THE INSTITUTE OF ONCOLOGY, LJUBLJANA* THE INSTITUTE OF HISTOLOGY AND EMBRIOLOGY, MEDICAL FACULTY,
LJUBLJANA**
CHEMOTHERAPY FOR HURTHLE CELL CARCINOMA BASED ON SEQUENTIAL
DNA MEASUREMENTS***
Auersperg M.*, Zorc R.**, Us-Krašovec M.**, Pogačnik A.*, Petrič G.*, Porenta-Vraspir O.**
Abstract — Inoperable or disseminated Hurthle celi carcinoma is a therapeutic challenge as therapy with I131, radiation or chemotherapy is usually ineffective. In order to find an effectivechemotherapy the influence of Vinblastine (VLB) (2 mg bolus or infusion over 6, 12, 24h) was studied in 5 patients (4 women, 1 man; age 43—69 years) and that of Cisplatinum (CDP) in 3 patients. CDP 20 mg was administred in two and CDP 50 mg/m2 in one patient. Four patients had distant metastases, one locoregional disease only. Thin-needle aspiration biopsies of tumors (1 primary, 4 metastases) were performed before and repeatedly after VLB or CDP application. The smears were stained afterFeulgen and were used for cytophotometric DNA measurements. VLB and CDP produced an increase of cells in S phase compartment. On the basis of changes produced in the DNA distribution pattern by the test dose of VLB or CDP chemotherapy was planned: either a sequence of 3 VLB infusions with individual intervals or a combination of VLB, Cisplatinum, Methotrexate, Bleomycin, 5-Fu and Adriamycin was used. All 5 patients responded — 1 CR, 3 PR, 1 MR. Chemotherapy was combined with surgery in one, and with radiation in 3 patients. Two out of 5 patients have NED 43 +, 29 + months after treatment, one patient is living with disease 41 months after chemotherapy, one patient died of tumor 5 months after chemotherapy and one patient 6 weeks after start of chemotherapy of pulmonary embolism with more than 90 necrosis in the tumor.
UDC: 616.441:615.277.3
Key words: thyroid neoplasms-drug therapy, carcinoma Hurthle celi, DNA-analysis Orig. sci. paper
Radiol. lugosl. 22 (3) 269—275, 1988
Introduction — Hurthle cell carcinoma is a thyroid tumor of moderate aggressivenes (9). It is relatively uncommon (6, 8). An incidence from about 3%—1 O% of all thyroid carcinomas was reported (6, 9). It is considered to be a variant of follicular carcinoma by some authors (4, 11). The
prognosis of Hurthle cell carcinoma, however, is much worse than that of follicular carcinoma (4). Surgery is the treatment of choice (8, 9, 12, 14). When inoperable or disseminated, Hurthle cell carcinoma is a difficult therapeutic problem as it usually does not concentrate 1131 (6, 8, 9, 10, 12)
Author		No. of cases	Chemotherapy	Response
Thompson	1974 (14)	2	Endoxan	not cited
Gottlieb	1974 ( 7)	9	Adria	2/5 PR
Droz	1981 ( 5)	3	Adria VP-16, 5-Fu Endoxan	1/3 MR
Gundry	1983 ( 8)	9	Adria	0
			Various	
Miller	1983 (12)	1	CDP	0
			Adria	
Har-EI	1986 ( 9)	4	not cited	3/4 0 1 /4 survived 5 yrs (CT + S + RTX)
Auersperg	1987	5	VLB sequence	
(present study)			VLB, MTX, 5-Fu, Bleo Adra, CDP Individualized CT	1 CR, 3 PR, 1 MR
CT = chemotherapy S = surgery Rtx = radiotherapy
Table 1 — Results of chemotherapy in Hurthle cell carcinoma
*** Presented at IXth Meeting of the European Association tor Cancer Research, Helsinki, 3—6 June 1987
Received: December 11, 1987 — Accepted: December 15, 1987
269
Auersperg M. et al.: Chemotherapy for Hurthle celi carcinoma based on sequential DNA measurements
and is relatively radioresistant (6, 8). Har-EI reported that paliation but never cure was achieved by radiation (9). Only few patients with HUrthle cell carcinoma were treated by chemotherapy (CT) (Table 1), therefore experience with this modality of treatment is scarce and most authors (5, 7, 8, 9, 12) consider the effect of chemotherapy in HUrthle celi carcinoma as disappointing. There are few reports on chemotherapy in thyroid malignancies — the schedule combining Adriamycin (Adria) and Cis-platinum (CDP) described by Shimaoka (13) is rather aggressive and therefore not easily applicalbe in patients with inoperable or disseminated HUrthle cell carcinoma. Those patients are usually in the older age group and have usually contraindications for agressive chemotherapy because of cardiovascular and other diseases.
