- Email: [email protected]
An ‘inflammatory’ variant of solar purpura: a simulant of leukocytoclastic vasculitis and neutrophilic dermatoses
Pathology (August 2013) 45(5), pp. 484–488
ANATOMICAL PATHOLOGY
An ‘inflammatory’ variant of solar purpura: a simulant of leukocytoclastic vasculitis and neutrophilic dermatoses BENJAMIN A. WOOD*{
AND
PHILIP E. LEBOIT{§
*PathWest Laboratory Medicine, QEII Medical Center and Sir Charles Gairdner Hospital, {School of Pathology and Laboratory Medicine, the University of Western Australia, Perth, Western Australia, Australia; Departments of zPathology, and §Dermatology, University of California, San Francisco, CA, United States
Summary Aims: To study the clinical and pathological features of cases of apparent solar purpura, with attention to the recently described phenomenon of inflammatory changes within otherwise typical lesions. Methods: We studied 95 cases diagnosed as solar purpura and identified 10 cases (10.5%) in which significant neutrophilic inflammation was present, potentially simulating a leukocytoclastic vasculitis or neutrophilic dermatosis. An additional three cases were identified in subsequent routine practice. The clinical features, including follow-up for subsequent development of vasculitis and histological features were studied. Results: In all cases the histological features were typical of solar purpura, with the exception of inflammatory changes, typically associated with clefting of elastotic stroma. Clinical follow-up information was available for all patients and none developed subsequent evidence of a cutaneous or systemic vasculitis or neutrophilic dermatosis. Conclusions: Inflammatory changes appear to be more frequent in solar purpura than is generally recognised. Awareness of this histological variation and correlation with the clinical findings and evolution is important in avoiding misdiagnosis. Key words: Inflammation, neutrophilic dermatosis, pseudovasculitis, secondary vasculitis, solar purpura. Received 15 November 2012, revised 17 January, accepted 29 January 2013
The pathophysiology of SP is thought to be related to a combination of minor (often unrecognised) trauma, combined with impaired tissue support and elasticity in skin with severe solar elastosis, leading to extravasation of blood from dermal vessels.1,4–6 In some cases additive predisposing factors such as the use of corticosteroids, anticoagulant or antiplatelet medications are identifiable.6 Although the majority of cases of SP are recognised as such clinically, biopsy specimens are commonly obtained by clinicians who practice in areas with fair skinned patients and sunny climes. The reasons for submission of these biopsies vary, but include concern for a melanocytic or vascular tumour and clinical consideration of the possibility of an alternative cause for purpura, such as vasculitis. Significant neutrophilic inflammation is generally not considered a feature of these lesions. In standard textbooks of dermatopathology, SP is classified as a ‘non-inflammatory’ type of purpura, and an inflammatory infiltrate is not described,1,4,5 or is specifically excluded.6 The presence of such an infiltrate naturally leads the pathologist to consider alternative causes of purpura, particularly vasculitis. In 2007 Sung and Jacobson presented in abstract form a series of 13 cases described as ‘solar purpura-related pseudovasculitis’.7 Following this description, we noted a number of cases of otherwise clinically and histologically typical solar purpura showing neutrophilic infiltrates. To date there has been no detailed publication exploring the incidence and features of this phenomenon.
INTRODUCTION
METHODS
Extravasation of red blood cells into the dermis correlates with the clinical finding of purpura. Its multiple causes include purpura due to inflammatory damage to vessel walls (e.g., in vasculitis), vascular leakage due to luminal occlusion by thrombi or cryoproteins, and purpura due to altered coagulation.1 So-called ‘senile’ purpura is a common condition, typically developing on the forearms of elderly patients with severely sun-damaged skin. There is a trend to eliminate use of the perjorative adjective ‘senile’, and the condition is also called ‘actinic purpura’, ‘solar purpura’ and ‘Bateman’s purpura’.2 Somewhat confusingly, the term solar purpura has also been used to describe a different spectrum of diseases characterised by rapidly developing purpuric lesions in response to acute solar irradiation,3 which are unrelated to the phenomenon of senile/solar purpura (SP) as used herein.
The files of the University of California San Francisco Dermatopathology Service (UCSF) were searched for all cases with a retrieval code of ‘solar purpura’ and ‘solar purpura-neutrophil rich’ between January 2009 and September 2011. The term ‘retrieval code’ designates that the pathologist who reviewed the sections referenced this term in the computer system for potential case retrieval, and does not necessarily indicate that the diagnosis on the report read as such. The slides of all cases were reviewed, and cases showing evidence of significant neutrophilic infiltration were selected for study. Significant neutrophilic inflammation was defined by the presence of a perivascular neutrophilic infiltrate, with or without evidence of vascular mural fibrin deposition, and/or a non-focal interstitial neutrophilic infiltrate. Cases showing trivial features such as neutrophil margination within vessel lumens or very focal entrapment of neutrophils in areas of fibrin deposition were not selected. A total of 106 cases of solar purpura were retrieved for review from the UCSF files. Eleven cases were excluded, due either to significant co-existing pathology or an absence of clear histological features of SP on the slides
Print ISSN 0031-3025/Online ISSN 1465-3931 DOI: 10.1097/PAT.0b013e3283632649
#
2013 Royal College of Pathologists of Australasia
Copyright © Royal College of pathologists of Australasia. Unauthorized reproduction of this article is prohibited.
