Diagnosis of Congenital Syphilis From Placental Examination: Comparison of Histopathology, Steiner Stain, and Polymerase Chain Reaction for Treponema pallidum DNA

DAVID R. GENEST, MD, SUNG R. CHOI-HONG, MD, JAMES E. TATE, MS, FAISAL QURESHI, MD, SUZANNE M. JACQUES, MD, AND CHRISTOPHER CRUM, MD

Congenital syphilis is often a presumptive diagnosis (based on serologies), because confirmation requires identification of Trepo- nema pallidum in fetal/neonatal tissues or in the placenta. Placental histological features associated with congenital syphilis include the triad of enlarged hypercellular villi, proliferative fetal vascular changes, and acute or chronic villitis. The authors blindly evaluated 49 formalin-fixed, paraffin-embedded placentas (38 with positive ma- ternal syphilis serologies; 11 with negative serologies) and compared results of histology, Steiner stain, and polymerase chain reaction (PCR) for T paUidum DNA. Histology was categorized as positive (triad present), suspicious (two thirds of triad present), or negative. Treponemal DNA was detected by amplifying a 189 base pair region of the 47 kd treponemal membrane antigen with 44 cycles of PCR; products were detected by Southern blot. Placentas from the 11 sero- negative mothers were all negative by histology, Steiner stain, and PCR. Among the 38 placentas from serologically positive mothers, 4 had positive histology (2 of 4 positive Steiner; 4 of 4 positive PCR); 6 had suggestive histology (0 of 6 positive Steiner; 1 of 6 positive

Primary, secondary, and congenital syphilis have all dramatically increased in prevalence over the past decade] '2 In one recent study, despite prior maternal penicillin therapy for primary or secondary syphilis, congenital syphilis developed in 49 of 108 (45%) neo- nateQ; possible explanations for this high rate of treat- ment failure include maternal reinfection, poor compli- ance, antibiotic resistance, and inadequate dosage. These data indicate an ominous trend that pediatric and perinatal pathologists will be confronted increas- ingly with patienfs at risk for congenital syphilis.

The diagnosis of congenital syphilis in the perina- tal period is often difficult, 4'5 in part, because neonatal serological test ing's unreliable and because Treponema pallidum cannot be cultivated on artificial medium. Ac- cording to the Centers for Disease Control guidelines, the diagnosis of congenital syphilis is confirmed only when treponemal organisms are identified in fetal, neo- natal, umbilical cord, or placental tissues. 6 Histological examination of the placenta may be helpful in this re- gard, because the placental histopathologic triad of en-

From the Division of Women's and Perinatal Pathology, Depart- ment of Pathology, Brigham and Women's Hospital, and Harvard Medical School, Boston; MA; and the Department of Pathology, Hut- zel Hospital and Wayne State University, Detroit, MI. Accepted for publication November 14, 1995.

Address correspondence and reprint requests to David R. Genest, MD, Department of Pathology, Brigham and Women's Hospital, 75 Francis St, Boston, MA 02115.

Copyright © 1996 by W.B. Saunders Company 0046-8177/96/2704-000955.00/0

PCR); and, 28 had negative histology (0 of 28 positive Steiner; 1 of 28 positive PCR). PCR identification of treponemal DNA was significantly associated with the triad (P = .0003), proliferative fetal vascular changes (P = .0003), acute villitis (P = .003), chronic villitis (P = .004), and spirochetes on Steiner stain (P = .01). These results (1) confirm a strong association between placental histopathologic features and congenital syphilis; (2) indicate that when such features are present, PCR of placental tissue may confirm the diagnosis of congenital syphilis; and (3) suggest that even when such features are absent, PCR of placental tissue may identify additional cases of histologically unsuspected congenital syphilis. HUM PATHOL 27:366-- 372. Copyright © 1996 by W.B. Saunders Company

Key words: polymerase chain reaction, congenital syphilis, Trepo- nema pallidum, placenta, villitls.

