- Email: [email protected]
Combined total body electron beam irradiation and chemotherapy for mycosis fungoides
I
IIII
I
II
II
II
I
I
II
I
Combined total body electron beam irradiation and chemotherapy for mycosis fungoides Irwin M. Braverman, M.D.,* N. Bruce Yager, M.D.,* Michael Chen, M.D., Ph.D.,** Edwin C. Cadman, M.D.,*** William N. Hait, M.D., Ph.D.,*** and Theresa Maynard, B.S., C.T.(ASCP)* New Haven, CT Since 1979 a protocol of total body electron beam therapy (3,600 rads; 6 MeV), followed by six monthly cycles of chemotherapy (doxorubicin, 30 mg/M2 given intravenously once monthly, and eyclophosphamide, 100 mg/Mz given orally each day for 14 days), has been used to treat fifty patients with mycosis fungoides (primarily Stages I and II). A group of twenty-four patients, treated by identical high-dose electron beam therapy alone, served as control subjects. Actuarial analysis by the Kaplan-Meier method and statistical analysis by the generalized Wilcoxon test of Gehan demonstrated a significant difference (p = 0.008) in the probability of Stages I and Ir patients' remaining in complete clinical remission when combination therapy was compared with high-dose electron beam therapy alone. No statistically significant difference was demonstrated in patients in Stages III and IV mycosis fungoides. Although 60% of patients were in "complete clinical remission," the longest follow-up being 75 months, all continued to show karyotypic abnormalities of circulating lymphocytes, and 70% had intermittently and abnormally elevated blood levels of Sgzary cells. (J AM ACADDERMATOL 1987; 16:45-60.)
Since 1974 we have been treating patients having mycosis fungoides with high-dose total body electron beam radiotherapy (6 MeV; 3,600 rads). Although the clinical remissions were longer than those following low-dose electron beam radiotherapy (3-6 MeV; 800-1,600 rads), which we had used before 1974, the survival time of patients had not increased. During 1974 to 1976 we treated three cases of mycosis fungoides (Stage III, Mycosis Fungoides Cooperative Group staging) with
From the Departments of Dermatology,* Therapeutic Radiotherapy,** and Internal Medicine (Medical Ontology),*** Yale University School of Medicine. Supported by National Institutes of Health Grant No. 21153. Accepted for publication June 19, 1986. Reprint requests to: Dr. Irwin M. Braverman, 333 Cedar St., New Haven, CT 06510/203-785-4092.
multiple-drug chemotherapy* for 3 to 4 months either before or after a course of high-dose electron beam radiotherapy. All three patients entered remission. Two have remained in remission for 10.5 and 11.0 years, respectively. The third patient died 3.5 years later from cancer of the colon, but no evidence of mycosis fungoides was found at autopsy. Encouraged by the experience with these three cases, we initiated a treatment protocol of high-dose electron beam therapy followed by 6 months of chemotherapy for all newly diagnosed and previously untreated cases of mycosis fungoides that were referred to the Yale-New Haven Medical Center. The disease of all patients was staged according to the Mycosis Fungoides Co*The multiple-drug chemotherapy consisted of cyelophosphamide, bleomycin, methotrexate, cytosine arabinoside, and doxorubicin.
45
46
Journal of the American Academy of Dermatology
Braverman et al
Table I. Mycosis Fungoides Cooperative Group stages
Stage
[
TNeategory
I
TiNo_l
II
T1N~+
Plaques, papules, or eczematous patches covering less than 10% of the surface and no more than one site of clinically enlarged peripheral lymph nodes Plaques, papules, or eczematous patches covering less than 10% of the surface and two or more sites of clinically enlarged peripheral lymph nodes Plaques, papules, or eczematous patches covering 10% or more of the skin surface and no more than one site of clinically enlarged peripheral lymph nodes Plaques, papules, or eczematous patches covering 10% or more of the skin surface and two or more sites of clinically enlarged peripheral lymph nodes Tumors (one or more); any number of clinically enlarged peripheral lymph node sites Generalized erythroderma; any number of clinically enlarged peripheral lymph node sites
T2No-I
IU
T~N~+
T3Naay IV
Definition
[
T4N~y
Tt: Plaque, papules, or eczematous patches coveting <:10% of the body; T2: same as above but covering ~10% of the body; T3: tumors, one or more with or without plaques or papules; T4: erythroderma; No: no nodal areas involved clinically; N~: one nodal area involved; N2: two nodal areas; N3; three nodal areas; N4 to N~: four to eight nodal areas.
Table H. Characterization by stage, age, sex, and group
Stage and age
Sex
Group 1: EB-Chemo
Group 2: High EB
Group 3: Low EB
I
M F M F M F M F M F
11 9 10 10 3 4 1 2 53 49
4 1 7 2 4 0 2 4 66 56
1 0 2 4 1 4 3 3 65 57
II III IV Average age (yr)
EB-Chemo: Combined electron beam radiation and chemotherapy; high EB: high-dose electron beam radiation; low EB: low-dose electron beam radiation.
operative Group clinical criteria. 1 In addition, special hematologic, cytogenetic, and histologic studies were performed on peripheral blood, lymph node biopsies, and cutaneous lesions. This article reports the results of this treatment protocol as of Sept. 30, 1985. The details of the cytogenetic and histologic studies are described elsewhere? ,~ MATERIALS AND METHODS The patient population consisted of three groups. Group I contained fifty patients who had been entered
into the study since 1979. All were newly diagnosed and had not received any significant prior therapy. These patients received high-dose electron beam radiotherapy followed by 6 months of two-drug chemotherapy. Group 2 consisted of twenty-four newly diagnosed and previously untreated individuals who received highdose electron beam radiotherapy alone in a manner identical to those in Group 1. Nineteen Of these patients were treated from 1974 to 1979, and five have been treated since 1979. Group 3 consisted of eighteen patients who were treated by low-dose electron beam radiotherapy from 1963 to 1974. In all patients the di-
Volume 16 Number 1, Part 1 January 1987
Electron beam irradiation and chemotherapy for MF 47
agnosis of mycosis fungoides was made on clinical grounds and confirmed by one or more skin biopsies that showed a polymorphous papillary dermal infiltrate containing lymphocytes with darkly staining convoluted nuclei ("mycosis fungoides" cells). Pautrier microabscesses containing "mycosis fungoides" cells were present in all cases. The disease of all the patients was staged according to the original clinical criteria described in the report of the Mycosis Fungoides Cooperative Group j (Table I). These criteria could be applied to the patients in Groups 2 and 3 because either photographs of their entire skin surface were available or the extent of their disease had been carefully documented in their clinical records. In addition, one of us (I. M. B.) had been involved in the care of every patient in Groups 1 and 2 and most of the patients in Group 3. The characterization of the patients by age, sex, Mycosis Fungoides Cooperative Group stage, and therapy groups is shown in Table II. By history, cutaneous lesions had been present for 3 to 483 months (mean • SD: 85.3 _-+ 104.9 months) before biopsy diagnosis was made in Group I patients. Therapy was begun within 4 to 6 weeks after biopsy diagnosis in forty-five of fifty patients. In the remaining five patients, therapy was initiated 12 to 13 months after biopsy diagnosis. These five patients either had spot orthovoltage radiation for single lesions or had been treated by topical steroids before being referred for further therapy. In Group 2, nineteen of twenty-four patients received only high-dose electron beam radiotherapy within 4 to 8 weeks of biopsy diagnosis, and five had been treated with topical steroids for 10 to 28 months before being referred for electron beam radiotherapy. Cutaneous lesions had been present for 9 to 447 months (mean +- SD: 99.9 +-- 104.8 months) before biopsy diagnosis had been made in Group 2 patients. There was no statistical difference between the patients in Groups 1 and 2 with regard to this parameter. In Group 3, nine of eighteen patients received low-dose electron beam therapy within 5 months of diagnosis, and in nine patients the interval was 11 to 96 months. Most of the Group 3 patients were treated from 1963 to 1965, when electron beam radiotherapy first became available as a therapeutic modality at the Yale-New Haven Medical Center. Prior to electron beam radiotherapy for the patients in Group 1, the following staging procedures were performed. The extent of the skin lesions was estimated by measuring with the entire palm of the hand, which is approximately 1% of the body surface. The following tests were performed: chest x-ray, liver-spleen radionuclide scan, abdominal computed tomography scan, skin testing with recall antigens (Candida, trichophytin,
purified protein derivative, and mumps), complete blood cell count, urinalysis, liver function tests, routine chemistry studies, and determination of the radionuclide cardiac ejection fraction. Additional tests were as follows: quantitation of serum immunoglobulins in the clinical immunology laboratory of the medical center by standard nephelometer methods (G, M, A), by radial immunodiffusion (D), and by radioimmunoassay (E); search for and enumeration of the number of S6zary cells (large and small variants) per 100 lymphocytes found in routine blood smears and in buffy coats embedded in Spurr's resin for examination in 1 p,m sections; skin biopsies of both normal-appearing, uninvolved skin and lesional skin for study by light and electron microscopy; biopsy of palpable nodes or a "blind" biopsy of axillary or inguinal nodes that drained mycosis fungoides skin lesions, performed in nineteen of fifty Group 1 patients; and cytogenetic studies with Giemsa G banding on lymphocytes derived from peripheral blood and biopsy specimens of lymph nodes. (The lymph nodes were interpreted in the department of surgical pathology at Yale.) Following completion of electron beam radiotherapy, the skin was rephotographed; the sites of previous lesionai skin and uninvolved skin were repeated with photographic documentation; serum immunoglobulins were remeasured; and cytogenetie studies on peripheral blood and a search for S6zary cells in routine smears and buffy coats prepared for 1-p,m sections were also repeated. After completion of chemotherapy, the same studies were repeated at 6- to 12-month intervals unless clinical events indicated an earlier reevaluation. If unusual skin lesions arose, they were photographed and biopsy specimens were obtained for light and electron microscopic examination. All patients were seen at lto 3-month intervals as their status required, at which times a complete physical examination, including inspection of the entire skin surface, was done and a search for palpable lymph nodes made. Of fifty-six consecutive mycosis fungoides patients, fifty agreed to participate in the protocol. Two men, 40 and 62 years old, both of whom had lesions of mycosis fungoides appearing as parapsoriasis and poikiloderma atrophicans vasculare involving only 1% and 2% of their body surface, respectively, elected to be treated by topical nitrogen mustard alone. Four patients were treated by high-dose electron beam radiotherapy alone (two women, 67 and 74 years old, and one man, 19 years old, refused chemotherapy; one man, 59 years old, had decreased cardiac function, precluding the use of chemotherapy). Electron beam radiotherapy for patients in Group 3 consisted of low total doses of radiation ranging from
48
Journal of the American Academy of Dermatology
Braverman et al
600 to 1,600 rads (mean: 1,100 rads) of mostly 2.5 to 3 MeV electrons. Since 1974 we have used the sixfield whole body technic of electron beam radiotherapy developed at Stanford University,4 but without the dual angulation of the beam for each field. The patients were treated with 6 MeV electrons produced by a Clinac-6 linear accelerator (u Associates Inc., Palo Alto, CA), in contrast to the 4 MeV electrons used at Stanford. We have calculated that these 6 MeV electrons degrade to an energy level of 4.4 MeV at the patient's skin surface. The patients received whole skin doses of 3,600 rads in 9 weeks at 100 rads/day, 4 days a week. A complete six-field treatment cycle requires two con: secutive treatment days. Booster doses of orthovoltage radiation (2,000 rads at 120 or 125 kVp) were given to the soles of the feet and to the perineum in a split course: 100 rads/day, 4 days a week, to 1,000 fads during the first and tast ten treatment days. If eyelid involvement by mycosis fungoides was present or suspected, internal lead eye shields were employed during most of the electron beam treatment. Within I month after the completion of electron beam therapy, chemotherapy was begun: doxorubicin, 30 mg/M 2 given intravenously once monthly, and oral cyelophosphamide, 100 mg/Mz daily for 14 days each month for 6 months. From 1974 to 1979, while treating patients in Group 2 with high-dose electron beam radiotherapy, we had observed the development of clinically nonspecific-appearing lesions, one to six in number, 4 to 46 months after the completion of electron beam therapy that had produced a complete remission. The lesions measured 5 to 20 mm and resembled patches of asteatotic eczema or pityriasis rosea. Histologically this eruption showed mycosis fungoides ceils in the epidermis and dermis without the usual accompanying dense, polymorphous, inflammatory cell infiltrate. In seven of eight patients, these lesions eventually evolved into typical mycosis fungoides lesions both clinically and histologically 4 to 23 months later. In the eighth patient the nonspecific lesions are still present and are unchanged 63 months later. In Group 1 patients, identical lesions have been observed at approximately the same rate, but a higher percentage have remained unchanged or have disappeared either spontaneously or following the use of topical steroids. These "nonspecific" lesions do not resemble the patients' original eruption. Other patients in Groups 1 and 2 have relapsed with lesions that clinically and histologically are characteristic of mycosis fungoides and resemble the initial eruption. We call these "specific" lesions. For Groups 1 and 2, we have precise information,
with biopsy and photographic documentation, of the times of onset of the nonspecific lesions and of their evolution into specific lesions if that event occurred. Except for one patient in Group 3, in whom nonspecific and specific lesions were commented on in the records, only specific lesions were recorded as a sign of clinical relapse following electron beam radiotherapy. None of the patients in these three groups had detectable visceral involvement at the time therapy was initiated. Life table analyses using the product-limit method of Kaplan and Meier* were performed on the data. Survival statistics were calculated from the initiation of electron beam therapy to death from any cause. Periods of clinical remission were measured in two ways: (1) from the end of electron beam therapy to the appearance of those nonspecific lesions that eventually evolved into specific lesions and (2) from the end o f electron beam therapy to the appearance of specific lesions. Because nonspecific lesions have spontaneously disappeared or have continued to persist for as long as 63 months without changing their clinical appearance or histologic features, we have not considered them to be end points for determining clinical relapse unless their clinical appearan.ce and histologic picture evolved into the diagnostic features of mycosis fungoides. The estimated probabilities of survival and o f a patient's remaining in complete clinical remission produced by the three forms of therapy were compared for patients in comparable stages. The generalized Wilcoxon test of Gehant for comparing two life tables was used to assess the statistical significance of the data. These studies were carded out the with approval o f the Yale School of Medicine Human Investigation Committee.