In our previous studies sequential DNA measurements in aspiration biopsies of tumors together with morphological studies of changes in tumor cells, induced by chemotherapy proved to be useful in development of rational nonaggressive chemotherapy for a variety of rare or che-moresistant tumors (1, 2, 3). The aim of this study was to find an effective and nonaggressive chemotherapy for inoperable or disseminated HUrthle celi tumors which would be applicable in
poor risk patients, or in patients after failure ot standard chemotherapy for thyroid cancer, Adria and CDP (13).
Material and methods — From 1984-1987 5 patients (1 male, 4 females; age 43--69 years) with inoperable or disseminated Hurthle cell carcinomas in whom standard CT for thyroid cancer (Adria + CDP) was not possibie, were treated with individualized chemotherapeutic schedules. Two patients were previously treated with CT; one with Adria and one with Adria + CDP. In 3 patients Adria and CDP in standard doses could not be given because of impaired renal function in 1, severe respiratory distress in 1 and cardiac arrhythmia in one patient.
The effect of low doses of VLB on tumor cell population 2 mg in bolus or i.v. infusion over 6, 12, 24h was tested in 4 patients and that of CDP 20 mg i.v. over 2 hours in 2 patients, and CDP 50 mg /m2 i.v. infusion over 12 hours in 1 patient. Thin needle aspiration biopsy of the primary tumor (1 patient) or metastases (4 patients) were performed before and at uneven intervals after VLB or CDP infusion. Half of smears were stained acc. May-Grunwald-Giemsa for morphological studies aid the other half by a Feulgen
Patient No.	Tumor site	Chemotherapy	Other therapy	Follow-up
M. Z. (1) female 69 yrs	soft tissue metastasis right hip 16 X 12cm	VLB CDP — PR VLB CDP Bleo 5-Fu MTX 2 courses — CR	radiation 5000 cGy	NED 43 + months
K. H. (2) female 56 yrs	locoregional tumor inoperable	VLB CDP VLB — PR CDP MTX Bleo — PR	0	died 6 weeks after CT, pulmonary embolism, —90% necrosis of the tumor
J. B. (3) female 60 yrs	metastasis clavicle 1:5 X 11cm	VLB CDP — PR VLB CDP VLB Bleo Adria Endoxan — PR	surgery	NED 29 + months
K. A. (4) male 60 yrs	Locoregional tu pulmonary metastases mediastinal lymph nodes	VLB VLB VLB — PR VLB Bleo MTX 5-Fu — PR At 21 months reinduction of CT VLB VLB VLB — PR	radisiicn + VLB 3000 cGy radiation 3000 cGy	alive with disease 41 months
G. M. (5) female 43 yrs	pulmonary metastases, mediastinal tu pleural, pericardial effusion, respiratory distress	VLB VLB VLB VLB Adria VLB CDP — MR 4.5 months	radiation 1000 cGy	dead 5 months
CT = chemotherapy
Table 2 — Patients characteristics
270
Radiol lugosl. 22 (3) 269-275, 1988
Auersperg M. et al.: Chemotherapy for Hijrthle cell carcinoma based on sequential DNA measurements
Fig. 1 — Patient No. 3, Hurthle-cell carcinoma, metastasis in the right clavicle. Aspiration biopsy sample, Giemsa, 100x. Large, round and polyhedral cells: ample, homogenous, slightly eosinophylic stained cytoplasm, round, regular nuclei, fine chromatine structure, occasional nucleoli.
