Continuous-specimen-flow, high-throughput, 1-hour tissue processing: a system for rapid diagnostic tissue preparation. (New Technology in Laboratory Medicine).
During the last 2 decades, as a consequence of the introduction of microwave energy into histology laboratories, progress has been made in the development of rapid processing methods and the elimination of toxic reagents. (6-10) Methods reported so far, however, are neither sufficiently practical nor shorten tissue-processing time to the extent required for their widespread acceptance.
We report a simple, practical, and fully automated method that reduces processing time to about 1 hour, and therefore enables pathologists to render a diagnosis in 2 to 3 hours. This method does not require prior fixation of tissue and obviates the need for formaldehyde and xylene during processing, leading to improved preservation of RNA.
MATERIALS AND METHODS
Approximately 32 000 surgical specimens are submitted annually to the Pathology Laboratory at the University of Miami/ Jackson Memorial Medical Center, Miami, Fla. About 27 000 of these specimens originate from surgical procedures at Jackson Memorial Hospital, and the remaining are from the University of Miami Hospital and Clinics, including the Sylvester Comprehensive Cancer Center. About 110 000 paraffin blocks are prepared every year from these tissues. Small biopsy specimens are received in the laboratory in containers partially filled with 10% sodium phosphate-buffered formalin (4% formaldehyde). Large specimens, such as uteri, breasts, intestines, and placentas, are usually received fresh or in various-sized containers with formalin. Until September 1997, the tissues were processed overnight following the steps outlined in Table 1. A few emergency biopsies were processed within 2 hours using the short cycle of the automated processor (Tissue-Tek VIP, Sakura Finetek, Torrance, Calif).
During the developmental stages of our continuous-throughput processing method (CTPM), we used samples from residual large surgical specimens for evaluation. As soon as the efficacy of the new methodology was established, we began routine processing of specimens with CTPM. The system uses either fresh or prefixed specimens. The tissue slices, however, should be no thicker than 1.5 mm. Consequently, small biopsy specimens require no trimming. Large specimens are trimmed on a dissecting board specifically developed to prepare uniform slices of desired thickness (Figure 1). This board was designed by inserting a 26.0 x 5.0-cm metallic plate on a 45.0 x 30.0 x 2.5-cm flat plastic slab. The metallic plate has two 1.5-mm-deep cutouts, measuring 3.0 x 2.5 cm and 2.5 x 1.8 cm. Tissues are placed over either of the cutouts and slid horizontally along the slicing plate. During slicing, specimens are held in place by the index finger or with a plastic holder applied to the exposed surface of the tissues. The use of this cutting board produces uniform, 1.5-mm-thick slices of tissue. To make this board equally useful for conventional processing, an additional 3.0-mm-deep slot was cut in the metallic plate, which can procure thicker sections if needed.
[FIGURE 1 OMITTED]
The CTPM as presently practiced in the Department of Pathology at the University of Miami/Jackson Memorial Medical Center is outlined in Table 2 and schematized in Figure 2. It is fully described elsewhere. (11) Briefly, we use mixtures of common histology reagents, such as isopropyl alcohol, acetone, polyethylene glycol, mineral oil, paraffin, and minute amounts of glacial acetic acid and dimethyl sulfoxide. Formaldehyde and xylene are not used during processing. During gross dissection, the sections are placed in regular plastic cassettes and immersed in a holding solution identical to that used in the first step (see "Step 1"), but without acetic acid. The sections are kept in this solution for about 5 to 15 minutes, depending on the time required for dissecting the surgical specimen. Because the first step of processing takes 15 minutes, this holding stage does not cause any delay in adding new samples to the system. Up to 30 cassettes are loaded into a plastic basket and subjected to the following steps:
[FIGURE 2 OMITTED]
Step 1.--The samples are microwaved at 62 [degrees] C for 15 minutes in 1000 mL of a solution of acetone, isopropyl alcohol, polyethylene glycol, glacial acetic acid, and dimethyl sulfoxide. Air bubbling produces agitation.
Step 2.--The samples are heated for 15 minutes at 62 [degrees] C in a second microwave in 1000 mL of a solution similar to that used in step 1, to which low-viscosity mineral oil has been added. The solution is agitated by bubbling.
Step 3.--The samples are heated for 5 minutes at 62 [degrees] C in a third microwave in 1000 mL of a solution similar to that used in step 2, but with a higher concentration of mineral oil. The solution is agitated by bubbling.
Steps 4 and 5.--The samples are incubated at 65 [degrees] C in 2 consecutive 1000-mL solutions of a mixture of low-viscosity mineral oil and paraffin. A vacuum of 640 mm Hg is applied for 5 minutes to each bath.
