Round Cell Pattern of Prostatic Stromal Tumor of Uncertain Malignant Potential: A Subtle Newly Recognized Variant Evita T. Sadimin MD, Jonathan I. Epstein MD PII: DOI: Reference:
S0046-8177(16)00034-4 doi: 10.1016/j.humpath.2016.01.002 YHUPA 3796
To appear in:
Human Pathology
Received date: Revised date: Accepted date:
6 November 2015 28 December 2015 6 January 2016
Please cite this article as: Sadimin Evita T., Epstein Jonathan I., Round Cell Pattern of Prostatic Stromal Tumor of Uncertain Malignant Potential: A Subtle Newly Recognized Variant, Human Pathology (2016), doi: 10.1016/j.humpath.2016.01.002
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT 1
Round Cell Pattern of Prostatic Stromal Tumor of Uncertain Malignant Potential:
1
RI P
T
A Subtle Newly Recognized Variant
Evita T. Sadimin, MD
1,2,3
SC
Jonathan I. Epstein, MD
Department of Pathology1, Urology2, and Oncology3
MA
NU
The Johns Hopkins Medical Institutions, Baltimore, MD
Running title: Round cell prostatic stromal tumor of uncertain malignant potential
Corresponding author:
CE
Jonathan I. Epstein, MD
PT
ED
Keywords: prostate stromal tumor; STUMP; round cell pattern
AC
Department of Pathology
The Johns Hopkins Hospital 401 N Broadway St, Weinberg Rm 2242 Baltimore, MD 21231 Phone : (410) 955-5043 Fax: (443) 287-3818 e-mail:
[email protected]
Conflicts of interest: The authors declare no conflict of interest associated with this work. Funding disclosures: The authors have no relevant financial relationships to disclose.
ACCEPTED MANUSCRIPT 2
ABSTRACT
RI P
T
Prostatic stromal tumor of uncertain malignant potential (STUMP) is a distinct entity which include several different patterns. Four patterns of STUMP have been described including
SC
stroma with: 1) degenerative atypia; 2) hypercellular spindle cells; 3) myxoid spindle cells; and 4) phyllodes-like pattern. The current study identified a novel round cell pattern. We searched
NU
our database from 1999 to 2015 and identified 7 patients with round cell pattern out of a total
MA
number of 98 patients with STUMP. All 7 cases showed mildly increased stromal cellularity with rounded nuclei, diagnosed on core biopsies in five cases, transurethral resection in one case
ED
and radical prostatectomy in one case. Some degree of glandular displacement was observed in four cases. In two of the cases, STUMP was not recognized histologically by the referring
PT
pathologists and was initially diagnosed as benign prostatic hyperplasia (BPH). As has been
CE
described with other patterns of STUMP, several cases showed associated epithelial proliferations that in some instances masked the neoplastic stromal process. The round cell
AC
pattern of STUMP is a new deceptively subtle pattern that may not be recognized as a neoplasm and may be misdiagnosed as BPH. While there was no direct evidence in our study that the round cell pattern of STUMP has the same behavior as other variants of STUMPs, increased recognition of this entity will hopefully lead to additional studies to further understand its malignant potential.
ACCEPTED MANUSCRIPT 3
INTRODUCTION Prior to the introduction of the term STUMP, prostatic stromal tumors have been referred
RI P
T
to in the literature as various other terms including atypical stromal hyperplasia, atypical stromal smooth muscle hyperplasia, cystosarcoma phyllodes, phyllodes tumor, and cystic epithelial-
SC
stromal tumors. Some of these terms indicate controversy whether these lesions are neoplastic or variants of stromal hyperplasia. Since the introduction of STUMP in 1998 [1], we have acquired
NU
a better understanding of the tumor as a distinct entity, supporting its neoplastic nature based on
MA
its behavior, association with stromal sarcoma, and molecular evidence that they have recurrent chromosomal aberrations [2]. Well-recognized patterns of STUMP include: 1) hypercellular
ED
stroma with degenerative atypia; 2) hypercellular spindle cells; 3) spindle cells in myxoid background; and 4) phyllodes-like [3]. In this study we introduce a fifth subtle pattern, the round
MATERIALS AND METHODS
AC
CE
PT
cell pattern.
