- Email: [email protected]
Postirradiation morphea: an underrecognized complication of treatment for breast cancer
Human Pathology (2008) 39, 1680–1688
www.elsevier.com/locate/humpath
Original contribution
Postirradiation morphea: an underrecognized complication of treatment for breast cancer Noreen Walsh MD, MRCPI, FRCP(C), FRCPath (UK)a,⁎, Dorianne Rheaume FRCPC (Internal Medicine), FRCPC (Radiation Oncology)b , Penelope Barnes MD, FRCP(C)a , Robert Tremaine MD, FRCP(C)c , Michael Reardon MD, FRCP(C)c a
Department of Pathology (Anatomical Pathology), Capital District Health Authority and Dalhousie University, Halifax, Nova Scotia, Canada B3H 1V8 b Department of Radiation Oncology, Capital District Health Authority and Dalhousie University, Halifax, Nova Scotia, Canada B3H 1V8 c Department of Medicine (Dermatology), Capital District Health Authority and Dalhousie University, Halifax, Nova Scotia, Canada B3H 1V8 Received 9 April 2008; revised 22 April 2008; accepted 23 April 2008
Keywords: Postirradiation morphea; Localized scleroderma after radiotherapy; Cutaneous complications of irradiation for breast cancer
Summary The most common cutaneous side effects of radiotherapy include radiodermatitis and radiation fibrosis. These are influenced by the type, dose, and pattern of delivery of the treatment. Distinct from these is postirradiation morphea (localized scleroderma), an idiosyncratic treatmentrelated phenomenon. Within the last 20 years, approximately 31 examples of postirradiation morphea after treatment for breast cancer were reported. We describe 5 new cases of this entity and integrate our findings with those in the literature. The mean age of the patients at the time of diagnosis of cancer was 58 years; all were left-sided and treated by local excision of the tumor, ipsilateral axillary lymph node dissection, and local radiotherapy. After an interval of 4 to 12 years, the patients developed morphea in the radiation portals, with extension beyond it in one instance. Recurrent breast carcinoma was suspected clinically in 2 cases. Microscopically, changes of morphea involved the dermis in all cases and the subcutis in 2. There was associated lichen sclerosus et atrophicus in 2 cases. Our data about management and outcome are limited, but 1 patient treated with potent topical steroids experienced gradual softening of the affected skin over a 5-year period, whereas another had a mastectomy for relief of painful induration of the breast. Our findings support existing theories about the pathogenesis of this condition and link it to those of sclerodermoid graft-versus-host disease. The purpose of our communication is to draw attention to this underrecognized complication of treatment for breast cancer. © 2008 Elsevier Inc. All rights reserved.
1. Introduction ⁎ Corresponding author. Division of Anatomical Pathology, Halifax, Nova Scotia, Canada B3H 1V8. E-mail address: [email protected] (N. Walsh). 0046-8177/$ – see front matter © 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.humpath.2008.04.010
That morphea ranks among the local cutaneous side effects of radiotherapy was recognized shortly after the discovery of x-
Postirradiation morphea Table 1
1681
Breast cancer data in current series
Case Age Site no. (y)
Histopathologic type of cancer
Axillary Treatment LN status
1
69
Left breast
Invasive ductal carcinoma LN° (ER+; PR+)
2
51
Left breast
Invasive ductal carcinoma LN° (ER−; PR+)
3
58
Left breast
Invasive ductal carcinoma LN° (ER+ and PR+)
4
57
Left breast
5
53
Left breast
Invasive carcinoma,mixed LN+1 lobular and ductal (ER+; PR−) Invasive ductal carcinoma LN° (ER+; PR+)
Local excision; axillary LN dissection; radiotherapy Local excision; axillary LN dissection; radiotherapy Local excision; axillary LN dissection; radiotherapy Local excision; axillary LN dissection; radiotherapy Local excision; axillary LN dissection; radiotherapy
Outcome
Duration of follow-up (y)
No known metastasis
13
Right hip metastasis and death— 14 14 years after diagnosis of breast cancer. No known metastasis 10
No known metastasis
17
No known metastasis
4
Abbreviations: LN, lymph node; ER, estrogen receptor; PR, progesterone receptor.
rays by Roentgen in 1885 and noted in textbooks of the same era [1]. It received little subsequent attention until the report by Colver et al [2] in 1989, describing the phenomenon in 9 patients, of whom 7 had received radiotherapy for breast cancer. Occurring as an idiosyncratic reaction, independent of the severity of acute radiodermatitis, its temporal relationship to treatment has been inconsistent and its main significance lies in its clinical mimicry of locally recurrent carcinoma. With the benefit of emerging data, it has become clear that this complication of radiotherapy can also develop as a sequel to treatment of cancers at sites other than the breast [2-4] and that it can, on occasion, extend beyond the portal of irradiation [2,48]. Moreover, the related entity, lichen sclerosus et atrophicus (LS&A), observed not infrequently in association with morphea and considered to be part of the same spectrum of disease, has been documented to occur in this clinical setting [9,10]. The objective of this report is to outline the clinical and microscopic characteristics of postirradiation morphea in 5 patients treated for breast cancer and to distill the literature on the subject. Our aims are (і) to highlight the diagnostic challenges presented by this entity, (іі) to expand existing knowledge of a rare but important phenomenon, and (ііі) to contribute to existing hypotheses about its pathogenesis.
2. Methods Five cases of postirradiation morphea were encountered between 1996 and 2008 in the pathology practices of 2 of the authors. Case retrieval was via the laboratory information system (Cerner Classic). Clinical data pertaining to the patients' breast cancers, diagnosed between 1991 and 2004, related treatment and subsequent skin disease were obtained with the assistance of clinicians. These included a radiation oncologist, 2 dermatologists, and a general surgeon. Three control cases of idiopathic morphea were selected from the
pathology files for the purpose of comparative histopathologic and immunohistochemical evaluation. Ten tissue samples (9 skin biopsies and 1 mastectomy specimen) from the 5 study patients and 4 biopsies from the 3 control cases were available for review. Categorization of the morphological findings in each biopsy, as (i) morphea alone or (ii) morphea in combination with LS&A, was based on routine histopathologic sections. Moreover, the stage of morphea, as (a) inflammatory, (b) inflammatory and sclerosing, or (c) sclerosing, was documented. Selected histochemical stains were performed to better explore stromal features of the biopsies. These included an Alcian blue stain (Ventana automated stainer; pH 2.5; VEntana, Tuscon, AZ) in search of excess interstitial dermal mucin and the modified Hart stain (Ventana automated stainer) to highlight the pattern of dermal elastic fibers. A panel of immunohistochemical studies was applied, mainly to determine the phenotype of the cellular infiltrate. The panel consisted of stains for CD3 (Ventana; prediluted; CC1 pretreatment × 30 min), CD20 (Ventana; prediluted; CC1 pretreatment × 8 min), CD4 (Labvision, Fremont, CA; 1/20; CC1 pretreatment × 60 min), CD8 (Labvision; 1/100; CC1 pretreatment × 30 min), CD30 (Dako, Mississauga, Ontario; 1/ 50; protease pretreatment × 4 min), CD68 (Dako; 1/100; no pretreatment), CD34 (Ventana; prediluted; no pretreatment), CD31 (Dako; 1/100; CC1 pretreatment × 8 min), c-kit (Ventana; prediluted; CC1 pretreatment × 30 mns), kappa and lambda light chains (Dako; 1/20,000; CC1 pretreatment × 30 min), and S100 protein (Dako; 1/5000; no pretreatment).
3. Results 3.1. Clinical findings Clinical data related to the 5 patients, their breast cancers, radiotherapeutic regimens, and postirradiation morphea are
1682
N. Walsh et al.
Table 2
Details of radiotherapy for breast cancer in current series
Case no.
Radiation dose (cGy)
Fractions
Dose/ Fraction
Total treatment time (d)
Biologic effective dose (cGy)
Energy (MV)
Patient separation (cm)
Max dose (%)
Boost (cGy/fr)
Observations during radiotherapy
1
4000
15
267
22
75.6
23.5
107
250/2
2
5000
25
200
45
83.5
20.5
102
1000/5
3
5000
25
200
37
83.5
1.25 (cobalt) 1.25 (cobalt) 6
17.2
4
4000
16
250
22
73.2
16.3
Not known 103
5
4240
16
265
Not known
Not known
28.2
108.4
No boost No boost 900/4
Induration in the operative area Mild-Moderate radiation erythema Slight nipple and breast tenderness Minimal radiation erythema Significant erythema and moist desquamation around nipple
1.25 (Cobalt) 6 and 18
outlined in Tables 1-3, respectively. They are summarized here. At the time of the diagnosis of breast cancer, the patients ranged in age from 51 to 69 years (mean = 58). All 5 cancers involved the left breast, 4 being invasive ductal carcinomas and 1 an invasive mixed lobular and ductal carcinoma. All 5 tumors were estrogen and/or progesterone receptor–positive. Only in 1 case was metastatic disease detected in a regional lymph node at the outset, and the management of all patients included local excision of the tumor, axillary lymph node excision, and radiotherapy. Based on the hormone sensitivity of the cancers, it is highly probable that the patients received antiestrogen therapy; however, documentation of this was obtained in only 1 case. In a follow-up period ranging from 4 to 17 years, 1 of the 5 patients is known to have died of metastatic carcinoma.
