may arise where microscopic dimensions are used for staging purposes, such as when the grossly measured tumor mass has only a minor invasive component or the carcinoma extends microscopically beyond the grossly assessed mass lesion. In addition, the College of American Pathologists recommends microscopic reassessment of tumor size for all tumors smaller than 1.0 cm [3]. Given these inconsistencies in approach to measurement, it is important to acknowledge that fixation and processing has the potential to change overall tissue size and shape [4-10]. Many authors have demonstrated an overall shrinkage during this process in a variety of tissues. For example, Boonstra et al [4] demonstrated a 15% decrease in uterine cervical measurements after standard fixation and processing. Similarly, Dobrin [5] demonstrated 19% to 25% shrinkage of arterial cross sections from the gross specimen to the mounted glass slides. Weight changes of up to 34% have also been observed in various organs after formalin fixation [11]. Considering the variability of shrinkage demonstrated by these studies, it is possible that the composition and type of the tissue (eg, amount of fat or fibrous tissue) may impact the overall amount of shrinkage. Therefore, it follows that studies addressing change in size should be organ-specific and account for tissue constitution. To the best of our knowledge, there are no studies examining the effect of fixation and processing on breast cancer size measurement.
2. Materials and methods To determine the potential influence of fixation and processing on breast tumor size, specimens from 2 male and 48 female patients aged 32 to 81 years (mean, 60 years) with invasive breast cancer were identified during routine tissue processing in the pathology department of our institution. In each study specimen, the largest
757
1-dimensional measurement in a single, controlled plane of sectioning for each tumor was obtained in 3 states: fresh, after overnight fixation in 10% formalin, and after routine processing and mounting on a glass slide. The plane measured in the fresh and fixed state was embedded in paraffin in a manner to preserve the same measurement plane on histological sections. For inclusion in this study, the tumor had to form a welldefined mass, consist microscopically of at least 75% invasive carcinoma, and be small enough so that a section of the tumor containing the largest dimension could be mounted on a single glass slide. The percentage of invasive cancer was performed by visual estimation from the final mounted specimen. All gross and microscopic measurements were taken in millimeters using standard methods (handheld ruler or microscope ocular) to simulate actual pathology practice. Only the mass-forming lesion was measured microscopically because this corresponded to the gross lesion measured. The fresh and fixed measurements were supervised by the same individual to ensure that the correct section of the gross specimen was measured. An independent pathologist, without prior knowledge of the fresh and fixed measurements, obtained the microscopic measurement on the same tissue block. The type of tumor (eg, ductal, lobular) was recorded for each case, and the percentage of change in tumor size was calculated for each specimen. To determine the potential influence of tumor composition on size change during the histological process, the ratio of fibrosis to glands was calculated by visual estimation from the final mounted specimen.
3. Results The study tumor characteristics, including changes in size, are presented in Table 1. The histological tumor types consisted of 36 ductal, 3 lobular, 6 mixed ductal/lobular, and 5 others. Gross tumor size ranged from 4 to 20 mm (mean, 12.1 mm; median, 12 mm) and tumors contained 10% to 90% fibrous tissue (mean, 52.8%; median, 55%). All tumors contained at least 75% invasive tumor, and 94% contained at least 90% invasive tumor. All tumors infiltrated with a relatively compact invasion front so that under low microscopic power (2 objective; original magnification 20), a virtual circle could be drawn around the tumor, serving as the boundary for microscopic measurement. In 48 cases (96%), there was no difference in measured size between fresh and fixed states. In the remaining 2 cases, the measured size decreased by 1 and 2 mm after overnight fixation in formalin. These tumors contained 30% and 40% fibrous tissue, respectively. After final processing and mounting, a decrease in size from initial fresh measurement was noted in 40% of cases (20 cases; mean difference, 2.4 mm; maximum difference, 7 mm). Twenty-one cases
758 Table 1 Patient 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50