Abbreviations: PCR~ polymerase chain reaction; H&E, hematoxy- lin-eosin; RPR, rapid plasma reagin card test; FrA-ABS, fluorescent treponemal antibody absorption test.

larged hypercellular villi, proliferative fetal vascular changes, and acute or chronic villitis has been de- scribed as characteristic of placental syphilitic infec- tion, 7,s although spirochetes are often difficult to locate with silver stains. ~'9"1°

Recently, sensitive molecular techniques have been proposed for diagnosing congenital syphilis. The polymer- ase chain reaction (PCR) has been used to detect trepone- mal DNA in human amniotic fluid, neonatal cerebrospinal fluid, and fetal and neonatal seFa. 11'12 PCR has also been used to identify treponemal DNA in formalin-fixed, paraf- fin-embedded, experimentally infected rabbit testis. 13 For the pathologist, a sensitive detection assay for bacterial DNA in formalin-fixed, paraffin-embedded tissues is opti- mal for confirming spirochetes in placental tissues. The authors, therefore, used this molecular technique to study placentas of potential cases of congenital syphilis. To com- pare histopathology with PCR, the authors studied archival placentas by conventional histopathology (using hematoxy- lin-eosin [H&E] and Steiner stains) and molecular biology (using PCR for treponemal DNA). These results substanti- ate a strong association between previously reported pla- cental histopathologic findings TM and congenital syphilis, but also suggest that PCR may identify additional, histo- pathologically inapparent, cases of congenital syphilis.

METHODS Case Selection

Forty-nine archival placental specimens were selected f rom the files of Hutzel Hospital (1990 to 1993); 11 placentas

366

PLACENTA IN CONGENITAL SYPHILIS (Genest et al)

were from rapid plasma reagin card test (RPR)-negative mothers, and 38 placentas were from RPR fluorescent trepo- nemal antibody absorption test (FTA-ABS) -positive mothers. Forty-six of the 49 placentas were derived from second- or third-trimester deliveries (35 of 46 from RPR/FrA-ABS-posi- tive women, and 11 of 46 from RPR-negative women), and three placentas represented first-trimester spontaneous abor- tions, all from RPR/FFA-ABS-positive patients. Two patholo- gists (F.Q. and S.MJ.) from Hutzel Hospital selected one rep- resentative block of placental parenchyma per placenta; subsequently, the placental samples were blindly reassessed histologically by two pathologists at the Brigham and Wom- en's Hospital (D.R.G. and S.R.C.H.), without knowledge of the serological status of the mother, the initial histopathologic diagnosis, or the PCR results.

Histological Assessment

Each paraffin block was serial sectioned, routinely processed, and stained for standard light microscopy with H&E and Steiner stain] 4 H&E-stained placentas were assessed for the following histological features: (1) enlarged, hypercellular villi (defined as immature appearing villi with prominent stromal fibroblasts or macrophages); (2) proliferative vascular change of fetal vascula- ture (defined as thickened fetal capillaries, arterioles, or venules, with prominent perivascular connective tissue or vascular smooth muscle); (3) villitis or perivillifis (characterized as acute or chronic); (4) plasma cell deciduitis; and (5) acute chorioamnio- nitis. Specimens with the histological triad of enlarged hypercellu- lar villi, proliferative fetal vascular changes, and villitis were catego- rized as histologically positive for congenital syphilis; specimens with two of the three triad features were categorized as histologi- cally suspicious. Steiner stains were assessed, with appropriate con- trols, at ×40 magnification; structures suggestive of spirochetes were confirmed by ×100 magnification under oil immersion.

Identification of T pallidum DNA

Serial sections from each placental specimen were placed in a digestion buffer containing 50 m m o l / L Tris hydrochloric acid pH 8.5, in 1 m m o l / L ethylenediaminetetra-acetic acid 200 # g / m L proteinase K, and 0.5% Tween-20, and were incu- bated at 62°C for 16 hours. Chelex-100 (BioRad, Richmond, CA) was added to 30%, and the specimens were boiled for 10 minutes, then centrifuged for 5 minutes in a microcentri- fuge. 15 The aqueous phase below the paraffin was placed in fresh tubes, and 1 #L of each sample was amplified by PCR in a 50-#L reaction containing the primers TP47A2 and TP47B2, which were designed from the 47-kD membrane antigen of T pallidum ~6'I7 to amplify a 189-base pair product. The se- quence of primer TP47A2 is 5 ' -CAT/GGT/TGA/CAG/CGA/ GGA/ATA/CA-3'; the sequence of primer TP47B2 is 5 ' -CGT/ GCA/GAA/AAA/CTA/TCC/TCA/GTG/A-3 ' . PCRs were performed according to the manufacturer 's specifications (Cetus, Emoryville, CA) in a thermocycler with an initial dena- turation at 95°C for 5 minutes, followed by 44 cycles of 95°C for 30 seconds, 44°C for 30 seconds, and 72°C for 1 minute, and a final 5-minute extension at 72°C.

