Treatment and outcomes for primary cutaneous extramedullary plasmacytoma : a case series

Background Primary cutaneous plasmacytoma (pcp) is a rare disease, with few studies to guide therapy. Our primary study objective was to define treatments used for pcp; a secondary objective was to describe outcomes of patients, including disease recurrence and death. Methods An institutional cancer registry was used to identify cases for retrospective chart review. In a systematic review, treatments for, and outcomes of, all known cases of pcp were described. Results Three eligible cases identified at our institution; each patient had a solitary pcp. The systematic review identified 66 patients. Radiotherapy was the most commonly used primary treatment modality (31% of all patients; 42% for patients with solitary lesions), followed by surgery (28% of all patients; 36% for patients with solitary lesions). Median survival for all patients was 10.4 years [95% ci: 4.3 years to not reached], with a trend toward a decreased risk of death with solitary lesions compared with multiple lesions (hazard ratio: 0.37; 95% ci: 0.13 to 1.08; p = 0.059). For patients with solitary lesions, the median and recurrence-free survivals were, respectively, 17.0 years (95% ci: 1.7 years to not reached) and 11.0 years (95% ci: 2 years to not reached); for patients with multiple lesions, they were 4.3 years (95% ci: 1.3 to not reached) and 1.4 years (95% ci: 0.6 years to not reached). Disease recurrence, including progression to multiple myeloma, was the most common cause of death. Conclusions Compared with patients having multiple pcp lesions, those presenting with a single pcp lesion might experience longer overall survival. Local therapy (radiation or surgery) is a reasonable curative treatment for a solitary pcp lesion.


INTRODUCTION
Primary cutaneous plasmacytoma (pcp) is a rare diagnosis, with fewer than 50 cases estimated to be have been reported and compiled in the literature [1][2][3] .Cutaneous deposits of plasma cells are more commonly seen in the context of multiple myeloma, where such deposits are associated with poor prognosis 4,5 .However, because of good long-term outcomes, treatment for plasmacytoma is typically approached with curative intent; for solitary lesions, local treatments such as radiotherapy (rt) or surgery play an important role 6 .As well, compared with osseous plasmacytomas, soft-tissue plasmacytomas are associated with a lower rate of progression to multiple myeloma 7 .Plasmacytoma of the integumentary system (a non-osseous organ) is classified as an extramedullary plasmacytoma.
Because of pcp's rarity, little is known about its treatment, which can vary widely, ranging from surgical excision to systemic therapy and rt 1,2 .Furthermore, details of treatment are poorly described; many case reports focus only on the pathologic diagnosis 3,[8][9][10][11] .Although extramedullar y plasmacytomas are associated with good clinical outcomes, patients with pcp can develop local recurrence, distant skin recurrence, or systemic progression to multiple myeloma 1,2 .Some authors have suggested an association with improved clinical outcomes in patients with a solitary pcp lesion compared with multiple pcp lesions 1,[12][13][14] .

Statistics
The rfs was defined as time from pathologic diagnosis of the index lesion to recurrence or progression of plasmacytoma, development of multiple myeloma, or death.
Persistent disease after treatment was counted as a recurrence at time of treatment.The os was defined as time from pathologic diagnosis of the index lesion to death.Patients lost to follow-up were censored.Kaplan-Meier curves were constructed for rfs and os.Univariate hazard ratios with 95% confidence intervals (cis) were calculated using a Cox proportional hazards model.All statistical analyses were performed using the SAS (version 9.4: SAS Institute, Cary, NC, U.S.A.) and R (version 3.2.2:The R Foundation, Vienna, Austria) software applications.

Chart Review
For the 3 eligible patients identified from the cancer registry (Table i), median age was 48 years (range: 48-69 years).Each patient had a solitary pcp.One patient received surgery followed by adjuvant rt; the other two patients received definitive rt alone.

