DERMATOPATHOLOGY

Histopathologic features distinguishing

secondary syphilis from its mimickers
Alexandra Flamm, MD,a Veronica Merelo Alcocer, MD,a Viktoryia Kazlouskaya, MD, PhD,b
Eun Ji Kwon, MD,c and Dirk Elston, MDd
Hershey, Pennsylvania; Brooklyn and Port Chester, New York; and Charleston, South Carolina

Background: Syphilis is often misdiagnosed clinically, and biopsies might be required.

Objective: To determine histopathologic features that distinguish secondary syphilis from pityriasis
lichenoides (PL), pityriasis rosea (PR), and early mycosis fungoides (MF).

Methods: Histopathologic features of 100 cases of syphilis, 110 cases of PL, 72 cases of PR, and 101 cases of
MF were compared.

Results: Elongated rete ridges and interstitial inflammation favor syphilis over PL (likelihood ratios 3.44
and 2.72, respectively), but no feature reliably distinguishes between them. Secondary syphilis and PR can
be distinguished by neutrophils in the stratum corneum, plasma cells, interface dermatitis with lymphocytes
and vacuoles, and lymphocytes with ample cytoplasm. Plasma cells and lymphocytes with ample
cytoplasm are rare in early MF and can be used as distinguishing features.

Conclusions: Histopathologic features characteristic of syphilis can be seen in PL, PR, and early MF.
Distinguishing syphilis from PL can be difficult histologically, and a high index of suspicion is required.
Although elongation of rete and interstitial inflammation favor syphilis, plasma cells (historically considered
a significant feature of syphilis) are often encountered in PL. Vacuolar interface dermatitis with a
lymphocyte in every vacuole is considered characteristic of PL, but this feature appears to be more
common in syphilis. ( J Am Acad Dermatol 2020;82:156-60.)

Key words: dermatopathology; mycosis fungoides; pityriasis lichenoides; pityriasis rosea; secondary
syphilis; syphilis.

S yphilis is the great imitator, both clinically and
histopathologically. Although histopathologic
features of syphilis have been described, large
comparative studies are lacking. Several studies have
shown that endothelial swelling, interstitial inflam-
mation, irregular acanthosis, elongated rete ridges,
lichenoid pattern, and presence of plasma cells are
the most common features of secondary syphilis, but
how reliably these features can be used to differen-
tiate syphilis from other inflammatory dermatoses is
unclear.1-3
We compared histopathologic features of second-
ary syphilis with other inflammatory dermatoses that
can mimic syphilis clinically or histologically
(pityriasis lichenoides [PL], pityriasis rosea [PR], and
mycosis fungoides [MF]). PL is known to have some
overlapping features with secondary syphilis,
including interface dermatitis and the presence of
neutrophils in the stratum corneum.4 We have
shown previously that secondary syphilis lesions
mimicking PR on the trunk tended to have fewer
diagnostic histopathologic features compared with
syphilitic lesions at other sites, which might
confound the diagnosis.2 Lymphocytic atypia might
also be observed in syphilis, raising the possibility of
MF.5-7 In addition, a pattern of interface dermatitis
with a lymphocyte in every vacuole is common in
both secondary syphilis and MF.8

From the Department of Pathology and Dermatology, Penn State
Hershey Medical Centera; Department of Dermatology, SUNY
Downstate Medical Center, Brooklynb; Dermpath Diagnostics
New York, Port Chesterc; and Department of Dermatology,
University of South Carolina, Charleston.d
Funding sources: None.
Conflicts of interest: None disclosed.
Accepted for publication July 8, 2019.
Reprint requests: Alexandra Flamm, MD, Department of
Dermatology, Penn State Hershey Medical Center, Mail Code
HU14, 500 University Dr, Hershey, PA 17033. E-mail: aflamm@
pennstatehealth.psu.edu.
Published online July 12, 2019.
0190-9622/$36.00
Ó 2019 by the American Academy of Dermatology, Inc.
https://doi.org/10.1016/j.jaad.2019.07.011