Ninety eight cases of STUMP from 1999 to 2015 were retrieved from our surgical pathology database at the Johns Hopkins Hospital. Eighty seven of these cases were consult cases and eleven were in-house cases. Seven patients were identified to have a round cell pattern. The age, clinical presentation, PSA value and treatments the patients received were reviewed in each case. All procedures were performed in compliance with relevant laws and institutional guidelines, and has been approved by the appropriate institutional committee.
ACCEPTED MANUSCRIPT 4
RESULTS
RI P
T
Of all the cases of STUMP with a round cell pattern, six were sent in as consults from other institutions and one was an in-house case. Patients ranged in age from 26 to 70 years at the
SC
time of presentation, with the majority being older than 50 years old (Table 1). Presenting symptoms included urinary obstructive symptoms, hematuria, elevated serum PSA levels and
NU
abnormal digital rectal examination (DRE). Serum PSA levels were increased (4 ng/ml) in 6 of
MA
the 7 cases. In five cases the lesions were diagnosed on core biopsies, in one case it was diagnosed on transurethral resection, and in one other case on radical prostatectomy. The extent
ED
of the disease varied; in four cases they were solitary nodules, in two cases affecting one side of
PT
the prostate, and in one case affecting the entire prostate.
The patient in case 1 underwent complete surgical resection of the nodule by
CE
transurethral resection with subsequent biopsies showing no evidence of recurrence 2.5 years
AC
after surgery. In case 5, radical prostatectomy was performed and the patient has not shown any evidence of disease at 2 years. He was also incidentally found to have a small focus of organconfined prostatic adenocarcinoma, Gleason score 3+3=6.
In two of the cases (cases 3 & 6), they were initially diagnosed as benign prostatic hyperplasia (BPH) on biopsies. In both cases, the patients subsequently underwent repeat biopsies after they were clinically treated without any improvement. The repeat biopsies were sent to our institution by the pathologists for consultations on the stromal proliferations. In both cases, the initial and repeat biopsies showed the same stromal lesion.
ACCEPTED MANUSCRIPT 5
The other two patients (cases 4 and 7) elected for surveillance, and did not show disease progression, though the obstructive symptoms persisted, both requiring catheterization. In case 4
T
the decision was made by personal preference of the patient, who continued to be closely
RI P
monitored by imaging. In case 7, the patient was being managed for his metastatic poorly differentiated squamous cell carcinoma of the esophagus at the time he was diagnosed with
SC
STUMP. He died two years after his STUMP diagnosis from his squamous cell carcinoma.
NU
Morphologically, all cases showed increased stromal cellularity where in contrast to
MA
normal prostate stroma cells or other patterns of STUMP, the vast majority of the nuclei were round. The nuclei lacked hyperchromasia or mitotic figures (figures 1A,B). In one case, the
ED
stromal lesion affecting one side of his prostate was very subtle and was associated with normal glands (figures 1C,D). The lesion was more evident when it was compared to his normal stroma
PT
from the contralateral side (figure 1E). In two cases the glands were significantly displaced
CE
(figure 1F). In all cases, with the exception of one case described below with very focal coexisting degenerative atypia, the lesions were entirely composed of stroma with a round cell
AC
pattern without spindle cells.
The patient in case 2 originally presented with palpable nodule on DRE. MRI at that time showed a distinct 2 cm nodule. Over the next 9 years, the patient underwent five biopsies. The former four biopsies done at the outside institutions were reported as benign. The patient was then followed by our institution. By this time, MRI showed that the nodule had progressed into a distinct 4.2 cm nodule with irregular borders closely approaching the posterior edge of the prostate suspicious for extraprostatic extension. A biopsy was performed at our institution and was reported by one of our general pathologists with a focus of prostatic adenocarcinoma,
ACCEPTED MANUSCRIPT 6
Gleason 3+3=6 being the main findings. However, the prostatic adenocarcinoma was not at the
T
area of the nodule.