B. Pritt et al. Tumor characteristics and change in measurement from fresh to final processed/mounted state Tissue state Fresh (mm)
Fixed (mm)
Processed (mm)
Size change from fresh to final (%)
Estimated % fibrosis
Histological type
15 14 6 7 19 9 8 11 15 9 15 17 15 5 7 11 7 18 9 11 11 18 20 5 14 19 7 20 14 17 8 13 19 12 13 5 4 19 14 9 4 20 10 12 8 13 11 16 13 10
14 12 6 7 19 9 8 11 15 9 15 17 15 5 7 11 7 18 9 11 11 18 20 5 14 19 7 20 14 17 8 13 19 12 13 5 4 19 14 9 4 20 10 12 8 13 11 16 13 10
14 12 5 7 17 10 8 11 17 8 14 14 15 7 7 13 10 18 7 11 11 16 19 5 14 16 7 14 14 13 7 14 18 10 11 5 4 12 14 11 4 16 10 13 8 13 11 14 14 10
7 14 17 0 11 + 11 0 0 + 13 11 7 18 0 + 40 0 + 18 + 43 0 22 0 0 11 5 0 0 16 0 30 0 24 13 +8 5 17 15 0 0 37 0 + 22 0 20 0 +8 0 0 0 13 +8 0
30 40 70 60 50 70 80 20 20 60 40 50 40 70 40 80 60 40 20 80 40 70 20 80 60 50 70 30 50 70 70 50 50 40 20 30 80 10 90 90 80 10 60 70 30 60 60 40 60 80
D D D D D D/L D M D D AC D/L D D D D D D D/L D D D D D D D D D D D D D D D D D D/M D D L D Med D/L D D/L L D D/L D D
Abbreviations. D, ductal; L, lobular; M, mucinous; Med, medullary; AC, adenoid cystic.
(42%) showed no change in measurement from fresh to final processed state, whereas in 9 cases (18%), the measured size increased on average by 1.7 mm (range, 1-3 mm) after processing/mounting. Most cases were pure
classic ductal-type (76%), whereas 8 cases (16%) had lobular histology or features. The remaining 4 cases had adenoid cystic, medullary, mucinous, and mixed ductal/ mucinous histology.
The effect of tissue fixation and processing on breast cancer size
4. Discussion Despite the potential for size discrepancy among fresh, fixed, and processed states, there is no bofficialQ agreement or recommendation on which state should be used for pathological staging of breast cancers. To the best of our knowledge, this study specifically addresses these issues for the first time to determine which state(s) most accurately reflects measured tumor size. Our results indicate that most carcinomas do not change in size after overnight fixation in 10% formalin. However, substantial changes in tumor size (increases and decreases) were noted after processing of the paraffin block. In 40% of our cases, shrinkage occurred with processing, and a mean decrease of 2.4 mm was noted. Although this difference would not always change the patient’s stage, the pathologist should keep this in mind when considering microscopic tumor measurements. Notably, 9 cases increased in measured dimension after paraffin embedding, mounting, and staining. Because only the mass-forming lesion was measured microscopically, size increase cannot be attributed to microscopic extension of tumor into the tissue surrounding the mass. Although counterintuitive, the most likely explanation for these changes in tumor size is that the tumor sections expanded during block preparation, cutting, or mounting. We cannot entirely exclude the possibility of minor size variations because the gross and microscopic measurements were obtained by different methodology (ruler versus ocular measurement). However, one would expect this variation to affect all specimens of this series to the same extent. Similarly, minimal deviation from the gross measurement could occur because of change in overall tumor size as the microtome sections through the tissue. However, for this particular study, efforts were made to mount the first available tissue section from the paraffin block; thus, the section on the resulting microscopic slide corresponded most closely to what was measured grossly. It is important to recognize that potential causes of both artifactual tissue shrinkage and expansion exist throughout the process of block manufacture and slide preparation. These causes may explain the above findings of tumor size increases and decreases after processing. To begin with, alterations may occur during fixation and dehydration of tissue in the automated processors. Next, during embedding, the temperature of the paraffin is important for regulating the size of the tissue. Paraffin that is too warm will shrink more than expected during cooling, which may compress the tissue and cause it to expand less predictably when sectioned on the microtome. Furthermore, during microtomy, incorrect knife angle or loose knives can cause tissue chatter or compression, which affects size during mounting. Poorly fixed and processed tissue can also behave unpredictably during microtomy and in the mounting water bath. Finally, water bath temperature affects tissue size and must be calibrated to the paraffin
759