156
J AM ACAD DERMATOL Flamm et al 157
VOLUME 82, NUMBER 1

We sought to determine which histopathologic included given the histologic findings and any pro-
features are statistically most helpful in distinguish- vided clinical history.
ing secondary syphilis from these common clinical In total, 100 cases of syphilis, 110 cases of PL, 72
and histologic mimickers. cases of PR, and 101 cases of early MF were included
in the study. Histologic features in each case were
tabulated by 2 dermatopathologists (Dr Elston and
METHODS Dr Flamm) and statistical
Specimen selection was evaluation was performed
performed by using elec- with Statistica TIBCO (Palo
tronic records from the CAPSULE SUMMARY Alto, CA). The x 2 test
Ackerman Academy of was used to compare differ-
Dermatopathology, SUNY d
Secondary syphilis can mimic pityriasis ences between the 4 groups
Downstate Medical Center, lichenoides, pityriasis rosea, and mycosis (Table I). Calculations of
Kings County Medical fungoides clinically and histologically. specificity, sensitivity, and
Center, Dermpath Diagno- likelihood ratios (LRs) were
d
Presence of interstitial inflammation can
stics New York, University of performed with online statis-
help distinguish syphilis from pityriasis
Pennsylvania, and Penn State tics calculator www.
lichenoides and pityriasis rosea but not
Hershey Medical Center. The vassarstats.net. A positive
mycosis fungoides. Plasma cells can be
study was determined to be LR [ 1 indicated an associa-
seen in both syphilis and pityriasis
exempt by the institutional tion of the feature with syph-
lichenoides and, therefore, is not a
review boards of these ilis. A positive LR [ 10
helpful distinguishing feature.
institutions. indicated the feature had a
Histopathologic features strong association, 5-10 a
of the syphilis cases were moderate association, and \5 a weak association.9
previously reported, and 6 cases were deleted from Absence of the feature, or a negative LR of \0.1, was
the current statistical evaluation as the specimens strongly suggestive against a syphilis diagnosis.10
were no longer available for review.2 All syphilis
cases were previously confirmed either with immu-
nohistochemistry or serologic studies.2 We selected RESULTS
PL, PR, and MF cases by searching the pathology Secondary syphilis features were statistically less
electronic medical record using a natural language frequently seen in other conditions (see Table I).
search for the diagnoses to be studied. To be Syphilis and PL were the most difficult to distinguish,
included, the case had to have a final diagnosis and there was no unique feature strongly associated
consistent with PL, PR, or MF, or 1 of these diagnoses with syphilis (Table I). Interstitial inflammation and
had to be strongly favored in the comment section. If elongated narrow or pointed rete ridges, which are
the diagnosis was strongly favored in the comment generally a thinner width than expected for rete
section, 2 authors had to agree the case could be ridges in that anatomic area, were noted ;3 times as

Table I. Histopathologic features of syphilis, pityriasis lichenoides, pityriasis rosea, and mycosis fungoides
Syphilis, Pityriasis Pityriasis Mycosis
n = 100, lichenoides, rosea, fungoides,
Features n (%) n = 110, n (%) n = 72, n (%) n = 101, n (%)
Neutrophils in the stratum corneum 29 (29) 60 (54.55) 1 (1.39)* 8 (7.92)
Presence of plasma cells 71 (71) 62 (56.36) 2 (2.78)* 7 (6.93)
Endothelial swelling 88 (88) 53 (48.18) 14 (19.44)* 12 (11.88)
Effacement of the rete ridges 20 (20) 66 (60.0) 6 (8.33)* 31 (30.69)
Irregular acanthosis 77 (77) 53 (48.18) 29 (40.28)* 34 (33.66)
Psoriasiform acanthosis 3 (3) 78 (70.91) 2 (2.78)* 3 (2.78)
Elongated rete ridges 75 (75) 24 (21.82) 29 (27.36)* 36 (36.36)
Vacuolar interface with vacuolar predominance 11 (11) 60 (54.55) 2 (2.78)* 4 (3.96)
Vacuolar interface with equal numbers of 73 (73) 54 (49.09) 1 (1.39)* 88 (87.13)
lymphocytes and vacuoles
Interstitial inflammation 94 (94) 38 (34.55) 16 (22.22)* 64 (63.37)
Lymphocytes with ample cytoplasm 57 (57.00) 58 (52.73) 2 (2.78)* 3 (4.95)

*P \ .0001 with x2 test.

