Justification for Use of a Single Trichrome Stain as the Sole Means for Routine Detection of Intestinal Parasites in Concentrated Stool Specimens

Abstract

Of 12,321 stool samples analyzed over a 6-year interval, 870 (7.1%) were positive for a total of 1,019 parasites, of which 1,011 (99.2%) were found in trichrome-stained smears of unconcentrated specimens while only 479 (47.0%) were detected in iodine-stained smears of concentrated samples. Stool specimens were next analyzed by trichrome staining of both unconcentrated and concentrated specimens preserved in either mercury-polyvinyl alcohol (PVA) or cupric PVA. Of 2,198 specimens, 171 (7.8%) were positive for a total of 208 parasites, 192 (92.3%) and 204 (98.1%) of which were found in the unconcentrated and concentrated specimens, respectively (P < 0.05). In our patient population, examination of a single trichrome-stained smear of a concentrated stool specimen is a cost-effective alternative to routinely analyzing both concentrated and unconcentrated specimens for parasites.

In the past, a thorough examination of stool specimens for detection of parasites involved four microscopic analyses: direct saline and iodine wet mounts of fresh (unpreserved) stool, an iodine wet mount of a concentrated sample, and a trichrome-stained smear of an unconcentrated sample (1, 4, 8). Because motile trophozoites cannot be detected in direct wet mounts of preserved specimens, many laboratories are now routinely performing only two microscopic analyses on stool specimens submitted for detection of parasites: trichrome staining of the unconcentrated sample and iodine staining of a wet mount of the concentrated specimen (2, 4, 14). In a continuing effort to maintain test sensitivity while reducing material and labor costs as well as exposure to toxic substances, including mercury and formaldehyde, there is an increasing interest in reevaluating the efficiency of older microscopic approaches and concepts, looking for more-rapid, less expensive, potentially less toxic, and equally sensitive alternatives. In 1996, Hale et al. (9) documented the accuracy of using for parasite detection a single, permanent, modified-trichrome-stained slide prepared from a concentrated specimen preserved in cupric polyvinyl alcohol (PVA).

The present study was performed in two phases and had two objectives. The sensitivity in our patient population of parasite detection by trichrome staining of unconcentrated specimens preserved in mercury-PVA was retrospectively compared with that of iodine wet mounts of concentrated specimens preserved in formalin. In addition, the sensitivities of trichrome staining of unconcentrated and concentrated specimens that had been preserved in either mercury-PVA or cupric PVA were prospectively compared.

Retrospective determination of sensitivity of trichrome staining of unconcentrated specimens versus that of iodine wet mounts of concentrated specimens.

Prior to 1997, all stool specimens submitted for ova and parasite analysis were inoculated into a two-vial preservative kit (Meridian Diagnostics, Cincinnati, Ohio). One vial contained mercuric chloride-PVA and the other contained 10% buffered formalin. Trichrome staining was performed on the unconcentrated specimen in mercury-PVA, and an iodine wet mount of the concentrated specimen in formalin was done (4, 11). At least 200 to 300 fields of the trichrome-stained smears were microscopically examined under 10× or 20× and then 100× objectives for a total of approximately 8 min for each specimen. The entire coverslip of each formalin smear was examined under the 10×, 40×, and 100× objectives for approximately 7 min.

Prospective comparison of sensitivities of trichrome stainings of unconcentrated and concentrated specimens.