Steps 6 through 9.--Impregnation is accomplished by incubation at 65 [degrees] C in 4 baths of molten paraffin applying a vacuum of 640 mm Hg. Tissue sections are transferred from 1 paraffin bath to the next at 5-minute intervals, for a total impregnation time of 20 minutes.
Because step 1 lasts 15 minutes, the system can be accessed with additional samples every 15 minutes, as more specimens become available.
Although the physical agents and reagents have remained fairly constant since the processing system was adapted in 1997, the manner in which these agents are applied has been modified. Initially, commercial microwave tissue processors (H2500 or H2800, Energy Beam Sciences, Inc, Agawan, Mass) were used for the first 3 steps of the procedure. Toward the development of a fully automated system, a transitional phase resulted in the creation of a specifically designed microwave applicator (Microwave Materials Technologies, Inc, Knoxville, Tenn), which was used from late 1999 to early 2000. With either system, the samples were manually transferred from one microwave to another. Treatment of tissue samples with the mineral oil/paraffin mixture (steps 5 and 6) and the 4 final impregnation steps were carried out in a large desiccator resting in a glycerin bath. In January 2000, we began using a robotic instrument that fully automated the processing system. Main features of this instrument are (1) a robotic arm that transfers the samples from one chamber to the next; (2) 3 microwave units, the chambers of which serve as the container for the 1000-mL solutions; and (3) paraffin-impregnation stations providing heat and vacuum. The microwave units allow for uniform distribution of the microwave energy, thus avoiding damage to the processed tissue.
Following paraffin impregnation and embedding, the sections are cut at 3 to 5 [micro]m and stained with hematoxylin-eosin. When indicated, histochemical and immunohistochemical stains are performed following standard techniques.
For RNA extraction, 20 ribbons (20 [micro]m thick) of paraffin-embedded tissue were deparaffinized with xylene twice for 15 minutes and washed in 100% ethanol 2 times for 10 minutes. After brief air drying, the tissue sections were homogenized with 1 mL of Trizol reagent (GIBCO BRL, Grand Island, NY) using power homogenizer. Two microliters of Pellet Paint (Novagen, Madison, Wis) fluorescent dye-labeled coprecipitant was added to each sample. The RNA was isolated following the Trizol protocol (GIBCO), isopropanol precipitation, air drying, and dissolution in 40 mL ribonuclease-free sterile water. Concentration was estimated by spectrophotometry, and the integrity of total RNA was tested by assessing the resolution of 28S and 18S ribosomal RNA in 0.8% agarose gel stained with ethidium bromide. The complementary DNA was generated from 100 ng of total RNA with random hexamer primers using a Gene-Amp RNA polymerase chain reaction (PCR) kit (Perkin Elmer, Foster City, Calif) and was amplified by reverse transcriptase (RT)-PCR using glyceraldehyde phosphate dehydrogenase (GAPDH) primers with AmpliTag DNA polymerase (Perkin Elmer). The cycle parameters for the PCR were 30 cycles at 94 [degrees] C for 1 minute, 55 [degrees] C for 1 minute, and 72 [degrees] C for 1 minute with a 1-time final extension at 72 [degrees] C for 7 minutes.
Comparison With CPM
To compare CTPM and CPM, 44 duplicate sets of tissue slices were taken from 38 randomly selected surgical specimens. One set was processed by CPM and the other by CTPM. Histologic sections were double-mounted on the same slide; the CTPM sections were placed toward the frosted end of the slide and the CPM sections on the opposite end. Thus, tissue sections were subjected to identical conditions during staining with hematoxylin-eosin. Histochemical and immunohistochemical assays were similarly performed using these double-mounted slides.
Hematoxylin-eosin-stained slides were evaluated for integrity of tissue architecture, as well as for details of nuclei and nucleoli, preservation of cytoplasmic structure, cellular secretions, stroma/ epithelium interface, and possible presence of crenation of red blood cells and hemolysis.
We also evaluated the impact of the new method on our surgical pathology turnaround times by comparing the release of final reports during a period of 6 months (January 1 to June 30, 1997) that preceded implementation of the new procedure with that of a comparable recent period (January 1 to June 30, 2000).