RESULTS The combination of electron beam irradiation and two-drug chemotherapy was well tolerated. Mild neutropenia was observed in all patients (3,000-4,500), but none experienced severe neutropenia or life-threatening infections. Neither clinical compromise in cardiac function nor hemorrhagic cystitis was seen. There were no significant side effects beyond those normally seen with electron b e a m radiotherapy. Electron beam therapy produces complete body hair loss beginning *Kaplan EL, Meier P: Nonparametricestimationsfrom incomplete observations. Am Stat Assoc J 53.457-481, 1958. "['GelaanEA: A generalizedWilcoxontest for comparingarbitrarily singly-censoredsamples.Biometrika52:203-223, 1965.
Volume 16 Number 1, Part 1 January 1987
Electron beam irradiation and chemotherapy for MF
in the third week of therapy. Hair began to regrow within 6 to 8 weeks after the completion of electron beam therapy in spite of the chemotherapy regimen. Ankle edema developed to variable degrees during the last 2 weeks of electron beam therapy, being most severe in patients with leg vein excisions required for coronary artery bypass surgery. Erythema, hyperpigmentation, and dry skin with exfoliation occurred to a variable degree in all patients. The skin returned to normal color and texture within 6 to 12 weeks after the completion of electron beam therapy. Neither radiodermatitis nor ectropion has developed in any patient. Most patients developed moderate generalized hypohidrosis affecting the entire skin surface except for the axillae and scalp. In most of the affected patients the return of sweating has been incomplete even after 3 to 4 years. Forty-nine of fifty patients in Group 1 had either a complete response or 95% clearing of their skin lesions at the end of electron beam radiotherapy. One patient, who had erythroderma but not classic S6zary syndrome, did not respond and is currently receiving experimental chemotherapy. In four of five patients with 95% clearing, the lesions resolved during chemotherapy. In the patient whose lesions failed to resolve completely, a relapse occurred 4 months after the completion of electron beam therapy. Of the fifty patients, twenty-two have remained free of skin lesions (nine for 30-75 months, ten for 14-26 months, and three for 2-10 months). Eighteen of the fifty patients developed nonspecific cutaneous lesions 6 to 23 months after the end of electron beam therapy, which appeared as 1-2 cm, ill-defined patches of dry, scaly, sometimes hypopigmented skin indistinguishable from skin with xerosis or pityriasis alba or from oval, slightly scaly patches similar to those of pityriasis rosea or guttate parapsoriasis. There were only one to six such lesions when they first appeared, in nine individuals the lesions have either waxed and waned spontaneously, disappeared spontaneously, or disappeared with the use of topical steroids during an observation period of 39 to 63 months. In the other nine, they increased in number over the course of 4 to 6 weeks. Biopsy of the lesions in all eighteen patients at the time the lesions arose
49
showed the "tolerance" histologic picture described in paragraph 14 of this section. In the nine individuals whose nonspecific lesions increased, the lesions became infiltrated and more numerous and eventually evolved into mycosis fungoides both clinically and histologically 5 to 20 months after their initial appearance. In none of the eighteen individuals did the nonspecific lesions resemble the initial eruption of mycosis fungoides. Nine patients relapsed with specific lesions identical to their original ones without going through the stage of having nonspecific lesions. These relapses occurred from 1 to 13 months after the end of electron bean therapy except for one relapse at 19 months. Eight of these nine individuals had nodules, tumors, or erythroderma as the presenting signs of mycosis fungoides, rather than the eczematous patches and poikilodermatous lesions characteristic of most of the other patients. Five patients have died: one died from staphylococcal sepsis and aplastic anemia and a second from staphylococcal sepsis resulting from bone marrow replacement by S6zary cells but with clinically and histologically clear skin; two with visceral mycosis fungoides died from infection; one died from a pulmonary embolus while the disease was under control by chemotherapy; and one died from congestive heart failure resulting from arteriosclerotic heart dsease complicated by a cardiac ventricular aneurysm. This last patient had minimal skin involvement and no visceral involvement by mycosis fungoides. None of these five deaths occurred during the chemotherapy phase immediately following electron beam therapy. Of the fifteen living patients with specific skin lesions, topical nitrogen mustard or treatment with psoralens and ultraviolet A is keeping eight under excellent control. Only seven patients have progressive disease requiring multiple-drug chemotherapy. In eight of twenty-four patients in Group 2 who were treated by the identical electron beam therapy regimen but without chemotherapy, the same types of nonspecific lesions with "tolerance" histologic features developed 4 to 46 months after the end of electron beam therapy. In seven patients, evolution into classic mycosis fungoides lesions both morphologically and histologically occurred 4 to 23 months later. The eighth patient still has non-
50
Journal of the American Academy of Dermatology
Braverman et al
-%
89
.9
~'T I
1
.Z-
I,,-4 ~ . .6-
re
(.,9 .~r'l
,4_
$0 kd
.2"
re
.... 9
.
,
20
,
40
.... . ,
,
60
.
100
,
le0
.
,
..........
140
,
160
.
;
18
.
,
~00
MONTHS Fig. 1. Actuarial curves of the estimated probabilities of survival for combined clinical Stages I and II in all three treatment groups. The numbers of patients at risk at specific break points were as follows: with combination therapy (COMB RX), 40 patients at 0 month, 34 at 14 months, 24 at 24- months, 21 at 27 months, 13 at 42 months, 7 at 54 months, 5 at 64 months, 2 at 68 months, and 1 at 77 months; with high-dose electron beam (EB) radiation, 14 patients at 0 month, 13 at 3 months, 8 at 38 months, 7 at 47 months, 5 at 63 months, 2 at 98 months, and 2 at 105 months; and with low-dose electron beam radiation, 7 patients at 0 month, 6 at 1 month, 5 at 6 months, 4 at 34 months, 3 at 41 months, 2 at 82 months, 2 at 134 months, and 1 at 180 months.
specific lesions that have remained unchanged for 63 months. Eight others in Group 2 had a clinical relapse characterized by classic mycosis fungoides skin lesions, with the typical histopathologic picture of mycosis fungoides, 5 to 18 months after the start of electron beam therapy without going through the stage of nonspecific lesions. For the patients in Group 3 treated with lowdose electron beam therapy (3 MeV; 800-1,600 rads), we have reliable information only on the duration of clinical remission, based on relapse with specific lesions, and the date of death. Nonspecific lesions were described in only one patient. Figs. 1 and 2 show the actuarial curves of the estimated probability of survival for all three treatment groups. Table III presents the levels of sta-
tistical significance for the comparisons between the various life table curves. Combination therapy, in comparison with low-dose electron beam therapy alone, has increased the survival rates for pa-. tients in all stages. For early mycosis fungoides (Stages I and II), high-dose electron beam therapy, in comparison with low-dose electron beam therapy, produces an improvement in survival rates approaching a level of significance (p = 0.06). No effect was seen in late mycosis fungoides (Stages III and IV). However, when the estimated probability of a patient's remaining in complete clinical remission was calculated with specific lesions used as end points and with comparable stages compared, a significant difference was demonstrated between
Volume 16 Number 1, Part I January 1987
Electron beam irradiation and chemotheraPY for MF
51
.92
~
.7-
~.
.6-
cz:
~
:D r
~
.'~
21.,., 0
10
I
L~
v
"
!
~)
I
40
!
~
60
'1"
~
I
70
r
B0
~
'
9~
I
I
lO0 110
9
f
1~0
!
l~O
'
t
140
'
!
l~
MONTHS -'44-'N- CBCB IO/,-(H}-- NlfiH s
~ r
L[]tJ EB
Fig. 2. Actuarial curves of the estimated probabilities of survival for combined clinical Stages III and IV in all three treatment groups. The numbers of patients at risk at specific break points were as follows: with combination therapy (COMB RX), 10 patients at 0 month, 4 at 23 months, and 1 at 33 months; with high-dose electron beam (EB) radiation, 10 patients at 0 month, 9 at 5 months, 7 at 19 months, 2 at 37 months, and 1 at 78 months; and with low-dose electron beam radiation, 9 patients at 0 month, 9 at 3 months, 7 at 10 months, 5 at 23 months, and 2 at 48 months. the results produced by high-dose electron beam therapy and those produced by low-dose electron beam therapy for advanced disease (Fig. 3; Table IV) but not for early disease (Fig. 4). Figs. 5 and 6 show the life table curves of the probability of a patient's remaining in complete remission when a combination of electron beam therapy and chemotherapy is compared with highdose electron beam therapy alone (Stages 1 and II and Stages III and IV, respectively) (Table V). Nonspecific lesions were used as end points in both comparisons. There is a significant difference when Stages I and II are compared (p = 0.008) but not when Stages III and IV are compared. When specific lesions are used as end points (Figs. 3 and 4), a significant difference is found when Stages I and II are compared (p = 0.01) but not when Stages III and IV are compared
HI. Levels of statistical significance for estimated probabilities of survival Table
Stage
I, II I, II I, II III, IV III, IV III, IV
I
Therapy
EB-Chemo vs high EB High EB vs low EB EB-Chemo vs low EB EB-Chemo vs high EB High EB vs low EB EB-Chemo vs low EB
I
pValue
0.193 0.06 0.002 0.07 0.09 0.004
See Table II for explanation of abbreviations.
(p = 0.36). Combination therapy increases the probability of a patient's remaining in complete remission for early-stage disease (Stages I and 1I) regardless of end point used. The actuarial curves for patients in Stages [ and II were identical, as were the curves for patients
52
Journal of the American Academyof Dermatology
Braverman et al
~
"7" r'-I
.e-
III
(")
.7-
I,I C~
.6-
]
.~I.,-,I
..I 0
.4-
t:~ []
.:~"
CL. (.,,)
Id
~'
!
10
'"
I
eO
'
I
aO
1
I
40
'
I
"
I
60
~0
"
I
70
,
I
80
,,T
I
90
'
I
xoo
'
I
110
'
I
' I
'
I
1~0 130 240
MONTH8 HIGH E~ ~ ~ LOW El
Fig. 3. Actuarial curves of the estimated probabilities of clinical remissions for combined Stages HI and IV in all three treatment groups, with specific lesions used as end points. The numbers of patients at risk at specific break points were as follows: with combination therapy (COMB RX), 9 patients at 0 month, 7 at 4 months, 4 at 12 months, 1 at 17 months, and 1 at 31 months; with high-dose electron beam (EB) radiation, 9 patients at 0 month, 5 at 4 months, 3 at 12 months, and 1 at 17 months; and with low-dose electron beam radiation, 2 patients at 0 month. Table IV. Levels of statistical significance for estimated probabilities of clinical remission, with specific lesions used as end points Stage
Therapy
I, II III, IV I, II
High EB vs low EB High EB vs low EB EB-Chemo vs high EB
]
p Value
0.08 0.0002 0.01
See Table II for explanation of abbreviations.
in Stages III and IV in all of the analyses described here. Therefore the data for Stages I and II and for Stages III and IV were pooled. Fig. 7 is a composite graph of the estimated probabilities of a patient's remaining in complete clinical remission for Stages I and II following combination therapy and high-dose electron beam
therapy. The actuarial curve from Group 2 is superimposable on the curves from three other studies in which high-dose electron beam therapy alone was used for patients with comparably staged disease (Hoppe et al, 5 Fig. 5; Tadros et al, 6 Fig. 3; Hamminga et al, 7 Fig. 7). The detailed histologic studies of the "nonspecific" lesions are reported elsewhere 3 and will be described only briefly here. At the time of initial diagnosis, all patients showed the classic histologic features of mycosis fungoides: a papillary dermal infiltrate composed of normal-appearing lymphocytes and histiocytes admixed with a few plasma cells and eosinophils. Within the infiltrate there were varying numbers of mononuclear cells with
deeply stained hyperconvoluted nuclei (mycosis fungoides cells). The "mycosis fungoides" cells were present singly and in groups (Pautrier microabscesses) within the epidermis. They were
V o l u m e 16
Number I, Part 1 January 1987
Electron beam irradiation and chemotheraPY for MF
53
.9
7"
r-!