During 1997, each stool specimen submitted for ova and parasite detection was inoculated into a single vial containing either mercuric chloride-PVA (for approximately 12% of the patients) or copper sulfate-PVA (for about 88% of the patients), as recommended by Hale et al. (9). A portion of each PVA-preserved specimen was used to prepare a smear for trichrome staining of the unconcentrated sample. The staining procedure used was a modification of the Wheatly trichrome stain procedure incorporating Hemo-De (1). Preserved specimens were then concentrated by a saline–Hemo-De procedure (1, 9, 12) in which formalin was replaced by saline and ethyl acetate was replaced by Hemo-De. Smears of the concentrates were made by adding 2 drops of the centrifuged pellet of a specimen (resuspended in 0.9% saline) to 1 drop of cupric PVA or mercury-PVA (depending on the chemicals in which the specimen was received) on a glass slide. Smears were air dried at room temperature overnight or at 37°C for 4 h. The trichrome staining procedure was identical to that used for unconcentrated-specimen smears. Trichrome-stained smears were microscopically analyzed for smear adequacy and for the presence and internal structure of parasites, using 10× or 20× and then 100× objectives and scanning 200 to 300 fields for approximately 8 min.

The Z test for differences in proportions for independent samples was used for statistical analysis of results (13). A P value of <0.05 was selected as the minimum level determining significance.

Detection of parasites in trichrome-stained smears of unconcentrated specimens versus iodine wet mounts of concentrated specimens.

From 13 February 1990 through 31 August 1996, 12,321 preserved specimens from a total of 9,366 patients were analyzed for ova and parasites. A total of 1,019 parasites were detected in 870 (7.1%) of the specimens from 767 (8.2%) of the patients. Of the 1,019 parasites found, 540 (53.0%), 8 (0.8%), and 471 (46.2%) were found only in trichrome-stained smears of unconcentrated specimens, only in iodine wet mounts of concentrated specimens, and in both the unconcentrated-specimen trichrome-stained smears and the concentrated-specimen smears, respectively (Table 1). Trichrome staining of unconcentrated specimens permitted the detection of 99.2% (1,011 of 1,019) of the total parasites found, while iodine wet mounts of concentrated specimens allowed the detection of only 47.0% (479 of 1,019) of the total parasites. This difference was significant (P < 0.001).

TABLE 1.

Detection of parasites by trichrome staining of unconcentrated specimens and by iodine staining of wet mounts of concentrated specimens

Parasite (n)	No. (%) of parasites detected in:
	Unconcentrated, trichrome-stained specimens alone	Iodine-stained concentrated specimens alone	Both systems
B. hominis (424)	256 (60.4)	0	168 (39.6)
Chilomastix mesnili (8)	7 (87.5)	0	1 (12.5)
Cyclospora spp. (3)	0	0	3 (100)
D. fragilis (87)	73 (83.9)	0	14 (16.1)
Endolimax nana (73)	60 (82.2)	0	13 (17.8)
Entamoeba spp.
E. coli (46)	23 (50.0)	1 (2.2)	22 (47.8)
E. hartmanni (11)	7 (63.6)	0	4 (36.4)
E. histolytica (25)	25 (100)	0	0
Enterobius vermicularis (2)	0	1 (50.0)	1 (50.0)
G. lamblia (321)	79 (24.6)	6 (1.9)	236 (73.5)
Iodamoeba butschlii (13)	9 (69.2)	0	4 (30.8)
Strongyloides stercoralis (3)	0	0	3 (100)
Trichuris trichiura (3)	1 (33.3)	0	2 (66.7)
Total (1,019)	540 (53.0)	8 (0.8)	471 (46.2)

Determination of recovery of parasites by trichrome staining of unconcentrated and concentrated specimens.

From 1 January 1997 through 31 December 1997, 2,198 preserved specimens from 1,623 patients were analyzed for parasites. A total of 208 parasites were detected in 171 (7.8%) of the specimens and in 133 (8.2%) of the patients. Of 182 parasites detected after using cupric PVA, 167 (91.8%) were found in trichrome-stained smears of the unconcentrated specimens while 178 (97.8%) were found in trichrome-stained smears of the concentrated specimens (P < 0.05) (Table 2). Of 26 parasites detected when mercuric PVA was used, 25 (96.2%) and 26 (100%) were found in trichrome-stained smears of the unconcentrated and concentrated specimens, respectively (not significant) (Table 3).

TABLE 2.

