Use of All-trans Retinoic Acid in the Treatment of Acute Promyelocytic Leukemia
Huang Meng-er 1, Ye Yu-chen 1, Chen Shu-rong 1, Chai Jin-ren 1, Lu Jia-xiang 2, Zhao Lin 3, Gu Long-jun 1, and Wang Zhen-yi 1    Hmatol. Bluttransf. Vol 32

1 Shanghai Institute of Hematology, Shanghai Second Medical University, Shanghai,People's Republic of China
2 Shanghai Zhong-Shan Hospital
3 Shanghai Children's Hospital, Shanghai, People's Republic of China

A. Introduction

Acute promyelocytic leukemia (APL) is considered to be a distinct entity among the acute myeloid leukemias (AMLs). Hemorrhagic diathesis often occurs and results in a rapid fatal outcome. The bleeding episodes are usually attributed to thrombocytopenia and/or disseminated intravascular coagulation (DIC), which is thought to result from the release of a procoagulant factor from the promyelocyte granules [1]. The use of daunorubicin in the induction therapy and improvement in the supportive therapy has greatly raised the rate of complete remission (CR) in APL [2, 3]. However, the increased mortality during induction therapy is higher in APL than in other forms of AML [3, 4] and there are still cases which continue to be refractory to induction chemotherapy. DIC remains a common lethal complication. Induction of differentiation may be an alternative approach to the treatment of APL. Retinoic acid (RA), an analog of vitamin A, is one of the many agents which can induce differentiation and terminal cell division of leukemic cells in vitro [5]. At the present time, several cases of APL treated with 13-cis RA have been reported with encouraging results [6-9]. In this paper, we report the in vitro studies and therapeutic trials of 24 APL patients using all-trans RA.

B. Materials and Methods

I. Patients

The diagnosis of APL was made according to the criteria of the French-American-British (FAB) cooperative study group [10]. Every patient presenting to our hospitals since early 1986 with a diagnosis of APL was included in this stud y .The clinical characteristics of the 24 patients with APL are shown in Table 1. There were 11 females and 13 males, with a mean age of 35.5 years (range = 5-69 years). The total white blood cell counts ranged from 0.5 x 109/liter to .8 x 109/liter, including 20 cases (83.3%) with less than 3x10 high9/liter, 3 cases (12.5%) with between 3 x 10 high9 and 10 x 10 high9/liter, and 1 case (4.1 %) with more than lOx 109/liter. The hemoglobin concentrations ranged from 41 g/liter to 121 g/liter including 8 cases (33.3%) wi th less than 60 g/li ter , 12 cases ( 50% ) between 60 and 90 g/liter and 4 cases (16.7%) with more than 90 g/liter. Platelet counts ranged from 10 x 109/liter to 337 x 10 high9/liter including 15 cases (62.5%) with less than 50 x 10 high9/liter, 7 cases (29.2%) between 50 x 109 and lOO x 10 high9/liter and 2 cases (8.3%) with more than lOO x 109/liter. The percentage of promyelocytes in the marrow ranged from 15.6% to 94%, with 22 patients having more than 30% and the remaining 2 patients having between 15.6% and 30%. Of these 24 studied patients, 16 had never been treated. The other eight (case # 1-# 8) had previously been treated with chemotherapy [HOAP 1(1 H, Harringtonin (0.02-0.07 mg/kg/day); 0, oncovin (0.02-0.03 mg/kg/day); A, cytosine arabinoside ara -C); P, prednisone; C, cyclophosphamide) , HOP, OH, COH, H]. Of the eight treated patients, three were in relapse after 1- 30 months of CR and five were resistant to or could not tolerate the chemotherapy (5-62 days of treatment). Twenty-two of the patients showed mild to moderate hemorrhagic manifestations (purpura, gingivorrhagia, gastrointestinal bleeding) but no laboratory evidence of DIC prior to treatment with RA except for a positive plasma protamin sulfate paracoagulation (3P+) in three of the cases.