Patient 1
A 48-year-old man presented in May 1979 with a single lesion over the right scapula that had doubled in size since March.No constitutional symptoms were reported.The patient had a non-contributory past medical history, but was a heavy smoker and drinker.The subcutaneous tumour was surgically removed by wide local excision on 9 May 1979.The biopsy was signed out as a "malignant plasmacytoma with infiltration of skin and subcutaneous adipose tissue from back," with negative margins.Diagnostic investigations were negative for multiple myeloma, with no Bence Jones protein in urine, and normal calcium (exact value not available), albumin (45 g/L), total protein (exact value not available), creatinine [1.0 mgm% (88 mmol/L)], and hemoglobin (141 g/L).Serum protein electrophoresis noted a very small kappa immunoglobulin G component before surgery.Bone marrow biopsy and skeletal survey were normal.
Adjuvant rt was recommended to reduce the risk of local recurrence.A total of 20 Gy in 5 daily fractions using a 250 kVp orthovoltage beam was prescribed [1.1 mm Cu hvl (copper half-value layer), 1 field; Figure 1 Diagnostic investigations were negative for multiple myeloma, with no Bence Jones protein in urine, and normal serum calcium, albumin, total protein, creatinine (exact values not available), and hemoglobin (143 g/L).Serum protein electrophoresis did note a lambda immunoglobulin A spike; no bone marrow biopsy was obtained, however.Skeletal survey was normal.
Definitive rt was recommended, to a total dose of 25 Gy in 10 daily fractions, using cobalt-60 external-beam rt.Treatment was delivered using a single anterior field encompassing the right orbit, with a pencil eye shield to block the anterior chamber [Figure 1(B)].Radiation was completed on 19 August 1983.
During follow-up, the patient experienced complete clinical response of the right eyelid lesion.Normal extraocular movements were noted, with no scarring over the eyelid; however, grade 1 dry eye was documented in 1989.Serum immunoglobulin G (normal range: 5-16 g/L) rose from 14 g/L (1986) to 21 g/L (1988) and to 24 g/L (1990).The patient remained asymptomatic, with normal complete blood count, calcium, and albumin.No bone marrow biopsies were obtained.The patient was lost to follow-up after last contact on 17 July 1991, 96 months after initial diagnosis.

Patient 3
A 69-year-old man presented in November 2014 with a solitary 4.5×1.5-cmcutaneous lesion on the left lateral thoracic wall [Figure 2(A)].The patient had no myalgia, bone pain, or constipation; he had a Zubrod performance score of 0. A prior history of coronary artery disease and asthma was documented.A biopsy on 15 December 2014 showed "dense lymphoplasmacytic infiltrate."No immunohistochemistry staining was done.A repeat biopsy on 13 January 2015 showed "monotypic plasma cell infiltrate" positive for CD138 and kappa immunoglobulin G, with "only few cells positive for lambda." Diagnostic investigations in February 2015 were negative for multiple myeloma, with no Bence Jones protein in urine, and normal calcium [2.38 mmol/L (normal range: 2.20-2.62mmol/L)], albumin [39 g/L (normal range: 38-50 g/L)], creatinine [74 μmol/L (normal range: ≤109 μmol/L)], and hemoglobin [158 g/L (normal range: 148-180 g/L)].Serum protein electrophoresis did not reveal any monoclonal protein spike; β 2 -microglobulin was 2.9 mg/L (normal range: 0.6-2.3mg/L), and the free kappa/ lambda ratio was 1.38 (normal range: 0.26-1.65).Bone marrow biopsy was normal, with less than 5% plasma cells; no clonal restriction was observed, although incidental megakaryocytic thrombocytopenia was noted.Computed tomography imaging showed no lymphadenopathy or lytic lesions; skeletal survey showed no lytic lesions.
The patient received definitive rt to 35 Gy in 10 daily fractions delivered using orthovoltage X-rays (225 kVp).