‘INFLAMMATORY’ VARIANT OF SOLAR PURPURA
available. Of the remaining 95 cases, 10 lesions (10.5%) showed a significant neutrophilic inflammatory infiltrate. An additional three cases were prospectively gathered from the routine practice of one of the authors (BAW) over the period from November 2011 to October 2012, giving a total of 13 cases. All available haematoxylin and eosin (H&E) stained sections that had been prepared by standard techniques were reviewed, comprising two slides for Cases 1 and 13 and one slide each for the remaining 11 lesions. The cases were reviewed at a mutiheaded microscope by both authors (Cases 1–10) or by one author microscopically and shared via digital imaging (Cases 11–13). Five features were assessed as being absent or present, and in the latter circumstance graded in a semi-quantitative fashion on a three-point scale, with 1 representing a minor inconspicuous feature and 3 representing a marked or prominent finding. These features were: perivascular inflammation; interstitial inflammation; leukocytoclastic debris; tissue clefts; vascular mural fibrin deposition. Clinical data were retrieved from the pathology files and clinical follow-up information was sought by telephone contact with the referring clinician. Specifically, detailed information regarding the clinical presentation and clinical evolution subsequent to biopsy was sought, with particular reference to excluding development of a cutaneous or systemic vasculitis. Additionally, a medication history was obtained with regard to synchronous use of corticosteroids, anti-coagulant or anti-platelet medications.
RESULTS The 13 lesions studied occurred in 12 patients, ranging in age from 57 to 76 years (mean 64 years), including nine females and three males (Table 1). Six lesions were present on the forearms, while the hands/wrists were the site of three lesions, the upper arms involved two cases and the lower leg involved two cases. One patient with disease on the lower extremity (Case 7) also had small numbers of lesions on the forearms which were considered clinically to represent typical SP and were not biopsied. The lesions were variously described as haemorrhagic, violaceous, pigmented or purpuric lesions, in most cases with the appearance of macules 10–20 mm in maximum dimension (Fig. 1). A single, larger 50 mm patch and smaller 5 mm papule were each described. Two lesions were described as irregular pigmented macules, with clinical concern for melanoma. In no case was there clinical evidence of bulla formation. Table 1 Lesion
485
Detailed information regarding the presence of other lesions was available for all 12 patients from the referring clinician. In eight patients the lesion biopsied was solitary (in two of these patients a history of a solitary lesion recurrent at a single site was elicited). In the other cases, multiple lesions of similar clinical morphology were present on the arms and forearms. No patient other than Cases 7 and 12 (in whom leg lesions were biopsied) had purpuric lesions on the legs. One patient (Case 3) was taking warfarin, one patient (Case 6) was taking acetylsalicylic acid and one patient (Case 12) was taking aspirin and clopidogrel when the lesions developed. The other patients did not have a history of anticoagulant/antiplatelet medication or systemic corticosteroid use. Clinical follow-up information was available for all cases, with duration of follow-up ranging from 7 days to 30 months (median 3 months). All seven patients with solitary lesions were without similar lesions on follow-up, either as a result of complete clinical removal by biopsy (5 cases) or resolution of clinical residuum of the biopsied lesion, the latter occurring without further treatment. In one patient with bilateral single forearm lesions (Case 1), both the residuum of the lesion biopsied and a contralateral clinically similar lesion resolved with the use of topical cortisone. The latter treatment was instituted by the treating clinician on an empirical basis due to a pathological diagnosis of ‘solar purpura with a neutrophil rich infiltrate’. None of the patients developed evidence of a cutaneous or systemic vasculitis during follow-up, although evaluation of the latter possibility was limited to clinical history and examination in most cases. The histological features are summarised in Table 2 and illustrated in Fig. 2–6. By definition, the sections from all cases showed severe solar elastosis and evidence of dermal haemorrhage. In all cases selected for this study, dermal haemorrhage was evidenced by extravasation of red blood cells. No cases were identified which showed resolving haemorrhage (sparse lymphohistiocytic infiltrate associated with haemosiderin deposition) with concomitant neutrophilic inflammation. All cases showed both perivascular and
Clinical features and follow-up details of patients with SP showing neutrophilic inflammation Age/Gender
Location
Clinical morphology
Other clinical features Bilateral forearm lesions only; 1 month F/U; no clinical evidence of vasculitis 4 recurrent lesion at same site over 1 year; 4 months F/U; no clinical evidence of vasculitis 6 recurrent lesion at same site over 2 years; 10 days F/U; no clinical evidence of vasculitis Bilateral forearm lesions; 28 months F/U; no clinical evidence of vasculitis Same patient as lesion 4 Bilateral multiple forearm lesions; 2 months F/U; no clinical evidence of vasculitis Clinically typical multiple SP on forearms; 30 months F/U; no clinical evidence of vasculitis Solitary lesion; 9 months F/U; no clinical evidence of vasculitis Solitary lesion; 13 months F/U; no clinical evidence of vasculitis
1 2
58/F 62/F
R forearm L forearm
50 mm purpura 10 mm purpura
3
59/F
L index finger
15 mm purpura
4 5 6
60/F 60/F 62/F
L wrist L forearm L forearm
10 mm purpura 10 mm purpura 10–20 mm purpura
7
76/F
R leg
8 9
75/F 72/M
R forearm L upper arm
10 11 12 13
60/F 59/M 57/F 63/M
R upper arm L dorsal hand L leg R forearm
Ecchymotic macule, size not available 10 mm purpura 10 mm irregular pigmented lesion 5 mm violaceous papule 6 mm pigmented macule 5 mm pigmented macule 12 mm irregular pigmented macule
Solitary Solitary Solitary Solitary
lesion; lesion; lesion; lesion;
1 month F/U; no clinical evidence of vasculitis 9 months F/U; no clinical evidence of vasculitis recent itch; 2 months F/U; no clinical evidence of vasculitis 7 days F/U; no clinical evidence of vasculitis
Lesions 4 and 5 developed and were biopsied synchronously in a single patient. F, female; F/U, follow up; L, left; M, male; R, right.