II. Marrow Preparation and Culture

A modification of the method of Flynn et al. [6] for short-term suspension culture was used. Marrow cells were aspirated from the iliac crest, layered onto Ficoll-Hypaque (specific gravity 1.077), and centrifuged at 800 g for 15 min. Interface cells were collected, washed with McCoy's 5A medium, and resuspended at a concentration of 5 x 10 high5 cells/mI in McCoy's 5A medium containing 15% fetal calf serum. All-trans RA (Shanghai No.6 Pharmaceutical Factory, Shanghai) was dissolved in absolute ethanol to a concentration of 1 mM and further diluted with the medium so that the final ethanol concentration in the cultures was 0.1% and the final RA concentration 1M. Controls were cultured in medium alone. (It had been previously demonstrated that 0.1% ethanol had no effect on cell growth and on differentiation of HL-60 cells [11].) All cultures were incubated at 37 C in a 5% CO2 atmosphere for up to 7 days. Cell density was determined by hemacytometer and cell viability by the trypan blue dye exclusion method. Aliquots of cells were removed for morphological examination on the 2nd, 4th, and 6th day of culture.

III. Morphological Studies

Differential counts were performed on cell smears stained with Wright's solution. Chloroacetate esterase and alphanaphthyl acetate esterase stains were performed using standard techniques [12]. Samples from four cases were prepared for transmission electron microscopic study. The nitroblue tetrazolium (NET) reduction assay was performed as described by Francis et al. [13]. The percentage of cells containing intracellular blue-black deposits was determined in 200 cells on Wright's stained slide preparations.

IV. Colony Formation Assay

Blast cell colonies were grown as described by Minden et al. [14]. Conditioned medium was prepared from leukocytes (106 cells/mI) incubated at 37 C for 7 days in McCoy's 5A medium with 10% fetal calf serum and 1% (v/v) phytohemagglutinin-P (PHA-P) (DIFCO) and stored at 4 C until used. The preparation was termed PHA-LCM. Bone marrow cells were plated at 1 x 10 high6 cells/ml using McCoy's 5A medium supplemented with 0.3% agar, 20% fetal calf serum, and 25% (v/v) PHA-LCM. GMCFU was determined according to the technique of Pike and Robinson [15] for colony growth in agar. Briefly, 2 x 10 high5 marrow cells were plated in 35-mm tissue culture dishes over a feeder layer of 1 x 10 high6 leukocytes from healthy donors. The plates were incubated at 37 C in a humidified 5% CO2 atmosphere. L-CFU colonies (more than 20 cells) were scored on day 8 and GM-CFU colonies (more than 40 cells) on day 10.

V. Treatment of Patients

The 24 patients in this series received alltrans RA (45-100 mg/m/day) as the remission induction therapy. Informed consent was obtained from all patients ( or their parents ). Peripheral blood counts, bone marrow aspiration, and co agulation parameters (in 21 cases) including thrombin time, prothrombin time, plasma protamin sulfate paracoagulation test, euglobulin lysis test, and fibrinogen levels were determined before the start of therapy and at regular intervals thereafter. CR is defined as less than 5% blasts plus promyelocytes in a normal cellular marrow with a normal peripheral blood count and an absence of the signs and symptoms of leukemia on physical examination [16]. Partial remission (PR) is defined as less than 5% blasts plus promyelocytes in a normal cellular marrow but with a clinically moderate anemia. Blood transfusion and antibiotics were given as supportive treatment when necessary.


Table 2. Response of promyelocytes to RA in suspension culture

VI. Continuation Therapy Following Complete Remission

Twenty-three patients were followed after attaining CR. Further therapy was as follows: (1) Maintained by RA, 2030 mg/m/day (six cases), (2) maintained by RA, 20-30 mg/m/day plus low-dose ara-C (10 mg i.m. every 12 h) or low-dose Harringtonin (0.5 mg/m i.v. daily) in rotation (four cases), (3) maintained by low-dose ara-C, 10 mg i.m. every 12 h (five cases), (4) consolidated by chemotherapy (HOAP) and maintained by 6mercaptopurine (2 mg/kg daily p.o.) and methotrexate (10 mg/m, i.v. weekly), or cyclophosphamide (200 mg/m, i.v. weekly) (nine cases).