Detection of parasites in trichrome-stained smears of unconcentrated and concentrated specimens preserved in Cu-PVA

Parasite (n)	Total no. (%) of parasites detected in trichrome-stained smears of:
	Unconcentrated specimens	Concentrated specimens
B. hominis (64)	60 (93.8)	63 (98.4)
Cyclospora spp. (1)	1 (100)	1 (100)
D. fragilis (23)	19 (82.6)	22 (95.7)
Endolimax nana (21)	21 (100)	19 (90.5)
Entamoeba coli (6)	6 (100)	6 (100)
G. lamblia (63)	57 (90.5)	63 (100)
Iodamoeba butschlii (4)	3 (75.0)	4 (100)
Total (182)	167 (91.8)	178 (97.8)

TABLE 3.

Detection of parasites in trichrome-stained smears of unconcentrated and concentrated specimens preserved in Hg-PVA

Parasite (n)	Total no. (%) of parasites detected in trichrome-stained smears of:
	Unconcentrated specimens	Concentrated specimens
B. hominis (12)	12 (100)	12 (100)
D. fragilis (1)	1 (100)	1 (100)
Entamoeba coli (2)	2 (100)	2 (100)
Enterobius vermicularis (1)	0	1 (100)
G. lamblia (10)	10 (100)	10 (100)
Total (26)	25 (96.2)	26 (100)

The size, shape, and staining characteristics of each parasite found, with the exception of Blastocystis hominis, were indistinguishable regardless of whether smears were stained following preservation in cupric PVA or in mercury-PVA. When smears were stained after preservation with cupric PVA, the features of B. hominis were not as distinctive as they were after treatment with mercury-PVA. However, with experience, laboratory personnel could reliably detect this parasite when either fixative was used. Material and labor costs for processing and reading trichrome-stained smears of both a concentrated and an unconcentrated sample from each patient were $10.20, or $22.420 for 2,198 specimens processed during the course of the year. Reading only the single trichrome-stained smears of concentrated specimens would have resulted in a savings, during the year, of $11,210.

The retrospective analysis of 6 1/2 years of recovery data indicated that in our patient population, routine use of both an iodine-stained smear prepared from the formalin concentrate and a trichrome-stained direct smear (the two methods currently recommended for routine detection of parasites [4, 11]) was not cost-effective. Garcia et al. (3) have reported similar findings. In their study, 2,204 (99.8%) and only 644 (29.2%) of pathogenic protozoa were detected by trichrome staining of unconcentrated samples and wet mounts of formalin-ether concentrates, respectively. Those authors showed that both the cyst and trophozoite forms of Entamoeba histolytica, Entamoeba hartmanni, and Giardia lamblia, as well as the trophozoite form of Dientamoeba fragilis, were easily missed if the laboratory examination was limited to only a microscopic analysis of a wet mount of the concentrated sediment. On the basis of the results of the present study, this laboratory no longer routinely collects stool specimens in formalin but relies on single specimens that are individually preserved in a vial of PVA.

During our prospective study of parasite detection by trichrome staining of specimens preserved in either cupric or mercuric PVA, our rates of detection (7.8% of the specimens and 8.2% of the patients were positive) were almost identical to our rates of detection in prior years, when both trichrome staining of unconcentrated samples and iodine wet mounts of concentrated specimens were routinely performed (7.1% of the specimens and 8.2% of the patients were positive during the earlier interval). Furthermore, of the 208 parasites detected during the prospective study, 204 (98.1%) were found in trichrome-stained smears of concentrated specimens while only 192 (92.3%) were detected by similar staining of the unconcentrated samples. Therefore, a single trichrome-stained smear of a concentrated specimen, as suggested by Hale et al. (9), appears to offer an acceptably sensitive, cost-effective alternative to processing and microscopically analyzing both unconcentrated and concentrated stool specimens from our patient population.