Copyright © Royal College of pathologists of Australasia. Unauthorized reproduction of this article is prohibited.
486
Pathology (2013), 45(5), August
WOOD and LEBOIT
Fig. 1 Purpuric lesions on left forearm (Case 5).
interstitial neutrophilic inflammation, which was generally mild to moderate in degree (Fig. 2 and 3), with a single case showing prominent interstitial inflammation (Fig. 4). Leukocytoclastic debris was identified in all but one case, although generally this was not a prominent feature. Eleven of the 13 cases showed moderate or prominent clefting within the elastotic upper reticular dermis, and in many cases the neutrophilic infiltrate appeared to surround the areas of stromal clefting (Fig. 5). In four cases vascular mural fibrin deposition was present (Fig. 6). Notably, all of these cases showed clear evidence of clefting within elastotic dermal tissue. There was no evidence of vascular amyloid deposition on H&E staining.
DISCUSSION Both clinically and histologically, purpura is conceptually divided into inflammatory and non-inflammatory causes. The former is typified by vasculitis, in which inflammatory damage to vessel walls leads to red blood cell extravasation. SP is generally considered a prototypical form of noninflammatory purpura, in which simple mechanical factors (weakened blood vessels and dermal connective tissue, subject to minor shearing forces) underlie the development of haemorrhage without inflammation. The only description of SP with neutrophilic inflammation of which we are aware was presented in abstract form by Sung and Jacobson at the Annual Meeting of the American Society of Dermatopathology in 2007.7 They reported that, in a study of 92 cases of solar purpura, 13 cases (14%) were identified which Table 2 Lesion 1 2 3 4 5 6 7 8 9 10 11 12 13
Fig. 2 All cases showed red cell extravasation on a background of severe solar elastosis. In most cases a mild to moderate infiltrate of neutrophils was present in an interstitial and perivascular distribution (Case 3, H&E).
showed features ‘reminiscent of small vessel leukocytoclastic vasculitis’. The incidence of this phenomenon, clinical distribution and histological features are remarkably similar to those identified in our series of cases. To the best of our knowledge, there is no other formal description of this phenomenon in the published literature, and indeed standard descriptions of solar purpura implicity or explicitly exclude this possibility.1,4–6 There is some histological similarity between the cases identified in this series and lesions previously reported as ‘bullous solar elastosis’ (BSE).8–10 The latter is described as an uncommon morphological variant of solar elastosis, with only six patients reported in the literature, though it is likely that this underestimates the true incidence of that phenomenon. The clinical descriptions and images from these reports show some overlap with those of cases reported in our series, with the presence of haemorrhagic lesions on the forearms. Additionally, stromal clefting in an elastotic dermis which is described as the histological correlate of clinical BSE was identified in 85% of our cases. Descriptions of BSE have noted varying degrees of inflammatory change, fibrin deposition and red cell extravasation surrounding areas of dermal clefting.8–10 In contrast to lesions described as BSE, none of our patients presented with clinical bullae. Nevertheless, the histological and pathophysiological similarities between these processes
Histopathological features of 13 lesions of SP showing neutrophilic inflammation Perivascular neutrophils
Interstitial neutrophils
Leukocytoclastic debris
Vascular mural fibrin
Stromal clefting
1 2 2 2 1 1 1 1 2 1 1 1 1
2 1 2 2 1 3 1 1 1 1 2 2 1
0 1 2 2 1 1 2 2 2 1 2 1 2
0 0 2 1 0 0 1 0 0 0 0 0 2
0 2 3 2 2 2 3 3 0 3 3 1 2
Scoring: 0, absent; 1, inconspicuous; 2, moderately prominent; 3, prominent.
Copyright © Royal College of pathologists of Australasia. Unauthorized reproduction of this article is prohibited.
‘INFLAMMATORY’ VARIANT OF SOLAR PURPURA
Fig. 3 Extravasated blood present within clefted spaces was identified in many cases, in this area with admixed neutrophils and a small amount of leukocytoclastic debris (Case 4, H&E).
suggest that they may represent patterns on a continuum of change in severely elastotic skin, rather than discrete phenomena. In contrast to leukocytoclastic vasculitis, which typically shows numerous purpuric lesions with involvement of the lower limbs, cases of SPI in our series showed a distribution similar to typical solar purpura, with most cases involving exclusively the upper limb and either solitary or oligolesional disease. The clinical history and presenting morphology in these patients were also otherwise typical of solar purpura, and clinical follow-up, in some cases for more than 2 years, revealed no evidence of cutaneous or systemic vasculitis. The major histological differential diagnostic considerations in these cases are leukocytoclastic vasculitis and a neutrophilic dermatosis such as Sweet’s syndrome occurring in sundamaged skin. In regard to the former possibility, mural fibrin deposition, a characteristic feature of leukocytoclastic vasculitis, was only seen in four of our cases. Although mural fibrin deposits are a criterion for the histopathological diagnosis of leukocytoclastic vasculitis, they can also occur in ‘secondary vasculitis’, i.e., vasculitis in which damage to vessel walls is caused by a non-vascular actor, such as ulceration, herpesvirus infection or folliculitis. In regard to the possibility of a neutrophilic dermatosis, our cases of SP with inflammation showed more prominent dermal haemorrhage relative to
Fig. 4 More florid inflammation associated with dermal haemorrhage was present in one case (Case 6, H&E).
