Recurring Chemotherapy-Associated Alopecia Areata: Case Report and Literature Review

Susannah E. Motl, Pharm.D., Christopher Fausel, Pharm.D.


Pharmacotherapy. 2003;23(1) 

In This Article


Alopecia is a well-recognized adverse effect of chemotherapy.[7] To our knowledge, however, cyclic alopecia areata of this pattern has not been reported in a patient with cancer. Alopecia areata can be characterized by a sudden onset of total hair loss in a well-demarcated area.[1] Close examination reveals poorly formed hair shafts in the affected region. The differential diagnosis of alopecia areata in the general population involves numerous medical and psychosocial factors. However, in our patient, laboratory test results within normal range excluded the possibility of thyroid or parathyroid disorders, lupus erythematosus, and nutrition or iron deficiencies. In addition, emotional stress was not likely a direct cause of her hair loss, although an indirect association could not be excluded.

Alopecia neoplastica, with an appearance similar to that of alopecia areata, has been linked to systemic manifestation of cancer recurrence or initial diagnosis of the disease.[8,9,10] It develops from a hematogenous spread of the cancer to the skin and generally is well demarcated with a smooth pink surface.[1] A biopsy is required to confirm alopecia neoplastica. Breast carcinoma is the most highly reported source of skin metastases.[8] Because our patient's serum tumor markers for ovarian cancer (e.g., CA 125) consistently remained normal after resection of her disease, metastatic spread from ovarian cancer was unlikely. In addition, negative findings from bone scan, ultrasound, and mammography confirmed the absence of bony and breast metastases.

Hair loss has been linked to certain drugs.[11] Our patient's drug regimen consisted of naproxen 250 mg at bedtime for back pain and zolpidem 5 mg as needed for sleep; these drugs have not been shown to cause alopecia.[11] In addition, the patient received paclitaxel and carboplatin for cancer treatment. Paclitaxel is associated with hair loss in virtually all patients treated, and in 87% of those receiving single-agent therapy.[6] In contrast, carboplatin-induced alopecia has been reported in only 2-3% of patients receiving single-agent therapy.[12] However, delayed alopecia, which our patient experienced, is not expected with these agents, as the elimination half-life is 13-52 hours for paclitaxel[6] and 1-6 hours for carboplatin.[12] The manufacturer of paclitaxel and carboplatin has received no reports of delayed, recurring alopecia.[6,12]

Our patient's experience 25 years earlier with alopecia diffusa of the scalp resulting from oral contraceptives is noteworthy. However, her current episodes of alopecia areata did not appear to be related (e.g., unaffected scalp hair). Two possible reasons for the lack of scalp hair loss with her later episodes of alopecia are lack of pigment in her scalp hair, and the characteristic growth cycle of scalp hair compared with that of other body hair.

First, unpigmented (white) hair tends to be unaffected by alopecia areata.[2] One case report describes a 31-year old man who experienced progressive loss of all pigmented body hair, with thinning of axillary and pubic hair, similar to our patient's pattern of hair loss.[2] Only white and gold hairs were preserved.

Second, hair growth occurs in three phases: anagen (growth), catagen (resting), and telogen (lack of growth).[1] Alopecia generally occurs in anagen and telogen phases. The growth phases of scalp and nonscalp hair differ in length. Compared with scalp hair, the anagen phase in eyebrow, eyelash, arm, and leg hair is shorter, and the telogen phase is longer.[1] Specifically, the anagen phase lasts 1-6 years for scalp hair and 2-6 months for nonscalp hair. This explains why scalp hair has the capability to grow the longest. If our patient's alopecia areata had occurred during a specific phase of hair growth, her scalp hair may not have been synchronized with her body hair and may have protected her scalp. The growth of scalp hair is synchronized in some animals, but not in humans.

A previously published report describes a patient who, similar to our patient, experienced alopecia universalis that did not affect the scalp hair.[13] Of interest, this patient had a history of previous occipital alopecia areata (loss of scalp hair at the posterior part of the head) that had resolved 10 years before this onset. The authors of this report did not hypothesize as to why the scalp hair was unaffected; however, this patient's hair was pigmented.