.8-
.7-
La.I , 6 -
l-.4
_1 (..)
.4-
m ,3.
I'-
.2-
Ld ,1
'
1
l0
''
I
~1
'
I'
~0
''t''
I
4{I
I
~
'
I
60
'
('""'
"l
70
'
~
I
'
-1""
c/O 100
]
llO
~
l
le0
r-"r'-v'-~
1~
140
MONTH8 Kew:
L; ~, COHBI~ ~ ~ HIC.d,4s
~ f.
LI]WEB
Fig. 4. Actuarial curves of the estimated probabilities of clinical remissions for combined stages I and II in all three treatment groups, with specific lesions used as end points. The numbers of patients at risk at specific break points were as follows: with combination therapy (COMB RX), 40 patients at 0 month, 34 at 8 months, 21 at 19 months, 14 at 31 months, 6 at 49 months, 2 at 67 months, 1 at 72 months, and 1 at 75 months; with highdose electron beam (EB) radiation, 14 patients at 0 month, 8 at 7 months, 4 at 16 months, 3 at 31 months, 1 at 59 months, and 1 at 96 months; and with low-dose electron beam radiation, 7 patients at 0 month, 4 at 5 months, 2 at 24 months, and 1 at 134 months.
also present singly and in clusters in the dermis both in perivascular and nonperivascular sites. In all patients in Groups 1 and 2, repeat biopsy of a lesional site that had returned to normal after either electron beam therapy alone or electron beam therapy with chemotherapy disclosed a unique picture. Single "mycosis fungoides" cells were still present both in the dermis and epidermis, and only rarely were they found in clusters in the epidermis. The usually dense, accompanying polymorphous infiltrate of normal inflammatory cells was either absent or minimal. Further, in eight patients, biopsies of normal uninvolved skin, taken both before therapy and after therapy, showed occasional "mycosis fungoides" ceils in the epidermis and dermis without any accompa-
Table V. Levels of statistical significance for estimated probabilities of clinical remission, with nonspecific lesions used as end points Stage
Therapy
I,.p Value
l, II HI, IV
EB-Chemo vs high EB EB-Chemo vs high EB
0.008 0.166
See Table II for explanation of abbreviations.
nying inflammatory infiltrate. We had never seen such findings in the skin of healthy individuals. Biopsies of the nonspecific cutaneous lesions that appeared post therapy showed histologic features similar to those found in the posttreatment
54
Journal o f the American A c a d e m y of Dermatology
Braverman et al
,9-
7" .er"l r(f) 3 I.M .6-
I,==I
.-I ixl [] FI I.-G~
.2-
W ,I-
o
9
I
'
I
'
I
'
I'
'
I
'
'
' ~I~
o
'
I
,
I
'
I
'
'
MONTHS Fig. 5. Actuarial curves of the estimated probabilities of clinical remissions for combined Stages I and II in the combination therapy (COMB RX) and high-dose electron beam (EB) therapy groups. Nonspecific lesions were used as end points. The numbers of patients at risk at specific break points were as follows: with combination therapy, 40 patients at 0 month, 36 at 6 months, 23 at 16 months, 18 at 22 months, 6 at 44 months, 2 at 67 months, and 1 at 75 months; and with high-dose electron beam radiation, 14 patients at 0 month, 7 at 7 months, 3 at 13 months, 2 at 46 months, 1 at 52 months, and 1 at 96 months. lesional sites9 "Mycosis fungoides" cells were more numerous, but an inflammatory infiltrate was still absent or minimal. We have referred to this phenomenon as the "tolerance" histologic picture or as the "recapitulative" phase of the disease9 In eight patients we were able to follow the evolution of nonspecific lesions into specific ones with serial biopsies. The "mycosis fungoides" cells became more numerous, and a polymorphous inflammatory infiltrate eventually reappeared, allowing for the histologic diagnosis of mycosis fungoides to be made. In the patients whose nonspecific lesions have not evolved, the histologic features have remained unchanged. These studies, reported elsewhere, 3 revealed that 60% to 80% of the "mycosis fungoides" cells in the epidermis and 10% to 25% of the "mycosis fungoides" cells in the dermis in nonspecific lesions were Langerhans and indeter-
minate cells and not T lymphocytes. In successfully treated lesions the density of Langerhans cells and indeterminate cells in the epidermis was still markedly decreased 1 month after the completion of electron beam therapy. Results of the following tests were normal in all fifty Gro.up 1 patients before therapy: chest xray, liver-spleen radionuclide scan, and abdominal computed tomography scan searching for enlarged abdominal nodes. Forty-eight of fifty patients reacted to at least one of the recall antigens. In one patient the results of the Candida skin tests became positive after completion of electron beam therapy, but the other patient remained anergic. Results of routine hematologic tests, urinalyses, liver function tests, and routine chemistry studies were normal in all patients. The white blood cell count ranged from 4,500 to 10,000. Anemia (hematocrit
V o l u m e 16
Number 1, Part 1 January 1987
Electron beam irradiation and chemotheraPY for MF
55
1.9-
[] r
.7-
D,,,,,,I
._1 I--
lJ n
I--
.2-
i,I
100
,d ld
MONTH~ K~m
,; ~
COI~IB~
0 ~
HIGH
Fig. 6. Actuarial curves of the estimated probabilities of clinical remissions for combined Stages III and IV in the combination therapy (COMB RX) and high-dose electron beam (EB) therapy groups. Nonspecific lesions were used as end points. The numbers of patients at risk at specific break points were as follows: with combination therapy, 10 patients at 0 month, 6 at 4 months, 3 at 10 months, and 1 at 31 months; and with high-dose electron beam radiation, 10 patients at 0 month, 5 at 4 months, and 3 at 12 months.
<38%) and thrombocytopenia were not found in any patient prior to therapy. The values for serum immunoglobulins, except for IgE, were normal, The only abnormalities found were those related to the serum levels of IgE, the percentage of S6zary cells per 100 lymphocytes, the karyotypic analyses of lymphocytes in the peripheral blood and lymph nodes, and the skin and lymph node biopsies. Following therapy, only the levels of serum immunoglobulins and circulating S6zary cells and the peripheral karyotypes were routinely measured.
IgE levels Before treatment, only twelve of fifty patients had abnormally elevated serum levels of IgE (>200 IU). The values ranged from 220 to 1400 IU. These same patients continued to have elevations after therapy. Fluctuations of IgE levels occurred over a twofold to threefold range post
therapy in three patients without any obvious clinical correlation. In two patients with normal levels of [gE before therapy, the levels rose from 44 to 430 IU immediately afterward and from 18 to 384 IU 21/2 years after the completion of electron beam therapy, respectively. Circulating Stzary ceils In forty-five normal control subjects, S6zarylike ceils of the small cell type constituted 0% to 5% of circulating lymphocytes both in routine smears and in buffy coats (mean • SD: 0.69 • 1.5). In ten psoriasis patients and in the five patients with diffuse eczema the percentage varied from 2% to 8% (mean • SD: 3.2 +_ 2.3). We used 10% as the dividing line between normal and abnormal levels of circulating Stzary cells in normal subjects and mycosis fungoides patients. We compared the percentage of circulating
56
Journal o f the American Academy of Dermatology
Braverman et al
1
.'9
~1
.8-
1-11
I-,.( t~y
.6-
.5-.I t.O
.4.
. [] tat
.3-
I1.
"~.i_t
I---'E_ . . . . .
.
I-- . ~ I.d
9T A HO
.l
I
1
I
I
I
I
|
|
I
I
10
la0
:30
40
~0
60
70
80
90
100
MDNTHS Kevz
--~Ct~'lB
RX ST i & ::'
D O HIGH EB ST t I, 2
Fig. 7. Actuarial curves of the estimated probabilities of clinical remissions for combined Stages I and II following combination therapy and high-dose electron beam therapy. The actuarial curves from three other studies (HO, reference 5; TA, reference 6; HA, reference 7), in which high-dose electron beam therapy alone was used, are superimposed. Specific lesions were used as end points. Sbzary cells found in 1-1~m sections of buffy coats to that found in normal smears of twenty-three mycosis fungoides patients. One hundred fifteen pairs were available for comparison. The correlation coefficient of this comparison was r = 0.815; p < 0.001. (There were only twelve instances, all post therapy, in which a marked discrepancy occurred, making it impossible to determine whether the percentage was greater or less than 10%). In general, it was not difficult to identify and quantitate the number of Sbzary cells per 100 lymphocytes in routine smears. However, the routine differential smear examination of 100 white blood cells was not helpful in indicating whether there was an elevated level of circulating Sbzary ceils. One needs to count 100 lymphocytes. Virtually all the SCzary cells were of the small cell type--identical in size to small lymphocytes. Before therapy, nineteen of forty-seven patients had 10% circulating Sbzary cells per 100 lympho-
cytes (range, 10% to 30%). After electron beam radiation alone and electron beam radiation with chemotherapy, elevated levels o f Sbzary cells were intermittently present in nineteen of twenty-six patients in the range of 12% to 30%. We could not relate these elevations to the eventual disease courses in these patients.