RI P
Subsequently, the patient underwent radical prostatectomy. On cross section, the nodule was very evident, affecting virtually an entire side of the prostate (figure 2A), yet was still
SC
circumscribed. Histological examination of the nodule showed marked glandular hyperplasia with infolding and basal cell hyperplasia surrounded by subtle stroma with round nuclei (figure
NU
2B,C) consistent with round cell STUMP. A less subtle degenerative atypia pattern of STUMP
MA
was also present, however only seen very focally on 1 slide (figure 2D). Based on this finding, the last biopsy was re-reviewed by the senior author, which revealed that the round cell STUMP
ED
was indeed present in two out of the twelve cores. In those cores, the affected area could be appreciated under low magnification (right side of figure 2E) when compared with the normal
PT
area (left side of figure 2E). Under higher magnification, the unique stroma with round nuclei
CE
was more evident yet still subtle (figure 2F).
AC
The radical prostatectomy in case 5 showed a distinct well-circumscribed nodule with marked hyperplastic glands (figure 3A). There was an abrupt transition between the neoplastic cellular round cell stroma and the surrounding hypocellular spindled non-neoplastic stroma (figure 3B). Under higher magnification, the bland round cell hypercellular stroma intimately surrounded the hyperplastic glands (figures 3C,D). Some of the hyperplastic glands had prominent basal cells (figure 3E). In areas, the markedly hyperplastic glands obscured the stroma (figure 3F).
ACCEPTED MANUSCRIPT 7
DISCUSSION The current study described a new deceptively subtle pattern of STUMP that may not be
RI P
T
recognized as a lesion or misdiagnosed as BPH, especially on core biopsies when the mass effect may not be obvious. Stromal nodular hyperplasia is the most common “soft tissue lesion” seen in
SC
our consult material where the concern is to rule out STUMP. Its increased frequency in part reflects that BPH is one of the most common disorders in men over 50 years of age and that
NU
prostate biopsies are frequently performed to assess an elevated serum PSA level. Stromal
MA
nodular BPH may also be extensive either on TURP, enucleation, or on needle biopsy suggestive of a stromal neoplasm. In contrast to round cell STUMP, the stroma in BPH may be
ED
normocellular or slightly hypercellular yet is spindled. An additional characteristic finding in prostatic stromal nodules of BPH, not seen in STUMPs, is that the vessels within the nodule
PT
show thickened, hyalinized walls out of proportion to the size of the vessels (figures 4A, B).
CE
Another feature of stromal nodular BPH that is difficult to appreciate on needle biopsy is its multifocal nodularity. STUMPs may grow as a single large nodule but within the nodule the
AC
stromal proliferation is diffuse.
The round cell pattern differs from the other previously described patterns of STUMP [4], lacking degenerative cytologically atypical cells, myxoid background or phyllodes-like growth pattern. Compared to the hypercellular pattern of STUMP, the round cell pattern shows plump round cells instead of elongated spindle cells and lacks the distinctive eosinophilic cytoplasm of the hypercellular spindled pattern. A common feature to all STUMPs, including our round cell cases, is the associated epithelial proliferation [5]. Some of the more commonly seen changes are glandular crowding, a prominent basal cell layer, and prominent papillary infolding. In cases where the epithelial proliferation obscures the stroma, diagnosis on core biopsies may be
ACCEPTED MANUSCRIPT 8
difficult. Therefore it is important to evaluate the stroma very closely when there is complex glandular proliferation on core biopsies.
RI P
T
The PSA values of the patients in our cases did not show any correlation with the extent of disease. Other than the presence of a small focus of prostatic adenocarcinoma, Gleason score
SC
3+3=6 in case 5, the other cases did not have other histological findings such as inflammation or BPH to account for the increased PSA value. Limited clinical information precluded definite
MA
findings could have contributed to the value.
