• hydroxyurea;
  • mucocutaneous toxicity;
  • myeloproliferative neoplasms;
  • partial resolution


  1. Top of page
  2. Abstract


The current study was conducted to evaluate severe mucocutaneous toxicity during treatment with hydroxyurea (HU) in a large cohort of patients with Philadelphia chromosome-negative myeloproliferative neoplasms (MPN).


Among 993 consecutive patients newly diagnosed with MPN at 4 centers in Rome between January 1980 and December 2009, 614 patients (277 men and 337 women with a median age of 64.4 years [interquartile range (IR), 54.4 years-72.7 years]) received HU. HU was administered as first-line treatment in 523 patients (85.2%) and as ≥ second-line treatment in 91 patients (14.8%).


Mucocutaneous toxicity was reported in 51 patients (8.3%) after a median period from the initiation of HU treatment of 32.1 months (IR, 10.5 months-74.6 months) and a mean HU dose of 1085 mg (± 390 mg); 30 patients (58.8%) developed a painful ulcerative skin toxicity, mainly located in the perimalleolar area; 11 patients (21.6%) had oral aphthous ulcers; and 10 patients (19.6%) developed a nonulcerative skin toxicity with erythema and skin infiltration. After the mucocutaneous toxicity occurred, HU treatment was continued at the same dose in 5 patients (9.8%), reduced in 12 patients (23.5%), and temporarily discontinued in 7 patients (13.7%); the remaining 27 patients (52.9%) required a permanent drug discontinuation. After a median period of 4.3 months (IR, 2.4 months-9.0 months) from the onset of the skin toxicity, 39 patients (76.5%) had a complete resolution and 12 patients (23.5%) had improvement without complete resolution.


Mucocutaneous toxicity during HU treatment is more common than expected and may present with different clinical features. Moreover, it often requires a permanent drug discontinuation and only a partial resolution is reported to occur in approximately 25% of patients. Cancer 2011;. © 2011 American Cancer Society.

Hydroxyurea (HU) was first synthesized in 1869 by Dresler and Stein,1 but initial testing of this substance in animal experiments was only performed in 1928 by Rosenthal, who discovered that HU can determine both leukocytopenia and megaloblastic anemia.2

The antitumor activity of HU is based on inhibition of the enzyme ribonucleotide reductase and hence the inhibition of DNA synthesis, specifically affecting the S phase of the cell cycle.3

In the late 1950s, this drug was tested in several mouse tumor models such as L 1210 and L 5178A leukemia, T 241 and MA 387 sarcoma, and Ridgway osteogenic sarcoma as well as against adenocarcinoma and mammary carcinomas,4, 5 and demonstrated clear activity and efficacy.

Because of the proven efficacy of HU in these tumor models, this substance was tested in clinical trials on a wide range of advanced solid tumors, acute and chronic leukemias, and myeloproliferative disorders. It was promptly demonstrated that the drug was a highly effective cytostatic agent in chronic myeloproliferative disorders rather than in solid tumors.

In the era of tyrosine kinase inhibitors, HU remains the treatment of choice only in Philadelphia chromosome-negative myeloproliferative neoplasms (MPN) such as polycythemia vera,6 essential thrombocythemia,7 and symptomatic myelofibrosis.8

The most common side effects of HU are mild, and include granulocytopenia and anemia, especially after long-term treatment. Other common side effects include fatigue, headache, nausea, vomiting, diarrhea, or fever that seldom cause discontinuation of the drug.

The occurrence of painful leg ulcers is a rare complication that has been described in patients receiving high-dose, long-term HU treatment, often leading to drug discontinuation.9 To the best of our knowledge, to date only single case reports10-18 or single-center experiences19-22 have been published regarding this phenomenon, with a reported incidence ranging from 5% to 10% of the treated patients. A 5% incidence of skin ulcers during HU treatment also has been reported in a large randomized trial in patients with essential thrombocythemia.23

In addition to skin ulcers, several other mucocutaneous alterations have been reported during treatment with HU, ranging from very mild (hyperpigmentation, nail discoloration, alopecia, and scaling) to severe and uncomfortable toxicities (oral aphtosis, dermatomyositis-like eruptions). However, to our knowledge, complete and exhaustive descriptions of these skin toxicities other than leg ulcers are still lacking in literature.

The “Gruppo Laziale SMPC Ph1 negative” has reviewed Philadelphia chromosome-negative MPN databases from 4 hematological centers to analyze the incidence, characteristics, and prognostic factors related to symptomatic mucocutaneous toxicities occurring during HU treatment in a large cohort of unselected patients. The results of this analysis are reported in the current study.