To compare the preservation of RNA using CTPM and CPM, 6 specimens were procured fresh, immediately upon surgical removal from patients. These specimens consisted of normal testis, normal ovary, leiomyoma of the uterus, hyperplastic prostate gland, and 2 breast excisions, one with carcinoma and the other with fat necrosis. One sample from each of these surgical specimens was processed by CTPM and CPM. An additional sample from each specimen was also processed by CTPM, but with 0.05% diethyl pyrocarbonate (DEPC) added to the solutions in steps 1, 2, and 3. The total RNA was isolated and amplified by RT-PCR using primer for the housekeeping gene GAPDH (see "Processing") and run on the agarose gel.
The CTPM method reduced the processing of both fresh and fixed tissue samples to approximately 60 minutes. Compared to the conventional method, blocks of tissue prepared by the CTPM are relatively softer, hence easier to cut and make ribbons. Evaluation of the quality of histology sections and staining properties of the double-mounted slides revealed comparable results with CTPM and CPM (Figures 3 through 6). In general, however, sections prepared by CTPM exhibited brighter staining with eosin and stronger reaction with hematoxylin. We have since reduced the exposure times for hematoxylin-eosin to one half of what was required with CPM. Other differences observed included widely patent uterine lymphatics in tissue processed using CTPM (Figure 3). The overall tissue architecture, stroma, secretory products, cell morphology, and nuclear morphology, however, appeared practically the same. Histochemical stains were also comparable in the double-mounted slides of tissues processed by CPM and CTPM (Figure 5). The intensity of immunohistochemical reactions and pattern of distribution of the antigens were for the most part identical in slides processed by CPM and CTPM (Figure 6).
[FIGURES 3-6 OMITTED]
Table 3 summarizes the impact of CTPM on turnaround times for our surgical pathology reports. The reported turnaround times apply to all cases, including those requiring deeper sections, histochemical or immunohistochemical stains, and intradepartmental consultations.
The integrity of extracted RNA following CTPM, particularly with addition of DEPC, was significantly improved, although 28S and 18S ribosomal bands were degraded in most samples. Nevertheless, when 100 ng of the same RNA was used for RT-PCR amplification (using GAPDH primers) the expected 500-base pair (bp) band was obtained in all RNA specimens extracted following CTPM, but in only 3 of the 6 specimens extracted after CPM. When amplification was obtained from CPM-processed tissue, the intensity of amplified product on ethidium bromide gel was significantly weaker than amplification product obtained in CTPM-processed tissue (Figure 7).
[FIGURE 7 OMITTED]
Because of the 8 hours or longer presently required to prepare tissues for histology, diagnosis of biopsy specimens on the day after receipt of the specimen is customary. The diagnosis of complex specimens may take even longer. Figure 2 schematizes the various steps, relevant methods, and their approximate time requirements from receipt of tissues in the laboratory to availability of microscopic sections for diagnosis. The CPM is particularly accountable for delaying the evaluation of surgically removed tissue, as this is the lengthiest of the stages of diagnostic tissue preparation. The various steps required by CPM are responsible for this long process, that is, incubation in separate solutions of formalin for fixation, a series of increasing concentrations of alcohol for dehydration, and xylene for clearing tissue of alcohol before impregnation. This 8-hour or longer process is usually performed overnight in automated instruments. The pathologist, therefore, is unable to render a diagnosis based on microscopic examination of tissue sections until the next day at the earliest, almost 24 hours after the specimen arrives in the laboratory.
Attempts are being made to reduce the time for tissue processing, either by shortening the cycle of automated CPM processors or by using microwave-assisted methods. A shortened method of the CPM is presently practiced. To that effect, tissues are usually fixed for a minimum of 30 minutes in formalin before placing them in automated processors, which are programmed for a cycle that lasts from about 70 to 150 minutes (Figure 2). However, only small specimens are amenable to this procedure. Moreover, the processor cannot be accessed with additional samples until completion of the entire cycle. Because of these constraints, the shortened method is used almost exclusively for the handling of emergency biopsies.
Several investigators, notably Boon and colleagues, (7-9) Leong, (6) and Visinoni et al, (10) have pioneered the use of microwaves to expedite fixation and/or processing of tissue for microscopy. Because dielectric solutions and tissue sections absorb microwave energy, it is translated into heat, both in the solutions and within the tissues immersed in those solutions. The resulting effect is accelerated diffusibility of the solutions and denaturation of proteins; both phenomena lead to shortened processing time. Boon and colleagues (7-9) and Visinoni and associates (10) have developed procedures that reduce the number of steps and the time customarily required to process fixed biopsies and larger samples. This reduction was accomplished primarily by combining vacuum and microwave energy and using mixtures of reagents different from those used in conventional methods. In these procedures, the processing time for fixed tissue varies from 30 minutes to 2 hours, depending on the size or thickness of the samples. These methods use a specially designed microwave oven into which the processing solutions are manually brought in and out.