Table 3
Details of the radiotherapeutic regimens used are outlined in Table 2. The dose of radiation ranged from 4000 to 5000 cGy, delivered in fractionated doses of 200 to 267 cGy. The total treatment time ranged from 22 to 45 days, and 3 of the 5 patients received a boost, as indicated when the tumor closely approaches a resection margin. Other technical details reflect the fact that in the 2 more recent cases (cases 3 and 5), treated within the last 10 years, high-energy radiation was used as is now the norm for deep-seated tumors. In the remaining 3 cases (cases 1, 2, and 4), treated more than 10 years ago, Cobalt, with the delivery of lowerenergy radiation, was employed in accordance with historical standards of practice. Calculations of the maximum dose and patient separation reflect the planning process for radiotherapy, geared to optimize the amount and localization of
Details of skin disease after radiotherapy for breast cancer in current series
Case Interval (y) Site of skin disease Review by Clinical no. since diagnosis of dermatologist differential breast cancer diagnosis 1
6
Radiation field only Yes
2
4
Radiation field only Yes
3
8
Radiation field and extension to right side of chest
4
12
Radiation field only No
5
4
Radiation field only No
Yes
Histopathologic diagnosis
Treatment
Unknown LS&A and morphea (inflammatory and sclerosing stage) Recurrent breast Morphea (inflammatory Potent topical carcinoma and sclerosing stage) steroid (Dermovate) None Radiation recall Morphea (inflammatory stage) dermatitis; carcinoma en cuirasse Unknown Not provided LS&A and morphea (inflammatory and sclerosing stage) Not provided Morphea (inflammatory Mastectomy and sclerosing stage) LS&A
Outcome
Unknown
Softening of skin over 5 y Unknown
Unknown
Insufficient follow-up to determine outcome
Postirradiation morphea
1683 chest, generating a differential diagnosis of radiation recall dermatitis versus carcinoma en cuirasse. Of the remaining 2 cases, a detailed clinical description of the skin change was not provided in one (case 4), and in the other (case 5), there was painful, induration of the irradiated breast. One patient (case 2), treated with a potent topical steroid, experienced gradual softening of the involved skin over a 5-year period. The dermatologist was not convinced that the clinical improvement was therapy related. Another (case 3) received no treatment, and a third (case 5) was managed by mastectomy for symptomatic relief. The management and outcome of the 2 remaining patients are unknown.
3.2. Histopathologic and immunohistochemical findings in skin
Fig. 1 A clinical photograph in case 2 reveals a mottled erythematous plaque involving the left previously irradiated breast, in the area superior and medial to the nipple (A). A close-up shot of the plaque, above the nipple, displays a puckered peau d'orange appearance, as well as the site of a previous punch biopsy (B).
the treatment. Acute radiation dermatitis was observed in 3 patients during treatment, being mild in 2 cases and more pronounced in the third. Induration at the operative site was noted in 1 patient, and another experienced slight nipple and breast tenderness. Clinical data pertaining to the development of morphea in the radiation portals (and beyond it in 1 instance) are outlined in Table 3. This occurred 4 to 12 years after the diagnosis and treatment of breast cancer. Of the 5 patients, 3 were seen by a dermatologist. In one of these (case 1), a description of the clinical lesion was not provided, but the impression was of LS&A. In another (case 2), the skin of the irradiated breast exhibited an erythematous indurated plaque with a peau d'orange appearance, leading to a clinical suspicion of locally recurrent carcinoma (Fig. 1). In the third (case 3), an erythematous edematous plaque was seen to involve the portal of irradiation, with extension to the right side of the
The microscopic changes in biopsies from the study patients were similar to those from the controls, and pertinent findings are outlined in Table 4. To briefly comment on the stages of morphea, this usually evolves from an inflammatory to a sclerosing phase, with an intervening mixed inflammatory and sclerosing phase. The inflammatory changes are characterized by a superficial and deep perivascular and interstitial lymphoplasmacytic infiltrate, whereas the sclerosing changes are represented by thickening of collagen bundles in the reticular dermis. Involvement of the subcutis, referred to as morphea profundus, is typified by collagenous thickening of septa with accompanying lymphoplasmacytic inflammation. On the other hand, LS&A is characterized by edema and hyalinization of the papillary dermis, an underlying nodular to band-like lymphocytic infiltrate, and epidermal thinning with hyperkeratosis. To subcategorize the morphological findings in our group, 2 of the 5 study cases showed features of combined morphea and LS&A and 3 of morphea alone. Of the 3 control cases, 2 exhibited changes of combined morphea and LS&A and 1 of morphea alone. With the exception of 1 study case (case 3) in which the morphea was in the inflammatory phase, the process was in the mixed inflammatory and sclerosing stage in all cases and controls. In the 2 study cases with LS&A (cases 1 and 4), a horizontally oriented space was evident in the edematous/hyalinized papillary dermis, consistent with bullous LS&A. Insufficient detail was available in those cases to determine whether the process was vesiculobullous clinically. A rare multinucleated histiocyte with elastophagocytosis was noted in 1 study and 1 control case. The pattern of coarse elastic fibers in the reticular dermis was not appreciably altered in any of the samples, but a decrease in density of the delicate fibers in the papillary dermis was observed in the setting of LS&A. Excess interstitial dermal mucin deposition was not observed. The phenotypic characteristics of the inflammatory infiltrate were similar in biopsies from the study patients and the controls. It was T-cell predominant, with CD3positive cells, in general, accounting for more than 90% and
1684 Table 4
N. Walsh et al. Postirradiation morphea: summary of histopathologic findings
Data No. of Patients (biopsies) Categorization of findings Main characteristics of dermal infiltrate T cells (CD3+) B cells (CD20+) b T-helper/T-suppressor (CD4:CD8) Plasma cells Distribution pattern of T cells T-helper (CD4) T-suppressor (CD8+) Distribution pattern of CD34+ dermal dendritic cells Pattern of panniculitis observed Sequential observations of panniculitis (case 2)
Study cases
Control cases
a
5 (10) 1 morphea (I); 2 morphea (I & S); 2 morphea (I & S) + LS&A
3 (4) 1 morphea (I & S ); 2 morphea (I & S) + LS&A
≥90% ≤10% 60:40 Scattered/polyclonal
≥90% ≤10% 60:40 Scattered/polyclonal
Perivascular Perivascular, interstitial and DE junction Absent in sclerotic zones, reticular dermis
Perivascular Perivascular, interstitial and DE junction Absent in sclerotic zones, reticular dermis
Septal and lobular (case 2) c predominantly septal (case 5) Sclerotic replacement of subcutis. Increased proportion of plasma cells to lymphocytes.
N/A N/A
Abbreviations: I, inflammatory; I & S, inflammatory and sclerosing; DE junction, dermoepidermal junction; N/A, not applicable. a Of the 10 tissue samples, 1 was a mastectomy specimen. b Only in 1 study case (case 5) did the B-cell component of the infiltrate exceed 10% and amount to approximately 30%. c The pattern of panniculitis observed in this case resembled that previously reported as “pseudosclerodermatous panniculitis after irradiation.”