Chromosomal studies The results of chromosomal studies are reported in detail elsewhere, z Clonal abnormalities were found once in only one patient. These abnormalities were not observed again in three subsequent tests during the succeeding 24 months. In no patient were the same chromosomal abnormalities found in both the lymph node and the peripheral blood specimens. In no instance have any of the original karyotypic abnormalities present before therapy been found after therapy. Karyotypic analyses have remained abnormal in all patients with
Volume 16 Number 1, Part 1 January I987
Electron beam irradiation and chemotherapy for MF
more than one follow-up test after treatment. Chromosomal abnormalities of circulating lymphocytes and intermittently elevated levels of S6zary cells continue to be present post therapy in most patients, without any apparent change in their states of clinical remission. Lymph nodes Of nineteen patients with lymph node biopsies, only six had clinically significant palpable lymphadenopathy. Sixteen patients had identical histologic findings on biopsy: dermatopathic changes with scattered S6zary-like cells in the paracortical area. The nodal architecture was preserved, with moderately active, normally placed follicular centers present. A marked expansion of the nodal T cell zones was accompanied by a moderate proliferation of small postcapillary venules, prominent hyperplasia of Langerhans macrophages, and a cellular population relatively rich in small lymphoid cells with hyperconvoluted nuclei. Sinus histiocytosis with melanin-containing macrophages was present. Three of the six patients with palpable nodes had involvement by mycosis fungoides in the excised nodes. These patients had four or more areas of nodal involvement and were in Stage III or IV. The only difference between the excised nonpalpable nodes and the excised palpable nodes in the other sixteen instances was the greater size of the latter. DISCUSSION Our studies confirm those from Stanford University 4 that electron beam radiation in doses of 3,000 rads can increase the probability of a patient's remaining in comlete clinical remission in late-stage mycosis fungoides in comparison with electron beam doses <1,600 fads. Cotter et als also found that a dose of at least 3,000 rads is needed for adequate local control of mycosis fungoides lesions. Both our combination electron beam radiation and chemotherapy and our highdose electron beam radiotherapy have produced an unusual syndrome in approximately 35% of patients who have had a complete response to the electron beam radiation: 1-2 cm patches, usually one to six in number and resembling xerosis or parapsoriasis, have appeared 4 to 43 months after
57
the completion of combination therapy or highdose electron beam therapy. Either they have evolved over 6 to 24 months into frank mycosis fungoides, or they have waxed and waned, spontaneously disappeared, or remained unchanged both in appearance and histologic features for up to 63 months. We have referred to the histologic changes as "tolerance" histologic changes or the "recapitulative" phase because "mycosis fungoides" cells are present but the normal accompanying inflammatory infiltrate is absent. One interpretation is that the electron beam has altered the microenvironment of the skin in some way so that the abnormal T cells ("mycosis fungoides" cells) do not evoke an inflammatory infiltrate (i. e., a tolerance to the abnormal cells has been created). Another interpretation is that the pathogenetic process underlying mycosis fungoides has been interrupted and the disease recapitulates itself through the manifestations of the nonspecific lesions. The observation that of eight nonspecific lesions arising after high-dose electron beam therapy alone, seven evolved into specific mycosis fungoides lesions, but that only nine of eighteen nonspecific lesions arising after combined therapy have done so, raises the possibility that chemotherapy may promote the "tolerance" or "recapitulative" phase. Regardless of whether we used specific mycosis fungoides lesions to indicate relapse after a complete response or nonspecific lesions that we know evolved into mycosis fungoides lesions, the results have been the same. Combination therapy has produced a statistically significant increased probability of a patient's remaining in complete clinical remission in Stage I or II mycosis fungoides, in comparison with the use of high-dose electron beam radiation alone. However, in late-stage disease (Stages III and IV), neither treatment modality has yet been shown to be superior to the other. These results illustrate the increased efficacy of adjuvant chemotherapy. Presumably the use of chemotherapy was less effective in advanced stages because these patients were left with a greater tumor burden, albeit clinically silent, following electron beam therapy than those with earlier stages of the disease. The group of twenty-four patients treated by
58 Braverman et al
high-dose electron beam radiation alone represents a historical control group except for the five patients who were added after 1979. Although we have accurate data in the form of photographs of the entire skin surface of these patients before and after therapy and detailed clinic notes on the course of their disease, a/lowing us to stage their disease by mycosis fungoides cooperative group criteria, there is always the concern that a retrospective historical group can never be used with confidence as a control group. However, as we have shown in Fig. 7, three other groups of investigators57 have published life table curves of clinical remissions following high-dose electron beam therapy for early mycosis fungoides that are superimposable on our own curve. This makes us confident that Group 2 is a valid control group. The actuarial curves generated by these three independent studies and ours may be a useful baseline against which other modalities can be judged. Since the nonspecific lesions in nine of eighteen patients have not yet evolved into mycosis fungoides lesions after 39 to 63 months of observation, we feel justified in considering the individuals with only one to six nonspecific lesions as being in clinical remission. Our data indicate that the combination of electron beam radiation and chemotherapy has produced a significantly increased frequency of sustained, complete clinical remission in early mycosis fungoides (Stages I and II) in comparison with highdose electron beam radiation alone. However, this combination therapy has not been as beneficial in late-stage disease (Stages III and IV). Griem et al 9 treated fourteen patients with mycosis fungoides with a combination chemotherapy and electron beam radiation protocol and produced an actuarial curve of clinical remissions that is virtually identical to our own. We have used the original Mycosis Fungoides Cooperative Group criteria for staging because we believe that it is more flexible in accommodating the various clinical features of the disease. In our experience the number of clinically involved nodal areas has correlated with the presence of lymph nodes involved by mycosis fungoides. We have taken biopsy specimens from nineteen nodes for staging purposes. Thirteen nodes were nonpalpable and six were palpable. All thirteen nonpalpable nodes from sites draining mycosis fungoides le-
Jourllal of the American Academy of Dermatology
sions showed dermatopathic lymphadenopathy with scattered S6zary-like cells in the paracortical area. Three of the six palpable nodes showed involvement by mycosis fungoides, and the remaining three showed identical dermatopathic changes. The only histologic difference between the excised nonpalpable nodes and the excised palpable nodes with dermatopathic changes was the greater size of the latter. In addition, the mycosis fungoidespositive nodes all occurred in Stage III and Stage IV patients. We have discontinued blind biopsies of nonpalpable nodes in the staging procedure. Therefore staging classifications that use nonpalpable nodes and the presence of dermatopathic changes as independent criteria for staging may not be entirely valid. In the future, the use of specific gene rearrangements to identify nodes involved with mycosis fungoides cells will become a more sensitive staging device. ~0 Seventy-three percent of the patients in Group 1 with more than two follow-up examinations after the completion of the electron beam radiationchemotherapy protocol had intermittent elevations of S6zary cells (small-cell variant) (11% to 30% of 100 lymphocytes) regardless of whether they were in clinical remission or not. Seventeen patients who had one to four serial chromosome analyses of their peripheral blood after completion of combination therapy continued to have abnormal karyotypes regardless of whether their skin was clear or whether they had nonspecific or specific lesions. Our studies also confirm those of Bunn et al, H who found abnormal karyotypes in cultures of peripheral blood lymphocytes even though the number of circulating S6zary-like cells was within the normal range and vice versa. The persistence of cytogenetic abnormalities and elevated levels of circulating S6zary cells post therapy, when the skin has clinically returned to normal, implies that the underlying pathogenetic mechanism is still operating. Pretreatment elevated IgE levels, which have been proposed as a sign of chronic antigenic stimulation in mycosis fungoides, lz remained unchanged after therapy in our series. In addition, the normal-appearing posttherapy lesional skin still showed occasional mycosis fungoides cells in the epidermis and dermis albeit without an accompanying inflammatory cell infiltrate. The nonspecific lesions showed similar
Volume 16 Number 1, Part 1 January 1987
Electron beam irradiation and chemotherapy for MF
histologic features that may or may not have evolved into classic mycosis fungoides over the course of many months. Thus we have the anomolous situation wherein visual inspection and routine laboratory examinations indicate that the patient is in a clinical remission or a disease-free interval. However, we have not produced a disease-free interval in any of our patients, even though by conventional clinical and laboratory criteria we would be permitted to place them in this category. Operationally, it would still be proper t~ call this state a clinical remission because the patient is free of pruritus and diagnostic skin lesions. Even when the patient has developed the clinically nonspecific eruptions, he or she usually is still asymptomatic. The nonspecific lesions have probably been overlooked in other studies because their numbers are usually small and have a nondescript clinical appearance. On the basis of our observations in eighteen cases (from this series and those of other patients treated between 1974 and 1979) the histologically verifiable nonspecific eruption has to be considered as a potential early sign of relapse in mycosis fungoides and may have to be used as an indicator in evaluating therapeutic responses. However, this nonspecific eruption does not invariably evolve into mycosis fungoides. It may disappear spontaneously or following the use of topical steroids. Our preliminary findings in two cases indicate that topical nitrogen mustard will produce a complete clinical resolution of such lesions and eliminate almost completely the "mycosis fungoides" cells from these sites. The prognostic significance of the nonspecific lesions has not yet been determined. The normal-appearing skin of mycosis fungoides patients, even before therapy, has single "mycosis fungoides" cells within the epidermis and dermis, indicating that mycosis fungoides is generalized in the skin at the earliest times of diagnosis. The histologic features of the nonspecific eruption, which appears as subtle macular lesions that may evolve into clinically recognizable mycosis fungoides lesions after the development of a normal inflammatory cell infiltrate, emphasize the point that nontumorous mycosis fungoides skin lesions are primarily caused by the accompanying normal cellular infiltrate and not by the "mycosis fungoides" cells themselves, which constitute
$9
only a minority of the cells in mycosis fungoides lesions. This point has also been emphasized by Sanchez and Ackerman. ~3 Our studies indicate that mycosis fungoides is a generalized disease occurring in the skin and extracutaneously at the earliest times of biopsy diagnosis. It has been argued that therapy for mycosis fungoides is generally unsuccessful because only the skin is treated. The lymph nodes and viscera, which are therapeutically neglected, then serve as the sites from which reseeding of the skin occurs. However, on the basis of our observations of "mycosis fungoides" cells in the normal-appearing, uninvolved mycosis fungoides skin, in the sites of successfully treated lesions, and in the "tolerance" or "recapitulative" histologic picture of nonspecific lesions, along with the persistence of cytogenetic abnormalities and the elevated levels of circulating S6zary cells during apparent clinical remissions or disease-free intervals, we suggest an alternate possibility. Regardless of how successfully we may lower the tumor load in the skin and nodes by electron beam radiation and chemotherapy, we are unable to stop the continuing underlying pathogenetic process that is evident during the "clinical remission." It is this unchecked process that may be responsible for eventual cutaneous recurrences and not simply metastatic seeding of the skin from sanctuary sites, such as viscera and nodes, that are unaffected by our primary combination therapy. We should be trying to control the underlying pathogenetic mechanism in addition to killing all the tumor cells. The "tolerance" or "recapitulative" histologic picture may be the morphologic expression of the partial inhibition of this pathogenetic mechanism. This concept is dealt with in more detail in another article.3 During the course of these studies, we observed three patients with early mycosis fungoides who developed acute biopsy-proved mycosis fungoides lesions, eczematous and papular, that disappeared spontaneously over the course of 2 to 4 weeks. This behavior is consistent with the concept of underlying host resistance. The same behavior was seen in the waxing and waning of some of the nonspecific lesions post therapy. Although it is argued that the presence of circulating "mycosis fungoides" or S6zary cells indicates extracutaneous involvement, there is no
60
Journal of the American Academy of Dermatology
B r a v e r m a n et al
evidence that such circulating cells, by t h e m s e l v e s , either p r o m o t e a high risk f o r mycosis fungoides tumor formation in v i s c e r a l organs or reduce survival time because o f visceral i n v o l v e m e n t with organ dysfunction. M y c o s i s fungoides t u m o r formation in the viscera is an u n c o m m o n finding both in the literature and in our personal experience, t4 Our current view is that prolonged survival and perhaps eventual cure will be possible either b y eradicating all the " m y c o s i s f u n g o i d e s " ceils with more effective treatment regimens or b y reducing the n u m b e r of " m y c o s i s f u n g o i d e s " cells to a m i n i m u m , so that whatever natural host resistance is present might have an opportunity to be effective. T h e three cures w e described at the beginning o f this report were associated with a g g r e s s i v e electron b e a m radiation and chemotherapy. REFERENCES
1. Lamberg SI, Green SB, Byar DP, et at: Status report of 376 mycosis fungoides patients at 4 years: Mycosis Fungoides Cooperative Group. Cancer Treat Rep 63:701707, 1979. 2. Braverman IM, Maynard T: Cytogenetic studies in mycosis fungoides. (Submitted for publication.) 3. Braverman IM, Klein S, Grant A: Electron microscopic and immunolabeling studies of the lesional and normal skin of patients with mycosis fungoides treated by total body electron beam irradiation. J AM ACAD DERMATOL 1987; 16:61-74. 4. Hoppe RT, Fuks Z, Bagshaw M_A: The rationale for curative radiotherapy in mycosis fungoides. Int J Radiat Oneol Biol Phys 2:843-851, 1977.
5. Hoppe RT, Cox TS, Fuks Z, et al: Electron-beam therapy for mycosis fangoides: The Stanford University experience. Cancer Treat Rep 63:691-700, 1979. 6. Tadros AAM, Tepperman BS, Hryniuk WM, et al: Total skin electron irradiation for mycosis fungoides: Failure analysis and prognostic factors, hat J Radiat Oncol Biol Phys 9:1279-1287, 1983. 7. Hamminga L, Hermans J, Noordijk EM, et al: Cutaneous T-cell lymphoma: Clinicopathological relationships, therapy and survival in ninety-two patients. Br J Dermatol 107:145-156, 1982. 8. Cotter GW, Baglan RJ, Wasserman TH, Mill W: Palliafive radiation treatment of cutaneous mycosis fungoides: A dose response. Int J Radiat Oncol Biol Phys 9:14771480, 1983. 9. Griem ML, Tokars RP, Petras V, et al: Combined therapy for patients with mycosis fungoides. Cancer Treat Rep 63:655-657, 1979. 10. Weiss LM, Hu E, Wood GS, et al: Clonal rearrangements of T-cell receptor genes in mycosis fungoides and dermatopathic lymphadenopathy. N Engl J Med 313:539544, t985. 11. Bunn PA Jr, Huberman MS, Whang-Peng J, et al: Prospective staging evaluation of patients with cutaneous Tcell lymphomas: Demonstration of a high frequency of extracutaneous dissemination. Ann Intern Med 93:223230, 1980. 12. Tan RA, Butterworth CM, McLaughlin H, et al: Mycosis fungoides: A disease of antigen persistence. Br J Dermatol 91:607-616, 1974. 13. Sanchez JL, Ackerman AB: The patch stage of mycosis fungoides: Criteria or histologic diagnosis. Am J Dermatopathol 1:5-26, 1979. 14. Rappaport H, Thomas LB: Mycosis fungoides: The pathology of extracutaneous involvement. Cancer 34:11981229, 1974.
ABSTRACTS Cutaneous lesions of sinus histiocytosis with massive iymphadenopathy: (Destombes-Rosai-Dor fman) associated with mycetoma. Description of a case
Rongioletti F, Heid E, Grosshans EG. G Ital Dermatol Venereol 1985;120:419-23 (Italian) A 22-year-old Tunisian woman when first seen had cutaneous lesions in which histopathologicexaminationdemonstratedthe typical
microscopic features of sinus histiocytosis with massive lymphadenopathy, tt is important to recognize that this condition, despite its resemblance to malignant histioeytosis, has a benign course, and to avoid treating these patients with dangerous drugs. Ten percent of all patients have akin involvement. Fehudi M. Felrnan, M.D.
Zinsser-Cole-Engman syndrome: Two cases transmitted by an autosomal dominant gene
Gasparini G, Sambuani N, Guidarelli C, Sarchi GD, Petro A, Raimondi E, Romagoni MM, et al. G Ital Dermatol Venereol 1985;120:429-33 (Italian) Two patients, a father and son, with Zinsser-Cole-Engman syndromes are reported. When first seen both had the typical symptoms triad of poikilomelanodermia, leukoplakia, and nail dystrophy. No other significant abnormalities were present. Genetic transmission prognosis and treatment of congenital dyskeratosis are discussed. Yehudi M. Felman, M.D.