487
Fig. 5 Stromal clefting in severely elastotic dermis with fibrin and red cell extravasation and a surrounding neutrophilic infiltrate (Case 8, H&E).
interstitial neutrophilic inflammation and a relative paucity of leukocytoclasis when compared to prototypical Sweet’s syndrome. Perhaps more critically, with the exception of an inflammatory infiltrate, all of our cases showed a histological pattern typical of solar purpura, with the inclusion of prominent clefts within elastotic material in most cases. The clefts are usually the centre of the process, with extravasated erythrocytes around them, and fibrin sometimes lining their interior margins. In these cases there appeared to be some tendency of the neutrophilic infiltrate to localise around the margins of areas of dermal clefting, although this distribution was by no means absolute. This also was true of inflamed vessels, which seldom were far from the clefts. We saw no cases with evidence of inflammation involving medium sized vessels or vessels beneath the mid-dermis, indicating that histological overlap with other vasculitic processes is unlikely. The recruitment of neutrophils to the process may be an effect of tissue injury from shearing forces that disrupt a fragile, elastotic dermis. This series may overestimate the incidence of SP with inflammation, due to the bias of a large component of external consultations in the practice at UCSF. We are aware that two
Fig. 6 In a minority of cases fibrin was present within vessel walls in areas of neutrophilic inflammation, mimicking leukocytoclastic vasculitis (Case 3, H&E).
Copyright © Royal College of pathologists of Australasia. Unauthorized reproduction of this article is prohibited.
488
Pathology (2013), 45(5), August
WOOD and LEBOIT
cases in this series were referred by the original reporting pathologist (in one case a dermatopathologist) because of concern for a vasculitis. Nevertheless, the incidence in our material is similar to that reported in the only other formal examination of this phenomenon, and the subsequent accrual of cases in routine practice suggests that SP with inflammation is not a rare finding in clinical practice in a region with high incidence of severe solar elastosis (Perth, Western Australia). Extravasation of blood leads to activation of coagulation, inflammatory and repair pathways, mediated by a complex interaction of circulating factors and cytokines. In addition, the pathogenesis of SP likely includes traumatic disruption of weakened elastotic dermal tissue, histologically evidenced in many cases by stromal clefting with surrounding fibrin and inflammatory changes. We hypothesise that these processes may lead to the recruitment of significant numbers of neutrophils to sites of dermal injury in a minority of cases of SP. In addition, we suggest that clefting of elastotic stroma and a tendency towards localisation of neutrophilic infiltration to the margins of such clefting are clues to recognising this phenomenon. It seems probable that neutrophilic inflammation in SP is a transient phenomenon, but accurate information regarding the duration of lesions prior to biopsy was difficult to obtain in our cases. It remains to be investigated whether similar inflammatory changes can occasionally be detected in other forms of purpura traditionally considered ‘noninflammatory’. It is important that pathologists are aware of the possibility of identifying neutrophilic inflammation, leukocytoclastic debris and even occasionally vascular mural fibrin deposition in cases of otherwise typical solar purpura in order to prevent misdiagnosis as a cutaneous vasculitis or neutrophilic dermatosis. This phenomenon, while not described in standard textbooks, appears to be relatively common, in our series
occurring in slightly over 10% of cases, and in the only other description of which we are aware, occurring in 14% of cases.7 In such circumstances, correlation with the clinical presentation and subsequent evolution and attention to otherwise typical histological features of solar purpura are important features in identifying the hitherto under-recognised phenomenon of solar purpura with inflammation. Conflicts of interest and sources of funding: The authors state that there are no conflicts of interest to disclose.
Address for correspondence: Dr P. E. LeBoit, Dermatopathology Section, University of California, San Francisco, CA, 1701 Divisadero St, Suite 280, San Francisco, CA 94563, USA. E-mail: [email protected]
References 1. Weedon D. Weedon’s Skin Pathology. 3rd ed. Philadelphia: Elsevier, 2010; 197. 2. Bolognia JL, Jorizzo JL, Rapini RP, editors. Dermatology. 2nd ed. St Louis: Mosby Elsevier, 2008; Chapter 23. 3. Waters AJ, Sandhu DR, Green CM, Ferguson J. Solar capillaritis as a cause of solar purpura. Clin Exp Dermatol 2009; 34: e821–4. 4. Elder D, editor. Lever’s Histopathology of the Skin. 10th ed. Philadelphia: Lippincott Williams and Wilkins, 2009; 231. 5. Barnhill RL, Crowson AN, Magro CM, et al. Dermatopathology. 3rd ed. New York: McGraw Hill, 2010; 209. 6. McKee PH, Calnje E, Granter SR. Pathology of the Skin. 3rd ed. St Louis: Elsevier Mosby, 2005; 773. 7. Sung K, Jacobson M. Solar purpura related pseudovasculitis: incidence and histopathology. 44th American Society of Dermatopathology Meeting, Baltimore, Oct 19 2007, Abstract. 8. Mataix J, Banuls J, Jimenez MJ. Bullous solar elastosis: An exceedingly rare clinical presentation of actinic elastosis. J Cutan Pathol 2007; 37: 605–6. 9. Lugo A, Montalvo L, Gonzalez JR, Sanchez JL. Bullous solar elastosis. Am J Dermatopathol 2005; 27: 34–5. 10. Williams BT, Barr RJ, Dutta B. Bullous solar elastosis. J Am Acad Dermatol 1996; 34: 856–8.