CD20-positive cells for less than 10% of the infiltrate. Only in case 5, where lymphoid follicles and germinal centers were present, was the B-cell component more prominent, amounting to an estimated 30% of the infiltrate. The approximate T-helper (CD4) to T-suppressor (CD8) cell ratio was 60:40. The T-helper cells were largely confined to perivascular zones, whereas the T-suppressors were perivascular and interstitial in distribution and also involved the dermoepidermal junction. Small numbers of plasma cells (b5% of the infiltrate) were generally present, and these showed expression of both kappa and lambda light chains. In 1 case (case 2), successive biopsies revealed an increase in the plasma cell component over time. C-kit–positive mast cells and CD68-positive histiocytes were scattered throughout the dermis in an unremarkable manner. No CD30positive cells were observed. The endothelial cells of dermal blood vessels, highlighted by stains for CD31 and CD34, appeared normal. With the exception of an increase in the number of capillaries in the papillary dermis in the cases of combined morphea and LS&A, no vascular abnormalities were appreciated. In both the study and control cases, the CD34-positive subpopulation dermal dendritic cells were reduced or absent exclusively in sclerotic zones. S100 protein–positive Langerhans cells were present in normal numbers in the epidermis and were also scattered among perivascular inflammatory cells. Five biopsies taken from a single patient (case 2) at 4 different points in time over an 18-month period were of interest. These had been obtained because of a persistent clinical suspicion of recurrent breast cancer despite histo-
pathologic findings of an inflammatory process. The diagnosis of postirradiation morphea had not been established until the last biopsy was examined. From the outset of sampling, the disease process was in the mixed inflammatory and sclerosing phase, affecting the lower two thirds of the reticular dermis and the subcutis. The mixed pattern persisted over 18 months, but more advanced sclerosis became evident with time as did an increase in the ratio of plasma cells to lymphocytes. In 2 of the 5 biopsies separated by a 16-month interval, there was sufficient subcutaneous fat to enable study of the disease process in this compartment. Unlike the dominance of septal involvement typically seen in idiopathic morphea profundus, the pattern of panniculitis observed in the first of the 2 biopsies from this study patient, with postirradiation morphea, revealed a mixed septal and lobular pattern. This was characterized by a lobular infiltrate of lymphocytes, plasma cells, and foamy macrophages, in addition to fibrous septal thickening. In the biopsy taken 16 months later, there was marked hyalinization of the subcutaneous compartment, accompanied by a plasma cell predominant infiltrate. The availability of abundant tissue from the patient treated by mastectomy, and generous sampling, yielded an opportunity to extensively study the pattern of the disease in this case (case 5). It affected the skin and superficial subcutis of the whole irradiated breast. In each of the 12 sections examined, the morphea was in the mixed inflammatory and sclerosing phase and involved the lower two thirds of the reticular dermis and the subcutaneous fat. In contrast to findings in the other cases and the controls, the inflammatory
Postirradiation morphea
1685 component in this instance revealed scattered lymphoid follicles with germinal centers. The pattern of involvement of subcutaneous fat, in foci, resembled the mixed lobular and septal pattern noted in case 2, but the septal pattern characteristic of morphea profundus predominated (Fig. 2). The underlying mammary tissue appeared normal.
4. Discussion
Fig. 2 A photomicrograph, at scanning magnification (magnification ×5.5), of one of the sections from the mastectomy specimen in case 5 shows condensation of collagen in the lower two thirds of the reticular dermis, fibrous thickening of interlobular fat septa in the superficial subcutis, and scattered lympoid follicles (A). At medium power (magnification ×31.25), the lymphoid follicles are seen near the thickened and sclerosed septa, which encroach on lobules (B). Under high magnification (magnification ×250), plasma cells are seen to be a component of the inflammatory infiltrate, involving the interstitium to the left of the lymphoid follicle (C).
A summary of the 31 published cases of postirradiation morphea in patients treated for breast cancer is provided in Table 5 [2-8,11-17]. The estimated incidence of this complication of radiotherapy is 1:500 patients [14], in contrast to that of morphea (of any etiology), which is 2.7 per 100 000 of the general population per year [7]. A difficulty in determining the exact frequency of the problem lies in the rate of accuracy of the diagnosis, and it is likely mistaken for chronic radiodermatitis or radiation fibrosis on occasion. There are no predictable risk factors for the development of postirradiation morphea, and its onset can vary from an interval of weeks to as many as 32 years after treatment. Beginning in and affecting predominantly the radiation portal, the process has, on occasion, extended to other sites. Serological studies in some patients with this condition have revealed weak positivity for antinuclear antibodies, but more specific scleroderma-related antibodies have not been found. The clinical manifestations of the cases in this series are comparable to those previously described, and the radiotherapeutic regimens used were within the standards of practice in a North American setting. Although all of our patients had invasive carcinomas of the left breast, the reported laterality of cancers in this context has been variable, and we do not attribute special significance to the site distribution in our group. The interval between the delivery of radiotherapy and the onset of morphea in the literature is broad, and the time lag of 4 to 12 years observed in our series falls within this spectrum. Extension of morphea beyond the radiation portal was observed in 1 (20%) of our 5 patients, the proportion in the literature being 29%. It is known that early lesions of postirradiation morphea are characterized by an erythematous edematous plaque, which frequently leads to a clinical suspicion of cellulitis. In our patient with early morphea (inflammatory phase), the clinical differential diagnosis included radiation recall dermatitis versus carcinoma en cuirasse. More advanced disease, characterized by induration and violaceous discoloration of the affected skin, with a peau d'orange appearance, usually has lead to a clinical impression of recurrent carcinoma, as observed in one of our patients. The outcome of postirradiation morphea in the literature has been variable [12-15]. Some cases have shown spontaneous gradual softening of the skin, others have shown a partial or complete response to corticosteroid
1686 Table 5
N. Walsh et al. Published reports of postirradiation morphea since 1989 Interval (y) between Skin disease beyond radiation port radiation and skin disease
Reference
No. of Cancer type cases
Radiation dose (Gy)/Fractions/ Boost (Gy)
Colver et al [2]
9
43-59
10-27
2-15
Forbes et al [5] Trattner et al [6] Winkelmann et al [11] Davis et al [12] Smith et al [3]
1 1 4 6 1
Gollob et al [13] Bleasel et al [14] Schaffer et al [15] Arden-Jones et al (2007) Ullen et al [4] Reddy et al [8] Dubner et al [16] Dancey et at [17]
1.5-10
4/9
ND 50 46-50 46-55 44
ND 25 23-27 25-30 22
ND 10 14-20 ND 0
b1 b1 b1 b1 b1
1/1 1/1 0/4 0/6 0/1
1 4 2 1
7 breast; 1 axilla (adenocarcinoma of unknown origin); 1 endocervical 1 breast 1 breast 4 breast 6 breast 1 subcutaneous lymphoma (HIV+) 1 breast 4 breast 2 breast 1 breast
42.5 45-50.4 43-46 ND
16 25-28 20-23 ND
0 10-16 0-18 ND
b1 b1 6.5-32 13
0/1 0/4 0/2 1/1
1 1 1 1
1 1 1 1
50 and 40 ND 50.4 ND
25 and 25 ND 28 ND
0 and 0 b1 and 5 ND b1 10 3 ND b1
breast and endometrial breast breast breast
therapy, and still others have proven recalcitrant. One of the patients in this series treated with potent topical steroids did experience gradual softening of the affected skin, although the dermatologist doubted that this was attributable to the treatment. In only rare cases in the literature was surgery contemplated as a reasonable approach to management; however, one of our patients did undergo mastectomy for symptomatic reasons. The concurrence of LS&A and morphea as seen in 2 of our cases has been reported previously in this clinical setting [9,10]. One of the cases in the literature was characterized by development of a clinically bullous lesion of LS&A, superimposed on known postirradiation morphea [10]. Although the clinical characteristics of the 2 lesions in our group are unknown, both had microscopic features suggestive of bullous LS&A. It may be that alteration of local lymphatics by the radiotherapy, possibly abetted by axillary lymphadenectomy, predisposes LS&A to become bullous in this setting. With regard to the pattern of involvement of subcutaneous fat in 2 of our cases, both showed concomitant inflammatory and sclerosing changes of morphea in the lower two thirds of the reticular dermis. Interestingly, one exhibited a lobular component to the panniculitis. This shared features in common with the entity previously reported as ‘pseudosclerodermatous panniculitis after irradiation’ [11,18]. In sequential biopsies from this patient, progressive sclerosis of the affected lobules, accompanied by a plasma cell–rich infiltrate, was observed. A minor degree of lobular panniculitis was also noted in the mastectomy specimen from the other patient, but the dominant pattern in this instance was septal, as seen typically in morphea profundus. Whether the lobular
1/1 1/1 0/1 0/1