  1. Top of page
  2. Abstract

Patient Population

From January 1980 to December 2009, a total of 993 patients (483 men and 510 women with a median age of 60.5 years [interquartile range (IR), 20.0 years -92.7 years]) with newly diagnosed MPN were registered in 4 hematological institutions in Rome, Italy.

The diagnosis of MPN was made according to Polycythemia Vera Study Group criteria until January 2002 and according to World Health Organization revised criteria after that date. There were 476 patients (47.9%) with essential thrombocythemia, 347 patients (34.9%) with polycythemia vera, 140 patients (14.1%) with primary myelofibrosis, and 30 patients (3.1%) with unclassifiable chronic myeloproliferative disorders.

Definition of Mucocutaneous Toxicity

After the exclusion of other known causes of skin toxicity, a symptomatic HU-dependent mucocutaneous toxicity was defined as the appearance of any cutaneous or oral mucosal manifestation of clinical significance (ie, with clinical symptoms such as pruritus or local pain) during treatment with HU.

Statistical Analysis

Data were expressed as the mean ± the standard deviation (normally distributed data), the median and IR (non-normally distributed data), or as percentage frequencies. Within-patient comparisons were made using the Student t test for paired data and the chi-square test as appropriate, at significance levels of P < .05.

All calculations were made using a standard statistical software package (SPSS for Windows Version 15.0; SPSS Inc, Chicago, Ill).


  1. Top of page
  2. Abstract

Among the 993 registered and evaluable patients, 614 (61.8%) received treatment with HU during the course of their disease. A total of 523 patients (85.2%) received HU as first-line treatment whereas 91 patients (14.8%) received the drug as second- or third-line treatment.

The development of a mucocutaneous toxicity during HU treatment was observed in 51 of 614 patients (8.3%) after a median period of 32.1 months (IR, 10.5 months-74.6 months) from the initiation of HU and a mean HU dose of 1085 mg (± 390 mg). The clinical features of these patients are reported and compared with those of patients without mucocutaneous toxicities in Table 1.

Table 1. Clinical Characteristics at Time of Diagnosis in Patients Receiving HU
CharacteristicAll Patients Treated With HUPatients Without Skin ToxicityPatients With Skin ToxicityP
  1. Abbreviations: CMPD-u, unclassifiable chronic myeloproliferative disorders; ET, essential thrombocythemia; Hb, hemoglobin; Ht, hematocrit; HU, hydroxyurea; IR, interquartile range; JAK-2, Janus kinase-2; MPN, myeloproliferative neoplasms; PLTS, platelets; PMF, primary myelofibrosis; PV, polycythemia vera; WBC, white blood cell count.

No. of patients61456351 
Males/females (%)277/337 (45.1/54.9)253/310 (44.9/55.1)24/27 (47.0/53.0)0.88
Median age (IR), y64.4 (54.6-72.8)64.3 (54.6-72.9)64.9 (54.4-69.7)0.84
MPN type    
 ET309 (50.3%)284 (50.4%)25 (49.0%) 
 PV214 (34.9%)196 (34.8%)18 (35.3%)0.53
 PMF73 (11.9%)65 (11.6%)8 (15.7%) 
 CMPD-u18 (2.9%)18 (3.2%) 
Median Hb (IR), g/dL14.9 (13.0-17.2)14.9 (13.0-17.2)14.8 (13.0-17.79)0.47
Median Ht (IR), %45.5 (40.5-53.1)45.5 (40.4-53.0)46.0 (40.4-55.3)0.4
Median WBC (IR), ×109/L9.6 (7.6-12.8)9.6 (7.6-12.8)9.7 (7.8-13.2)0.57
Median PLTS (IR), ×109/L703 (464-923)700 (474-923)708 (400-928)0.5
JAK-2 mutational status    
 Evaluable patients42138338 
 JAK-2 wild-type130 (30.9%)121 (31.6%)9 (23.7%)0.41
 JAK-2 V617F291 (69.1%)262 (68.4%)29 (76.3%) 
HU treatment    
 First line523 (85.2%)482 (85.6%)41 (80.4%)0.42
 ≥Second line91 (14.8%)81 (14.4%)10 (19.6%) 

According to the different clinical features, 3 types of mucocutaneous toxicities were recognized and considered.