used. A too-warm water bath may melt the paraffin, diminishing support for the tissue leading to expansion, whereas a too-cool water bath may not allow the tissue to evenly spread after the natural compression of microtomy. Simple quality checks should be incorporated into the histological process. In our laboratory, the histotechnologist visually compares the size of the mounted section to the size of the tissue within the paraffin block. Major discrepancies in size would prompt preparation of new sections before review by a pathologist. This policy may have limited the range of observed changes on this study. However, despite this safeguard, changes in size were observed in most cases (58%). The percentage of desmoplastic fibrous tissue associated with the invasive tumor was determined for each case by visual estimation at the microscope and did not appear to correlate with the degree of change in measured dimension (Table 1). At the onset of our study, we hypothesized that the fibrosis-to-gland ratio may influence the degree of change in tumor size throughout processing; therefore, we chose a method of measurement that could easily be incorporated into daily pathology practice. Lack of correlation of percentage of fibrosis and degree of overall size change suggests that the observed size changes do not depend on overall tissue composition, but reflects more on general processing artifact in the histology laboratory. Histopathological staging of breast cancers is based on the size of the invasive tumor component, but many invasive carcinomas are associated with various amounts of in situ carcinoma. It is critical to adjust maximum tumor size in such cases after microscopic review because extensive in situ carcinoma can be forming a mass that greatly exceeds the size of the invasive tumor. The current case series included only tumors with a dominant (N75%) invasive tumor component. As per our departmental reporting policy, the percentage of in situ carcinoma is determined for each invasive breast cancer by visual approximation from the glass slide and recorded in the final report. This allowed for easy extrapolation of the percentage of invasive cancer that we confirmed for each case during review for our study, excluding less than 5% of total cases that had an invasive tumor component of below 75%. Further studies are needed to determine the influence, if any, of an in situ component on tissue shrinkage or expansion. The possibility of some degree of selection bias cannot entirely be excluded because this study was not designed to sample nonselected consecutive cases. The accuracy, repeatability, and reliability of our measurements have not been specifically determined as part of the study but rather reflect daily practice. However, we excluded interobserver variability in the gross measurements by assigning only one investigator to this part of the study. The 50 cases of this study represent approximately 20% of resection specimens for invasive carcinoma (excisional biopsies and partial and total mastectomies with or without axillary dissection) processed by our pathology department in a
760 1-year period (2003-2004). The goal of this study was to determine potential tumor size changes during histological processing that would have a significant impact on subsequent tumor staging. Thus, only lesions 2 cm in size or smaller and where a section of the entire lesion fits onto 1 glass slide were included. Specimens were also excluded when the accuracy of the histological diagnosis could potentially have been jeopardized by inclusion in the current study. This concerned a significant number of cases that were smaller than 5 mm and/or where existing clinical study protocols required extensive tumor tissue sampling for other purposes. Finally, it is possible that tumors of varying histological subtype (eg, ductal, lobular) will behave differently during histological processing. This is beyond the scope of our study and merits further investigation. Based on our data, one can arguably measure the size of an invasive breast cancer from either the fresh or fixed state (gross measurements) without affecting accuracy. We therefore recommend that staging measurements be taken from 1 of these 2 states. Furthermore, we advise that the pathologist verify the gross measurements microscopically when possible. In most cases, it is our opinion that the gross measurement will provide the most accurate staging information. However, in instances where significant associated in situ disease is present or when the invasive tumor extends microscopically beyond the grossly measured mass, then the microscopic measurements of the invasive component are more accurate and should be used for staging. We strongly advise against decreasing or increasing tumor staging based solely on discrepant microscopic measurements because processing artifact may lead to