Although the aforementioned microwave-assisted and conventional shortened methods reduce processing time significantly, they are attendant to a number of technical difficulties that would seem to preclude their widespread acceptance. The requirement to fix specimens for a number of hours prior to entering the processing stage in microwave methods would make them impractical for same-day diagnostic tissue preparation. Furthermore, none allow for adding specimens until the entire cycle is complete; specimen batching is therefore required. It is also cumbersome to adjust the timing of each step according to the size of the specimens, as required by some of the protocols mentioned. (9,10) Moreover, microwave methods heretofore reported are not automated, another significant impediment for the daily practice of histology.
The method we describe provides for continuous flow of specimens from dissection to wax impregnation. This system can accept new specimens every 15 minutes, leading to availability of microscopic slides for diagnosis throughout the day and, most importantly, on the same day that those specimens are received. Since the implementation of CTPM, more than one third of our surgical pathology specimens are diagnosed and reported on the same day that they are received. This rate compares very favorably with our previous practice and with that of others as reported in surveys of departments of pathology. (1,2) This important improvement in our practice has caused no compromise in the quality of the diagnostic material, because routine histochemical and immunohistochemical stains obtained by CTPM are comparable to CPM.
Another advantage of CTPM is that it requires only about 10% of the total volume of reagents, which also are considerably less toxic than those used in CPM. Furthermore, lack of requirement for prefixation and complete elimination of formalin and xylene from the system are important for the safety of personnel; the method also minimizes the cost of recycling or disposal of these chemicals. Moreover, there is no need to monitor the instrument overnight, as is the case with CPM because of the possibility of instrument failure. Because the method is composed of well-defined steps with standard volumes, times, pressure, and temperatures, it has led to the development of a robotic processing instrument, thus converting a previously manual procedure to a fully automated one. A comparison of CPTM with other currently used processing methods is summarized in Table 4 and Figure 2.
Notwithstanding the aforementioned benefits provided by CTPM, we are far from realizing the full potential of this method. One of its most important attributes may reside in providing high-quality slides for histopathologic diagnosis together with superior RNA preservation. Although the integrity of RNA in our preliminary work was partially compromised, it was of sufficient quality that segments up to 500 bases could be consistently amplified, provided that the processed tissue was submitted fresh, immediately after surgical removal. As most PCR-amplified fragments in molecular pathology are smaller than 500 bases, it suggests that CTPM-RNA can be used for most PCR amplifications. A prerequisite for RNA preservation in CTPM-processed tissue is to either obtain freshly removed surgical specimens or use a special preservative that maintains morphologic integrity and RNA quality. Such a solution is currently being evaluated for transportation and storage of surgical specimens.
In summary, we have developed and are now using a method for preparing tissues for diagnosis that offers many advantages over the long-established tissue-processing procedures. Because this method facilitates continuous flow of tissue samples, it may eventually lead to the automation of all steps of diagnostic tissue preparation, akin to the handling of liquid tissue samples in the clinical laboratory.