To maximize sensitivity, PCR products were detected by Southern blot analysis. ~8 Briefly, PCR products were electro- phoresed in 2% agarose with 0.5 # g / m L ethidium bromide, visualized and photographed under ultraviolet light, and transferred to Nytran nylon membranes (Schleicher & Schuell, Keene, NH) by capillary blot overnight. Blots were probed overnight at 44°C, washed at 67°C, and exposed to autoradiography film. All positive specimens had an aurora-

diograph band that migrated to the expected size of 189 base pairs when aligned with the gel photographs.

To exclude contamination, PCR reactions were run with negative reagent and tissue controls. To verify that the tissue DNA was suitable for amplification, samples were also ana- lyzed for human sequences (globin). To verify specificity, PCR reactions were run with a positive tissue control. The positive control was a Steiner-positive skin biopsy from a RPR/YI'A- ABS-positive patient with secondary syphilis. The specificity of the treponemal DNA from this sample was verified by di- gesting a ~ZP-labeled 430-base pair PCR product of the 47- kd antigen from this sample with Rsa I, Hae III, and Hha I. The expected sizes of 160, 119, 91, and 60 base pairs were observed after polyacrylamide gel electrophoresis and autora- diography (data not shown). This Steiner-positive biopsy was also used to generate the probe for the Southern blots. After amplifying the 189-base pair product in PCR and electropho- resing through 2% agarose, the product band was excised, purified by Gene Clean (Bio 101, LaJolla, CA), and labeled with [a-~zP]-deoxycytidine triphosphate (dCTP) using ran- dom prime-labeling and Klenow enzyme (Boehringer Mann- heim Biochemical, Indianapolis, IN).

Statistical Analysis

The relationships between placental/decidual histology, Steiner stain, and PCR for treponemal DNA were compared using Fisher's exact two-tailed test. ]9 A P value of <.05 was considered statistically significant.

RESULTS

T h e re la t ionsh ip o f ma t e rna l serological status to the p lacenta l h i s topa thologic triad, S te iner stain for spi- rochetes , and PCR for t r e p o n e m a l DNA is s u m m a r i z e d in Tab le 1.

Serologically Negative Mothers

A m o n g the 11 p lacentas f r o m RPR negat ive moth - ers, no spec imen was histologically positive ( tr iad pres- ent) or suspicious (two o f th ree tr iad fea tures present ) for congeni ta l syphilis. Ste iner stain for sp i rochetes and PCR for t r e p o n e m a l DNA were negat ive in all 11 speci- mens . O n e p lacen ta (9%) h a d p l a sma cell deciduitis, one p l acen ta (9%) h a d chron ic villitis, and four p lacen- tas (36%) showed acute chor ioamnioni t i s .

Serologically Positive Mothers

A m o n g the th ree f irs t- tr imester s p o n t a n e o u s abor- t ion spec imens f r o m RPR/FFA-ABS-pos i t ive women , th ree spec imens (100%) showed p lasma cell deciduitis, no p lacen ta was histologically positive, bu t one p lacen ta (33%) was histologically suspicious (en la rged hypercel- lular villi with focal acute villitis). S te iner stain for spiro- chetes and PCR for t r e p o n e m a l DNA were negat ive in all s p o n t a n e o u s abor t ion spec imens .