B. Pritt et al. significant tissue shrinkage or expansion after processing and embedding.
References [1] Breast. In: Greene FL, Page DL, Fleming ID, Fritz AG, Balch CH, Haller DG, et al, editors. American Joint Committee on Cancer (AJCC) cancer staging manual. 6th ed. New York7 Springer; 2002. p. 223 - 40. [2] Rosen PP. Rosen’s breast pathology. Philadelphia7 Lippincott Williams & Wilkins; 2001. [3] Fitzgibbons PL, Page DL, Weaver D, Thor AD, Allred DC, Clark GM, et al. Prognostic factors in breast cancer: College of American Pathologists Consensus Statement 1999. Arch Pathol Lab Med 2000;124:966 - 78. [4] Boonstra H, Oosterhuis JW, Oosterhuis AM, Fleuren GJ. Cervical tissue shrinkage by formaldehyde fixation, paraffin wax embedding, section cutting and mounting. Virchows Arch A Pathol Anat Histopathol 1983;402:195 - 201. [5] Dobrin PB. Effect of histologic preparation on the cross-sectional area of arterial rings. J Surg Res 1996;61:413 - 5. [6] Friedman NB. On formalin fixation. Hum Path 1992;23:1440 - 1. [7] Goldstein NS, Soman A, Sacksner J. Disparate surgical margin lengths of colorectal resection specimens between in vivo and in vitro measurements. The effects of surgical resection and formalin fixation on organ shrinkage. Am J Clin Pathol 1999;111:349 - 51. [8] Lane J, Ralis ZA. Changes in dimensions of large cancellous bone specimens during histological preparation as measured on slabs from human femoral heads. Calcif Tissue Int 1983;35:1 - 4. [9] Schned AR, Wheeler KJ, Hodorowski CA, Heaney JA, Ernstoff MS, Amdur RJ, et al. Tissue-shrinkage correction factor in the calculation of prostate cancer volume. Am J Surg Pathol 1996;20:1501 - 6. [10] Siu KF, Cheung HC, Wong J. Shrinkage of the esophagus after resection for carcinoma. Ann Surg 1986;203:173 - 6. [11] Schremmer CN. Gewichts7nderungen verschiedener Gewebe nach Formalinfixierung. Frankf Z Pathol 1967;77:299 - 304.
www.elsevier.com/locate/humpath
The effect of tissue fixation and processing on breast cancer sizeB,BB Bobbi Pritt MDa, Joseph J. Tessitore BHSa, Donald L. Weaver MDa,b, Hagen Blaszyk MDa,* a
Department of Pathology, University of Vermont College of Medicine, Burlington, VT 05405, USA Vermont Cancer Center, University of Vermont College of Medicine, Burlington, VT 05405, USA
b
Accepted 25 April 2005
Keywords: Breast; Cancer; Fixation; Measurement; Formalin
Summary Precise measurement of an invasive breast cancer is crucial for pathological staging and subsequent patient management. Formalin fixation and histological processing may change tissue size, but there is no agreement on which state of the specimen, fresh or fixed, should be used for final tumor measurement. To determine the influence of fixation and processing on breast tumor size, a specific 1-dimensional measurement from 50 invasive breast tumors was recorded in fresh, fixed, and processed/ mounted states. Tumors varied in maximum measured dimension from 4 to 20 mm and contained 10% to 90% estimated fibrous tissue (mean, 52.8%). In 96% of cases, there was no difference in measured size between fresh and fixed states. After final processing and mounting, a decrease in size from initial fresh measurement was noted in 40% of cases (mean difference, 2.4 mm; maximum difference, 7 mm). In 9 cases (18%), the measured size increased by a maximum of 3 mm (mean, 1.7 mm) after processing/ mounting. Twenty-one cases (42%) showed no change in measurement during the entire fixation and processing protocol. Increases in measured size were attributed largely to tissue expansion during histological sectioning/mounting. One can arguably measure the size of an invasive breast cancer from either the fresh or fixed state without affecting accuracy, but caution should be exercised in relying solely on the microscopic measurements. D 2005 Elsevier Inc. All rights reserved.