NU
correlation between the round cell STUMP and the PSA values since other non-histological
In our experience, all the other types of STUMP have been shown to have the potential to
ED
transform into a sarcoma [6]. When STUMP is diagnosed on biopsy or TURP, where only representative parts of the stromal lesion are present, radical resection may be recommended for
PT
removal of the remaining lesion. There can be discordance between STUMPs on biopsy where
CE
there is the potential for concurrent or subsequent sarcoma and on complete resection where the
AC
lesion is found to be entirely benign. This is an inherent problem with the aspect of "uncertain malignant potential" in the name "STUMP". Nonetheless, "STUMP" has been in use for many years and is well-entrenched in the literature. In order to account for the differences between biopsy and resection, a note should be added to the report that when STUMP is completely removed by radical prostatectomy the "uncertain malignant potential" aspect of the diagnosis no longer applies and that the lesion is cured as long as no sarcoma component is present. The transformation from STUMP to sarcoma is characterized morphologically by significant hypercellularity, atypia, mitotic activity and necrosis, and clinically with the potential to recur and metastasize [1,7]. We have not seen similar transformation in this series, though the
ACCEPTED MANUSCRIPT 9
study had short clinical follow-up of limited number of cases. Description of this entity will allow pathologists to recognize additional cases so that larger series can be carried out in the
T
future. Until more information is available, close follow up is warranted and possible
RI P
consideration should be given to radical resection depending on factors such as to whether there
SC
is a mass present on digital rectal examination or on imaging and the age of the patient. Of the various patterns of STUMP, both the newly described round cell variant and the
NU
cellular spindled pattern most closely resemble BPH stroma. Pathologists are more likely to
MA
underdiagnose these two variants of STUMP as BPH, or be concerned that BPH stromal nodules represent these subtle patterns of STUMP. Some pathologists may question whether the lesion
ED
depicted herein even is a neoplasm, as opposed to merely stromal hyperplasia. The same doubts were initially expressed for the other patterns of STUMP, although now they are definitively
PT
accepted as neoplastic [6]. Evidence that the pattern described herein is not just a peculiar pattern
CE
of BPH is: 1) formation of discrete nodules or involvement of only one side of the prostate without co-existing diffuse nodular BPH; 2) in one case round cell and more classic degenerative
AC
atypia patterns of STUMP co-existing in the same nodule; 3) some cases having the typical finding of florid epithelial proliferations localized to the STUMP; 4) we have never seen this round cell pattern in usual bilateral diffuse BPH; and 5) there are too many rounded stromal nuclei to merely represent a tangential section of spindled BPH stroma. While it may be difficult to establish a diagnosis of this pattern of STUMP on needle biopsy, the finding of cellular prostatic stroma composed of cells with round nuclei, especially on multiple cores, should be correlated with imaging studies to determine if there is a distinct mass that is inconsistent with BPH. If so, then either a more definitive diagnosis of STUMP should be rendered or the case be sent off for a second opinion.
ACCEPTED MANUSCRIPT 10
No immunohistochemical stains were performed as the immunoreactivity of all STUMPs is the same as non-neoplastic normal or hyperplastic prostatic stroma, such that
T
immunohistochemistry is not useful in this differential diagnosis. Molecular studies may help
RI P
resolved difficult cases as we have previously shown that specialized stromal tumors of the prostate, including STUMPs and stromal sarcomas have common chromosomal aberrations,
SC
most commonly loss of chromosomes 10, 13, and 14 [8]. We are in the process of an expanded
NU
comparative genomic hybridization study of these tumors, including the pattern of STUMP described within the current study. Increased recognition of this entity will hopefully lead to
AC
CE
PT
ED
MA
additional studies on this new pattern of STUMP to assess its malignant potential.
ACCEPTED MANUSCRIPT 11
LEGENDS
T
Figure 1. Case 3. Round cell STUMP, composed of hypercellular stroma with rounded cells
RI P
partially displacing other glands (A, H&E 10x; B, H&E 20x). Case 4. A subtle case where the round cell STUMP surrounded normal non-displaced glands (C, H&E 20x; D, H&E 40x). The
SC
lesion is more evident when compared to other uninvolved areas of the same case (E, H&E 20x).
NU
Case 7. STUMP where the glands on a core were completely displaced by the cellular round cell
MA
stroma. As in many STUMPs, the cytoplasm is very eosinophilic (F, H&E left 10x, right 40x).
Figure 2. Case 2. Cross section of the radical prostatectomy showing a distinct nodule affecting
ED
virtually an entire side of the prostate (A). STUMP more cellular round cell stroma (right) compared to uninvolved stroma (lower left) (B, H&E 2x). Higher magnification with more round
PT
cell cellular stroma in between crowded benign prostate glands (C, H&E 10x). Focal area with
CE
degenerative stromal atypia (D, H&E 40x). Needle biopsy prior to radical prostatectomy shows increased cellularity (right) contrasted to uninvolved stroma (left) (E, H&E 20x). Higher
20x).
