Indian Journal of Cancer Medknow Publications on behalf of Indian Cancer Society ISSN: 0019-509X EISSN: 1998-4774 Vol. 44, Num. 3, 2007, pp. 93-98

Indian Journal of Cancer, Vol. 44, No. 3, July-September, 2007, pp. 93-98

Original Article

Clinico-biologic profile of Langerhans cell histiocytosis: A single institutional study

Narula G, Bhagwat R, Arora B, Banavali SD, Pai SK, Nair CN, Seth T, Laskar S, Muckaden MA, Kurkure PA, Parikh PM

Department of Medical Oncology, Tata Memorial Hospital, Parel, Mumbai - 400 012, Maharashtra
Correspondence Address:Department of Medical Oncology, Tata Memorial Hospital, Parel, Mumbai - 400 012, Maharashtra, kurkurepa@tmcmail.org

Code Number: cn07015

Abstract

Keywords: Clinical profile, Langerhans cell histiocytosis, outcome

Introduction

Materials and Methods

Results

Fifty-two patients with a confirmed diagnosis of LCH were eligible. The median age was 3 years and median follow-up was 16 months (range 1-183 months) with 35 cases followed for more than 6 months. Patient characteristics are shown in [Table - 2].

Clinical features

The skeletal system was the most commonly involved system. The classical lytic lesions, often with adjacent soft tissue swelling, were detected either incidentally or on investigation of a swelling by imaging. Pain occurred in only two patients with lesions in weight-bearing bones. The calvarium accounted for 63% of skeletal involvement followed by lesions of the base of skull bones, maxilla and mandible. "CNS risk" lesions were present in five patients, usually as intracranial extensions of surrounding bony lesions. Two of these subsequently developed diabetes insipidus (DI). The other flat and long bones were infrequently involved. Bone scan done in selected cases did not have any added diagnostic value over conventional skeletal survey. The frequency of various organ involvement is given in [Table - 2].

Clinically significant lymph nodes were the second commonest disease site, with cervical nodes involved in 45% patients followed by axillary and inguinal (one-third of cases each) involvement. Mediastinal lymph nodes were enlarged in three patients. Majority (60%) of cases had histopathological confirmation on lymph node biopsy, while another 20% were reported as "Reactive" Lymph nodes. The remaining 20% cases with significant lymphadenopathy did not undergo a node biopsy either due to deep location (mediastinal/abdominal) and/or availability of another easily accessible biopsy site.

Dermal involvement was seen commonly in scalp (75%), face, limbs and trunk (40-60%). Papular rash was the commonest lesion, while seborrheic dermatitis was less frequent. Diffuse skin pigmentation and mucocutaneous ulcers were seen in two patients each. Approximately half of the cases could be confirmed histopathologically.

Splenomegaly was frequent in our patients (25%). Nineteen cases (36.5%) had hematological involvement based on deranged peripheral blood parameters. Anemia as defined by grouping criteria [Table - 1] was common to all, while one had thrombocytopenia. Five had bone marrow involvement on morphological assessments of aspirate smears or marrow biopsy specimens. CD1a studies were not done on any marrow specimen.

Lungs (based on suggestive imaging) were involved in eight patients. Liver involvement was common. Three patients had obstructive hepatitis, of which one went on to develop cirrhosis. Intestinal involvement suggested by profuse chronic diarrhea was seen in one patient.

Endocrine involvement manifested by DI was seen in five cases. One of these patients also had growth retardation suggestive of possible growth hormone deficiency. Other manifestations included chronic suppurative otitis media seen in six, while constitutional symptoms were seen in two patients.

The onset of involvement of various organ systems with relation to time of presentation showed trends reflective of the biology of the disease [Figure - 1]. Most of the patients had consistent temporal pattern of disease progression, with skin and skeletal system being the earliest to get involved and other visceral organs later in the neoplastic continuum heralding an aggressive phase of disease.