component of the panniculitis observed by us and by others represents a variation on the theme of postirradiation morphea profundus or a separate entity is unclear. In the series of Winkelmann et al [11] and in our series, it was intimately associated with changes of morphea, and our impression is that it is part of the same process. However, in the cases reported by Carrasco et al [18], apart from concomitant fibrous septal thickening, a convincing association with morphea was not demonstrated. The pathogenesis of radiation-induced morphea is unknown, but the fact that it is distinct from and unrelated to radiodermatitis and radiation fibrosis is established. These predictable side effects of irradiation lack significant inflammatory infiltrates. Acute radiodermatitis, represented clinically by erythema, edema, epilation, and pigmentary change, usually occurs during or within weeks of treatment. It is characterized at the microscopic level by fibrin thrombi in small vessels, red blood cell extravasation, edema, and apoptosis of epidermal keratinocytes [19]. Chronic radiodermatitis, observed months to years after treatment, exhibits poikilodermatous clinical features. These are reflected microscopically by epidermal atrophy with basal vacuolar change, telangiectasia, sclerosis of the dermis with scattered atypical ‘radiation fibroblasts,’ and loss of adnexal structures [19]. Radiation fibrosis, likely part of the same spectrum of reaction as chronic radiodermatitis but involving deeper tissue, can begin as early as 10 weeks after radiotherapy. It is characterized by progressive fibrosis of deep subcutaneous tissues, such as fascia and muscle, in the radiation portal. It usually spares the skin [15]. In contrast to acute and chronic radiodermatitis, the entity described by LeBoit [19] in 1989, titled ‘subacute radiation
Postirradiation morphea dermatitis’ may in fact have a pathogenic relationship to postirradiation morphea. This entity, represented clinically by either persistent erythema or poikilodermatous change, is characterized microscopically by an interface dermatitis, which mimics acute cutaneous graft-versus-host disease. The potential link between this entity and postirradiation morphea lies in the fact that cutaneous graft-versus-host disease also has a chronic sclerodermoid stage. Moreover, the concurrence of LS&A and morphea in radiation-induced disease may constitute a link between the epidermal involvement observed in subacute radiation dermatitis and the dermal and subcutaneous involvement by morphea. To date, evolution of subacute radiation dermatitis to postirradiation morphea has not been documented. It would be overly simplistic to draw firm conclusions about the nature of the inflammatory process solely from our morphological observations at a single point in time. That said, the concurrence of T-lymphocytes and plasma cells suggests a role for both the cell-mediated and humoral immune responses in the setting of morphea. Moreover, the interplay between these systems likely changes over time. The relative prominence of T-suppressor cells, and their distribution, in a pattern which mimics that of the inflammatory stage of morphea, suggests a significant pathogenic role for this subset of T cells. Of interest, the loss of the CD34+ subpopulation of dermal dendrocytes in sclerotic zones, noted in our cases and controls, has been documented previously in morphea and scleroderma [20]. Although the specific function of these cells is unknown, their selective depletion in this setting has led to speculation that they may have pathogenic relevance. If, for example, they were to play a role in modulation of normal collagen synthesis, their loss, possibly due to an autoimmune attack, could result in the enhanced sclerotic change characteristic of morphea. It is interesting to compare this finding with those in another fibrosing disorder of the skin, nephrogenic fibrosing dermopathy (nephrogenic systemic fibrosis). In that setting, indications are that circulating CD34+, procollagen-1+ fibrocytes selectively home to the skin, thus contributing to the characteristic scleromyxedema-like change [21]. Also of interest in considering the pathogenesis of postirradiation morphea and its clinical differential diagnostic contender, radiation fibrosis, is the fact that a previous study of radiation fibroblasts has revealed that in only 1 of 16 cases examined did these cells express CD34 [22]. That the pathogenic mechanisms at play in cutaneous graft-versus-host disease may be similar to those in subacute radiation dermatitis, and postirradiation morphea seems plausible. In the past, a unifying concept to explain interface dermatitides, such as that seen in acute graft-versus-host disease, has been inappropriate expression of foreign antigens by altered epidermal cells or development of altered antigenic coding for self, stimulating an inflammatory response [19]. Likewise, those who have studied radiationinduced morphea have postulated that radiation-induced
1687 neoantigens in dermal fibroblasts or endothelial cells may stimulate a lymphocytic (T or B cell) response, leading to release of cytokines and growth factors, which in turn stimulate fibroblasts to generate excess collagen. Transforming growth factor (TGF)-beta, known to be (i) secreted by activated T-helper cells, (ii) to stimulate fibroblasts and endothelial cells, and (iii) to produce features of morphea/ scleroderma in experimental settings, has been particularly implicated in this hypothesis. The lag phase in the development of morphea after irradiation has been ascribed to the timing of exposure to an environmental antigen, such as an infection, which might stimulate immune crossreactivity with the radiation-induced neoantigens in the skin [15]. Just as the pathogenesis of cutaneous graft-versushost disease is complex, involving several effector cell populations [23], the same is likely true of postirradiation morphea. What seems clear is that the direct effect alone of radiation on fibroblasts, stimulating increased collagen synthesis, is unlikely to explain the inflammatory character of postirradiation morphea [24]. With the increasing emphasis on a breast-conserving approach to management of mammary cancer, it is likely that postirradiation morphea will continue to emerge as a side effect of treatment. An awareness of this phenomenon should help in its distinction from chronic radiodermatitis and radiation fibrosis. This would be of benefit in assuaging clinical suspicions of recurrent carcinoma after appropriate sampling. Moreover, cognizance of the entity among pathologists should enhance the accuracy of diagnosis. This, in turn, would afford an opportunity to better explore treatment options. It is conceivable that in a more confident diagnostic setting, therapeutic trials of sustained corticosteroid therapy might benefit patients from both a cosmetic and symptomatic perspective. Furthermore, other modalities of therapy, such as psoralen ultraviolet A (PUVA), reported to have a favorable effect on some cases of idiopathic morphea, might be worth exploring.
Acknowledgments The authors gratefully acknowledge the clerical assistance of Mrs Andrena Brown. They also thank Mr S. Whitefield, who took the photomicrographs, in the Department of Anatomy and Neurobiology, Dalhousie University.
References [1] Crocker HR. Diseases of the skin. Philadelphia: P Blakistons Son & Co; 1905. p. 633. [2] Colver GB, Rodger A, Mortimer PS, et al. Post-irradiation morphoea. Br J Dermatol 1989;I20:831-5. [3] Smith KJ, Yeager J, Skelton HG. Letter to the editor: localized scleroderma in breast cancer patients treated with supervoltage
1688
[4]
[5]
[6] [7]
[8]
[9] [10] [11]
[12]
[13]
external beam radiation: radiation port scleroderma. J Am Acad Dermatol 1997;37:806-7. Ullen H, Bjorkholm E. Localized scleroderma in a woman irradiated at two sites for endometrial and breast carcinoma: a case history and a review of the literature. Int J Gynecol Cancer 2003;13:77-82. Forbes AM, Woodrow JC, Verbov JL, et al. Carcinoma of breast and scleroderma: four further cases and a literature review. Br J Rheumatol 1989;28:65-9. Trattner A, Figer A, David M, et al. Circumscribed scleroderma induced by postlumpectomy radiation therapy. Cancer 1991;68:2131-3. Ardern-Jones MR, Black MM. Widespread morphea following radiotherapy for carcinoma of the breast. Clin and Exper Dermatol 2003;28:160-2. Reddy SM, Pui JC, Gold LI, et al. Postirradiation morphea and subcutaneous polyarteritis nodosa: case report and literature review. Semin Arthritis Rheum 2005;34:728-34. Yates VM, King CM, Dave VK. Lichen sclerosus et atrophicus following radiation therapy. Arch Dermatol 1985;121:1044-7. Trattner A, David M, Sandbank M. Bullous morphea: a distinct entity? Am J Dermatopathol 1994;16:414-7. Winkelmann RK, Grado GL, Quimby SR, et al. Pseudosclerodermatous panniculitis after irradiation: an unusual complication of megavoltage treatment of breast carcinoma. Mayo Clin Proc 1993;68:122-7. Davis DA, Cohen PR, McNeese MD, et al. Localized scleroderma in breast cancer patients treated with supervoltage external beam radiation: radiation port scleroderma. J Am Acad Dermatol 1996;35:923-8. Gollob MH, Dekoven JG, Bell MJ, et al. Postradiation morphea. J Rheumatol 1998;25:2267-9.
N. Walsh et al. [14] Bleasel NR, Stapleton KM, Commens C, et al. Radiation-induced localized scleroderma in breast cancer patients. Australas J Dermatol 1999;40:99-102. [15] Schaffer JV, Carroll C, Dvoretsky I, et al. Postirradiation morphea of the breast. Presentation of two cases and review of the literature. Dermatology 2000;200:67-71. [16] Dubner S, Bovi J, White J, et al. Postirradiation morphea in a breast cancer patient. The Breast Journal 2006;12:173-6. [17] Dancey AL, Waters RA. Morphea of the breast. Two case reports and discussion of the literature. J Plast Reconstr Aesthetic Surg 2006;59: 1114-7. [18] Carrasco L, Moreno C, Pastor MA, et al. Postirradiation pseudosclerodermatous panniculitis. Am J Dermatopathol 2001;23:283-7. [19] LeBoit PE. Subacute radiation dermatitis: a histologic imitator of acute cutaneous graft-versus-host disease. J Am Acad Dermatol 1989;20: 236-41. [20] McNiff JM, Glusac EJ, Lazova RZ, et al. Morphea limited to the superficial reticular dermis: an underrecognized histologic phenomenon. Am J Dermatopathol 1999;21(4):315-9. [21] Cowper SE, Su LD, Bhawan J, et al. Nephrogenic fibrosing dermopathy. Am J Dermatopathol 2001;23:383-93. [22] Meehan SA, LeBoit PE. An immunohistochemical analysis of radiation fibroblasts. J Cutan Pathol 1997;24:309-13. [23] Horn TD. Effector cells in cutaneous graft-versus-host disease. Who? What? When? Where? How? Br J Dermatol 1999;141:779-82. [24] Riekki R, Jukkola A, Sassi ML, et al. Modulation of skin collagen metabolism by irradiation: collagen synthesis is increased in irradiated human skin. Br J Dermatol 2000;142:874-80.