AC
magnification with subtle increase of stroma with round cells consistent with STUMP (F, H&E
Figure 3. Case 5. Radical prostatectomy showing a distinct nodule with marked hyperplastic glands (A, H&E 1x). There is an abrupt transition between the neoplastic cellular round cell stroma and the surrounding hypocellular spindled non-neoplastic stroma (B, H&E 2x). Under higher magnification, the stroma is very hypercellular (C, H&E 10x). Individual stromal cells have rounded nuclei (D, H&E 20x). Some of the hyperplastic glands have prominent basal cells (E, H&E 10x). In areas, the markedly hyperplastic glands obscure the neoplastic stroma (F, H&E 10x).
ACCEPTED MANUSCRIPT 12
Figure 4. A case not from this study where the stroma in BPH is slightly hypercellular, with spindled cells, and associated with small capillaries with relatively thick muscle walls (A, H&E
AC
CE
PT
ED
MA
NU
SC
RI P
T
10x; B, H&E 20x).
ACCEPTED MANUSCRIPT 13
REFERENCES
Gaudin PB, Juan R, Epstein JI. Sarcomas and related proliferative lesions of specialized prostatic stroma: A clinicopathologic study of 22 cases. Am. J. Surg. Pathol. 1998;22:148–162.
2.
Pan C-C, Epstein JI. Common chromosomal aberrations detected by array comparative genomic hybridization in specialized stromal tumors of the prostate. Mod. Pathol. 2013;26:1536–43.
3.
Hansel DE, Herawi M, Montgomery E, et al. Spindle cell lesions of the adult prostate. Mod. Pathol. 2007;20:148–58.
4.
Herawi M, Epstein JI. Specialized stromal tumors of the prostate: A clinicopathologic study of 50 cases. Am. J. Surg. Pathol. 2006;30:694–704.
5.
Nagar M, Epstein JI. Epithelial proliferations in prostatic stromal tumors of uncertain malignant potential (STUMP). Am. J. Surg. Pathol. 2011;35:898–903.
6.
Hossain D, Meiers I, Qian J, et al. Prostatic stromal hyperplasia with atypia: follow-up study of 18 cases. Arch Pathol Lab Med 2008;132:1729–33.
7.
Bostwick DG, Hossain D, Qian J, et al. Phyllodes tumor of the prostate: long-term followup study of 23 cases. J. Urol. 2004;172:894–9.
8.
Pan CC, Epstein JI. Common chromosomal aberrations detected by array comparatic genomic hybridization in specialized stromal tumors of the prostate. Mod Pathol. 2013: 26: 1536-43.
AC
CE
PT
ED
MA
NU
SC
RI P
T
1.
ACCEPTED MANUSCRIPT
AC
CE
PT
ED
MA
NU
SC
RI P
T
14
Figure 1
ACCEPTED MANUSCRIPT
AC
CE
PT
ED
MA
NU
SC
RI P
T
15
Figure 2
ACCEPTED MANUSCRIPT
AC
CE
PT
ED
MA
NU
SC
RI P
T
16
Figure 3
ACCEPTED MANUSCRIPT
NU
SC
RI P
T
17
AC
CE
PT
ED
MA
Figure 4
ACCEPTED MANUSCRIPT 18
51
4
58
elevated PSA
6
62
elevated PSA
7
69
obstructive symptoms
PT
5
1.5 cm nodule 4 cm nodule entire left side entire right side 6.5 cm nodule 2 cm nodule entire prostate
ED
3
nodule palpated on DRE obstructive 51 symptoms obstructive 58 symptoms
0.9 ng/ml 4.1 ng/ml 7.0 ng/ml 4.7 ng/ml 13.1 ng/ml 5.8 ng/ml 16.1 ng/ml
Treatment
RI P
2
hematuria
Extent
transurethral resection of entire nodule
Follow up 2.5 years
SC
26
PSA
radical prostatectomy
N/A
BPH treatment, surveillance
1 year
NU
1
Presentation
MA
Case Age
T
TABLE 1. Summary of clinical findings
surveillance
1.5 years
radical prostatectomy
2 years
BPH treatment, surveillance
4.5 years
surveillance
2 years
AC
CE
DRE – digital rectal examination; PSA – prostate specific antigen; BPH – benign prostatic hyperplasia; N/A – not available