1. Introduction B
The views expressed in this article are solely those of the authors and do not necessarily represent the official views of the National Cancer Institute or the federal government. The data were presented in part at the United States and Canadian Academy of Pathology 93rd Annual Meeting, March 8, 2004, Vancouver, British Columbia. BB This work was supported by grant P30-CA22435 from the National Cancer Institute. T Corresponding author. E-mail address: [email protected] (H. Blaszyk). 0046-8177/$ – see front matter D 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.humpath.2005.04.018
Precise measurement of a primary invasive breast cancer is crucial for accurate staging and subsequent patient management. In some instances, the difference of only 1 mm in maximum size could change the final staging category in which a patient is placed and alter patient management. For example, changing T classification from pT1a to pT1b (0.5 versus 0.6 cm maximum tumor size) could make axillary evaluation probable, and changing T classification from pT1b to pT1c (1.0 versus 1.1 cm
The effect of tissue fixation and processing on breast cancer size maximum tumor size) could make the patient a candidate for adjuvant therapy. The American Joint Committee on Cancer staging manual [1] sets guidelines for assigning pathological T classification in breast cancer based on maximum measurement of the invasive tumor. However, the state of the specimen (fresh, formalin-fixed, or processed) from which the measurements are to be obtained is not specified. The 3-dimensional measurements are most easily obtained from the fresh gross specimen [2]. However, the opportunity exists to measure maximum tumor dimensions at several other stages during the normal course of processing in the pathology laboratory, such as after overnight fixation in formalin and after processing and mounting on a glass slide. Indeed, many pathology laboratories serving smaller health-care institutions obtain gross tumor measurements after formalin fixation because of geographic or temporal constraints in the logistics of tissue processing. Furthermore, situations may arise where microscopic dimensions are used for staging purposes, such as when the grossly measured tumor mass has only a minor invasive component or the carcinoma extends microscopically beyond the grossly assessed mass lesion. In addition, the College of American Pathologists recommends microscopic reassessment of tumor size for all tumors smaller than 1.0 cm [3]. Given these inconsistencies in approach to measurement, it is important to acknowledge that fixation and processing has the potential to change overall tissue size and shape [4-10]. Many authors have demonstrated an overall shrinkage during this process in a variety of tissues. For example, Boonstra et al [4] demonstrated a 15% decrease in uterine cervical measurements after standard fixation and processing. Similarly, Dobrin [5] demonstrated 19% to 25% shrinkage of arterial cross sections from the gross specimen to the mounted glass slides. Weight changes of up to 34% have also been observed in various organs after formalin fixation [11]. Considering the variability of shrinkage demonstrated by these studies, it is possible that the composition and type of the tissue (eg, amount of fat or fibrous tissue) may impact the overall amount of shrinkage. Therefore, it follows that studies addressing change in size should be organ-specific and account for tissue constitution. To the best of our knowledge, there are no studies examining the effect of fixation and processing on breast cancer size measurement.
2. Materials and methods To determine the potential influence of fixation and processing on breast tumor size, specimens from 2 male and 48 female patients aged 32 to 81 years (mean, 60 years) with invasive breast cancer were identified during routine tissue processing in the pathology department of our institution. In each study specimen, the largest
757
1-dimensional measurement in a single, controlled plane of sectioning for each tumor was obtained in 3 states: fresh, after overnight fixation in 10% formalin, and after routine processing and mounting on a glass slide. The plane measured in the fresh and fixed state was embedded in paraffin in a manner to preserve the same measurement plane on histological sections. For inclusion in this study, the tumor had to form a welldefined mass, consist microscopically of at least 75% invasive carcinoma, and be small enough so that a section of the tumor containing the largest dimension could be mounted on a single glass slide. The percentage of invasive cancer was performed by visual estimation from the final mounted specimen. All gross and microscopic measurements were taken in millimeters using standard methods (handheld ruler or microscope ocular) to simulate actual pathology practice. Only the mass-forming lesion was measured microscopically because this corresponded to the gross lesion measured. The fresh and fixed measurements were supervised by the same individual to ensure that the correct section of the gross specimen was measured. An independent pathologist, without prior knowledge of the fresh and fixed measurements, obtained the microscopic measurement on the same tissue block. The type of tumor (eg, ductal, lobular) was recorded for each case, and the percentage of change in tumor size was calculated for each specimen. To determine the potential influence of tumor composition on size change during the histological process, the ratio of fibrosis to glands was calculated by visual estimation from the final mounted specimen.