Systematic Review
The systematic review of the literature (Figure 3) identified 66 eligible cases of pcp.Including the 3 patients from our chart review, 69 patients-35 of whom had a solitary pcp lesion-were analyzed (Table ii).Table iv provides rt and chemotherapy details.The rt dose was specified for 15 of the 27 patients (56%) who received rt as a component of their pcp treatment.Three patients received palliative-intent rt (total dose: ≤20 Gy).Two patients received surgery and adjuvant radiation as primary treatment: the dose for one patient was unknown 17 , and it was 20 Gy in 5 fractions for the other (patient 1 in the present report).One patient received 50 Gy in 25 fractions to a left cheek primary, with no response; salvage surgery was subsequently performed 32 .As with radiation doses, chemotherapy regimens ranged widely; the most commonly used regimen was melphalan combined with a corticosteroid.
Of the patients with a known disease response to chemotherapy or rt, most experienced a complete clinical response (Table v).Of patients with known recurrence details, most had no disease recurrence (Table vi).Among patients with a solitary pcp, 9 recurrences were observed.
Local recurrences in 2 patients were both initially treated with surgery 23,31 .Distant skin recurrences in 5 patients were initially treated with surgery alone 32 , surgery and rt 17 , rt alone (2 patients) 2,64 , or chemotherapy alone 45 .Finally, 2 patients experienced a distant recurrence; those patients were initially treated with rt 44 or observation 58 .Among the 19 patients for whom information about time-to-recurrence was available, median time to recurrence was 7.0 months (range: 1-132 months), and mean time to recurrence was 17.4 months.
Among the 69 analyzed patients, 27 events (recurrence of plasmacytoma, development of multiple myeloma, or death) were reported (Figure 4).Median rfs was 2.7 years (95% ci: 1.4 years to not reached), with 1-year, 2-year, and 5-year rfs estimates of 63%, 53%, and 44% respectively.Of those 27 events, 11 were observed in patients with a solitary pcp, with a median rfs of 11.0 years (95% ci: 2 years to not reached).The other 16 events were seen in patients with multiple lesions, with a median rfs of 1.4 years (95% ci: 0.6 years to not reached).The hazard ratio for rfs for patients with a solitary pcp lesion compared with those having multiple lesions was 0.47 (95% ci: 0.21 to 1.08; p = 0.069).No other variables were associated with rfs or os on univariate analysis (Table vii).
Among the 19 patient deaths reported, 18 had a documented time from diagnosis to death (Figure 4).The remaining patient was censored at last known follow-up 51 .Median os was 10.4 years (95% ci: 4.3 years to not reached), with 1-year, 2-year, and 5-year os estimates of 80%, 64%, and 59% respectively.In patients with a solitary pcp, 7 deaths were observed, with a median os of 17.0 years (95% ci: 1.7 years to not reached).In patients with multiple pcp lesions, 12 deaths were observed, with a median os of 4.3 years (95% ci: 1.3 years to not reached).The hazard ratio for os in patients with a solitary pcp lesion compared with those having multiple lesions was 0.37 (95% ci: 0.13 to 1.08; p = 0.059).
Table vi reports causes of death.Multiple myeloma was the most common cause of death, followed by recurrent plasma cell disease not meeting diagnostic criteria for multiple myeloma.

DISCUSSION
The present study represents the largest published single-institution case series of solitar y pcp, com-          prising 3 cases in total.Yamamoto described 3 patients with multiple pcp lesions 52 .Three other articles have each described 2 patients with pcp 2,46,54 .Cases of reactive pcp, which are excluded from the present analysis, have also been reported, associated with stimulation from infectious agents such as herpes simplex virus or insect bites 67,68 .The present systematic review summarizes all published nonreactive cases of pcp.The most recent comprehensive literature reviews of pcp were published by Muscardin et al. 1 in 2000 and Kazakov et al. 2 in 2002.Since those reviews, 14 new cases have been published, which are included here.With those cases, analyses of rfs and os were possible (Figure 4), although the data are limited by short follow-up and a small sample size.