Copyright © Royal College of pathologists of Australasia. Unauthorized reproduction of this article is prohibited.
ANATOMICAL PATHOLOGY
An ‘inflammatory’ variant of solar purpura: a simulant of leukocytoclastic vasculitis and neutrophilic dermatoses BENJAMIN A. WOOD*{
AND
PHILIP E. LEBOIT{§
*PathWest Laboratory Medicine, QEII Medical Center and Sir Charles Gairdner Hospital, {School of Pathology and Laboratory Medicine, the University of Western Australia, Perth, Western Australia, Australia; Departments of zPathology, and §Dermatology, University of California, San Francisco, CA, United States
Summary Aims: To study the clinical and pathological features of cases of apparent solar purpura, with attention to the recently described phenomenon of inflammatory changes within otherwise typical lesions. Methods: We studied 95 cases diagnosed as solar purpura and identified 10 cases (10.5%) in which significant neutrophilic inflammation was present, potentially simulating a leukocytoclastic vasculitis or neutrophilic dermatosis. An additional three cases were identified in subsequent routine practice. The clinical features, including follow-up for subsequent development of vasculitis and histological features were studied. Results: In all cases the histological features were typical of solar purpura, with the exception of inflammatory changes, typically associated with clefting of elastotic stroma. Clinical follow-up information was available for all patients and none developed subsequent evidence of a cutaneous or systemic vasculitis or neutrophilic dermatosis. Conclusions: Inflammatory changes appear to be more frequent in solar purpura than is generally recognised. Awareness of this histological variation and correlation with the clinical findings and evolution is important in avoiding misdiagnosis. Key words: Inflammation, neutrophilic dermatosis, pseudovasculitis, secondary vasculitis, solar purpura. Received 15 November 2012, revised 17 January, accepted 29 January 2013
The pathophysiology of SP is thought to be related to a combination of minor (often unrecognised) trauma, combined with impaired tissue support and elasticity in skin with severe solar elastosis, leading to extravasation of blood from dermal vessels.1,4–6 In some cases additive predisposing factors such as the use of corticosteroids, anticoagulant or antiplatelet medications are identifiable.6 Although the majority of cases of SP are recognised as such clinically, biopsy specimens are commonly obtained by clinicians who practice in areas with fair skinned patients and sunny climes. The reasons for submission of these biopsies vary, but include concern for a melanocytic or vascular tumour and clinical consideration of the possibility of an alternative cause for purpura, such as vasculitis. Significant neutrophilic inflammation is generally not considered a feature of these lesions. In standard textbooks of dermatopathology, SP is classified as a ‘non-inflammatory’ type of purpura, and an inflammatory infiltrate is not described,1,4,5 or is specifically excluded.6 The presence of such an infiltrate naturally leads the pathologist to consider alternative causes of purpura, particularly vasculitis. In 2007 Sung and Jacobson presented in abstract form a series of 13 cases described as ‘solar purpura-related pseudovasculitis’.7 Following this description, we noted a number of cases of otherwise clinically and histologically typical solar purpura showing neutrophilic infiltrates. To date there has been no detailed publication exploring the incidence and features of this phenomenon.
INTRODUCTION
METHODS
Extravasation of red blood cells into the dermis correlates with the clinical finding of purpura. Its multiple causes include purpura due to inflammatory damage to vessel walls (e.g., in vasculitis), vascular leakage due to luminal occlusion by thrombi or cryoproteins, and purpura due to altered coagulation.1 So-called ‘senile’ purpura is a common condition, typically developing on the forearms of elderly patients with severely sun-damaged skin. There is a trend to eliminate use of the perjorative adjective ‘senile’, and the condition is also called ‘actinic purpura’, ‘solar purpura’ and ‘Bateman’s purpura’.2 Somewhat confusingly, the term solar purpura has also been used to describe a different spectrum of diseases characterised by rapidly developing purpuric lesions in response to acute solar irradiation,3 which are unrelated to the phenomenon of senile/solar purpura (SP) as used herein.
The files of the University of California San Francisco Dermatopathology Service (UCSF) were searched for all cases with a retrieval code of ‘solar purpura’ and ‘solar purpura-neutrophil rich’ between January 2009 and September 2011. The term ‘retrieval code’ designates that the pathologist who reviewed the sections referenced this term in the computer system for potential case retrieval, and does not necessarily indicate that the diagnosis on the report read as such. The slides of all cases were reviewed, and cases showing evidence of significant neutrophilic infiltration were selected for study. Significant neutrophilic inflammation was defined by the presence of a perivascular neutrophilic infiltrate, with or without evidence of vascular mural fibrin deposition, and/or a non-focal interstitial neutrophilic infiltrate. Cases showing trivial features such as neutrophil margination within vessel lumens or very focal entrapment of neutrophils in areas of fibrin deposition were not selected. A total of 106 cases of solar purpura were retrieved for review from the UCSF files. Eleven cases were excluded, due either to significant co-existing pathology or an absence of clear histological features of SP on the slides
Print ISSN 0031-3025/Online ISSN 1465-3931 DOI: 10.1097/PAT.0b013e3283632649
#
2013 Royal College of Pathologists of Australasia
Copyright © Royal College of pathologists of Australasia. Unauthorized reproduction of this article is prohibited.