3. Results The study tumor characteristics, including changes in size, are presented in Table 1. The histological tumor types consisted of 36 ductal, 3 lobular, 6 mixed ductal/lobular, and 5 others. Gross tumor size ranged from 4 to 20 mm (mean, 12.1 mm; median, 12 mm) and tumors contained 10% to 90% fibrous tissue (mean, 52.8%; median, 55%). All tumors contained at least 75% invasive tumor, and 94% contained at least 90% invasive tumor. All tumors infiltrated with a relatively compact invasion front so that under low microscopic power (2 objective; original magnification 20), a virtual circle could be drawn around the tumor, serving as the boundary for microscopic measurement. In 48 cases (96%), there was no difference in measured size between fresh and fixed states. In the remaining 2 cases, the measured size decreased by 1 and 2 mm after overnight fixation in formalin. These tumors contained 30% and 40% fibrous tissue, respectively. After final processing and mounting, a decrease in size from initial fresh measurement was noted in 40% of cases (20 cases; mean difference, 2.4 mm; maximum difference, 7 mm). Twenty-one cases
758 Table 1 Patient 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50
B. Pritt et al. Tumor characteristics and change in measurement from fresh to final processed/mounted state Tissue state Fresh (mm)
Fixed (mm)
Processed (mm)
Size change from fresh to final (%)
Estimated % fibrosis
Histological type
15 14 6 7 19 9 8 11 15 9 15 17 15 5 7 11 7 18 9 11 11 18 20 5 14 19 7 20 14 17 8 13 19 12 13 5 4 19 14 9 4 20 10 12 8 13 11 16 13 10
14 12 6 7 19 9 8 11 15 9 15 17 15 5 7 11 7 18 9 11 11 18 20 5 14 19 7 20 14 17 8 13 19 12 13 5 4 19 14 9 4 20 10 12 8 13 11 16 13 10
14 12 5 7 17 10 8 11 17 8 14 14 15 7 7 13 10 18 7 11 11 16 19 5 14 16 7 14 14 13 7 14 18 10 11 5 4 12 14 11 4 16 10 13 8 13 11 14 14 10
7 14 17 0 11 + 11 0 0 + 13 11 7 18 0 + 40 0 + 18 + 43 0 22 0 0 11 5 0 0 16 0 30 0 24 13 +8 5 17 15 0 0 37 0 + 22 0 20 0 +8 0 0 0 13 +8 0
30 40 70 60 50 70 80 20 20 60 40 50 40 70 40 80 60 40 20 80 40 70 20 80 60 50 70 30 50 70 70 50 50 40 20 30 80 10 90 90 80 10 60 70 30 60 60 40 60 80
D D D D D D/L D M D D AC D/L D D D D D D D/L D D D D D D D D D D D D D D D D D D/M D D L D Med D/L D D/L L D D/L D D
Abbreviations. D, ductal; L, lobular; M, mucinous; Med, medullary; AC, adenoid cystic.
(42%) showed no change in measurement from fresh to final processed state, whereas in 9 cases (18%), the measured size increased on average by 1.7 mm (range, 1-3 mm) after processing/mounting. Most cases were pure
classic ductal-type (76%), whereas 8 cases (16%) had lobular histology or features. The remaining 4 cases had adenoid cystic, medullary, mucinous, and mixed ductal/ mucinous histology.