158 Flamm et al J AM ACAD DERMATOL
JANUARY 2020

was noted in almost 75% of cases of syphilis, a

Abbreviations used:
LR: likelihood ratio
MF: mycosis fungoides
PL: pityriasis lichenoides
PR: pityriasis rosea
often in syphilis as opposed to PL. Interestingly,
although plasma cells are historically considered a
significant feature of syphilis, they were often
encountered in PL, although in smaller numbers
(Fig 1). Of note, almost 30% of the cases of syphilis
lacked plasma cells and could also lack endothelial
swelling (Fig 2).
Although PR and syphilis are similar clinically,
there were histologic differences. The most helpful
features, with a stronger association with syphilis
than PR, were neutrophils in the stratum corneum,
plasma cells, vacuolar interface with lymphocytes,
and lymphocytes with ample cytoplasm. Although
true vacuolar interface was rare in PR, some cases
had lymphocytosis in the areas of spongiosis near the
dermoepidermal junction that mimicked an interface
process.
Although all studied features were seen in some
MF cases, presence of plasma cells and lymphocytes
with ample cytoplasm were features mostly seen in
syphilis (Table II).
DISCUSSION
Although our data can be used to help distinguish
secondary syphilis from clinical and histologic mim-
ickers, our data also indicated that distinguishing
syphilis from PL histologically might be difficult.11 An
interface pattern with a lymphocyte in most vacuoles
feature that overlaps with PL.2 Neutrophils in the
stratum corneum are common in both conditions.1
There is no single distinguishing feature that differ-
entiates reliably between these 2 conditions.
Presence of interstitial inflammation and elongated
rete ridges, however, stand out as the most important
features to suggest syphilis from PL. Of note,
although plasma cells are classically noted in sec-
ondary syphilis, they were also present in a substan-
tial number of PL cases.
The eruption of secondary syphilis frequently
presents on the trunk with a clinical differential
diagnosis of PR. In addition, cases of secondary
syphilis mimicking PR clinically tend to have fewer
syphilis diagnostic features on biopsy than other
presentations of secondary syphilis.2 In view of the
increasing incidence of syphilis, pathologists should
maintain a high degree of suspicion when encoun-
tering a truncal eruption with a differential diagnosis
including PR.
PR does have histologic features that can mimic
secondary syphilis. Specifically, exocytosis in cases
of PR can mimic interface dermatitis, and endothelial
swelling is noted in almost 20% of cases of PR, which
could raise concerns about the possibility of syphilis.
Silver and immunostaining have a significant rate of
false-negative results, and clinicians are often left to
decide how far to proceed with other diagnostic
testing to exclude the possibility of syphilis.12,13 Our
data suggests that histologic features that suggest
syphilis in this setting include neutrophils in the
stratum corneum, plasma cells, interface dermatitis
lymphocytes and vacuoles, and the presence of
lymphocytes with ample cytoplasm. These features
can be used reliably to differentiate syphilis from PR.

Fig 1. A, Interstitial infiltrate in pityriasis lichenoides. B, Endothelial swelling and small

collections of plasma cells in the inflammatory infiltrate. (A and B, Hematoxylin-eosin stain;
original magnification: A, x100; B, x400.)
J AM ACAD DERMATOL Flamm et al 159
VOLUME 82, NUMBER 1

Fig 2. Irregular acanthosis in syphilis. (Hematoxylin-eosin stain; original magnification: x100.)