IIII
I
II
II
II
I
I
II
I
Combined total body electron beam irradiation and chemotherapy for mycosis fungoides Irwin M. Braverman, M.D.,* N. Bruce Yager, M.D.,* Michael Chen, M.D., Ph.D.,** Edwin C. Cadman, M.D.,*** William N. Hait, M.D., Ph.D.,*** and Theresa Maynard, B.S., C.T.(ASCP)* New Haven, CT Since 1979 a protocol of total body electron beam therapy (3,600 rads; 6 MeV), followed by six monthly cycles of chemotherapy (doxorubicin, 30 mg/M2 given intravenously once monthly, and eyclophosphamide, 100 mg/Mz given orally each day for 14 days), has been used to treat fifty patients with mycosis fungoides (primarily Stages I and II). A group of twenty-four patients, treated by identical high-dose electron beam therapy alone, served as control subjects. Actuarial analysis by the Kaplan-Meier method and statistical analysis by the generalized Wilcoxon test of Gehan demonstrated a significant difference (p = 0.008) in the probability of Stages I and Ir patients' remaining in complete clinical remission when combination therapy was compared with high-dose electron beam therapy alone. No statistically significant difference was demonstrated in patients in Stages III and IV mycosis fungoides. Although 60% of patients were in "complete clinical remission," the longest follow-up being 75 months, all continued to show karyotypic abnormalities of circulating lymphocytes, and 70% had intermittently and abnormally elevated blood levels of Sgzary cells. (J AM ACADDERMATOL 1987; 16:45-60.)
Since 1974 we have been treating patients having mycosis fungoides with high-dose total body electron beam radiotherapy (6 MeV; 3,600 rads). Although the clinical remissions were longer than those following low-dose electron beam radiotherapy (3-6 MeV; 800-1,600 rads), which we had used before 1974, the survival time of patients had not increased. During 1974 to 1976 we treated three cases of mycosis fungoides (Stage III, Mycosis Fungoides Cooperative Group staging) with
From the Departments of Dermatology,* Therapeutic Radiotherapy,** and Internal Medicine (Medical Ontology),*** Yale University School of Medicine. Supported by National Institutes of Health Grant No. 21153. Accepted for publication June 19, 1986. Reprint requests to: Dr. Irwin M. Braverman, 333 Cedar St., New Haven, CT 06510/203-785-4092.
multiple-drug chemotherapy* for 3 to 4 months either before or after a course of high-dose electron beam radiotherapy. All three patients entered remission. Two have remained in remission for 10.5 and 11.0 years, respectively. The third patient died 3.5 years later from cancer of the colon, but no evidence of mycosis fungoides was found at autopsy. Encouraged by the experience with these three cases, we initiated a treatment protocol of high-dose electron beam therapy followed by 6 months of chemotherapy for all newly diagnosed and previously untreated cases of mycosis fungoides that were referred to the Yale-New Haven Medical Center. The disease of all patients was staged according to the Mycosis Fungoides Co*The multiple-drug chemotherapy consisted of cyelophosphamide, bleomycin, methotrexate, cytosine arabinoside, and doxorubicin.
45
46
Journal of the American Academy of Dermatology
Braverman et al
Table I. Mycosis Fungoides Cooperative Group stages
Stage
[
TNeategory
I
TiNo_l
II
T1N~+
Plaques, papules, or eczematous patches covering less than 10% of the surface and no more than one site of clinically enlarged peripheral lymph nodes Plaques, papules, or eczematous patches covering less than 10% of the surface and two or more sites of clinically enlarged peripheral lymph nodes Plaques, papules, or eczematous patches covering 10% or more of the skin surface and no more than one site of clinically enlarged peripheral lymph nodes Plaques, papules, or eczematous patches covering 10% or more of the skin surface and two or more sites of clinically enlarged peripheral lymph nodes Tumors (one or more); any number of clinically enlarged peripheral lymph node sites Generalized erythroderma; any number of clinically enlarged peripheral lymph node sites
T2No-I
IU
T~N~+
T3Naay IV
Definition
[
T4N~y
Tt: Plaque, papules, or eczematous patches coveting <:10% of the body; T2: same as above but covering ~10% of the body; T3: tumors, one or more with or without plaques or papules; T4: erythroderma; No: no nodal areas involved clinically; N~: one nodal area involved; N2: two nodal areas; N3; three nodal areas; N4 to N~: four to eight nodal areas.
Table H. Characterization by stage, age, sex, and group
Stage and age
Sex
Group 1: EB-Chemo
Group 2: High EB
Group 3: Low EB
I
M F M F M F M F M F
11 9 10 10 3 4 1 2 53 49
4 1 7 2 4 0 2 4 66 56
1 0 2 4 1 4 3 3 65 57
II III IV Average age (yr)
EB-Chemo: Combined electron beam radiation and chemotherapy; high EB: high-dose electron beam radiation; low EB: low-dose electron beam radiation.
operative Group clinical criteria. 1 In addition, special hematologic, cytogenetic, and histologic studies were performed on peripheral blood, lymph node biopsies, and cutaneous lesions. This article reports the results of this treatment protocol as of Sept. 30, 1985. The details of the cytogenetic and histologic studies are described elsewhere? ,~ MATERIALS AND METHODS The patient population consisted of three groups. Group I contained fifty patients who had been entered
into the study since 1979. All were newly diagnosed and had not received any significant prior therapy. These patients received high-dose electron beam radiotherapy followed by 6 months of two-drug chemotherapy. Group 2 consisted of twenty-four newly diagnosed and previously untreated individuals who received highdose electron beam radiotherapy alone in a manner identical to those in Group 1. Nineteen Of these patients were treated from 1974 to 1979, and five have been treated since 1979. Group 3 consisted of eighteen patients who were treated by low-dose electron beam radiotherapy from 1963 to 1974. In all patients the di-
Volume 16 Number 1, Part 1 January 1987
Electron beam irradiation and chemotherapy for MF 47
agnosis of mycosis fungoides was made on clinical grounds and confirmed by one or more skin biopsies that showed a polymorphous papillary dermal infiltrate containing lymphocytes with darkly staining convoluted nuclei ("mycosis fungoides" cells). Pautrier microabscesses containing "mycosis fungoides" cells were present in all cases. The disease of all the patients was staged according to the original clinical criteria described in the report of the Mycosis Fungoides Cooperative Group j (Table I). These criteria could be applied to the patients in Groups 2 and 3 because either photographs of their entire skin surface were available or the extent of their disease had been carefully documented in their clinical records. In addition, one of us (I. M. B.) had been involved in the care of every patient in Groups 1 and 2 and most of the patients in Group 3. The characterization of the patients by age, sex, Mycosis Fungoides Cooperative Group stage, and therapy groups is shown in Table II. By history, cutaneous lesions had been present for 3 to 483 months (mean • SD: 85.3 _-+ 104.9 months) before biopsy diagnosis was made in Group I patients. Therapy was begun within 4 to 6 weeks after biopsy diagnosis in forty-five of fifty patients. In the remaining five patients, therapy was initiated 12 to 13 months after biopsy diagnosis. These five patients either had spot orthovoltage radiation for single lesions or had been treated by topical steroids before being referred for further therapy. In Group 2, nineteen of twenty-four patients received only high-dose electron beam radiotherapy within 4 to 8 weeks of biopsy diagnosis, and five had been treated with topical steroids for 10 to 28 months before being referred for electron beam radiotherapy. Cutaneous lesions had been present for 9 to 447 months (mean +- SD: 99.9 +-- 104.8 months) before biopsy diagnosis had been made in Group 2 patients. There was no statistical difference between the patients in Groups 1 and 2 with regard to this parameter. In Group 3, nine of eighteen patients received low-dose electron beam therapy within 5 months of diagnosis, and in nine patients the interval was 11 to 96 months. Most of the Group 3 patients were treated from 1963 to 1965, when electron beam radiotherapy first became available as a therapeutic modality at the Yale-New Haven Medical Center. Prior to electron beam radiotherapy for the patients in Group 1, the following staging procedures were performed. The extent of the skin lesions was estimated by measuring with the entire palm of the hand, which is approximately 1% of the body surface. The following tests were performed: chest x-ray, liver-spleen radionuclide scan, abdominal computed tomography scan, skin testing with recall antigens (Candida, trichophytin,
purified protein derivative, and mumps), complete blood cell count, urinalysis, liver function tests, routine chemistry studies, and determination of the radionuclide cardiac ejection fraction. Additional tests were as follows: quantitation of serum immunoglobulins in the clinical immunology laboratory of the medical center by standard nephelometer methods (G, M, A), by radial immunodiffusion (D), and by radioimmunoassay (E); search for and enumeration of the number of S6zary cells (large and small variants) per 100 lymphocytes found in routine blood smears and in buffy coats embedded in Spurr's resin for examination in 1 p,m sections; skin biopsies of both normal-appearing, uninvolved skin and lesional skin for study by light and electron microscopy; biopsy of palpable nodes or a "blind" biopsy of axillary or inguinal nodes that drained mycosis fungoides skin lesions, performed in nineteen of fifty Group 1 patients; and cytogenetic studies with Giemsa G banding on lymphocytes derived from peripheral blood and biopsy specimens of lymph nodes. (The lymph nodes were interpreted in the department of surgical pathology at Yale.) Following completion of electron beam radiotherapy, the skin was rephotographed; the sites of previous lesionai skin and uninvolved skin were repeated with photographic documentation; serum immunoglobulins were remeasured; and cytogenetie studies on peripheral blood and a search for S6zary cells in routine smears and buffy coats prepared for 1-p,m sections were also repeated. After completion of chemotherapy, the same studies were repeated at 6- to 12-month intervals unless clinical events indicated an earlier reevaluation. If unusual skin lesions arose, they were photographed and biopsy specimens were obtained for light and electron microscopic examination. All patients were seen at lto 3-month intervals as their status required, at which times a complete physical examination, including inspection of the entire skin surface, was done and a search for palpable lymph nodes made. Of fifty-six consecutive mycosis fungoides patients, fifty agreed to participate in the protocol. Two men, 40 and 62 years old, both of whom had lesions of mycosis fungoides appearing as parapsoriasis and poikiloderma atrophicans vasculare involving only 1% and 2% of their body surface, respectively, elected to be treated by topical nitrogen mustard alone. Four patients were treated by high-dose electron beam radiotherapy alone (two women, 67 and 74 years old, and one man, 19 years old, refused chemotherapy; one man, 59 years old, had decreased cardiac function, precluding the use of chemotherapy). Electron beam radiotherapy for patients in Group 3 consisted of low total doses of radiation ranging from
48
Journal of the American Academy of Dermatology
Braverman et al
600 to 1,600 rads (mean: 1,100 rads) of mostly 2.5 to 3 MeV electrons. Since 1974 we have used the sixfield whole body technic of electron beam radiotherapy developed at Stanford University,4 but without the dual angulation of the beam for each field. The patients were treated with 6 MeV electrons produced by a Clinac-6 linear accelerator (u Associates Inc., Palo Alto, CA), in contrast to the 4 MeV electrons used at Stanford. We have calculated that these 6 MeV electrons degrade to an energy level of 4.4 MeV at the patient's skin surface. The patients received whole skin doses of 3,600 rads in 9 weeks at 100 rads/day, 4 days a week. A complete six-field treatment cycle requires two con: secutive treatment days. Booster doses of orthovoltage radiation (2,000 rads at 120 or 125 kVp) were given to the soles of the feet and to the perineum in a split course: 100 rads/day, 4 days a week, to 1,000 fads during the first and tast ten treatment days. If eyelid involvement by mycosis fungoides was present or suspected, internal lead eye shields were employed during most of the electron beam treatment. Within I month after the completion of electron beam therapy, chemotherapy was begun: doxorubicin, 30 mg/M 2 given intravenously once monthly, and oral cyelophosphamide, 100 mg/Mz daily for 14 days each month for 6 months. From 1974 to 1979, while treating patients in Group 2 with high-dose electron beam radiotherapy, we had observed the development of clinically nonspecific-appearing lesions, one to six in number, 4 to 46 months after the completion of electron beam therapy that had produced a complete remission. The lesions measured 5 to 20 mm and resembled patches of asteatotic eczema or pityriasis rosea. Histologically this eruption showed mycosis fungoides ceils in the epidermis and dermis without the usual accompanying dense, polymorphous, inflammatory cell infiltrate. In seven of eight patients, these lesions eventually evolved into typical mycosis fungoides lesions both clinically and histologically 4 to 23 months later. In the eighth patient the nonspecific lesions are still present and are unchanged 63 months later. In Group 1 patients, identical lesions have been observed at approximately the same rate, but a higher percentage have remained unchanged or have disappeared either spontaneously or following the use of topical steroids. These "nonspecific" lesions do not resemble the patients' original eruption. Other patients in Groups 1 and 2 have relapsed with lesions that clinically and histologically are characteristic of mycosis fungoides and resemble the initial eruption. We call these "specific" lesions. For Groups 1 and 2, we have precise information,
with biopsy and photographic documentation, of the times of onset of the nonspecific lesions and of their evolution into specific lesions if that event occurred. Except for one patient in Group 3, in whom nonspecific and specific lesions were commented on in the records, only specific lesions were recorded as a sign of clinical relapse following electron beam radiotherapy. None of the patients in these three groups had detectable visceral involvement at the time therapy was initiated. Life table analyses using the product-limit method of Kaplan and Meier* were performed on the data. Survival statistics were calculated from the initiation of electron beam therapy to death from any cause. Periods of clinical remission were measured in two ways: (1) from the end of electron beam therapy to the appearance of those nonspecific lesions that eventually evolved into specific lesions and (2) from the end o f electron beam therapy to the appearance of specific lesions. Because nonspecific lesions have spontaneously disappeared or have continued to persist for as long as 63 months without changing their clinical appearance or histologic features, we have not considered them to be end points for determining clinical relapse unless their clinical appearan.ce and histologic picture evolved into the diagnostic features of mycosis fungoides. The estimated probabilities of survival and o f a patient's remaining in complete clinical remission produced by the three forms of therapy were compared for patients in comparable stages. The generalized Wilcoxon test of Gehant for comparing two life tables was used to assess the statistical significance of the data. These studies were carded out the with approval o f the Yale School of Medicine Human Investigation Committee.