The effect of tissue fixation and processing on breast cancer size
4. Discussion Despite the potential for size discrepancy among fresh, fixed, and processed states, there is no bofficialQ agreement or recommendation on which state should be used for pathological staging of breast cancers. To the best of our knowledge, this study specifically addresses these issues for the first time to determine which state(s) most accurately reflects measured tumor size. Our results indicate that most carcinomas do not change in size after overnight fixation in 10% formalin. However, substantial changes in tumor size (increases and decreases) were noted after processing of the paraffin block. In 40% of our cases, shrinkage occurred with processing, and a mean decrease of 2.4 mm was noted. Although this difference would not always change the patient’s stage, the pathologist should keep this in mind when considering microscopic tumor measurements. Notably, 9 cases increased in measured dimension after paraffin embedding, mounting, and staining. Because only the mass-forming lesion was measured microscopically, size increase cannot be attributed to microscopic extension of tumor into the tissue surrounding the mass. Although counterintuitive, the most likely explanation for these changes in tumor size is that the tumor sections expanded during block preparation, cutting, or mounting. We cannot entirely exclude the possibility of minor size variations because the gross and microscopic measurements were obtained by different methodology (ruler versus ocular measurement). However, one would expect this variation to affect all specimens of this series to the same extent. Similarly, minimal deviation from the gross measurement could occur because of change in overall tumor size as the microtome sections through the tissue. However, for this particular study, efforts were made to mount the first available tissue section from the paraffin block; thus, the section on the resulting microscopic slide corresponded most closely to what was measured grossly. It is important to recognize that potential causes of both artifactual tissue shrinkage and expansion exist throughout the process of block manufacture and slide preparation. These causes may explain the above findings of tumor size increases and decreases after processing. To begin with, alterations may occur during fixation and dehydration of tissue in the automated processors. Next, during embedding, the temperature of the paraffin is important for regulating the size of the tissue. Paraffin that is too warm will shrink more than expected during cooling, which may compress the tissue and cause it to expand less predictably when sectioned on the microtome. Furthermore, during microtomy, incorrect knife angle or loose knives can cause tissue chatter or compression, which affects size during mounting. Poorly fixed and processed tissue can also behave unpredictably during microtomy and in the mounting water bath. Finally, water bath temperature affects tissue size and must be calibrated to the paraffin
759
used. A too-warm water bath may melt the paraffin, diminishing support for the tissue leading to expansion, whereas a too-cool water bath may not allow the tissue to evenly spread after the natural compression of microtomy. Simple quality checks should be incorporated into the histological process. In our laboratory, the histotechnologist visually compares the size of the mounted section to the size of the tissue within the paraffin block. Major discrepancies in size would prompt preparation of new sections before review by a pathologist. This policy may have limited the range of observed changes on this study. However, despite this safeguard, changes in size were observed in most cases (58%). The percentage of desmoplastic fibrous tissue associated with the invasive tumor was determined for each case by visual estimation at the microscope and did not appear to correlate with the degree of change in measured dimension (Table 1). At the onset of our study, we hypothesized that the fibrosis-to-gland ratio may influence the degree of change in tumor size throughout processing; therefore, we chose a method of measurement that could easily be incorporated into daily pathology practice. Lack of correlation of percentage of fibrosis and degree of overall size change suggests that the observed size changes do not depend on overall tissue composition, but reflects more on general processing artifact in the histology laboratory. Histopathological staging of breast cancers is based on the size of the invasive tumor component, but many invasive carcinomas are associated with various amounts of in situ carcinoma. It is critical to adjust maximum tumor size in such cases after microscopic review because extensive in situ carcinoma can be forming a mass that greatly exceeds the size of the invasive tumor. The current case series included only tumors with