A m o n g the 35 second- and th i rd- t r imester p lacen- tas f r o m RPR/FTA-ABS-pos i t ive mothers , four p lacen- tas (11%) were histologically positive ( tr iad presen t ) (Figs 1 to 3); five p lacentas (14%) were histologically suspicious (two o f th ree tr iad fea tures p resen t ) ; a n d 26 p lacentas (74%) were histologically negative. O t h e r

367

TABLE 1.

HUMAN PATHOLOGY Volume 27, No. 4 (April 1996)

Comparison of Maternal Serology, Placental Histology, Steiner Stain for Spirochetes, and PCR for Treponemal DNA

Maternal Syphilis Serology Placental Histology Steiner Stain Positive

for Spirochetes PCR Positive for

Treponemal DNA

11 with negative serology

38 with positive serology

Triad present: 0/11 - - - - Triad absent: 11/11 0/11 0/11 Triad present: 4/38 2 /4 4 /4 Triad absent: 34/38 0/34 2/34

Abbreviation: PCR, polymerase chain reaction.

histological findings in the 35 placentas included plasma cell deciduitis (five specimens; 14%) and acute chorioamnionitis (11 specimens; 31%).

Among the four histologically positive second- and third-trimester placentas (triad present), two of four placentas (50%) had spirochetes identified by Steiner stain (Fig 4); in each specimen, fewer than 10 spiro- chetes were found, either as single organisms or small groups of organisms, always in the fibrotic stroma of villi involved by acute villitis/perivillitis. Among the four histologically positive specimens, four of four placentas

FIGURE I . Villous enlargement with prominent stromal hyper- cellularity is one feature of the placental histopathologic triad associated with congenital syphilis. (H&E stain; original magni- fication ×60.)

(100%) had treponemal DNA identified by PCR (Fig 5, lanes 1, 3, 4, and 21).

Among the five histologically suspicious second- and third-trimester placentas (two of three triad features pres- ent), all Steiner stains were negative, but one of five placen- tas (20%) had treponemal DNA identified by PCR. This histologically suspicious/PCR-positive placenta showed en- larged hypercellular villi and proliferative fetal vascular changes but no viUitis. This placenta was one of dichori- onic twin placentas; of interest, the other twin's placenta, also included in this study, showed histological and molec- ular evidence of congenital syphilis (ie, it contained the triad and was positive for treponemal DNA by PCR).

Among the 26 histologically negative second- and third-trimester placentas, all Steiner stains were negative, but one of 26 placentas (4%) had treponemal DNA identi- fied by PCR. This histologically negative/PCR positive pla- centa showed acute chorioamnionitis and proliferative fetal vascular changes; in the absence of additional triad fea- tures, the fetal vascular abnormalities had been interpreted as a nonspecific changes after fetal death, z°'21

Clinical and pathological features of the six PCR- positive second- and third-trimester placentas from se- rologically positive mothers are summarized in Table 2. Three surviving infants were treated with penicillin for definite (two) or presumed (one) congenital syphi- lis. The other three were perinatal deaths; congenital syphilis was not evaluated because autopsies were not performed. The relationship of placental histology, Steiner stain, and PCR for treponemal DNA was as- sessed for the 35 second- and third-trimester placentas from serologically positive mothers (Table 3). Three placental histological features (triad, acute villitis, and proliferative fetal vascular changes) were significantly associated with the identification of syphilitic organisms on Steiner stain. Six histological features (triad, >~ two of three triad features, >1 one of three triad features, proliferative fetal vascular changes, acute villitis, and chronic villitis) were significantly associated with the identification of treponemal DNA by PCR. The relation- ship between PCR treponemal positivity and Steiner stain positivity was also assessed; PCR positivity was pres- ent in two of two (100%) placentas with a positive Steiner stain versus four of 47 (9 %) placentas with nega- tive Steiner stain (P = .01, Fisher's exact two-tailed test).

DISCUSSION

Several placental pathological abnormalities associ- ated with congenital syphilis have been desc r ibed . 7-9'99-97

368

PLACENTA iN CONGENITAL SYPHILIS (Genest et al)

FIGURE 2. Proliferative vas- cular change of fetal ves- sels, a second feature of the placental histopathologic triad associated with con- genital syphilis, involves muscularized vessels of stem villi (A) as well as capillaries of terminal villi (B); the latter is characterized by concen- tric pericapillary stromal fi- brosis (B and arrows). (H&E stain; original magnification: A, ×60, B, ×250).