Ulcerative Skin Toxicity

Thirty patients (58.8% of the patients with a mucocutaneous toxicity and 4.9% of all patients treated with HU) developed a painful ulcerative skin toxicity (UST) that was located in the perimalleolar area in 26 of 30 cases (86.7%) (Fig. 1 Top). In the remaining patients, the UST affected the head (2 patients) or feet (2 patients).

thumbnail image

Figure 1. Different types of mucocutaneous toxicities developing during therapy with hydroxyurea are shown: (Top) perimalleolar ulcer, (Middle) oral aphthous ulcer, and (Bottom) nonulcerative skin toxicity.

Download figure to PowerPoint

Oral Aphtosis

A total of 11 patients (21.6% of the patients with a mucocutaneous toxicity and 1.8% of all patients treated with HU) developed oral aphthous ulcers (Fig. 1 Middle).

Nonulcerative Skin Toxicity

Ten patients (19.6% of the patients with a mucocutaneous toxicity and 1.6% of all patients treated with HU) developed a nonulcerative skin toxicity (NUST), characterized by erythema and skin infiltration (diffuse in 2 patients and localized to the head or extremities in 5 patients and 3 patients, respectively) (Fig. 1 Bottom).

Table 2 shows some relevant characteristics of these different types of skin toxicities. There was no difference with regard to the mean HU dose; however, oral aphthous ulcers were found to be an early toxicity (median time from the initiation of HU therapy, 2.1 months; IR, 1.6 months-9.4 months), whereas UST and NUST were found to be late complications (median time from the initiation of HU therapy, 37.6 months [IR, 24.7 months-91.5 months] and 36.1 months [IR, 17.0 months-64.5 months], respectively) (P = .001).

Table 2. Features of Different Mucocutaneous Toxicities During HU Treatment
FeatureUSTOral AphtosisNUSTP
  1. Abbreviations: HU, hydroxyurea; IR, interquartile range; NUST, nonulcerative skin toxicity; UST, ulcerative skin toxicity.

No. of patients (%)30 (58.8)11 (21.6)10 (19.6) 
Mean HU dose, mg1080105511250.9
Median time from HU initiation (IR), mo37.6 (24.7-91.5)2.1 (1.6-9.4)36.1 (17.0-64.5)0.001
HU modification    
 No modification221 
 Dose reduction7230.29
 Temporary discontinuation232 
 Permanent discontinuation1944 
Median toxicity duration (IR), mo6.0 (3.1-16.5)3.3 (0.8-6.6)3.1 (3.0-8.4)0.04
Toxicity resolution    
 Complete (%)23 (76.6)8 (72.7)8 (80)0.34
 Partial (%)7 (33.4)3 (27.3)2 (20) 

After the occurrence of symptomatic mucocutaneous toxicities, HU treatment was continued at the same dose in 5 patients (9.8%) and at a reduced dose in 12 patients (23.5%), and was temporarily discontinued in 7 patients (13.7%); the remaining 27 patients (52.9%) required a permanent drug discontinuation (Table 2). In this latter group, HU was replaced by pipobroman (18 patients), anagrelide (6 patients), melphalan (2 patients), or interferon-α (1 patient).

After a median of 4.3 months (IR, 2.4 months-9.0 months) from the onset of the mucocutaneous toxicity to the better response achieved, 39 patients (76.5%) achieved a complete resolution and 12 patients (23.5%) demonstrated an improvement without complete resolution (Table 2); for example, it was common to observe a stable or recurring insignificant skin ulcer without pain or other local symptoms in patients with an UST who achieved an improvement without complete resolution. There was no difference noted in the rate of complete and partial resolution among the different types of toxicity, but oral aphthous ulcers and NUST were found to have a shorter median time of resolution compared with UST (3.3 months [IR, 0.8 months-6.6 months] and 3.1 months [IR, 3.0 months-8.4 months] vs 6.0 months [IR, 3.1 months-16.5 months], respectively) (P = .04).

Several clinical and biological features (age at diagnosis, gender, type of MPN, Janus kinase [JAK]-2 mutational status, white blood cell count and platelet count at the time of diagnosis, concomitant diabetes and dyslipidemia, smoking history, previous thrombotic events, first-line or second-line HU treatment, and HU dose) were compared on univariate analysis between patients with and those without symptomatic mucocutaneous toxicity; no predictive factor for the development of such complications was detected (Table 1).