Fig. 2 — The same patient as in Fig. 1 5-4 hours after 2mg VLB in 12-hour intravenous infusion, aspiration biopsy sample, Giemsa, 100x. Few cells with slightly enlarged nuclei and vacuoles within the cytoplasm (1).
Patient No.
Chemotherapy
Response
test VLB 2mg 24h interval 96h CDP 70mg 12h (1) CT VLB 2 mg 24h interval 28h CDP 70mg 12h interval 33h MTX 50mg 18h Bleo 15mg 8h 5-Fu 500mg 18h 2 courses
PR CR
(2)	test CT	VLB 2 mg bolus VLB 2 mg 12h CDP 20 mg 2h interval 30h VLB 2mg 12h Bleo 15mg 1 2h MTX 50mg 16h VLB 2mg 12h Bleo 15mg 12h	PR PR PR/90% tu necrosis at autopsy
(3)	test CT	VLB 2 mg 12h CDP 20mg 2h VLB 2 mg 12h interval 120h CDP 70mg 7h interval 1Oh Bleo 15mg 1Oh Adria 40 mg 4h Endoxan 400 mg 4h VLB 2mg 12h Bleo 15mg 12h	R R PP
(4)	test CT	VLB 2 mg 12h VLB 2 mg 6h VLB 2 mg bolus VLB 2 mg 6h Bleo 15 mg 6h MTX 50 mg 12h VLB 2mg 12h 1)	5-Fu 750mg 12h Bleo 1 5mg 12h 2)	VLB 2 mg 12h interval 3(3h VLB 2mg 12h interval 36h VLB 2mg 12h (test and CT 2 better than CT 1)	PR PR
(5)	CT	1)	VLB 2 mg 1 2h interval 24h VLB 2mg 12h interval 118h Adria 70 mg 7h 2)	CT 2 = 1 without Adria 3)	CT 3 = 1	
CT = Chemotherapy
Table 3 — Chemotherapy
Radiol. lugosl. 22 (3) 269-275, 1988	271
Auersperg M. et al.: Chemotherapy (or Hurthle cell carcinoma based on sequential DNA measurements
Fig. 3 — To same patient 72 hours after 2 mg VLB in 12-hour infusion. Aspiration biopsy sample, Giemsa, 100 x. Small picnotic (1) and hypochromatic, pale stained nuclei (2) and nucleoli (3).
staining procedure including acid hydrolysis in 4 NHCI at 28 °C for 60 min. DNA measurements were carried out on a Vickers 85 scanning microdensitometer at wavelenght 560 nm and processed by a computer. In each smear 150-250 tumor cells and 25-100 leukocytes were measured. The DNA value for leukocytes served as a reference value.
Data obtained by DNA measurement were used in planning of time sequence of cytostatics. It was attemped to infuse drugs predominantly effective in S-phase (MTX, VLB, Adria, Endoxan, 5-Fu) and G2-phase (Bleo, VLB, Adria) during the time after VLB or CDP when DNA histograms showed a maximum in relative increase in S-and G2-compartments.
A block of cells at the beginning of S-phase which was found after VLB was a basis for a combination of VLB and irradiation which was used in 2 patients after CT (patients No. 4 and 5, Table 2).
CT was individualized according to the results of DNA measurements and cytomorphological studies. Data on drugs, dosage and time sequence of cytostatics are in Table 3.
Results - Results of DNA studies — DNA
distribution pattern afterVLB was individual and seems to depend on the tumor sensitivity and time of exposure to the drug. In patients 1 and 4 (table 3) the number of cells with high DNA values was reduced after VLB. In patient No. 3 (Table 3) the cells in S-phase accumulated from 0—48\ at later hours, the cells in were probably blocked at the beginning of S-phase (Fig. 5, A and B). In patient No. 2 a difference in the DNA distribution pattern depending on time of the exposure to the drug was demonstrated (Fig. 6). In
Fig. 4 — Patient No. 1, soft tissue metastasis of an anaplastic clear cell carcinoma (12 X 16 cm) before CT (outer circle), after test dose of VLB 2 mg/24h and CDP 70 mg/12h (inner circle). A CR was achieved after the first course of CT with VLB 2 mg, CDP 70 mg, MTX 50 mg, Bleo 15 mg and 5-Fu 500 mg.
patient No. 1 (Table 3) there was an impressive tumor reduction already after the test dose of VLB. The DNA histograms showed, however, only minor changes — endoreduplication — and later on elimination of cells with high DNA values.