Surgery
As in the present analysis, earlier reviews of patients with pcp have reported a mean age of 60, with male predominance 1,2,31 .The lesions are often described as reddish or purple nodules or plaques without ulceration 2,12,31 .The most consistently reported prognostic factor is the number of lesions, with outcomes being better with solitary lesions than with multiple lesions at presentation 1,[12][13][14] .Solitary lesions tended to have less locoregional or distant dissemination and remained confined to the skin 1,31 ; however, large solitary lesions might have a higher risk of Multiple, not specified 25 ( 36)  progression to metastatic disease (although a size cut-off is not defined) 1 .Patients with multiple lesions had a higher likelihood of progression to multiple myeloma and subsequent death 2,46,68 .Muscardin et al. 1 did not report any association of paraproteinemia with outcome.The present study accords with the prior finding that paraproteinemia is not associated with survival, although information about other variables associated with progression to multiple myeloma such as immunoglobulin levels or serum free light-chain ratio were not available.
Treatment of pcp should be individualized to the patient and tumour characteristics.In general, authors agreed that curative-intent local treatment-which can entail surgery, rt, or surgery followed by adjuvant rt-is appropriate for solitary pcp lesions 1,9,23,31,46 .Green et al. 68 cautioned that local recurrence or lymph node metastases could develop after local therapy.Patients with multiple pcp lesions are optimally treated with systemic chemotherapy 9 .
In the present analysis, it was not possible, because the wide range of dose and fractionations used, to draw conclusions about the efficacy of a specific rt dose.Recurrence data and site of recurrence were available but challenging to interpret because of the many patients whose recurrence status was unknown.Nonetheless, it is notable that the only local recurrences observed in patients with a solitary pcp lesion happened in patients initially treated with surgery alone.
The only clinical variable associated with rfs and os was the number of lesions (solitary vs. multiple), which showed a trend toward statistical significance.We observed a large difference in median survival (17 years vs. 4.3 years) and rfs (11.0 years vs. 1.4 years) between patients with solitary lesions and those with multiple lesions, albeit with overlapping cis.Based on those data, patients with solitary lesions are likely to have better outcomes, although that probability cannot be demonstrated to a p value of less than 0.05 because of the small number of events and the lack of long-term follow-up.Therapeutic approach (surgery vs. rt vs. chemotherapy) was not associated with rfs or os; however, our study was underpowered to detect a difference  between treatments in the overall study population or in subgroups of patients with solitary or multiple lesions.
The present study contributes to the literature by describing 3 additional patients with a solitary pcp (2 of whom have long-term follow-up) and by updating the systematic review of a rare disease to 69 total cases.However, the analysis is limited by the retrospective collection of published data and the extraction of details from articles.Given the elapsed time, contacting authors directly for further clinical data was infeasible.Epidermal and dermal involvement could not be distinguished.It is also possible that benign plasmacytoses were included in the analysis because information about monoclonality was not available for all reported cases; that limitation is acknowledged by Muscardin et al. 1 in their earlier review.Indeed, 3 of the historical patients in the review 15,16,18 could potentially have had benign tumours with plasma cell infiltrates or plasmacytic granulomas 23,26,31,32,49,68 ; diagnosis requires careful histologic and immunohistochemical analysis 9 .Exclusion of the former 3 patients does not change the survival analysis, because follow-up for them was unavailable.
The results of this systematic review are subject to publication bias.A future approach could involve the use of population-level databases to provide accurate information about survival, but granularity of clinical data and treatment details is often unavailable with such an approach.

CONCLUSIONS
Primary cutaneous plasmacytoma is a rare diagnosis, with limited information to guide treatment.Here, 3 patients with a solitary pcp lesion are reported, with complete details about treatment.With the accompanying systematic review, a total of 69 patients with pcp are now reported in the literature.Most patients with a solitary pcp received rt as a component of treatment.A trend toward improved median rfs (11 years vs. 1.4 years) and os (17 years vs. 4.3 years) was observed in patients with a solitary pcp compared with those having multiple pcp lesions.Local treatment (such as rt or surgery) should be considered for patients with a solitary pcp; patients with multiple pcp lesions are likely to be best served by systemic chemotherapy.Treatment of pcp should be individualized depending on factors such as patient performance status, lesion count (solitary vs. multiple), and expected ability to tolerate treatment.The combination of surgery and radiation was excluded because of the small count (2 patients).HR = hazard ratio; CI = confidence interval.

2 A 48 -
(A)].Radiation was completed on 15 June 1979.The patient experienced a complete clinical response to surgery and adjuvant rt.No recurrence of pcp or multiple myeloma was noted.During a hospital admission in 2000 for congestive heart failure, stroke, pneumonia, and atrial flutter, this patient was found to have an incidental 2×3.8cm anterior mediastinal mass.The lesion was clinically and radiologically consistent with a thymoma; no biopsy was taken, and the mass was observed.A benign right parotid Warthin tumour was excised in 2001.The patient died of right hemorrhagic stroke on 20 May 2002, 276 months after the initial pcp diagnosis.Patient year-old woman presented in March 1983 with a solitary right upper eyelid mass in the skin.Past medical history was non-contributory.Computed tomography of the orbit demonstrated a soft-tissue mass in the right upper eyelid with no bony erosion.A biopsy on 20 July 1983 demonstrated malignant plasmacytoma.The specimen was signed out as "plasma cell tumour," with monoclonal lambda immunoglobulin A expression.
Radiation was completed on 21 April 2015.At follow-up in June and December 2015, hyperpigmentation of the skin surrounding the lesion was noted, with complete clinical response and residual scar in the area of the biopsy [Figure 2(B,C)].Lab work was normal.The patient remains alive and well 11 months after initial diagnosis.