‘INFLAMMATORY’ VARIANT OF SOLAR PURPURA
available. Of the remaining 95 cases, 10 lesions (10.5%) showed a significant neutrophilic inflammatory infiltrate. An additional three cases were prospectively gathered from the routine practice of one of the authors (BAW) over the period from November 2011 to October 2012, giving a total of 13 cases. All available haematoxylin and eosin (H&E) stained sections that had been prepared by standard techniques were reviewed, comprising two slides for Cases 1 and 13 and one slide each for the remaining 11 lesions. The cases were reviewed at a mutiheaded microscope by both authors (Cases 1–10) or by one author microscopically and shared via digital imaging (Cases 11–13). Five features were assessed as being absent or present, and in the latter circumstance graded in a semi-quantitative fashion on a three-point scale, with 1 representing a minor inconspicuous feature and 3 representing a marked or prominent finding. These features were: perivascular inflammation; interstitial inflammation; leukocytoclastic debris; tissue clefts; vascular mural fibrin deposition. Clinical data were retrieved from the pathology files and clinical follow-up information was sought by telephone contact with the referring clinician. Specifically, detailed information regarding the clinical presentation and clinical evolution subsequent to biopsy was sought, with particular reference to excluding development of a cutaneous or systemic vasculitis. Additionally, a medication history was obtained with regard to synchronous use of corticosteroids, anti-coagulant or anti-platelet medications.
RESULTS The 13 lesions studied occurred in 12 patients, ranging in age from 57 to 76 years (mean 64 years), including nine females and three males (Table 1). Six lesions were present on the forearms, while the hands/wrists were the site of three lesions, the upper arms involved two cases and the lower leg involved two cases. One patient with disease on the lower extremity (Case 7) also had small numbers of lesions on the forearms which were considered clinically to represent typical SP and were not biopsied. The lesions were variously described as haemorrhagic, violaceous, pigmented or purpuric lesions, in most cases with the appearance of macules 10–20 mm in maximum dimension (Fig. 1). A single, larger 50 mm patch and smaller 5 mm papule were each described. Two lesions were described as irregular pigmented macules, with clinical concern for melanoma. In no case was there clinical evidence of bulla formation. Table 1 Lesion
485
Detailed information regarding the presence of other lesions was available for all 12 patients from the referring clinician. In eight patients the lesion biopsied was solitary (in two of these patients a history of a solitary lesion recurrent at a single site was elicited). In the other cases, multiple lesions of similar clinical morphology were present on the arms and forearms. No patient other than Cases 7 and 12 (in whom leg lesions were biopsied) had purpuric lesions on the legs. One patient (Case 3) was taking warfarin, one patient (Case 6) was taking acetylsalicylic acid and one patient (Case 12) was taking aspirin and clopidogrel when the lesions developed. The other patients did not have a history of anticoagulant/antiplatelet medication or systemic corticosteroid use. Clinical follow-up information was available for all cases, with duration of follow-up ranging from 7 days to 30 months (median 3 months). All seven patients with solitary lesions were without similar lesions on follow-up, either as a result of complete clinical removal by biopsy (5 cases) or resolution of clinical residuum of the biopsied lesion, the latter occurring without further treatment. In one patient with bilateral single forearm lesions (Case 1), both the residuum of the lesion biopsied and a contralateral clinically similar lesion resolved with the use of topical cortisone. The latter treatment was instituted by the treating clinician on an empirical basis due to a pathological diagnosis of ‘solar purpura with a neutrophil rich infiltrate’. None of the patients developed evidence of a cutaneous or systemic vasculitis during follow-up, although evaluation of the latter possibility was limited to clinical history and examination in most cases. The histological features are summarised in Table 2 and illustrated in Fig. 2–6. By definition, the sections from all cases showed severe solar elastosis and evidence of dermal haemorrhage. In all cases selected for this study, dermal haemorrhage was evidenced by extravasation of red blood cells. No cases were identified which showed resolving haemorrhage (sparse lymphohistiocytic infiltrate associated with haemosiderin deposition) with concomitant neutrophilic inflammation. All cases showed both perivascular and
Clinical features and follow-up details of patients with SP showing neutrophilic inflammation Age/Gender
Location
Clinical morphology
Other clinical features Bilateral forearm lesions only; 1 month F/U; no clinical evidence of vasculitis 4 recurrent lesion at same site over 1 year; 4 months F/U; no clinical evidence of vasculitis 6 recurrent lesion at same site over 2 years; 10 days F/U; no clinical evidence of vasculitis Bilateral forearm lesions; 28 months F/U; no clinical evidence of vasculitis Same patient as lesion 4 Bilateral multiple forearm lesions; 2 months F/U; no clinical evidence of vasculitis Clinically typical multiple SP on forearms; 30 months F/U; no clinical evidence of vasculitis Solitary lesion; 9 months F/U; no clinical evidence of vasculitis Solitary lesion; 13 months F/U; no clinical evidence of vasculitis
1 2
58/F 62/F
R forearm L forearm
50 mm purpura 10 mm purpura
3
59/F
L index finger
15 mm purpura
4 5 6
60/F 60/F 62/F
L wrist L forearm L forearm
10 mm purpura 10 mm purpura 10–20 mm purpura
7
76/F
R leg
8 9
75/F 72/M
R forearm L upper arm
10 11 12 13
60/F 59/M 57/F 63/M
R upper arm L dorsal hand L leg R forearm
Ecchymotic macule, size not available 10 mm purpura 10 mm irregular pigmented lesion 5 mm violaceous papule 6 mm pigmented macule 5 mm pigmented macule 12 mm irregular pigmented macule
Solitary Solitary Solitary Solitary
lesion; lesion; lesion; lesion;
1 month F/U; no clinical evidence of vasculitis 9 months F/U; no clinical evidence of vasculitis recent itch; 2 months F/U; no clinical evidence of vasculitis 7 days F/U; no clinical evidence of vasculitis
Lesions 4 and 5 developed and were biopsied synchronously in a single patient. F, female; F/U, follow up; L, left; M, male; R, right.