Advantages of processing only one slide, a trichrome-stained smear of a concentrated stool specimen, when routine ova and parasite detection is ordered include the use of only one transport vial per specimen (instead of two); the elimination of toxic chemicals, including mercury (if cupric PVA is used) and formalin; increased detection of parasites and a greater number of organisms per positive slide when the concentrated sediment is stained instead of the unconcentrated smear; the availability of a permanently stained slide that may be filed; and the obvious cost savings, in both material and labor, when only one slide per specimen is stained and analyzed (9).

Some parasites, such as Isospora belli, Cryptosporidium spp., and Cyclospora spp., may stain poorly by the modified trichrome stain procedure. Examination of a modified-acid-fast-stained smear of the PVA concentrate may be necessary to confirm the presence of these parasites. Hale et al. (9) reported that helminth ova with thick outer shells were easily identified by using the modified trichrome stain. For helminth ova with thin shells, including Hymenolepis spp. and Schistosoma spp., examination of an iodine wet mount of the PVA concentrate aided in species identification. One previous study reported that 14.6% of helminth ova or larvae were detected only by trichrome staining of concentrated specimens (15). Some parasites, including G. lamblia, hookworm eggs, and Trichuris eggs, may not concentrate as well from specimens preserved in PVA as from those preserved in formalin (4). Laboratory personnel should be aware of the prevalence of parasites in their own populations and select their diagnostic methods accordingly.

Copper sulfate has been added to Schaudinn’s fixative as a less toxic, more environmentally compatible alternative to mercuric chloride. Its use for detection of parasites, however, has been controversial. In 1981, Horen (10) reported that smears containing protozoa fixed with copper sulfate (but not cobalt chloride) exhibited satisfactory nuclear and cytoplasmic morphological detail following trichrome staining. Hale et al., in 1996, found that protozoa exhibited less shrinkage and more-distinct internal structure with trichrome staining after saline concentration of cupric-PVA-preserved specimens, compared to the morphology observed in stained smears of unconcentrated specimens in cupric PVA (9). However, Garcia et al. (6) reported in 1983 that the overall morphology of protozoa preserved in PVA with copper sulfate was not as good as that achieved when the protozoa were preserved in PVA with mercuric chloride. In the present study, there was no apparent difference in the background colors or the morphological or staining characteristics of protozoa detected when copper was used instead of mercury, with the exception of one species: B. hominis. The differing results reported when copper sulfate-containing fixatives were used could be a function of their sources. Commercially available copper sulfate-PVA today, at least that from the source used during the present study, appears to be an adequate, although not perfect, alternative to mercuric chloride-PVA.

Other mercury-free fixatives that have been recently reported to perform satisfactorily include zinc sulfate-PVA for trichrome-stained smears (7) and EcoFix for both trichrome- and EcoStain-stained smears (5). Many of these alternatives may result in some staining of protozoa which lacks the color intensity or precise nuclear and cytoplasmic detail seen when the organisms are stained following fixation in mercuric chloride. As long as these alternatives permit accurate, reliable detection and identification of the parasites, however, laboratory personnel may have to accept the fact that protozoa may not always be as crisply stained as they were when mercuric chloride was used (7), and with experience, they will most likely become accustomed to small differences in staining characteristics.

Routine laboratory tests designed to detect and identify parasitic pathogens should be selected after taking into consideration the prevalence of different pathogens in the specific laboratory’s population. In some parts of the United States with large populations of immigrants or AIDS patients, a single trichrome-stained smear of a concentrated specimen may not adequately detect parasites that normally cause disease in those patients. In our population in central Pennsylvania, as well as in that of Hale et al. (9) (Salt Lake City), the routine use of an iodine wet mount prepared from a formalin–Hemo-De concentrate was not cost-effective. In both populations, the use of a single trichrome-stained smear of a concentrated stool specimen preserved in cupric PVA offered very acceptable sensitivity for routine parasite detection as well as a substantial savings in terms of materials and labor. Performance of iodine wet mounts or modified acid-fast staining of the concentrated specimens when necessary and appropriate permits the detection or confirmation of suspected parasites, including helminths, I. belli, Cryptosporidium spp., and Cyclospora spp.