Fig. 1 A -D. Morphological maturation of leukemic cells of case 10 in vitro and in vivo. A Cells cultured without RA, consisting of promyelocytes with characteristic cytoplasmic granules, x 1000. B Cells cultured with RA, showing maturation of granulocytes, x 1000. C Bone marrow before RA treatment. The predominance of promyelocytes (76% ) indicates typical APL. D Bone marrow after 5 weeks of RA treatment. Promyelocyte level under 2%, and restoration of normal hemopoiesis without an aplasia phase are consistent with differentiation induction

C. Results

I. In Vitro Studies

Leukemic bone marrow cells derived from 15 patients and incubated for 7 days in suspension culture, with or without all-trans RA (1 M), showed little change in cell density. Viability of both control and RA-treated cells was consistently greater than 75%. Leukemic promyelocytes from 14 patients showed morphologic and functional maturation when cultured in the presence of RA (Table 2), (Fig. 1 A, B). The percentage of promyelocytes in the con trol group versus the RA-treated group was 83.5% +-12.8% and 5.9% +-5.0% respectively. The percentage of mature cells (metamyelocytes + bands + PMNs) was 4.7% +-4.5% and 53.9% +-15.4% respectively. The rate of NET reduction in RA-treated cells was 42.0% +-7.5% , significantly higher than that of the control group (4.2% +-3.5%) . 2 Results represent the data from the patients studied and are expressed as mean % +- standard deviation. To examine the progression of cellular differentiation, we incubated cells from four patients with 1 M RA for various time intervals. After 48 h, morphologically recognizable changes in the promyelocytes could be observed. The nu cleus became larger and fewer primary granules were observed in the cytoplasm. On the 4th day of culture, these cells gave rise to myelocytes which contained specific, or secondaIy, granules. The nuclear chromatin was more condensed and the nucleoli were either vague or no longer visible. There was an elevated population of metamyelocytes which had indented or horseshoe-shaped nuclei and cytoplasm filled with both primary and secondary granules appearing by day 6, as well as some band and fully mature granulocytes. When the cultures were continued for 7- 8 days, the relative number of mature granulocytes increased. Cytochemical analysis showed that in the control cells chloroacetate esterase activity varied from mildly to moderately positive, while in the RA-treated cultures intensely positive granules were seen, either diffusely scattered or accumulated in some portion of the cytoplasm. The majority of control cells showed weak nonspecific esterase activity while RA-treated cells had a stronger reaction. Transmission electron microscopic examination of four cultures confirmed that in the presence of RA the cells had been differentiated to mature granulocytes. Condensation of the heterochromatin became evident and the nucleus had often been changed to a bean-shaped or even a segmented form. Neutrophilic granules were smaller and diffusely scattered throughout the cytoplasm. Azurophilic granules were markedly decreased.

II. Clinical Studies

Twenty-four patients were treated with all-trans RA as a single agent. All achieved both PR and CR except the one patient ( # 24) whose cells were not inducible when cultured with RA in vitro. Subsequent bone marrow examination of this patient revealed a continuing proliferation of leukemic promyelocytes. When ara-C (10 mg) was added intramuscularly every 12 h, the patient achieved CR in 98 days (Table 1). In the 12 patients studied who responded to the induction differentiation effect of RA, L-CFU growth was predominant (163.3+- 129.0 colonies) and GM-CFU suppressed (0.63+- 1.3 colonies) prior to treatment. GM-CFU reached normal levels (100.2+- 55.1 colonies) with little or no growth of LCFU after CR was achieved.