Table II. Sensitivity, specificity, and LRs of syphilis features compared with pityriasis lichenoides, pityriasis
rosea, and mycosis fungoides
Condition and feature Sensitivity Specificity LR (1) LR (e) P value
Pityriasis lichenoides
Neutrophils in the stratum corneum 0.29 0.46 0.53 1.56 \.05*
Presence of plasma cells 0.71 0.44 1.26 0.66 \.05*
Endothelial swelling 0.88 0.52 1.82 0.23 \.05*
Effacement of the rete ridges 0.20 0.4 0.33 2 \.05*
Irregular acanthosis 0.77 0.52 1.60 0.44 \.05*
Psoriasiform acanthosis 0.03 0.29 0.04 3.33 \.05*
Elongated rete ridges 0.75 0.78 3.44 0.32 \.05*
Vacuolar interface with vacuolar predominance 0.11 0.45 0.20 1.96 \.05*
Vacuolar interface with equal numbers of lymphocytes and vacuoles 0.73 0.51 1.49 0.53 \.05*
Interstitial inflammation 0.94 0.65 2.72 0.09* \.05*
Lymphocytes with ample cytoplasm 0.57 0.47 1.08 0.90 .53
Pityriasis rosea
Neutrophils in the stratum corneum 0.29 0.97 20.89* 0.72 \.05*
Presence of plasma cells 0.71 0.97 25.56* 0.30 \.05*
Endothelial swelling 0.88 0.81 4.53 0.15 \.05*
Effacement of the rete ridges 0.20 0.92 2.4 0.88 \.05*
Irregular acanthosis 0.77 0.60 1.91 0.38 \.05*
Psoriasiform acanthosis 0.03 0.97 1.08 1.00 .65
Elongated rete ridges 0.75 0.60 1.86 0.42 \.05*
Vacuolar interface with vacuolar predominance 0.11 0.97 3.96 0.92 \.05*
Vacuolar interface with equal numbers of lymphocytes and vacuoles 0.73 0.99 52.56* 0.27 \.05*
Interstitial inflammation 0.94 0.78 4.23 0.08* \.05*
Lymphocytes with ample cytoplasm 0.57 0.97 20.52* 0.44 \.05*
Mycosis fungoides
Neutrophils in the stratum corneum 0.29 0.92 3.66 0.77 \.05*
Presence of plasma cells 0.71 0.93 10.24* 0.31 \.05*
Endothelial swelling 0.88 0.89 7.41 0.14 \.05*
Effacement of the rete ridges 0.20 0.69 0.65 1.15 .09
Irregular acanthosis 0.77 0.66 2.29 0.34 \.05*
Psoriasiform acanthosis 0.03 0.97 1.01 1.00 .9
Elongated rete ridges 0.75 0.64 2.10 0.39 \.05*
Vacuolar interface with vacuolar predominance 0.11 0.96 2.78 0.93 .06
Vacuolar interface with equal numbers of lymphocytes and vacuoles 0.73 0.13 0.84 2.10 \.05*
Interstitial inflammation 0.94 0.37 1.48 0.16 \.05*
Lymphocytes with ample cytoplasm 0.57 0.97 19.19* 0.44 \.05*

LR, Likelihood ratio.