www.elsevier.com/locate/humpath
Original contribution
Postirradiation morphea: an underrecognized complication of treatment for breast cancer Noreen Walsh MD, MRCPI, FRCP(C), FRCPath (UK)a,⁎, Dorianne Rheaume FRCPC (Internal Medicine), FRCPC (Radiation Oncology)b , Penelope Barnes MD, FRCP(C)a , Robert Tremaine MD, FRCP(C)c , Michael Reardon MD, FRCP(C)c a
Department of Pathology (Anatomical Pathology), Capital District Health Authority and Dalhousie University, Halifax, Nova Scotia, Canada B3H 1V8 b Department of Radiation Oncology, Capital District Health Authority and Dalhousie University, Halifax, Nova Scotia, Canada B3H 1V8 c Department of Medicine (Dermatology), Capital District Health Authority and Dalhousie University, Halifax, Nova Scotia, Canada B3H 1V8 Received 9 April 2008; revised 22 April 2008; accepted 23 April 2008
Keywords: Postirradiation morphea; Localized scleroderma after radiotherapy; Cutaneous complications of irradiation for breast cancer
Summary The most common cutaneous side effects of radiotherapy include radiodermatitis and radiation fibrosis. These are influenced by the type, dose, and pattern of delivery of the treatment. Distinct from these is postirradiation morphea (localized scleroderma), an idiosyncratic treatmentrelated phenomenon. Within the last 20 years, approximately 31 examples of postirradiation morphea after treatment for breast cancer were reported. We describe 5 new cases of this entity and integrate our findings with those in the literature. The mean age of the patients at the time of diagnosis of cancer was 58 years; all were left-sided and treated by local excision of the tumor, ipsilateral axillary lymph node dissection, and local radiotherapy. After an interval of 4 to 12 years, the patients developed morphea in the radiation portals, with extension beyond it in one instance. Recurrent breast carcinoma was suspected clinically in 2 cases. Microscopically, changes of morphea involved the dermis in all cases and the subcutis in 2. There was associated lichen sclerosus et atrophicus in 2 cases. Our data about management and outcome are limited, but 1 patient treated with potent topical steroids experienced gradual softening of the affected skin over a 5-year period, whereas another had a mastectomy for relief of painful induration of the breast. Our findings support existing theories about the pathogenesis of this condition and link it to those of sclerodermoid graft-versus-host disease. The purpose of our communication is to draw attention to this underrecognized complication of treatment for breast cancer. © 2008 Elsevier Inc. All rights reserved.
1. Introduction ⁎ Corresponding author. Division of Anatomical Pathology, Halifax, Nova Scotia, Canada B3H 1V8. E-mail address: [email protected] (N. Walsh). 0046-8177/$ – see front matter © 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.humpath.2008.04.010
That morphea ranks among the local cutaneous side effects of radiotherapy was recognized shortly after the discovery of x-
Postirradiation morphea Table 1
1681
Breast cancer data in current series
Case Age Site no. (y)
Histopathologic type of cancer
Axillary Treatment LN status
1
69
Left breast
Invasive ductal carcinoma LN° (ER+; PR+)
2
51
Left breast
Invasive ductal carcinoma LN° (ER−; PR+)
3
58
Left breast
Invasive ductal carcinoma LN° (ER+ and PR+)
4
57
Left breast
5
53
Left breast
Invasive carcinoma,mixed LN+1 lobular and ductal (ER+; PR−) Invasive ductal carcinoma LN° (ER+; PR+)
Local excision; axillary LN dissection; radiotherapy Local excision; axillary LN dissection; radiotherapy Local excision; axillary LN dissection; radiotherapy Local excision; axillary LN dissection; radiotherapy Local excision; axillary LN dissection; radiotherapy
Outcome
Duration of follow-up (y)
No known metastasis
13
Right hip metastasis and death— 14 14 years after diagnosis of breast cancer. No known metastasis 10
No known metastasis
17
No known metastasis
4
Abbreviations: LN, lymph node; ER, estrogen receptor; PR, progesterone receptor.
rays by Roentgen in 1885 and noted in textbooks of the same era [1]. It received little subsequent attention until the report by Colver et al [2] in 1989, describing the phenomenon in 9 patients, of whom 7 had received radiotherapy for breast cancer. Occurring as an idiosyncratic reaction, independent of the severity of acute radiodermatitis, its temporal relationship to treatment has been inconsistent and its main significance lies in its clinical mimicry of locally recurrent carcinoma. With the benefit of emerging data, it has become clear that this complication of radiotherapy can also develop as a sequel to treatment of cancers at sites other than the breast [2-4] and that it can, on occasion, extend beyond the portal of irradiation [2,48]. Moreover, the related entity, lichen sclerosus et atrophicus (LS&A), observed not infrequently in association with morphea and considered to be part of the same spectrum of disease, has been documented to occur in this clinical setting [9,10]. The objective of this report is to outline the clinical and microscopic characteristics of postirradiation morphea in 5 patients treated for breast cancer and to distill the literature on the subject. Our aims are (і) to highlight the diagnostic challenges presented by this entity, (іі) to expand existing knowledge of a rare but important phenomenon, and (ііі) to contribute to existing hypotheses about its pathogenesis.
2. Methods Five cases of postirradiation morphea were encountered between 1996 and 2008 in the pathology practices of 2 of the authors. Case retrieval was via the laboratory information system (Cerner Classic). Clinical data pertaining to the patients' breast cancers, diagnosed between 1991 and 2004, related treatment and subsequent skin disease were obtained with the assistance of clinicians. These included a radiation oncologist, 2 dermatologists, and a general surgeon. Three control cases of idiopathic morphea were selected from the
pathology files for the purpose of comparative histopathologic and immunohistochemical evaluation. Ten tissue samples (9 skin biopsies and 1 mastectomy specimen) from the 5 study patients and 4 biopsies from the 3 control cases were available for review. Categorization of the morphological findings in each biopsy, as (i) morphea alone or (ii) morphea in combination with LS&A, was based on routine histopathologic sections. Moreover, the stage of morphea, as (a) inflammatory, (b) inflammatory and sclerosing, or (c) sclerosing, was documented. Selected histochemical stains were performed to better explore stromal features of the biopsies. These included an Alcian blue stain (Ventana automated stainer; pH 2.5; VEntana, Tuscon, AZ) in search of excess interstitial dermal mucin and the modified Hart stain (Ventana automated stainer) to highlight the pattern of dermal elastic fibers. A panel of immunohistochemical studies was applied, mainly to determine the phenotype of the cellular infiltrate. The panel consisted of stains for CD3 (Ventana; prediluted; CC1 pretreatment × 30 min), CD20 (Ventana; prediluted; CC1 pretreatment × 8 min), CD4 (Labvision, Fremont, CA; 1/20; CC1 pretreatment × 60 min), CD8 (Labvision; 1/100; CC1 pretreatment × 30 min), CD30 (Dako, Mississauga, Ontario; 1/ 50; protease pretreatment × 4 min), CD68 (Dako; 1/100; no pretreatment), CD34 (Ventana; prediluted; no pretreatment), CD31 (Dako; 1/100; CC1 pretreatment × 8 min), c-kit (Ventana; prediluted; CC1 pretreatment × 30 mns), kappa and lambda light chains (Dako; 1/20,000; CC1 pretreatment × 30 min), and S100 protein (Dako; 1/5000; no pretreatment).
3. Results 3.1. Clinical findings Clinical data related to the 5 patients, their breast cancers, radiotherapeutic regimens, and postirradiation morphea are
1682
N. Walsh et al.
Table 2
Details of radiotherapy for breast cancer in current series
Case no.
Radiation dose (cGy)
Fractions
Dose/ Fraction
Total treatment time (d)
Biologic effective dose (cGy)
Energy (MV)
Patient separation (cm)
Max dose (%)
Boost (cGy/fr)
Observations during radiotherapy
1
4000
15
267
22
75.6
23.5
107
250/2
2
5000
25
200
45
83.5
20.5
102
1000/5
3
5000
25
200
37
83.5
1.25 (cobalt) 1.25 (cobalt) 6
17.2
4
4000
16
250
22
73.2
16.3
Not known 103
5
4240
16
265
Not known
Not known
28.2
108.4
No boost No boost 900/4
Induration in the operative area Mild-Moderate radiation erythema Slight nipple and breast tenderness Minimal radiation erythema Significant erythema and moist desquamation around nipple
1.25 (Cobalt) 6 and 18
outlined in Tables 1-3, respectively. They are summarized here. At the time of the diagnosis of breast cancer, the patients ranged in age from 51 to 69 years (mean = 58). All 5 cancers involved the left breast, 4 being invasive ductal carcinomas and 1 an invasive mixed lobular and ductal carcinoma. All 5 tumors were estrogen and/or progesterone receptor–positive. Only in 1 case was metastatic disease detected in a regional lymph node at the outset, and the management of all patients included local excision of the tumor, axillary lymph node excision, and radiotherapy. Based on the hormone sensitivity of the cancers, it is highly probable that the patients received antiestrogen therapy; however, documentation of this was obtained in only 1 case. In a follow-up period ranging from 4 to 17 years, 1 of the 5 patients is known to have died of metastatic carcinoma.