Copyright © Royal College of pathologists of Australasia. Unauthorized reproduction of this article is prohibited.
486
Pathology (2013), 45(5), August
WOOD and LEBOIT
Fig. 1 Purpuric lesions on left forearm (Case 5).
interstitial neutrophilic inflammation, which was generally mild to moderate in degree (Fig. 2 and 3), with a single case showing prominent interstitial inflammation (Fig. 4). Leukocytoclastic debris was identified in all but one case, although generally this was not a prominent feature. Eleven of the 13 cases showed moderate or prominent clefting within the elastotic upper reticular dermis, and in many cases the neutrophilic infiltrate appeared to surround the areas of stromal clefting (Fig. 5). In four cases vascular mural fibrin deposition was present (Fig. 6). Notably, all of these cases showed clear evidence of clefting within elastotic dermal tissue. There was no evidence of vascular amyloid deposition on H&E staining.
DISCUSSION Both clinically and histologically, purpura is conceptually divided into inflammatory and non-inflammatory causes. The former is typified by vasculitis, in which inflammatory damage to vessel walls leads to red blood cell extravasation. SP is generally considered a prototypical form of noninflammatory purpura, in which simple mechanical factors (weakened blood vessels and dermal connective tissue, subject to minor shearing forces) underlie the development of haemorrhage without inflammation. The only description of SP with neutrophilic inflammation of which we are aware was presented in abstract form by Sung and Jacobson at the Annual Meeting of the American Society of Dermatopathology in 2007.7 They reported that, in a study of 92 cases of solar purpura, 13 cases (14%) were identified which Table 2 Lesion 1 2 3 4 5 6 7 8 9 10 11 12 13
Fig. 2 All cases showed red cell extravasation on a background of severe solar elastosis. In most cases a mild to moderate infiltrate of neutrophils was present in an interstitial and perivascular distribution (Case 3, H&E).
showed features ‘reminiscent of small vessel leukocytoclastic vasculitis’. The incidence of this phenomenon, clinical distribution and histological features are remarkably similar to those identified in our series of cases. To the best of our knowledge, there is no other formal description of this phenomenon in the published literature, and indeed standard descriptions of solar purpura implicity or explicitly exclude this possibility.1,4–6 There is some histological similarity between the cases identified in this series and lesions previously reported as ‘bullous solar elastosis’ (BSE).8–10 The latter is described as an uncommon morphological variant of solar elastosis, with only six patients reported in the literature, though it is likely that this underestimates the true incidence of that phenomenon. The clinical descriptions and images from these reports show some overlap with those of cases reported in our series, with the presence of haemorrhagic lesions on the forearms. Additionally, stromal clefting in an elastotic dermis which is described as the histological correlate of clinical BSE was identified in 85% of our cases. Descriptions of BSE have noted varying degrees of inflammatory change, fibrin deposition and red cell extravasation surrounding areas of dermal clefting.8–10 In contrast to lesions described as BSE, none of our patients presented with clinical bullae. Nevertheless, the histological and pathophysiological similarities between these processes
Histopathological features of 13 lesions of SP showing neutrophilic inflammation Perivascular neutrophils
Interstitial neutrophils
Leukocytoclastic debris
Vascular mural fibrin
Stromal clefting
1 2 2 2 1 1 1 1 2 1 1 1 1
2 1 2 2 1 3 1 1 1 1 2 2 1
0 1 2 2 1 1 2 2 2 1 2 1 2
0 0 2 1 0 0 1 0 0 0 0 0 2
0 2 3 2 2 2 3 3 0 3 3 1 2
Scoring: 0, absent; 1, inconspicuous; 2, moderately prominent; 3, prominent.
Copyright © Royal College of pathologists of Australasia. Unauthorized reproduction of this article is prohibited.
‘INFLAMMATORY’ VARIANT OF SOLAR PURPURA
Fig. 3 Extravasated blood present within clefted spaces was identified in many cases, in this area with admixed neutrophils and a small amount of leukocytoclastic debris (Case 4, H&E).
suggest that they may represent patterns on a continuum of change in severely elastotic skin, rather than discrete phenomena. In contrast to leukocytoclastic vasculitis, which typically shows numerous purpuric lesions with involvement of the lower limbs, cases of SPI in our series showed a distribution similar to typical solar purpura, with most cases involving exclusively the upper limb and either solitary or oligolesional disease. The clinical history and presenting morphology in these patients were also otherwise typical of solar purpura, and clinical follow-up, in some cases for more than 2 years, revealed no evidence of cutaneous or systemic vasculitis. The major histological differential diagnostic considerations in these cases are leukocytoclastic vasculitis and a neutrophilic dermatosis such as Sweet’s syndrome occurring in sundamaged skin. In regard to the former possibility, mural fibrin deposition, a characteristic feature of leukocytoclastic vasculitis, was only seen in four of our cases. Although mural fibrin deposits are a criterion for the histopathological diagnosis of leukocytoclastic vasculitis, they can also occur in ‘secondary vasculitis’, i.e., vasculitis in which damage to vessel walls is caused by a non-vascular actor, such as ulceration, herpesvirus infection or folliculitis. In regard to the possibility of a neutrophilic dermatosis, our cases of SP with inflammation showed more prominent dermal haemorrhage relative to
Fig. 4 More florid inflammation associated with dermal haemorrhage was present in one case (Case 6, H&E).