Grouping

Forty-four percent patients ( n = 23) were classified as Group I with involvement of one or more "risk" organs [Table - 1]. Hematopoietic system involvement accounted for the majority ( n = 19). Significantly, six patients were classified in this group only on the basis of anemia as defined in [Table - 1]. Group II, III and IV lesions formed more than half the patients [Table - 2].

Prior treatment

Twenty-nine percent patients had prior treatment at a median of 2 months before presentation (range −27 to −1 month). In half of these, surgery - usually curettage of an isolated bony lesion - had been done. One patient had received radiotherapy. Five patients with multi-system disease had received chemotherapy and three of these had progressive disease at presentation.

Initial treatment

Sixty percent of our patients received chemotherapy as initial treatment. One-fifth ( n = 10) received radiotherapy, while nine were kept on surveillance. One patient each underwent surgery or intra-lesional steroids as definitive therapy. Etoposide-based chemotherapy (150 mg/m ² /day over 3 days 4 weekly, 1-6 cycles, median 3) was the commonest regime ( n = 20) usually along with methyl prednisolone (30 mg/kg/day infusion for 3 days 4 weekly, 1-6 cycles, median 3) ( n = 14) or other drugs ( n = 6). Cyclophosphamide (400 mg/m ² IV 3 weekly), vincristine (1.5 mg/m ² 3 weekly) and prednisolone (40 mg/m ² x 5 days) (COP regime) were used in another 10 patients. A median of 6 cycles was received by all (range 1-6). Twenty-one patients received second-line treatment consisting of radiotherapy ( n = 4) and polychemotherapy ( n = 17). The latter group included three cases that received COP, five each of vinblastine (6 mg/m ² ) or etoposide with or without steroids (1-6 cycles, median 4) and two each of cyclosporine (5 mg/kg/day for 1-6 month, mean 3.5 month) and cladiribine (0.15-0.35 mg/kg/day x 5 days). Two patients received cyclosporine after failing second-line treatment.

Disease state and response

While 40% had good response and inactive disease at the end of 6 cycles/month from initial therapy, 33% had progressive disease. Of the 21 patients requiring second-line treatment, about 50% showed a complete response and clinically inactive disease, 30% had progression and the rest had stable disease.

Outcome

Majority of patients (58%) were in CR at last follow-up, 13% had stable disease and the remaining 29% had progressive disease. Of these, five died of the disease in hospital and the remaining patients opted for best supportive care only. Overall survival of the study cohort was 62.78% and mean survival was 118 months [Figure - 2].

Patients with Group I disease had the worst outcome, with almost all the fatalities belonging to this group. Fatalities included cases classified in this group on the basis of anemia alone. The projected overall survival at 10 years in Group I was 35% as compared to 68% in Group II and almost 100% in the other groups. The outcome of non-Group I patients remained excellent irrespective of the treatment offered [Figure - 3].

The outcome was also assessed according to the initial treatment offered, as this was more reflective of the risk assessment at the time of diagnosis. Predictably, the chemotherapy group did the worst, as this was offered to those perceived to be most at risk [Figure - 4].

Thirty-four surviving patients who were followed-up for at least six months were evaluated for sequelae. At a median follow-up of 40 months (range 6-183), six patients (17%) had sequelae. DI occurred in two patients. Both were among five patients who had "CNS risk" disease at diagnosis. Two patients had suboptimal scholastic performance. Both had received radiotherapy for skeletal lesions in the calvarium. One of these patients also had obesity. One patient each had pulmonary (restrictive lung disease) and hepatic sequelae (cirrhosis). In the latter, death occurred as a result of the sequelae.