RESULTS The combination of electron beam irradiation and two-drug chemotherapy was well tolerated. Mild neutropenia was observed in all patients (3,000-4,500), but none experienced severe neutropenia or life-threatening infections. Neither clinical compromise in cardiac function nor hemorrhagic cystitis was seen. There were no significant side effects beyond those normally seen with electron b e a m radiotherapy. Electron beam therapy produces complete body hair loss beginning *Kaplan EL, Meier P: Nonparametricestimationsfrom incomplete observations. Am Stat Assoc J 53.457-481, 1958. "['GelaanEA: A generalizedWilcoxontest for comparingarbitrarily singly-censoredsamples.Biometrika52:203-223, 1965.
Volume 16 Number 1, Part 1 January 1987
Electron beam irradiation and chemotherapy for MF
in the third week of therapy. Hair began to regrow within 6 to 8 weeks after the completion of electron beam therapy in spite of the chemotherapy regimen. Ankle edema developed to variable degrees during the last 2 weeks of electron beam therapy, being most severe in patients with leg vein excisions required for coronary artery bypass surgery. Erythema, hyperpigmentation, and dry skin with exfoliation occurred to a variable degree in all patients. The skin returned to normal color and texture within 6 to 12 weeks after the completion of electron beam therapy. Neither radiodermatitis nor ectropion has developed in any patient. Most patients developed moderate generalized hypohidrosis affecting the entire skin surface except for the axillae and scalp. In most of the affected patients the return of sweating has been incomplete even after 3 to 4 years. Forty-nine of fifty patients in Group 1 had either a complete response or 95% clearing of their skin lesions at the end of electron beam radiotherapy. One patient, who had erythroderma but not classic S6zary syndrome, did not respond and is currently receiving experimental chemotherapy. In four of five patients with 95% clearing, the lesions resolved during chemotherapy. In the patient whose lesions failed to resolve completely, a relapse occurred 4 months after the completion of electron beam therapy. Of the fifty patients, twenty-two have remained free of skin lesions (nine for 30-75 months, ten for 14-26 months, and three for 2-10 months). Eighteen of the fifty patients developed nonspecific cutaneous lesions 6 to 23 months after the end of electron beam therapy, which appeared as 1-2 cm, ill-defined patches of dry, scaly, sometimes hypopigmented skin indistinguishable from skin with xerosis or pityriasis alba or from oval, slightly scaly patches similar to those of pityriasis rosea or guttate parapsoriasis. There were only one to six such lesions when they first appeared, in nine individuals the lesions have either waxed and waned spontaneously, disappeared spontaneously, or disappeared with the use of topical steroids during an observation period of 39 to 63 months. In the other nine, they increased in number over the course of 4 to 6 weeks. Biopsy of the lesions in all eighteen patients at the time the lesions arose
49
showed the "tolerance" histologic picture described in paragraph 14 of this section. In the nine individuals whose nonspecific lesions increased, the lesions became infiltrated and more numerous and eventually evolved into mycosis fungoides both clinically and histologically 5 to 20 months after their initial appearance. In none of the eighteen individuals did the nonspecific lesions resemble the initial eruption of mycosis fungoides. Nine patients relapsed with specific lesions identical to their original ones without going through the stage of having nonspecific lesions. These relapses occurred from 1 to 13 months after the end of electron bean therapy except for one relapse at 19 months. Eight of these nine individuals had nodules, tumors, or erythroderma as the presenting signs of mycosis fungoides, rather than the eczematous patches and poikilodermatous lesions characteristic of most of the other patients. Five patients have died: one died from staphylococcal sepsis and aplastic anemia and a second from staphylococcal sepsis resulting from bone marrow replacement by S6zary cells but with clinically and histologically clear skin; two with visceral mycosis fungoides died from infection; one died from a pulmonary embolus while the disease was under control by chemotherapy; and one died from congestive heart failure resulting from arteriosclerotic heart dsease complicated by a cardiac ventricular aneurysm. This last patient had minimal skin involvement and no visceral involvement by mycosis fungoides. None of these five deaths occurred during the chemotherapy phase immediately following electron beam therapy. Of the fifteen living patients with specific skin lesions, topical nitrogen mustard or treatment with psoralens and ultraviolet A is keeping eight under excellent control. Only seven patients have progressive disease requiring multiple-drug chemotherapy. In eight of twenty-four patients in Group 2 who were treated by the identical electron beam therapy regimen but without chemotherapy, the same types of nonspecific lesions with "tolerance" histologic features developed 4 to 46 months after the end of electron beam therapy. In seven patients, evolution into classic mycosis fungoides lesions both morphologically and histologically occurred 4 to 23 months later. The eighth patient still has non-
50
Journal of the American Academy of Dermatology
Braverman et al
-%
89
.9
~'T I
1
.Z-
I,,-4 ~ . .6-
re
(.,9 .~r'l
,4_
$0 kd
.2"
re
.... 9
.
,
20
,
40
.... . ,
,
60
.
100
,
le0
.
,
..........
140
,
160
.
;
18
.
,
~00
MONTHS Fig. 1. Actuarial curves of the estimated probabilities of survival for combined clinical Stages I and II in all three treatment groups. The numbers of patients at risk at specific break points were as follows: with combination therapy (COMB RX), 40 patients at 0 month, 34 at 14 months, 24 at 24- months, 21 at 27 months, 13 at 42 months, 7 at 54 months, 5 at 64 months, 2 at 68 months, and 1 at 77 months; with high-dose electron beam (EB) radiation, 14 patients at 0 month, 13 at 3 months, 8 at 38 months, 7 at 47 months, 5 at 63 months, 2 at 98 months, and 2 at 105 months; and with low-dose electron beam radiation, 7 patients at 0 month, 6 at 1 month, 5 at 6 months, 4 at 34 months, 3 at 41 months, 2 at 82 months, 2 at 134 months, and 1 at 180 months.
specific lesions that have remained unchanged for 63 months. Eight others in Group 2 had a clinical relapse characterized by classic mycosis fungoides skin lesions, with the typical histopathologic picture of mycosis fungoides, 5 to 18 months after the start of electron beam therapy without going through the stage of nonspecific lesions. For the patients in Group 3 treated with lowdose electron beam therapy (3 MeV; 800-1,600 rads), we have reliable information only on the duration of clinical remission, based on relapse with specific lesions, and the date of death. Nonspecific lesions were described in only one patient. Figs. 1 and 2 show the actuarial curves of the estimated probability of survival for all three treatment groups. Table III presents the levels of sta-
tistical significance for the comparisons between the various life table curves. Combination therapy, in comparison with low-dose electron beam therapy alone, has increased the survival rates for pa-. tients in all stages. For early mycosis fungoides (Stages I and II), high-dose electron beam therapy, in comparison with low-dose electron beam therapy, produces an improvement in survival rates approaching a level of significance (p = 0.06). No effect was seen in late mycosis fungoides (Stages III and IV). However, when the estimated probability of a patient's remaining in complete clinical remission was calculated with specific lesions used as end points and with comparable stages compared, a significant difference was demonstrated between
Volume 16 Number 1, Part I January 1987
Electron beam irradiation and chemotheraPY for MF
51
.92
~
.7-
~.
.6-
cz:
~
:D r
~
.'~
21.,., 0
10
I
L~
v
"
!
~)
I
40
!
~
60
'1"
~
I
70
r
B0
~
'
9~
I
I
lO0 110
9
f
1~0
!
l~O
'
t
140
'
!
l~
MONTHS -'44-'N- CBCB IO/,-(H}-- NlfiH s
~ r
L[]tJ EB
Fig. 2. Actuarial curves of the estimated probabilities of survival for combined clinical Stages III and IV in all three treatment groups. The numbers of patients at risk at specific break points were as follows: with combination therapy (COMB RX), 10 patients at 0 month, 4 at 23 months, and 1 at 33 months; with high-dose electron beam (EB) radiation, 10 patients at 0 month, 9 at 5 months, 7 at 19 months, 2 at 37 months, and 1 at 78 months; and with low-dose electron beam radiation, 9 patients at 0 month, 9 at 3 months, 7 at 10 months, 5 at 23 months, and 2 at 48 months. the results produced by high-dose electron beam therapy and those produced by low-dose electron beam therapy for advanced disease (Fig. 3; Table IV) but not for early disease (Fig. 4). Figs. 5 and 6 show the life table curves of the probability of a patient's remaining in complete remission when a combination of electron beam therapy and chemotherapy is compared with highdose electron beam therapy alone (Stages 1 and II and Stages III and IV, respectively) (Table V). Nonspecific lesions were used as end points in both comparisons. There is a significant difference when Stages I and II are compared (p = 0.008) but not when Stages III and IV are compared. When specific lesions are used as end points (Figs. 3 and 4), a significant difference is found when Stages I and II are compared (p = 0.01) but not when Stages III and IV are compared
HI. Levels of statistical significance for estimated probabilities of survival Table
Stage
I, II I, II I, II III, IV III, IV III, IV
I
Therapy
EB-Chemo vs high EB High EB vs low EB EB-Chemo vs low EB EB-Chemo vs high EB High EB vs low EB EB-Chemo vs low EB
I
pValue
0.193 0.06 0.002 0.07 0.09 0.004
See Table II for explanation of abbreviations.
(p = 0.36). Combination therapy increases the probability of a patient's remaining in complete remission for early-stage disease (Stages I and 1I) regardless of end point used. The actuarial curves for patients in Stages [ and II were identical, as were the curves for patients
52
Journal of the American Academyof Dermatology
Braverman et al
~
"7" r'-I
.e-
III
(")
.7-
I,I C~
.6-
]
.~I.,-,I
..I 0
.4-
t:~ []
.:~"
CL. (.,,)
Id
~'
!
10
'"
I
eO
'
I
aO
1
I
40
'
I
"
I
60
~0
"
I
70
,
I
80
,,T
I
90
'
I
xoo
'
I
110
'
I
' I
'
I
1~0 130 240
MONTH8 HIGH E~ ~ ~ LOW El
Fig. 3. Actuarial curves of the estimated probabilities of clinical remissions for combined Stages HI and IV in all three treatment groups, with specific lesions used as end points. The numbers of patients at risk at specific break points were as follows: with combination therapy (COMB RX), 9 patients at 0 month, 7 at 4 months, 4 at 12 months, 1 at 17 months, and 1 at 31 months; with high-dose electron beam (EB) radiation, 9 patients at 0 month, 5 at 4 months, 3 at 12 months, and 1 at 17 months; and with low-dose electron beam radiation, 2 patients at 0 month. Table IV. Levels of statistical significance for estimated probabilities of clinical remission, with specific lesions used as end points Stage
Therapy
I, II III, IV I, II
High EB vs low EB High EB vs low EB EB-Chemo vs high EB
]
p Value
0.08 0.0002 0.01
See Table II for explanation of abbreviations.
in Stages III and IV in all of the analyses described here. Therefore the data for Stages I and II and for Stages III and IV were pooled. Fig. 7 is a composite graph of the estimated probabilities of a patient's remaining in complete clinical remission for Stages I and II following combination therapy and high-dose electron beam
therapy. The actuarial curve from Group 2 is superimposable on the curves from three other studies in which high-dose electron beam therapy alone was used for patients with comparably staged disease (Hoppe et al, 5 Fig. 5; Tadros et al, 6 Fig. 3; Hamminga et al, 7 Fig. 7). The detailed histologic studies of the "nonspecific" lesions are reported elsewhere 3 and will be described only briefly here. At the time of initial diagnosis, all patients showed the classic histologic features of mycosis fungoides: a papillary dermal infiltrate composed of normal-appearing lymphocytes and histiocytes admixed with a few plasma cells and eosinophils. Within the infiltrate there were varying numbers of mononuclear cells with
deeply stained hyperconvoluted nuclei (mycosis fungoides cells). The "mycosis fungoides" cells were present singly and in groups (Pautrier microabscesses) within the epidermis. They were
V o l u m e 16
Number I, Part 1 January 1987
Electron beam irradiation and chemotheraPY for MF
53
.9
7"
r-!