After CDP 20 mg in 2h there was an accumulation of cells in S- and G2 + M compartment (Fig. 5 B) in patient No. 1 (Table 3) again — after CDP 70 mg in 12 hours — the cells with high DNA values were eliminated. The tumor in this patient regressed rapidly — 20 hours after the first chemotherapy no more malignant cells were obtained by aspiration biopsy.
Results of morphological study — Morphological changes of tumor cells after chemotherapy were found in aspiration biopsies of tumors in patients No. 1—4. In patient No. 5 the cell sample was inadequate. Different degenerative changes were caused by chemotherapy. The intensity of changes and the time of their onset seems to depend on the tumor sensitivity and the time of exposure to the drug. In the patients No. 2 and 4 the changes of tumor cells were more pronounced and were detected earlier afterVLB 2 mg in 12h infusion in comparison with VLB 2 mg in bolus. In patient No. 3 the first morphological changes were observed 54 hoursafterVLB 2 mg in 12 hours administration. Among morphologically unchanged tumor cells, cells with slightly
272
Radiol. lugosl. 22 (3) 269-275, 1988
Auersperg M. et al.: Chemotherapy for HUrthle cell carcinoma based on sequential DNA measurements
enlarged nuclei with pycnotic or regular chromatin structure and vacuoles within the cytoplasm could be found. More diffuse, however discrete changes appeared 72 hours after VLB administration; hypochromatic, pale stained nuclei, wrinkled nuclear membrane, pale stained nucleoli, degenerated cells and naked nuclei.
On general, the morphological changes appeared rather late 54—72 hours after chemotherapy and were rather discrete. In patient No. 4 (Table 3) the morphological changes appeared only 5 months after chemotherapy.
Clinical results — Four out of 5 patients responded, one patient had complete response (CR), 3 patients partial response (PR), one patient had minimal response (Mr) and subjective improvement lasting 4, 5 months. This patient died 5 months after CT in respiratory distress. One patient died 6 weeks after CT of pulmonary embolism. At autopsy more than 90 % of the tumor was necrotic. Two patients are alive with no evidence of disease (NED) 29 and 43 months after therapy, one is living with disease 41 months after CT. CT was combined with surgery in one patient and with irradiation in 3 patients — the
duration of the effect of CT alone could be evaluated only in patient No. 4 (Table 3). The PR obtained after test dose ofVLB and 1 st combined CT lasted 20 months. After the 1 st CT this patient refused further treatment and had reinduction of chemotherapy only after tumor progression 21 months later.
Chemotherapy was well tolerated with no major side effects.
The patient No. 1 had malaise and loss of appetite during CT. Patients No. 2 and 3 had palpitations during infusion ofVLB. Patient No. 2 developed diarrhea on month after CT. The fatal pulmonary embolism two weeks later in this case could be connected with dehydration.
Discussion — DNA studies in aspiration biopsies of Hurthle cell tumors yielded useful information for planning of CT. On the basis of changes in the DNA distribution pattern after VLB an effective regimen consisting of sequence of VLB infusions with individualized intervals between infusions was designed. The second and third infusion ofVLB were applied during the time of accumulation of cells in S-phases. Another possibility was sequence of VLB and CDP
OKA arbitrary units
JO h VLB
A.
rH ' 1 1 1 . 1 1 i J	i i i i i i
' ^ i p i i i I r-1	i i i u
Pi	1 u
I1 1 ! J	1 1 1 1 1
120 h alter VLB Before COP
DKAarbitrarr units
20 h
B
Fig. 5 — Patient No. 3, well differantiated Hurthle-cell carcinoma — metastasis in the clavicle. DNA histograms of histograms of aspiration biopsies specimens before, O, 6, 24, 30, 48 and 72 hours after VLB 2 mg in 12-hour intravenous infusion (A). The DNA distribution pattern shows a progression delay through the S phase. The relative number of cells in S phase is increased from 0—48", at 72—120 hours (B) the cells are probably blocked at the beginning of S phase. At 30" there is a shift of DNA values toward left. After CDP 20 mg in 2" infusion (B) there is a block of cells in the S phase and G2 + M compartment.