Discussion

LCH is a rare childhood malignancy. Only 52 cases were registered over the last 17 years at our center, which gets at least 20,000 new cases of malignancies per year in all age groups (Hospital Registry). There has been unprecedented progress in the understanding of this enigmatic disease over the past few years by means of collaborative trials and large single institutional studies. ^{[1],[5],[6],[7],[8]} Foremost is that multi-organ involvement has worse outcome than single-organ disease and needs some form of cytotoxic therapy and among those with multi-organ disease, some patients are more "at risk" for disease progression, sequelae and death. ^{[5],[6],[7],[8],[9],[10],[11]}

Clinical features in our patients were similar in nature and extent to those reported in most large series. Skeletal involvement with or without other sites occurs in 80 to 100% of patients in most large series and was 92% in our series. ^{[10],[12],[13]} Radionuclide scanning provided no additional benefit as against its suggested complementary role in recent studies. ^{[14],[15],[16],[17]}

Dermal involvement (25%) was less common in contrast to 35-50% cases in most series. ^[12],[17] This is likely to be due to underdiagnosis of minor lesions such as seborrheic dermatitis. Lymphadenopathy was more common as compared to less than 10% in literature. ^[18] This, to some extent, may have been contributed by the high incidence of reactive adenopathy in our country, but warrants further investigation.

Similarly, splenic involvement was much higher (25%) in our patients as compared to 5% at presentation in a French series. ^[12] Conversely, pulmonary involvement (15%) was less frequent, compared to 50% of patients in a large series that also included patients with isolated abnormal pulmonary function tests. ^[19]

Hepatomegaly is often present in systemic disease and pathological changes might be present in the liver on histology, even in the absence of liver dysfunction. ^[20] In our series, a quarter of the patients with multi-systemic disease had hepatomegaly. Only one with associated hepatic dysfunction developed cirrhosis.

DI afflicts nearly one-third of all patients with multi-system LCH and can develop at any time during the course of the illness. ^{[5],[6],[13],[18],[21],[22]} This was reflected in our patients where it occurred in five patients at a variable time during the disease course (median 11 months; range 20-31 months). The regular and prospective use of water-deprivation test in our study cohort might have ensured maximum detection of cases. CNS risk lesions predated DI in two of five cases.

In the present study, risk perception remained a matter of interpretation that varied over time and observer. We, therefore, retrospectively applied uniform risk and response criteria as per the current understanding to make the analysis more meaningful. The projected overall survival of all cases at 10 years was just over 60% with a mean observation time of 118.15 months [Figure - 2]. The survival of patients with multi-system risk disease (Group I) was just over 30% [Figure - 3]. The occurrence of patients classified in this group solely for anemia underlines the importance of this finding in LCH. This group compared poorly to Group II patients with 68% survival, the only other group to have any mortality at all. Here too, the only two patients, who died shortly after presentation, had one of the risk-organs involved pre-terminally. It is likely that involvement of these systems was missed at initial assessment. Prior treatment is unlikely to have impacted the outcome as two-thirds were low-risk at presentation and had received only local therapy. In the analysis by treatment group too, early mortalities occurred in the radiotherapy/other therapies group only in two patients with risk-organ systems involved [Figure - 4]. In our study, as in literature, approximately one-third of patients had poor prognosis, with survival just above 30%, irrespective of the type of systemic therapy given. These patients were correctly identified using the current criteria of high-risk-organ system involvement, despite the limitations of applying such criteria retrospectively. ^{[5],[6],[7],[8],[9],[10],[11]}

In the earliest European studies, patients with multi-system disease and evidence of organ dysfunction were considered "at risk" and had an overall survival of 81%. However, dissimilarity in the grouping criteria prevents any meaningful comparison with these studies. ^[23] Furthermore, the stratification of patients into responders and non-responders early in their treatment at 6 weeks closely identified poor prognostic groups. ^{[5],[7],[12],[24]} The importance of early response could not be assessed in our study due to its retrospective nature.

Sequelae in LCH have usually been limited to bony deformities, teeth loss and endocrine disorders. ^[8],[13] No bony deformities were seen in our patients, while endocrine disorders were seen in a few patients.

Acknowledgment

References

Copyright 2007 - Indian Journal of Cancer

The following images related to this document are available:

Photo images