.8-
.7-
La.I , 6 -
l-.4
_1 (..)
.4-
m ,3.
I'-
.2-
Ld ,1
'
1
l0
''
I
~1
'
I'
~0
''t''
I
4{I
I
~
'
I
60
'
('""'
"l
70
'
~
I
'
-1""
c/O 100
]
llO
~
l
le0
r-"r'-v'-~
1~
140
MONTH8 Kew:
L; ~, COHBI~ ~ ~ HIC.d,4s
~ f.
LI]WEB
Fig. 4. Actuarial curves of the estimated probabilities of clinical remissions for combined stages I and II in all three treatment groups, with specific lesions used as end points. The numbers of patients at risk at specific break points were as follows: with combination therapy (COMB RX), 40 patients at 0 month, 34 at 8 months, 21 at 19 months, 14 at 31 months, 6 at 49 months, 2 at 67 months, 1 at 72 months, and 1 at 75 months; with highdose electron beam (EB) radiation, 14 patients at 0 month, 8 at 7 months, 4 at 16 months, 3 at 31 months, 1 at 59 months, and 1 at 96 months; and with low-dose electron beam radiation, 7 patients at 0 month, 4 at 5 months, 2 at 24 months, and 1 at 134 months.
also present singly and in clusters in the dermis both in perivascular and nonperivascular sites. In all patients in Groups 1 and 2, repeat biopsy of a lesional site that had returned to normal after either electron beam therapy alone or electron beam therapy with chemotherapy disclosed a unique picture. Single "mycosis fungoides" cells were still present both in the dermis and epidermis, and only rarely were they found in clusters in the epidermis. The usually dense, accompanying polymorphous infiltrate of normal inflammatory cells was either absent or minimal. Further, in eight patients, biopsies of normal uninvolved skin, taken both before therapy and after therapy, showed occasional "mycosis fungoides" ceils in the epidermis and dermis without any accompa-
Table V. Levels of statistical significance for estimated probabilities of clinical remission, with nonspecific lesions used as end points Stage
Therapy
I,.p Value
l, II HI, IV
EB-Chemo vs high EB EB-Chemo vs high EB
0.008 0.166
See Table II for explanation of abbreviations.
nying inflammatory infiltrate. We had never seen such findings in the skin of healthy individuals. Biopsies of the nonspecific cutaneous lesions that appeared post therapy showed histologic features similar to those found in the posttreatment
54
Journal o f the American A c a d e m y of Dermatology
Braverman et al
,9-
7" .er"l r(f) 3 I.M .6-
I,==I
.-I ixl [] FI I.-G~
.2-
W ,I-
o
9
I
'
I
'
I
'
I'
'
I
'
'
' ~I~
o
'
I
,
I
'
I
'
'
MONTHS Fig. 5. Actuarial curves of the estimated probabilities of clinical remissions for combined Stages I and II in the combination therapy (COMB RX) and high-dose electron beam (EB) therapy groups. Nonspecific lesions were used as end points. The numbers of patients at risk at specific break points were as follows: with combination therapy, 40 patients at 0 month, 36 at 6 months, 23 at 16 months, 18 at 22 months, 6 at 44 months, 2 at 67 months, and 1 at 75 months; and with high-dose electron beam radiation, 14 patients at 0 month, 7 at 7 months, 3 at 13 months, 2 at 46 months, 1 at 52 months, and 1 at 96 months. lesional sites9 "Mycosis fungoides" cells were more numerous, but an inflammatory infiltrate was still absent or minimal. We have referred to this phenomenon as the "tolerance" histologic picture or as the "recapitulative" phase of the disease9 In eight patients we were able to follow the evolution of nonspecific lesions into specific ones with serial biopsies. The "mycosis fungoides" cells became more numerous, and a polymorphous inflammatory infiltrate eventually reappeared, allowing for the histologic diagnosis of mycosis fungoides to be made. In the patients whose nonspecific lesions have not evolved, the histologic features have remained unchanged. These studies, reported elsewhere, 3 revealed that 60% to 80% of the "mycosis fungoides" cells in the epidermis and 10% to 25% of the "mycosis fungoides" cells in the dermis in nonspecific lesions were Langerhans and indeter-
minate cells and not T lymphocytes. In successfully treated lesions the density of Langerhans cells and indeterminate cells in the epidermis was still markedly decreased 1 month after the completion of electron beam therapy. Results of the following tests were normal in all fifty Gro.up 1 patients before therapy: chest xray, liver-spleen radionuclide scan, and abdominal computed tomography scan searching for enlarged abdominal nodes. Forty-eight of fifty patients reacted to at least one of the recall antigens. In one patient the results of the Candida skin tests became positive after completion of electron beam therapy, but the other patient remained anergic. Results of routine hematologic tests, urinalyses, liver function tests, and routine chemistry studies were normal in all patients. The white blood cell count ranged from 4,500 to 10,000. Anemia (hematocrit
V o l u m e 16
Number 1, Part 1 January 1987
Electron beam irradiation and chemotheraPY for MF
55
1.9-
[] r
.7-
D,,,,,,I
._1 I--
lJ n
I--
.2-
i,I
100
,d ld
MONTH~ K~m
,; ~
COI~IB~
0 ~
HIGH
Fig. 6. Actuarial curves of the estimated probabilities of clinical remissions for combined Stages III and IV in the combination therapy (COMB RX) and high-dose electron beam (EB) therapy groups. Nonspecific lesions were used as end points. The numbers of patients at risk at specific break points were as follows: with combination therapy, 10 patients at 0 month, 6 at 4 months, 3 at 10 months, and 1 at 31 months; and with high-dose electron beam radiation, 10 patients at 0 month, 5 at 4 months, and 3 at 12 months.
<38%) and thrombocytopenia were not found in any patient prior to therapy. The values for serum immunoglobulins, except for IgE, were normal, The only abnormalities found were those related to the serum levels of IgE, the percentage of S6zary cells per 100 lymphocytes, the karyotypic analyses of lymphocytes in the peripheral blood and lymph nodes, and the skin and lymph node biopsies. Following therapy, only the levels of serum immunoglobulins and circulating S6zary cells and the peripheral karyotypes were routinely measured.
IgE levels Before treatment, only twelve of fifty patients had abnormally elevated serum levels of IgE (>200 IU). The values ranged from 220 to 1400 IU. These same patients continued to have elevations after therapy. Fluctuations of IgE levels occurred over a twofold to threefold range post
therapy in three patients without any obvious clinical correlation. In two patients with normal levels of [gE before therapy, the levels rose from 44 to 430 IU immediately afterward and from 18 to 384 IU 21/2 years after the completion of electron beam therapy, respectively. Circulating Stzary ceils In forty-five normal control subjects, S6zarylike ceils of the small cell type constituted 0% to 5% of circulating lymphocytes both in routine smears and in buffy coats (mean • SD: 0.69 • 1.5). In ten psoriasis patients and in the five patients with diffuse eczema the percentage varied from 2% to 8% (mean • SD: 3.2 +_ 2.3). We used 10% as the dividing line between normal and abnormal levels of circulating Stzary cells in normal subjects and mycosis fungoides patients. We compared the percentage of circulating
56
Journal o f the American Academy of Dermatology
Braverman et al
1
.'9
~1
.8-
1-11
I-,.( t~y
.6-
.5-.I t.O
.4.
. [] tat
.3-
I1.
"~.i_t
I---'E_ . . . . .
.
I-- . ~ I.d
9T A HO
.l
I
1
I
I
I
I
|
|
I
I
10
la0
:30
40
~0
60
70
80
90
100
MDNTHS Kevz
--~Ct~'lB
RX ST i & ::'
D O HIGH EB ST t I, 2
Fig. 7. Actuarial curves of the estimated probabilities of clinical remissions for combined Stages I and II following combination therapy and high-dose electron beam therapy. The actuarial curves from three other studies (HO, reference 5; TA, reference 6; HA, reference 7), in which high-dose electron beam therapy alone was used, are superimposed. Specific lesions were used as end points. Sbzary cells found in 1-1~m sections of buffy coats to that found in normal smears of twenty-three mycosis fungoides patients. One hundred fifteen pairs were available for comparison. The correlation coefficient of this comparison was r = 0.815; p < 0.001. (There were only twelve instances, all post therapy, in which a marked discrepancy occurred, making it impossible to determine whether the percentage was greater or less than 10%). In general, it was not difficult to identify and quantitate the number of Sbzary cells per 100 lymphocytes in routine smears. However, the routine differential smear examination of 100 white blood cells was not helpful in indicating whether there was an elevated level of circulating Sbzary ceils. One needs to count 100 lymphocytes. Virtually all the SCzary cells were of the small cell type--identical in size to small lymphocytes. Before therapy, nineteen of forty-seven patients had 10% circulating Sbzary cells per 100 lympho-
cytes (range, 10% to 30%). After electron beam radiation alone and electron beam radiation with chemotherapy, elevated levels o f Sbzary cells were intermittently present in nineteen of twenty-six patients in the range of 12% to 30%. We could not relate these elevations to the eventual disease courses in these patients.