Heavy placental weight 24 and decreased fetal-to-placental weight ratio 2v are the simplest, but least specific, of such placental findings. Necrotizing funisitis is an unusual um- bilical cord lesion that has been variously characterized as either strongly 2s or sporadically 25'28'29 associated with con- genital syphilis; this abnormality is characterized, both grossly and histologically, by distinctive pefivascular tings of necrosis and inflammation within Wharton's jelly.

Most placental pathological studies of congenital syphilis have emphasized villous and decidual histologi- cal abnormalities, 7-9'~2'26'27 particularly the histopatho- logic triad of enlarged hypercellular villi, proliferative fetal vascular changes, and acute or chronic villitis. One study systematically described placental pathology in 25 RPR/FTA-ABS-positive pregnancies, of which seven in- fants developed congenital syphilis, and 18 infants did

FIGURE 3. Villitis is a third feature of the placental histopathologic triad associated with congenital syphilis. Acute villitis (A and B) is typically multifocal, with lesions ranging in severity from tiny subtrophoblastic collections of neutrophils (A), to larger microabscesses, often with central fibrotic villi (B); in the fibrotic stroma of one villus (arrow), three spirochetes were found by Steiner stain (Fig 4A). Chronic villitis (C) is often subtle and focal; an inconspicuous perivascular lymphohistiocytic infiltrate was present in several foci in this placenta. (H&E stain; all panels, original magnification ×250).

3 6 9

HUMAN PATHOLOGY Volume 27, No. 4 (April 1996)

FIGURE 4. In two placentas, Steiner stain revealed scattered single organisms (arrows) in rare fibrotic villi involved by acute villitis. (Steiner stain; original magnification ×625.)

not. 8 Several other reports described placental findings in a total of 23 infants who developed congenital syphi- lis. 7'9"~'~6'27 Collectively, among the 30 previously re- ported placentas of infants with congenital syphilis, the complete histopathologic triad was present in 43% (13 of 30 placentas) ; two features of the triad were observed in 47% (14 of 30 placentas) (ie, 12 placentas had en- larged hypercellular villi and villitis, and two placentas had enlarged hypercellular villi and proliferative vascu- lar changes); and one feature of the triad (enlarged hypercellular villi) was found in the remaining 10% (three of 30 placentas). This suggests that although the complete triad may be characteristic of congenital syphilis, many (if not most) placentas do not display the full triad.

Placental spirochetes may be widely scattered and 7 9 1 0 difficult to locate in congenital syphilis. ' ' Using conven-

tional silver staining (Dieterle, Steiner, or Warthin-Starry stains), spirochetes were identified in 11 of 16 (69%) placentas studied in three prior reportsT-9; most placentas showed the complete histological triad. Treponemal iden- tification was described as difficult in several instances (ie, lengthy search for bacteria was necessary in one placenta7; only rare "atypical, fragmented bacteria" were found in

9 another specimen ). Additional limitations of silver stain- ing include technical difficulty, inconsistent results (even with positive controls), and frequent artifacts (such as

14.30 31 elastic fibers in normal tissues). ' ' These limitations of conventional silver staining have prompted the evalua- tion of ancillary methods of spirochete identification including immunohistochemical ~° and immunofluo- rescent staining. 31~4

Because microscopic identification of placental spi- rochetes can be difficult, the authors wished to assess whether the placental histopathologic triad alone was sufficient to confirm the diagnosis of congenital syphi- lis. An important related question was: How often might congenital syphilis exist without manifesting the placen- tal histological triad? To answer these questions, a sensi-

tive and reliable method of identifying treponemal or- ganisms in placental tissues is required. Because PCR for treponemal DNA has previously been shown to be sensitive (capable of identifying suspensions containing as few as one to 10 organisms),l.~ the authors used PCR (in addition to conventional silver staining) to address these issues.