*Statistically significant.
160 Flamm et al
Lymphocyte atypia might be seen in syphilis, and
a vacuolar-interface pattern with a lymphocyte in
every vacuole is common in both syphilis and MF.
Papillary dermal fibrosis might not be prominent in
cases of MF, creating a diagnostic dilemma. Our data
additionally shows that interstitial inflammation is
noted in 64% of cases of MF and endothelial
swelling in 12%, further complicating the differen-
tial diagnosis. Our data suggest that the most helpful
features to distinguish syphilis from MF are the
presence of lymphocytes with ample cytoplasm and
the presence of plasma cells. Of note, lymphocytes
with ample cytoplasm have normal-appearing
nuclei, which distinguishes this entity from true
lymphocyte atypia. In addition, these lymphocytes
retain their nuclear characteristics, which distin-
guished them from histiocytes, which have paler
vesicular nuclei with less dense chromatin.
Immunohistochemical stains can also be used to
distinguish lymphocytes with ample cytoplasm
from histiocytes and confirm that the cells in
question are lymphocytes. Presence of plasma cells
has previously been mentioned as a helpful clue to
differentiate syphilis from MF.6,7 While plasma cells
might be present in later stages on MF, they are
rarely encountered in early MF.
Although most of the features studied had
P values \ .05, a P value is simply the probability
of a similar or more extreme result if the null
hypothesis is true and does not address the magni-
tude of a difference. To interpret our data, it is
important to distinguish statistical significance from
medical relevance. LRs are a better measure of the
clinical relevance of our results.
In sum, while many of the histologic features
mentioned can be used to distinguish syphilis from
PR and MF, no feature is reliable in distinguishing
syphilis from PL. Specifically, although plasma cells
are classically associated with syphilis, they are also
found in many cases of PL. Additionally, syphilis is
likely to demonstrate a vacuolar interface dermatitis
J AM ACAD DERMATOL
JANUARY 2020
with a lymphocyte in every vacuole, a pattern also
strongly associated with PL. Therefore, correlation
with clinical morphology, serology, and temporality,
as well as ancillary immunostaining or silver staining,
should be considered in cases where these findings
are encountered histologically.
We thank Kruti Parikh and Quian Xie for helping
analyze previously published data.
REFERENCES
1. Jeerapaet P, Ackerman AB. Histologic patterns of secondary
syphilis. Arch Dermatol. 1973;107(3):373-377.
2. Flamm A, Parikh K, Xie Q, Kwon EJ, Elston DM. Histologic
features of secondary syphilis: a multicenter retrospective
review. J Am Acad Dermatol. 2015;73(6):1025-1030.
3. Alessi E, Innocenti M, Ragusa G. Secondary syphilis. Clinical
morphology and histopathology. Am J Dermatopathol. 1983;
5(1):11-17.
4. Bowers S, Warshaw EM. Pityriasis lichenoides and its subtypes.
J Am Acad Dermatol. 2006;55(4):557-572; quiz 73-6.
5. Levin DL, Greenberg MH, Hasegawa J, Roenigk HH Jr. Sec-
ondary syphilis mimicking mycosis fungoides. J Am Acad
Dermatol. 1980;3(1):92-94.
6. McComb ME, Telang GH, Vonderheid EC. Secondary syphilis
presenting as pseudolymphoma of the skin. J Am Acad
Dermatol. 2003;49(2 Suppl Case Reports):S174-S176.
7. Reddy K, Bhawan J. Histologic mimickers of mycosis fun-
goides: a review. J Cutan Pathol. 2007;34(7):519-525.
8. Guitart J, Peduto M, Caro WA, Roenigk HH. Lichenoid changes
in mycosis fungoides. J Am Acad Dermatol. 1997;36(3 Pt 1):
417-422.
9. McGee S. Simplifying likelihood ratios. J Gen Intern Med. 2002;
17(8):647-650.
10. Sonis J. How to use and interpret interval likelihood ratios.
Fam Med. 1999;31(6):432-437.
11. Lee EH, Lee SE, Kim DH, Yoon MS, Kim SC. Demonstration of
Treponema pallidum by immunohistochemistry in a patient
with secondary syphilis mimicking pityriasis lichenoides
chronica. J Dermatol. 2012;39(6):567-568.
12. Buffet M, Grange PA, Gerhardt P, et al. Diagnosing Treponema
pallidum in secondary syphilis by PCR and immunohistochem-
istry. J Invest Dermatol. 2007;127(10):2345-2350.
13. Wenhai L, Jianzhong Z, Cao Y. Detection of Treponema
pallidum in skin lesions of secondary syphilis and character-
ization of the inflammatory infiltrate. Dermatology. 2004;
208(2):94-97.