50
1 h COP
50
51 h
Radiol. lugosl. 22 (3) 269-275. 1988
273
Auersperg M. et al.: Chemotherapy for Hurthle cell carcinoma based on sequential DNA measurements
H	i
Before treatment
16 h VLB bolus
36 h
Before treatment
18 h VLB

n
X
36 h
jr.
DNA arbitrary units
Fig. 6 — Patient No. 2, low differentiated HUrthle-celi carcinoma — inoperable primary tumor. DNA histograms of aspiration biopsies specimens before, 18 and 36 hours after VLB 2 mg in bouls and VLB 2 mg in 12-hour intravenous infusion. The DNA distribution patterns are different at 18 hours: After VLB in bolus there is an accumulation of cells at the beginning of S phase, whereas after VLB infusion the cells accumulate in the middle of S and G2 + M compartment.
infusion — CDP beeing applied during the block of tumor cells in the beginning of S-phase. When tested in the same patient, sequence of VLB or VLB and CDP infusions seemed to be more effective than combined chemotherapy with more drugs. The most impressive clinical response was achieved in patient No. 1 (Table 3). In this patient the first chemotherapy course resulted in a CR. The DNA histograms of the tumor in this patient, however, showed only minor changes. Apparently, the cells remained in a frozen state untill they were eliminated. Another interesting observation is the rather slow tumor regression and late onset of morphological changes after chemotherapy in patient No. 4, with long lasting effect of CT.
DNA measurements were an useful aid in planning CT. There is, however, not enough experience yet to assess the predictive value of changes in the DNA distribution pattern of tumors tor the outcome of CT.
Conclusions — Low doses of VLB or CDP are effective modulators of cellular kinetics in Hurthle cell tumors. Cytophotometric DNA measurements were useful in designing non agressive effective chemotherapeutic schedules. A
sequence of low doses ofVLB and CDP infusion was effective in Hurthle cell carcinoma. DNA distribution patterns of tumors after VLB infusion and bolus are different.
Povzetek
KEMOTERAPIJA PRI KARCINOMU HURTHLOVIH
CELIC. PLANIRANJE NA OSNOVI MERITEV DNK
Zdravljenje inoperabilnega ali diseminiranega kar-cinoma HUrthlovih celic je težko, ker je zdravljenje z J,131 obsevanjem ali kemoterapijo največkrat neuspešno. Z namenom, da bi izdelali uspešno in ne-agresivno kemoterapijo smo preučevali učinek nizkih doz Vinblastina (VLB) in Cisplatinuma (CDP) na karci-nom HUrthlovih celic pri 5 bolnikih. Zasledovali smo razporeditev vrednosti dezoksiribonukleinske kisline (DNK) v vzorcih aspiracijskih biopsij tumorjev (barvanje po Feulgenu, meritve na posameznih celicah) pred in po infuziji VLB in CDP ter zasledovali učinek VLB in CDP na morfologijo tumorskih celic. Po VLB in CDP smo ugotavljali relativno kopičenje celic v S fazah celičnega ciklusa, kar smo izkoristili za načrtovanje kemoterapije. Uporabili smo infuzijo 3 zaporednih odmerkov VLB z individualnimi presledki ali pa kombinacijo VLB, CDP, Methotrexata, Bleomycina, 5-Fluorou-racila in Adriamycina. Pri vseh 5 bolnikih smo dosegli zmanjšanje tumorja (1 popolen regres, 3 delni regresi, 1 minimalen regres. Kemoterapijo smo kombinirali z operacijo pri 1 bolniku in z obsevanjem pri 3 bolnikih. Dva bolnika sta brez bolezni 43 in 29 mesecev, 1 živi z boleznijo 41 mesecev. Dve bolnici sta umrli, ena 5 mesecev, druga 6 tednov po začetku kemoterapije. Ta bolnica je umrla zaradiplučne embolije, v tumorju ščitnice smo našli 90 % nekroze po kemoterapiji.