III. Pattern of Clinical Response to trans- Retinoic Acid

Systematic observation of the peripheral blood counts during RA treatment of the previously untreated patients revealed some specific patterns of change. There was a progressive rise in the total white blood cell count which started with ini tiation of treatment and which reached a peak between 7 and 14 days. After this, the white blood cell count fell with the progressive maturation of granulocytes. Increase in platelets was most prominent after 3 weeks. Elevation of the hemoglobin concentration appeared reluctant and slow. Bone marrow aspirate revealed that hypercellularity existed throughout the RA treatment. Partial remission could be expected within 1 month (Fig. 1 C, D). Therapy with oral all-trans RA was accompanied by mild toxicity that consisted of dryness of the lips and skin (100%), headache (25%), nausea or vomiting (20.8%), moderate bone or joint pain (12.5%), and mild exfoliation (8.3%). Two patients had elevated SGPT. All of these side effects were well tolerated or alleviated when the dosage of oral RA was reduced.

IV. Disseminated Intravascular Coagulation

Coagulation parameters, including thrombin time, prothrombin time, plasma protamin sulfate paracoagulation test (3P), euglobulin lysis test, and fibrinogen levels, were measured simultaneously, in 21 patients, at the beginning of RA therapy and throughout the course of treatment. Of these patients, 18 who were normal in coagulation parameters prior to the start of RA therapy showed no changes during treatment. The other three patients who had been previously treated and who were 3P( + ) became negative 7-10 days after RA. Therefore, DIC or other hemorrhagic complications did not occur when patients with APL were induced to remission with RA.

V. Duration of Clinical Remission

Twenty-three patients were followed after induction of CR (Table 1 ). Of the six patients maintained on RA alone, four were still in remission for a period of 5 10 months. Two patients relapsed in 2 and 4 months. Among the four patients maintained on RA with either low-dose ara-C or low-dose Harringtonin in rotation, three relapsed within a period of 4- 5 months. Of the five patients maintained on low-dose ara-C alone, one case ( * 2) was lost to follow-up, one relapsed in 4 months, and the other three remained in CR for 1 + to 5 + months. Of the remaining nine patients who were consolidated by chemotherapeutic regimens and maintained on 6-mercaptopurine, methotrexate, or cyclophosphamide, two relapsed and seven have been in CR from 1 + to 8 + months. The new population of APL promyelocytes at relapse differed morphologically from those present at the start of treatment and were resistant to all-trans RA induction of differentiation in vitro.