  1. Top of page
  2. Abstract

HU is a well-known cytostatic drug that is used to treat patients with many hematological diseases. Its safe toxicity profile has led to the use of HU in the treatment of elderly and frail patients as palliative therapy, as well as in some non-neoplastic settings such as sickle cell anemia. In patients with high-risk MPN, in whom a long-lasting treatment is required, HU currently appears to be the most effective and safe first-line choice, as has been shown also in randomized trials23; HU cytotoxicity in particular does not appear to translate into a higher risk of second therapy-related neoplasia.

The skin constitutes a specific target of HU toxicity; the majority of the skin alterations observed during HU treatment are mild and do not lead to dose modifications or discontinuations. However, there are some rare types of mucocutaneous toxicity that are symptomatic and often require a permanent discontinuation of HU therapy. Among them, painful skin ulcers have been reported for many years in several single cases or unicentric small cohorts of patients.10-22 Currently, the occurrence of skin ulcers during HU treatment is considered a treatment failure24; nonetheless, a full and specific evaluation of symptomatic mucocutaneous toxicity during HU treatment in a large multicentric study focusing on incidence, clinical features, and follow-up is still lacking.

The results of the current study demonstrate that 8.3% of MPN patients receiving HU treatment developed a severe mucocutaneous toxicity, with no difference noted with regard to MPN type, clinical features, JAK-2 mutational status, and predisposing conditions. It is worth noting that skin ulcers, often reported in previous small studies, represented 58.8% of the severe mucocutaneous toxicities; the remaining 41.2% of toxicities were comprised of oral aphtosis and NUST, which are less recognized complications but represented a significant percentage of the phenomenon.

Although oral aphtosis was an early complication and was never observed after the first 12 months of treatment, UST and NUST developed after a median treatment period of 36 months and therefore are to be considered as late complications.

On the basis of single, fully described case reports, a general agreement exists concerning the complete healing of mucocutaneous toxicities, nearly always after permanent discontinuation of HU. However, in the current study cohort, there was a percentage of patients who did not achieve complete regression of mucocutaneous toxicities even after interruption of HU, with rates ranging from 20% of patients with NUST to 30% of patients with UST; moreover, permanent HU discontinuation was often but not always needed, and at least one-third of patients achieved skin healing while continuing HU therapy at the same or a reduced dose.

Considering that only 8.3% of the patients in the current study receiving HU were affected by skin complications, a major issue of this study was to identify predisposing factors. Our analysis, which took into account several clinical features, was unsuccessful in demonstrating any predisposing factor. Therefore, it is not possible at the current time to select patients at high risk of developing skin toxicities in whom HU should be avoided.

The pathogenesis of mucocutaneous toxicities developing during HU treatment is still unclear. There was no relation noted between HU dose and the development of a skin toxicity; furthermore, the mean HU dose in patients with a mucocutaneous toxicity was unexpectedly low. Both these features point to the importance of still unknown individual biologic characteristics predisposing toward skin toxicity rather than a direct dose-dependent toxic effect of the drug. In particular, possible candidates for the development of a skin toxicity may be genetic polymorphisms of the genes involved in drug metabolism as well as in skin growth.

In conclusion, symptomatic mucocutaneous toxicity during HU treatment is more common than expected, with an incidence of approximately 8%, and may present with different clinical features; moreover, it often requires a permanent drug discontinuation and in approximately 25% of patients there is only a partial resolution. Thus, further studies are warranted to highlight the pathogenesis, as well as individual predisposing factors that could be different in the various types of mucocutaneous toxicity.