Chromosomal studies The results of chromosomal studies are reported in detail elsewhere, z Clonal abnormalities were found once in only one patient. These abnormalities were not observed again in three subsequent tests during the succeeding 24 months. In no patient were the same chromosomal abnormalities found in both the lymph node and the peripheral blood specimens. In no instance have any of the original karyotypic abnormalities present before therapy been found after therapy. Karyotypic analyses have remained abnormal in all patients with
Volume 16 Number 1, Part 1 January I987
Electron beam irradiation and chemotherapy for MF
more than one follow-up test after treatment. Chromosomal abnormalities of circulating lymphocytes and intermittently elevated levels of S6zary cells continue to be present post therapy in most patients, without any apparent change in their states of clinical remission. Lymph nodes Of nineteen patients with lymph node biopsies, only six had clinically significant palpable lymphadenopathy. Sixteen patients had identical histologic findings on biopsy: dermatopathic changes with scattered S6zary-like cells in the paracortical area. The nodal architecture was preserved, with moderately active, normally placed follicular centers present. A marked expansion of the nodal T cell zones was accompanied by a moderate proliferation of small postcapillary venules, prominent hyperplasia of Langerhans macrophages, and a cellular population relatively rich in small lymphoid cells with hyperconvoluted nuclei. Sinus histiocytosis with melanin-containing macrophages was present. Three of the six patients with palpable nodes had involvement by mycosis fungoides in the excised nodes. These patients had four or more areas of nodal involvement and were in Stage III or IV. The only difference between the excised nonpalpable nodes and the excised palpable nodes in the other sixteen instances was the greater size of the latter. DISCUSSION Our studies confirm those from Stanford University 4 that electron beam radiation in doses of 3,000 rads can increase the probability of a patient's remaining in comlete clinical remission in late-stage mycosis fungoides in comparison with electron beam doses <1,600 fads. Cotter et als also found that a dose of at least 3,000 rads is needed for adequate local control of mycosis fungoides lesions. Both our combination electron beam radiation and chemotherapy and our highdose electron beam radiotherapy have produced an unusual syndrome in approximately 35% of patients who have had a complete response to the electron beam radiation: 1-2 cm patches, usually one to six in number and resembling xerosis or parapsoriasis, have appeared 4 to 43 months after
57
the completion of combination therapy or highdose electron beam therapy. Either they have evolved over 6 to 24 months into frank mycosis fungoides, or they have waxed and waned, spontaneously disappeared, or remained unchanged both in appearance and histologic features for up to 63 months. We have referred to the histologic changes as "tolerance" histologic changes or the "recapitulative" phase because "mycosis fungoides" cells are present but the normal accompanying inflammatory infiltrate is absent. One interpretation is that the electron beam has altered the microenvironment of the skin in some way so that the abnormal T cells ("mycosis fungoides" cells) do not evoke an inflammatory infiltrate (i. e., a tolerance to the abnormal cells has been created). Another interpretation is that the pathogenetic process underlying mycosis fungoides has been interrupted and the disease recapitulates itself through the manifestations of the nonspecific lesions. The observation that of eight nonspecific lesions arising after high-dose electron beam therapy alone, seven evolved into specific mycosis fungoides lesions, but that only nine of eighteen nonspecific lesions arising after combined therapy have done so, raises the possibility that chemotherapy may promote the "tolerance" or "recapitulative" phase. Regardless of whether we used specific mycosis fungoides lesions to indicate relapse after a complete response or nonspecific lesions that we know evolved into mycosis fungoides lesions, the results have been the same. Combination therapy has produced a statistically significant increased probability of a patient's remaining in complete clinical remission in Stage I or II mycosis fungoides, in comparison with the use of high-dose electron beam radiation alone. However, in late-stage disease (Stages III and IV), neither treatment modality has yet been shown to be superior to the other. These results illustrate the increased efficacy of adjuvant chemotherapy. Presumably the use of chemotherapy was less effective in advanced stages because these patients were left with a greater tumor burden, albeit clinically silent, following electron beam therapy than those with earlier stages of the disease. The group of twenty-four patients treated by
58 Braverman et al
high-dose electron beam radiation alone represents a historical control group except for the five patients who were added after 1979. Although we have accurate data in the form of photographs of the entire skin surface of these patients before and after therapy and detailed clinic notes on the course of their disease, a/lowing us to stage their disease by mycosis fungoides cooperative group criteria, there is always the concern that a retrospective historical group can never be used with confidence as a control group. However, as we have shown in Fig. 7, three other groups of investigators57 have published life table curves of clinical remissions following high-dose electron beam therapy for early mycosis fungoides that are superimposable on our own curve. This makes us confident that Group 2 is a valid control group. The actuarial curves generated by these three independent studies and ours may be a useful baseline against which other modalities can be judged. Since the nonspecific lesions in nine of eighteen patients have not yet evolved into mycosis fungoides lesions after 39 to 63 months of observation, we feel justified in considering the individuals with only one to six nonspecific lesions as being in clinical remission. Our data indicate that the combination of electron beam radiation and chemotherapy has produced a significantly increased frequency of sustained, complete clinical remission in early mycosis fungoides (Stages I and II) in comparison with highdose electron beam radiation alone. However, this combination therapy has not been as beneficial in late-stage disease (Stages III and IV). Griem et al 9 treated fourteen patients with mycosis fungoides with a combination chemotherapy and electron beam radiation protocol and produced an actuarial curve of clinical remissions that is virtually identical to our own. We have used the original Mycosis Fungoides Cooperative Group criteria for staging because we believe that it is more flexible in accommodating the various clinical features of the disease. In our experience the number of clinically involved nodal areas has correlated with the presence of lymph nodes involved by mycosis fungoides. We have taken biopsy specimens from nineteen nodes for staging purposes. Thirteen nodes were nonpalpable and six were palpable. All thirteen nonpalpable nodes from sites draining mycosis fungoides le-
Jourllal of the American Academy of Dermatology
sions showed dermatopathic lymphadenopathy with scattered S6zary-like cells in the paracortical area. Three of the six palpable nodes showed involvement by mycosis fungoides, and the remaining three showed identical dermatopathic changes. The only histologic difference between the excised nonpalpable nodes and the excised palpable nodes with dermatopathic changes was the greater size of the latter. In addition, the mycosis fungoidespositive nodes all occurred in Stage III and Stage IV patients. We have discontinued blind biopsies of nonpalpable nodes in the staging procedure. Therefore staging classifications that use nonpalpable nodes and the presence of dermatopathic changes as independent criteria for staging may not be entirely valid. In the future, the use of specific gene rearrangements to identify nodes involved with mycosis fungoides cells will become a more sensitive staging device. ~0 Seventy-three percent of the patients in Group 1 with more than two follow-up examinations after the completion of the electron beam radiationchemotherapy protocol had intermittent elevations of S6zary cells (small-cell variant) (11% to 30% of 100 lymphocytes) regardless of whether they were in clinical remission or not. Seventeen patients who had one to four serial chromosome analyses of their peripheral blood after completion of combination therapy continued to have abnormal karyotypes regardless of whether their skin was clear or whether they had nonspecific or specific lesions. Our studies also confirm those of Bunn et al, H who found abnormal karyotypes in cultures of peripheral blood lymphocytes even though the number of circulating S6zary-like cells was within the normal range and vice versa. The persistence of cytogenetic abnormalities and elevated levels of circulating S6zary cells post therapy, when the skin has clinically returned to normal, implies that the underlying pathogenetic mechanism is still operating. Pretreatment elevated IgE levels, which have been proposed as a sign of chronic antigenic stimulation in mycosis fungoides, lz remained unchanged after therapy in our series. In addition, the normal-appearing posttherapy lesional skin still showed occasional mycosis fungoides cells in the epidermis and dermis albeit without an accompanying inflammatory cell infiltrate. The nonspecific lesions showed similar
Volume 16 Number 1, Part 1 January 1987
Electron beam irradiation and chemotherapy for MF
histologic features that may or may not have evolved into classic mycosis fungoides over the course of many months. Thus we have the anomolous situation wherein visual inspection and routine laboratory examinations indicate that the patient is in a clinical remission or a disease-free interval. However, we have not produced a disease-free interval in any of our patients, even though by conventional clinical and laboratory criteria we would be permitted to place them in this category. Operationally, it would still be proper t~ call this state a clinical remission because the patient is free of pruritus and diagnostic skin lesions. Even when the patient has developed the clinically nonspecific eruptions, he or she usually is still asymptomatic. The nonspecific lesions have probably been overlooked in other studies because their numbers are usually small and have a nondescript clinical appearance. On the basis of our observations in eighteen cases (from this series and those of other patients treated between 1974 and 1979) the histologically verifiable nonspecific eruption has to be considered as a potential early sign of relapse in mycosis fungoides and may have to be used as an indicator in evaluating therapeutic responses. However, this nonspecific eruption does not invariably evolve into mycosis fungoides. It may disappear spontaneously or following the use of topical steroids. Our preliminary findings in two cases indicate that topical nitrogen mustard will produce a complete clinical resolution of such lesions and eliminate almost completely the "mycosis fungoides" cells from these sites. The prognostic significance of the nonspecific lesions has not yet been determined. The normal-appearing skin of mycosis fungoides patients, even before therapy, has single "mycosis fungoides" cells within the epidermis and dermis, indicating that mycosis fungoides is generalized in the skin at the earliest times of diagnosis. The histologic features of the nonspecific eruption, which appears as subtle macular lesions that may evolve into clinically recognizable mycosis fungoides lesions after the development of a normal inflammatory cell infiltrate, emphasize the point that nontumorous mycosis fungoides skin lesions are primarily caused by the accompanying normal cellular infiltrate and not by the "mycosis fungoides" cells themselves, which constitute
$9
only a minority of the cells in mycosis fungoides lesions. This point has also been emphasized by Sanchez and Ackerman. ~3 Our studies indicate that mycosis fungoides is a generalized disease occurring in the skin and extracutaneously at the earliest times of biopsy diagnosis. It has been argued that therapy for mycosis fungoides is generally unsuccessful because only the skin is treated. The lymph nodes and viscera, which are therapeutically neglected, then serve as the sites from which reseeding of the skin occurs. However, on the basis of our observations of "mycosis fungoides" cells in the normal-appearing, uninvolved mycosis fungoides skin, in the sites of successfully treated lesions, and in the "tolerance" or "recapitulative" histologic picture of nonspecific lesions, along with the persistence of cytogenetic abnormalities and the elevated levels of circulating S6zary cells during apparent clinical remissions or disease-free intervals, we suggest an alternate possibility. Regardless of how successfully we may lower the tumor load in the skin and nodes by electron beam radiation and chemotherapy, we are unable to stop the continuing underlying pathogenetic process that is evident during the "clinical remission." It is this unchecked process that may be responsible for eventual cutaneous recurrences and not simply metastatic seeding of the skin from sanctuary sites, such as viscera and nodes, that are unaffected by our primary combination therapy. We should be trying to control the underlying pathogenetic mechanism in addition to killing all the tumor cells. The "tolerance" or "recapitulative" histologic picture may be the morphologic expression of the partial inhibition of this pathogenetic mechanism. This concept is dealt with in more detail in another article.3 During the course of these studies, we observed three patients with early mycosis fungoides who developed acute biopsy-proved mycosis fungoides lesions, eczematous and papular, that disappeared spontaneously over the course of 2 to 4 weeks. This behavior is consistent with the concept of underlying host resistance. The same behavior was seen in the waxing and waning of some of the nonspecific lesions post therapy. Although it is argued that the presence of circulating "mycosis fungoides" or S6zary cells indicates extracutaneous involvement, there is no
60
Journal of the American Academy of Dermatology
B r a v e r m a n et al
evidence that such circulating cells, by t h e m s e l v e s , either p r o m o t e a high risk f o r mycosis fungoides tumor formation in v i s c e r a l organs or reduce survival time because o f visceral i n v o l v e m e n t with organ dysfunction. M y c o s i s fungoides t u m o r formation in the viscera is an u n c o m m o n finding both in the literature and in our personal experience, t4 Our current view is that prolonged survival and perhaps eventual cure will be possible either b y eradicating all the " m y c o s i s f u n g o i d e s " ceils with more effective treatment regimens or b y reducing the n u m b e r of " m y c o s i s f u n g o i d e s " cells to a m i n i m u m , so that whatever natural host resistance is present might have an opportunity to be effective. T h e three cures w e described at the beginning o f this report were associated with a g g r e s s i v e electron b e a m radiation and chemotherapy. REFERENCES
1. Lamberg SI, Green SB, Byar DP, et at: Status report of 376 mycosis fungoides patients at 4 years: Mycosis Fungoides Cooperative Group. Cancer Treat Rep 63:701707, 1979. 2. Braverman IM, Maynard T: Cytogenetic studies in mycosis fungoides. (Submitted for publication.) 3. Braverman IM, Klein S, Grant A: Electron microscopic and immunolabeling studies of the lesional and normal skin of patients with mycosis fungoides treated by total body electron beam irradiation. J AM ACAD DERMATOL 1987; 16:61-74. 4. Hoppe RT, Fuks Z, Bagshaw M_A: The rationale for curative radiotherapy in mycosis fungoides. Int J Radiat Oneol Biol Phys 2:843-851, 1977.
5. Hoppe RT, Cox TS, Fuks Z, et al: Electron-beam therapy for mycosis fangoides: The Stanford University experience. Cancer Treat Rep 63:691-700, 1979. 6. Tadros AAM, Tepperman BS, Hryniuk WM, et al: Total skin electron irradiation for mycosis fungoides: Failure analysis and prognostic factors, hat J Radiat Oncol Biol Phys 9:1279-1287, 1983. 7. Hamminga L, Hermans J, Noordijk EM, et al: Cutaneous T-cell lymphoma: Clinicopathological relationships, therapy and survival in ninety-two patients. Br J Dermatol 107:145-156, 1982. 8. Cotter GW, Baglan RJ, Wasserman TH, Mill W: Palliafive radiation treatment of cutaneous mycosis fungoides: A dose response. Int J Radiat Oncol Biol Phys 9:14771480, 1983. 9. Griem ML, Tokars RP, Petras V, et al: Combined therapy for patients with mycosis fungoides. Cancer Treat Rep 63:655-657, 1979. 10. Weiss LM, Hu E, Wood GS, et al: Clonal rearrangements of T-cell receptor genes in mycosis fungoides and dermatopathic lymphadenopathy. N Engl J Med 313:539544, t985. 11. Bunn PA Jr, Huberman MS, Whang-Peng J, et al: Prospective staging evaluation of patients with cutaneous Tcell lymphomas: Demonstration of a high frequency of extracutaneous dissemination. Ann Intern Med 93:223230, 1980. 12. Tan RA, Butterworth CM, McLaughlin H, et al: Mycosis fungoides: A disease of antigen persistence. Br J Dermatol 91:607-616, 1974. 13. Sanchez JL, Ackerman AB: The patch stage of mycosis fungoides: Criteria or histologic diagnosis. Am J Dermatopathol 1:5-26, 1979. 14. Rappaport H, Thomas LB: Mycosis fungoides: The pathology of extracutaneous involvement. Cancer 34:11981229, 1974.
ABSTRACTS Cutaneous lesions of sinus histiocytosis with massive iymphadenopathy: (Destombes-Rosai-Dor fman) associated with mycetoma. Description of a case
Rongioletti F, Heid E, Grosshans EG. G Ital Dermatol Venereol 1985;120:419-23 (Italian) A 22-year-old Tunisian woman when first seen had cutaneous lesions in which histopathologicexaminationdemonstratedthe typical
microscopic features of sinus histiocytosis with massive lymphadenopathy, tt is important to recognize that this condition, despite its resemblance to malignant histioeytosis, has a benign course, and to avoid treating these patients with dangerous drugs. Ten percent of all patients have akin involvement. Fehudi M. Felrnan, M.D.
Zinsser-Cole-Engman syndrome: Two cases transmitted by an autosomal dominant gene
Gasparini G, Sambuani N, Guidarelli C, Sarchi GD, Petro A, Raimondi E, Romagoni MM, et al. G Ital Dermatol Venereol 1985;120:429-33 (Italian) Two patients, a father and son, with Zinsser-Cole-Engman syndromes are reported. When first seen both had the typical symptoms triad of poikilomelanodermia, leukoplakia, and nail dystrophy. No other significant abnormalities were present. Genetic transmission prognosis and treatment of congenital dyskeratosis are discussed. Yehudi M. Felman, M.D.