D. Discussion

Recent approaches in treatment of leukemia include the use of "differentiation-inducing agents" such as RA, vitamin D3, or low-dose ara-C [17-19]. Numerous studies both in vitro and in vivo have revealed that RA is a potent inducer of myeloid differentiation, both in the promyelocytic cell line HL-60 as well as in fresh promyelocytes from patients with APL, and at a concentration that was pharmacologically obtainable in man [11,20]. 13-cis RA and all-trans RA were equally effective in induction of dif ferentiation in vitro [5]. Our studies confirm that in vitro leukemic promyelocytes could be induced by all-trans RA to differentiate toward mature granulocytes. One exception was that the cells from patient * 24 were resistant to RA induction. The morphological characteristics of these RA-resistant cells revealed a scanty cytoplasm with less-prominent coarse granulation. The differences in sensitivity to RA may be due to the heterogeneous entities of APL [21, 22]. In 1983, Flynn et al. [6] described the first case of APL treated with 13-cis- RA. Unfortunately, this patient died from disseminated candidiasis although there was a marked elevation in his peripheral granulocyte count after 2 weeks of treatment. Nilsson [7] reported a 30-year-old woman with APL in relapse for 10 months; she was treated by 13-ci,s RA (1 mg/kg) and began to respond after 1 month, and normal blood and bone marrow pictures continued for 11 months. Daenen et al. [8] reported a 33-year-old patient with refractory APL complicated by fibrinolysis and aspergillus pneumonia. He was treated with 13-cis RA (80 mg/day) alone, and attained a CR after 7 weeks. Recently, Fontana et al. [9] reported one case of refractory APL treated with 13-ci,s RA (lOO mg/m) which resulted in CR after 13 days. In vitro studies of this patient's leukemic blasts showed differentiation in the presence of RA. Sampi et al. [23] reported a 58-year-old Japanese man who also had relapsed APL and failed to respond to etretinate (a form of retinoid) and dactinomycin, although the leukemic cells were sensitive to all- tran,s RA ( 10 -6 10- 7 M) in vitro. We have treated our patients with all-trans RA and found that all-trans RA was not only effective in patients who had been refractory to chemotherapy, but also effective in those with "de novo" APL. Moreover, we were able to find predictive value in the in vitro differentiation studies. The single patient who was resistant to RA induction failed to show marrow improvement when treated with RA as the sole agent. According to most authors, the main disquieting problem of APL is death during induction treatment [3, 4, 24], especially because of intracerebral hemorrhage. DIC is the most common complication of APL. Its severity and frequency are often aggravated by chemotherapy despite the use of heparin. In this study we report no aggravation of hemorrhagic manifestation or appearance of coagulation parameter abnormalities suggesting DIC during the course of RA treatment. This would be one of the striking advantages over aggressive chemotherapy which could destroy the leukemic cells and cause the release of procoagulant factors from the azurophilic granules into the circulation. It is possible that the leukemic cells are not destroyed during treatment ofAPL with RA, but that they have differentiated, undergone terminal cell division, and lost the capacity to release these coagulant factors during this process. The fact that there was no decrease, but rather an increase, of marrow cellularity during induction therapy supports this possibility. The role of all-trans RA in the maintenance of remission is undetermined. Two cases of APL, reported by Daenen et al. [8] and Fontana et al. [9], relapsed in 6 and 12 months respectively. In our series, the patients were further treated with four different regimens after CR was induced, but it is too early to conclude which of these is the most effective. From the data obtained from both our clinical survey and the cytogenetic studies showing the persistence of abnormal clones (unpublished data), we suggest that intensive chemotherapy after CR may be beneficial. The knowledge about the side effects of oral RA is mainly from the dermatological literature. Our data are compatible with other reports on the toxicity of oral all-trans RA [25]. The toxicity of 13cis RA has been shown to be relatively lower than all-trans RA [25], but in our experience the side effects were well tolerated by the patients, some of whom have been taking RA for more than 10 months with no severe untoward effects. Based on these observations, we conclude that all-trans RA is an effective agent for obtaining CR in APL. How to maintain and prolong the duration of the CR, however, requires further study.


We are grateful to Professor Samuel Waxman of Mount Sinai School of Medicine for his kind comments and suggestions in editing this manuscript, and to other physicians and hematologists at Shanghai Rui-Jin Hospital, Shanghai Zhong-Shan Hospital, Shanghai Chang-Zhen Hospital, and Shanghai Institute of Pediatrics for providing samples and care of patients studied. We are also grateful to Zhao Jin-chai for expert technical assistance.


1. Jones ME, Saleem A (1978) Acute promyelocytic leukemia: a review of literature. Am J Med 65: 673-677

2. Bernard J, Weil M, Boiron M, Jacquillat C, Gemon MF (1973) Acute promyelocytic leukemia: results of treatment by Daunorubicin. Blood 41: 489 -496

3. Drapkin RL, Timothy SG, Dowling MD, Arlin Z, Mckenzie S, Kempin S, Clarkson B (1978) Prohylactic heparin therapy in acute promyelocytic leukemia. Cancer 41:2484-2490

4. Cordonnier C, Vernant JP, Brun B, Heilmann MG, Kuentz M, Bierling P, Farcet JP, Rodet M, Duedari N, Imbert M, Jouault H, Mannoni P, Reyes F, Dreyfus B, Rochant H (1985) Acute promyelocytic leukemia in 57 previously untreated patients. Cancer 55: 18- 25

5. Koeffler HP (1983) Induction of differentiation of human acute myelogenous leukemia cells: therapeutic implications. Blood 62: 709- 721

6. Flynn P, Miller W, Weisdorf D, Arthur D, Banning R, Branda R (1983) Retinoic acid treatment of acute promyelocytic leukemia: in vitro and in vivo observations. Blood 62: 1211-1217