This study shows a strong association between the placental histopathologic triad and the presence of treponemal organisms in the placenta, identified both by Steiner stain (P = .01) and PCR for treponemal DNA (P = .0003). Among the four placentas showing the histological triad, treponemal DNA was confirmed in all four specimens (100%); among the 31 placentas without the complete histological triad, treponemal DNA was confirmed in two specimens (6%). Although the authors studied a small series of placentas, these results suggest that the placental histopathologic triad may have a high specificity but a lower sensitivity for the presence of treponemal DNA. Among 29 serologically positive specimens without t reponemal DNA, no pla- centa had the complete histological triad (ie, specificity of triad 100%). Among six placentas with treponemal DNA, the complete histological triad was present in four placentas (ie, sensitivity of triad 67%).

Of note, among four placentas with the complete histological triad in this study, although treponemal DNA was confirmed by PCR in every specimen, spiro- chetes were identified with Steiner stain in only two instances, despite careful, prolonged searching. This finding suggests that a presumptive diagnosis of con- genital syphilis may be warranted when the complete placental histopathologic triad is present in serologi- cally positive patients, even when spirochetes cannot be confirmed by conventional methods. Alternatively,

1 2 3 4 A

- 4,

B 21 - ÷ ÷

C 3 3 ÷ -- ÷

FIGURE 5. PCR/Southern blot analysis of placental extracts. DNA was extracted and amplified in PCR with primers de- signed from the 47-kD membrane antigen of Tpallidum. Prod- ucts were electrophoresed in 2% agarose, transferred to nylon membrane and probed with a 32P-labeled DNA fragment am- plified from positive controls. Samples 1 to 4, 21, and 33 and the positive controls show strong amplification of the 47-kd antigen gene.

3 7 0

PLACENTA IN CONGENITAL SYPHILIS (Genest et al)

TABLE 2. Six Placentas Positive for Treponemal DNA by PCR: Clinical and Pathological Findings

Case Steiner Birth No. Placental Histology Stain Pregnancy Weight (g) Outcome

1 Triad (with acute and chronic + 40 wk 2353 villitis) ; chorioamnionifis

2 Triad (with acute and chronic + 32 wk 2650 villitis); chorioamnionitis; plasma cell deciduitis

Triad (with acute and chronic villitis); chorioamnionitis

4 Triad (with chronic villitis); - plasma cell deciduitis

5 Two of triad (enlarged villi - and fetal vascular changes)

6 One of triad (fetal vascular - changes); chorioamnionitis

31 wk; no prenatal care; 2041 crack cocaine use; fetal death

26 wk; dichorionic twins 680 (twin of case no. 5)

26 wk; dichorionic twins 670 (twin of case no. 4)

26 wk; ruptured 765 membranes; crack cocaine use; fetal death

Pediatric diagnosis: definite congenital syphilis--treated with penicillin

Pediatric diagnosis: definite congenital syphilis (+CSF) -- t reated with penicillin

Stillbirth; no autopsy

Pediatric diagnosis: possible congenital syphilis--treated with penicillin

Neonatal death (1 day old); no autopsy

Stillbirth; no autopsy

Abbreviations: PCR, polymerase chain reaction; CSF, cerebrospinal fluid.

confirmation of congenital syphilis could be sought by a more sensitive technique, such as PCR.

In this study, adequate precautions and controls were used to ensure against false positives by PCR. How- ever, the number of false negatives (ie, the specificity of this assay) is not known, without an independent, reliable method of detecting all treponemal infections. Exponential amplification makes PCR a sensitive tech- nique, but a potential limitation is scarcity of the target sequences in a sample. These histopathologic results, together with previous reports, 7'9'~° indicate that placen- tal spirochetes may be scarce in congenital syphilis. Be- cause the authors extracted DNA from at most 10 8-#m sections, and because only 1 of 200 #L was used as template for PCR, it is possible that this small aliquot contained insufficient treponemal DNA to be detected by PCR in some infected placentas. Increased sensitivity might be gained by purifying and concentrating DNA

from a larger sample of tissue, or using two rounds of PCR with nested primers and fresh reagents each round. However, such increases in sensitivity could come at a cost of specificity, unless this issue is carefully addressed with negative tissue controls.