Table 1. Conventional Processing Method Heat, Concentra- Time, [degrees] C Step Solution tion, % min (Convective) 1 Formalin 10 120 40 2 Formalin 10 120 40 3 Alcohol 80 30 40 4 Alcohol 95 30 40 5 Alcohol 95 45 40 6 Alcohol 100 45 40 7 Alcohol 100 45 40 8 Alcohol 100 45 40 9 Xylene 100 45 40 10 Xylene 100 45 40 11 Paraffin ... 30 60 12 Paraffin ... 30 60 13 Paraffin ... 30 60 14 Paraffin ... 30 60 Pressure/ Agita- Volume, Step Vacuum tion L 1 Yes Yes 4 2 Yes Yes 4 3 Yes Yes 4 4 Yes Yes 4 5 Yes Yes 4 6 Yes Yes 4 7 Yes Yes 4 8 Yes Yes 4 9 Yes Yes 4 10 Yes Yes 4 11 Yes Yes 4 12 Yes Yes 4 13 Yes Yes 4 14 Yes Yes 4 Table 2. Continuous-Throughput Processing Method Heat, [degrees] C Time, Convec- Micro- Step Liquids min tive wave 1 Solution I * 15 ... 62 2 Solution II 15 ... 62 ([dagger]) 3 Solution II 5 ... 62 ([dagger]) 4 Mineral oil/ paraffin 5 65 ... 5 Mineral oil/ paraffin 5 65 ... 6 Paraffin 5 65 ... 7 Paraffin 5 65 ... 8 Paraffin 5 65 ... 9 Paraffin 5 65 ... Volume, Step Vacuum Agitation L 1 ... Bubbling 1 2 ... Bubbling 1 3 ... Bubbling 1 4 Yes ... 1 5 Yes ... 1 6 Yes ... 1 7 Yes ... 1 8 Yes ... 1 9 Yes ... 1 * Solution of isopropyl alcohol, acetone, polyethylene glycol, glacial acetic acid, and dimethyl sulfoxide. ([dagger]) Solution containing same reagents as step 1, with the addition of low-viscosity mineral oil. Table 3. Comparison of Diagnostic Turnaround Times of Surgical Specimens * January-June 1997 January-June 2000 CPM, No. (%) CTPM, No. (%) Same day 56 (0) 4919 (36) 1 day 5139 (44) 4198 (31) 2 days 3029 (26) 2027 (15) 3 days 1887 (16) 1006 (7) >3 days 1534 (13) 1440 (11) Total 11 645 13 590 * CPM indicates conventional processing method; CTPM, continuous-throughput processing method. Table 4. Comparison of Continuous-Throughput Processing Method (CTPM) With Other Currently Used Processing Methods * Other Microwave- Based Regular Short Methods CTPM CPM CPM ([dagger]) Prefixation required No Yes Yes Yes Specimens added every 15 min Yes No No No Fresh tissue to paraffin in 1 h Yes No No No Formalin used in processing No Yes Yes No Xylene used in processing No Yes Yes No High-quality HC and IHC Yes Yes Yes Yes Relative RNA integrity Yes No No Not reported Complete automation Yes Yes Yes No * HC indicates histochemistry; IHC, immunohistochemistry; and CPM, conventional processing method. ([dagger]) Please see references 6-10.
(1.) Zarbo RJ, Gephardt GN, Howanitz PJ. Intralaboratory timeliness of surgical pathology reports. Arch Pathol Lab Med. 1996;120:234-244.
(2.) Novis DA, Zarbo RJ, Saladino AJ. Interinstitutional comparison of surgical biopsy diagnosis turnaround time. Arch Pathol Lab Med. 1998;122:951-956.
(3.) OSHA Regulations (Standards--29 CFR). Medical surveillance--formaldehyde. Standard Number 1910.1048AppC.
(4.) Warmington AR, Wilkinson JM, Riley CB. Evaluation of ethanol-based fixatives as a substitute for formalin in diagnostic clinical laboratories. J Histotechnol. 2000;23:299-308.
(5.) Liotta L, Petricoin E. Molecular profiling of human cancer. Nature Rev. 2000;1:48-55.
(6.) Leong AS-Y. Microwave irradiation in histopathology. Pathol Annu. 1998; 23(pt 2):213-233.
(7.) Boon ME, Kok LP, Ouwerkerk-Noordam E. Microwave-stimulated diffusion for fast processing of tissue: reduced dehydrating, clearing and impregnation times. Histopathology. 1986;10:303-309.
(8.) Kok LP, Boon ME. Ultrarapid vacuum-microwave histoprocessing. Histochem J. 1995;27:412-419.
(9.) Boon ME, Wals-Paap C, Visinoni FA, et al. The two-step vacuum-microwave method of histoprocessing. Eur J Morphol. 1995;33:349-358.
(10.) Visinoni F, Milios J, Leong AS-Y, et al. Ultra-rapid microwave/variable pressure induced histoprocessing: description of a new tissue processor. J Histotechnol. 1998;21:219-224.
(11.) Essenfeld E, Essenfeld H, Morales AR, inventors; University of Miami, Miami, Fla, assignee. High quality, continuous throughput, tissue fixation--dehydration--fat removal--impregnation method. US patent 6 207 408, B1. March 2001.
Accepted for publication January 2, 2002.
From the Department of Pathology, University of Miami/Jackson Memorial Medical Center, Miami, Fla (Drs Morales, Vincek, and Nadji and Ms Duboue); and the Essenfeld Pathology Laboratory, Policlinica Metropolitana, Caracas, Venezuela (Drs H. Essenfeld and E. Essenfeld).
Dr Morales, Dr E. Essenfeld, and Dr H. Essenfeld participate in a Sponsored Research Program with Sakura Finetek USA through the University of Miami and are listed as inventors of US patent 6 207 408. The University of Miami has licensed this patent to Sakura Finetek USA, and the inventors will receive a percentage of the proceeds gained by the University of Miami.
Reprints: Azorides R. Morales, MD, Department of Pathology (D-33), University of Miami School of Medicine, PO Box 016960, Miami, FL 33101 (e-mail: firstname.lastname@example.org).