7. Nilsson B (1984) Probable in vivo induction of differentiation by retinoic acid of promyelocytes in acute promyelocytic leukemia. Br J Haematol 57: 365-371

8. Daenen S, Vellenga E, Van Dobbenbugh OA, Halie MR (1986) Retinoic acid as antileukemic therapy in a patient with acute promyelocytic leukemia and aspergillus pneumonia. Blood 67: 559-561

9. Fontana JA, Roger JS, Durham JP (1986) The role of 13-cis retinoic acid in the remission induction of a patient with acute promyelocytic leukemia. Cancer 57: 209-217

10. Bennett JM, Catovsky D, Daniel MT, Flandrin G, Galton DAG, Gralnick HR, Sultan C (1976) Proposals for the classification of the acute leukemia. Br J Haematol 33: 451-458

11. Breitman TR, Selonick SE, Collins SJ (1980) Induction of differentiation of the human promyelocytic leukemia cell line (HL-60) by retinoic acid. Proc Natl Acad Sci USA 77: 2936-2940

12. Yam LT, Li CY, Crosby WH (1971) Cytochemical identification of monocytes and granulocytes. Am J Clin Pathol 55: 283 290

13. Francis GE, Guimaraes JETE, Berney 11, Wing MA (1985) Synergistic interaction between differentiation inducers and DNA synthesis inhibitors: a new approach to differentiation induction in myelodysplasia and acute myeloid leukemia. Leuk Res 9: 573581

14. Minden MD, Buick RN, McCulloch EA ( 1979) Separation of blast cell and T lymphocyte progenitors in the blood of patients with acute myeloblastic leukemia. Blood 54: 186-195

15. Pike BL, Robinson WR (1970) Human bone marrow culture in agar gel. 1 Cell Physiol 76: 77 -84

16. Vogler WR (1985) Post-remission therapy for acute myelogenous leukemia. In: Bloomfield CD (ed) Chronic and acute leukemias in adults. Martinus Nijhoff, Boston, p 209-228

17. Gold El, Mettelsmann RH, Itri LM, Gee T, Arlin Z, Kempin S, Clarkson B, Moore MAS (1983) Phase I clinical trial of 13-cis retinoic acid in myelodyplastic syndromes. Cancer Treat Rep 67: 981-986

18. Koeffler HP, Hirji K, Itri L (1985) 1,25Dihydroxyvitamin D3: in vitro and in vivo effects on human preleukemic and leukemic cells. Cancer Treat Rep 69: 1399 1407

19. Degos L, Castaigne S, Tilly H, Sigaux F, Daniel MT (1985) Treatment of leukemia with low-dose Ara-C: a study of 160 cases. Semin Oncol 12: 196 199 [Suppl 3]

20. Breitman TR, Collins SJ, Keene BR (1981) Terminal differentiation of human promyelocytic leukemic cells in primary culture in response to retinoic acid. Blood 57: 1000-1004

21. Golomb HM, Rowley 1D, Vardiman 1W, Testa 1R, Butler A (1980) "Microgranular" acute promyelocytic leukemia: a distinct clinical, ultrastructural, and cytogenetic entity. Blood 55: 253- 259

22. Tomonaga M, Yoshida Y, Tagawa M, 1innai I, Kuriyama K, Amenomori T, Yoshioka A, Matsuo T, Nonaka H, Ichimaru M (1985) Cytochemistry of acute promyelocy tic leukemia (M 3): leukemic promyelocytes exhibit heterogeneous patterns in cellular differentiation. Blood 66: 350-357

23. Sampi K, Honam Y, Hozumi M, Sakurai M (1985) Discrepancy between in vitro and in vivo induction of differentiation by retinoids of human acute promyelocytic leukemia cells in relapse. Leuk Res 9: 1475- 1478

24. Ruggero D, Baccarani M, Guarini A ( 1977) Acute promyelocytic leukemia: results or therapy and analysis of 13 cases. Acta Haematol (Basel) 58:108-119

25. Windhorst DB, Nigra T (1982) General clinical toxicology of oral retinoids. J Am Acad Dermatol 6: 675-682