Table 3
Details of the radiotherapeutic regimens used are outlined in Table 2. The dose of radiation ranged from 4000 to 5000 cGy, delivered in fractionated doses of 200 to 267 cGy. The total treatment time ranged from 22 to 45 days, and 3 of the 5 patients received a boost, as indicated when the tumor closely approaches a resection margin. Other technical details reflect the fact that in the 2 more recent cases (cases 3 and 5), treated within the last 10 years, high-energy radiation was used as is now the norm for deep-seated tumors. In the remaining 3 cases (cases 1, 2, and 4), treated more than 10 years ago, Cobalt, with the delivery of lowerenergy radiation, was employed in accordance with historical standards of practice. Calculations of the maximum dose and patient separation reflect the planning process for radiotherapy, geared to optimize the amount and localization of
Details of skin disease after radiotherapy for breast cancer in current series
Case Interval (y) Site of skin disease Review by Clinical no. since diagnosis of dermatologist differential breast cancer diagnosis 1
6
Radiation field only Yes
2
4
Radiation field only Yes
3
8
Radiation field and extension to right side of chest
4
12
Radiation field only No
5
4
Radiation field only No
Yes
Histopathologic diagnosis
Treatment
Unknown LS&A and morphea (inflammatory and sclerosing stage) Recurrent breast Morphea (inflammatory Potent topical carcinoma and sclerosing stage) steroid (Dermovate) None Radiation recall Morphea (inflammatory stage) dermatitis; carcinoma en cuirasse Unknown Not provided LS&A and morphea (inflammatory and sclerosing stage) Not provided Morphea (inflammatory Mastectomy and sclerosing stage) LS&A
Outcome
Unknown
Softening of skin over 5 y Unknown
Unknown
Insufficient follow-up to determine outcome
Postirradiation morphea
1683 chest, generating a differential diagnosis of radiation recall dermatitis versus carcinoma en cuirasse. Of the remaining 2 cases, a detailed clinical description of the skin change was not provided in one (case 4), and in the other (case 5), there was painful, induration of the irradiated breast. One patient (case 2), treated with a potent topical steroid, experienced gradual softening of the involved skin over a 5-year period. The dermatologist was not convinced that the clinical improvement was therapy related. Another (case 3) received no treatment, and a third (case 5) was managed by mastectomy for symptomatic relief. The management and outcome of the 2 remaining patients are unknown.
3.2. Histopathologic and immunohistochemical findings in skin
Fig. 1 A clinical photograph in case 2 reveals a mottled erythematous plaque involving the left previously irradiated breast, in the area superior and medial to the nipple (A). A close-up shot of the plaque, above the nipple, displays a puckered peau d'orange appearance, as well as the site of a previous punch biopsy (B).
the treatment. Acute radiation dermatitis was observed in 3 patients during treatment, being mild in 2 cases and more pronounced in the third. Induration at the operative site was noted in 1 patient, and another experienced slight nipple and breast tenderness. Clinical data pertaining to the development of morphea in the radiation portals (and beyond it in 1 instance) are outlined in Table 3. This occurred 4 to 12 years after the diagnosis and treatment of breast cancer. Of the 5 patients, 3 were seen by a dermatologist. In one of these (case 1), a description of the clinical lesion was not provided, but the impression was of LS&A. In another (case 2), the skin of the irradiated breast exhibited an erythematous indurated plaque with a peau d'orange appearance, leading to a clinical suspicion of locally recurrent carcinoma (Fig. 1). In the third (case 3), an erythematous edematous plaque was seen to involve the portal of irradiation, with extension to the right side of the
The microscopic changes in biopsies from the study patients were similar to those from the controls, and pertinent findings are outlined in Table 4. To briefly comment on the stages of morphea, this usually evolves from an inflammatory to a sclerosing phase, with an intervening mixed inflammatory and sclerosing phase. The inflammatory changes are characterized by a superficial and deep perivascular and interstitial lymphoplasmacytic infiltrate, whereas the sclerosing changes are represented by thickening of collagen bundles in the reticular dermis. Involvement of the subcutis, referred to as morphea profundus, is typified by collagenous thickening of septa with accompanying lymphoplasmacytic inflammation. On the other hand, LS&A is characterized by edema and hyalinization of the papillary dermis, an underlying nodular to band-like lymphocytic infiltrate, and epidermal thinning with hyperkeratosis. To subcategorize the morphological findings in our group, 2 of the 5 study cases showed features of combined morphea and LS&A and 3 of morphea alone. Of the 3 control cases, 2 exhibited changes of combined morphea and LS&A and 1 of morphea alone. With the exception of 1 study case (case 3) in which the morphea was in the inflammatory phase, the process was in the mixed inflammatory and sclerosing stage in all cases and controls. In the 2 study cases with LS&A (cases 1 and 4), a horizontally oriented space was evident in the edematous/hyalinized papillary dermis, consistent with bullous LS&A. Insufficient detail was available in those cases to determine whether the process was vesiculobullous clinically. A rare multinucleated histiocyte with elastophagocytosis was noted in 1 study and 1 control case. The pattern of coarse elastic fibers in the reticular dermis was not appreciably altered in any of the samples, but a decrease in density of the delicate fibers in the papillary dermis was observed in the setting of LS&A. Excess interstitial dermal mucin deposition was not observed. The phenotypic characteristics of the inflammatory infiltrate were similar in biopsies from the study patients and the controls. It was T-cell predominant, with CD3positive cells, in general, accounting for more than 90% and
1684 Table 4
N. Walsh et al. Postirradiation morphea: summary of histopathologic findings
Data No. of Patients (biopsies) Categorization of findings Main characteristics of dermal infiltrate T cells (CD3+) B cells (CD20+) b T-helper/T-suppressor (CD4:CD8) Plasma cells Distribution pattern of T cells T-helper (CD4) T-suppressor (CD8+) Distribution pattern of CD34+ dermal dendritic cells Pattern of panniculitis observed Sequential observations of panniculitis (case 2)
Study cases
Control cases
a
5 (10) 1 morphea (I); 2 morphea (I & S); 2 morphea (I & S) + LS&A
3 (4) 1 morphea (I & S ); 2 morphea (I & S) + LS&A
≥90% ≤10% 60:40 Scattered/polyclonal
≥90% ≤10% 60:40 Scattered/polyclonal
Perivascular Perivascular, interstitial and DE junction Absent in sclerotic zones, reticular dermis
Perivascular Perivascular, interstitial and DE junction Absent in sclerotic zones, reticular dermis
Septal and lobular (case 2) c predominantly septal (case 5) Sclerotic replacement of subcutis. Increased proportion of plasma cells to lymphocytes.
N/A N/A
Abbreviations: I, inflammatory; I & S, inflammatory and sclerosing; DE junction, dermoepidermal junction; N/A, not applicable. a Of the 10 tissue samples, 1 was a mastectomy specimen. b Only in 1 study case (case 5) did the B-cell component of the infiltrate exceed 10% and amount to approximately 30%. c The pattern of panniculitis observed in this case resembled that previously reported as “pseudosclerodermatous panniculitis after irradiation.”