Normal hair loss from chemotherapy is associated with anagen effluvium (hair loss in the anagen phase), as chemotherapy affects living, rapidly dividing cells.[11] However, telogen effluvium (hair loss in the telogen phase) also may play a role in alopecia. The greatest differential factor between the two effluvia is the relationships between the insult to the hair and the occurrence of hair loss.[1] Anagen effluvium generally occurs 1-4 weeks after the insult, whereas telogen effluvium occurs within 2-4 months.

Onset of our patient's cyclic alopecia areata began 3 months after completion of chemotherapy, which is consistent with telogen effluvium.[1] Normal hair loss in the telogen phase occurs when a new hair shaft is produced under the base of the existing hair.[11] The new hair pushes the older hair out of the follicle, and hair loss occurs. Generally, this results in no net hair loss because the old hairs are replaced with the new. This is similar to the pattern of eyelash and eyebrow hair loss that our patient experienced. However, it does not explain loss of arm, leg, armpit, and pubic hair. The catagen phase in these areas may have been lengthened, causing a stagnant growth period.[1]

Our patient's condition most likely reflects overlapping types of alopecia; the patterns of her hair loss and the time frames for regrowth were not consistent across affected areas. Whereas loss of eyebrow, eyelash, and armpit hair was patchy, loss of arm and pubic hair was very diffuse, and loss of leg hair was total. In addition, regrowth was stagnant for a month or longer, except for her eyelashes and eyebrows, which grew back immediately.

A possible hypothesis for delayed chemotherapy-induced alopecia relies on the autoimmune theory of alopecia areata. Autoimmune diseases, such as lupus erythematosus, vitiligo, and Hashimoto's thyroiditis, have been associated with alopecia areata.[14] Here, antibody attack on specific autoantigens normally found within hair follicles may cause hair loss.[15,16] These hair autoantigens are not normally exposed; however, follicular damage presents them to the immune system. This explains why healthy individuals with follicle-associated antibodies do not lose hair; the autoantigens are hidden from attack.[16] The CD8+ and CD4+ T cells have been implicated as the effectors and helpers of hair loss, respectively.[15]

Proof backing this autoimmune theory of alopecia areata is the positive regrowth of hair in affected patients who are treated with immune stimulating agents, such as glucocorticoids, cyclosporine, and topical immunotherapy.[15,17,18] Animal experiments have shown that transplantation of alopecia-affected human scalp to nude mice induces hair regrowth with treatment, since these mice have functioning immune systems.[14] This suggests that the alopecia areata is created humorally rather than environmentally. Affected patients still have the ability to grow hair, as demonstrated by spontaneous regrowth in our patient.

Additional experiments have shown that the injection of T lymphocytes (from affected human scalp) can induce alopecia areata in immunodeficient mice with hair grafts from affected human scalp, again suggesting an autoimmune factor.[15] However, injection of nonscalp T lymphocytes does not induce alopecia in these same mice. Finally, antigens associated with melanocytes of patients with alopecia areata function as autoantigens and promote hair loss when transferred to immunodeficient mice.[19] This may explain why unpigmented hair is not affected in many patients with alopecia areata.

We hypothesized that our patient's cyclic hair loss was a result of a chemical alteration of the hair follicles; chemotherapy may have altered the follicular makeup, resulting in autoantigen exposure. Chemotherapy has altered the p53 protein, which is responsible for apoptosis in hair follicles.[20,21] Microbial superantigens, localized microtrauma, neurogenic inflammation, and infectious agents all have been implicated in follicular modifications.[16]

Our patient experienced cyclic attacks on her hair follicles, similar to the cyclic nature of her chemotherapy. Hair loss seemed to occur every 10 weeks. Recurrent loss of hair from an autoimmune reaction may be explained by the presence of primed T lymphocytes from previous insults.[16] Of interest, both our patient and the other patient described,[13] with a history of alopecia universalis affecting the scalp, experienced alopecia areata the second time with scalp hair remaining intact.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.