In conclusion, these data confirm a strong associa- tion between the placental histopathologic triad and congenital syphilis. It also suggests that treponemal or- ganisms may be present in the placenta even when the histopathologic triad is absent. Thus, in the placentas of serologically positive patients, it is reasonable to search for spirochetes with silver stains even when the complete triad is absent, particularly when acute villitis is found. Furthermore, when silver stains are negative for spirochetes, it is possible that sensitive molecular techniques of treponemal identification, such as PCR, may increase the sensitivity of the placenta for the diag- nosis of congenital syphilis. However, the effectiveness

TABLE 3. Thrity-Five Third-Trimester Placentas From VDRL/FTA-ABS-Positive Women: Histological Findings Compared With Results of Steiner Stain and PCR for Treponemal DNA

Steiner Stain PCR

+ - + -

Histological Feature N N = 2 N = 33 P N = 6 N = 29 P

Triad 3 features (positive histology) 4 2 (100%) 2 (6%) .01 4 (67%) 0 />2 features (suspicious or positive histology) 9 2 (100%) 7 (21%) NS 5 (83%) 4 (14%) I>1 feature 17 2 (100%) 15 (45%) NS 6 (100%) 11 (38%)

Individual triad features Villous enlargement/hypercellularity 15 2 (100%) 13 (39%) NS 5 (83%) 10 (34%) Chronic villitis 10 2 (100%) 8 (24%) NS 5 (83%) 5 (17%) Acute villitis 3 2 (100%) 1 (3%) .005 3 (50%) 0 Proliferative fetal vascular changes 7 2 (100%) 5 (15%) .04 5 (83%) 2 (7%)

Other histological findings Plasma cell deciduitis 5 1 (50%) 4 (12%) NS 2 (33%) 3 (10%) Acute chorioamnionitis 11 2 (100%) 9 (27%) NS 4 (67%) 7 (24%)

.0003

.002

.007

NS .004 .003 .0003

NS NS

Abbreviations: FFA-ABS, fluorescent treponemal antibody absorption test; PCR, polymerase chain reaction; NS, not statistically significant. * P > .05.

371

HUMAN PATHOLOGY Volume 27, No. 4 (April 1996)

of this approach can be determined by only prospective studies.

REFERENCES

1. Webster LA, Rolfs RT, Nakashima AK, et al: Regional and temporal trends in the surveillance of syphilis, United States, 1986- 1990. MMWR Morb Mortal Wkly Rep 40:29-33, 1991

2. Zenker PN, Berman SM: Congenital syphilis: Trends and rec- ommendations for evaluation and management. Pediatr Infect DisJ 10:516-522, 1991

3. McFarlin BL, Bottoms SF, Dock BS, et al: Epidemic syphilis: Maternal factors associated with congenital infection. Am J Obstet Gynecol 170:535-540, 1994

4. Srinivasan G, Ramamurthy RS, Bharathi A, et al: Congenital syphilis: A diagnostic and therapeutic dilemma. Pediatr Infect Dis J 2:436.441, 1983

5. Mascola L, Pelosi R, BlountJH, et al: Congenital syphilis revis- ited. AmJ Dis Child 139:575-580, 1985

6. Centers for Disease Control: Congenital syphilis: New York City, 1986.1988. MMMWR Morb Mortal Wkly Rep 38:825-829, 1989

7. Russell P, Altshuler G: Placental abnormalities of congenital syphilis. A m J Dis Child 128:160-163, 1974

8. Qureshi F, Jacques SM, Reyes MP: Placental histopathology in syphilis. HUM PATHOL 24:779-784, •993

9. Walter P, Blot P, IvanoffB: The placental lesions in congenital syphilis. Virchows Arch A Pathol Anat Histopathol 397:313-326, 1982

10. Benirschke K, Kaufman P: Pathology of the Human Placenta (ed 2). New York, NY, Springer-Verlag, 1990, pp 575-578

11. Grimprel E, Sanchez PJ, Wendel GD, et al: Use of polymerase chain reaction and rabbit infectivity testing to detect Treponema pal lidum in amniotic fluid, fetal and neonatal sera, and cerebrospinal fluid. J Clin Microbiol 29:1711-1718, 1991