487
Fig. 5 Stromal clefting in severely elastotic dermis with fibrin and red cell extravasation and a surrounding neutrophilic infiltrate (Case 8, H&E).
interstitial neutrophilic inflammation and a relative paucity of leukocytoclasis when compared to prototypical Sweet’s syndrome. Perhaps more critically, with the exception of an inflammatory infiltrate, all of our cases showed a histological pattern typical of solar purpura, with the inclusion of prominent clefts within elastotic material in most cases. The clefts are usually the centre of the process, with extravasated erythrocytes around them, and fibrin sometimes lining their interior margins. In these cases there appeared to be some tendency of the neutrophilic infiltrate to localise around the margins of areas of dermal clefting, although this distribution was by no means absolute. This also was true of inflamed vessels, which seldom were far from the clefts. We saw no cases with evidence of inflammation involving medium sized vessels or vessels beneath the mid-dermis, indicating that histological overlap with other vasculitic processes is unlikely. The recruitment of neutrophils to the process may be an effect of tissue injury from shearing forces that disrupt a fragile, elastotic dermis. This series may overestimate the incidence of SP with inflammation, due to the bias of a large component of external consultations in the practice at UCSF. We are aware that two
Fig. 6 In a minority of cases fibrin was present within vessel walls in areas of neutrophilic inflammation, mimicking leukocytoclastic vasculitis (Case 3, H&E).
Copyright © Royal College of pathologists of Australasia. Unauthorized reproduction of this article is prohibited.
488
Pathology (2013), 45(5), August
WOOD and LEBOIT
cases in this series were referred by the original reporting pathologist (in one case a dermatopathologist) because of concern for a vasculitis. Nevertheless, the incidence in our material is similar to that reported in the only other formal examination of this phenomenon, and the subsequent accrual of cases in routine practice suggests that SP with inflammation is not a rare finding in clinical practice in a region with high incidence of severe solar elastosis (Perth, Western Australia). Extravasation of blood leads to activation of coagulation, inflammatory and repair pathways, mediated by a complex interaction of circulating factors and cytokines. In addition, the pathogenesis of SP likely includes traumatic disruption of weakened elastotic dermal tissue, histologically evidenced in many cases by stromal clefting with surrounding fibrin and inflammatory changes. We hypothesise that these processes may lead to the recruitment of significant numbers of neutrophils to sites of dermal injury in a minority of cases of SP. In addition, we suggest that clefting of elastotic stroma and a tendency towards localisation of neutrophilic infiltration to the margins of such clefting are clues to recognising this phenomenon. It seems probable that neutrophilic inflammation in SP is a transient phenomenon, but accurate information regarding the duration of lesions prior to biopsy was difficult to obtain in our cases. It remains to be investigated whether similar inflammatory changes can occasionally be detected in other forms of purpura traditionally considered ‘noninflammatory’. It is important that pathologists are aware of the possibility of identifying neutrophilic inflammation, leukocytoclastic debris and even occasionally vascular mural fibrin deposition in cases of otherwise typical solar purpura in order to prevent misdiagnosis as a cutaneous vasculitis or neutrophilic dermatosis. This phenomenon, while not described in standard textbooks, appears to be relatively common, in our series
occurring in slightly over 10% of cases, and in the only other description of which we are aware, occurring in 14% of cases.7 In such circumstances, correlation with the clinical presentation and subsequent evolution and attention to otherwise typical histological features of solar purpura are important features in identifying the hitherto under-recognised phenomenon of solar purpura with inflammation. Conflicts of interest and sources of funding: The authors state that there are no conflicts of interest to disclose.
Address for correspondence: Dr P. E. LeBoit, Dermatopathology Section, University of California, San Francisco, CA, 1701 Divisadero St, Suite 280, San Francisco, CA 94563, USA. E-mail: [email protected]
References 1. Weedon D. Weedon’s Skin Pathology. 3rd ed. Philadelphia: Elsevier, 2010; 197. 2. Bolognia JL, Jorizzo JL, Rapini RP, editors. Dermatology. 2nd ed. St Louis: Mosby Elsevier, 2008; Chapter 23. 3. Waters AJ, Sandhu DR, Green CM, Ferguson J. Solar capillaritis as a cause of solar purpura. Clin Exp Dermatol 2009; 34: e821–4. 4. Elder D, editor. Lever’s Histopathology of the Skin. 10th ed. Philadelphia: Lippincott Williams and Wilkins, 2009; 231. 5. Barnhill RL, Crowson AN, Magro CM, et al. Dermatopathology. 3rd ed. New York: McGraw Hill, 2010; 209. 6. McKee PH, Calnje E, Granter SR. Pathology of the Skin. 3rd ed. St Louis: Elsevier Mosby, 2005; 773. 7. Sung K, Jacobson M. Solar purpura related pseudovasculitis: incidence and histopathology. 44th American Society of Dermatopathology Meeting, Baltimore, Oct 19 2007, Abstract. 8. Mataix J, Banuls J, Jimenez MJ. Bullous solar elastosis: An exceedingly rare clinical presentation of actinic elastosis. J Cutan Pathol 2007; 37: 605–6. 9. Lugo A, Montalvo L, Gonzalez JR, Sanchez JL. Bullous solar elastosis. Am J Dermatopathol 2005; 27: 34–5. 10. Williams BT, Barr RJ, Dutta B. Bullous solar elastosis. J Am Acad Dermatol 1996; 34: 856–8.
Copyright © Royal College of pathologists of Australasia. Unauthorized reproduction of this article is prohibited.