CD20-positive cells for less than 10% of the infiltrate. Only in case 5, where lymphoid follicles and germinal centers were present, was the B-cell component more prominent, amounting to an estimated 30% of the infiltrate. The approximate T-helper (CD4) to T-suppressor (CD8) cell ratio was 60:40. The T-helper cells were largely confined to perivascular zones, whereas the T-suppressors were perivascular and interstitial in distribution and also involved the dermoepidermal junction. Small numbers of plasma cells (b5% of the infiltrate) were generally present, and these showed expression of both kappa and lambda light chains. In 1 case (case 2), successive biopsies revealed an increase in the plasma cell component over time. C-kit–positive mast cells and CD68-positive histiocytes were scattered throughout the dermis in an unremarkable manner. No CD30positive cells were observed. The endothelial cells of dermal blood vessels, highlighted by stains for CD31 and CD34, appeared normal. With the exception of an increase in the number of capillaries in the papillary dermis in the cases of combined morphea and LS&A, no vascular abnormalities were appreciated. In both the study and control cases, the CD34-positive subpopulation dermal dendritic cells were reduced or absent exclusively in sclerotic zones. S100 protein–positive Langerhans cells were present in normal numbers in the epidermis and were also scattered among perivascular inflammatory cells. Five biopsies taken from a single patient (case 2) at 4 different points in time over an 18-month period were of interest. These had been obtained because of a persistent clinical suspicion of recurrent breast cancer despite histo-
pathologic findings of an inflammatory process. The diagnosis of postirradiation morphea had not been established until the last biopsy was examined. From the outset of sampling, the disease process was in the mixed inflammatory and sclerosing phase, affecting the lower two thirds of the reticular dermis and the subcutis. The mixed pattern persisted over 18 months, but more advanced sclerosis became evident with time as did an increase in the ratio of plasma cells to lymphocytes. In 2 of the 5 biopsies separated by a 16-month interval, there was sufficient subcutaneous fat to enable study of the disease process in this compartment. Unlike the dominance of septal involvement typically seen in idiopathic morphea profundus, the pattern of panniculitis observed in the first of the 2 biopsies from this study patient, with postirradiation morphea, revealed a mixed septal and lobular pattern. This was characterized by a lobular infiltrate of lymphocytes, plasma cells, and foamy macrophages, in addition to fibrous septal thickening. In the biopsy taken 16 months later, there was marked hyalinization of the subcutaneous compartment, accompanied by a plasma cell predominant infiltrate. The availability of abundant tissue from the patient treated by mastectomy, and generous sampling, yielded an opportunity to extensively study the pattern of the disease in this case (case 5). It affected the skin and superficial subcutis of the whole irradiated breast. In each of the 12 sections examined, the morphea was in the mixed inflammatory and sclerosing phase and involved the lower two thirds of the reticular dermis and the subcutaneous fat. In contrast to findings in the other cases and the controls, the inflammatory
Postirradiation morphea
1685 component in this instance revealed scattered lymphoid follicles with germinal centers. The pattern of involvement of subcutaneous fat, in foci, resembled the mixed lobular and septal pattern noted in case 2, but the septal pattern characteristic of morphea profundus predominated (Fig. 2). The underlying mammary tissue appeared normal.
4. Discussion
Fig. 2 A photomicrograph, at scanning magnification (magnification ×5.5), of one of the sections from the mastectomy specimen in case 5 shows condensation of collagen in the lower two thirds of the reticular dermis, fibrous thickening of interlobular fat septa in the superficial subcutis, and scattered lympoid follicles (A). At medium power (magnification ×31.25), the lymphoid follicles are seen near the thickened and sclerosed septa, which encroach on lobules (B). Under high magnification (magnification ×250), plasma cells are seen to be a component of the inflammatory infiltrate, involving the interstitium to the left of the lymphoid follicle (C).
A summary of the 31 published cases of postirradiation morphea in patients treated for breast cancer is provided in Table 5 [2-8,11-17]. The estimated incidence of this complication of radiotherapy is 1:500 patients [14], in contrast to that of morphea (of any etiology), which is 2.7 per 100 000 of the general population per year [7]. A difficulty in determining the exact frequency of the problem lies in the rate of accuracy of the diagnosis, and it is likely mistaken for chronic radiodermatitis or radiation fibrosis on occasion. There are no predictable risk factors for the development of postirradiation morphea, and its onset can vary from an interval of weeks to as many as 32 years after treatment. Beginning in and affecting predominantly the radiation portal, the process has, on occasion, extended to other sites. Serological studies in some patients with this condition have revealed weak positivity for antinuclear antibodies, but more specific scleroderma-related antibodies have not been found. The clinical manifestations of the cases in this series are comparable to those previously described, and the radiotherapeutic regimens used were within the standards of practice in a North American setting. Although all of our patients had invasive carcinomas of the left breast, the reported laterality of cancers in this context has been variable, and we do not attribute special significance to the site distribution in our group. The interval between the delivery of radiotherapy and the onset of morphea in the literature is broad, and the time lag of 4 to 12 years observed in our series falls within this spectrum. Extension of morphea beyond the radiation portal was observed in 1 (20%) of our 5 patients, the proportion in the literature being 29%. It is known that early lesions of postirradiation morphea are characterized by an erythematous edematous plaque, which frequently leads to a clinical suspicion of cellulitis. In our patient with early morphea (inflammatory phase), the clinical differential diagnosis included radiation recall dermatitis versus carcinoma en cuirasse. More advanced disease, characterized by induration and violaceous discoloration of the affected skin, with a peau d'orange appearance, usually has lead to a clinical impression of recurrent carcinoma, as observed in one of our patients. The outcome of postirradiation morphea in the literature has been variable [12-15]. Some cases have shown spontaneous gradual softening of the skin, others have shown a partial or complete response to corticosteroid
1686 Table 5
N. Walsh et al. Published reports of postirradiation morphea since 1989 Interval (y) between Skin disease beyond radiation port radiation and skin disease
Reference
No. of Cancer type cases
Radiation dose (Gy)/Fractions/ Boost (Gy)
Colver et al [2]
9
43-59
10-27
2-15
Forbes et al [5] Trattner et al [6] Winkelmann et al [11] Davis et al [12] Smith et al [3]
1 1 4 6 1
Gollob et al [13] Bleasel et al [14] Schaffer et al [15] Arden-Jones et al (2007) Ullen et al [4] Reddy et al [8] Dubner et al [16] Dancey et at [17]
1.5-10
4/9
ND 50 46-50 46-55 44
ND 25 23-27 25-30 22
ND 10 14-20 ND 0
b1 b1 b1 b1 b1
1/1 1/1 0/4 0/6 0/1
1 4 2 1
7 breast; 1 axilla (adenocarcinoma of unknown origin); 1 endocervical 1 breast 1 breast 4 breast 6 breast 1 subcutaneous lymphoma (HIV+) 1 breast 4 breast 2 breast 1 breast
42.5 45-50.4 43-46 ND
16 25-28 20-23 ND
0 10-16 0-18 ND
b1 b1 6.5-32 13
0/1 0/4 0/2 1/1
1 1 1 1
1 1 1 1
50 and 40 ND 50.4 ND
25 and 25 ND 28 ND
0 and 0 b1 and 5 ND b1 10 3 ND b1
breast and endometrial breast breast breast
therapy, and still others have proven recalcitrant. One of the patients in this series treated with potent topical steroids did experience gradual softening of the affected skin, although the dermatologist doubted that this was attributable to the treatment. In only rare cases in the literature was surgery contemplated as a reasonable approach to management; however, one of our patients did undergo mastectomy for symptomatic reasons. The concurrence of LS&A and morphea as seen in 2 of our cases has been reported previously in this clinical setting [9,10]. One of the cases in the literature was characterized by development of a clinically bullous lesion of LS&A, superimposed on known postirradiation morphea [10]. Although the clinical characteristics of the 2 lesions in our group are unknown, both had microscopic features suggestive of bullous LS&A. It may be that alteration of local lymphatics by the radiotherapy, possibly abetted by axillary lymphadenectomy, predisposes LS&A to become bullous in this setting. With regard to the pattern of involvement of subcutaneous fat in 2 of our cases, both showed concomitant inflammatory and sclerosing changes of morphea in the lower two thirds of the reticular dermis. Interestingly, one exhibited a lobular component to the panniculitis. This shared features in common with the entity previously reported as ‘pseudosclerodermatous panniculitis after irradiation’ [11,18]. In sequential biopsies from this patient, progressive sclerosis of the affected lobules, accompanied by a plasma cell–rich infiltrate, was observed. A minor degree of lobular panniculitis was also noted in the mastectomy specimen from the other patient, but the dominant pattern in this instance was septal, as seen typically in morphea profundus. Whether the lobular
1/1 1/1 0/1 0/1