FIGURE 1 (
FIGURE 1 (A) Radiotherapy field for patient 1. (B) Radiotherapy field for patient 2 (dark circle represents a pencil eye-shield).

FIGURE 2
FIGURE 2 Clinical photographs for patient 3. (A) After biopsy, but before treatment (dots surrounding the lesion are clinical mark-up for the orthovoltage field edge).(B) Two months after completion of radiotherapy.Mild hyperpigmentation and erythema are noted.(C) Eight months after completion of radiotherapy.

FIGURE 3
FIGURE 3Flow diagram of articles and patients included in the systematic review.POEMS = polyneuropathy, organomegaly, endocrinopathy, monoclonal gammopathy, and skin changes; AESOP = adenopathy and extensive skin patch overlying a plasmacytoma.
Yes (cervical lymph nodes), timing unknownLast known status: Unknown; unknown follow-up Volk, 193616 UnknownLast known status: Unknown; unknown follow-up Raznatovsky et al., 1996 48 (case details obtained from: Kazakov et al., No Last known status: Alive; 2 months of follow-up Vyas et al., Unknown Last known status: Alive; 2 months of follow-up Miyamoto et al., No Last known status: Alive at 43 months of follow-up; cervical, axillary, inguinal lymph nodes and cutaneous plaques recurred at 43 months, but were polyclonal; patient died of heart failure at time unknown Yamamoto et al., Unknown Last known status: Alive; 7 months of follow-up e640 Current Oncology, Vol. 23, No. 6, December 2016 © 2016 Multimed Inc. Yes (chest and gingiva), 6 months Last known status: Alive; 7 months of follow-up Ak and Gülbas, Unknown Last known status: Unknown; unknown follow-up Dhouib Sellami et al., Unknown Last known status: Unknown; unknown follow-up Fitzhugh et al., Unknown Last known status: Unknown; unknown follow-up Gambichler et al., Unknown Last known status: Unknown; unknown follow-up Ioannidis et al., No Last known status: Alive; 10 months of follow-up Arias-Santiago et al., Unknown Last known status: Unknown; unknown follow-up Koletsa et al., Unknown Last known status: Unknown; unknown follow-up TREATMENTS AND OUTCOMES FOR PRIMARY CUTANEOUS PLASMACYTOMA, Tsang et al. e641 Current Oncology, Vol. 23, No. 6, December 2016 © 2016 Multimed Inc.

TABLE I
Clinical information and diagnostic investigations for three Princess Margaret Cancer Centre patients a Reported as "normal"; exact value not available.b Reported as normal [1.0 mgm% (88 mmol/L)].c <5% Plasma cells.IgA = immunoglobulin A; NA = not available.e632 Current Oncology, Vol. 23, No. 6, December 2016 © 2016 Multimed Inc.
Table iii lists the characteristics of those patients.Median follow-up duration was 2 years.

TABLE II Continued
Last known status: Alive; 9 months of follow-up; lesion was present for 20 years TREATMENTS AND OUTCOMES FOR PRIMARY CUTANEOUS PLASMACYTOMA, Tsang et al. e637 Current Oncology, Vol. 23, No. 6, December 2016 © 2016 Multimed Inc.

TABLE III
Demographics and clinical characteristics of all patients with primary cutaneous plasmacytoma

TABLE V
Treatments given for primary cutaneous plasmacytoma and responses to treatment No Bence Jones protein was detected at diagnosis.Routine lab work, protein electrophoresis, and bone marrow biopsy were normal.This patient was treated with 50 Gy over 5 weeks with little shrinkage; surgery was then performed, but distant skin recurrence was seen 4 months after surgery.At the time of recurrence, bone marrow biopsy and serum protein electrophoresis remained normal.Chemotherapy was initiated; the patient progressed and died 22 months after diagnosis 8 .
bThis patient had a 6-cm mass overlying the left cheek, biopsied as plasmacytoma.

TABLE VI
Recurrences and causes of death

TABLE VII Univariate
Cox proportional hazards analyses of variables associated with recurrence-free survival and overall survival in patients with primary cutaneous plasmacytoma (pCP) a Performed for patients with available follow-up information (n = 52).bScore test.The resulting p values were very similar to those from the likelihood ratio and log-rank tests.c