  1. Top of page
  2. Abstract
  • 1
    Dresler WF, Stein R. Ueber den Hydroxylharnstoff. Justus Liebig's Ann Chem Pharm. 1869; 150: 242-252.
  • 2
    Rosenthal F. Ueber die Beziehungen von schwersten Blutgiften zu Abbauprodukten des Eiweisses: Ein Beitrag zum Enstehungsmechanismus der perniziosen Anamie. Klin Wochenschr. 1928; 7: 972-977.
  • 3
    Kovacic P. Hydroxyurea (therapeutics and mechanism): metabolism, carbamoyl nitroso, nitroxyl, radicals, cell signaling and clinical applications. Med Hypotheses. 2011; 76: 24-31
  • 4
    Tarnowski GS, Stock CC. Chemotherapy studies on the RC and S 790 mouse mammary carcinomas. Cancer Res. 1958; 18( 8 pt 2): 1-48.
  • 5
    Brockman RW, Shaddix S, Laster WR, Schabel FM Jr. Inhibition of ribonucleotide reductase, DNA synthesis, and L1210 leukemia by guanazole. Cancer Res. 1970; 30: 2358-2368.
  • 6
    Finazzi G, Barbui T. How I treat patients with polycythemia vera. Blood. 2007; 109: 5104-5111.
  • 7
    Barbui T, Barosi G, Grossi A, et al. Practice guidelines for the therapy of essential thrombocythemia. A statement from the Italian Society of Hematology, the Italian Society of Experimental Hematology and the Italian Group for Bone Marrow Transplantation. Haematologica. 2004; 89: 215-232.
  • 8
    Mesa RA. How I treat symptomatic splenomegaly in patients with myelofibrosis. Blood. 2009; 113: 5394-5400.
  • 9
    Montefusco E, Alimena G, Gastaldi R, Carlesimo OA, Valesini G, Mandelli F. Unusual dermatologic toxicity of long-term therapy with hydroxyurea in chronic myelogenous leukemia. Tumori. 1986; 72: 317-321.
  • 10
    Radaelli F, Calori R, Faccini P, Maiolo AT. Early cutaneous lesions secondary to hydroxyurea therapy. Am J Hematol. 1998; 58: 82-83.
  • 11
    Best PJ, Daoud MS, Pittelkow MR, Petitt RM. Hydroxyurea-induced leg ulceration in 14 patients. Ann Intern Med. 1998; 128: 29-32.
  • 12
    Stagno F, Guglielmo P, Consoli U, Fiumara P, Russo M, Giustolisi R. Successful healing of hydroxyurea-related leg ulcers with topical granulocyte-macrophage colony-stimulating factor. Blood. 1999; 94: 1479-1480.
  • 13
    Tarumoto T, Imagawa S, Hotta T, et al. A case of bilateral heel ulcers associated with hydroxyurea therapy for chronic myelogenous leukemia. Jpn J Clin Oncol. 2000; 30: 159-162.
  • 14
    Sastre JL, Bravo A, Tembras S, Gomez R, Ulibarrena C. Leg ulcers associated with hydroxyurea therapy. Haematologica. 2003; 88: EIM01.
  • 15
    Lankipalli R, Konkle B, Mohler ER 3rd. An unusual case of toe ulceration. Vasc Med. 2006; 11: 35-37.
  • 16
    Romanelli M, Dini V, Romanelli P. Hydroxyurea-induced leg ulcers treated with a protease-modulating matrix. Arch Dermatol. 2007; 143: 1310-1313.
  • 17
    Hwang SW, Hong SK, Kim SH, Seo JK, Lee D, Sung HS. A hydroxyurea-induced leg ulcer. Ann Dermatol. 2009; 21: 39-41.
  • 18
    Martorell-Calatayud A, Requena C, Nagore-Enguidanos E, Guillen-Barona C. Multiple, painful, treatment-resistant leg ulcers associated with dermatomyositis-like lesions over the interphalangeal joints induced by hydroxyurea [in Spanish]. Actas Dermosifiliogr. 2009; 100: 804-807.
  • 19
    Najean Y, Rain JD. Treatment of polycythemia vera: the use of hydroxyurea and pipobroman in 292 patients under the age of 65 years. Blood. 1997; 90: 3370-3377.
  • 20
    Sirieix ME, Debure C, Baudot N, et al. Leg ulcers and hydroxyurea: forty-one cases. Arch Dermatol. 1999; 135: 818-820.
  • 21
    Randi ML, Ruzzon E, Luzzatto G, Tezza F, Girolami A, Fabris F. Safety profile of hydroxyurea in the treatment of patients with Philadelphia-negative chronic myeloproliferative disorders. Haematologica. 2005; 90: 261-262.
  • 22
    Ruzzon E, Randi ML, Tezza F, Luzzatto G, Scandellari R, Fabris F. Leg ulcers in elderly on hydroxyurea: a single center experience in Ph- myeloproliferative disorders and review of literature. Aging Clin Exp Res. 2006; 18: 187-190.
  • 23
    Harrison CN, Campbell PJ, Buck G, et al; United Kingdom Medical Research Council Primary Thrombocythemia 1 Study. Hydroxyurea compared with anagrelide in high-risk essential thrombocythemia. N Engl J Med. 2005; 353: 33-45.
  • 24
    Barosi G, Birgegard G, Finazzi G, et al. A unified definition of clinical resistance and intolerance to hydroxycarbamide in polycythaemia vera and primary myelofibrosis: results of a European LeukemiaNet (ELN) consensus process. Br J Haematol. 2010; 148: 961-963.