12. Sanchez PJ, Wendel GD, Grimprel E, et al: Evaluation of molecular methodologies and rabbit infectivity testing for the diagno- sis of congenital syphilis and central nervous system invasion by Trepo- nema pallidum. J Infect Dis 167:148-157, 1993

13. BurstainJM, Grimprel E, Lukehart SA, et al: Sensitive detec- tion of Treponema pallidum by using the polyrnerase chain reaction. J Clin Microbiol 29:62-69, 1991

14. Sparks CK, Genest DR, Crum CP: Rapid reliable microwave modification of Steiner silver impregnation method, for diagnosis of spirochetae and non-filamentous bacteria. J Histotechnol 18:315-318, 1995

15. TateJE, Mutter GL, Prasad CJ, et al: Analysis of HPV-positive

and -negative vulvar carcinomas for alterations in c-myc, Ha-, Ki-, and N-ras genes. Gynecol Oncol 53:78-83, 1994

16. Dallas WS, Ray PH, Leong J, et al: Identification and purifi- cation of a recombinant Treponema pallidum basic membrane protein antigen expressed in Escherichia coll. Infect Immun 55:1106-1115, 1987

17. Hsn P, Chamberlain NR, Orth K, et al: Sequence analysis of the 47-kilodalton major integral membrane immunogen of Treponema paUidum. Infect Immun 57:196-203, 1989

18. SambrookJ, Fritsch EF, Maniatis T: Analysis and cloning of eukaryotic genomic DNA, in Molecular Cloning: A Laboratory Man- ual (ed 2). Cold Spring Harbor, NY, Cold Spring Harbor Laboratory Press, 1989, pp 9.31-9.59

19. Gustafson TL: True Epistat Manual. Richardson, TX, Epistat Services, 1991, pp 8/8-8/11

20. Fox H: Morphologic changes in the human placenta follow- ing fetal death. J Obstet Gynaecol Br Commonw 88:224-229, 1968

21. Genest DR: Estimating the time of death in stillborn fetuses. II. Histologic evaluation of the placenta: A study of 71 stillborns. Obstet Gynecol 80:585-592, i992

22. Brannstein H: Congenital syphilis in aborted second trimes- ter fetus: Diagnosis by histopathological study. J Clin Pathol 31:265- 267, 1978

23. Fojaco RM, Hensley GT, Moskowitz L: Congenital syphilis and necrotizing funisitis. JAMA 26I:1788-1790, 1989

24. Malan AF, Woods DL, Van der Elst CW, et al: Relative placen- tal weight in congenital syphilis. Placenta 11:3-6, 1990

25. Jacques SM, Qureshi F: Necrotizing funisitis: A study of 45 cases. HUM PaTHOL 23:1278-1283, 1992

26. Samsom GR, Meyer MP, Blake DRB, et al: Syphilitic placenti- tis: An immunopathy. Placenta 15:67-77, 1994

27. Young SA, Crocker DW: Occult congenital syphilis in macer- ated stillborn fetuses. Arch Pathol Lab Med 118:44-47, 1994

28. Benirschke K. Congenital syphilis and necrotizing funisitis. JAMA 262:904, 1989 (letter)

29. Craver RD, Baldwin VJ: Necrotizing funisitis. Obstet Gynecol 79:64-70, 1992

30. Beckett JH, Bigbee JW: Immunoperoxidase localization of Treponema pallidum. Arch Pathol Lab Med 103:135-138, 1979

31. Epstein H, King CR: Diagnosis of congenital syphilis by im- munofluorescence following fetal death in utero. AmJ Obstet Gyne- col 152:689-690, 1985

32. Hatter CA, Benirschke K: Fetal syphilis in the first trimester. AmJ Obstet Gynecol 124:705-711, 1976

33. Mote PT, Hunter EF, Van Orden AE, et al: Immunofluores- cence and Treponevna infection. Arch Pathol Lab Med 106:295-297, 1982

34. Hunter EF, Greer PW, Swisher BL, et al: Immunofluorescent staining of Treponema in tissues fixed with formalin. Arch Pathol Lab Med 108:878-880, 1984

372