a dominant (N75%) invasive tumor component. As per our departmental reporting policy, the percentage of in situ carcinoma is determined for each invasive breast cancer by visual approximation from the glass slide and recorded in the final report. This allowed for easy extrapolation of the percentage of invasive cancer that we confirmed for each case during review for our study, excluding less than 5% of total cases that had an invasive tumor component of below 75%. Further studies are needed to determine the influence, if any, of an in situ component on tissue shrinkage or expansion. The possibility of some degree of selection bias cannot entirely be excluded because this study was not designed to sample nonselected consecutive cases. The accuracy, repeatability, and reliability of our measurements have not been specifically determined as part of the study but rather reflect daily practice. However, we excluded interobserver variability in the gross measurements by assigning only one investigator to this part of the study. The 50 cases of this study represent approximately 20% of resection specimens for invasive carcinoma (excisional biopsies and partial and total mastectomies with or without axillary dissection) processed by our pathology department in a
760 1-year period (2003-2004). The goal of this study was to determine potential tumor size changes during histological processing that would have a significant impact on subsequent tumor staging. Thus, only lesions 2 cm in size or smaller and where a section of the entire lesion fits onto 1 glass slide were included. Specimens were also excluded when the accuracy of the histological diagnosis could potentially have been jeopardized by inclusion in the current study. This concerned a significant number of cases that were smaller than 5 mm and/or where existing clinical study protocols required extensive tumor tissue sampling for other purposes. Finally, it is possible that tumors of varying histological subtype (eg, ductal, lobular) will behave differently during histological processing. This is beyond the scope of our study and merits further investigation. Based on our data, one can arguably measure the size of an invasive breast cancer from either the fresh or fixed state (gross measurements) without affecting accuracy. We therefore recommend that staging measurements be taken from 1 of these 2 states. Furthermore, we advise that the pathologist verify the gross measurements microscopically when possible. In most cases, it is our opinion that the gross measurement will provide the most accurate staging information. However, in instances where significant associated in situ disease is present or when the invasive tumor extends microscopically beyond the grossly measured mass, then the microscopic measurements of the invasive component are more accurate and should be used for staging. We strongly advise against decreasing or increasing tumor staging based solely on discrepant microscopic measurements because processing artifact may lead to
B. Pritt et al. significant tissue shrinkage or expansion after processing and embedding.
References [1] Breast. In: Greene FL, Page DL, Fleming ID, Fritz AG, Balch CH, Haller DG, et al, editors. American Joint Committee on Cancer (AJCC) cancer staging manual. 6th ed. New York7 Springer; 2002. p. 223 - 40. [2] Rosen PP. Rosen’s breast pathology. Philadelphia7 Lippincott Williams & Wilkins; 2001. [3] Fitzgibbons PL, Page DL, Weaver D, Thor AD, Allred DC, Clark GM, et al. Prognostic factors in breast cancer: College of American Pathologists Consensus Statement 1999. Arch Pathol Lab Med 2000;124:966 - 78. [4] Boonstra H, Oosterhuis JW, Oosterhuis AM, Fleuren GJ. Cervical tissue shrinkage by formaldehyde fixation, paraffin wax embedding, section cutting and mounting. Virchows Arch A Pathol Anat Histopathol 1983;402:195 - 201. [5] Dobrin PB. Effect of histologic preparation on the cross-sectional area of arterial rings. J Surg Res 1996;61:413 - 5. [6] Friedman NB. On formalin fixation. Hum Path 1992;23:1440 - 1. [7] Goldstein NS, Soman A, Sacksner J. Disparate surgical margin lengths of colorectal resection specimens between in vivo and in vitro measurements. The effects of surgical resection and formalin fixation on organ shrinkage. Am J Clin Pathol 1999;111:349 - 51. [8] Lane J, Ralis ZA. Changes in dimensions of large cancellous bone specimens during histological preparation as measured on slabs from human femoral heads. Calcif Tissue Int 1983;35:1 - 4. [9] Schned AR, Wheeler KJ, Hodorowski CA, Heaney JA, Ernstoff MS, Amdur RJ, et al. Tissue-shrinkage correction factor in the calculation of prostate cancer volume. Am J Surg Pathol 1996;20:1501 - 6. [10] Siu KF, Cheung HC, Wong J. Shrinkage of the esophagus after resection for carcinoma. Ann Surg 1986;203:173 - 6. [11] Schremmer CN. Gewichts7nderungen verschiedener Gewebe nach Formalinfixierung. Frankf Z Pathol 1967;77:299 - 304.