component of the panniculitis observed by us and by others represents a variation on the theme of postirradiation morphea profundus or a separate entity is unclear. In the series of Winkelmann et al [11] and in our series, it was intimately associated with changes of morphea, and our impression is that it is part of the same process. However, in the cases reported by Carrasco et al [18], apart from concomitant fibrous septal thickening, a convincing association with morphea was not demonstrated. The pathogenesis of radiation-induced morphea is unknown, but the fact that it is distinct from and unrelated to radiodermatitis and radiation fibrosis is established. These predictable side effects of irradiation lack significant inflammatory infiltrates. Acute radiodermatitis, represented clinically by erythema, edema, epilation, and pigmentary change, usually occurs during or within weeks of treatment. It is characterized at the microscopic level by fibrin thrombi in small vessels, red blood cell extravasation, edema, and apoptosis of epidermal keratinocytes [19]. Chronic radiodermatitis, observed months to years after treatment, exhibits poikilodermatous clinical features. These are reflected microscopically by epidermal atrophy with basal vacuolar change, telangiectasia, sclerosis of the dermis with scattered atypical ‘radiation fibroblasts,’ and loss of adnexal structures [19]. Radiation fibrosis, likely part of the same spectrum of reaction as chronic radiodermatitis but involving deeper tissue, can begin as early as 10 weeks after radiotherapy. It is characterized by progressive fibrosis of deep subcutaneous tissues, such as fascia and muscle, in the radiation portal. It usually spares the skin [15]. In contrast to acute and chronic radiodermatitis, the entity described by LeBoit [19] in 1989, titled ‘subacute radiation
Postirradiation morphea dermatitis’ may in fact have a pathogenic relationship to postirradiation morphea. This entity, represented clinically by either persistent erythema or poikilodermatous change, is characterized microscopically by an interface dermatitis, which mimics acute cutaneous graft-versus-host disease. The potential link between this entity and postirradiation morphea lies in the fact that cutaneous graft-versus-host disease also has a chronic sclerodermoid stage. Moreover, the concurrence of LS&A and morphea in radiation-induced disease may constitute a link between the epidermal involvement observed in subacute radiation dermatitis and the dermal and subcutaneous involvement by morphea. To date, evolution of subacute radiation dermatitis to postirradiation morphea has not been documented. It would be overly simplistic to draw firm conclusions about the nature of the inflammatory process solely from our morphological observations at a single point in time. That said, the concurrence of T-lymphocytes and plasma cells suggests a role for both the cell-mediated and humoral immune responses in the setting of morphea. Moreover, the interplay between these systems likely changes over time. The relative prominence of T-suppressor cells, and their distribution, in a pattern which mimics that of the inflammatory stage of morphea, suggests a significant pathogenic role for this subset of T cells. Of interest, the loss of the CD34+ subpopulation of dermal dendrocytes in sclerotic zones, noted in our cases and controls, has been documented previously in morphea and scleroderma [20]. Although the specific function of these cells is unknown, their selective depletion in this setting has led to speculation that they may have pathogenic relevance. If, for example, they were to play a role in modulation of normal collagen synthesis, their loss, possibly due to an autoimmune attack, could result in the enhanced sclerotic change characteristic of morphea. It is interesting to compare this finding with those in another fibrosing disorder of the skin, nephrogenic fibrosing dermopathy (nephrogenic systemic fibrosis). In that setting, indications are that circulating CD34+, procollagen-1+ fibrocytes selectively home to the skin, thus contributing to the characteristic scleromyxedema-like change [21]. Also of interest in considering the pathogenesis of postirradiation morphea and its clinical differential diagnostic contender, radiation fibrosis, is the fact that a previous study of radiation fibroblasts has revealed that in only 1 of 16 cases examined did these cells express CD34 [22]. That the pathogenic mechanisms at play in cutaneous graft-versus-host disease may be similar to those in subacute radiation dermatitis, and postirradiation morphea seems plausible. In the past, a unifying concept to explain interface dermatitides, such as that seen in acute graft-versus-host disease, has been inappropriate expression of foreign antigens by altered epidermal cells or development of altered antigenic coding for self, stimulating an inflammatory response [19]. Likewise, those who have studied radiationinduced morphea have postulated that radiation-induced
1687 neoantigens in dermal fibroblasts or endothelial cells may stimulate a lymphocytic (T or B cell) response, leading to release of cytokines and growth factors, which in turn stimulate fibroblasts to generate excess collagen. Transforming growth factor (TGF)-beta, known to be (i) secreted by activated T-helper cells, (ii) to stimulate fibroblasts and endothelial cells, and (iii) to produce features of morphea/ scleroderma in experimental settings, has been particularly implicated in this hypothesis. The lag phase in the development of morphea after irradiation has been ascribed to the timing of exposure to an environmental antigen, such as an infection, which might stimulate immune crossreactivity with the radiation-induced neoantigens in the skin [15]. Just as the pathogenesis of cutaneous graft-versushost disease is complex, involving several effector cell populations [23], the same is likely true of postirradiation morphea. What seems clear is that the direct effect alone of radiation on fibroblasts, stimulating increased collagen synthesis, is unlikely to explain the inflammatory character of postirradiation morphea [24]. With the increasing emphasis on a breast-conserving approach to management of mammary cancer, it is likely that postirradiation morphea will continue to emerge as a side effect of treatment. An awareness of this phenomenon should help in its distinction from chronic radiodermatitis and radiation fibrosis. This would be of benefit in assuaging clinical suspicions of recurrent carcinoma after appropriate sampling. Moreover, cognizance of the entity among pathologists should enhance the accuracy of diagnosis. This, in turn, would afford an opportunity to better explore treatment options. It is conceivable that in a more confident diagnostic setting, therapeutic trials of sustained corticosteroid therapy might benefit patients from both a cosmetic and symptomatic perspective. Furthermore, other modalities of therapy, such as psoralen ultraviolet A (PUVA), reported to have a favorable effect on some cases of idiopathic morphea, might be worth exploring.
Acknowledgments The authors gratefully acknowledge the clerical assistance of Mrs Andrena Brown. They also thank Mr S. Whitefield, who took the photomicrographs, in the Department of Anatomy and Neurobiology, Dalhousie University.
References [1] Crocker HR. Diseases of the skin. Philadelphia: P Blakistons Son & Co; 1905. p. 633. [2] Colver GB, Rodger A, Mortimer PS, et al. Post-irradiation morphoea. Br J Dermatol 1989;I20:831-5. [3] Smith KJ, Yeager J, Skelton HG. Letter to the editor: localized scleroderma in breast cancer patients treated with supervoltage
1688
[4]
[5]
[6] [7]
[8]
[9] [10] [11]
[12]
[13]
external beam radiation: radiation port scleroderma. J Am Acad Dermatol 1997;37:806-7. Ullen H, Bjorkholm E. Localized scleroderma in a woman irradiated at two sites for endometrial and breast carcinoma: a case history and a review of the literature. Int J Gynecol Cancer 2003;13:77-82. Forbes AM, Woodrow JC, Verbov JL, et al. Carcinoma of breast and scleroderma: four further cases and a literature review. Br J Rheumatol 1989;28:65-9. Trattner A, Figer A, David M, et al. Circumscribed scleroderma induced by postlumpectomy radiation therapy. Cancer 1991;68:2131-3. Ardern-Jones MR, Black MM. Widespread morphea following radiotherapy for carcinoma of the breast. Clin and Exper Dermatol 2003;28:160-2. Reddy SM, Pui JC, Gold LI, et al. Postirradiation morphea and subcutaneous polyarteritis nodosa: case report and literature review. Semin Arthritis Rheum 2005;34:728-34. Yates VM, King CM, Dave VK. Lichen sclerosus et atrophicus following radiation therapy. Arch Dermatol 1985;121:1044-7. Trattner A, David M, Sandbank M. Bullous morphea: a distinct entity? Am J Dermatopathol 1994;16:414-7. Winkelmann RK, Grado GL, Quimby SR, et al. Pseudosclerodermatous panniculitis after irradiation: an unusual complication of megavoltage treatment of breast carcinoma. Mayo Clin Proc 1993;68:122-7. Davis DA, Cohen PR, McNeese MD, et al. Localized scleroderma in breast cancer patients treated with supervoltage external beam radiation: radiation port scleroderma. J Am Acad Dermatol 1996;35:923-8. Gollob MH, Dekoven JG, Bell MJ, et al. Postradiation morphea. J Rheumatol 1998;25:2267-9.
N. Walsh et al. [14] Bleasel NR, Stapleton KM, Commens C, et al. Radiation-induced localized scleroderma in breast cancer patients. Australas J Dermatol 1999;40:99-102. [15] Schaffer JV, Carroll C, Dvoretsky I, et al. Postirradiation morphea of the breast. Presentation of two cases and review of the literature. Dermatology 2000;200:67-71. [16] Dubner S, Bovi J, White J, et al. Postirradiation morphea in a breast cancer patient. The Breast Journal 2006;12:173-6. [17] Dancey AL, Waters RA. Morphea of the breast. Two case reports and discussion of the literature. J Plast Reconstr Aesthetic Surg 2006;59: 1114-7. [18] Carrasco L, Moreno C, Pastor MA, et al. Postirradiation pseudosclerodermatous panniculitis. Am J Dermatopathol 2001;23:283-7. [19] LeBoit PE. Subacute radiation dermatitis: a histologic imitator of acute cutaneous graft-versus-host disease. J Am Acad Dermatol 1989;20: 236-41. [20] McNiff JM, Glusac EJ, Lazova RZ, et al. Morphea limited to the superficial reticular dermis: an underrecognized histologic phenomenon. Am J Dermatopathol 1999;21(4):315-9. [21] Cowper SE, Su LD, Bhawan J, et al. Nephrogenic fibrosing dermopathy. Am J Dermatopathol 2001;23:383-93. [22] Meehan SA, LeBoit PE. An immunohistochemical analysis of radiation fibroblasts. J Cutan Pathol 1997;24:309-13. [23] Horn TD. Effector cells in cutaneous graft-versus-host disease. Who? What? When? Where? How? Br J Dermatol 1999;141:779-82. [24] Riekki R, Jukkola A, Sassi ML, et al. Modulation of skin collagen metabolism by irradiation: collagen synthesis is increased in irradiated human skin. Br J Dermatol 2000;142:874-80.