References
1.	Auersperg M., Zorc R., Us-Krasovec M., Porenta-Vraspir O.: DNA measurements used tor individual planning of chemotherapy in rare or chemoresistant tumors. International meeting on clinical and experimental cancer research, Trieste 1985. Abstracts, 6.
2.	Auersperg M., Zorc R., Us-Krasovec M., Porenta-Vraspir O., Petric J., Vodnik A., Novak J., Senear M.: DNA measurements used tor planning of multimodal treatment in sarcomas. 14th international cancer congress, Budapest 1986. Abstracts, Vol. 2: 637.
3.	Auersperg M., Zorc R., Us-Krasovec M., Vraspir-Porenta O.: dNa measurements and cytomorphology — a basis tor planning chemotherapy in anaplastic giant-cell carcinoma ofthe thyroid. Radiol. lugosl. 21: 47-53, 1987.
4.	Crile G., Jr., Pontius K. l., Hawk W. A.: Factors influencing the survival of patients with folicular carcinoma ofthe thyroid gland. Surg. Gynec. Obstet. 160 (5): 409-413, 1985.
5.	proz J. P., Charbord P., Rougier Ph., Parmentier CI.: Échec de la chimiothérapie des cancers de la thyroïde. Buli. Cancer 68 (4): 350-352, 1981.
6.	Goldenberg l. S.: Hürthle cell carcinoma. Arch. Surg. 67: 495-501, 1953.
7.	Gottlieb J. A., Hill C. S., Jr.: Chemotherapy of thyroid cancer with adriamycin. Experience with 30 patients. New Engl. J. Med. 290 (4): 193-197, 1974.
8.	Gundry S. R., Burney R. É.,Thompson N. W., Lloyd R.: Total thyroidenctomyfor Hürthle celi neplasm ofthe thyroid. Arch. Surg. 118 (5): 529-532, 1983.
infusion
274
Radiol. lugosl. 22 (3) 269-275, 1988
Auersperg M. et al.. Chemotherapy for Hurthle ceii carcinoma based on sequential DNA measurements
9.	Har-EI G., Hadar T., Segal K., Levy R., Sidi J.: Hurthle celi carcinoma of the thyroid gland. A tumor of moderate malignancy. Cancer 57 (8): 1613—1627, 1986.
10.	Krishnamurthy G. T., Blahd W. H.: Radioiodine 1-131 therapy in the management of thyroid cancer. A prospective study. Cancer 40 (1): 195—202, 1977.
11.	LiVolsi V. A.: Pathology of thyroid cancer. U: Thyroid cancer (Greenfield L. D., ed.), CRS Press, Palm Beach, Floride, 1978, (85—141).
12.	Miller R. H., Estrada R., Sneed W. F., Mace M. L.: Hurthle celi tumors of the thyroid gland. Laryngoscope 93 (7): 884-888, 1983.
13.	Shimaoka K., Schoenfeld D. A., DeWys W. D., Creech R. H., DeConti R.: A randomized trial of doxorubicin versus doxorubicin plus cisplatin in patients with advanced thyroid carcinoma. Cancer 56 (9): 2155-2160, 1985.
14.	Thompson N. W., Dunn E. L., Batsakis J. G., Nis-hiyama R. H.: Hurthle cell lesions of the thyroid gland. Surg. Gynec. Obstet. 139: 555-560, 1974.
Author's address: Prof. dr. Marija Auersperg, Onkološki inštitut, Zaloška 2, 61105 Ljubljana.
E^D TOSAMA
Proizvaja in nudi kvalitetne izdelke:
Komprese vseh vrst Gazo sterilno in nesterilno Elastične ovoje Virfix mrežo Micropore obliže Obliže vseh vrst Gypsona in mavčene ovoje Sanitetno vato PhJ III Zdravniške maske in kape Sanitetne torbice in omarice Avtomobilske apoteke
Radiol. lugosl. 22 (3) 269-275, 1988
275